Ardila - Ostrosky Solis - (Ed) - 1989 PDF

Brain Organization
of Language and
Cognitive Processes
Critical Issues in Neuropsychology
Series Editors:
Cecil R. Reynolds Antonio E. Puente
Texas A&M University University of North Carolina, Wilmington
Editorial Advisory Board:

Erin D. Bigler, University of Texas, Austin
Raymond S. Dean, Ball State University
Hans J. Eysenck, University of London
Charles J. Golden, Drexel University
John Gruzelier, University of London
Lawrence C. Hartlage, Fairfax University
Merrill Hiscock, University of Saskatchewan
Lawrence Majovski, Huntington Memorial Hospital
Francis J. Pizzolo, Baylor University School of Medicine
Karl Pribram, Stanford University
ASSESSMENT ISSUES IN CHILD NEUROPSYCHOLOGY

Edited by Michael G. Tramontana and Stephen R. Hooper
BRAIN ORGANIZATION OF LANGUAGE AND COGNITIVE PROCESSES

Edited by Alfredo Ardila and Feggy Ostrosky-Solis
HANDBOOK OF CLINICAL CHILD NEUROPSYCHOLOGY

Edited by Cecil R. Reynolds and Elaine Fletcher-Janzen
MEDICAL NEUROPSYCHOLOGY: The Impact of Disease on Behavior

Edited by Ralph E. Tarter, David H. Van Thiel, and Kathleen L. Edwards
NEUROPSYCHOLOGICAL FUNCTION AND BRAIN IMAGING

Edited by Erin D. Bigler, Ronald A. Yeo, and Eric Turkheimer
NEUROPSYCHOLOGY, NEUROPSYCHIATRY, AND BEHAVIORAL

NEUROLOGY
R. Joseph
RELIABILITY AND VALIDITY IN NEUROPSYCHOLOGICAL

ASSESSMENT
Michael D. Franzen
A Continuation Order Plan is available for this series. A continuation order will bring delivery of
each new volume immediately upon publication. Volumes are billed only upon actual shipment.
For further information please contact the publisher.
Brain Organization
of Language and
Cognitive Processes
Edited by
ALFREDO ARDILA
Miami Institute of Psychology
Miami, Florida
and
FEGGY OSTROSKY-SOLIS
National Autonomous University of Mexico
Mexico D.F., Mexico
Plenum Press • New York and London

Library of Congress Cataloging in Publication Data
Brain organization of language and cognitive processes.
(Critical issues in neuropsychology)
Includes bibliographies and index.
1. Language disorders. 2. Brain damage. 3. Neuropsychology. 4. Cognition. I. Ar-
dila, Alfredo. II. Ostrosky-Solis, Feggy. III. Series. [DNLM: 1. Brain Injuries-
physiopathology. 2. Cognition-physiology. 3. Language Disorders-physiopathology.
4. Neuropsychology. WL 103 B81453]
RC423.B657 1989 616.85'5 89-16051
ISBN-13:978-1-4612-8088-0 e-ISBN-13:978-1-4613-0799-0
DOI:1O.1007/978-1-4613-0799-0
© 1989 Plenum Press, New York

Softcover reprint of the hardcover lst edition 1989
A Division of Plenum Publishing Corporation
233 Spring Street, New York, N.Y. 10013
All rights reserved
No part of this book may be reproduced, stored in a retrieval system, or transmitted
in any form or by any means, electronic, mechanical. photocopying, microfilming,
recording, or otherwise, without written permission from the Publisher
In memoriam
A. R. Luria
Professor and friend
Contributors
ALFREDO ARDILA, Konrad Lorenz Foundation University, Bogota, Co-

lombia, and Miami Institute of Psychology, Miami, Florida
33166-6612
D. FRANK BENSON, The Neurobehavior Unit, West Los Angeles Veterans
Administration Medical Center, Departments of Neurology and Psy-
chiatry and Behavioral Sciences, UCLA School of Medicine, Los
Angeles, California 90024
AMY M. BIHRLE, The Salk Institute for Biological Studies, University of
California-San Diego, and San Diego State University, San Diego, Cal-
ifornia 92138
JASON W. BROWN, Department of Neurology, New York University
Medical Center, New York, New York 10016
HUGH W. BUCKINGHAM, Program in Linguistics and Communication
Disorders, Louisiana State University, Baton Rouge, Louisiana 70803
JEFFREY L. CUMMINGS, The Neurobehavior Unit, West Los Angeles
Veterans Administration Medical Center, Departments of Neurology
and Psychiatry and Behavioral Sciences, UCLA School of Medicine,
Los Angeles, California 90024
DEAN C. DELIS, San Diego Veterans Administration Medical Center, De-
partment of Psychiatry, University of California-San Diego, School of
Medicine, San Diego, California 92161
RENE DRUCKER-COLIN, Department of Neurosciences, Institute of Cel-
lular Physiology, National University of Mexico, Mexico D.F. 4510,
Mexico
JORGE ESLAVA-COBOS, Department of Child Neurology, Neurological
Institute of Colombia, Bogota, Colombia
ROBERT E. HANLON, Department of Psychology, City College of the City
University of New York, New York, New York 10031
vii
viii CONTRIBUTORS
ANDREW KERTESZ, University of Western Ontario, St. Joseph's Health

Centre, London, Ontario N6A 4V2, Canada
MARIA VICTORIA LOPEZ, Colombian Association of Neuropsychology,
Bogota, Colombia
IGNACIO MADRAZO, Department of Neurosurgery, Specialties Hospital,
Centro Medico "La Raza," IMSS, Mexico D.F. 02990, Mexico
LYDA MEJIA, Colombian School of Rehabilitation, University of Rosario,
Bogota, Colombia
FEGGY OSTROSKY-SOLIS, Department of Psychophysiology, Faculty of
Psychology, National University of Mexico, Mexico D.F. 4510,
Mexico
OSCAR PINZON, San Juan de Dios Hospital, National University of Co-
lombia, Bogota, Colombia
ROBERTO A. PRADO-ALCALA, Department of Physiology, Faculty of
Medicine, National University of Mexico, Mexico D.F. 4510, Mexico
LUIS QUINTANAR, Department of Psychophysiology, Faculty of Psy-
chology, National University of Mexico, Mexico D.F. 4510, Mexico
MONICA ROSSELLI, Konrad Lorenz Foundation University, Bogota, Co-
lombia, and Miami Institute of Psychology, Miami, Florida 33166-
6612
SIDNEY J. SEGALOWITZ, Department of Psychology, Brock University,
St. Catharines, Ontario L2S 3A1, Canada
EUGENIA SOLANO, Neurologic Institute of Colombia, Bogota, Colombia
DONALD G. STEIN, Dean of the Graduate School and Associate Provost
for Research, Rutgers University, Newark, New Jersey 07102
Preface
Neuropsychology has presented a particularly formidable array of devel-

opments during recent years. The number of methods, theoretical ap-
proaches, and publications has been steadily increasing, permitting a
step-by-step approach to a deeper understanding of the tremendously
complex relationships existing between brain and behavior.
This volume was planned as a collection of papers that, in one way or
another, present new research and clinical perspectives or interpretations
about brain-behavior relationships. Some chapters present new research
in specific topics, others summarize the evidence for a particular the-
oretical position, and others simply review the area and suggest new
perspectives of research. Consistent with the spirit in which the book was
planned, the authors present and propose new avenues for developing
neuropsychology and understanding the organization of cognitive activity.
Part I is devoted to basic theoretical and technical approaches in
studying brain organization of cognitive processes. Hanlon and Brown
("Microgenesis: Historical Review and Current Studies") present an over-
view of some clinical and experimental work from the standpoint of mi-
crogenetic theory. Microgenesis is considered to be the structural devel-
opment of a cognition through qualitatively different stages. The authors
discuss the growing dissatisfaction with both the old center and pathway
theories and the newer modular or componental accounts. They also ex-
plore how micro genesis can be extended to the interpretation of symp-
toms of brain damage in developing a structural model of hierarchic levels
through which the process of cognitive function unfolds.
Delis and Bihrle ("Fractionation of Spatial Cognition following Focal
and Diffuse Brain Damage") analyze visuospatial impairments that occur
subsequent to brain damage. They emphasize that spatial cognition, like
language, fractionates in precise ways following focal and diffuse brain
damage, and that selective deficits in visual attention, perception, and
visuoconstructional abilities can be distinguished.
Kertesz's chapter ("Anatomical and Physiological Correlations and
Neuroimaging Techniques in Language Disorders") details some of the
advantages and limitations of various localizing techniques and summa-
ix
x PREFACE
rizes the findings of the relatively new science of in vivo localization. The
author considers traditional (autopsy, head injury, surgical procedure,
cortical stimulation, EEG) and modern (isotope localization, comput-
erized axial tomography, cerebral blood flow, positron emission tomogra-
phy, single-photon computerized tomography, magnetic resonance imag-
ing) techniques for localizing lesions in an in-depth analysis focusing on
the issue of the extent to which a function is localizable and the variables
that influence its symptomatology.
Segalowitz ("ERPs and Advances in Neurolinguistics") presents a
review of some event-related potentials (ERP) techniques and technical
contributions directly related to questions of research in neurolinguistics.
He examines the application of ERP in the area of lexical semantics,
syntaxis, reading, and primary and secondary language acquisition, and in
the study of patients with disorders of language resulting from brain
damage. Segalowitz emphasizes the advantages of the latter paradigm as an
alternative technique to supplement the traditional behavioral measures.
Part II deals with language development and dissolution. Eslava-
Cobos and Mejia ("Disorders in Language Acquisition and Cerebral Matu-
ration: A Neurophysiological Perspective") present a historical review of
concepts of disorders in language acquisition. They discuss the embryo-
logic development of the nervous system with regard to normal language
acquisition. In addition, they review Latin American studies, especially
those carried out by Quiros and Azcoaga, emphasizing Pavlov's theory of
analyzers as applied to language by Bechtereva. The authors also discuss
abnormal language organization, trying to support a system of neu-
rophysiological classification.
Cummings and Benson ("Speech and Language Alterations in De-
mentia Syndromes") present a broad clinical study focusing on speech
and language in Alzheimer's dementia, Parkinson's disease dementia, and
multi-infarction dementia. They analyze distinguishing features for each
dementia and discuss differential profiles of preserved and impaired
functions in each dementia syndrome.
Part III deals with oral and written language disorders. Buckingham
("Mechanisms Underlying Aphasic Transformations") outlines a lan-
guage production model based on normality, underlining the fact that,
without some notion of how a normal linguistic productive system oper-
ates, there is little hope of understanding aphasic errors. He points out
that aphasic breakdowns arise through derailments of computation errors
that are corrected in normal language production but augmented and
sustained in aphasics. The author emphasizes that only through models of
normal language production is it possible to characterize mechanisms
underlying aphasic transformations.
Ardila, Rosselli, and Pinzon ("Alexia and Agraphia in Spanish
Speakers: CAT Correlations and Interlinguistic Analysis") present a large-
PREFACE xi
scale study of a brain-damaged population with regard to alterations in

reading and writing. The authors emphasize the necessity of developing
new models of alexias and agraphias and the importance of considering
particular characteristics of each language.
Ardila, Lopez, and Solano ("Semantic Aphasia Reconsidered"), in a
departure from some clinical cases, analyze the so-called semantic apha-
sia, which was postulated by Head and systematized by Luria. They pro-
pose that semantic aphasia and Gerstmann's syndrome conform to a sin-
gle syndrome and that underlying cognitive deficits are common.
Part IV is devoted to discussion of some relations of the basal ganglia
of the brain to cognitive activity. Ostrosky-Solis, Madrazo, Drucker-Colin,
and Quintanar ("Cognitive Effects of Adrenal Autografting in Parkinson's
Disease") present neuropsychological findings in patients with Parkin-
son's disease who have undergone adrenal autograft in Mexico City; those
patients had an extensive pre- and postoperatory cognitive evaluation.
The authors examine the changes observed in these patients, the cognitive
area in which modifications are observed, and the factors with which they
are correlated, and propose some explanatory hypotheses for the results
observed.
Prado-Alcala ("The Striatum as a Temporary Memory Store") pro-
poses that long-term memory implies the participation of different neu-
roanatomical and neurochemical systems that may be sequentially in-
volved. He discusses the view that the caudate nucleus and striatonigral
GABA neurons play a major role in the process underlying recent
memory.
Part V is devoted to recovery from brain damage. Stein ("Develop-
ment and Plasticity in the Central Nervous System: Organismic and En-
vironmental Influences") examines in detail some fundamental and the-
oretical problems with regard to plasticity of the nervous system and
recovery from brain damage, and analyzes contextual and environmental
factors influencing recovery. He focuses on the effect of different vari-
ables, such as the moment of the lesion, and individual differences in
recovery from brain damage, as well as the concept of plasticity.
Although many different topics are considered in this volume, all of
them in one way or another enlighten (from the conceptual, experimental,
or clinical point of view) the problem of the complex set of relationships
between language and cognitive organization of activity in the brain.
The main contemporary trends in neuropsychology are clearly re-
flected in this book: the emphasis on cognition and the necessity to carry
out an analytical approach to normal and pathological cognitive activity
under both normal and abnormal conditions; the effort to develop in-
creasingly sophisticated technical procedures to study psychological pro-
cesses; the interest in correlating neurological conditions with linguistic
events in the development as well as the dissolution of language; the
xii PREFACE
analysis and differential characteristics of dementia syndromes; the ap-

plication of complex linguistic concepts to neuropsychological analysis;
the emphasis on language similarities and differences when analyzing
language disorders; the current interest in studying participation of sub-
cortical structures in cognitive activity; the enormous interest in develop-
ing procedures to overcome brain dysfunction; and the myriad remaining
conceptual and practical problems.
We sincerely hope this collection of papers will further our develop-
ing knowledge of the brain's cerebral organization of language and cog-
nitive activity.
Alfredo Ardila
Feggy Ostrosky-Solis
Contents
Part I: THEORETICAL AND TECHNICAL APPROACHES
Chapter 1
Microgenesis: Historical Review and Current Studies 3
Robert E. Hanlon and Jason W. Brown
Early Studies ................................................ 3

Perceptgenesis .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Clinical Applications ........................................ 6
Experimental Studies ........................................ 10
Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
References .................................................. 13
Chapter 2
Fractionation of Spatial Cognition following Focal and Diffuse
Brain Damage ............................................... 17
Dean C. Delis and Amy M. Bihrle
Historical Perspective ........................................ 18

Visual Attention. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Parts and Wholes ............................................ 20
Studies of Visuospatial Processing Using Hierarchical Stimuli ... 21
Frequency and Severity of Visuospatial Deficits ................ 25
Intrahemispheric Dissociations in Visuospatial Processing ...... 27
Toward an Integration of Intra- and Interhemispheric
Dissociations in Visuospatial Processing .................. 28
Fractionation of Spatial Cognition in Diffuse Brain Pathology 29
xiii
xiv CONTENTS
Conclusion 32
References 32
Chapter 3
Anatomical and Physiological Correlations and Neuroimaging
Techniques in Language Disorders ............................ 37
Andrew Kertesz
Autopsy Location. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Localization through Head Injuries ............................ 40
Surgical Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Cortical Stimulation ......................................... 41
Electroencephalography and Evoked Potentials . . . . . . . . . . . . . . . . . 43
Isotope Localization of Lesions ............................... 44
Computerized Tomographic (CT) Lesion Localization ........... 45
Cerebral Blood Flow (CBF) Localization ....................... 48
Positron Emission Tomography (PET) ......................... 49
Single-Photon Computerized Tomography (SPECT) ............. 50
Magnetic Resonance Imager (MRI) .. . . .. . .. . .. . . ... . . .. . . . ..... 50
Discussion and Conclusion ................................... 52
References .................................................. 56
Chapter 4
ERPs and Advances in Neurolinguistics 61
Sidney J. Segalowitz
Overview of ERP Methodology .. , .................... '" . ..... 61

ERP versus Reaction Time .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
ERPs and Some Research Issues in Neurolinguistics ............ 69
Lexical Semantics ......................................... 69
Semantics versus Syntax ................................... 70
Reading in Normals.. .. . . . . .... . . . ..... . . .. .... . . .. ... ..... 71
Development of Reading Skills. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Language Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Second-Language Acquisition· .............................. 75
Issues in Clinical Neurolinguistics. . . . . . . . . . . . . . . . . . . . . . . . . . . 75
CONTENTS XV
Summary................................................... 77
References .................................................. 77
Part II: DEVELOPMENT AND DISSOLUTION OF LANGUAGE
Chapter 5
Disorders in Language Acquisition and Cerebral Maturation: A
Neurophysiological Perspective ............................... 85
Lyda Mejia and Jorge Eslava-Cobos
Organic and Functional Syndromes ........................... 88

Basic Embryologic Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Analyzers and Stereotypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Normal Language Organization: Physiologic Stable States ....... 93
Abnormal Language Organization: Pathologic Stable States ...... 98
Kinesthetic Motor Verbal Analyzer. . . . . . . . . . . . . . . . . . . . . . . .. . 98
Verbal Analyzer ........................................... 100
Conclusions ................................................. 102
References .................................................. 102
Chapter 6
Speech and Language Alterations in Dementia Syndromes 107
Jeffrey 1. Cummings and D. Frank Benson
Methodology... .. . . . . . . . . . . .... ... . . . ... . . . .. . . .. . . .. .. . . .. . 108

Assessment ............................................... 108
Participants ............................................... 109
Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Normal Control Subjects ................................... 110
Dementia of the Alzheimer Type. . . . . . .... .. .. .. .. . . .. . . . . .. 110
Parkinson's Disease ...................................... . . 112
Multi-Infarct Dementia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Discussion .................................................. 115
Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
References .................................................. 119
xvi CONTENTS
Part III: ORAL AND WRITTEN LANGUAGE DISORDERS
Chapter 7
Mechanisms Underlying Aphasic Transformations. . . . . . . . . . . . . . 123
Hugh W. Buckingham
The Garrett Model ........................................... 124

The Augmentations of the Garrett Model ...................... 128
Lexical Substitutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Lexical Blends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Phonemic Substitution ....................................... 134
Phonemic Transposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Constraints on Errors Conditioned by
Representative/Computational Knowledge .. . . . . . . . . . . . . . . . . 139
Summary ................................................... 142
References .................................................. 143
Chapter 8
Alexia and Agraphia in Spanish Speakers: CAT Correlations and
Interlinguistic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Alfredo Ardila, Monica Rosselli, and Oscar Pinzon
Method..................................................... 150
Subjects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Procedure ................................................. 150
Instrument ................................................ 154
Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Population with Cerebral Damage and Normal Population... .. 155
Patients with Right-Hemisphere Lesions and Patients
with Left-Hemisphere Lesions ............................ 155
Intrahemispheric Comparisons. . . . . . . . . . . . . . .. . . . . . . . . . . . . . . 155
Quantification of the Type of Errors. . . . .. . . . . . . . . . . . . . . . . . . . 160
Scores on the Rey-Osterrieth Complex Figure ................ 162
Superimposition of Scans .................................. 163
Discussion .................................................. 168
Patients with Right and Left Lesions ........................ 168
Intrahemispheric Comparisons. . . . . . . . . . . . . . . . . . . . . . . . . . .. . . 168
Topography of the Clinical Findings ........................ 172
An Interlinguistic Consideration ............................ 173
References .................................................. 174
CONTENTS xvii
Chapter 9
Semantic Aphasia Reconsidered 177
Alfredo Ardila, Maria Victoria Lopez, and Eugenia Solano
Case Reports ................................................ 180

Patient 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Patient 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Testing Procedure ......................................... 182
Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Discussion .................................................. 186
References .................................................. 192
Part IV: BASAL GANGLIA AND COGNITIVE ACTIVITY
Chapter 10
Cognitive Effects of Adrenal Autografting
in Parkinson's Disease ....................................... 197
Feggy Ostrosky-Solis, Ignacio Madrazo, Rene Drucker-Colin,
and Luis Quintanar
Pathology and Pathogenesis ",.".".,.,",.,",.",.,',."" 198

Treatment. , , .. , , , , . , , , , , .... , . , , , . , .. , . , . , . , , ...... , .... , . , . 201
Neurosurgical Treatment .,., .. ,.',."., ....... , ..... , .. ,',. 202
Pharmacological Treatment, . , ... , .. , ...... , .. , , . , ...... , .. , 202
Tissue Transplants ......... , , , . , ..... , .. , ..... , ........ , , , , 203
Adrenal Medullary Grafts in PD Patients .................. , ... , 205
Surgical Technique ." .. ,.,"' .. ," .... ,"', .......... ,.... 206
Neuropsychological Findings .,",.,""""", ..... ".,...... 206
General Considerations, , , , , . , , . , , , , , , , , . , , , .. , .. , .. , . , , . , , . , . 212
References " .. ".,., .. ,.,',.,"',.,",.,., ..... , .. ".,., .. ,. 214
Chapter 11
The Striatum as a Temporary Memory Store 219
Roberto A. Prado-Alcala
Involvement of the Neostriatum in Memory,. ,.,. , ........ , , ,., 219

The Overtraining Effect ., .. ,', .. , .... , .. , .. ,.,",.,., ... ,"" 219
Conclusions, , ............................ , , . . . . . . . . . . . . . . . . . 223
References .................................................. 224
xviii CONTENTS
Part V: RECOVERY FROM BRAIN DAMAGE
Chapter 12
Development and Plasticity in the Central Nervous System:
Organismic and Environmental Influences ..................... 229
Donald G. Stein
Plasticity and Levels of Analysis of CNS Functions. . . . . . . . . . . . . 234

Environmental and Behavioral Factors
"Shape" CNS Morphology. . . . . . . . . . . . . . . .. . . . . . . . . . . . . .. . 236
Temporal Factors also Important in Predicting
the Outcome of Brain Damage ........................... 238
Adaptative and Maladaptative Anomalous Growth
in Response to CNS Injury ............................... 239
Behavioral Consequences of Lesions Early in Life
vis-a-vis Injuries in Maturity ............................. 241
Important Individual Differences in Response to Brain Injury
That Are Not Adequately Studied ........................ 244
Summary and Concluding Remarks ........................... 249
References .................................................. 250
Index........................................................ 253
Brain Organization
of Language and
Cognitive Processes
I
Theoretical and Technical

Approaches
1
Microgenesis
Historical Review and Current Studies
ROBERT E. HANLON and JASON W. BROWN
Currently, there is renewed interest in process-oriented approaches to the

study of mental phenomena. In this chapter we review one of the more
established approaches of this type, that of microgenesis. Originating
from assumptions on the structural development of thought by the early
experimental psychologists of Wiirzburg and Leipzig, the fundamental
principles of microgenesis have been progressively elaborated.
First, a definition of the term, or rather a set of definitions, to specify
and delimit the process that this term denotes. Microgenesis refers to the
structural development of a cognition (idea, percept, act) through qualita-
tively different stages. The temporal period of this development extends
from the inception of the cognition to its final representation in .con-
sciousness or actualization (expression) in behavior. In theoretical work
on micI:ogenesis, two major schools of thought can be distinguished, in
terms of experimental and quasi-experimental methods utilized to dem-
onstrate this process. These two schools relate to the microgeny of
thoughts and the microgeny of percepts. Flavell and Draguns (1957) noted
this methodological division but cautioned against any assumptions re-
garding a functional dichotomy.
EARLY STUDIES
Approaches to the analysis of the microgenesis of thought predomi-

nantly entail the description and interpretation of the microtemporal se-
ROBERT E. HANLON· Department of Psychology, City College of the City University of

New York, New York, New York 10031. JASON W. BROWN· Department of Neurology,
New York University Medical Center, New York, New York 10016.
3
4 ROBERT E. HANLON and JASON W. BROWN
ries of conceptual stages that unfold, following the presentation of a

stimulus, to a behavioral response. This approach lends itself to clinical
application and to the interpretation of thought disorders and cognitive
disorganization in various pathological populations. Such microgenetic
models have been explored by a small group of theorists including Brown
(1977), Ey (1950), Schilder (1951, 1953), and Werner (1956). Werner, in
fact, is credited with coining the term microgenesis in translation of the
German word Aktualgenese. The microprocessing approach of these
clinicians incorporates the Jacksonian principle of hierarchical,organiza-
tion within the stage model of micro genetic progression. Investigations in
perceptual microgenesis have been conducted in a very different manner
with a stronger emphasis on stimulus properties and control. The classic
experimental paradigm consists in the repeated presentation of a stimulus
after an initially obscured or fragmented stage, through subsequent stages
of increasing clarity, until the stimulus becomes clearly differentiated and
complete. Draguns (1984, p. 5) notes that this approach "involves progres-
sions of stimuli from maximal or pronounced information deprivation to
the presentation of adequate or optimal amounts of information for the
response or the decision at hand." While this methodological design was
originally developed by Sander (1928), later experimentalists who have
utilized and expanded on this technique with normals and pathological
groups include Draguns (1967), Draguns and Multari (1961), Flavell
(1956), Froelich (1978, 1984), Kragh (1955), and Smith (1956, 1984).
This discussion will begin with a review of the historical develop-
ment of microgenetic theory and experimentation (see also Brown, 1972,
1977, 1988a; Flavell & Draguns, 1957). Despite the methodological draw-
backs of the introspective approach to the study of psychological pro-
cesses, the work of the Wiirzburg group around the turn of the century
produced the conceptual precursors of microgenesis. Humphrey (1963)
reviewed in detail early experiments by Mayer and Orth (1901) and Marbe
(1901). Though crude, these experiments led to the description of a non-
sensory or imageless event or state of consciousness immediately follow-
ing the presentation of a stimulus that could not be analyzed, termed the
Bewusstseinslage. In an attempt to elucidate the Bewusstseinslage of
meaning, Messer (1906) delineated a tripartite model of meaning from
subjects' responses following the presentation of a word. According to
Messer, the Bewusstseinslage constituted an intermediate or transitional
stage between the initial direct derivation of the word's meaning from its
presentation and the accessory image of meaning.
Ach (1905) expanded the Wiirzburg doctrine through studies on
awareness or the experience of knowing (Wissen). Through elaborate ex-
periments on sensory reactions, Ach proposed a second stage of imageless
knowing, which he termed Bewusstheit. This stage of awareness and ex-
pectation represented a phase in the process characterized by willed and
directed thought. In other words, following the impalpable Bewusstseins-
MICROGENESIS 5
lage, during which there occurred only the mere apprehension of a

thought, the process advances to the Bewusstheit, at which point there is a
distinct, yet vague, awareness, which constitutes a state of imageless
knowing.
Following these early Wurzburger proposals on the dynamic nature of
imageless thought and particularly the transitional stages of thought forma-
tion, Pick (1913) and Jung (1919) presented theoretical models of thought
development based on temporal hierarchies. From clinical experience in
aphasia, Pick delineated four hypothetical stages through which thlilught
develops in the process of language production. The first of the two pre-
linguistic stages consisted of the initial formulation of the thought as a
loose structural assembly of components. A stage of structural thought then
emerges during which the predicative aspects are arranged. The third stage
involves the development of the pattern for the utterance, and finally, word
choice in conformity with this pattern (see De Bleser, 1987).
Jung's hierarchical model developed in relation to work on word
associations. According to Jung, in the "apperceptive" process cognition
progressed from an immediate superficial response that is ordinarily sup-
pressed, through a contextual stage such as that of the phrase, and finally
to a semantic level of denotative and connotative response. Pick (1931)
described a comparable three-stage model, the first stage consisting of the
initial recognition of the physical attributes of the word, the second of
access to the word's "meaning sphere" or its position in conceptual
space, and third, the determination of its grammatical character.
PERCEPTGENESIS
Until this time, process-oriented studies and theories were concerned

with the micro development of thought and language. The influence of the
Gestalt movement, however, gave rise to experiments on the temporal
development of percepts. Sander (1928) presented a theory of perception
in which this perceptual micro development was referred to as Aktual-
genese (momentary genesis). The basic premise of this theory was that the
process of perception consisted of several phases, which develop from an
initial stage of a diffuse, undifferentiated percept through subsequent
stages of differentiation and discrimination, leading to a distinct configu-
ration. Sander described four discrete phases of perceptual microdevelop-
ment. Initially, there is a stage of an amorphous undifferentiated percept.
Next, the percept achieves course figure-ground properties. The third
stage is pre-Gestalt (Vorgestalt), or preconfigurational, and is viewed as a
labile, preparatory phase for the subsequent derivation of a veridical ob-
ject. This stage is accompanied by a transient unpleasant anticipation of
structural completion.
Sander was instrumental in the development and application of ex-
perimental techniques to demonstrate microgenetic phenomena. These

paradigms involved the presentation of a stimulus under obstructive or
inadequate conditions (e.g., brief tachistoscopic exposure, poor illumina-
tion, peripheral location). The initial presentation was followed by a se-
ries of presentations in which there was a successive increase in clarity,
distinctiveness, and/or complexity. As a result of experimental evidence
provided by the Sander group, clinicians again began to theorize on pa-
thology within a micro genetic framework. Two pathological conditions
that received considerable attention were aphasia and schizophJenia.
CLINICAL APPLICATIONS
Building primarily on the findings of Sander, Conrad (1947) present-

ed a microgenetic interpretation of aphasia. In accordance with Sander's
four-phase progression in perceptual microgeny, Conrad described four
levels of disorder that can appear in aphasia. Aphasic symptoms, particu-
larly word-finding difficulty, represented an arrest in the Gestalt forma-
tion process. The highest level of disorder in this process resulted in a
fully developed end-Gestalt that was qualitatively distorted. The second
level, which coincides with Sander's Vorgestalt, consisted of incomplete
structural differentiation. Next is a level of blurred figure-ground discrim-
ination, while the lowest level is characterized by an absence of formative
processing. Conrad argued that cognition is a double-edged process of
differentiation and integration, and either one or both of these compo-
nents are disrupted in aphasia.
Arguably the most imaginitive micro genetic framework for thought
formation was provided by Schilder (1951), who drew heavily on the
work of the Wurzburgers, particularly Ach ana Messer, and studies of
imageless thought. Schilder's theory was derived from studies of psycho-
pathology, particularly schizophrenia and the effects of brain damage. He
insisted that thought disorders "can all be understood as abortive forma-
tions produced in the course of the differentiation-process of thought" (p.
513). This notion is based on the idea that a "presentation" or a vague
content is progressively incorporated within the developing thought
structure. On this view, thought originates in a diffuse concept of the
direction the thought is to take in its structural development. The deter-
mination of this directionality conforms to the overall goal of the thought
development, which is roughly established at the moment of conception.
Following the establishment of a goal and the determination of the direc-
tion that must be taken in order to reach this goal, the thought develops
through a series of preliminary stages, constituting the preparatory phase
of thought. During this preparatory phase, mental contents-presenta-
tions-impinge on and influence the thought formation. These presenta-
MICROGENESIS 7
tions, which may take the form of images or ideas, then function to aid in
the process of constructing the thought.
Though not explicit as to the manner in which some presentations
influence and are integrated in the process of thought development while
others are discarded, Schilder maintained that the dual processes of sim-
ilarity and contiguity playa governing role-that is, "a presentation
evolves through other presentations which are associated with it by sim-
ilarity and contiguity, primarily by clang and external associations" (p.
499). The idea of inclusion versus exclusion entails that presentations
holding a logical relevance to the developing thought will be incorporated
into its structure, while those without a logical relationship fade into the
background or periphery.
One of the more intriguing aspects of Schilder's work is the conten-
tion that these early, primitive presentations that contribute to the devel-
oping thought are symbolic images that aid in the apprehension of mean-
ing. Moreover, these preparatory presentations have a tendency to fuse
with associated presentations, as well as the capacity for alteration and
transformation in response to emotional influences.
The direction of this process is a progressive evolvement toward ob-
ject-directed, reality-oriented thought. Two major phases were differenti-
ated within this unidirectional process, along with the primary determi-
nants of the progression within each phase. Subjective, wish-determined
symbolic images of the preparatory stages were referred to as the sphere.
The thought structures that arise from this sphere show a progressive
striving for objectivity in a manner that increasingly conforms to the rules
of logic and the nature of reality. In this way the structure finally achieves
the form of a concept. This conceptual phase represents the elaboration
and structuration of the symbolic images of the sphere toward objectivity,
resulting in the clear realization of the meaning content of the thought.
As mentioned, Schilder's theory of microgenesis is vast and encom-
passing, especially with regard to the interpretation of different forms of
impaired cognitive processing in various pathological conditions. A fun-
damental idea in his writings is that the process of thought development
recapitulates the phyloontogeny of cognition. In addition, and in keeping
with the Jacksonian idea of hierarchical organization, Schilder (1953)
concludes that this developmental progression occurs through the psy-
chic "energy-exchanges" between the neural levels in the hierarchy.
Schizophrenic thought disorders, expressive and receptive aphasia, the
cognitive impairments such as dementia, agnosia, apraxia-essentially
the entire realm of neuropsychology-reflects an incomplete develop-
ment through the micro genetic levels of cognition.
Another major microgenetic theorist was Werner (1956), who set
forth two basic assumptions: "First, the functions underlying abnormal
behavior are in their essence not different from those underlying normal
behavior. Second, any human activity such as perceiving, thinking, act-

ing, etc., is an unfolding process, and this unfolding of microgenesis,
whether it takes seconds or hours or days, occurs in developmental se-
quence" (p. 347). Drawing on the work of Pick and Schilder, as well as his
own experiments, he described the semantic spheres through which lin-
guistic processes develop in microgenetic progression. In early tachisto-
scopic experiments, subjects reported that verbal stimuli aroused feelings
of meanings or experiences of the semantic sphere of a stimulus prior to
the articulation of the word. Werner elaborated on the dynamic character
of this sphere of cognition in terms of "organismic involvement."
These organismic factors were revealed during the genesi.s of a per-
cept-perceptgenesis-prior to actual word recognition. One such factor
involves a vague awareness or impression of the meaning sphere of
words, such as the feeling of warmth aroused by the initial, brief ta-
chistoscopic exposure of the stimulus "gentle wind." Aspects of direc-
tional or intensity properties of the stimulus word then follow. Contextual
or situational aspects of the experience of word meaning may also devel-
op prior to full realization and discrimination of word form.
Werner extended these microdevelopmental findings in normals on
the word recognition paradigm to aphasics in confrontation naming. He
noted that "in many instances, aphasic cognitions emerge in form of
organismic-bodily reactions which seem to be comparable to the re-
sponses of our normal subjects" (p. 349). Responses reflecting these orga-
nismic-bodily factors entailed verbalizations consistent with the action
sphere of the stimulus item (Le., "blowing" in response to the picture of a
trumpet). Similarly, the impression of a direction or vector generated in
normals on word recognition were also formed in aphasics in object nam-
ing. While unable to generate the accurate name, aphasics were able to
access and describe the "vectorial-physiognomonic" character of the
word. In accordance with the third feature of spheric cognition demon-
strated in the percept-genetic studies-the achievement of "atmospheric-
contextual" properties of words-a comparable picture was seen in apha-
sics. In Werner's view this feature was exemplified by the achievement of
the categorical sphere to which the object belonged. Further, depending
on task requirements, the contextual nature of the stimulus words may
appear in production of either the semantic or the grammatical category of
the item. In addition, he believed that the form of a given paraphasia
reflects the spheric character at which derailment in the micro develop-
mental progression occurred.
In studies of perceptual disorders such as agnosia, P6tzl (1960) noted
the recurrence (intrusion) of unreported elements in subsequent object
descriptions. For example, a green asparagus stalk that was not reported
on one task might recur in confabulatory form as a green tie, in the de-
scription of a person a few moments later. P6tzl confirmed this effect in
MICROGENESIS 9
normal subjects with tachistoscopic methods. He briefly presented scenes

to subjects and found that unreported fragments were integrated into
dreams and could be recovered in morning dream reports. Fisher (1960)
confirmed these subliminal effects. The implications of these findings,
that subconscious residues are linked to early stages in object perception,
and that the symptoms of object breakdown can be reproduced through
experiments in percept formation, were a major theme in P6tzl's writings
(see Brown, 1988b) and played a prominent role in subsequent research
on perceptgenesis.
The organodynamic theory of Ey (1950) represents another process
model based on the idea of evolutionary levels of cognition. Symptomatic
manifestations of psychosis and dementia reflect various levels of disor-
ganization of consciousness. Evans (1972) contributed a detailed summa-
ry of Ey's theory. Drawing on the work of Jackson and Freud, Ey incorpo-
rates their concepts into an organodynamic theory of consciousness. The
concept of dissolution of function as the basis of psychopathological
symptomatology was derived from H. Jackson and Herbert Spencer. The
process of dissolution entails both evolution and hierarchic order. Ey took
the concept of repression from psychoanalytic theory but considered it
conscious activity necessary for the organization of actuality. Ey believed
consciousness to be a dynamic, nonlocalized activity in which the various
cognitive processes participate, not merely an end result of this evolution-
ary progression. Rather, it is the actualization of the collective contribu-
tion of each level within this hierarchical process.
The structure of consciousness represents a power of legislation. of control, of
order and direction. The pathology of mental illness is a process of regression.
But the process is an organically determined one in the same way that the
symbolic elaboration of dreams occurs in and by sleep. Symptomatology mere-
ly indicates the level of dissolution at one point in time and to consider a
symptom in isolation from the pathological process clouds the issue. (Evans.
1972. p. 418)
Ey elaborated three levels of disorganization of consciousness in psy-

chosis that, together with the clinical expression of their respective disor-
ders, conform to a hierarchical gradient of conscious organization. The
highest level of psychotic disorganization within this model, the manic-
depressive level, represents a breakdown in the temporal structure of
consciousness. The development of control over psychological time is the
most recent ontogenetic acquisition of consciousness. The next level in
the hierarchy, that of delusion and hallucination, represents a dissolution
of spatial orientation within the field of consciousness. This spatial com-
ponent refers to existential space or the manner in which one represents
the world. This conscious representation of the world develops through
the structural process of thought formation that is based on the construc-
tion and incorporation of reality-oriented images. When this representa-
tional process breaks down, the subjective and objective components of

imagery become confused. Delusions and hallucinations reflect the erup-
tion of subjective imaginal components into the field of consciousness
and a general disorganization of existential space. There is a detachment
and projection of the content of thought.
Ey's lowest level of disorganization of consciousness is the confu-
sional, oneiroid level. Consciousness is no longer constructed into a phe-
nomenal field and is unable to represent a world of reality. Oneirism, the
pathological dream state, is therefore a manifestation of the deepest level
of conscious disorganization, which may be characterized as a disruption
at a very primitive stage of thought formation. The structural components
of developing consciousness are undifferentiated and fragmented, result-
ing in a miasma of confusing images.
More recently, Brown (see 1988a) has extended the microgenetic con-
cept to the interpretation of symptoms of brain damage in developing a
structural model of hierarchic levels through which the process of cog-
nitive formation unfolds. In this model, the microgeny of acts and per-
cepts is held to be identical to that for language, or affect, while other
cognitive functions, such as thought or memory, are interpreted in funda-
mentally the same way. A basic assumption is that the symptoms of brain
damage, like those of psychopathology, represent a premature exposure of
preliminary levels in the microstructure of cognition that are normally
transformed. On this view, symptoms take on a new meaning; they are not
aberrations of brain injury but pieces of preliminary mentation that can be
used to reconstruct the sequence of normal cognitive processing. Another
assumption is that cognitions unfold in a vertical, unidirectional (bottom-
up) manner over distributed anatomical systems, and that the sequence of
this unfolding is the same as that in which the anatomical systems ap-
peared in the course of forebrain evolution. In other words, microgeny
retraces the phyletic history of brain development. The significance of
this work is that it provides a neurological framework for what has been to
date largely a psychological theory, and attempts to fill in the detail of
level-to-level transformation using the clinical material as a guide to the
processing that is taking place.
EXPERIMENTAL STUDIES
Returning to experimental psychology and the classic microgenetic

methods handed down from Sander, there was comparatively little work
on perceptgenesis from the 1930s through the mid-1950s. This lack of
microgenetic and other process-oriented approaches was due chiefly to
the predominance of behaviorism throughout this time. However, with
renewed interest in information processing and mediational activity,
MICROGENESIS 11
there has been a modest renewal of micro genetic research (Draguns,

1984). Still, there are only a limited number of researchers actively in-
volved in micro genetic experimentation. This research continues to be
conducted mainly in Europe, most notably by Froelich and his group at
the University of Mainz, Smith at Lund University, and Kragh at the
University of Oslo.
Except for Werner's comparative-developmental studies (1948) at
Clark University, Dragun's continuation of earlier work with Flavell (Fla-
vell & Draguns, 1957), and Rimoldi's (1960, 1967) work on problem solv-
ing, experimental techniques and micro genetic theory have not found a
firm foothold in America.
Froelich (1984) addressed the functional frame of reference for micro-
genetic procedures with respect to objectives and alternative methods. As
outlined, the standard method of stimulus presentation entails the pre-
sentation of a patterned stimulus in a manner that provides for a sequen-
tial (trial-to-trial) increase in clarity, distinctiveness, or complexity. The
subject must cope with an obscured, obstructed, too brief, or fragmented
informational pattern that delays the formation of a stable cognitive re-
sponse. Froelich (1978) proposed that the motivational force underlying
the micro development may be conceptualized as "effort after coping with
ambiguity." In terms of information processing, these conditions induce a
fractionated evolvement through a series of qualitatively different stages.
With respect to information processing, one can distinguish two main
experimental procedures. Merogenetic techniques entail a progression of
stimulus presentations from figural fragmentation to an intact configura-
tional whole. This procedure elicits imaginal activity, guessing and hy-
pothesizing through the process of sequential comparison. Hologenetic
techniques are based on a progressive change in the energy level of pre-
sentation of a figurally intact stimulus (Le., luminosity, exposure time,
clarity). These procedures reveal the evolvement of figure-ground dis-
criminations through the stages of meaning assessment (Froelich, 1984).
In Scandinavia, primarily through the work of the Lund group
(Kragh, 1955; Kragh & Smith, 1970; Smith & Danielsson, 1982), micro-
genetic research has progressed within the domain of percept-genetic ap-
proaches to personality. Kragh (1955) described changes in thematic con-
tent on merogenetic procedures from a psychodynamic perspective.
These changes were consistent with micro genetic stages progressing from
ambiguity to stabilization. More recently, Smith and his group have em-
ployed experimental procedures based on microgenetic principles in the
assessment and differentiation of various psychopathological populations
within a psychoanalytic frame of reference. Experimentation with serial
afterimages (Smith, Fries, Andersson, & Reid, 1971; Smith & Kragh, 1967)
demonstrated that successive change in the size of microgenetic after-
image serials could discriminate schizophrenia, depression, and anxiety
states. Two tachistoscopic procedures, the Meta-Contrast Technique

(MCT; Kragh & Smith, 1970) and the Defense Mechanism Test (DMT;
Kragh, 1969; Westerlundh, 1976, 1984), in which incongruent and/or
threatening stimuli are successively presented in order to induce anxiety
and perceptual defenses, have been utilized in the assessment of normal
and psychopathological personality organization.
Smith and Danielsson (1982) have applied these experimental tech-
niques in percept-genetic studies of defensive strategies in Swedish
schoolchildren. Incorporating psychoanalytic concepts and Piagetian
ideas on cognitive development within a "dynamic-developmental per-
spective," they revealed "a series of transitions in anxiety manifestations
and defensive strategies from early preschool age to adolescence" (p. 190).
Johanson, Risberg, Silfverskiold, and Smith (1986) have also used MCT
anxiety responses as activation stimuli during rCBF measures in patients
with anxiety neurosis. Anxiety provocation resulted in a significant in-
crease in regional blood flow in the left frontoorbital area.
Cegalis and colleagues at Syracuse (Cegalis, 1973; Cegalis & Leen,
1977; Cegalis & Young, 1974) studied the role of conflict in microgenesis.
On the assumption that "disequilibration is a necessary condition for
extended microgenetic development" (Cegalis, 1984, p. 108), Cegalis has
utilized different types of stimulus distortions as dis equilibrating situa-
tions in the demonstration of individual differences in the reactions to
conflict.
CONCLUSION
This is a brief overview of some clinical and experimental work from

a microgenetic perspective. Currently, there is little work from this the-
oretical angle but there is considerable interest in its applicability, pro-
voked in large part by such findings as those of "blindsight," early seman-
tic activation, behavior without awareness, and so on, which cannot
easily be resolved within existing models. One also senses a growing
dissatisfaction with old center and pathway theories and the newer mod-
ular or componential accounts to which they have given rise, for these
also cannot deal with the richness and variety of normal and abnormal
behavior, and they are fatally uncoupled from evolutionary and other
process-oriented studies of brain function. In the final analysis, however,
the test of the micro genetic model will rest on brain-imaging, electrical,
magnetoencephalographic, or other studies that have the capacity to re-
cord the sequence of entrainment or engagement of specific brain systems
in a brief creative or behavioral performance. This is an important goal for
future research.
Ultimately, our concept of the mind is linked to our concept of the
MICROGENESIS 13
world around us. This is no less true for cognitivist or AI theories of mind
than for the micro genetic theory. Thus, microgenesis entails that objects
in the world are the end products of a chain of events in the mind, not a
starting point in a piecemeal constructive process. In this sense, the mi-
crogenetic idea seems to be bound up with an idealist stand on the nature
of externality. Microgenesis takes a strong position with regard to this
question, so that the data supporting micro genetic theory can be taken as
evidence for the correctness of its epistemological basis.
Similarly, the notion that representations develop out of subsurface
stages, levels beneath awareness, has implications for our understanding
of what a mental representation is. What it definitely is not is its proposi-
tional or perceptual content, for this is only the tip of its structure. On the
micro genetic view, conscious representations are not the instigators of
behavior but are rather outcomes of the processing rising from below. This
clearly has profound consequences for our views on volition and agency.
In fact, current physiological work lends support to the idea that actions
develop prior to our awareness of them, and that the sense of volition does
not stand behind and direct behavior but is given to awareness on the
heels of the action development. These philosophical issues, and their
admittedly somewhat unpalatable solutions, are deeply woven into the
fabric of micro genetic thinking, perhaps accounting for some of the re-
sistance to this concept. But the very fact that our idea of what the world
is depends so importantly on our theory of mind and brain is precisely the
reason why this theory is worth a closer look.
REFERENCES
Ach, N. (1905). Uber die Willenstatigheit und das Denken. Gottingen: Vandenhoeck and
Ruprecht.
Brown. J. W. (1972). Aphasia. apraxia and agnosia: Clinical and theoretical aspects.
Springfield. IL: Thomas.
Brown. J. W. (1977). Mind, brain. and consciousness. New York: Academic Press.
Brown, J. W. (1988a). The life of the mind: Selected papers. Hillsdale, NJ: Erlbaum.
Brown. J. W. (1988b). Classics in neuropsychology: Selected papers on agnosia and apraxia.
Hillsdale. NJ: Erlbaum.
Cegalis. J. A. (1973). Prism distortion and accommodtive change. Perception and Psycho-
physics, 13, 494-498.
Cegalis, J. A. (1984). On the role of conflict in microgenesis. In W. Froehlich, G. Smith. J.
Dr':lguns, & U. Hentschel (Eds.), Psychological processes in cognition and personality.
Washington, DC: Hemisphere.
Cegalis, J. A., & Leen, D. (1977). Individual differences in responses to induced perceptual
conflict. Perceptual and Motor Skills, 44, 991-998.
Cegalis, J. A, & Young, R. (1974). The effect of inversion-induced conflict on field-depen-
dence. Journal of Abnormal Psychology, 83, 373-379.
Conrad. K. (1947). Dber den Begriff der Vorgestalt und seine Bedeutung fur die Hirnpathol-
ogie. Nervenarzt, 18, 289-293.
De Bleser, R. (1987). From agrammatism to paragrammatism: German aphasiological tradi-

tions and grammatical disturbances. Cognitive Neuropsychology, 4, 187-256.
Draguns, J. (1967). Affective meaning of reduced stimulus input: A study by means of the
semantic differential. Canadian Journal of Psychology, 21, 231-241.
Draguns, J. (1984). Microgenesis by any other name ... In W. Froehlich, G. Smith, J. Drag-
uns, & U. Hentschel (Eds.), Psychological processes in cognition and personality.
Washington, DC: Hemisphere.
Draguns, J., & Multari, G. (1961). Recognition of perceptually ambiguous stimuli in grade
school children. Child Development, 32, 541-550.
Evans, P. (1972). Henri Ey's concepts of the organization of consciousness and its disor-
ganization: An extension of Jacksonian theory. Brain, 95, 413-440.
Ey, H. (1950-1954). Etudes psychiatriques (Vols. 1-3). Paris: Desclee de Brouwer.
Fisher, C. (1960). Introduction: Preconscious stimulation in dreams, associations, and im-
ages. Psychological Issues, 2, 1-40.
Flavell, J. H. (1956). Abstract thinking and social behavior in schizophrenia. Journal of
Abnormal Social Psychology, 52, 208-211.
Flavell, J. H., & Draguns, J. (1957). A microgenetic approach to perception and thought.
Psychological Bulletin, 54, 197-217.
Froehlich, W. D. (1978). Stress, anxiety, and the control of attention: A psychophysiological
approach. In C. D. Spielberger & 1. G. Sarason (Eds.), Stress and anxiety (Vol. 5). Wash-
ington, DC: Hemisphere.
Froehlich, W. D. (1984). Microgenesis as a functional approach to information processing
through search. In W. Froehlich, G. Smith, J. Draguns, & U. Hentschel (Eds.), Psychologi-
cal processes in cognition and personality. Washington, DC: Hemisphere.
Humphrey, G. (1963). Thinking: An introduction to its experimental psychology. New York:
Wiley.
Johanson, A. M., Risberg, J., Silfverskiold, P., & Smith, G. (1986). Regional changes in cere-
bral blood flow during increased anxiety in patients with anxiety neurosis. In U.
Hentschel, G. Smith, & J. G. Draguns (Eds.), The roots of perception. Amsterdam:
Elsevier North-Holland.
Jung, C. G. (1919). Studies in word association (M. D. Eder, Trans.). New York: Moffat.
Kragh, U. (1955). The actual-genetic model of perception-personality. Lund: Gleerup;
Copenhagen: Munksgaard.
Kragh, U. (1969). DMT-defense mechanism test. Stockholm: Skandinaviska Testforlaget.
Kragh, U., & Smith, G. J. W. (Eds.). (1970). Percept-genetic analysis. Lund: Gleerup.
Marbe, K. (1901). Experimentell-psychologische Untersuchungen iiber das Urteil. Leipzig:
Engelmann.
Mayer, A., & Orth, J. (1901). Zeitschriftfiir Psychologie und Physiologie der Sinnesorgan, 26.
Messer, A. (1906). Archives Gesamte Psychologie, B, 1-224.
Pick, A. (1913). Die agrammatischen Sprachstorungen. Berlin: Springer.
Pick, A. (1931). Aphasie. In A. Bethe & G. von Bergmann (Eds.), Handbuch der normalen
und pathologischen Physiologie (Vol. 15). Berlin: Springer.
P6tzl, O. (1960). The relationship between experimentally induced dream images and indi-
rect vision. Psychological Issues, 2, 41-120.
Rimoldi, H. J. A. (1960). Problem solving as a process. Educational and Psychological
Measurement, 20, 249-260.
Rimoldi, H. J. A. (1967). Thinking and language. Archives of General Psychiatry, 17, 568-
576.
Sander, F. (1928). Experimentelle Ergebnisse der Gestaltpsychologie. In E. Becher (Ed.), 10
Kongres bericht experimentelle Psychologie. Jena: Fischer.
Schilder, P. (1951). On the development of thoughts. In D. Rapaport (Ed.), Organization and
pathology of thought. New York: Columbia University Press.
MICROGENESIS 15
Schilder, P. (1953). Medical psychology. New York: International Universities Press.

Smith, G. J. W. (1956). Review of U. Kragh's "The actual-genetic model of perception-
personality." Theoria, 22, 61-69.
Smith, G. J. W. (1984). Stabilization and a'ltomatization of perceptual activity over time. In
W. Froelich, G. Smith, J. Draguns, & U. Hentschel (Eds.), Psychological processes in
cognition and personality. Washington, DC: Hemisphere.
Smith, G. J. W., & Danielsson, A. (1982). Anxiety and defensive strategies in childhood and
adolescence. New York: International Universities Press.
Smith, G. J. W., Fries, I., Andersson, A. L., & Reid, J. (1971). Diagnostic exploitation of visual
after-effect measures in a moderately depressive patient group. Scandinavian Journal of
Psychology, 12, 67-79.
Smith, G. J. W., & Kragh, U. (1967). A serial afterimage experiment in clinical diagnostics.
Scandinavian Journal of Psychology, 8, 52-64.
Werner, H. (1948). Comparative psychology of mental development (2nd ed.). New York:
International Universities Press.
Werner, H. (1956). Microgenesis and aphasia. Journal of Abnormal Social Psychology, 52,
347-353.
Westerlundh, B. (1976). Aggression, anxiety and defense. Lund: Gleerup.
Westerlundh, B. (1984). Perceptgenesis and the experimental study of conflict and defense.
In W. Froehlich, G. Smith, J. Draguns, & U. Hentschel (Eds.), Psychological processes in
cognition and personality. Washington, DC: Hemisphere.
2
Fractionation of Spatial Cognition

following Focal and Diffuse Brain
Damage
DEAN C. DELIS and AMY M. BIHRLE
Visuospatial dysfunction denotes a constellation of perceptual and con-

structional deficits in which the ability to organize visual features into
perceptual wholes is lost or impaired. Despite the high incidence of
visuospatial deficits following brain damage, there have been relatively
few neuropsychological investigations of visuospatial dysfunction rela-
tive to the large number of studies of aphasia. Several reasons may under-
lie this disparity. While the structure of language is discrete and readily
observable (Luria, 1966), the structure of visuospatial cognition is much
more elusive. Visuospatial structure, unlike linguistic structure, does not
contain temporally discrete entities, such as phonemes and words, but
must be inferred from perceptual judgments and visuospatial construc-
tions. This diffuse quality hinders the development of a grammar of visu-
ospatial thinking (Chomsky, 1965; Fillmore, 1971). Additionally, specific
brain lesions often interrupt aspects of linguistic processes in well-de-
fined ways, enabling the development of reliable taxonomies of aphasias
(Geschwind, 1970; Goodglass & Kaplan, 1972). In contrast, when visuo-
spatial abilities are disrupted by brain damage, the nature of the impair-
ment is difficult to characterize in a precise manner. Not surprisingly, one
rarely refers to visuospatial syndromes resulting from specific brain le-
sions; instead, difficulties on visuospatial tasks tend to be amalgamated
DEAN C. DELIS· San Diego Veterans Administration Medical Center, Department of Psy-
chiatry, University of California-San Diego, School of Medicine, San Diego, California
92161. AMY M. BIHRLE • The Salk Institute for Biological Studies, University of Cal-
ifornia-San Diego, and San Diego State University, San Diego, California 92138.
17
18 DEAN C. DELIS and AMY M. BIHRLE
into one category and referred to as "visuospatial dysfunction," "con-

structional apraxia," or "constructional difficulties."
HISTORICAL PERSPECTIVE
Visuospatial difficulties in neurological patients were first noted by

Hughlings Jackson in 1876. He coined the term imperception to refer to a
wide range of spatial deficits, including the inability to recognize familiar
faces, find one's way around the environment, and dress oneself. Jackson
implicated the posterior region of the right hemisphere in spatial process-
ing. In 1934 Kleist introduced the term constructional apraxia, which, he
posited, resulted from a disconnection between the perceptual and motor
regions of the cortex. The term constructional apraxia was later extended
beyond its original meaning of disconnection to include any visuospatial
impairment. Critchley (1953), for example, defined constructional apraxia
as "an executive defect within a visuospatial domain." Benson and Bar-
ton (1970) later cautioned against the misapplication of the term and
proposed that constructional disability be used as the more general term
for visuospatial disorders, with constructional apraxia being reserved for
cases in which a disconnection exists between preserved perceptual and
motor skills. More recently, Benton (1982) proposed a classificatory sche-
ma based on tasks used to clinically assess different aspects of spatial
cognitive processing (e.g., visuospatial, visuoconstructive).
Extensive studies of visuospatial dysfunction did not emerge until as
late as the 1960s. Initial studies concluded that the right hemisphere
subserved nonverbal, visuospatial processing while the left hemisphere
was specialized for language processing. This dichotomy was bolstered by
studies demonstrating that right-hemisphere-damaged (RHD) patients
performed poorly on tests of perceptual closure (Benton & Van Allen,
1968; Newcombe & Russell, 1969; Warrington & James, 1967), visual lo-
calization (Hannay, Varney, & Benton, 1976; Warrington & Rabin, 1970),
and line perception (Benton, Hannay, & Varney, 1975), whereas patients
with left-hemisphere damage (LHD) performed at the same level as neu-
rologically intact normal control subjects. Other investigations, however,
challenged the simple verbal/nonverbal dichotomy for hemispheric spe-
cialization. These studies found that LHD patients were also impaired on
visuospatial tasks, though not as severely or as frequently as RHD pa-
tients, suggesting that both hemispheres have visuospatial capabilities,
with the right hemisphere commanding the superior role (De Renzi,
1982). This conclusion was consistent with studies suggesting that the
right hemisphere is involved in aspects of language functioning, includ-
ing, for example, comprehension of nonliterallanguage such as metaphor
(Gardner, Brownell, Wapner, & Michelow, 1983).
FRACTIONATION OF SPATIAL COGNITION 19
Until recently, few neuropsychological studies utilized constructs

developed in cognitive psychology to characterize visuospatial dysfunc-
tion following brain damage. Within the past few years, however, an
increasing number of researchers have begun to investigate whether focal
brain lesions disrupt specific aspects of visuospatial functioning in a pre-
dictable manner. The continued development of sophisticated theories of
normal cognitive processes has assisted neuropsychologists in more pre-
cisely delineating the neural substrate of cognitive operations. Concur-
rently, findings in neuropsychology have provided support for theories of
normal cognition (Cermak & Butters, 1972) and have provided evidence
for the independence of component processes by their "separability" in
the brain.
VISUAL ATTENTION
Visual attention, one aspect of visuospatial processing, has been the

subject of intensive study by Posner and his colleagues. Visual detection
experiments with normal subjects have demonstrated that attention can
be dissociated from eye movements (Posner, 1980). Furthermore, Posner
and his colleagues have provided evidence for different operations under-
lying attention. For example, detection of a target in the periphery is
facilitated if the target is preceded by a peripheral cue; however, if no
target occurs in the periphery following the peripheral cue and the subject
must then detect a target at the original fixation point, an inhibition of
detection occurs (Posner & Cohen, 1984). Because different aspects of
attention can be facilitated or inhibited, Posner and Cohen have posited
specific elementary operations in normal attention, such as moving to-
ward a target and disengaging from a target in order to move toward a new
target.
With this model as a guide, the attentional capacities of patients with
unilateral parietal lobe lesions have been investigated (Posner, Walker,
Friedrich, & Rafal, 1984, 1987). In earlier neuropsychological investiga-
tions, attention was regarded as a unitary function that was either spared
or impaired. Posner and his colleagues provided convincing evidence that
attention is disrupted in a highly selective manner following parietal lobe
lesions. The parietal lobe patients' difficulty in detecting targets was not
in moving toward initial target sites but in disengaging from one site in
order to detect another target. Furthermore, this deficit was most apparent
when the patients were cued to attend to a target on the same side of the
lesion (ipsilesion) and then were presented with a target on the opposite
side (contralesion). If these studies had found that all the proposed ele-
mentary operations of visual attention were equally impaired following
parietal lobe lesions, then one might have concluded that attention is
disrupted in a unitary manner, a result that would have been of little

theoretical import for cognitive psychologists. However, the findings in
the Posner et al. (1984,1987) studies give a neuropsychological reality to
the notion of separate operations in normal visual attention. In addition,
these findings enhance our understanding of neuroanatomical correlates
of attentional processes by more precisely describing the role of the par-
ietal lobe in the operations of visual attention.
PARTS AND WHOLES

Historically, visuospatial dysfunction has resisted fractionation into
meaningful spared and impaired components (Newcombe, 1985). A dis-
tinction between processing of parts and wholes, however, has been a
central focus of theories of normal perception. The possibility that the
brain has evolved to differentially process parts or details versus wholes
or configurations in different cerebral regions was initially proposed in
the 1940s (Patterson & Zangwill, 1944). Subsequent research with split-
brain patients following commissurotomy, unilateral brain-damaged pa-
tients, and neurologically intact subjects in laterality experiments has
produced convergent evidence indicating that the two cerebral hemi-
spheres represent a division of labor in processing parts and wholes
(Levy-Agristi & Sperry, 1968; Sergent & Bindra, 1982; Warrington, James,
& Kinsbourne, 1966). For example, Kaplan (1983) has reported that the
drawings of LHD patients often lack internal details, whereas the draw-
ings of RHD patients frequently have omitted or distorted outer configura-
tions. Similarly, on the Block Design subtest of the WAIS-R, Kaplan
(1983) reports that LHD patients have difficulty orienting individual
blocks, frequently making internal, single-block rotational errors but
maintaining the outer square configuration. RHD patients, in contrast,
often break the two by two or three by three square configuration of the
designs. These studies have concluded that both hemispheres are essen-
tial for successful performance on visuospatial tasks, with each hemi-
sphere contributing qualitatively different processes. In short, the left
hemisphere is superior for detailed, "analytic" processing, whereas the
right hemisphere is superior for configural, "holistic processing" (see
Bradshaw & Nettleton, 1981, for a review).
Despite the heuristic appeal of the analytic/holistic theory of hemi-
spheric specialization, this distinction has been criticized for both meth-
odological and theoretical reasons (see Brownell & Gardner, 1981; Delis,
Robertson, & Efron, 1986; Moscovitch, 1979). A major weakness of the
distinction is that "detailed and analytic" and "configural and holistic"
are imprecisely and vaguely defined. Researchers tend to infer that a
particular task involves more analytic or holistic processing without ex-
plicitly operationalizing these constructs. Furthermore, investigations

with both unilateral brain-damaged and split-brain patients typically use
stimuli (e.g., block designs, incomplete figures, facial stimuli; see cri-
tiques by Brownell & Gardner, 1981; De Renzi, 1982; Moscovitch, 1979) in
which the boundary between features perceived as "parts" and "wholes"
is unclear.
Recent investigations, however, have begun to provide a more precise
characterization of hemispheric asymmetry for visuospatial processing by
adopting paradigms from cognitive psychology. These studies suggest
that the two hemispheres are differentially specialized for processing of
different structural levels of visual hierarchical stimuli. A key feature of a
series of studies that we have conducted is the use of visual hierarchical
stimuli that have explicit demarcation between features perceived as
"parts" and "wholes." These hierarchical stimuli consist of a large (high-
er-level) letter or shape constructed from numerous smaller (lower-level)
letters and shapes. These two-leveled stimuli are methodologically supe-
rior to those traditionally used in visuospatial studies because (1) the two
levels of structure are explicitly demarcated and (2) the same type of form
(letter or shape) can exist at different levels of hierarchical structure,
enabling a tightly controlled test of differential response at the two levels.
Studies in which these stimuli were presented centrally to normal sub-
jects have found evidence for cognitive operations differentially sensitive
to higher- and lower-level analysis (Kinchla & Wolfe, 1979; Navon, 1977).
They have also found that it is relative rather than absolute size that
determines whether stimuli are analyzed as higher- or lower-level forms
(Navon, 1977). The use of these stimuli with brain-damaged patients is
particularly important in that the ability to organize visual features into
perceptual wholes is a critical aspect of perception.
STUDIES OF VISUOSPATIAL PROCESSING USING

HIERARCHICAL STIMULI
In one of our first studies (Delis et aI., 1986), we presented visual

hierarchical stimuli consisting of a large (global) form made up of numer-
ous smaller (local) forms to unilateral brain-damaged patients and normal
control subjects. Stimuli were presented individually for 2 seconds to the
central visual field, followed by a 15-second distractor task. Immediately
after the distractor task, a forced-choice recognition test was given by
presenting four alternative hierarchical stimuli (see Figure 1). The results
indicated that RHD patients were selectively impaired in recognizing
forms at the global structural level, whereas LHD patients were selectively
defective at recognizing forms at the local level (see Figure 2). These
findings are consonant with hemispheric asymmetries reported in two
o 0 0 0 Q Q Q Q
0 Q
0 Q
o 0 Q Q
0 Q
0 Q
o 0 o 0 Q
o 0 0 0 Q Q Q Q
0 Q
0 Q
o 0 Q Q
0 Q
0 Q
0 Q Q Q Q
r
r
r
c c c c c c r r r r r r
r
r
r
r
r
c C t r r r
r
FIGURE 1. Examples of (a) linguistic
r
hierarchical stimulus set and (b) non-
linguistic hierarchical stimulus set
b used in the experiment.
.,
~
E
~ .80
E
~ .60
.,
~
~ .40
c
...J
~ .20
~
0
~
E -.20
~
E -.40
~ FIGURE 2. Error analysis in recognizing
-.60 larger and smaller forms for the three
..!!!
-0 subject groups. The LHD patients made
E -.80
en
more errors in recognizing smaller forms
I!?
0
:::;: relative to larger forms, whereas the RHD
Left- Control Right-
Hemisphere Subjects Hemisphere patients made more errors in recognizing
Patients Patients larger forms relative to smaller forms.
studies with normal subjects in which visual hierarchical stimuli were

presented tachistoscopically to their visual half-fields (Martin, 1979; Ser-
gent, 1982). For example, Sergent (1982) presented hierarchical letter
stimuli to the visual half-fields of normal subjects and found a left visual-
field advantage (putatively, a right-hemisphere superiority) for identifica-
tion of the global letter, and a right visual-field advantage (putatively, a
left-hemisphere superiority) for identification of the local level.
In a subsequent investigation (Delis, Kiefner, & Fridlund, 1988), we
explored whether or not differential deficits in hierarchical processing
following unilateral brain damage occur on a constructional task. This
study was designed to tap impairments in both hierarchical and hemi-
spatial processing within the same task across a broad range of stimuli.
Deficits in hemispatial processing are common sequelae of brain damage,
and other investigations have demonstrated consistently that unilateral
brain damage impairs processing of visual stimuli presented in con-
tralateral hemispace (Benton, Levin, & Van Allen, 1974; De Renzi,
Faglioni, & Scotti, 1970; Heilman, Watson, & Valenstein, 1985; Samuels,
Butters, & Goodglass, 1970). Three types of stimuli were used at the global
and local levels: letters, shapes with established names, and shapes with-
out established names. The stimuli pair were presented simultaneously in
left and right hemispace to investigate the interaction between hier-
archical and hemispatial processing. Unilateral brain-damaged patients
and normal control subjects were first asked to draw the pairs of stimuli
from immediate memory, and then to copy them. The findings revealed a
pronounced dissociation: The RHD patients were selectively impaired in
recalling the global forms relative to the local forms, whereas the LHD
patients demonstrated the opposite pattern (see Figures 3 and 4). Al-
though the unilateral brain-damaged patients were most impaired when
drawing forms presented in the contralateral hemispace, the interaction
between hierarchical level (global vs. local) and side of hemispace (right
vs. left) was not significant, suggesting functional independence between
these two aspects of visuospatial processing. Stimulus category did not
differentiate the groups. This procedure thus revealed clear dissociations
in global-local and hemispatial components of visuospatial processing.
The results also suggest that for right-handers, the cerebral hemispheres
may be functionally specialized to mediate global or local processing on
both perceptual and constructive tasks.
The finding that the right hemisphere is specialized for global pro-
cessing and the left for local processing has also been confirmed following
commissurotomy. We had a rare opportunity (Delis, Kramer, & Kiefner,
1988) to administer hierarchical stimuli to a split-brain patient pre- and
postcommissurotomy. Prior to surgery, the patient was able to represent
both global and local forms in his drawings using either hand. After sur-
gery, he was able to draw only local forms with his right hand (controlled
J
J J t! t! t!
J t! t!
J t! t!
J t! t!
J J t! t!
J t! " " " t!
a
S b
FIGURE 3. Examples of (a) pair of hier-
archical stimuli, (b) drawing by an LHD
patient illustrating correct construction
of higher-level form presented in left
hemispace, and (c) drawing by an RHD
patient showing correct construction of
lower-level forms presented in right
c hemispace.
primarily by his ltlft hemisphere) and only global forms with his left hand
(controlled primarily by his right hemisphere) (see Figures 5 and 6). His
performance on a forced-choice recognition task was consistent with his
drawing results: He was more accurate in recognizing global forms when
pointing with his left hand and local forms when pointing with his right
hand. These results strongly suggest that components of visuospatial pro-
cessing can be disconnected in the same person following resection of the
corpus callosum.
Up to this point, all of the procedures involved a memory or construc-
tional component. We thus conducted a study to test whether the global-
local dissociation is revealed on a purely perceptual task without memory
or constructional demands (Robertson & Delis, 1986). In this experiment,
a task developed by Palmer (1980) was presented to unilateral brain-
damaged patients and normal control subjects. The subjects were asked to
indicate in which direction an equilateral triangle appeared to point when
it was either presented individually or aligned within other equilateral
triangles or circles. This task enables an analysis of the influence of a
more global reference frame (i.e., the aligned triangles) on the perception
of more local elements (i.e., a single triangle). The procedure is well
suited for brain-damaged populations because there is no time limit and it
places minimal demand on motor manipulation and memory skills. The
results of this experiment were consistent with our other findings: The
RHD patients were influenced significantly less by the global alignment
than the normal control subjects, whereas the LHD patients were influ-
Immediate Recall • Controls
7 o LHD Patients
o RHD Patients
~
6
j
0 ,,
,, /0
"I ,
-..>--
/
/' ,
/
CY
/
-- -- , "-
"-
3 "" ""
'0 '0
.."8 2
<II
>-
a c
..'"
U
:l
U Copy
«
U
10
o, ,
8
, /0
, /
" / ,/
/
0---------:8
7
/ , ------
/ / "0 0--
/
/
/
6 0
jL-____ ~ ________L __ _ _ __ L_ _ _ _ ~ _ _ _ _ _ _ _ __ L_ _ _ __ _
b d
Higher Lower Left Right
Level Level Hemispace Hemispace
Forms Forms Forms Forms
FIGURE 4. Drawing accuracy for (a and b) forms at higher and lower levels of hierarchical
stimuli, and (c and d) forms presented in left and right hemispace, for controls (e), LHD
patients (0), and RHD patients (0).
enced significantly more than the normal control subjects. These findings,
in concert with other data, suggest that differential processing of levels of
visual hierarchical structure has a biological basis in the brain.
FREQUENCY AND SEVERITY OF VISUOSPATIAL DEFICITS
Past investigations have been equivocal concerning the frequency

and severity of visuospatial dysfunction following unilateral brain dam-
age. As noted above, one group of studies reported that visuospatial im-
yyyyyyyyy
y
y
y
y
y
y
FIGURE 5. Examples of (a) target hierarchical stim-

ulus. (h) drawing done with patient's right hand
postcommissurotomy illustrating accurate representa-
tion of only the lower-level form. and (c) drawing
done with his left hand postcommissurotomy il-
lustrating accurate representation of only the higher-
level form
8000008
o 0
c--------)
rIGURE 6. Examples of (a) target hierarchical stimulus.
:h) drawing done with the patient's right hand postcom-
nissurotomy illustrating accurate representation of
mly the lower-level form. and (c) drawing done with
iJ.is left hand postcommissurotomy illustrating accurate
representation of only the higher-level form.
pairment occurs primarily after RHD (Arrigoni & De Renzi, 1964; Benton,
1967; Benton & Fogel, 1962; De Renzi and Faglioni, 1967; Hannay et al.,
1976; Piercy & Smyth, 1962), whereas others have concluded that damage
to either hemisphere results in equivalent visuospatial dysfunction (Ben-
ton, 1973; Benson & Barton, 1970; Black & Strub, 1976; Colombo, De
Renzi, & Faglioni, 1976). These contradictory findings have raised the
question of whether visuospatial functioning is subserved primarily by
the right hemisphere or both hemispheres (Benton, 1985; De Renzi, 1982).
Delis, Kiefner, and Fridlund (1988) investigated severity of drawing
impairment and found that RHD and LHD patients were equally impaired
when the results were collapsed across global and local level forms pre-
sented in right and left hemispace. However, when drawing performance
was analyzed in terms of hierarchical levels and sides of hemispace, the
results clearly indicated that both hemispheres contribute qualitatively
different and complementary processes to visuospatial functioning (see
also Kaplan, 1983; Levy, 1974; Nebes, 1973; Sergent, 1982). At least two
reasons may account for past reports of greater visuospatial dysfunction
in RHD patients: (1) Impaired global processing may be manifested more
saliently than deficient local processing, especially when studies employ
visual stimuli that have minimal or no internal details (e.g., the outline of
a geometric shape); and (2) RHD is associated with defective processing in
both sides of hemispace, whereas LHD is associated with impaired pro-
cessing primarily in contralateral hemispace (Heilman et al., 1985).
Further evidence indicating that both hemispheres mediate qualita-
tively different visuospatial processes comes from studies of facial per-
ception. Traditionally it was thought that the right hemisphere is superior
for facial processing (Benton & Van Allen, 1968). It has recently been
demonstrated that the right hemisphere is superior only when holistic
aspects of faces must be considered. When individual featural aspects of
faces are important to the task, the left hemisphere is dominant (Parkin &
Williamson, 1987; Patterson & Bradshaw, 1975; Sergent, 1982). Thus it
appears that facial processing may be either primarily a right- or a left-
hemisphere function, depending on the task demands.
INTRAHEMISPHERIC DISSOCIATIONS IN VISUOSPATIAL

PROCESSING
Up to this point, we have discussed dissociations in visuospatial

processing between the two hemispheres. Visuospatial dissociations,
however, have also been reported following lesions within the right hemi-
sphere. Newcombe and colleagues found that patients with focal damage
involving either a superior occipital-parietal or inferior occipital-tem-
poral pathway within the right hemisphere demonstrated different pro-

files of scores on clinical visuoperceptual and visuospatial tasks (New-
combe, Ratcliff, & Damasio, 1987; Newcombe & Russell, 1969).
Specifically, patients with superior occipital-parietal damage were im-
paired on some visuospatial tasks (e.g., maze learning) but not on others
(e.g., Mooneys visual closure task). Conversely, patients with inferior oc-
cipital-temporal damage demonstrated the opposite profile. An important
next step will be to characterize explicitly the component processes asso-
ciated with these performance differences on clinical tests. Newcombe's
findings are consistent with physiological data based on ablation studies
in monkeys (Ungerleider & Mishkin, 1982) suggesting two cortical visual
processing pathways. Ungerleider and Mishkin (1982) proposed that the
function of the inferior temporal pathway is visual discrimination and
thus is chiefly concerned with the shape of the visual stimulus. In con-
trast, the function of the superior parietal pathway is analysis of spatial
relations and the location of objects in space.
TOWARD AN INTEGRATION OF INTRA- AND

INTERHEMISPHERIC DISSOCIATIONS IN VISUOSP ATIAL
PROCESSING
Robertson, Lamb, and Knight (in press) have recently conducted an

important study aimed at integrating findings in the neuropsychological
and cognitive literatures on visuospatial processing. As noted earlier, it
has been well established that unilateral brain damage results in a dis-
sociation between global and local processing, with RHD disrupting
global processing and LHD disrupting local processing. Furthermore, the
posterior parietal region plays a major role in attention (Posner et 01.,
1984) and spatial relations, while temporal lobe damage disrupts one's
ability to identify objects (Newcombe et 01., 1987; Newcombe & Russell,
1969). The relative contribution of attentional and/or perceptual mecha-
nisms to the global-local dissociation found following unilateral right-
and left-hemisphere damage has heretofore been unexplored.
Robertson et 01. (in press) employed a procedure developed by
Kinchla and his colleagues (Kinchla, Solis-Macias, & Hoffman, 1983) to
test whether the global-local dissociation in unilateral brain-damaged
patients could be attributed to controlled attentional processes, passive
perceptual processes, or both. The type of attention examined in this
study was defined as "controlled distribution of attention over the visual
field," and perceptual processing was defined as "direct automatic en-
coding of stimulus features." In this investigation, unilateral brain-
damaged patients and normal control subjects were presented with hier-
archical stimuli and asked to identify one of two targets presented at
either the global or local levels. Stimuli were presented in three experi-
mental conditions: a No Bias condition in which there was an equal
probability of the target's appearing at either the global or the local level, a
Local Bias condition in which there was a 75% probability of the target's
appearing at the local level, and a Global Bias condition in which there
was a 75% probability of the target's appearing at the global level. Predict-
ably, subjects who are able to distribute attention between the global and
local levels should show improved performance in the biasing conditions.
The unilateral brain-damaged patients included a group with damage
restricted to the right temporal-parietal region (RTP) and two groups of
left-hemisphere-damaged patients with lesions in either the superior tem-
poral gyrus (LSTG) or the rostral inferior parietal lobule (LIPL) as deter-
mined by CT scan.
The authors report three important findings. First, attentional pro-
cesses in the LIPL patients were disrupted; that is, their performance did
not vary as a function of biasing condition. In contrast, the LSTG group
did not demonstrate an attentional disturbance. This finding is consonant
with studies indicating that the parietal region of higher primates and
humans plays a special role in attention (Bushnell, Goldberg, & Robinson,
1981; Mountcastle, 1978; Posner et a1., 1984). More specifically, the pre-
sent study suggests that the inferior parietal region is critical for the con-
trolled distribution of attention over the visual field and between hier-
archical levels. Second, overall response times were significantly lower
for both global and local forms in the LTPG group than for the LIPL group
and normal control group, suggesting a deficit in visual discrimination
associated with temporal damage. Such an interpretation is in accord
with the functional roles attributed to the two visual pathways. Third,
there was a global-local dissociation between LSTG and RTP groups, but
not between LIPL and RTP groups. The LSTG group showed a strong
baseline advantage in identifying targets at the global level, whereas the
RTP group demonstrated the opposite pattern. This finding suggests that
the glocal-Iocal dissociation previously reported with unilateral brain-
damaged patients is related to perceptual, rather than attentional, mecha-
nisms. It is also now clear that the hemispheric asymmetry observed in
hierarchical processing occurs in the first few hundred milliseconds of
analyzing a stimulus.
FRACTIONATION OF SPATIAL COGNITION IN DIFFUSE

BRAIN PATHOLOGY
Dissociations in spatial cognition are much rarer and infrequently

reported in cases of diffuse, relative to focal, brain damage. Recently,
however, we have encountered a striking exception to this pattern. We
conducted a study of visuospatial processing in two groups of mentally

retarded subjects between the ages of 9 and 18 years; children with
Williams syndrome and with Down syndrome (Bihrle, Bellugi, Delis, &
Marks, in press). Although these two groups perform comparably on
global measures of intelligence (mean IQ of 57) and on clinical visuo-
constructive tests (e.g., block design subtest of the WISC-R), the qualita-
tive features of their constructions are quite distinct. The Williams syn-
drome subjects' drawings typically show excessive attention to the parts
of an object with no integration of the parts to form a coherent whole. In
contrast, the Down syndrome subjects' drawings represent the outer
configuration of an object, but internal details are frequently lacking.
These characterizations have been systematically substantiated in a study
using hierarchical stimuli. In both memory and copy conditions, the
Williams syndrome subjects, like unilateral RHD patients, were signifi-
cantly more impaired in global relative to local analysis, producing local
forms with no attempt at representing the global figure (see Figure 7). In
contrast, the Down syndrome subjects, like unilateral LHD patients, fre-
quently produced only the global figure using a solid line and omitted the
local figures completely. Thus, the interaction between subject group and
hierarchical level (global-local) was highly significant, although the two
groups did not differ in overall performance collapsed across global and
local conditions.
These robust findings are important for a number of reasons. First,
they suggest that sharp dissociations in visuospatial processing can occur
in cases of diffuse brain damage, just as they occur following unilateral
focal lesions. Neuroradiologic studies do not suggest that Williams syn-
drome or Down syndrome results in unilateral focal lesions Uernigan,
Tallal, & Bellugi, 1988). Furthermore, although it is well known that clus-
ters of spared and impaired components of cognition occur following
damage to particular regions of the brain, such patterns are infrequently
reported in cases of global deterioration and may be theoretically impor-
tant. Unilateral RHD typically results in relatively preserved language
functioning and a proclivity for local processing. Perhaps it is not coinci-
dental that Williams syndrome, also characterized by selectively pre-
served language ability relative to other cognitive functioning (Bellugi,
Marks, Bihrle, & Sabo, in press; Bennett, LaVeck, & Sells, 1978; Udwin,
Yule & Martin, 1986; von Armin & Engel, 1964), shows a local processing
bias. In contrast, unilateral LHD is associated with language dysfunction
and relatively preserved global processing. Down syndrome, as well, is
marked by language ability disproportionately impaired relative to other
cognitive functions (Fowler, in press) and a propensity for global process-
ing. Further exploration of these neuropsychological profiles may aid us
in understanding the interrelationship between spatial and language pro-
cessing and their representation in the brain.
"'l
a NORMAL CONTROL NORMAL CONTROL
b ~
C"l
."i 'I Y't!,. yyyy Y yn g

I
V .,. Y '( ~ I Z
Y y ?
Y ~ y >->-3
'I ~ y " -------- {-------
YyyyY Y - --------- \
'(~ yvv Y '( y , 5
rY,(,r "- , z
NC Age 11 NC Age 16 NC Age 16
0
NC Age 11 NC Age 16 NC ~ge 16 "'l
CIl
"C
>-
~
>-r
DOWN SYNDROME DOWN SYNDROME (j
0
c:l
~
5
0 Eo: H f=l z
Age 11 OS Age 11
os
D OS Age 16
DOS Age 18 OS Age 16 OS Age 18
WILLIAMS SYNDROME WILLIAMS SYNDROME

'(
y'
'I ----
I ~yyyy\"rtr- I I rrY,YYrr -----
- --- , .... c...: _
rVrif'( - - II \.t:G;";;-- - -
~---
> -'----- WS Age 16 WS Age 16

WS Age 11 WS Age 18 I WS Age 18 WS Age 11
FIGURE 7. Examples of drawings of (a) letter and (b) shape hierarchical stimuli by normal control, Down syndrome, and Williams
syndrome subjects in the memory condition. w
....
Results from a pilot study we recently conducted suggest that asym-

metric visuospatial profiles may also occur in some patients with Alz-
heimer disease. We administered the same global-local task used in the
study cited above to six Alzheimer patients who showed lateralized cog-
nitive deficits on other clinical tests. Five patients displayed normal per-
formance on the Boston Naming Test and impaired performance on the
WAIS-R Block Design subtest, whereas the other patient showed the op-
posite pattern. The Alzheimer patients with normal naming recalled pri-
marily local forms, a performance similar to unilateral right CVA patients
(Delis et al., 1988). In contrast, the patient with normal block construc-
tions recalled only global forms, a performance similar to left CVA pa-
tients. These results are consistent with the finding that Alzheimer pa-
tients show significantly more lateral asymmetry of brain glucose metabo-
lism than age-matched normal subjects (Haxby, Duara, Grady, Cutler, &
Rapoport, 1985). A third group of Alzheimer patients ~ho were impaired
in both naming and block construction (but matched with the other pa-
tients in overall dementia severity; Mattis, 1976) showed less discrepancy
in their recall of global and local forms with no global-local dissociation.
Our results suggest that even in some cases of diffuse pathology, distinct
visuospatial profiles may emerge.
CONCLUSION
Although it is difficult to characterize different aspects of visuo-

spatial impairments, the studies cited in this chapter have been able to
demonstrate selective deficits in visual attention, perception, and visuo-
construction. Recent research in neuropsychology, guided by work in
cognitive psychology, has suggested that spatial cognition, like language,
fractionates in precise ways following focal and diffuse brain damage.
ACKNOWLEDGMENT
This research was supported by the Medical Research Service of the

Veterans Administration.
REFERENCES
Arrigoni, G., & De Renzi, E. (1964). Constructional apraxia and hemispheric locus of lesion.
Cortex, 1, 180-197.
Bellugi, U., Marks, S., Bihrle, A., & Sabo, H. (in press). Dissociation between language and
cognitive functions in Williams syndrome. In K. Mogford & D. Bishop (Eds.), Language
development in exceptional circumstances. London: Churchill Livingstone.
Bennett, F. c., LaVeck, B., & Sells, C. J. (1978). The Williams elfin facies syndrome: The
psychological profile as an aid in syndrome identification. Pediatrics, 61, 303-305.
Benson, D. F., & Barton, M. 1. (1970). Disturbances in constructional ability. Cortex, 6, 19-46.
Benton, A. 1. (1967). Constructional apraxia and the minor hemisphere. Confinia Neu-
rologica, 19, 1-16.
Benton, A. 1. (1973). Visuoconstructive disability in patients with cerebral disease: Its rela-
tionship to side of lesion and aphasic disorder. Documenta Opthal. 34, 67-706.
Benton, A. 1. (1982). Spatial thinking in neurological patients: Historical aspects. In M.
Potegal (Ed.), Spatial abilities: Development and physiological foundations. New York:
Academic Press.
Benton, A. L. (1985). Visuoperceptual, visuospatial, and visuoconstructive disorders. In K.
M. Heilman & E. Valenstein (Eds.), Clinical neuropsychology. New York: Oxford.
Benton, A. 1., & Fogel, M. L. (1962). Three-dimensional constructional praxis. Archives of
Neurology, 7, 347-354.
Benton, A. L., Hannay, J., & Varney, N. R. (1975). Visual perception of line direc;tion in
patients with unilateral brain disease. Neurology, 25, 907-910.
Benton, A. 1., Levin, H. S., & Van Allen, M. W. (1974). Geographic orientation in patients
with unilateral cerebral disease. Neuropsychologia, 12, 183-191.
Benton, A. 1., & Van Allen, M. W. (1968). Impairment in facial recognition in patients with
cerebral disease. Cortex, 4, 344-358.
Bihrle, A. M., Bellugi, U., Delis, D. C., & Marks, S. (in press). Seeing either the forest or the
trees: Dissociations in visuospatial dysfunction. Brain and cognition.
Black, F. W., & Strub, R. 1. (1976). Constructional apraxia in patients with discrete missile
wounds of the brain. Cortex, 12, 212-220.
Bradshaw, J. 1., & Nettleton, N. C. (1981). The nature of hemispheric specialization in man.
Behavioral and Brain Sciences, 4, 51-91.
Brownell, H. H., & Gardner, H. (1981). Hemispheric specialization: Definitions, not incanta-
tions. Behavioral and Brain Sciences, 4, 64-65.
Bushnell, M. c., Goldberg, M. E., & Robinson, D. 1. (1981). Behavioral enhancement of visual
responses in monkey cerebral cortex. L Modulation in posterior parietal cortex related to
selected visual attention. Journal of Neurophysiology, 46, 755-772.
Cermak,1. S., & Butters, N. (1972). The role of interference and encoding in the short-term
memory deficits of Korsakoff patients. Neuropsychologia, 10, 89-95.
Chomsky, N. A. (1965). Aspects of the theory of syntax. Cambridge, MA: M.LT. Press.
Colombo, A., De Renzi, E., & Faglioni, P. (1976). The occurrence of visual neglect in patients
with unilateral cerebral disease. Cortex, 12, 221-231.
Critcheley, M. (1953). The parietal lobes. London: Edward Arnold.
Delis, D. C., Kiefner, M., & Fridlund, A. J. (1988). Visuospatial dysfunction following uni-
lateral brain damage: Dissociations in hierarchical and hemispatial analysis. Journal of
Clinical and Experimental Neuropsychology, 10, 421-431.
Delis, D. C., Kramer, J. H., & Kiefner, M. G. (1988). Visuospatial functioning before and after
commissurotomy: Disconnection in hierarchical processing. Archives of Neurology, 45,
123-130.
Delis, D. C., Robertson, 1. C., & Efron, R. (1986). Hemispheric specialization of memory for
visual hierarchical stimuli. Neuropsychologia, 24, 205-214.
De Renzi, E. (1982). Disorders of spatial exploration and cognition. New York: Wiley.
De Renzi, E., & Faglioni, P. (1967). The relationship between visuospatial impairment and
constructional apraxia. Cortex, 3, 327-342.
De Renzi, E., Faglioni, P., & Scotti, G. (1970). Hemispheric contribution to exploration of
space through the visual tactile modality. Cortex, 6, 191-203.
Fillmore, C. J. (1971). Types of lexical information, In D. D. Steinberg & 1. A. Jakobouits
(Eds.), Semantics: An interdisciplinary reader in philosophy, linguistics, and psychol-
ogy. Cambridge: Cambridge University Press.
Fowler, A. E. (in press). The development of language structure in children with Down
syndrome. In D. Cicchetti & M. Beeghly (Eds.). Down syndrome: The developmental
perspective. New York: Cambridge University Press.
Gardner, H., Brownell, H. H., Wapner, W., & Michelow, D. (1983). Missing the point: The
role of the right hemisphere in the processing of complex linguistic materials. In E.
Perecman (Ed.). Cognitive processes in the right hemisphere. (pp. 169-191). New York:
Academic Press.
Geschwind, N. (1970). The organization of language and the brain. Science, 170,940-944.
Goodglass, H., & Kaplan, E. (1972). The assessment of aphasia and related disorders.
Philadelphia: Lea & Febiger.
Hannay, H. J., Varney, N. R., & Benton, A. L. (1976). Visual localization in patients with
unilateral brain disease. Journal of Neurology, Neurosurgery and Psychiatry, 39, 307-
313.
Haxby, J. V., Duara, R., Grady, C. L., Cutler, N. R., & Rapoport, S. 1. (1985). Relations between
neuropsychological and cerebral metabolic asymmetries in early Alzheimer's disease.
Journal of Cerebral Blood Flow and Metabolism, 5, 193-200.
Heilman, K. M., Watson, R. T., & Valenstein, E. (1985). Neglect and related disorders. In K.
M. Heilman & E. Valenstein (Eds.). Clinical neuropsychology. New York: Oxford Uni-
versity Press.
Jackson, J. H. (1876). Case of large cerebral tumour without optic neuritis and with left
hemiplegia and imperception. Royal Opthalmological Hospital Reports, B, 434-444.
Jernigan, T., Tallal, P., & Bellugi, U. (1988). Cerebral morphology of magnetic resonance
imaging (MRI) in developmental cognitive disorders. Paper presented at the Interna-
tional Neuropsychological Society annual meeting, New Orleans, LA.
Kaplan, E. (1983). Process and achievement revisited. In S. Wapner & B. Kaplan (Eds.),
Toward a holistic developmental psychology. Hillsdale, NJ: Erlbaum.
Kinchla, R. A., Solis-Macias, V., & Hoffman, J. (1983). Attending to different levels of struc-
ture in a visual image. Perception and Psychophysics, 33, 1-10.
Kinchla, R. A., & Wolfe, J. M. (1979). The order of visual processing: Top-down, bottom-up,
or middle-out. Perception and Psychophysics, 25, 225-231.
Kleist, K. (1934). Gehirnpathologie. Leipzig: Barth.
Levy, J. (1974). Cerebral asymmetries as manifested in split-brain man. In M. Kinsbourne &
W. 1. Smith (Eds.), Hemispheric disconnection and cerebral function. Springfield, IL:
Thomas.
Levy-Agristi, J., & Sperry, R. W. (1968). Differential perceptual capacities in major and minor
hemispheres. Proceedings of the National Academy of Sciences, 61, 1151.
Luria, A. R. (1966). Higher cortical functions in man. New York: Basic Books.
Martin, M. (1979). Local and global processing: The role of sparsity. Memory and Cognition,
6,476-484.
Mattis, S. (1976). Mental status examination for organic mental syndrome in the elderly
patient. In 1. Bellack & T. B. Karusu (Eds.), Geriatric psychiatry. New York: Grune &
Stratton.
Moscovitch, M. (1979). Information processing and the cerebral hemispheres. In M. S.
Gazzaniga (Ed.). The handbook of behavioral neurobiology. New York: Plenum Press.
Mountcastle, V. B. (1978). Brain mechanisms for directed attention. Journal of the Royal
Society of Medicine, 71, 14-28.
Navon, D. (1977). Forest before trees: The precedence of global features in visual perception.
Cognitive Psychology, 9, 353-383.
Nebes, R. D. (1973). Perception of spatial relationships by the right and left hemispheres of
commisurotomized man. Neuropsychologia, 11, 285-289.
Newcombe, F. (1985). Neuropsychology qua interface. Journal of Clinical and Experimental
Neuropsychology, 7, 663-681.
Newcombe, F., Ratcliff, G., & Damasio, H. (1987). Dissociable impairments of visual and
spatial processing following right posterior cerebral lesions: Clinical, neuropsycholo-
gical, and anatomic evidence. Neuropsychologia, 25, 149-161.
Newcombe, F., & Russell, W. R (1969). Dissociated visual perceptual and spatial deficits in
focal lesions of the right hemisphere. Journal of Neurology, Neurosurgery and Psychia-
try, 32, 73-81.
Palmer, S. E. (1980). What makes triangles point: Local and global effects in configuration of
ambiguous triangles. Cognitive Psychology, 12, 285-305.
Parkin, A. J., & Williamson, P. (1987). Cerebral lateralisation at different stages of facial
processing. Cortex, 23, 99-110.
Patterson, K., & Bradshaw, J. 1. (1975J. Differential hemispheric mediation of nonverbal
stimuli. Journal of Psychology: Human Perception and Performance, 1,246-252.
Patterson, A., & Zangwill, O. 1. (1944). Disorders of visual space perception associated with
lesions of the right cerebral hemisphere. Brain, 67, 331-358.
Piercy, M., & Smith, V. O. G. (1962). Right hemisphere dominance for certain nonverbal
intellectual skills. Brain, 85, 775-790.
Posner, M. I. (1980). Orienting of attention. Quarterly Journal of Experimental Psychology,
32,3-25.
Posner, M.I., & Cohen, Y. (1984). Components of visual orienting. In H. Bouma & D. Bowhuis
(Eds.), Attention and performance. Hillsdale, NJ: Erlbaum.
Posner, M. I., Walker, J. A., Friedrich, F. J., & Rafa!' RD. (1984J. Effects of parietal injury on
covert orienting of visual attention. Journal of Neuroscience, 4, 1863-1874.
Posner, M. I., Walker, J. A., Friedrich, F. J., & Rafa!, R D. (1987J. How do the parietal lobes
direct covert attention? Neuropsychologia, 25, 135-147.
Robertson, L. C., & Delis, D. C. (1986J. "Part-whole" processing in unilateral brain damaged
patients: Dysfunction of hierarchical organization. Neuropsychologia, 24, 363-370.
Robertson, 1. C., Lamb, M. R, & Knight, R T. (in press). Effects of lesions of temporal-
parietal junction on perceptual and attentional processing in humans. Journal of Neu-
roscience,
Samuels, I., Butters, N., & Goodglass, H. (1970). Visual memory deficits following cortical
and limbic lesions: Effect of field of presentation. Neuropsychologia, 6, 447-452.
Sergent, J. (1982). The cerebral balance of power; Confrontation or cooperation? Journal of
Experimental Psychology: Human Perception and Performance, 8, 253-272.
Sergent, J., & Bindra, D. (1982). Differential hemispheric processing of faces: Methodological
considerations and reinterpretation. Psychological Bulletin, 89, 541-554.
Udwin, 0., Yule, W., & Martin, N. D. T. (1986). Age at diagnosis and abilities in idiopathic
hypercalcemia. Archives of Disease in Childhood, 61, 1164-1167.
Ungerleider, L. G., & Mishkin, M. (1982). Two cortical visual pathways. In D. J. Ingle, M. A.
Goodale, & R J. W. Mansfield (Eds.), Analysis of visual behavior, (pp. 549-585).
Cambridge, MA: M.I.T. Press.
Von Armin, G., & Enge!, P. (1964). Mental retardation related to hypercalcemia. Develop-
mental Medicine and Child Neurology, 6, 366-377.
Warrington, E. K., & James, M. (1967). An experimental investigation of facial recognition in
patients with unilateral lesions. Cortex, 3, 317-326.
Warrington, E. K., James, M., & Kinsbourne, M. (1966). Drawing disability in relation to
laterality of cerebral lesion. Brain, 89, 53-82.
Warrington, E. K., & Rabin, P. (1970). Perceptual matching in patients with cerebral lesions.
Neuropsychologia, 8, 475-487.
3
Anatomical and Physiological

Correlations and Neuroimaging
Techniques in Language Disorders
ANDREW KERTESZ
Localization of brain lesions has undergone a revolutionary development

in the last 10 years, yet some workers in cognitive science ignore the
wealth of new information that has direct bearing on brain function. The
most important achievement is the appearance of in vivo techniques of
localization, which allows the collection of localizing information, at the
same time as the neuropsychological examination of the brain-damaged
patient, in a prospective, planned detail. For nearly 100 years since
Broca's epoch-making clinicopathological correlation, neurological scien-
tists relied on the rarely available postmortem examination of patients
who were assessed during life, often long before they came to autopsy. In
this chapter, I intend to detail some of the advantages and limitations of
various localizing techniques and summarize the findings of the relatively
new science of in vivo localization.
AUTOPSY LOCALIZATION
This oldest of techniques has been further perfected by the discovery

of fixative and staining techniques, as well as optical enlarging and pho-
tographic instruments. Initially, alcohol fixation allowed the dissection of
fiber tracts (Gall & Spurtzheim, 1810; Reil, 1809). Formalin fixation, and
myelin and neuronal staining, around the turn of this century, resulted in
ANDREW KERTESZ· University of Western Ontario, St. Joseph's Health Centre, London,
Ontario N6A 4V2, Canada.
37
38 ANDREW KERTESZ
the creation of cortical maps based on cellular differentiation (Brodmann,

1909; Economo & Koskinas, 1925). The study of human cytoarchitectonics
and myeloarchitectonics diminished considerably after the 1920s in favor
of experimental neuroanatomy in animals bolstered by autoradiography
and axonal tracers, such a horseradish peroxidase (HRP). Recent advances
in pigment cytoarchitectonics and the application of silver stains to ante-
rograde and retrograde degeneration of tracts in human lesions opened
new opportunities to study human cerebral connectivity and cortical or-
ganization (Galaburda & Mesulam, 1983; Sanides, 1962).
The accuracy of postmortem localization makes the method the stan-
dard with which others are compared. Autopsy examination of the human
brain allows topographical localization of gyral and sulcal structures, and
fiber tracts and the accurate determination of the extent of lesions. In
addition to gross photography, the myelin and cellular stains allow micro-
scopic analysis of neuronal structures and lesions. Economo and Kos-
kinas (1925) standardized the nomenclature of the sulcal patterns and
described the common variations in conjunction with cytoarchitectonic
maps. Very little information has been accumulated since then concern-
ing the extent of variations in surface geography. Recently, integration of
anatomical landmarks with the vascular territories of the brain has been
achieved (Talairach & Szikla, 1967; von Keyserlingk, De Bleser, & Poeck,
1983).
The description of the etiology of lesions and the surrounding tissue
changes allows conclusions to be drawn about the possible effects of these
changes on the brain at the time of clinical observation.
The major disadvantage of the autopsy method is that the brain usu-
ally becomes available only long after the patient has been examined
clinically. Only occasionally is it possible to have the postmortem exam-
ination of the brain shortly after detailed, clinical examination was per-
formed. Patients who die after a cerebral event are usually so severely
affected that they could not be examined in the acute premortem illness in
any detail. If they die after a downhill course, such as in the case of a brain
tumor, by the time they come to autopsy the brain is severely distorted
and will not reflect the clinical state that was previously studied.
Although many circumscribed lesions have come to autopsy, the cor-
responding clinical examinations have often been scanty and incomplete.
Unfortunately, the reverse is also true. When the clinical examination has
been conducted in great length, the patient often recovers; when death
eventually occurs for some other reason, permission to carry out a
postmortem examination is not granted or the patient dies in another
location and the opportunity to establish correlation is lost.
The occasional situation where detailed examination is followed by
an autopsy report fresh after the clinical observations remains a very
important method of localization, even though it will provide only a
ANATOMICAL AND PHYSIOLOGICAL CORRELATIONS 39
random and rare opportunity for single cases and precludes the collection
of groups for generalizations to be made. Attempts to collect a large
number of postmortem examinations correlated with clinical observa-
tions have been made in the early 20th century, but they are plagued with
the heterogenity of the material and the lack of reliable detail in the
neurological and psychological examinations. Especially prominent are
the works of Moutier (1908), who wrote a monograph on Broca's aphasia
and associated syndromes. He described a systematic aphasia examina-
tion as an appendix to his book. He borrowed many cases, especially from
Bernheim (1900), and this method of collecting cases from the literature to
augment a few new descriptions became popular and is widely used even
today. It is not unusual to trace a case description through several series
(sometimes losing quite a bit in translation). Marie and Moutier (his
pupil) entered a vigorous debate concerning the localization of lesions in
Broca's aphasia and singled out the "quadrilateral space" central to the
insula as crucial for language, opposing Dejerine and his anatomist wife,
Klumpke, who favored the frontal operculum, also on the basis of autopsy
evidence. The most monumental collection of autopsy material is that of
Henschen (1922), who in seven large volumes collected 1,337 cases with
descriptions and illustrated more than 300 of them, often including mye-
lin-stained sections and photographs of specimens. The turn of the cen-
tury autopsy studies not only described the anatomy of lesions but drew
elaborate diagrams (resembling at times the more recent processing mod-
els) to illustrate their particular theory of language production as inferred
from the lesion evidence. This activity was dampened by the criticism of
Henry Head, who derided the "diagram makers." Von Monakow's (1914)
influential work pointed out the problems of distant effects of tumors and
the effects of depression of surrounding or functionally connected struc-
tures with acute stroke diaschisis. Nielson (1946) and Kleist (1962) wrote
more recent monographs on postmortem examinations and clinical patho-
logical correlation with language function and cognition. They espoused
rather extreme views of equating specific anatomy with function. Since
then, mostly single case reports have appeared with a variable quality of
clinical and psychological detail.
Improvements in autopsy studies may come about by the systematic
examination of stroke populations with the more sophisticated cytoarchi-
tectonic and connectivity techniques described above. Stroke registries
established at several centers allow more careful follow-up of patients.
The topographical analysis of the brain regions has been limited largely to
sulcal and gyral patterns, but cytoarchitectonic methods may correspond
more to functional differentiation. Our knowledge of functional dif-
ferences between children and adults, left and right hemispheres, sexes,
handedness, and cultural influences on cerebral organization may further
improve the sophistication of interpretations.
40 ANDREW KERTESZ
LOCALIZATION THROUGH HEAD INJURIES
The first group studies of live patients with neuropsychological defi-

cits were carried out using localization information provided by the radi-
ographs or surgical explorations of head-injured patients. Most of these
head injury studies relied on a postwar population. Marie and Foix
(1917), in France, used the overlap method of estimating the extent of
injuries in various types of aphasia, distinguishing five syndromes with
distinct localizations, contradicting somewhat Marie's earlier dogma of
the unitary nature of the aphasic syndrome. Even Henry Head (1926)
constructed an elaborate wire grid to localize the lesions from radiographs
based on anatomical specimens. After World War II, Schiller (1947), in
Germany, provided anatomical overlaps in a well-studied population
based on skull X rays. Further large series were carried out by Luria
(1970), in Russia, and Russell and Espir (1961), in England.
The major disadvantage of this method is that penetrating head inju-
ries often cause damage distant from the entry wound, and radiographs of
the skull defects are too inaccurate to draw reliable conclusions as to the
exact location and depth of the damage. Nonpenetrating head traumas,
even though they are more common in peacetime, are not usable because
they cause diffuse, shearing injuries and widespread disconnection in the
brain, as well as severe hemorrhage and destruction, with edematous
brain tissue that is, at times, infarcted irreversibly. Unfortunately, diffuse
head injury is often added to penetrating trauma, further detracting from
the information about localization. An advantage of the head injury popu-
lation is their relative youth and freedom from cerebral atrophy or repeat-
ed vascular insults.
SURGICAL LOCALIZATION
Neurosurgical resection of tumors and surgery for epilepsy, and at

times operation on penetrating head injuries, provide a variable measure
of localization. Tumor resection is not the most suitable material because
slowly growing tumors compress the brain in an insidious fashion and
often there is a great deal of compensation for the slowly incurred deficit.
In fact, this may result in false negative conclusions concerning the role of
certain areas in the brain. Malignant tumors infiltrate the brain tissue
beyond resection, and it is often difficult to determine how far they actu-
ally exert their influence. The distant effects of edema, vascular insuffi-
ciency, and displacement of the tissues are difficult to quantitate on sur-
gical explorations. Epilepsy surgery is often performed in scarred brains
and the removed tissue may not be functional at all. The extent of the
excision is difficult to judge because of the limitation of the exposure and
the variability of the surface anatomy.
Burkhardt (1891) removed Broca's area and posterior Tl and T2 with-

out permanent aphasia. Dandy (1931), after three cases, concluded that
Broca's area is where Broca described it, except that it may be a little more
posterior. He also said that visual speech (reading) is distinct, and that it
is impaired by occipital lobectomy including the angular and su-
pramarginal excisions. Mettler (1949) reported that bilateral removal of
Broca's area did not produce aphasia, but the actual areas were not identi-
fied by stimulation or autopsy, and detailed language examination was
lacking. Penfield and Roberts (1959) described language impairment in
581 left-hemisphere operations, and limited removal was compatible with
recovery even in Broca's area. Persistent defects were associated with
larger lesions. In a study of 214 surgically verified cases of mostly tumor
and trauma, Hecaen and Angelergues (1964) found that disturbances of
fluency, articulation, and naming were related to temporal lesions.
Hemispherectomies have been used to eradicate malignant tumors in
adults (Smith, 1966) and for the excision of severely scarred epileptogenic
tissue in infants and children. Some of these were extensively studied.
Left hemispherectomies in adults produced global aphasia, but they re-
tained emotional and automatic utterances, articulated even at a sentence
level, and some comprehension of single nouns. The limitations on study-
ing adult hemispherectomies are the rarity of the operation (most sur-
geons do not consider it worthwhile) and the usually relentless spread of
the neoplasm into the other hemisphere. Infantile hemispherectomies
have been well studied, indicating almost complete recovery of function
on either side. However, sophisticated language measures have indicated
less than normal development in verbal IQ and in grammatical compe-
tence in children who had left hemispherectomy in their infancy.
Sections of the corpus callosum performed for intractable epilepsy
provides an opportunity to study the function of each hemisphere (Sper-
ry, Gazzaniga, & Bogen, 1969). In this respect, the method is not a surgical
resection but a disconnection of two functional areas. The results have
illuminated many right-hemisphere functions, including a certain capaci-
ty for comprehending language. (A more detailed review can be found in
Gazzaniga, 1970.) Some of the operations have been less than complete, as
recent MRI studies have indicated, and this may have led to some false
conclusions concerning function.
CORTICAL STIMULATION
Cortical stimulation developed for epilepsy surgery is a major contri-

bution to cerebral localization. Initially it was used by physiologists to
map the function of the cerebral cortex in animals, and subsequently
neurosurgeons, such as Otto Foerster in the 1920s, have tried to identify
cortical function with stimulation in preparation for their excisions. Sen-
42 ANDREW KERTESZ
sory, motor, and language areas can be spared using this technique. Pen-
field and Roberts (1959) mapped the language areas and Van Buren,
Fedio, and Frederick (1978), Ojemann and Whitaker (1978), and Ojemann
and Mateer (1979) have added further information concerning cortical
localization.
The advantage of the method is that it can be applied in the awake,
cooperative patient who can perform selected functions during stimula-
tion. In this respect, it resembles other so-called functional methods of
localization, such as cerebral blood flow (CBF) and positron emission
tomography (PET). However, stimulation interferes with functions, inter-
rupts or alters them, and, less frequently, elicits positive phenomena. In
this respect, it is more like the lesion method, but stimulation involves a
much smaller area, allowing a greater resolution for mapping. Several
functions can be examined by repeated stimulation of the same area.
Limitations of the technique include the logistics of a lengthy surgical
procedure, difficulty reproducing the stimulation on the same spot, and
the brevity of tasks that can be employed during the short period of stim-
ulation (maximum about 12 seconds). The subjects usually have epilepsy
or tumors with reorganized brains, and the conclusions may not always be
generalizable.
The results of cortical stimulation reveal a striking discreteness of
function that is different from the conclusions drawn from lesion studies.
Naming, for example, may be affected over one gyrus, but not a few milli-
meters next to it. On the other hand, naming errors could be elicited from
a wide variety of cortical areas. The most consistent area to produce
speech arrest was in Broca's area, the posterior inferior frontal cortex, in
front of the face motor cortex (Ojemann & Whitaker, 1978). The posterior
superior temporal cortex seemed to be the next most important area in-
volved in language. Cortical mapping in bilingual patients showed sites
where one language was impaired but not the other. Naming in the less
competent and more recently acquired language was disrupted from a
wider area of stimulation. Some dissociations were also observed between
oral and sign language. When multiple language functions are tested over
the same site, only one may be involved. Grammar, semantics, and artic-
ulation seem to be dissociated in a highly variable fashion across patients.
Grammatical errors often occurred in the frontal lobe without articulatory
deticit. Semantic errors were from more widely distributed sites of stim-
ulation. Memory-related sites surround the perisylvian (usually parietal)
cortex at some distance from it. Individual variability may be related to
anatomical and educational differences and IQ (Mateer, 1983).
The localization of nonverbal functions in the nondominant hemi-
sphere was also studied with cortical stimulation using line orientation,
face-matching, and facial recognition tasks. Perceptual errors were ob-
served at the parietooccipital and also at the frontal sites. Recognition
errors occured in the posterior portion of the superior temporal gyrus and
parietal lobe. The different spatial tasks, such as face matching and line
orientation, were affected at similar sites, but dissociation was observed
between the sites for perception and recognition (Mateer, 1983). The ten-
tative conclusion reached was that visuospatial functions in the nondomi-
nant hemisphere appeared to be as discretely localized as verbal functions
in the dominant hemisphere. This is somewhat contrary to the conclu-
sions drawn from lesion studies by Semmes, Weinstein, Ghent, and Teu-
ber (1960) and Kertesz and Dobrowolski (1981).
ELECTROENCEPHALOGRAPHY AND EVOKED POTENTIALS
This physiological method of localization utilizes the small electric

potentials generated by neuronal activity that can be detected by surface
electrodes on the scalp or on the brain surface during operations. Elec-
troencephalograms have been used for more than 50 years. However, the
method suffers from relatively low specificities, since the electric poten-
tial changes often present a summation of remote effects. The localization
of cognitive function in aphasia with EEG has been reported by Ma-
rinesco, Sager, and Kreindler (1936) and Tikofsky, Kooi, and Thomas
(1960). Galin and Ornstein (1972) observed suppression of the alpha
rhythm in the dominant hemisphere during verbal tasks and in the non-
dominant side during spatial tasks. An attempt to assess intercortical
connections was made by measuring the degree of correlation between
pairs of electrodes, called cortical coupling (Callaway & Harris, 1974).
The introduction of signal-averaging computers allowed the measure-
ment of event-related potentials (ERP) by averaging out the background
EEG activity. Differences in ERP lateralization were observed by using
verbal and nonverbal stimuli (Buchsbaum & Fedio, 1970), words versus
nonsense syllables (Shelburne, 1972), and contextual meaning (Brown,
Marsh, & Smith, 1973; Teyler et al., 1973). Long-latency ERPs are sen-
sitive to task relevance and expectancy. The 300-msec positive compo-
nent is called P300, which is considered endogenous and less dependent
on stimulus characteristics. A potential that precedes movement or cogni-
tion is called the contingent negative variation (CNV). This has been cor-
related with hemispheric dominance and has been useful in various cog-
nitive studies (Low, Wada, & Fox, 1973). Another example of negative
potentials associated with processing is the N400 potential, which is re-
lated to the occurrence of context incongruity in serial semantic tasks
(Kutas & Hillyard, 1980; Nevill, Kutas, & Schmidt, 1982).
Cerebral event-related potentials are technically difficult because of
the many artifacts that interfere with the study. Some of these technical
difficulties have been discussed by Desmedt (1977). Another problem
44 ANDREW KERTESZ
with the method is the time restriction on the stimuli. The stimulation has
to be very short and have a definite onset, duration, and offset, in order to
be connected with cerebral event. In a way, it is the opposite to cerebral
blood flow studies (see below), which require more sustained cerebral
activity to be analyzed. The major advantage of the technique is that it
reflects a physiological event connected with actual cerebral processing,
although the resolution of localization is poor and the potentials are usu-
ally over a large area of the scalp and the brain. The origin and the genera-
tion of evoked potentials are often obscure, and their interpretation is
often subject to argument and to change over the course of years. For
instance, P300 potentials, which were initially interpreted as reflecting
cerebral perceptual processing, are now considered a postdecisional po-
tential, signifying the closure of the perceptual process and target selec-
tion. Since the method is noninvasive and is becoming relatively less
expensive, future expansion of the technique can be expected. Its major
requirement and, at the same time, limiting factor is the necessity of
having a trained electrophysiologist who devotes full time to this activity.
ISOTOPE LOCALIZATION OF LESIONS

Isotope scanning made it possible to outline lesions with consider-
able accuracy for the first time in patients who could be tested coinciden-
tally. Many radioisotopes will not penetrate the blood brain barrier,
unless there is some pathology. First, it was obvious that brain tumors and
vascular lesions could be seen and the extent of the lesion could be esti-
mated quite adequately. Subsequently, it became evident that cerebral
infarcts will also take up the isotopes approximately a week after the
stroke and continue to do so for about a month. Therefore, the technique
is particularly applicable for the localization of acute lesions and the
differentiation of acute from chronic lesions. For example, an infarct on a
CT scan would appear as an area of decreased density that would be the
same in the acute or the chronic state. Enhancement after contrast injec-
tion on the CT scan may help to clarify the problem, but a radioisotope
scan can still be used as a confirmatory evidence for an acute stroke.
Initial radioisotope localization of aphasias demonstrated a clear an-
terior-posterior dichotomy (Benson, 1967). An isotope study confirmed
that transcortical motor aphasia was seen with lesions of the supplemen-
tary motor area and also of Broca's area (Rubens, 1976). An objective
method of tracing a series of radioisotope lesions and correlating them
with aphasic syndromes defined by cognitive scores was a forerunner of
subsequent CT studies (Kertesz, Lesk, & McCabe, 1977). Maximum over-
lap for Broca's aphasia was over Broca's area and at the foot of the rolan-
dic fissure. Wernicke's aphasia involved the posterior superior temporal
gyrus and the posterior perisylvian region, and conduction aphasia in-
volved a smaller and somewhat more anterior region between Wernicke's
and Broca's area. Lesions causing conduction aphasia appeared to cluster
bimodally into anterior and posterior groups. Lesions of anomic aphasics
were widely scattered, and transcortical aphasics were outside the per-
isylvian speech area. Transcortical motor aphasics overlapped in the sup-
plementary motor area in the superior frontal lobe, and transcortical sen-
sory aphasics were posterior to Wernicke's area. The severity of aphasia as
measured with a standardized test correlated inversely with lesion size.
The major disadvantage of isotope scanning is that the normal land-
marks are few, so that accurate anatomical localization is difficult and
depends on inference. The edges of the lesions are somewhat indistinct
and can be subject to misinterpretation. A certain portion of the lesions is
missed because the scan is undertaken either too early or too late. This
restriction in time is often a disadvantage and excludes localization in
many patients studied chronically. However, as discussed above, the spe-
cific interval during which the scan will be positive makes it useful as a
diagnostic tool. This method has been eclipsed by the CT scan and subse-
quently by MRI scanning. However, there are still conditions where this
method is superior to all others, such as in the localization of herpes
simplex encephalitis, where, in several examples, the CT scan was nega-
tive while the radioisotope study showed a clear-cut uptake in both tem-
porallobes. Another continuing advantage of the method is its relatively
high availability in just about every hospital and its relatively low ex-
pense.
COMPUTERIZED TOMOGRAPHIC (CT) LESION

LOCALIZATION
CT scanning is a significant breakthrough in radiological diagnosis

and has also become the most important method in localizing lesions in
the brain. The technique combines X-ray technology and computeriza-
tion. Modern CT scanners provide fairly accurate bony detail and delinea-
tion of CSF spaces. The latest generation scanners also show some gray
and white matter differentiation and allow us to see changes in density,
such as brain edema. The technique utilizes a rotating X-ray beam, which
circles the head rapidly with a paired detector; the computer calculates
the densities of the tissue being imaged. Contrast enhancement with ra-
dio-opaque material, such as organic iodine, increases the visualization of
vascular structures and the increased vascularity around an acute infarct.
CT shows lesions earlier than radioisotope scans, but not as early as MRI
scans, and the early changes tend to be not as distinct as the ones obtained
after about 3 weeks. One of the reasons for negative localization with a
46 ANDREW KERTESZ
clear-cut clinical syndrome could be related to the timing of the CT scan.

Alteration in tissue density, 1 to 2 weeks after the infarct, sometimes
produces a decrease in contrast, so-called fogging, and occasionally even
a temporary disappearance of the lesion. Contrast enhancement and iso-
tope study may be helpful to visualize infarcts at that stage. Hemorrhages
are quite dramatic, and brain tumors are also shown superiorly on CT
scans. Old infarcts are quite distinct, with sharp margins and lower densi-
ties than the surrounding brain, and enlargement of the ventricles and
sulci also provide an accurate measurement of atrophy.
CT scanning has become the standard method of localization of le-
sions not only in clinical neurology but also in neuropsychology. The
correlation of lesions with behavioral studies began soon after the CT
equipment became available. These studies differ in the quality of the
scans, the sophistication of localization, the number of patients included,
the actual neuropsychological measurements used, and the definition of
the syndromes. One of the most important considerations, which is miss-
ing from some of the studies, is the time from onset. Some patients with
large lesions may have recovered considerably, and this may lead to the
conclusion that the area involved plays no role in a function. It is es-
pecially misleading to mix acute and chronic patients. CT studies, never-
theless, have contributed a great deal to our knowiedge about localization
of lesions in syndromes of language and cognitive impairment.
Severe persisting Broca's aphasia is associated with large lesions that
include not only Broca's area but also the inferior parietal and often the
anterior temporal region (Kertesz, 1979; Mohr, 1976). Small lesions in-
volving Broca's area only are compatible with good recovery. Small
Broca's area lesions often produce a minor motor aphasia (called "cortical
motor aphasia," "pure motor aphasia," or "verbal apraxia") (Kertesz,
1979; Mohr, 1976). Although both acute and chronic Broca's aphasia is
associated with Broca's area lesion (Kertesz, Harlock, & Coates, 1979),
Levine and Mohr (1979) presented autopsy evidence in favor of the
inferior precentral gyrus. Pure motor aphasia or verbal apraxia has been
associated with anterior subcortical as well as cortical lesions (Kertesz,
1983). Subcortical lesions produce atypical aphasias that may vary
clinically, depending on their anterior posterior location and the extent of
involvement of various subcortical structures.
The study of lesions in fluent aphasia has revealed not only a cortical
location but also thalamic involvement. Neologistic jargon output is asso-
ciated with lesions of both the superior temporal and the inferior parietal
region (Kertesz, 1979). Thalamic lesions, on the other hand, produce fluc-
tuating jargon aphasia alternating with relatively nonfluent speech (Mohr,
Watters, & Duncan, 1975). Repetition is often preserved (Cappa & Vignolo,
1979). Conduction aphasia is associated with lesions in the arcuate fas-
ciculus, as well as in the insula (Damasio & Damasio, 1983). Transcortical
sensory aphasia is associated with inferior temporooccipital lesions (Ker-

tesz, Sheppard, & MacKenzie, 1982).
Right-hemisphere function appears to be much less well localized
than that on the left side. This conclusion was first suggested from cortical
excision studies by Semmes et 01. (1960) and later received support from a
CT study by Kertesz and Dobrowolski (1981). Line bisection drawing,
block design, and Raven's scores were not significantly different among
frontal, central, subcortical, parietal, and occipital lesions. Lesion size
was certainly a contributing variable. In the study by Hier, Mondlock, and
Caplan (1983) left neglect motor impersistence and anosognosia tended to
occur with rather large lesions extending beyond the right parietal lobe.
On a few occasions, small deep lesions also produced behavioral abnor-
malities comparable to the larger superficial cortical lesions.
The correlation of lesion size and recovery was found to be signifi-
cant in overall group studies (Kertesz et 01., 1979; Yarnell, Monroe, &
Sobol, 1976). Care must be taken to distinguish between outcome mea-
sures and recovery rates. Sometimes patients with a large lesion and great-
er initial severity show greater recovery rates because they have much
more room to improve than those with smaller lesions and a higher level
of initial deficit.
The major advantage of a CT scan is its ability to localize lesions with
high anatomical resolution, especially in the late model scanners. CT
scanning is especially useful in chronic lesions where the edges of in-
farcts are well outlined, and in horizontal cuts that are usually oriented 15
degrees above the orbitomeatal line. CT scans are now available in most
major centers, and the method has become standard for neurological in-
vestigation. The scanning time is brief, and there is no discomfort associ-
ated with it.
Disadvantages are few. A clinical indication is needed for scanning
because of radiation. Repeated, frequent exposures to X rays are consid-
ered harmful. Cortical landmarks are not usually seen, and one has to use
ventricular and bony landmarks for anatomical orientation. Variations in
head positioning, head size, ventricular size, extent of atrophy, and cere-
bral asymmetries are important to consider in accurate localization. The
variation in the sulcal and gyral pattern and the depth of the cortex in the
central portions of the brain is often underestimated in CT studies. Many
of the "subcortical" lesions often involve insular cortex, for example. The
initial appearance of a CT infarct may be negative, and it is in the chronic
state that localization is most reliable. CT provides some image of dynam-
ic changes around the edge of the infarct in the acute state if enhancement
is used, but this is a somewhat more invasive method, involving the
injection of iodinated contrast, which is not entirely without risks. The
chronic state only shows the static structural damage.
In summary, CT remains the standard of in vivo localization with
48 ANDREW KERTESZ
which all other localizing methods are compared. Because of its availabil-
ity and easy administration, it will likely retain this position for consider-
able time to come. Much neuropsychological work has been done with CT
and will continue to be relevant.
CEREBRAL BLOOD FLOW (CBF) LOCALIZATION
The CBF technique utilizes the physiological and pathological altera-

tions in regional blood supply that is measured in a resting state and can
be related to functional alterations. The technique is based 0:0 measuring
the clearance of radioactive isotopes from various regions of the brain
through surface detectors. The values are expressed as percentages of the
hemispheric average. Increase of blood flow are assumed to be associated
with increased neuronal and, therefore, functional activity. A sustained
repetitive task of at least 3 to 5 minutes' duration is required for activa-
tion.
The color-coded images of CBF have become popular in illustrating
that the brain is activated in multiple locations when, for instance, a
person reads or the right hemisphere also "lights" up when one speaks.
The resting pattern shows precentral high and postcentral low flows (Ing-
var & Schwartz, 1974). Simple repetitive movements of the mouth, hand,
or foot augment CBF in the contralateral sensorimotor area, outlining the
topography of the cortical homunculus.
Visual perceptual tasks increase CBF occipitally and in the frontal eye
field. The posterior parietal regions are involved in shape discrimination.
Tactile perception activates the anterior parietal (sensory) area. Auditory
perceptual tasks increase regional blood flow to the temporal region verbal
tasks on the left and nonverbal sound discrimination on the right. Verbal
answers during the tests were also assumed to cause the increased flow to
the frontal regions. A trimodality stimulation paradigm appears as asumma-
tion of the single modality activation, but in addition, the posterior superi-
or frontal cortex, which is considered a supramodal processing area, is also
activated. Widespread cortical systems, in addition to specific primary
areas, show activation during complex processing (Lassen & Roland, 1983).
The technique is also sensitive to general arousal, in addition to selective
activation of local regions. A more detailed review of the application of
CBF in neuropsychology was reviewed by Wood (1980).
The important advantage of the technique is that it reflects physiolog-
ical metabolic changes accompanying psychological function and it can
be used to study normal processes. It must be remembered, however, that
these changes are not directly measuring neuronal events. These tech-
niques are best suited for the study of sustained repetitive acts of cogni-
tion. Lesions, such as infarcts producing a deficit, may appear as areas of
low flow with an area of "luxury perfusion" of high flow surrounding

them.
The major disadvantage is a relatively poor resolution of the nonin-
vasive xenon inhalation method. It only measures blood flow on the
surface. The more accurate intraarterial injection is rarely used because it
is invasive and requires the indications of an angiogram. A more recent
modification of the technique combines xenon-133 inhalation with com-
puterized tomography, achieving three dimensional representations in
slices similar to those in PET scanning (see SPECT section below).
POSITRON EMISSION TOMOGRAPHY (PET)
PET scanning measures oxygen and glucose metabolism by using a

positron-emitting isotope and a computerized tomographic scanner. It can
also be used to study regional blood flow. This complex and expensive
technique is available only in a few centers equipped with a particle
accelerator (cyclotron) and a team of nuclear physicists, radiophar-
macists, computer experts, isotope specialists, clinicians, and experimen-
tal psychologists. The positron labeled metabolites, such as 18F-desoxy-
glucose, must be given immediately at their source because of their short
half-life. The nature of tracer kinetics require that the physiological ac-
tivity studied has to be sustained for 20 to 40 minutes. Currently, a bolus
administration of oxygen-15 shortens the period of observation and al-
lows repeated measurements in the same subjects in one session (Raichle,
Herscovitch, Mintun, Martin, & Power, 1984). A new device, the "super-
PETT-l" is capable of a temporal resolution of less than 1 minute (Ter-
Pogossian et al., 1984). This opens up the possibility of following brief
physiological events in the brain in the future.
PET observations in normal volunteers showed hemispheric later-
alization that varied with stimulus content in the visual and auditory
systems, and with the response strategy of subjects. Particularly interest-
ing were metabolic asymmetries (left greater than right) in sensory depri-
vation and also on auditory stimulation (Mazziotta & Phelps, 1984). Bilat-
eral activation of the thalamus was observed in auditory verbal stimula-
tion. Ambient test conditions influenced the results considerably, and
caution must be exercised in interpreting all PET results because of that.
After a stroke, two phases of "luxury perfusion" are seen surrounding
an area of hypometabolism. PET studies showed prominent hypometabo-
lism of the ipsilateral subcortical regions and the basal ganglia in cortical
focal lesions even though these structures remained intact on the CT scan
(Metter, Wasterlain, Kuhl, Hanson, & Phelps, 1981). There are often much
larger areas of hypometabolism around a lesion on PET than the de-
creased densities seen on CT. Following subcortical lesions, cortical ac-
50 ANDREW KERTESZ
tivation is also decreased, possibly because of lowered metabolic demand.

Future applications of the PET technique may reveal areas in the brain
that playa role in functional compensation for structural lesions. So far,
contralateral homologous structures have been implicated. In epilepsy,
cortical zones considered responsible for seizures were hypometabolic
without a structural lesion being demonstrated. In Huntington's disease, a
decrease in metabolism was seen in the caudate and putamen early, pre-
ceding bulk tissue loss. PET studies of Alzheimer's disease show wide-
spread cortical hypometabolism. Elderly normals also show a significant
decrease in cerebral glucose metabolism (Benson, Metter, Kuhl, & Phelps,
1983).
The major advantage of PET is that it measures cerebral metabolism,
therefore function, in normals and in disease. The major disadvantages
are large expense and general unavailability, poor anatomical resolution,
uncertainties between the temporal sequence of cerebral events and the
isotope kinetics in time, and the influence of unrelated cerebral activity.
SINGLE-PHOTON COMPUTERIZED TOMOGRAPHY (SPECT)
Regional blood flow can also be estimated by using a single-photon

technique that is much cheaper than PET scanning (Hill, 1980). This is the
latest of the functional localizing methods, but experience is rapidly ac-
cumulating because it is more generally available than PET. Single-pho-
ton tracers can be used to measure blood flow, blood volume, and also
blood brain barrier integrity. Disadvantages are that large amounts of cost-
ly xenon gas (in nearly anesthetic doses) or iodinated isopropyl ampheta-
mine (IMP) are needed. Sequential scans at each level mean repeated
exposure to X rays. The exact positioning is crucial, and this restricts the
study to cooperative subjects. In brain damage, IMP is not only a blood
flow tracer but also an index of impaired cellular function. It has a longer
biological half-life (13 hours) than xenon, which is inert. Studies in
neuropsychology with SPECT are just beginning to appear (Tikofsky et
aI., 1984).
MAGNETIC RESONANCE IMAGER (MRI)
MRI is the latest imaging modality, and it shows the greatest promise
in accurate localization of lesions without invasive radiation. The tech-
nique uses the inherent magnetic properties of spinning atomic nuclei by
placing the structure to be imaged in a large magnet and applying short-
wave radiofrequency pulses to produce a resonance signal that can be
quantified and computerized (Doyle et a1., 1981). Superior anatomical
detail can be achieved with excellent gray and white matter differentia-
tion and an accurate outline of the edge of the brain from CSF spaces. The
brain can be imaged in coronal and sagittal sections, in addition to the
horizontal ones, which is the usual plane obtained in other modalities.
This imaging flexibility, combined with anatomical accuracy, has already
established MRI as a useful clinical and research tool. The apparent lack
of biohazard allows it to be used to study normals without clinical indica-
tions, as well as a more frequent repetition of imaging in patients than is
possible in other modalities that use ionizing radiation.
Various pulse sequences can be used to probe the metabolic and
molecular environment of various regions of the brain, and this makes the
technique much more dynamic than the CT scan. Currently, two major
pulse sequences are used. One called "inversion recovery" emphasizes
gray and white matter differences and provides excellent anatomical de-
tail. The second, called "spin echo," is utilized to detect edema and other
metabolic changes associated with lesions. Spin echo sequences are often
superior to CT scanning in the early detection of cerebral infarct, thus
reducing the false negative rate obtained on CT in the first 2 or 3 days of a
stroke. The technique is particularly suited to detect demyelinating
plaques and some other degenerative diseases in addition to early strokes.
The evolution of an infarct follows a regular pattern, with a gradual clear-
ing of the central portion on the spin echo image that leaves a ring of
increased signal intensity around the infarct in the chronic state. This ring
appears to extend beyond the hypo density seen on CT. The selection of
various repetition times and pulse sequences permits a differentiated vi-
sualization of brain tissues, such as the CSF versus the brain, by altering
the influence of various tissue properties. The inversion recovery se-
quence emphasizes the T1 or longitudinal relaxation time, which is the
return of the spinning nuclei to their original axis in the large magnetic
field. The spin echo sequences are weighted for the T2 relaxation time,
which depends on the interaction of the spinning nuclei with each other.
MRI is a new technology that is currently in a relatively unstable state
of development. The expense is somewhat greater than that of scanners
for most prototype equipment. Recently, this modality has become avail-
able to most major centers. It will likely not displace, but will comple-
ment CT scanning. Clinical imaging can be accomplished in 15 minutes,
but for research purposes the pulse sequences can take an hour. Lesions
are well outlined, and the dynamic changes around or in the lesions can
be followed over time. Variations in the images are dependent on the
molecular environment and water density of the tissues. Actual func-
tional changes or psychological states cannot be examined with the cur-
rent equipment.
Neuropsychological studies with MRI include the localization of le-
sions in aphasia and a study of anatomical asymmetries utilizing the
52 ANDREW KERTESZ
excellent anatomical detail on the inversion recovery scan, which allows

the direct visualization of the opercular sulci in the planum temp orale
(Kertesz & Black, 1983; Kertesz, Black, & Howell, 1984). In the future, with
improved technology, it is likely that metabolic alterations such as
changes in high energy phosphates can be imaged and that physiological,
in addition to structural, changes in the brain can be investigated.
DISCUSSION AND CONCLUSION
The integration of structural alterations with functional changes pres-

ents a challenge that is at its greatest in the central nervous system. Cog-
nitive behavior and language in man is, to a large extent, tied to localiz-
able brain structures, and the purpose of this chapter was to summarize
the methods of localization. Any attempt to correlate function with le-
sions must cope with a multitude of theoretical and practical problems.
These issues in localization have undergone considerable change in the
last few years, not only because of revolutionary new technology in lo-
calization but also because of our increasing sophistication and knowl-
edge of neuropsychological phenomena.
One of the most important issues in localization is the extent to
which function is localizable. Primary sensory and motor areas in the
brain are clearly related to well-defined functions. Such a relationship is
exemplified by the recent physiological studies exploring the relationship
of columnar organization of cerebral cortex to visual feature analysis
(Hubel & Wiesel, 1965). Secondary and tertiary areas of the brain that are
associated with elaboration of cognitive and language function are much
less clearly localizable. Language appears to take an intermediate position
among cognitive processes in that it is well localized to the left side, in
most instances, and there exists a dichotomy of output and articulation,
and input and comprehension, in relation to the different structures in-
volved. There are similar dichotomies in agrammatism and paragram-
matism, phonemic and semantic paraphasias, and various dissociations
in reading and writing that may be related to differential localization.
Other cognitive processes, such as semantic elaboration and retrieval, and
multiple modality associations, appear to occupy a much less localizable
network that can be impaired from lesions in many areas of the left hemi-
sphere. Memory function also represents a less localizable component of
cognition, even though recent memory' deficits are clearly related to le-
sions of the limbic lobe, which itself is a widely distributed but localiza-
ble network. The diffuse reticular activating system and thalamocortical
projections underlie the widespread activation involved in attention and
in the performance of many cognitive tasks. Visuospatial functions are
elaborated, mainly by the right hemisphere, although some components
are clearly related to left-hemisphere function as well. Recent studies

have confirmed the right-hemisphere dominance for directed attention
and visuospatial function. Most of these functions are complex enough
that bilateral hemispheric integration is necessary for their performance.
The definition of what constitutes a function is often dynamic and
arbitrary. Complex alternative theories are offered analyzing certain func-
tions. A great deal of fractionation and the reduction of complex behav-
iors to its elements risks the loss of meaning and biological significance to
the organism. Some psychological concepts of functions may not be ap-
propriate to describe the brain function as we know it from the structural
point of view. On the other hand, some structural alterations cannot be
correlated with function, but only with the deficit syndrome that is seen
after a lesion. The behavior observed after a lesion may not be analyzable
in terms of normal function. This often prevents any direct conclusion
about normal mechanisms that is based on structural and pathological
observations. However, this does not diminish the usefulness of the
clinicopathological correlations. The analysis of behavior after a lesion
should be complemented by studies of function in normals. The function
lost after a circumscribed area of brain damage is clearly better analyzed if
the components of the function are known and can be identified to be
missing.
A major objection to relating lesions to function is that the observed
function after a lesion is related not to the area damaged but to other areas
or structures that take over function or become reorganized to substitute
for it. At times after a lesion, not only do deficit symptoms appear but new
behavior or positive symptoms can also be observed. These may represent
elements of neural activity that had been suppressed by the structure that
is destroyed by the lesion. This argument pertains not only to lesions but
to other methods as well, such as electrical stimulation or the recording of
single units, because these are all connected to the remaining system. This
consideration is important for the interpretation of any method, but it
should not eliminate them from the correlation of function with structure.
Most functions are represented in the brain as a network of intercon-
nected sites. It is also likely that these networks overlap and therefore
lesions of a certain cortical area are going to produce multiple deficits
(syndromes). Since some of the components of a certain function remain
intact when only one circumscribed area is damaged, recovery often oc-
curs by reorganization of the residual network. A severe and lasting im-
pairment of a function requires the multiple involvement of several com-
ponents, usually the majority of them. Some of the functional networks
are more widely distributed than others, and therefore the same function
can be impaired from several cortical areas that may be quite far from each
other.
The variability of deficits following similar lesions in the brain are
54 ANDREW KERTESZ
partly related to problems of localization that are corrected by improve-

ment in the techniques of localization itself. However, some factors are
related to the effect of recovery or reorganization in the nervous system.
Therefore, in all attempts to correlate deficit with lesion, the time from the
onset is crucial. Failure to consider this variable is a major source of
confusion in neuropsychology. The extent of substitution and reorganiza-
tion in the central nervous system is considerable, far beyond that which
can be attributed to anatomical regeneration. This, of course, complicates
efforts to localize function in any part of the nervous system.
The interconnection of functional networks results in distance effects
in sudden lesions, which is known as "diaschisis" (von Monakow, 1914).
This principle, in essence, explains the initial effect of a lesion on a
function that produces a more severe deficit, which eventually recovers as
the rest of the network becomes reorganized. This explains much of the
observed recovery and plasticity in central nervous system lesions. On the
other hand, slowly growing lesions produce relatively little functional
deficit. Therefore, the etiology of the lesion is a very important variable in
considering lesion localization. For instance, transcortical sensory apha-
sia or Wernicke's aphasia may appear with Alzheimer's disease, associ-
ated with diffuse neuronal degeneration that is not localizable with our
present methods to the same extent as a focal infarct causing those defi-
cits.
Differences between individuals are also factors that need to be taken
into consideration in the interpretation of the localization of function.
The plasticity of the young brain allows for greater compensation by ho-
mologous structures of the other hemisphere. There may be differences in
the distribution of the function between the left and right hemispheres in
individuals with various degrees of handedness or some other factor that
influences lateralization. Gender differences have also been postulated
concerning the functional organization of language and visuospatial
function.
The various methods of correlating cerebral function with structure
are summarized in Table 1. This table indicates the major characteristics
of the methods, such as their accuracy, relation to function, expense,
availability, and the various advantages and disadvantages. Table 2 sum-
marizes the areas of application of each method. This indicates the work
that has been done with the technique, as well as the suitability for future
studies. The investigator must have a great deal of familiarity with each
technology if the appropriate experiments or study paradiagms are to be
designed with these methods. However, a considerable amount of general
knowledge is needed also for those who want to evaluate the information
obtained with these complex techniques, which are ever increasing in
number.
TABLE 1. Localization Techniques
Anatomical Relation to Performance

resolution Availability function duration Invasiveness Expense Advantages Disad vantages
Autopsy The ultimate in Restricted to Indirect re- Only in the Postmortem Moderate Accuracy Retrospecti ve
accuracy some cases mote past
Skull wounds Poor, only an ap- Restricted group Indirect Unlimited af- Penetrating Minor In vivo exam Inaccurate
proximation ter lesion injury
Neurosurgery Accurate Restricted group Indirect Unlimited af- Neurosurgery Moderate In vivo exam Altered tissue
ter lesion
Cortical stim- Accurate surface Restricted to a Indirect 12 sec and Neurosurgery Moderate to Small areas Limited test-
ulation only few control expensive ing
EEG, evoked Poor (lobar only) Limited to phys- Expecta- Minutes and Noninvasive Moderate Physiological Low specif-
potentials iologists tions, at- millise- icity
tention conds
Isotope scans Poor (outline Everywhere Indirect Unlimited Noninvasive Minor In vivo exam Poor resolu-
only) acute stage tion
CT scans Accurate to a de- Almost every- Indirect Unlimited af- Minor radia- Moderate In vivo exam Limited land-
gree where ter lesion tion marks
CBF scans Poor, better ar- Restricted to Closely re- 5 min Minor (more Moderate to Related to Poor resolu-
terially stroke centers lated if arterial) expensive function tion
PET scans Poor but improv- Only few Direct (met- 20-60 min Minor radia- Veryexpen- Directly func- Resolution,
ing abolic) <1 min in tion sive tional expense
future
SPECT scans Poor but improv- Restricted to Closely re- 5-20 min Minor radia- Moderate to CBF in slices Resolution,
ing few centers lated tion expensive expense
MRI scans Accurate Expanding ma- Indirect Unlimited af- Noninvasive Moderate to Best resolution Not yet avail-
jor centers ter lesion expensive able
56 ANDREW KERTESZ
TABLE 2. The Areas of Application
Autopsy Aphasias, apraxias, agnosias, memory, neglect, constructional

and visuospatial syndromes
Skull wounds Frontal lobe syndromes, aphasias, visual field defects, memory
Neurosurgery Temporal, frontal, and occipital lobe function; memory; callo-
sal disconnection
Cortical stimulation Naming, articulation, grammar, semantics, gesture, polyglots,
face matching
EEG and ERP Cerebral lateralization, attention, expectation, task relevance,
central processing
Isotope scans Aphasias, agnosias, memory, neglect, superior for herpes ence-
phalitis
CT scans All lesion localization; superior for hemorrhages, tumors, and
chronic stroke
CBF scans Resting activity, normal physiology, perceptual functions, later-
alization, integrative processes
PET scans Perceptual activation, lateralization, distance effects of lesions,
epilepsy, subcortical lesions
SPECT scans Lateralization, processing, perceptual activation, lesion effect
MRI scans All lesion localization, superior for demyelination and early
strokes, anatomical studies
REFERENCES
Benson, D. F. (1967). Fluency in aphasia: Correlation with radioactive scan localization.
Cortex, 3, 373-394.
Benson, D. F., Metter, E. J., Kuhl, D. E., & Phelps, M. (1983). Positron-computed tomography
in neurobehavioral problems. In A. Kertesz (Ed.), Localization in neuropsychology. New
York: Academic Press.
Bernheim, F. (1900). De l'aphasie motrice. Paris: Carre et Naud.
Brodmann, K. (1909). Vergleichende Lokalisationslehre der Grosshirnrinde. Leipzig: Barth.
Brown, W. S., Marsh, J. T., & Smith, J. C. (1973). Contextual meaning effects on speech-
evoked potentials. Behavioral Biology, 9, 755-761.
Buchsbaum, M., & Fedio, P. (1970). Hemispheric differences in evoked potentials to verbal
and nonverbal stimuli in the left and right visual fields. Physiology and Behavior,S,
207-210.
Burkhardt, G. (1891). Ueber Rindenexcisionen, als Beitrag zur operativen Therapie der Psy-
chosen. Allgemeine Zeitschrift fiir Psychiatrie, 47, 463-548.
Callaway, E., & Harris, P. R. (1974). Coupling between cortical potentials from different
areas. Science, 183, 873-875.
Cappa, S. F., & Vignolo, L. A. (1979). "Transcortical" features of aphasia following left
thalamic hemorrhage. Cortex, 15, 121-130.
Damasio, A. R., & Damasio, H. (1983). Localization of lesions in achromatopsia and pros-
opagnosia. In A. Kertesz (Ed.), Localization in neuropsychology. New York: Academic
Press.
Dandy, W. (1931). The effect of total removal of the left temporal lobe in a right handed
individual: Localization of areas of the brain concerned with speech. Journal of Nervous
and Mental Disease, 74, 739-742.
Desmedt, J. E. (1977). Some observations on the methodology of cerebral evoked potentials

in man. In J. E. Desmedt (Ed.), Attention, voluntary contraction and event-related cere-
bral potentials, Progress in Clinical Neuropsychology (Vo!. 1). Basel: Karger.
Doyle, F. H., Gore, J. C., Pennock, J. M., Bydder, G. M., Orr, J. S., Steiner, R. E., Young, I. R.,
Burl, M., Clow, H., Gilderdale, D. J., Bailes, D. R., & Walters, P. E. (1981). Imaging of the
brain by NMR. Lancet, 8237, 53-57.
Economo, C., & Koskinas, G. N. (1925). Die Cytoarchitektonik der Hirnrinde des er-
wachsenen Menschen. Berlin: Springer-Verlag.
Galaburda, A. M., & Mesulam, M. M. (1983). Neuroanatomical aspects of cerebrallocaliza-
tion. In A. Kertesz (Ed.), Localization in neuropsychology. New York: Academic Press.
Galin, D., & Ornstein, R. (1972). Lateral specialization of cognitive mode: An EEG study.
Psychophysiology, 9, 412-418.
Gall, F. J., & Spurtzheim, G. (1810). Anatomie et physiologie du systeme nerveux en general
et du cerveau en particulier (Vol. 1). Paris: F. Schoel!.
Gazzaniga, M. S. (1970). The bisected brain. New York: Appleton-Century-Crofts.
Head, H. (1926). Aphasia and kindred disorders of speech. Cambridge: Cambridge Univer-
sity Press.
Hecaen, A., & Angelergues, R. (1964). Localization of symptoms in aphasia. In A. De Reuck &
M. O'Connor (Eds.)' Disorders of language. London: Churchill.
Henschen, S. E. (1922). Klinische und anatomische Beitrage zur Pathologie des Gehirns
(Vols. 5-7) Stockholm: Nordiska Bokhande!.
Hier, D. B., Mondlock, J., & Caplan, L. R. (1983). Behavioral abnormalities after right hemi-
sphere stroke. Neurology, 33, 337-344.
Hill, T. C. (1980). Single-photon emission computed tomography to study cerebral function
in man. Journal of Nuclear Medicine, 21, 1197-1199.
Hubel, D., & Wiesel, T. (1965). Receptive fields and functional architecture in two nonstriate
visual areas (18 and 19) of the cat. Journal of Neurophysiology, 28, 229-289.
Ingvar, D. H., & Schwartz, M. S. (1974). Blood flow patterns induced in the dominant
hemisphere by speech and reading. Brain, 97, 273-288.
Kertesz, A. (1979). Aphasia and associated disorders: Taxonomy, localization and recovery.
New York: Grune & Stratton.
Kertesz, A. (1983). Localization in neuropsychology. New York: Academic Press.
Kertesz, A., & Black, S. E. (1983). Lesion localization in aphasia. Paper presented at the
Academy of Aphasia meeting, Minneapolis.
Kertesz, A., Black, S. E., & Howell, J. (1984). Anatomical asymmetries on nuclear magnetic
resonance: Correlation with dissection and functionallateralization. Paper presented at
the Academy of Aphasia meeting, Los Angeles.
Kertesz, A., & Dobrowolski, S. (1981). Right hemisphere deficits, lesion size and location.
Journal of Clinical Neuropsychology, 4, 283-299.
Kertesz, A., Harlock, W., & Coates, R. (1979). Computer tomographic localization, lesion size
and prognosis in aphasia. Brain and Language, 8, 34-50.
Kertesz, A., Lesk, D., & McCabe, P. (1977). Isotope localization of infarcts in aphasia. Brain,
100, 1-18.
Kertesz, A., Sheppard, A., & MacKenzie, R. A. (1982). Localization in transcortical sensory
aphasia. Archives of Neurology, 39, 475-478.
Kleist, K. (1962). Sensory aphasia and amusia. In F. J. Fish & J. B. Stanton (Eds.), The
myeloarchitectonic basis. Oxford: Pergamon Press.
Kutas, M., & Hillyard, S. A. (1980). Reading between the lines: Event-related brain potentials
during natural sentence processing. Brain and Language, 11,354-373.
Lassen, N. A., & Roland, P. R. (1983). Localization of cognitive function with cerebral blood
flow. In A. Kertesz (Ed.), Localization in neuropsychology. New York: Academic Press.
58 ANDREW KERTESZ
Levine, D., & Mohr, J. (1979). Language after bilateral cerebral infarctions: Role of the minor
hemisphere in speech. Neurology, 29, 297-338.
Low, W., Wada, J. A., & Fox, M. (1973). Electroencephalographic localization of the conative
aspects of language production in the human brain. Transactions of American Neu-
rological Association, 98, 129-133.
Luria, A. R. (1970). Traumatic aphasia. The Hague: Mouton.
Marie, 1"., & Foix, C. (1917). Les aphasies de guerre. Revue Neurologique, 24, 53-87.
Marinesco, G., Sager, 0., & Kreindler, A. (1936). Etudes electroencephalographiques; elec-
troencephalogrammes dans l'aphasie. Bulletin of the Academy of Medicine of Paris,
116, 182.
Mateer, C. (1983). Localization of language and visuospatial functions by electrical stimula-
tion. In A. Kertesz (Ed.), Localization in neuropsychology. New York: Academic Press.
Mazziotta, J. C., & Phelps, M. E. (1984). Human sensory stimulation and· deprivation:
Positron emission tomographic results and strategies. Neurology, 14, 50-60.
Metter, E. J., Wasterlain, C. G., Kuhl, D. E., Hanson, W. R., & Phelps, M. E. (1981). 18 FDG
positron emission computed tomography in a study of aphasia. Annals of Neurology,
10, 173-183.
Mettler, R. (1949). Selective partial ablation of the frontal cortex: A correlative study of the
effects of the human psychotic subjects. New York: Hoeber.
Mohr, J. P. (1976). Broca's area and Broca's aphasia. In H. Whitaker & H. A. Whitaker (Eds.),
Studies in neurolinguistics (Vol. 1). New York: Academic Press.
Mohr, J. P., Watters, W. C., & Duncan, G. W. (1975). Thalamic hemorrhage and aphasia. Brain
and Language, 2, 3-17.
Moutier, F. (1908). L'aphasie de Broca. Paris: Steinheil.
Neville, H. J., Kutas, M., & Schmidt, A. (1982). Event-related potential studies of cerebral
specialization during reading. 1. Studies of normal adults. Brain and Language, 16, 300-
315.
Nielsen, J. M. (1946). Agnosia, apraxia, aphasia. New York: Hoeber.
Ojemann, G. A., & Mateer, C. (1979). Human language cortex: Localization of memory,
syntax and sequential motor-phoneme identification systems. Science, 250, 1401-1403.
Ojemann, G. A., & Whitaker, H. A. (1978). Language localization and variability. Brain and
Language, 6, 239-260.
Penfield, W., & Roberts, L. (1959). Speech and brain mechanisms. Princeton, NJ: Princeton
University Press.
Raichle, M. E., Herscovitch, P., Mintun, A. M., Martin, W. R. W., & Power, W. (1984).
Dynamic measurements of local blood flow and metabolism in the study of higher
cortical function in humans with positron emission tomography. Neurology, 14,48-49.
Rei!, J. C. (1809). Untersuchungen tiber den Bau des grossen Gehirns im Menschen. Archiv
fiir Anatomie und Physiologie, 9, 136-524.
Rubens, A. B. (1976). Transcortical motor aphasia. In H. Whitaker & H. A. Whitaker (Eds.),
Studies in neurolinguistics (Vol. 1). New York: Academic Press.
Russell, W. R., & Espir, M. L. E. (1961). Traumatic aphasia. London: Oxford University
Press.
Sanides, F. (1962). Die architektonik des menschlichen Stirnhirns. Berlin: Springer-Verlag.
Schiller, F. (1947). Aphasia studied in patients with missile wounds. Journal of Neurology,
Neurosurgery and Psychiatry, 10, 183.
Semmes, J., Weinstein, S., Ghent, L., & Teuber, H. 1. (1960). Somatosensory changes after
penetrating brain wounds in man. Cambridge, MA: Harvard University Press.
Shelburne, S. A. (1972). Visual evoked responses to word and nonsense syllable stimuli.
Cortex, 12, 325-336.
Smith, A. (1966). Speech and other functions after left (dominant) hemispherectomy.
Journal of Neurology, Neurosurgery and Psychiatry, 29, 467-471.
Sperry, R. W., Gazzaniga, M. S., & Bogen, J. E. (1969). Interhemispheric relationship: The
neocortical commissures; syndromes of hemispheric disconnection. In P. J. Vinken & G.
W. Bruyn (Eds.), Handbook of clinical neurology (Vol. 4). Amsterdam: North-Holland.
Talairach, J., & Szikla, G. (1967). Atlas of stereotaxic anatomy of the telencephalon. Paris:
Masson.
Ter-Pogossian, M. M., Ficke, D. C., Mintun, M. A., Herscovitch, P., Fox, P. T., & Raichle, M.
E. (1984). Dynamic cerebral positron emission tomographic studies. Neurology, 15, 46-
47.
Teyler, T. J., Roemer, R. A., Harrison, T. F., et al. (1973). Human scalp-recorded evoked-
potential correlates of linguistic stimuli. Bulletin of the Psychonomic Society, 1, 333-
324.
Tikofsky, R. S., Collier, B. D., Saxena, V. K., Hellman, R. S., Zielonka, J. A., & Palmer, D. W.
(1984). 1-123 iodoamphetamine (IMP) single photon emission computerized tomogra-
phy (SPECT) imaging in normals and aphasics. Paper presented at the Academy of
Aphasia meeting, Los Angeles.
Tikofsky, R. S., Kooi, K. A., & Thomas, M. H. (1960). Electroencephalographic findings and
recovery from aphasia. Neurology, 10, 154-156.
Van Buren, J., Fedio, P., & Frederick, G. (1978). Mechanism and localization of speech in the
parietotemporal cortex. Neurosurgery, 2, 233-239.
Von Keyserlingk, D. G., De Bleser, R., & Poeck, K. (1983). Stereographic reconstruction of
human brain CT series. Acta Anatomique, 115, 336-344.
Von Monakow, C. (1914). Die Lokalisation im Grosshirn und der Abbau der Funktionen
durch corticale Herde. Wiesbaden: Bergmann.
Wood, F. (1980). Theoretical, methodological, and statistical implications of the inhalation
rCBF technique for the study of brain-behavior relationships. Brain and Language, 9, 1-
8.
Yarnell, P. R., Monroe, M. A., & Sobel, 1. (1976). Aphasia outcome in stroke: A clinical and
neuroradiological correlation. Stroke, 7, 516-522.
4
ERPs and Advances in

N eurolinguistics
SIDNEY J. SEGALOWITZ
Neurolinguistics is the study of the brain correlates of language, and the

methods used have been drawn from experimental and clinical neuropsy-
chology as well as from experimental and descriptive linguistics. Tradi-
tionally, the database has drawn from clinical aphasiology (Caplan, 1987;
Lecours, Lhermitte, & Bryans, 1983), but experimental neuropsychology
has added a number of research paradigms that neurolinguists have found
useful (Beaumont, 1982; Bryden, 1982; Hannay, 1986).
The event-related potential (ERP) technique is beginning to prove
very useful in neurolinguistics, providing another metric with which to
attack some long-standing questions. In this chapter, I will summarize
some of the ways the ERP technique may prove useful in addressing
neurolinguistics research questions and have focused on neurolinguistic
issues that I think are amenable for use with the ERP paradigm. The
idiosyncratic mix for research questions reflects both current issues in the
field and my own particular interests. I will begin, however, with a review
of some ERP techniques and technical controversies that relate directly to
the research questions of interest.
OVERVIEW OF ERP METHODOLOGY
There are a great many reviews of ERP techniques for those intending
to set up an ERP laboratory (e.g., Callaway, Tueting, & Koslow, 1978;
Gaillard & Ritter, 1983). In this section, rather than outline these details, I
SIDNEY J. SEGALOWITZ • Department of Psychology, Brock University, St. Catharines,

Ontario L2S 3A1, Canada.
61
62 SIDNEY J. SEGALOWITZ
will describe in rough terms the philosophy underlying this methodology

for cognitive neuropsychology.
The basic assumption behind the measuring of event-related elec-
trical potential shifts on the scalp is that these shifts are sensitive to the
cognitive activity of the brain. As obvious as this assumption is, it is
useful to examine it in some detail to explicitly state its limitations and its
strengths. For example, this sensitivity is a mixed blessing, since the
resultant EEG wave form during any particular event is a reflection of the
many information processes and nervous system activities occurring dur-
ing the event. Most of these processes are not of interest to our research
question, especially those relating to movements idiosyncratic to the par-
ticular event. The way to eliminate the effects of these activities is, of
course, to average the EEG wave form over many trials with appropriate
control conditions, effectively averaging out any random nonevent-relat-
ed potential shifts and summating those that are truly event-linked. Be-
cause of restrictions designed to help eliminate the unwanted noise fac-
tors, we find that the ERP paradigm is not as ideally suited to measuring
brain events as we might like. More on this later.
ERP versus Reaction Time

One of the strengths of the cognitive experimental paradigm is the
imaginative ways that reaction times (RTs) can be used to examine com-
ponent processes of a task. For example, if we find that subjects con-
sistently require more time to solve Task A than Task B, we can safely
conclude that the tasks require different cognitive or motor resources.
This paradigm has been elaborated to a high degree of sophistication
(Posner, 1986). It is difficult, however, to rely on RT (or error-score) para-
digms when we have reason to believe that the cognitive processes are not
additive, or when we are interested in examining processing that occurs
during the event and not just as a result of its completion. The ERP
paradigm is especially useful here since the wave form is a reflection of
ongoing activity.
A second advantage of the ERP over the RT paradigm in neurolinguis-
tic research relates to issue of brain localization of function. There are
limits to how much visual half-field or dichotic listening paradigms will
aid our determining even hemispheric asymmetries of processing. In
these cases, we once again are limited by the response reflecting only the
finished product, as well by the attendant problem of the artifacts peculiar
to these paradigms (Bryden, 1982). Now, there are considerable limits on
the localizing potential of most ERP paradigms, but there are also reason-
able conclusions that can be made from information on brain localization
of the signals.
ERPs AND ADVANCES IN NEUROLINGUISTICS 63
FIGURE 1. Averaged ERP results from

P300
Segalowitz and MacGregor (1987) il-
lustrating a common configuration of
peaks. In this case, the subjects were
presented with English words on a CRT
screen to read aloud, one at a time. The
ERP duration is 500 msec. The sim-
ilarity across the hemispheres indicates
that functional asymmetries contribute
very little variance to the ERP, although LEn PARI ElAL
this variance can be consistent (see Fig-
ure 3).
Methods of Analysis
Rather than outline the evidence in favor of and against interpreta-
tions of the various wave peaks that have been propounded, I would like
to focus on the aspects of wave analysis that pertain to neurolinguistics
specifically. The first question we must address in cognitive electrophysi-
ology is whether the effect we find is attributable to the specific cognitive
processes of interest or to general attentional processes. For example, the
early negative (Nl00 and sometimes N200), late positive (P300), and later
negative (N400) components seem to be sensitive to attentional aspects of
the typical cognitive ERP paradigm, and the peaks have become quite
famous in the ERP literature (see Figure 1 for a simple example). The size
and latency of these components may vary depending on the stimuli or
the processing task, allowing us to conclude that there are processing
differentiations being made. For example, reading high-frequency words
may produce values quite different from reading low-frequency words. As
useful as these components are in cognitive electrophysiology (see Fried-
man, Simson, Ritter, & Rapin, 1975; Gaillard & Ritter, 1983; Kutas & Van
Petten, 1988, for reviews), we can never be sure that the cognitive compo-
nent we are interested in is reflected in these famous peaks, since the
wave contains much information beyond these particular points.
An alternate approach is to compare the entire wave shape across the
experimental conditions. This is done usually with one of two methods,
either with what is known as the peA-ANOVA technique (Donchin, 1966)
or by using Pearson correlations in particular time frames across the wave.
The former is a highly controversial technique that derives time-clustered
components by which to parse the wave shape, and then compares the
factor scores for these components across the independent variables. The
factor scores essentially reflect amplitude differences associated with the
independent variables. The controversy surrounds the issue of whether or
Factor 1
Factor 2 FIGURE 2. An example of a PCA out-
put from an ERP experiment
Factor 3 (Segalowitz & Cohen, 1988). The cen-
Factor 4 troid is the averaged response over
frontal, parietal, and temporal sites to a
Factor 5 series of consonant-vowel speech
Factor 6 sounds. The eight factors represent sta-
tistically orthogonal time clusters that
Factor 7 together account for 93% of the vari-
ance in the raw data. See text for further
Factor 8
discussion.
not the time-based components are true underlying factors in the makeup
of the aggregate averaged ERP (Mocks, 1986; Mocks & Verleger, 1986;
Wood & McCarthy, 1984). The true test of this technique is, like any other
technique, in replication and generalization, which has been occasionally
done (e.g., Molfese, 1978, vs. Segalowitz & Cohen, 1989), but much less
often than for the more traditional analysis of wave peaks. This situation
may be as much a product of the relatively few laboratories utilizing this
complex technique as of the inherent complications in using it (Kramer &
Donchin, 1987).
For example, one of the facets of this technique that deserves caution
is that in a single study, the averaged ERPs are subjected to a principal
components analysis that derives several factors, each representing some
time clustering (see Figure 2). For each factor, the analysis produces factor
scores for each combination of independent variables for each averaged
ERP entering the analysis. These factor scores indicate the relative weight
for the dependent measures-Leo the cell in the design-leading to an
analysis of variance for each factor. One may question the true Type I
error when, say, 10 such ANOVAs are performed, not exactly on the same
data set, but on scores derived from the same data set. We do not know
whether this should be a concern, and perhaps the test of time will be an
empirical one. For example, Segalowitz and Cohen (1988) tried to repli-
cate Molfese's (1978) study of cerebral asymmetries in ERPs produced to
certain phonological contrasts. The effect seemed to generalize well to the
new stimuli, contexts, and electrode sites used by Segalowitz and Cohen,
but the nagging problem of true alpha level remained. To test for this, they
performed 100 repeated ANOV As on the factor scores for the critical
condition that replicated Molfese for each of eight factors, each time ran-
domizing the factor score set. There were in this way nine chance replica-
tions of the original result. They concluded that the .025 level was more
than safe as a level of Type I error to accept for the replication.
It may be that the strength of the PCA-ANDV A technique is not sim-
ply in finding differentiations in the ERPs that are very difficult to make
from a visual scan of the data, but is to be found in the focusing of
attention to a particular portion of the wave. As long as we accept the idea
that the ERP is highly sensitive to all sorts of processing, we will be left
with the problem of deciding which subtle fluctuation in the waves are
reflections of our experimental design. To return to the example given
from Segalowitz and Cohen, after verifying that the effect replicating Mol-
fese was not a statistical quirk, they returned to the raw ERPs and exam-
ined the wave in the section indicated by the factor in question. If the
factor had been an emergent statistical product, there should be no corre-
spondence in the raw data. They found that the raw data produced the
same interaction. It was unlikely, though, that they would have thought to
look at that part of the wave if the PCA had not pointed to it, since it did
not represent one of the famous peaks. Similarly, Brown, Marsh, and Smith
(1979) found the same results using a PCA-ANOVA treatment of data
reported earlier using ERP correlations (Brown, Marsh, & Smith, 1976).
A simpler method of comparing the overall wave shape that would be
useful for neurolinguistic investigations is a straight correlational metric.
For example, Brown, Marsh, and Smith (1973, 1976) found that the Pear-
son correlation of the ERPs for noun and verb forms of ambiguous words
was higher over the right hemisphere than over the left hemisphere, and
that this difference was especially found for the anterior sites. The beauty
of this technique is that we do not have to interpret the specifics of the
wave characteristics, only that the waves resemble each other across con-
ditions over some sites and not over others. Refinement of this method
involves restricting the time frame for the calculation of the ERP.
Is a Mental Chronometry Possible?

One of the benefits of the ERP paradigm in cognitive electrophysi-
ology is that it allows us to examine ongoing processing. Does this really
allow a "mental chronometry," a specification of what information is
being processed when? Or barring that, does it at least allow us to say
something about the timing of the thinking, as opposed to the response?
An important demonstration that this may be the case is the classic study
of McCarthy and Donchin (1981), where it was shown that the P300 is not
simply a reflection of the impending motor response. That is, the P300
peak reflects in a systematic way something other than the preparation or
execution of the response. Donchin et a1. showed this by demonstrating
that the latency of the P300 peak is not correlated with the RT response,
and thus could not be simply due to some preparation to respond. This is
important because, without being assured of this separation, we could
conclude that the ERP did not reflect cognitive processes at all, but only
motor responses. This would have been easy to accept since movement
effects are very strong in the EEG record. Since, however, the P300 seems
to reflect something in the decision-making process, we can entertain the
possibility that the peaks in general reflect important steps in the cog-
nitive requirements of the tasks.
Now, if we find that a change in processing requirement causes a
prominent peak to lag or lead, we may wish to conclude that the process-
ing change occurs at this point in the task. This is a reasonable hypothesis
to form from such data, but are the data sufficient to conclude this? Clear-
ly the processing must occur at some time before the evidence of it ap-
pears in the ERP. The difficulty is that we do not know, and perhaps
cannot know, how much before. Thus, a mental chronometry is possible
only in terms of "at least by the point" constructs.
More elaborate mental chronometries can be constructed from the
PCA-ANDVA technique, since many research questions can be asked at
the same time of the same data, each with a timing figure. For example,
Segalowitz, Menna, and MacGregor (1987) presented the results of adult
subjects' making matching judgments on pairs of words, which either (1)
look similar and rhyme (e.g., look, book), (2) look similar and do not
rhyme (e.g., paid, said), (3) rhyme but do not look similar (e.g., make,
ache), or (4) do not rhyme or look similar (e.g., mind, wall), a paradigm
used by Polich, McCarthy, Wang, and Donchin (1983). A factor peaking at
272 msec indicates that the left hemisphere has reacted differentially to
the rhyme confusion list (list 3) compared with the two control lists (lists
1 and 4), whereas the right hemisphere does not. A factor at 370 msec
shows that the left hemisphere is clearly differentiating both confusion
lists (2 and 3) from the controls, while the right hemisphere is only differ-
entiating the visual confusion list (list 2). By 490 msec, both hemispheres
are sensitive to the two confusion lists. This simple story would be fine if
we could be confident that these times really reflected the decision point
in the mental progression in the task. Because of the difficulties men-
tioned earlier in verifying the objective reality of the component factors, it
is important to replicate these sorts of specific findings. The use of the
ERP technique to create a mental chronometry would certainly be a break-
through for cognitive electrophysiology. Whether the techniques outlined
will prove useful in the end is an empirical question.
There are, however, built-in limitations that should be acknowledged
at the outset. First of all, the generalizability of the results across laborato-
ries will always be difficult not because of instability of the results but
because of differing amplification and filtering systems. Each set of ampli-
fiers has to have some setting for the time constant (Duncan-Johnson &
Donchin, 1981), and this time constant effectively sets the lower limit of
the frequency response. Similarly, each laboratory must decide on the set
of signal filters that it will use, these usually being analogue filters built
into the amplifier system. Analogue filters, however, cause phase shifting.
For example, the peaks of slow waves can be delayed by half a cycle with
standard 30-Hz low pass filters, which can translate into as much as 50
msec or more. This shift can be easily measured for a pure signal, say a 10-
Hz sine wave, but is unpredictable for an irregular wave such as an EEG
signal. Since most laboratories use the upper cutoff of 30 Hz and a lower
cutoff of under 1 Hz, there may be fairly good correspondence in the
literature, but it does not mean that this will hold for all laboratories (cf.
Kramer & Donchin, 1987).
A second difficulty with obtaining a fine chronometry is the inherent
limitation of the filtered ERP signal. The usual 30-Hz band pass is used to
reduce the higher frequencies that stem from muscle artifacts, and indeed,
EEG signals filtered at 30 Hz are considerably cleaner to view. The diffi-
culty is that at 30-Hz filtering, defining a peak takes 33 msec, which
means that peaks cannot be fully resolved in a time period shorter than
this, although with averaging, apparently higher frequencies can be ob-
tained. Is 33 msec a long or a short time for cognitive processing? This
clearly depends on the processing we are concerned with, but we can
easily think of cases where much can be done in this time period. It is not
surprising, then, that the major ERP peaks that have been tied to cognitive
processing are separated by at least 100 msec. The technique as it is
currently used limits the temporal resolution.
Is Cortical Localization Possible?

Can we localize cognitive processes using electrical signals? This is
still a controversial issue. Gevins, Bressler, and Illes (1988), for example,
recently argued that misleading conclusions are drawn from localizing
attempts when too few electrodes are used. His illustrations involved at
least 128 electrode sites. One can use mathematical interpolation to help
localize the source of signal changes (Duffy, Burchfiel, & Lombroso, 1979).
This is dependent, of course, on the degree of detail of the input. But what
if we have far fewer sites than are needed for mathematical interpolation?
The research question, in such a case, usually focuses on either the issue
of anterior versus posterior location or that of left versus right hemi-
sphere. There are cases of each where we can suspect difficulty with such
localization. The recording at the scalp may be driven by a generator at
some distance if the distant brain surface is oriented toward the recording
site and the signal is well localized within that brain surface. For exam-
ple, if care is not taken, the electrical response a visual stimulus delivered
to the left hemisphere via the right visual half-field will be greater over the
right hemisphere, especially if recorded at the 01 and 02 sites (Wood,
1982). This is because the visual signal registers in the mesial section of
the occipital lobe, and the more centrally located the stimulus, the strong-
er this effect. One solution is to display stimuli laterally enough to have
them represented on the lateral occipital surface; another is not to record
occipitally when using visual stimuli. For example, Neville, Kutas, and
Schmidt (1982a) find ERP asymmetries in temporal and frontal regions
that are independent of the visual field of stimulus presentation. There
have been no suggestions, though, that this problem of hemisphere differ-
entiation arises from any cognitive activity outside of visual perception,
although presumably mesial primary sensory cortex of the parietal lobe
would produce similar results.
A similar situation holds for the N100 component of the ERP ob-
tained in an auditory attention task. This negative wave is found max-
imally over the frontocentral areas of the scalp. Yet it has been shown to
be generated in the auditory cortex of the temporal lobe (Scherg & Von
Craman, 1986). Notice that once again we are dealing with primary senso-
ry cortex.
The Problem of Electrode Reference

Another issue reputedly leading to difficulty in localizing hemi-
spheric asymmetries stems from the question of appropriate reference
electrode. This problem has an interesting history, and no commonly
acceptable solution. Katznelson (1981) has argued that using the usual
linked-ear reference sites reduces lateral asymmetries electrically because
of an equalizing of the potentials at the ears, and presumably linked
mastoids produces a similar situation. Thus, if one temporal lobe is gener-
ating a signal, linking the ears will reduce this asymmetry at the scalp.
Alternative reference sites can be complex (such as using a form of aver-
aged site; Hjorth, 1982) or simple (such as using Cz), but then introduce
vagaries of their own. The latter site itself is very highly responsive to
cognitive processing and therefore is perhaps not a good reference loca-
tion; the former method produces a result that is dependent on the partic-
ular electrode configuration used. Since different laboratories insist on
using different site configurations, this is problematical. The original
problem may be overstated by Katznelson, though. There have been many
studies finding lateral asymmetries with linked ears or mastoids as refer-
ences (e.g., Brown et a1., 1973, 1976, discussed below; see Kutas & Van
Petten, 1988; Molfese, 1983, for reviews). If Katznelson is correct, these
findings are interesting for neurolinguistics in that they may reflect a
relative lack of generators in the lateral temporal lobes, especially for
linguistic material.
ERPs AND ADVANCES IN NEUROUNGUISTICS 69
In summary, it may be that any attempt to localize cognitive compo-

nents with only a few electrode sites will be hampered by various factors.
However, simple laterality questions may be addressable still with mod-
est ERP paradigms. Clearly, however, the trend will be toward more elabo-
rate methods, with more elaborate data analysis.
ERPs AND SOME RESEARCH ISSUES IN NEUROLINGUISTICS
Neurolinguistics and psycholinguistics have burgeoned as fields of

research in the past two decades. It is now well-nigh impossible to keep
up with the books in the field, not to mention the proliferating journals.
Many advances have been made, but some questions remain, perhaps
clarified and sharpened by the attention they have received, but nonethe-
less available for further clarification. Some of the issues, on the other
hand, seem to be unresolvable, given traditional research paradigms, and
may have become more philosophical than empirical. The semantics/
syntax distinction falls into this position. However, the new ERP tech-
nology will allow us to open up some of these questions again for em-
pirical study in new ways and may help us sharpen once again the deep
questions being posed. In this section, I will address several research
areas in neurolinguistics that may benefit or have benefited from exam-
ination in the ERP paradigm.
Lexical Semantics
There has been a considerable increase in interest in the last few
years on the issue of lexical semantics, which for the purposes of this
chapter will be defined as the network of meaning relationships that is
mediated through the linguistic system (d. Job & Sartori, 1988). The in-
crease in interest stems no doubt from both cognitive and neuropsycholo-
gical sources. The development of semantic network systems over the
past two decades (e.g., Collins & Loftus, 1975; Johnson-Laird, Hermann, &
Chaffin, 1984) has spurred hope of understanding the structure of the web
of meaning in the human mind. Until recently, however, the neurolin-
guistic issue remained fairly simple: Linguistic semantics is entirely a
function of the left hemisphere, and the primary issue was whether or not
there were semantic-syntactic disruptions in anterior versus posterior
forms of aphasia (Zurif & Caramazza, 1976). Brown et a1. (1973, 1976), for
example, as mentioned earlier found that when subjects listen to a word
that could be interpreted as a noun or a verb depending on the context
(Le., fire, duck, 1ead/led), ERPs over the left hemisphere, and the anterior
presentation especially, differentiated the two grammatical forms.
More recently, however, patients with right-hemisphere damage have
been documented to have subtle communicative dysfunctions that could

be seen to involve lexical semantics. For example, Brownell, Potter,
Michelow, and Gardner (1984) examined the sensitivity of left- and right-
brain-damaged patients to connotative versus denotative facets of a
word's meaning. They found a double dissociation linking the right hemi-
sphere to connotation. Similarly, the right hemisphere has been linked in
clinical studies to certain aspects of verbal humor (Brownell, Michel,
Powelson, & Gardner, 1983), difficulty with negative sentences (D'Urso,
Denes, Testa, & Senenza, 1986), difficulty with logicogrammatioal rela-
tions (Hier & Kaplan, 1980), difficulty with organizing sentences into
coherent paragraphs (Delis, Wapner, Gardner, & Moses, 1983), difficulty
with two-term series problems, such as "Tom is taller than Fred. Who is
shorter?" (Caramazza, Gordon, Zurif, & DeLuca, 1976), and difficulty in
the acquisition of the meaning of new lexical items from context (Gross-
man & Carey, 1978). It would be interesting to see if ERP evidence of these
processes patterns itself similarly since one can imagine complex interac-
tions between the intact hemispheres such that the asymmetry model
suggested above is not adequate (Gevins, 1986).
Similarly, the ERP paradigm may prove useful as a way of examining
the semantic networks involved in the issues of polysemy, semantic prim-
ing, and activation. For example, Burgess and Simpson (1988) have pro-
posed an interesting neurolinguistic hypothesis. They suggest that when a
subject reads an ambiguous word, both major and minor meanings are
invoked but the minor meaning is quickly suppressed. They then show by
means of the visual half-field paradigm that this pattern is true for the left
hemisphere only, and that the right hemisphere indeed promotes the
subordinate meaning of the word, which they interpret as akin to the link
between metaphorical thinking and right-hemisphere processing. Van
Petten and Kutas (1987) conducted an analogous study using ERPs and
conclude on the contrary that both meanings are not activated at the same
time. Unfortunately, they did not utilize posterior lateral recording sites
and so did not examine the lateralization hypothesis. Clearly, however, it
would be interesting to examine the Burgess et al. finding more closely
with an ERP paradigm that allows for more normal reading processes.
Semantics versus Syntax

Separating semantics from syntax has been a long-standing contro-
versy in linguistics (Caplan, 1987; Crystal, 1988). The neurolinguistic
aspect of this difficulty is seen in the controversy over whether it is
possible for some aphasics to have a semantic deficit without concomitant
syntactic losses, and vice versa. There are many aspects to this problem,
and the traditional difficulty is that behavioral paradigms have usually
not simultaneously addressed the semantic and syntactic components of

the task. For example, Zurif and Caramazza (1976) concluded that there
are grammatical competence disruptions in frontal aphasics; however,
they do not claim to have assessed semantic comprehension. Similarly, in
an ERP study, Brown et al. (1973, 1976) do demonstrate that the anterior
areas of the brain are more sensitive to the distinction between verb and
noun, but the task does not separate out semantic processing. However,
Kutas and Hillyard (1983) have provided direct ERP evidence for some
separation by comparing the N400 component to semantic versus syntac-
tic anomalies (Le., stimuli representing an inappropriate semantic catego-
ry versus incorrect verb tense or number marking). They found that se-
mantic anomalies produce the N400 deflection whereas the syntactic ones
do not. Moreover, the degree of deflection varies with the unexpectedness
of the anomaly (Kutas & Hillyard, 1984). They also found (in the 1983
study) that the open- versus closed-class distinction was associated with
the more anterior sites. Garnsey (1985) also found that the ERPs were
sensitive to open- versus closed-class words but not RT.
Reading in Normals
Can we discern component processes in reading and use ERPs to
define them, time-track them, and localize them cerebrally? Current theo-
ries of reading focus on English, probably because most of the researchers
in the field work in English-speaking countries (e.g., Coltheart, 1987; Colt-
heart, Patterson, & Marshall, 1980; Henderson, 1984). The models of read-
ing English almost invariably include two routes: a phonologically based
route, which involves the grapheme-phoneme correspondence (GPC)
rules, and a so-called visual route, which makes use of visual information.
The need for the two routes in English is clear (Patterson, 1982). The
reading of new words is clearly first attempted by "sounding out," if the
spelling seems to be regular, and often pronunciation errors are based on
spelling-pronunciations. Visual information is also clearly used since En-
glish has many homophones that have distinct spellings (e.g., seas, seize,
sees). Both these situations obtain in French as well (Le., new words
pronounced by GPC analogies to known words, and the existence of hom-
onyms such as il LIE, le lit, je lis). The mapping of spelling onto sound and
vice versa differs in interesting ways, however, and so the psycholinguis-
tic processes in reading may differ across the two languages (N. S.
Segalowitz, 1986). Languages with more regular GPC, such as Spanish,
Italian, or Estonian, again may involve the reader with a different set of
cognitive strategies. These interlanguage differences are only beginning to
be studied (Patterson, Marshall, & Coltheart, 1985).
In any case, the existence of these two routes for English is well
established in the neurolinguistic literature, especially in the acquired

dyslexia domain (Marshall, 1985; Marshall & Newcombe, 1973). Some
have stated that whereas the phonological route must be left-hemisphere-
based, the visual route may involve right-hemisphere processes (Colt-
heart, 1980), although there is actually very little or no empirical support
for this neurolinguistic hypothesis. Visual half-field studies (Chiarello,
1985; Hatta, 1977, 1981; Leong, Wong, Wong, & Hiscock, 1985; Sasanuma,
Itoh, Kobayashi, & Mori, 1980; Sasanuma, Itoh, Mori, & Kobayashi, 1977)
suggest that the hemispheres may differ on the processes involved. The
ERP paradigm enables us to examine ongoing processing with more nor-
mal reading (centrally presented, unlimited viewing). So, for example,
Kramer and Donchin (1987) found that in a word-matching paradigm, a
specific ERP component (N200) appears in response to an orthographic
mismatch whether the subject is attending to orthography or not, while an
N200 is elicited by a phonological mismatch only if phonology is task-
relevant. That is, orthographic analysis is automatic while phonological is
not. This may be relevant to the issue of pre- versus postlexical phonology
in reading English.
Using this approach to examine the hemispheric specialization issue
directly, Segalowitz et a1. (1987) have found that with both nonword
matching of letter strings and reading of real words, leftlright-hemisphere
asymmetries can be manipulated by varying the decoding route primarily
used. This was done with nonword letter string matching (same/different
judgment) by making the letter strings either pronounceable or unpro-
nounceable. The pronounceable letter strings encouraged a phonological
coding strategy, which is probably the more natural one for fluent readers
(in this case, university undergraduates). The unpronounceable letter
string discourages the phonological coding, and the subjects reported a
visual strategy. The N1-P3 amplitude shift was significantly greater over
the left hemisphere for the pronounceable lists and over the right hemi-
sphere for the unpronounceable lists (see Figure 3).
Similarly, they asked university undergraduates to perform either a
rhyme judgment or a visual-similarity judgment (as in Polich et a1., 1983;
Kramer & Donchin, 1987, discussed above) on word pairs that either
rhymed and/or looked similar (i.e., looked as if they should rhyme). They
compared the general amplitude shift of the averaged ERPs of the control
nonrhyming, nonvisually similar list with the others. The visually confus-
ing list (i.e., the ones that look similar without rhyming, such as said and
paid) differentiated from the control list in sites over the right hemi-
sphere, while the phonologically confusing list (e.g., ache and make)
differentiated from the control list only after the left hemisphere. Thus,
there appears to be support for different hemisphere sensitivities to the
processes involved in the two decoding routes. No doubt the picture will
be much more complex when considering reading in context.
RIGHT HEMISPHERE
,
FIGURE 3. Averaged ERPs from left- and right- ,,
hemisphere sites taken to the second of a pair of LEFT HEM I SPHERE
letter strings (Segalowitz, Menna, & MacGregor,
1987). Subjects had to indicate by button press
whether the two letter strings, which were either
pronounceable (---) or un-pronounceable (-J, were
the same or different. These averages are for the
"same" trials only, and illustrate the slight but con-
sistent hemisphere by stimulus list interaction. 1;---+----+---+-----+---;j,5bo MSEC
Development of Reading Skills

Perhaps it is a bit too cynical to suggest that children learn to read in
spite of the pedagogical technique currently in vogue at the time they
enter public school, but one is tempted to look for some cognitive strat-
egies that children apply to the task whether the curriculum recommends
"whole-word" or "phonic" methods. This controversy as to the most
efficient technique for teaching reading may pertain more to the teaching
of English than other languages. Indeed, the number of children who
experience difficulty in phonic decoding is considerably higher in En-
glish-speaking countries than in those with more phonically regular writ-
ing systems, although it is hard to attribute the difference solely to the
writing system since there are also differences in SES, cultural attitudes
toward schooling, age at school entry, and so on (Taylor & Taylor, 1983).
The acquisitions of English, then, may involve a subtle balance between
the two decoding routes, as many specialists in the field have suggested
(e.g., Boder, 1973; Ellis, 1984; Mitterer, 1982). It would be interesting to
try to define these processes with ERP measurement techniques in chil-
dren using paradigms similar to those described above, and to explore
developmental shifts in how these processes are distributed.
Licht, Bakker, Kok, and Bouma (1988), for example, have collected
ERPs from a group of children for 4 consecutive years beginning in kin-
dergarten. The ERPs were to reading words in Dutch. The children were
taught explicitly with a whole-word method initially, and proceeded to
more phonic strategies later. The researchers performed a regression of
ERP variables onto reading performance scores at each age group. The
finding relevant to the present discussion is that as the children got older,
the amount of variance accounted for by the left temporal site increased
dramatically while that for the parietal sites, especially the right, de-
creased. Thus, there is support, circumstantially at least, for the neu-
rolinguistic right-to-left shift for reading. By the hypothesis given here,
this relationship may not hold for those languages where phonics is the
sole strategy used.
These researchers, however, suggest that the important issue here is
not the application of GPC rules but of automaticity. Many theorists (Ellis,
1984) suggest that beginning readers first decode words element by ele-
ment and gradually automatize the process so that larger groupings are
apprehended. This chunking is necessary, of course, for rapid, fluent
reading. Remaining research questions include, however, whether the
advanced reader truly abandons the more elemental process and whether
the automatized process represents a shift in cerebral functioning. Bakker
and Licht (1986), for instance, report that normally, beginning reading
children show a shift in greater ERP amplitude to reading words from the
right hemisphere to the left hemisphere. Poor readers who remain slow at
decoding seem to be stuck at the first element-by-element stage and do not
show such a shift. Poor readers who read rapidly with many mistakes-
that is, who seem to move to automatized reading before they have built
up the processing chunks adequately-make this ERP amplitude shift
sooner than normal. If these results are reliable, the neurolinguistic im-
plications for reading and for automatized processes in general are con-
siderable with pedagogical implications (Bakker, 1984).
Language Development
We expect that language acquisition, as a phonological guessing
game, is primarily left-hemisphere-based. Molfese, Freeman, and Palermo
(1975), in their classic study, found ERP hemisphere differentiation of
speech sounds versus musical sounds in young infants. Many similar
findings (Molfese & Betz, 1988; S. J. Segalowitz, 1983) have since rein-
forced the view that hemisphere specialization is congenital in most chil-
dren and is not gradually developed during childhood, as was formerly
thought (Lenneberg, 1967). ERPs have proved very useful for cognitive
studies with young infants because there are few behavioral paradigms
that will tap processing at such an early stage, especially in the auditory
modality. Molfese and Betz list 11 ERP studies with infants demonstrating
cerebral asymmetries. They then go on to describe their line of studies
that focus on the mechanisms of such early functional asymmetries, again
illustrating how the ERP paradigm is well suited for this research ques-
tion. For example, they have shown not only that certain stop consonant
distinctions are made from birth onward but that such data can be used to
predict later delay in language acquisition (Molfese & Molfese, 1985). As
well, they are using the ERP paradigm to examine the brain correlates of
word acquisition by comparing the ERP response to arbitrarily paired
nonsense syllables with a new object (Molfese et 01., 1985) and to real
known versus unknown words (Molfese, n.d.).
There are many other developmental neurolinguistic hypotheses
worthy of investigation using this technique. For example, the develop-
ment of lexical semantics is an area that has been well studied behav-
iorally over the last two decades. Meanwhile, neurolinguistic investiga-
tors have suggested that word acquisition requiring a new category
distinction may especially entail right-hemisphere mechanisms (Gross-
man, 1981; Grossman & Carey, 1978). From a developmental neu-
rolinguistic viewpoint, this is an important question, since we are still
exploring the relationship between lexical and nonlexical semantics.
Given that Molfese has shown that the ERP paradigm is sensitive to at
least some aspect of word acquisition, it would be interesting to examine
developmental data on these same parameters that were of interest in the
adult lexical semantic studies.
Second -Language Acquisition

All of the research questions in lexical semantics and language ac-
quisition have new aspects to them in studies of second-language acquisi-
tion. The study of the brain correlates of bilingualism is highly complex
(Albert & Obler, 1978; Paradis, 1983; Vaid, 1983; Vaid & Genesee, 1980),
and this is not the place to explore the details of the issues. However, it is
easy to see that many of the neurolinguistic hypotheses outlined in vari-
ous reviews (such as Vaid, 1983) are highly amenable to the ERP para-
digm.
Another (admittedly more complex) twist to this issue is that of the
acquisition of English as a second language by deaf individuals. For exam-
ple, Neville, Kutas, & Schmidt (1982b) found ERP results (using the N400
paradigm) indicating a lack of asymmetry in congenitally deaf adults
reading English, compared with the usual asymmetry found in normals.
These results are fully compatible with current behavioral work examin-
ing the reading skills and strategies of deaf individuals (Gibson, 1988;
Gibson & Segalowitz, 1986).
Issues in Clinical Neurolinguistics

Research in NL with clinical populations has utilized a variety of
paradigms, deriving from various information-processing procedures
(Beaumont, 1982; Hannay, 1986; Posner, 1986). It would be interesting to
apply ERP techniques to some of these issues. There are a variety of

obstacles, of course, making this application somewhat more problematic
than for the issues outlined earlier. First of all, as with behavioral para-
digms, there is the issue of compliance and consent. It is often the case
that clinical patients are undergoing enough stress in their lives not to
need or want the further stimulation of participating in basic research. In
addition, those caring for this group may not be interested in the research
questions enough to devote their own time to such projects.
Assuming, however, that these obstacles are overcome, there are re-
search design technicalities that must be attended to. The electrical field
configuration of the head may be altered among some clinical populations
owing to physical damage to the skull. Nunez (1981) presents the physical
problem that an irregularity, such as a break, in the skull theoretically
alters the electric field properties of the scalp. However, Campbell and his
colleagues (Campbell, Deacon-Eliott, & Proulx, 1986; Campbell, Deacon-
Eliott, Suffield, & Proulx, 1987; Deacon-Eliott, Campbell, Suffield, &
Proulx, 1987) report little difficulty in working with closed head injury
(CHI) patients, although one could argue that, since they do find ERP
differences between this group and normals, it is impossible to determine
whether the effect is due to the physical alteration of the skull or to the
consequences of CHI on the brain. Presumably, this is a testable hypoth-
esis given the known condition of the subjects, using a quasi-case study
design (Dywan & Segalowitz, 1986).
Neurolinguistic hypotheses involving populations with brain damage
of whatever origin are critical to the field. For example, we may ask
whether language recovery after severe left-hemisphere damage is done in
the right hemisphere or in the residual tissue of the left hemisphere. This
has been a long-debated issue, and the controversy continues. For exam-
ple, much has been made of the residual ideogram-reading skills of some
Japanese aphasics and of the semantic substitutions made by deep dyslex-
ics to support the notion of a right-hemisphere reading capacity (Colt-
heart, 1980). To the extent that various ERP paradigms allow for lateral
localization, we could address the issue. We could look for N400 asymme-
tries (with auditory presentation) to semantic incongruities in somewhat
recovered aphasics. Papanicolaou, Levin, and Eisenberg (1984) used the
ERP probe paradigm to test this hypothesis and indeed found support for
a right-hemisphere basis for language in recovered aphasics.
Further work on subtle aspects of recovered language is possible. We
may be interested in differences in the lexical semantics of patients who
show left-hemisphere or right-hemisphere recovery of language. For ex-
ample, Chapman, Bragdon, Chapman, and McCrary (1977) found ERP
correlates of the semantic differential structure of nouns. Is this structure
altered systematically in aphasics with differing symptoms (e.g., fluent vs.
nonfiuent), in dementia patients, or in CHI patients who seem to have

difficulty appreciating complexities in the world?
SUMMARY
The field of neurolinguistics is already well established and highly

active. Many of the research questions derive from the clinic's using be-
havioral paradigms that have some inherent limitations. This chapter
presents an outline of some of the advantages and disadvantages of the
ERP paradigm as an alternate technology to supplement the traditional
behavioral measures, and argues that a great many of the classic and new
research questions in neurolinguistics could benefit from this paradigm.
REFERENCES
Albert, M., & abler, 1. (1978). The bilingual brain. New York: Academic Press.
Bakker, D. J. (1984). The brain as a dependent variable. Journal of Clinical Neuropsychology,
6, 1-16.
Bakker, D. J., & Licht, R. (1986). Learning to read: Changing horses in midstream. In G. Th.
Pavlides & D. F. Fisher (Eds.), Dyslexia: Its neuropsychology and treatment. New York:
Riley.
Beaumont, J. G. (Ed.). (1982). Divided visual field studies of cerebral organization. London:
Academic Press.
Boder, E. (1973). Developmental dyslexia: A diagnostic approach based on three atypical
reading-spelling patterns. Developmental Medicine and Child Neurology, 15,663-687.
Brown, W. S., Marsh, J. T., & Smith, J. C. (1973). Contextual meaning effects on speech-
evoked potentials. Behavioral Biology, 9, 755-761.
Brown, W. S., Marsh, J. T., & Smith, J. C. (1976). Evoked potential wave form differences
produced by the perception of different meanings of an ambiguous phrase. Electroen-
cephalography and Clinical Neurophysiology, 41, 113-123.
Brown, W. S., Marsh, J. T., & Smith, J. C. (1979). Principal component analysis of ERP
differences related to the meaning of an ambiguous word. Electroencephalography and
Clinical Neurophysiology, 46, 709-714.
Brownell, H. H., Michel, D., Powelson, J., & Gardner, H. (1983). Surprise but not coherence:
Sensitivity to verbal humor in right-hemisphere patients. Brain and Language, 18, 20-
27.
Brownell, H. H., Potter, H. H., Michelow, D., & Gardner, H. (1984). Sensitivity to lexical
denotation and connotation in brain-damaged patients: A double dissociation? Brain
and Language, 22, 253-265.
Bryden, M. P. (1982). Laterality: Functional asymmetry in the intact brain. New York:
Academic Press.
Burgess, C., & Simpson, G. B. (1988). Cerebral hemisphere mechanisms in the retrieval of
ambiguous word meanings. Brain and Language, 33, 86-103.
Callaway, E., Tueting, P., & Koslow, S. H. (1978). Event-related brain potentials in man. New
Campbell, K, Deacon-Elliott, D., & Proulx, G. (1986). Electrophysiological monitoring of
closed head injury, I. Basic principles and techniques. Cognitive Rehabilitation, Sep-
tember-October, 26-32.
Campbell, K, Deacon-Elliott, D., Suffield, J. B., & Proulx, G. B. (1987). Electrophysiological
monitoring of closed head injury, II. Multimodal sensory evoked potentials. Cognitive
Rehabilitation, January-February, 26-33.
Caplan, D. (1987). Neurolinguistics and linguistic aphasiology: An introduction. Cambridge:
Cambridge University Press.
Caramazza, A., Gordon, J., Zurif, E. B., & DeLuca, D. (1976). Right-hemispheric damage and
verbal problem solving behavior. Brain and Language, 3, 41-46.
Chapman, R M., Bragdon, H. R, Chapman, J. A., & McCrary, J. W. (1977). Semantic meaning
of words and averaged evoked potentials. In J. E. Desmedt (Ed.), Recent developments in
the psychobiology of language: The cerebral evoked potential approach. London: Ox-
ford University Press.
Chiarello, C. (1985). Hemisphere dynamics in lexical access: Automatic and controlled
priming. Brain and Language, 26, 146-172.
Collins, A. M., & Loftus, E. F. (1975). A spreading-activation theory of semantic processing.
Psychological Review, 82, 407-428.
Coltheart, M. (1980). Deep dyslexia: A right-hemisphere hypothesis. In M. Coltheart, K.
Patterson, & J. C. Marshall (Eds.), Deep dyslexia. London: Routledge and Kegan Paul.
Coltheart, M. (1987). The psychology of reading. Hillsdale, NJ: Erlbaum.
Coltheart, M., Patterson, K, & Marshall, J. C. (1980). Deep dyslexia. London: Routledge and
Kegan Paul.
Crystal, D. (1988). Linguistic levels in aphasia. In F. C. Rose, R Whurr, & M. Ryke (Eds.),
Aphasia (pp. 23-45). London: Whurr.
Deacan-Eliott, D., Campbell, K. B., Suffield, J. B., & Proulx, G. B. (1987). Electrophysiological
monitoring of closed head injury. III. Cognitive evoked potentials. Cognitive Rehabilita-
tion, 12-21.
Delis, D. C., Wapner, W., Gardner, H., & Moses, J. A. (1983). The contribution of the right
hemisphere to the organization of paragraphs. Cortex, 19, 43-50.
Donchin, E. (1966). A multivariate approach to the analysis of averaged evoked potentials.
IEEE Transactions on Biomedical Engineering, BME-13(3), 131-139.
Duffy, F. H., Burchfiel, J. L., & Lambroso, C. T. (1979). Brain electrical activity mapping
(BEAM): A method for extending the clinical utility of EEG and evoked potential data.
Annals of Neurology, 5, 309-321.
Duncan-Johnson, C. c., & Donchin, E. (1981). The time constant in P300 recording. Psycho-
physiology, 16, 53-55.
D'Urso, V., Denes, G., Testa, S., & Semenza, C. (1986). The role of the right hemisphere in
processing negative sentences in context. Neuropsychologia, 24, 289-292.
Dywan, J., & Segalowitz, S. J. (1986). The role of the case study in neuropsychological
research. In J. Val siner (Ed.), The individual subject and scientific psychology. New
York: Plenum Press.
Ellis, A. W. (1984). Reading, writing and dyslexia: A cognitive analysis. London: Erlbaum.
Friedman, D., Simson, R, Ritter, W., & Rapin, I. (1975). The late positive component (P300)
and information processing in sentences. Electroencephalography and Clinical Neu-
rophysiology, 38, 255-262.
Gaillard, A. W. K, & Ritter, W. (Eds.). (1983). Tutorials in ERP research: Endogenous compo-
nents. Amsterdam: North-Holland.
Garnsey, S. (1985). Function words and context words: Reaction time and evoked potential
measures of word recognition. Cognitive Science Technical Report URCS-29, University
of Rochester, Rochester, New York.
Gevins, A. S. (1986). Quantitative human neurophysiology. In H. J. Hannay (Ed.), Experi-
mental techniques in human neuropsychology (pp. 419-456). New York: Oxford Uni-
versity Press.
Gevins, A. S., Bressler, S. L., & Illes, J. (1988, January 27-30). Effects of incipient fatigue on
higher cortical function. Paper presented at the 16th Annual International Neuropsy-
chological Society meeting, New Orleans.
Gibson, D. (1988). The impact of early developmental history on cerebral asymmetries:
Implications for reading ability in deaf children. In D. L. Molfese & S. J. Segalowitz
(Eds.), Developmental implications of brain lateralization. New York: Guilford Press.
Gibson, C., & Segalowitz, S. J. (1986). The impact of visual-spatial information on the devel-
opment of reading proficiency in deaf children. In J. L. Nespoulous, P. Perron, & A. R.
Lecours (Eds.), The biological foundations of gestures: Motor and semiotic aspects (pp.
215-227). Hillsdale, NJ: Erlbaum.
Grossman, M. (1981). A bird is a bird is a bird: Making reference within and .without
superordinate categories. Brain and Language, 12, 313-331.
Grossman, M., & Carey, S. (1978, October). Word-learning after brain damage. Paper present-
ed to the Academy of Aphasia, Chicago.
Hannay, J. H. (1986). Experimental techniques in human neuropsychology. New York: Ox-
ford University Press.
Hatta, T. (1977). Recognition of Japanese kaiji in the left and right visual fields. Neuropsy-
chologia, 15, 605-688.
Hatta, T. (1981). Differential processing of kaiji and kana stimuli in Japanese people: Some
implications from Stroop-test results. Neuropsychologia, 19, 87-93.
Henderson, L. (1984). Orthographies and readings; Perspectives from cognitive psychology.
Hillsdale, NJ: Erlbaum.
Hier, D. B., & Kaplan, J. (1980). Verbal comprehension deficits after RH damage. Applied
Psycholinguistics, 1, 279-294.
Hjorth, B. (1982). An adaptive EEG derivation technique. Electroencephalography and
Clinical Neurophysiology, 54, 654-661.
Job, R., & Sartori, G. (1988). The cognitive neuropsychology of visual and semantic process-
ing of concepts. Cognitive Neuropsychology, 5(1).
Johnson-Laird, P. N., Hermann, D. J., & Chaffin, R. (1984). Only connections: A critique of
semantic networks. Psychological Bulletin, 96, 292-315.
Katznelson, R. D. (1981). EEG recording, electrode placement, and aspects of generator
localization. In P. L. Nunez, Electrical fields of the brain. New York: Oxford University
Press.
Kramer, A. F., & Donchin, E. (1987). Brain potentials as indices of orthographiC and pho-
nological interaction during word matching. Journal of Experimental Psychology;
Learning, Memory and Cognition, 13, 76-86.
Kutas, M., & Hillyard, S. A. (1983). Event-related brain potentials to grammatical errors and
semantic anomolies. Memory and Cognition, 11, 539-550.
Kutas, M., & Hillyard, S. A. (1984). Brain potentials during reading reflect word expectancy
and semantic association. Nature, 307, 161-163.
Kutas, M., & Van Petten, C. (1988). Event-related brain potential studies of language. In P. K.
Ackles, J. R. Jennings, & M. G. H. Close (Eds.), Advances in psychophysiology (139-187).
Greenwich, CT: JAI Press.
Lecours, A. R., Lhermitte, F., & Bryans, B. (1983). Aphasiology. London: Balliere Tindall.
Lenneberg, E. H. (1967). Biological foundations of language. New York: Wiley.
Leong, C. K., Wong, S., Wong, A., & Hiscock, M. (1985). Differential cerebral involvement in
perceiving Chinese characters: Levels of processing approach. Brain and language, 26,
131-145.
Licht, R., Bakker, D. J., Kok, A., & Bouma, A. (1988). The development of lateral event-related
potentials (ERPs) related to word meaning: A four year longitudinal study. Neuropsy-
chologia, 26, 327-340.
Marshall, J. C. (1985). On some relationships between acquired and developmental dyslexia.
In F. H. Duffy & N. Geschwind (Eds.), Dyslexia: A neuroscientific approach to clinical
evaluation. Boston: Little, Brown.
Marshall, J. C., & Newcombe, F. (1973). Patterns of paralexia: A psycholinguistic approach.
Journal of Psycholinguistic Research, 2, 175-199.
McCarthy, G., & Donchin, E. (1981). A metric for thought: A comparison of P300 latency and
reaction time. Science, 211, 77-80.
Mitterer, J. O. (1982). There are at least two kinds of poor readers: Whole-word poor readers
and recoding poor readers. Canadian Journal of Psychology, 36, 445-461.
Mocks, J. (1986). The influence of latency jitter in principal component analysis of event-
related potentials. Psychophysiology, 23, 480-484.
Mocks, J., & Verleger, R. (1986). Principal component analysis of event-related potentials: A
note on misallocation of variance. Electroencephalography and Clinical Neurophysi-
ology, 65, 393-398.
Molfese, D. L. (1978). Left and right hemisphere involvement in speech perception: Elec-
trophysiological correlates. Perception and Psychophysics, 23, 237-243.
Molfese, D. M. (1983). Event related potentials and language processes. In A. W. K. Gaillard
& W. Ritter (Eds.), Tutorials in ERP research: Endogenous components (pp. 345-367).
Amsterdam: North Holland.
Molfese, D. M. (n.d.). Electrophysiological correlates of word meanings in 14-month-old
human infants. (Mimeo)
Molfese, D. M., & Betz, J. C. (1988). Electrophysiological indices of the early development of
lateralization for language and cognition and their implications for predicting later
development. In D. M. Molfese & S. J. Segalowitz (Eds.), Developmental implications of
brain lateralization (pp. 17-190). New York: Guilford Press.
Molfese, D., Freeman, R. B., & Palermo, D. S. (1975). The ontogeny of brain lateralization for
speech and non-speech stimuli. Brain and Language, 2, 356-368.
Molfese, D. L., & Molfese, V. J. (1985). Electrophysiological indices of auditory discrimina-
tion in newborn infants: The bases for predicting later language development? Infant
Behavior and Development, 8, 197-211.
Molfese, D. M., Wetzel, W. F., Linnville, S. E., Imbasciate, C., Leicht, D., Courtney, C.,
Baldwin, K., & Adams, C. A. (1985). Word recognition in 16-month-old infants: EJec-
trophysiological indices. Paper presented at the 57th annual meetings of the Mid-
western Psychological Association.
Neville, H. J., Kutas, M., & Schmidt, A. (1982a). Event-related potential studies of cerebral
specialization during reading. I. Studies of normal adults. Brain and Language, 16, 300-
315.
Neville, H. J., Kutas, M., & Schmidt, A. (1982b). Event-related potential studies of cerebral
specialization during reading. II. Studies of congenitally deaf adults. Brain and Lan-
guage, 16, 316-337.
Nunez, P. L. (1981). Electric fields of the brain. New York: Oxford University Press.
Papanicolaou, A. C., Levin, H. S., & Eisenbeg, H. (1984). Evoked potential correlates of
recovery from aphasia after focal left hemisphere injury in adults. Neurosurgery, 14,
412-415.
Paradis, M. (Ed.). (1983). Readings on aphasia in bilinguals and polyglots. Montreal: Didier.
Patterson, K. E. (1982). The relation between reading and phonological coding: Further
neuropsychological observations. In A. W. Ellis (Ed.), Normality and pathology in cog-
nitive functions (pp. 77-111). London: Academic.
Patterson, K. E., Marshall, J. C., & Coltheart, M. (1985). Surface dyslexia: Neuropsychological
and cognitive studies of phonological reading. Hillsdale, NJ: Erlbaum.
Polich, J., McCarthy, G., Wang, W. S., & Donchin, K (1983). When words collide: Ortho-
graphic and phonological interference during word processing. Biological Psychology,
16, 155-180.
Posner, M. 1. (1986). Chronometric explorations of mind. New York: Oxford University
Press.
Sasanuma, S., Hoh, M., Kobayashi, Y., & Mori, K. (1980). The nature of the task-stimulus
interaction in the tachistoscopic recognition of kana and kanji words. Brain and Lan-
guage, 9, 298-306.
Sasanuma, S., Hoh, M., Mori, K., & Kobayashi, Y. (1977). Tachistoscopic recognition of kane
and kanji words. Neuropsychologia, 15, 547-553.
Scherg, M., & Von Cramon, D. (1986). Evoked dipole source potentials of the human auditory
cortex. Electroencephalography and Clinical Neurophysiology, 65, 344-360.
Segalowitz, N. S. (1986). Skilled reading in the second language. In J. Vaid (Ed.), Language
processing in bilinguals: Psycholinguistic and neuropsychological perspectives. Hills-
dale, NJ: Erlbaum.
Segalowitz, S. J. (1983). Cerebral asymmetries for speech in infancy. In S. J. Segalowitz (Ed.),
Language functions and brain organization. New York: Academic Press.
Segalowitz, S. J., & Cohen, H. (1989). Right hemisphere EEG sensitivity to speech. Brain and
Language. (in press).
Segalowitz, S. J., & MacGregor, L. (1987, February 19-21). ERP correlates of reading pho-
nologically regular and non-regular English words. Paper presented at the International
Neuropsychological Society meeting, Washington, DC.
Segalowitz, S. J., Menna, R., & MacGregor, L. (1987, July 1-4). Left and right hemispheric
participation in normal adults' reading: Evidence from ERPs. Paper presented at the
European meeting of the International Neuropsychological Society, Barcelona, Spain.
Taylor, 1., & Taylor, M. M. (1983). The psychology of reading. New York: Academic Press.
Vaid, J. (1983). Bilingualism and brain lateralization. In S. J. Segalowitz (Ed.), Language
functions and brain organization. New York: Academic Press.
Vaid, J., & Genesee, F. (1980). Neuropsychological approaches to bilingualism: A critical
review. Canadian Journal of Psychology, 34, 417-445.
Van Petten, C., & Kutas, M. (1987). Ambiguous words in context: An event-related potential
analysis of the time course of meaning activation. Journal of Memory and Language, 26,
188-208.
Wood, C. C. (1982). Application of dipole localization methods to source identification of
human evoked potentials. Annals of the New York Academy of Sciences, 388, 139-155.
Wood, C. C., & McCarthy, G. (1984). Principal component analysis of event-related poten-
tials: Simulation studies demonstrate misallocation of variance across components.
Electroencephalography and Clinical Neurophysiology, 59, 249-260.
Zurif, K, & Caramazza, A. (1976). Psycho linguistic structures in aphasia: Studies in syntax
and semantics. In H. Whitaker & H. A. Whitaker (Eds.), Studies in neurolinguistics (Vol.
1). New York: Academic Press.
II
Development and Dissolution of

Language
5
Disorders in Language Acquisition

and Cerebral Maturation
A Neurophysiological Perspective
LYDA MEJIA and JORGE ESLAVA-COBOS
Since Broadbent (1872) described congenital aphasia in children with

brain lesions, many authors have expanded on the same topic. Neverthe-
less, a considerable portion of the published material has centered on the
discussion of the appropriate terminology more than on the basic mecha-
nisms that account for delays or involution of language organization in
children.
After Broadbent, Coen in 1886 for the first time used the expression
oudiomutism, which since then has been employed with such different
meanings by various authors (Ajuriaguerra & Marcelli, 1982; Borel-Mais-
onny, 1979; Quiros et 01., 1971; Quiros & Della Cella, 1971) that one of
them proposed it be abandoned until a consensus could be reached as to
its exact meaning.
Referring to Broadbent's "congenital aphasia," Hadden (1891),
Ucherman (1891), and Wyllie (1894) mentioned how brain lesions can
produce congenital aphasia alone or associated with paralysis and other
symptoms. Before 1900, however, Kerr (1897) disputed this hypothesis as
he showed how "congenital aphasia" could develop without past history
of brain lesion. In the late 1920s, Worster-Drought and Allen (1929) re-
ferred to "congenital auditory imperception" and emphasized how it
could also appear in children without a past history of brain lesion but
usually with other parents affected and male preponderance. They also
LYDA MEJIA • Colombian School of Rehabilitation, University of Rosario, Bogota, Colom-

bia. JORGE ESLAVA-COBOS • Department of Child Neurology, Colombian Institute of
Neurology, Bogota, Colombia.
85
86 LYDA MEJIA and JORGE ESLAVA-COBOS
pointed out how these children usually had difficulties in learning certain
skills.
Ley (1930) defined two types of difficulties in children's language.
The first, which he considered a delay in the praxic functions in language,
is one in which the child can pronounce sounds and syllables but has
great difficulties in the synthesis of these elements, which nevertheless
still allow an adequate expression of language. Interestingly enough, he
described in this problem an alteration of abstract concepts, which he
explained as due to the fact that they are acquired through the spoken
word. The other clinical entity he called "hearing agnosia of evolution,"
characterized by impairment in the comprehension of words and of writ-
ten material, which he considered due to difficulties in abstract thinking.
Worster-Drought (1943, 1968) returned to his analysis of "congenital
auditory imperception" and further defined its characteristics as impos-
sibility for the comprehension of language, poor expression (frequently as
idioglossia), normal nonverbal IQ, and normal audiogram; some of these
patients have a discrete hearing loss but not sufficiently severe as to
explain their language difficulties. He defined four types of imperception,
ranging from difficulty to comprehension of words to imperception of all
sounds-a situation extremely difficult to differentiate from deafness. At
the same time, he described delays in development of spoken language
without any difficulty in comprehension, called by him "developmental
executive aphasia."
Separating cases in which expression was impaired from those that
affected comprehension, Morley (1957) described "developmental apha-
sia," distinguishing three subtypes of expression impairment: expression
aphasia of evolution, evolutive dysarthria, and evolutive apraxia of
expression.
In the early 1950s, Landau, Goldstein, and Kleffner (1954) described
autopsy findings in a child whose language development started at age 6
and by the moment of his death-4 years later-was quite functional. They
found extensive bilateral destruction of the perisylvian cortex in the area of
the central sulcus and retrograde degeneration in both medial geniculate
nuclei, and concluded that his language had developed through different
connections from the usual primary thalamocortical projections. In a com-
mentaryto this case, Lenneberg (1967) says: "I cite this case to illustrate the
enormous plasticity of the human brain (or the lack of cortical specializa-
tion) with respect to language during the first years of life" (p. 183).
In the mid-1950s and 1960s, the discussion frequently centered on
terminology. Many rejected the term aphasia, because they felt it implied
loss of something that children had not acquired, and preferred dyspha-
sia. Others, Ajuriaguerra et a1. (1969) among them, preferred the term
aphasia for the most serious symtoms and dysphasia for "a population
who presents a disturbance in the integration of language without sensory
LANGUAGE ACQUISITION AND CEREBRAL MATURATION 87
or phonatory disability and who can nevertheless express themselves

verbally-although with difficulty-and whose mental level can be con-
sidered normal" (p. 119). He distinguished two groups of children: one
who choose actions or gestures as means for communication, and the
other who-in spite of their difficulties-still prefer language for commu-
nication. In this second category he recognized two subgroups: (1) the
"economomensurates," who always communicate verbally, although
with difficulty, and who he felt had a better prognosis, and (2) the "imper-
tinents," children who speak abundantly with poor language and who are
not interested in being understood.
In 1971 Quiros et al. published The So-called Aphasias of Child-
hood, where after a detailed revision of the literature they concluded that
child aphasia is "a symptomatic label which tries to include all deficien-
cies of language of neurologic origin, once sensory, intellectual or psychic
possible causes have been ruled out." They then proposed their own
classification. For what had been called "sensory aphasia of childhood"
they distinguished three syndromes: "aphasoidism," "auditive agnosia,"
and "troncular integrative deafness." For "mixed aphasia of childhood"
they substituted "subcortical syndromes." Finally, for "motor aphasia of
childhood" they differentiated the syndromes of "apractognosia of child-
hood" and "vestibular-proprioceptive disorganization." For each one of
these syndromes they tried to define a localization of injury and corre-
sponding language symptoms, which resulted in a classification similar to
that of the aphasias in the adult.
Lenneberg (1967) studied sequelae of lesions suffered at different
ages and concluded that lateralized brain lesions sustained before the
integration of language is accomplished, or in its first stages, do not pro-
duce significant language impairment or sequelae, and that only when the
injury occurs after age 11 will the aphasic symptoms be irreversible.
Azcoaga, Bello, Citrinovitz, Derman, and Frutos (1981) proposed that
language abnormalities of neurologic origin, congenital or acquired before
language begins, be called "language retardations." After a profound ex-
planation of the subjacent physiopathological mechanisms, he classified
them as "anartric retardation" and "aphasic retardation," since he con-
sidered that-even though the child is in the process of acquiring lan-
guage-the symptoms and pathophysiology found in each syndrome is
equivalent to that found in the corresponding syndromes of the adult. He
further described specific symptoms depending on whether the phys-
iopathological anomaly is a disequilibrium in the activity of excitatory or
inhibitory neuronal circuits. Each syndrome is characterized according to
the different stages of development of the infant and child. Azcoaga and
his colleagues considered that when the lesion occurs after language de-
velopment has started, the clinical entity recognizable is no longer a retar-
dation but a syndrome essentially similar to that of the adult, owing to
disruption in the activity of the particular analyzer involved. This distinc-

tion has important implications for therapy.
As regards childhood aphasia due to brain lesions that occur after
language has been organized, Brock and Krieger (1966) recalled 30 cases
published by Guttman (1942), including patients with abscess, trauma,
tumors, and thrombosis, all between 2 and 14 years of age. In left-hemi-
sphere lesions, aphasia was as frequent as in adults; one case with bilat-
era I lesions also had aphasia, while of 13 patients with right-hemisphere
lesions only 1 had aphasia. Temporal lesions produced difficulties in
both comprehension and expression; frontal lobe lesions only impaired
expression. Guttman considered that children with global aphasia recover
slowly and are left with permanent sequelae.
Landau and Kleffner (1957) published the association of acquired
aphasia with epilepsy-usually of the partial type (or in some patients
only with the EEG expression). This syndrome has since been extensively
studied (Billard et 01., 1981; Bishop, 1985; Brissaud & Richardet, 1974;
Deonna, Beaumanoir, Gaillard, & AssaI, 1977; Duloc, Billard, & Arthuis,
1983); language symptoms have been found to be extremely variable,
which has induced some to consider it a heterogenous syndrome.
ORGANIC AND FUNCTIONAL SYNDROMES
The foregoing review of the literature presents us with, among other

things, an important confusion in terms and concepts. But it could be no
other way if we remember that we are but heirs of a dualistic conception
that up until today dominates our literature with the distinction between
the "organic" and the "functional" (psychogenic, mental) (Eccles, 1970).
We have advanced to an understanding (or at least an intuition) that the
"functional" demands an "organic housing" (Popper, 1977), but not far
enough to accept that neuronal tissue and environment form one complex
and inseparable unit when the explanation of psychism is attempted.
Thus, neural structure in any given moment-understood in its full mor-
phological, chemical, and electrical aspects-is the sole determinant of
the different "constituents" of psychism, language included. But at the
same time, the afferent bioelectric input imposed on us by environmental
signals-among them those that as feedback come back from the same
neuropsychologic expression-are fundamental influences for the modi-
fication of that same structure (Horwitz, 1981; McLennan & Hendry, 1981;
Meisami & Mousavi, 1982; Mower, Burchfield, & Duffy, 1981). This is so
because embryologic development of the human brain is not independent
of environmental influences (Cowan, 1979). Right from its very begin-
nings in the induction of the differentiation of a neural plaque on the
ectoderm of the embryo (through the influence of substances that proba-

bly originate in the subjacent mesoderm) through well into adult life
when myelination is completed, many of the processes involved are un-
der the influence of signals that arise from the environment-inside or
outside the body.
We consider the structure as the neural elements (macroscopic, cel-
ular, subcellular, and chemical) and their interrelationships whose func-
tional expressions are the different constituents of psychism. At the same
time, we consider that structure as the result-always in a dynamic
changing process-of the embryologic development of the nervous sys-
tem, which depends on individual genetic conditioning as much as on
environmental influences.
Health can be understood as the "relatively stable state" in the course
of neural processes that results from the presence of "usual" structural
elements and also "usual" interrelationships between them (physiologic
stable state). Sickness (or disability) instead is the "relatively stable state"
in which structural elements or their interrelationships are "erroneous"
(pathologic stable state) (Bejtereva, 1984).
We shall therefore discuss initially the embryologic development of
the nervous system and then consider (only in reference to language) the
physiologic stable state in its evolution (language development in healthy
children). Finally, we will proceed to a discussion of the pathologic stable
states (language retardations and involution), including the acquisition of
reading and writing skills and their abnormalities, also from the perspec-
tive of stable states.
BASIC EMBRYOLOGIC CONSIDERATIONS
Volpe (1977) has made an excellent summary of neural embryo-

genesis. Dorsal induction spans from 3 to 4 gestational weeks; then comes
ventral induction from 5 to 6 weeks, followed by neuronal proliferation
from 2 to 4 months, then migration from 3 to 5 months, organization from
6 months to "years" postnatal, and, finally, myelination, which prolongs
itself at least until the third decade of extrauterine life.
The events described are not exactly limited to that timetable, and not
infrequently some overlapping is observed; nevertheless, it is valid as
well as convenient to consider the general process of maturation in terms
of a sequence of individual events. It should be clearly noted that birth is
not shown as a critical borderline because, in the process of neural em-
bryogenesis, birth is but an accident whose only significance is that it
influences the characteristics of environmental information.
Around the third gestational week a specialized tissue differentiates

on the ectoderm (the neural plaque), which then folds back until a neural
tube is formed. This is the origin of the whole nervous system from telen-
cephalon to spinal level Ll. This process-the first in neural embryo-
genesis-is called dorsal induction. Abnormalities in this stage give rise
to neural tube closure defects.
Two to 3 weeks later, the rostral portion of this neural tube estab-
lishes close interrelationships with the subjacent ecto- and mesoderm,
giving rise to craniofacial architecture in the process known as ventral
induction. Cerebral hemispheres, diencephalon, thalami, olfactory bulbs
and tracts, eyes and optic nerves, and some structures of the posterior
fossa are thus formed. It must be remembered that the architectonic ele-
ments with which these structures are formed are neural tissue; for this
reason, abnormalities in this stage will carry varying degrees of neurologic
sequelae (DeMyer, Zeman, & Palmer, 1964), ranging from severe malfor-
mations that invariably lead to death (e.g., alobar holoprosencephaly with
cyclopia) to subtle neuropsychologic defects in some minor fasciotelen-
cephalic malformations.
In the third stage of embryogenesis-proliferation-the primitive
periependymal neuroblasts must multiply themselves to the definite
number of neurons that the individual will have for the rest of his life,
which Eccles has calculated as approximately 10,000 million. Neverthe-
less, a considerable individual variability exists which has been difficult
to evaluate clinically in its full physiologic or pathologic significance due
to the impossibility of quantitating neuronal populations. Thus, in mega-
lencephalies (DeMyer, 1972) descriptions range from profound mental
retardation to the genius of Lord Byron (whose brain weighed 1800 gr).
The fourth event-migration-refers to the extraordinarily complex
phenomenon whereby thousands of millions of nerve cells move from
their sites of origin in the ventricular and subventricular zones to the loci
within the central nervous system where they will reside for life (Sidman
& Rakic, 1973). According to Volpe, this remarkable phenomenon is
bound to suffer some imperfections in every human being. In conse-
quence, minor alterations called neuronal heterotopias are probably pre-
sent to some degree in all of us. On the other hand, neuropsychological
defects have been attributed to these same migration "imperfections"
(Galaburda & Kemper, 1979). Abnormalities of the migration process can
therefore be understood as a continuum ranging from the most severe
forms (schizencephaly, lissencephaly) to subtle defects that produce dis-
crete neuropsychological deficits that impair learning or that may even
pass undetected in "normal" subjects, perhaps contributing in all of us to
our peculiar "neuropsychological imperfections."
The fifth event in neural embryogenesis is neural organization. Its
main developmental features include proper alignment, orientation and

layering of cortical neurons, elaboration of dendritic and axonal ramifica-
tions, establishment of synaptic contacts, and proliferation and differ-
entiation of glia. Quite evidently, the proliferation of synaptic contacts
sets the morphologic substrate for the future enrichment of neuronal cir-
cuits (Purpura, 1976), which in turn will be the physiologic substrate of
the different "analyzers" (Azcoaga et al., 1983). This stage is highly de-
pendent on environmental stimulation (Chang & Greenough, 1982; Kaye,
Mitchell, & Cynader, 1982) in such a way that intensely stimulated neu-
ronal circuits will enrich the number, volume, and electric functional
status (Horwitz, 1981) of their synapses, while those poorly stimulated
will suffer atrophy (Meisami & Mousavi, 1982), maybe even to the point of
functional disappearance of that circuit.
The last stage is myelination. In it, nude axons-both intra- and ex-
traaxial-gradually cover themselves with a myelin sheath that hastens
and improves the efficiency of transmission of membrane potentials, with
the consequent enhancement of the richness and harmony of the func-
tional expression of the activity of neuronal circuits. Myelination starts in
the motor roots of the peripheral nervous system, then proceeds with
intra- and extraaxial afferent systems, then with efferent intraaxial path-
ways, and finally-in a process that prolongs itself at least through the
third decade-with intra- and interhemispheric association fibers (Ya-
kovlev & Lecours, 1967). No abnormalities have yet been clearly defined
in the process of myelination in the human; in contrast, disease entities in
which myelin is destroyed or otherwise injured-the leukodystrophies
(Swaiman & Wright, 1982)-are characterized by severe progressive de-
rangement of neuropsychologic abilities-language included.
At some as yet ill-defined time in embryogenesis, an apparently
important amount of cell death occurs in some neuronal populations
(Levi-Montalcini, 1964; Rabinowicz, 1976). The role of this phenomenon
has not been clearly established, but it is tempting to speculate that it
might contribute to the "minimization of brain areas necessary and suffi-
cient to sustain a well formed and developed function that will not devel-
op further" (Bejtereva, 1984, pp. 66-67). This is how Bejtereva explain
the loss of potentiality in a given analyzer that at the same time both
stabilizes and makes it an automatic phenomenon with which the "neu-
ronal effort" to sustain a given functional stereotype is minimized.
It is perhaps important to recall at this stage that, as previously men-
tioned, neural embryogenesis by no means ends with birth. Consequently,
language organization and cerebral maturation occur during-and be-
cause of-the events previously mentioned. It follows, then, that abnor-
malities in these stages (whether originally due to biologic or environ-
mental influences) will cause disorders in language organization.
ANALYZERS AND STEREOTYPES
If we attempt a synthesis of the preceding discussion on neural

orthogenesis with certain tendencies in child neuropsychology (Azcoaga,
1981)-deeply rooted in Pavlov's theory of analyzers-and with neu-
rophysiology of language as has been explained mainly by Bejtereva
(1984), we could consider the central nervous system as an extremely
complex structure that conjugates biologic and environmental condition-
ants. If at the same time we refuse to explain language solely through the
study of the "external expressions" of cerebral activity (language ster-
eotypes) and rather prefer to describe the activity of neural circuits re-
sponsible for those stereotypes (or, in Pavlov's conception, the analyzers;
Azcoaga et 01., 1981), we could arrive at a novel conception of health and
sickness on the basis of the stable states theory-both physiologic and
pathologic (Bejtereva, 1984). We believe this approach, greatly influenced
by Azcoaga's and Bejtereva's contributions, will be extremely productive
both from the theoretical and the therapeutic perspective.
The whole embryologic process just described is aimed at the forma-
tion and then stabilization of neuronal circuits (analyzers) where the ex-
perience or the physiologic substrate of the capacity to react to this experi-
ence is "impressed" (Barbizet & Duizabo, 1978). Upon this basis, the
analyzer must then combine those elements "impressed" on its circuits
through the neurophysiologic mechanisms of neural excitation and inhi-
bition (Azcoaga et 01., 1983). "It is known that the phenomenon of lan-
guage (whose basics can be traced to the close interaction between two
signal systems: direct and verbal) obeys the same rules of excitation and
inhibition, irradiation and concentration as any other form of cortical
activity. The difference lies only in the fact that verbal processes, which
reflect reality through relations fixed in language, are accomplished
through a process of much greater complexity" (Luria, 1974, pp. 27). The
resulting "external expression" of activity of that analyzer-available to
the external observer-is the stereotype (Azcoaga et 01., 1983).
Environmental signals influencing CNS maturation allow the organi-
zation of analyzers that will sustain language and which, under normal
conditions, are formed along genetically predetermined pathways. Repe-
tition and coincidence over time of certain stimuli gradually stabilize
stereotypes that can be actualized at the "right" time for a "correct lan-
guage expression" (Azcoaga, 1977). If, as a consequence of structural de-
rangement or "mistaken" relationships between neural elements, these
"correct" stereotypes cannot be formed, those same environmental stim-
uli will influence the CNS, forcing the "impression" of different neuronal
circuits. As a consequence, analyzers will then have a different organiza-
tion, and stereotypes (their functional expression) will thus be abnormal.
These abnormal stereotypes can, of course, also be stabilized, and the
final consequence of this sequence of events is an abnormal language evo-

lution.
NORMAL LANGUAGE ORGANIZATION: PHYSIOLOGIC

STABLE STATES
Reflex or innate orofacial behaviors gradually acquire a commu-

nicative function, not yet linguistic, since they are not then symbols or
signs. In time they will be, signaling the initiation of language.
We shall describe the stages of development of communicative be-
havior in man according to Azcoaga's proposal as "first stage in commu-
nication (prelinguistic level)," "second stage in communication (first lin-
guistic level)," and "third stage in communication (second linguistic
level)" (Azcoaga et al., 1981).
At the prelinguistic level, spanning from 0 to 12 (15) months, it can be
observed how certain behavior, which originally had merely a survival
function such as crying, shouting, or even breathing, gradually puts itself
at the service of communication. Other behaviors such as sucking and
swallowing progressively refine themselves to allow the establishment of
well-coordinated patterns of movement. All these events establish the
basis for the fine orofacial movements and phonorespiratory coordination
that will be necessary to articulate sounds, syllables, and words. The role
of the adult as moderator and model in this process is almost unan-
imously accepted. The response the infant obtains will reinforce or stimu-
late their use with a communicative purpose. But the adult plays other
roles different from being a simple reinforcer. Bruner, according to
Owens, considers language development as "made possible by the pres-
ence of an interpreting adult who operates not so much as a corrector or
reinforcer but rather as a provider, an expander and idealizer of utterances
while interacting with the child" (Owens, 1984).
Siguan-Soler (1978) emphasizes the importance that early accentua-
tion, modulation, and gesture have in the full development of these initial
stages. The child is inmersed in a linguistic environment that continually
feeds him verbal messages; sensoriperceptive input then starts pairing
with verbal input that accompanies that particular situation, gradually
substituting the first signal system for the second one-which initiates
semantic recognition. This in turn will serve as basis for future verbal
comprehension and expression. It can be stated then that the infant recog-
nizes linguistic meanings before he can express himself in the oral code.
The sound emissions that start with crying, whimpering, or shouting
gradually differentiate toward vocal play, which basically consists in
playful repetition of phonation. This vocal play has two well-differenti-
ated stages: (1) "proprioceptive vocal play," in which sounds are emitted
94 LYDA MEJIA and JORGE ESLAV A-COBOS
because of the "pleasure" felt in the oropharynx by the child as he pro-

duces sounds (Azcoaga et al., 1981); (2) "auditory proprioceptive vocal
play" in which, as the child produces sounds, hearing aggregates inti-
mately to the proprioceptive afferences (Azcoaga et al., 1981). The lack of
auditory reinforcement "forces" the child to discard sounds not belong-
ing to his linguistic environment. On the contrary, the continuous rein-
forcement of sounds from his linguistic community trains him in the
production of those first babblings. It is this process of differentiation that,
guided by hearing, will allow the progressive recognition of the dis-
tinctive features of phonemes. "From then on, the child does not learn
sounds but phonemes; he no longer is guided by his own articulatory
instinct but rather submits to his environment and tries to imitate what he
hears" (Alarcos Llorach, 1976). From the adequate, coordinated synthesis
of auditory and proprioceptive afferences, two fundamental requisites for
language integration will arise: the recognition of acoustic features of the
phonic emissions from his linguistic environment, and the synthesis be-
tween the articulatory pattern and the acoustic image of the phonemes
produced by the same subject. These two accomplishments will result in
the phonematic stereotype, the first step in the process of language
organization.
The posterior synthesis of these phonematic stereotypes will result in
longer chains of emissions called motor verbal stereotypes, and these in
turn will eventually organize themselves in a syntactic ordainment. In
close relation to this process, the child recognizes and synthetizes acous-
tic features of his linguistic environment and ascribes meaning to them
because of the repetitive spatiotemporal coincidence of the emission with
a certain specific object or event. "The child observes that in a certain
situation the adults repeat the same expression and starts pairing them"
(Alarcos Lorach, 1976).
From the complex connections of those analyzers involved in the
reception and analysis of the various signals pertinent to the recognition
of meaning, verbal stereotypes gradually stabilize as functional ex-
pressions of the activity of the verbal analyzer. "At present, no one ac-
cepts that the meaning of words can be reduced to simple visual images
associated with their corresponding verbal designations. The meaning of
a word to a child is always coupled to some practical activity, in the
course of which abbreviation and differentiation gradually take place and
lead to that schematized 'concept of the object' that nominates a word"
(Luria, 1974).
As a concluding remark on this prelinguistic level, the importance of
the auditory control for the understanding of other aspects of communica-
tion should be emphasized. The auditory analyzer helps the child under-
stand and differentiate prosodic aspects of adult language and the corre-
sponding affective messages it transmits. At the same time he gradually

includes prosody as a communicative resource that accompanies gesture,
even though it does not yet have a true linguistic character. Even though
gesture will gradually be replaced by verbalization, it never will disap-
pear but rather will be used as a reinforcer of oral expression; indeed, in
times of great emotional excitement it can even replace language as a more
primitive-and occasionally more efficient-means of communication.
With the first verbal stereotypes, the second stage in communica-
tion-first linguistic level-starts. The initial event occurs when, through
the repetition of sounds that characterize vocal play, a meaningfuI.mono-
syllable appears. After that, the first words will gradually surge, in close
relation to the vital needs of the child, and of course imbued with a great
emotional significance. These words evoke something; they are the repre-
sentation through a differentiated significance (the word) of a significance
(the need) and initially have a generalizer role; because of it their meaning
is extended to many other objects or events circumstantially related. An
example of this situation is the use of the word bread to mean the act of
eating, or any food, the dining room, the dishes, the napkin, and similar
objects. Gradually, these words act as sentences in the stage known as
holosentence or word sentence, which essentially implies the use of a
single word with different possible meanings. At this moment the child
can randomly use holosentences or juxtaposed words; in the latter, in-
tensely meaningful words are linked together without any respect for
grammatical rules but with a strictly nominative sense, independently of
whether they are verbs, adjectives, or other parts of speech. Simple sen-
tences will then gradually appear, plagued with grammatical "errors"
since the child has not yet interiorized the "rules" that common use of
language dictates. Verbs, nouns, and adjectives appear more frequently
than prepositions and linkage elements because the latter require the
enchainment of the syntactic structure to highlight their full meaning,
since their significance is not linked to concrete objects or events. Verbs
in this stage do not respect inflections; nouns, adjectives, and adverbs can
likewise be misused. The child guides himself by generalizations of what
he hears, obviously without respect for the grammatical exceptions of his
native language. With the reinforcement and modeling offered by the
adult, he will gradually acquire them.
As the first linguistic level starts, the child shows the symbolic be-
havior inhering to the preoperational stage-preconceptual phase-
which denotes the establishment of a semiotic function defined by Piaget
as "being able to represent something (a 'significance' -object, event,
conceptual scheme, etc.) through a differentiated 'significant' which only
serves that representation: language, symbolic gesture, etc." (Piaget & In-
helder, 1981, p. 59). This symbolic behavior will dominate the activity of
the child until 4 years of age, when he will initiate intuitive thinking (still
in the preoperational stage). In spite of his having organized important
linguistic abilities, they are not yet sufficient to decentrate thinking,
which must then rely on perceptual clues. Reversibility is still absent,
hence preventing the child from understanding the relation of parts to the
totality, explaining conservations after a perceptual transformation, or
having categorical criteria for classifications.
The coincidence of this first linguistic level with preconceptual
thinking is responsible for the enrichment of semantic aspects of language
in this child, busy searching for different ways in which to represent the
world he is exploring through symbolic play. This latter is accompanied
by egocentric language, first expressed as repetition or echolalia, then as
monologue, and finally as collective monologue (Piaget & Inhelder, 1981).
The importance of this egocentric language cannot be sufficiently empha-
sized because it plays the role that interior language accomplishes in the
adult. "It not only escorts the activity of the child but serves his mental
orientation and conscious comprehension, and helps him in the difficul-
ties that continuously surge; it is the language for himself, intimately and
fruitfully linked with infantile thinking .... At the end it will transform
itself into interior language" (Vigotsky, 1962).
It is important to realize that the gradual enrichment of significances
contributes directly to syntactic progress. The handling of complex struc-
tures will in tum permit the expression of more abstract situations, pro-
gressively separated from purely perceptual aspects-as will be seen in
the third stage of communication. At the same time, that handling of
complex structures reveals to us a greater conceptual enrichment of the
individual. From the phonologic perspective, a gradual acquisition of
phonemes from his native language can be seen, but some with greater
difficulty than others, thus appearing later. While some explain this delay
through the need for finer articulatory movements that require greater
maturity of muscles involved (for those that appear later), others believe
that "the articulatory delay is rather due to an imperfect bond between the
acoustic and the motor image" (Alarcos Llorach, 1976)-that is, to an
instability of the kinesthetic motor verbal analyzer, whose functional ex-
ternal expressions are phonematic stereotypes.
By the end of this first linguistic level the child is in possession of all
the sounds of his native language, even though he sometimes fails in their
synthesis to form polysyllables with symphones. From the syntactic per-
spective he initiates his first complex sentences-basically of the coordi-
nated type-and has a verbal repertoire whose richness depends greatly
on his own peculiar environment. The connections and relations among
those significances from his repertoire that the child progressively makes
will allow the appearance of interior language.
At approximately 5 years of age the child initiates the third stage in

communication-second linguistic level-which can in turn be divided
in two substages: The first one coincides with intuitive thinking and ends
at 7 years of age; the second one, spanning from 7 through 12 years of age,
coincides with concrete operational thinking and the initiation of school-
ing.
In the first substage the phonologic aspects are rapidly perfected,
sentences gain in complexity, with great creativity in the use of linkage
elements and constructions in general, and a great many significances are
added; the child additionally finds new similitudes and differences be-
tween those significances and classifies them so that they will expand
that creative richness. Language is interiorized, monologue disappears as
a guider of activity, and this role is taken over by interior language. "Exte-
riorized decentrated language gradually diminishes, becomes fragmented
and appears only in a scattered, reduced, occasional form which some-
times becomes a mere whisper; it gradually passes to interior language. It
is an invariable part of the process of thinking" (Luria, 1979, p. 22).
We can now state that the oral code has been dominated; it will
further enrich itself with each new experience, a never-ending process
that essentially depends on the particular opportunities of each person.
At approximately 7 years of age the second substage of the second
linguistic level starts. A noteworthy event is the enrichment of interior
language, made possible by the coincident concrete operational structure
with the consequent possibility of categorization of significances, logic,
judgment, etc. A final consequence of all these phenomena is the perfect
mastery of complex subordinated and coordinated sentences, and height-
ened creativity and structural richness.
Learning of the written code begins. In that process the already inte-
riorized language will be raw material, but it will need the support of
gnosis and praxis different from those learned in oral language. Oro facial
praxis and auditory-proprioceptive coordination involved in oral ex-
pression are replaced by manual praxis and visuomanual coordination.
By the same token, temporoauditory gnosis is replaced by visuospatial
gnosis in the comprehension process. Phonetic material is replaced by
graphic elements, but with the additional difficulty imposed by the fact
that graphic symbols are not univocally and universally paired with the
same sound, and also by the necessity of substituting trained channels for
specific discrimination for other perceptual discriminatory modalities for
the same meaning. In learning these abilities the child spends a great part
of his time. By the end of this stage, at approximately 12 years of age, the
individual will be in possession of the two codes. Language will then be
his best tool for learning new skills and facts, communicating his ideas
and feelings, or talking with himself in those long monologues that will
help him solve problems, be artistically creative, or amuse himself. From

then on, he will enrich and specialize his language in direct relationship
to the characteristics of his sociocultural environment.
ABNORMAL LANGUAGE ORGANIZATION: PATHOLOGIC

STABLE STATES
Language derangement syndromes of children cannot be classified in

the same way as those of adults. Even though certain symptoms are equiv-
alent, it is well known how parenchymal lesions do not produce in the
child the clinical manifestations observable in the adult. Descriptions of
children severely affected initially but who gradually compensate for
those difficulties or (on the opposite) who apparently worsen progres-
sively in their linguistic abilities also abound.
As a general "rule of thumb," parents describe their children's diffi-
culties in two different ways: (1) The child understands and talks poorly.
(2) The child understands well but talks poorly or not at all. These de-
scriptions coincide well with what has been a constant in the historical
revision: difficulties in comprehension or expression of language. In
agreement with the preceding discussion, we consider that these children
suffer derangements of their verbal analyzer (semantic code) or their kin-
esthetic motor verbal analyzer (phonologic-syntactic code).
Abnormalities of language analyzer in children can be grouped basi-
cally under three different conditions: (1) retarded organization due to
delay in neurophysiologic mechanisms or poor environmental stimula-
tion, which results in the late initiation of the first linguistic level without
pathologic symptoms; (2) inadequate organization due to "abnormal ana-
lyzer architecture" because of neurologic derangements that occur before
the first linguistic level is under way; this will produce a delay in lan-
guage evolution which will progressively incorporate pathologic symp-
toms equivalent to those of the adult; and (3) disorganization of analyzers
because of neurologic lesions that occur after linguistic levels are well
under way, which will determine involution of language milestones al-
ready acquired and symptoms similar to those of the adult. We shall
attempt a discussion of these abnormalities through the separate study of
the verbal and kinesthetic motor verbal analyzers. Nevertheless, it should
be borne in mind that in certain conditions (e.g., mental retardation) both
analyzers are affected. Hearing impairment will not be discussed.
Kinesthetic Motor Verbal Analyzer

Derangements of the kinesthetic motor verbal analyzer produce diffi-
culties in language expression. A frequent condition is "simple language
retardation" characterized by a delay in verbal expression and prolonged

persistence of infantile talk without other symptoms, and adequate com-
prehension. In these children, expression milestones occur in their cor-
rect order, but slowly. By the end of the first linguistic level the delay is
compensated, and the child continues a normal evolution.
Multiple etiologies have been proposed. Ajuriaguerra and Marcelli
(1982) believe it is due to "cerebral inmaturity" or lack of verbal stimula-
tion, or both. Azcoaga et 01. (1981) propose chronologic, psychologic,
familial, and sociocultural factors. Inadequate or insufficient stimulation
in the pre linguistic level could delay connections between the analyzer
processing proprioceptive information from orofacial structures and hear-
ing, thereby impeding syntheses necessary for language expression, with
the consequent "simple retardation." Nevertheless, if this environmental
anomaly prolongs itself into the critical stages, poorly organized kines-
thetic motor verbal analyzers may consolidate engrams thus stabilizing
abnormal phonematic and motor verbal stereotypes in a pathologic stable
state. Otherwise, the almost mandatory enrichment of stimuli that occurs
in the first linguistic level corrects this retardation before this second
stage ends.
A familial history is frequently encountered. Nevertheless, this histo-
ry must only be one of simple retardation in acquiring milestones but
devoid of pathologic symptoms (Azcoaga et 01.,1981); otherwise the retar-
dation will most probably not be a "simple" one.
Quite different is the situation where anomalies in this analyzer are
due to neurologic derangements occurring before the first linguistic level
is under way. These children likewise suffer a lag in their language
milestones, but their expression is also plagued with pathologic symp-
toms in both phonologic and syntactic aspects. Occasionally the delay in
acquiring milestones is not initially evident, but only the pathologic
symptoms such as multiple dyslalias and poor syntactic organization ap-
pear; with time, the retardation gradually makes itself evident. In these
children the kinesthetic motor verbal analyzer must search for new ways
of organization, establishing and consolidating different connections that
will result in normal or pathologic stable stereotypes, depending on the
characteristics of environmental stimuli, and degree and magnitude of
neurologic derangement.
Lenneberg (1967) maintains that lateralized lesions suffered between
birth and 20 months of age do not interfere with normal language mile-
stones in half the cases, while in the other half, language is retarded but
otherwise normal; he likewise feels that prognosis and symptoms are
independent of which hemisphere is affected. In this case, the situation
would usually pass undetected or, at most, be considered a "simple retar-
dation." Only those with bilateral lesions would be clearly symptomatic.
If subtle aspects of linguistic abilities are considered, we believe most
children with early lesions will show language difficulties (Eslava-Cobos

& Mejia, 1985).
A clear genetic factor is advocated by many authors (Gutzmann, 1916;
Orton, 1930; Pfaendler, 1960). Brewer's statement is endorsed by Len-
neb erg (1967): "Congenital linguistic ability is probably a dominant fea-
ture, influenced by sex and with at least a regular penetrance" (p. 283).
When these children are left untreated, their symptoms progressively
worsen. What initially was only multiple dyslalia gradually evolves into a
poor syntactic structure by the end of the first linguistic level: Complex
sentences are impossible because of the difficulties in the syntheses
among different motor verbal stereotypes, dyslalias become fixed, and
abnormal phonematic stereotypes stabilize, hence stabilizing also aber-
rant motor verbal stereotypes (Azcoaga et al., 1981, p. 85). "[TJhe evo-
lutive clinical characteristics of anarthric retardation show well the de-
pendency of the grammatical system on the phonologic in the linguistic
level. ... By the same token, from the neurophysiologic perspective, the
inadequate organization of the analytic-synthetic activity at different lev-
els starts hampering phonematic stereotypes and ends jeopardizing the
whole syntactic grammatical organization" states Azcoaga, who refers to
this derangement of the kinesthetic motor verbal analyzer as "anarthric
retardation. "
The great influence of this deficient organization of the kinesthetic
motor verbal analyzer on the process of learning of the written code must
be emphasized. Azcoaga, Derman, and Iglesias (1979), Launay (1979), and
Quiros and Della Cella (1971), among others, have elaborated on the great
relationship that exists between learning disabilities and language ex-
pression retardations. Quite obviously, when the graphic element has to
be paired with the phoneme, this abnormal phonematic stereotype leads
the child to confusion with the written material as it is presented to him;
what he sees and hears does not match with what he has fixed as the
activity of the kinesthetic motor verbal analyzer.
Finally, in conditions that produce disorganization of this analyzer
because of neurologic lesions that occur after linguistic levels are well
under way, symptoms will be more similar to those of adults, the older the
child is by the time he suffers the lesion. Some children can lose ex-
pression altogether. The process may then reinitiate, fulfilling all its
stages in an abbreviated form with an almost total compensation. For
others, especially when the child has already attained the second lin-
guistic level, sequelae are usual, with important difficulties in articula-
tion and agrammatism, in spite of vigorous therapy.
Verbal Analyzer
"Simple" retardations, equivalent to the one previously described for
the kinesthetic motor verbal analyzer, are infrequent in verbal analyzer
derangements. Nevertheless, occasionally a child can be seen who starts

comprehension of oral language at older stages than what is accepted as
normal, not in the general framework of mental deficiency, and devoid of
pathologic symptoms. If this delay for verbal stereotypes is due to severe
environmental deficiencies, as for example in forlornness that prolongs
itself beyond 3 or 4 years of age, it usually is not compensated, especially
because other affective factors are also hindered. Not only language, but
the totality of communication is then jeopardized.
When the inadequate organization is due to "abnormal analyzer ar-
chitecture" because of neurologic derangements that occur before the first
linguistic level is under way, a delay in the acquisition of guidelines
related to language comprehension is seen. As the child progresses in the
first linguistic level, symptoms equivalent to those of the adult also ap-
pear: verbal paraphasias, anomias, agrammatism, defective articulation.
The greatest problem occurs when the primary level of analysis of
nonlinguistic characteristics of phonic material is hampered, since this
hinders not only connections essential for other levels involved in de-
codification but also the coordination of proprioceptive feedback of phon-
ic emissions with recognition of characteristic features of one's own
sounds. In this case, both the verbal and the kinesthetic motor verbal
analyzer are poorly organized, leading to a mixed language retardation
syndrome of poor prognosis. These children cannot perform an adequate
analysis and integration of sounds, in spite of normal hearing, a situation
referred to as "hearing agnosia" (Launay, 1979) or "congenital auditory
imperception" (Worster-Drought & Allen, 1929). A severe lack of com-
prehension of language occurs; the expression is confined to idioglossia
accompanied by gestures. These children are usually misdiagnosed as
deaf or-if a profound audiological evaluation rules out deafness-as
"dysphasic" at best. The resulting treatment consequently will emphasize
verbal symbolism without great concern for discriminatory training of
phonic material, a strategy doomed to failure.
In the more frequent, true aphasic retardation (Azcoaga et al., 1981),
the verbal analysis is the only one involved without derangement of the
acoustic analysis. It is thus named by Azcoaga as he insists that the phys-
iopathologic phenomena are equivalent of those in the aphasias of the
adult (differentiating expressive aphasias as "anarthria"). Symptoms are
consequently equivalent to those of the adult, although they gradually
change and evolve as the child's development progresses. In addition,
environmental pressure progressively highlights difficulties that might
have gone undetected in previous stages.
Derangements of the verbal analyzer eventually produce impairment
of expression. Not infrequently the child is brought for treatment because
of those expressive difficulties and not because of miscomprehension.
This situation is fostered by the relatively simple communicative environ-
ment the young child is exposed to, which allows nonlinguistic com-
prehension and therefore leaves the expression difficulties as the sole

apparent reason for concern. In the clinical setting, these children show
difficulty in understanding orders and complex verbalizations, with a
contextual-rather than verbal-comprehension. Multiple paraphasias,
latencies, and anomias are found, and occasionally language is even re-
stricted to idioglossia. Slight derangements of this analyzer may pass
undetected until the child enters school, when he will suffer learning
disabilities-not only reading and writing clumsiness, but difficulties in
other subjects as well, when they demand abstraction abilities that pull
him away from concrete references (Azcoaga et 01., 1979).
The disorganization of the activity of the verbal analyzer because of
neurologic lesions that occur after linguistic levels are well under way
leads to disintegration of language in syndromes gradually more akin to
those of the adult, the older the child is when he suffers the lesion. Even
when posttraumatic aphasias occur in the first linguistic level, where
language apparently fully recovers, academic learning processes are hin-
dered and children require therapeutic support, especially because of
difficulties in abstraction (Kiessling, Denckla, & Carlton, 1983).
CONCLUSIONS
The study of disorders in language acquisition in child~en is hin-
dered by the changes that maturation imposes on the clinical manifesta-
tions, thereby forcing their interpretation within a dynamic frame. They
must then be considered from the perspective of functional states-either
physiologic or pathologic-that progressively stabilize until a stable state
is reached. This evolution has its origins in neural embryogenesis, a com-
plex phenomenon, determined by both biologic and environmental fac-
tors, which probably extends itself at least until the third decade of life.
Language derangements can be grouped in two large categories: diffi-
culties in the phonologic-syntactic code or in the semantic code, depend-
ing on whether the kinesthetic motor verbal or the verbal analyzer is
involved. Conditions characterized by difficulties in the organization of
analyzers must be differentiated from those where involution is the basic
phenomenon, even though the process of stabilization of their stereotypes
is still in its initial stages. Derangements of both analyzers or associations
with difficulties in other cerebral functions can be frequently seen, lead-
ing to more complex pathologic conditions.
REFERENCES
Ajuriaguerra. J., Guignard. F .• Jaeggi, A., Kocher, F., Maquard. M .• Paunier. A .• Quinodoz. D .•
& Siotis. E. (1969). Organizacion psicologica y perturbaciones del desarrollo del
lenguaje. In J. Piaget, J. Ajuriaguerra, F. Bresson, P. Fraisse, B. Inhelder, & P. Oleron

Introduccion ala psicolinguistica (pp. 119-146). Buenos Aires: Proteo.
Ajuriaguerra, J., & Marcelli, D. (1982). Manual de psicopato10gia del nino. Barcelona: Toray
Mason.
Alarcos Llorach, K (1976). La adquisicion del Lenguaje por el nino. In A. Martinet, E.
Alarcos Llorach, S. Borel-Maisonny, H. Hecaen, D. Mandin, & P. Guiraud TLE. (pp. 7-
42). Buenos Aires: Nueva Vision.
Azcoaga, J. (1977). Trastornos de11enguaje. Buenos Aires: Ateneo.
Azcoaga, J. (1981). De11enguaje a1 pensamiento verbal. Mexico: Ateneo.
Azcoaga, J., Bello, J., Citrinovitz, J., Derman, B., & Frutos, W. (1981). Los retardos del
1enguaje en e1 nino. Barcelona: Paidos.
Azcoaga, J., Derman, B., & Iglesias, P. (1979). Alteraciones del aprendizaje esco1ar. Buenos
Aires: Paidos.
Azcoaga, J., Fainstein, J., Ferreres, A., Gonorasky, S., Kochen, S., Krynveniuk, M., &
Podliszewski, A. (1983). Las funciones cerebra1es superiores y sus a1teraciones en e1
nino y en e1 adu1to. Buenos Aires: Paidos.
Barbizet, J., & Duizabo, P. (1978). Manual de neuropsico10gia. Barcelona: Toray Mason.
Bejtereva, N. (1984). E1 cerebro humano sana y enfermo. Buenos Aires: Paidos.
Billard, C., Autret, A., Laffont, F., Degiovanni, K, Lucas, B., Santini, J., Duloc, 0., Ploin, P.
(1981). Aphasie acquise de l'enfant avec epilepsie. Revue E1ectroencepha10graphie et
Neurophysio10gie Clinic, 11,457-467.
Bishop, D. (1985). Age of onset and outcome in acquired aphasia with convulsive disorder
(Landau-Kleffner syndrome). Developmental Medicine and Child Neurology, 27, 705-
712.
Brissaud, K, & Richardet, J. (1974). Le syndrome des integration du langage et comitialite.
Journal Parisien du Paediatrie, 441-445.
Broadbent, W. (1872). On the cerebral mechanisms of speech and thought. Med. Chir. Trans.
15, 155.
Brock, S., & Krieger, H. (1966). Fundamentos de neuro10gia c1inica. Barcelona: Jims.
Chang, F., & Greenough, W. (1982). Lateralized effects of monocular training on dendritic
branching in adult split-brain rats. Brain Research, 232, 283-292.
Coen, R. (1880). Patho10gie und Therapie der Sprachanomalien. Viena: Urban und Schwarz-
enberg.
Cowan, W. (1979). The development of the brain. Scientific American, September, 112-133.
DeMyer, W. (1972). Megalencephaly in children: Clinical syndromes, genetic patterns and
differential diagnosis from other causes of megalocephaly. Neurology, 22, 634.
DeMyer, W., Zeman, W., & Palmer, C. (1964). The face predicts the brain: Diagnostic signifi-
cance of median facial anomalies for holoprosencephaly (arrinencephaly). Pediatrics,
34, 256.
Deonna, T., Beamanoir, A., Gaillard, F., & AssaI, G. (1977). Acquired aphasia in childhood
with seizure disorders: A heterogeneous syndrome. Neuropaediatrie, 8, 263-273.
Duloc, 0., Billard, C., & Arthuis, M. (1983). Aspects electro-cliniques et evolutifs de l'epilep-
sia dans Ie syndrome aphasie-epilepsie. Archives Francaises du Paediatrie, 40, 299-
308.
Eccles, J. (1970). Observando 1a rea1idad. Berlin: Springer-Verlag.
Eslava-Cobos, J., & Mejia, 1. (1985). Los trastornos de aprendizaje y la organizacion cerebral
de funciones. In P. Montanes (Ed.), Asimetria funciona1 cerebral (pp. 133-145). Bogota:
As. Col. Neuropsicologia.
Galaburda, A., & Kemper, T. (1979). Cytoarchitectonic abnormalities in developmental dis-
lexia: A case study. Annals of Neurology, 6, 94-100.
Guttman, K (1942). Aphasia in children. Brain, 65, 205-219.
Gutzmann, H. (1916). Die Vererbung der Sprachstoerungen. Leipzig: Thieme.
Hadden, W. (1891). On certain defects of articulation with cases ilustrating the results of
education on the oral system. Journal of Mental Sciences, 37, 96.
Horwitz, B. (1981). Neuronal plasticity: How changes in dendritic architecture can affect the
spread of postsynaptic potentials. Brain Rsearch, 224, 412-418.
Kaye, M., Mitchell, D., & Cynader, M. (1982). Depth perception, eye alignment and cortical
ocular dominance of dark reared cats. Developmental Brain Research, 2, 37-53.
Kerr, J. (1897). School hygiene in its mental, moral and physical aspects. Journal of the Royal
Statistics Society, 60, 613.
Kiessling, L., Denckla, M., & Carlton, M. (1983). Evidence for differential hemispheric func-
tion in children with hemiplegic cerebral palsy. Developmental Medicine and Child
Neurology, 25, 727-734.
Landau, W., Goldstein, R, & Kleffner, F. (1954). Congenital aphasia: A clinic pathologic
study. Neurology, 10, 915-921.
Landau, W., & Kleffner, F. (1957). Syndrome of acquired aphasia with convulsive disorder in
children. Neurology, 7, 523-530.
Launay, C. L. (1979). Graves en la adquisicion de lenguaje. In C. Launay & S. Burel-Maison-
ny, Trastornes del lenguaje la palabre y la voz en el nino (pp. 128-129). Barcelona:
Toray Masson.
Lenneberg, E. (1967). Biological foundations of language. New York: Wiley.
Levi-Montalcini, R (1964). Events in the developing nervous system. Progress in Brain
Research, 4, 1-29.
Ley, J. (1930). Les troubles du developpement du langage. Journal du Neurologie et Psychi-
atTie, 30, 78-90.
Luria, A. (1974). Cerebro y lenguaje. Barcelona: Fontanella.
Luria, A. (1979). El papel del lenguaje en el desarrollo de la conducta. Buenos Aires:
Cartago.
McLennan, 1., & Hendry, 1. (1981). Effects of peripheral deprivation on mammalian motoneu-
rone development. Developmental Brain Research, 1, 440-443.
Meisami, E., & Mousavi, R. (1982). Lasting effects of early olfactory deprivation on the
growth, DNA, RNA, and protein content, and Na- K-ATPase and ache activity of the rat
olfactory bulb. Developmental Brain Research, 2, 217-229.
Morley, M. (1957). The development and disorders of speech in childhood. Edinburgh:
Livingstone.
Mower, G., Burchfield, J., & Duffy, F. (1981). The effects of dark-rearing on the development
and plasticity of the lateral geniculate nucleus. Developmental Brain Research, 1,418-
424.
Orton, S. (1930). Familial occurrence of disorders in acquisition of language. Eugenics, 3,
140-147.
Owens, R (1984). Language development. Columbus: Merrill.
Pfaendler, V. (1960). Les vices de la parole dans l'optique du geneticien. Akt Problem der
Phoniatriks und Logopaedie, 1,35-40.
Piaget, J., & Inhelder, B. (1981). Psicologia del nino. Madrid: Morata.
Popper, K. (1977). Busqueda sin termino: Una autobiografia intelectual. Madrid: Tecnos.
Purpura, D. (1976). Structure dysfunction relations in the visual cortex of preterm infants. In
C. Brazier, (Ed.), Brain dysfunction in infantile febrile convulsions (pp. 223-240). New
York: Raven Press.
Quiros, J., D'Elia, N., Feldman, J., Galian, M., Lira, E., Lujan Marelli, L., Mercado, M.,
Pericoli, B., Rincon, M., Rosental, C., Schrager, 0., Soto, R, & Tormakh, E. (1971). Las
llamadas afasias infantiles. Buenos Aires: C.E.M.1.F.A.
Quiros, J., & Della Cella, M. (1971). La dislexia en la ninez. Buenos Aires: Paidos.
Rabinowicz, T. (1976). Morphological features of the developing brain. In C. Brazier (Ed.),
Brain dysfunction in infantile febrile convulsions (pp. 1-23). New York: Raven Press.
Sidman, R, & Rakic, P. (1973). Neuronal migration with special reference to developing
human brain: A review. Brain Research, 62, 1.
Siguan Soler, M. (1978). De la comunicacion gestual al lenguaje verbal. In J. Bronckart, P.
Malrieu, M. Siguan Soler, H. Sinclair, T. Slama-Cazacu, & A. Tabouret La genesis del
lenguaje (pp. 23-50). Madrid: Pablo del Rio.
Swaimann, D., & Wright, F. (1982). The practice of pediatric neurology. St. Louis: C. V.
Mosby.
Ucherman, V. (1891). The case of dumbness (aphasia) without deafness, paralysis or mental
debility. Archives of Otolaringology, 20, 321.
Vigotsky, 1. S. (1962). Thought and language. Cambridge, MA: M.LT. Press.
Volpe, J. (1977). Neurologia neonatal. Clinicas de Perinatologia, 4, 3-30.
Worster-Drought, C. (1943). Congenital auditory imperception (congenital word deafness)
and its relation to idioglossia and allied speech defects. Medical Press and Circu~ar, 210,
5460.
Worster-Drought, C. (1968). Speech disorders in children. Developmental Medicine and
Child Neurology, 10, 427-440.
Worster-Drought, C., & Allen, L (1929). Congenital auditory imperception. Journal of Neu-
rology and Psycopathology, 10, 193.
Wyllie, J. (1894). The disorders of speech. Edinburgh: Oliver & Boyd.
Yakovlev, P., & Lecours, A. (1967). The myelogenetic cycles of regional maturation of the
brain. In Minkowski, A. (Ed.) Regional development of the brain in early life (pp. 3-70).
Philadelphia: F. A. Davis.
6
Speech and Language Alterations

.
In
Dementia Syndromes
JEFFREY L. CUMMINGS and D. FRANK BENSON
Dementia is an etiologically nonspecific syndrome characterized by ac-

quired persistent disturbance of several areas of neuropsychological func-
tion. Once regarded as a syndrome of global intellectual loss, dementia is
increasingly approached as a group of disorders with distinct neuropsy-
chological patterns of preserved and impaired abilities reflecting disease-
related regional brain involvement. The topography of histologic, bio-
chemical, and metabolic alterations associated with individual central
nervous system (CNS) disorders can be correlated with disease-related
behavioral profiles. Conversely, behavioral observations are utilized to
infer the geography of brain dysfunction and the etiology of the neu-
rologic condition. Such brain-behavior correlations have reached an ac-
ceptable degree of reliability with regard to monosymptomatic deficit
syndromes such as the aphasias (Benson & Geschwind, 1985) and am-
nesias (Benson, 1978; Damasio, Graff-Redford, Eslinger, Damasio, & Kas-
sell, 1985) but are in a relatively primitive state of development in poly-
symptomatic disorders such as the dementias. A crude distinction has
been drawn between dementi as associated primarily with cortical dys-
function and those occurring with subcortical disturbances (Cummings &
Benson, 1983, 1984), but more refined distinctions are necessary both to
improve clinical diagnosis and management and to enhance understand-
ing of brain-behavior relationships in the dementia syndromes.
Memory dysfunction has received the most intensive study within
the dementias. In the Diagnostic and Statistical Manual of Mental Disor-
JEFFREY L. CUMMINGS and D. FRANK BENSON' The Neurobehavior Unit, West Los
Angeles Veterans Administration Medical Center, Departments of Neurology and Psychiatry
and Behavioral Sciences, UCLA School of Medicine, Los Angeles, California 90024.
107
108 JEFFREY L. CUMMINGS and D. FRANK BENSON
ders (DSM-III) (American Psychiatric Association, 1980) memory loss is a

mandatory criterion for the diagnosis of dementia. The pan-biographic
memory loss of the dementi as has been contrasted with the temporal
gradient present in isolated amnesia syndromes (Freedman, Rivoira, But-
ters, Sax, & Feldman, 1984; Wilson, Kaszniak, & Fox, 1981), and the learn-
ing deficit of the cortical dementias has been contrasted with the recall
disturbance of subcortical dementias (Caine, Ebert, & Weingartner, 1977;
Cummings, 1986; Cummings & Benson, 1984).
Abnormalities of speech and language occur in nearly all demented
patients, but these disturbances are less well recognized as providing
criteria for distinguishing between different dementia syndromes. Apha-
sia as a manifestation of focal left-hemisphere injury following stroke,
trauma, neoplastic involvement, or infection has been investigated and
categorized (Benson, 1979), and the language disturbance of dementia of
the Alzheimer type (DAT) has also been described (Appell, Kertesz, &
Fisman, 1982; Cummings, Benson, Hill, & Read, 1985). The value of lan-
guage profiles in differentiating among the various dementing disorders
remains to be determined. Likewise, dysarthria is well recognized as a
manifestation of motor system dysfunction, but the usefulness of motor
speech abnormalities in dementia diagnosis has not been established. In
this chapter, the application of a battery of speech and language tests to
normal control subjects and to patients with three varieties of dementia is
described. The usefulness of speech and language alterations in dis-
tinguishing dementing conditions is investigated and discussed. The de-
mentia syndromes are DAT, Parkinson's disease (PD) with overt demen-
tia, and multi-infarct dementia (MID).
METHODOLOGY
Assessment
The battery of speech and language tests used in these studies was
derived from the Boston Diagnostic Aphasia Examination (Goodglass &
Kaplan, 1976) and the Western Aphasia Battery (Kertesz, 1979). The battery
included 37 operationalized subtests and assessed elements of spon-
taneous speech, auditory comprehension, repetition, naming, paraphasia,
reading aloud, reading comprehension, automatic speech, sentence com-
pletion, word list generation (number of animals named in 1 minute),
writing, dysarthria, and reiterative speech disturbances. Each subtest re-
ceived a scale score of 0 (normal) to 6 (most abnormal) based on responses
to six (or a multiple of six) scorable questions (e.g., "Is your name Green?")
or graded by operationalized anchor points. Table 1 lists the elements of the
battery. The Mini-Mental State Examination (MMSE; Folstein, Folstein, &
McHugh, 1975) was used to determine the severity of the dementia syn-
SPEECH AND LANGUAGE ALTERATIONS IN DEMENTIA SYNDROMES 109
TABLE 1. Elements of Speech and Language

Assessed with the Dementia Speech
and Language Instrument
Spontaneous speech Reading comprehension

Information content Words
Melodic line Sentences
Phrase length Commands
Grammatical complexity Writing
Auditory comprehension Mechanics
Word discrimination Dictation
Yes/no Narrative
Sequential commands Automatic speech
Complex commands Alphabet recitation
Repetition Counting (1-20)
Words Completion
Numbers Sentences
Phrases Nursery rhymes
Naming Dysarthria
Paraphasia Loudness
Literal Pitch
Verbal Articulation
Neologistic Rate
Reading aloud Intelligibility
Words Reiteration
Sentences Stuttering
Commands Echolalia
Palilalia
Logoclonia
dromes. BMDP statistical software (Dixon, Brown, Engleman, & Hill, 1985)
was used for all statistical analyses; the Bonferroni adjustment for multiple
testing has been applied to all results reported as significant.
Participants
All participants had at least an eighth-grade education and spoke
English as their native language. Seventy normal control subjects were
tested and all had MMSE scores in the normal range.
All dementia patients were thoroughly evaluated to establish an ac-
curate clinical diagnosis. Each had a thorough historical review, physical
examination, neurologic examination, and mental status testing. Standard
laboratory studies, including complete blood count, erythrocyte sedimen-
tation rate, serum electrolytes, serum calcium and phosphorus, serum
vitamin B12 and folate levels, thyroid and liver function tests, and a
serological test for syphilis, were obtained and determined to be normal.
An electroencephalogram and computerized tomographic (CT) scan of the
110 JEFFREY 1. CUMMINGS and D. FRANK BENSON
head excluded hydrocephalus, subdural hematomas, abscesses, neo-

plasms, and other occult causes of mental impairment.
DAT subjects met DSM-III criteria for primary degenerative dementia
and the NINCDS-ADRDA criteria (McKhann et aI., 1984) for probable
Alzheimer's disease. In addition, all had an Ischemia Scale score (Hachin-
ski et al., 1975) of 4 or less, supportive of a diagnosis of a degenerative
disorder. Seventeen men and 13 women were considered to have DAT-
15 patients with onset before age 65 and 15 with onset after that age.
Patients with stroke-related dementias met DSM-UI criteria for MID and
had Ischemia Scale scores of 7 or higher. PD patients had bradykinetic-
rigid extrapyramidal syndromes consistent with idiopathic PD (Adams &
Victor, 1977; Cummings & Benson, 1983). Historical and clinical features
were used to exclude patients with progressive supranuclear palsy, post-
encephalitic PD, drug-induced PD, and atypical extrapyramidal syn-
dromes from the PD group. A total of 21 patients with MID and 52 with PD
(18 with overt dementia and 34 without overt dementia) were studied. All
comparisons between dementia syndromes are based on patient groups
with equally severe dementia syndromes as reflected in their mean MMSE
scores (the number of participants in each comparison is shown in the
tables).
RESULTS
Normal Control Subjects

Control subjects ranged in age from 20 to 75 years and had a mean age
of 42.4 years. Twenty members of the control group were between the ages
of 50 and 75, the age range of the patient group. The mean MMSE score for
controls was 29.50. There was no significant difference between the
MMSE scores of young and old control subjects.
Individuals in the control group performed almost perfectly on the
speech and language tests. At least one of the 70 controls made errors on
the following subtests: yes/no questions, complex commands, complex
comprehension, alphabet recitation, nursery rhyme completion. Using a
one-sample t test, however, the means for the performance of the controls
did not differ significantly from zero (normal). Young and old controls
performed similarly, and the entire group was used for comparison with
the patient groups.
Dementia of the Alzheimer Type

Results of DAT patients have been reported previously (Cummings et
al., 1985). The mean age of the DAT patients was 71.34 and their mean
MMSE score was 10.50.
Mean DAT patient scores on all subtests were abnormal and dis-
TABLE 2. Characteristics of Verbal Output

of Patients with Dementia of the Alzheimer Type
Sample characteristics
Age (mean with SEM) 71.33 (1.64)
MMSE 10.50 (1.3)
N 30
Score
Most impaired features (scale score> 3/6)
Narrative writing 4.63 (0.39)
Information content of speech 3.83 (0.29)
Word list generation (animals/min) 3.37 (0.59)
Nursery rhyme completion 3.17 (0.42)
Comprehension of complex commands 3.09 (0.43)
Naming 3.03 (0.34)
Least impaired abilities (scale score < 2/6)
Reading sentences aloud 1.97 (0.47)
Phrase repetition 1.87 (0.40)
Sentence completion 1.83 (0.46)
Phrase length 1.80 (0.38)
Grammatical complexity 1.53 (0.38)
Reading words aloud 1.47 (0.45)
Melodic line 1.43 (0.34)
Verbal paraphasia 1.43 (0.32)
Neologistic paraphasia 1.40 (4.40)
Echolalia 1.33 (0.25)
Loudness 1.28 (0.29)
Intelligibility 1.17 (0.27)
Literal paraphasia 1.10 (0.26)
Counting 1.07 (0.38)
Articulation 1.07 (0.24)
Pitch 1.00 (0.23)
Number repetition 0.97 (0.38)
Rate 0.90 (0.24)
Word repetition 0.80 (0.32)
Palilalia 0.70 (0.27)
Logoclonia 0.70 (0.31)
Stuttering 0.33 (0.14)
tinguished the patient from the control population. After Bonferroni ad-
justment for multiple testing, six subtests no longer differed from the
control group: word and number repetition, reading words aloud, stutter-
ing, palilalia, and logoclonia. The-most marked disturbances (scale scores
> 3/6) involved information content of spontaneous speech, comprehen-
sion of complex commands, naming, word list generation, writing to dic-
tation, narrative writing, and completion of nursery rhymes (Table 2).
Relatively preserved abilities (scale scores of < 2/6) included counting,
sentence completion, phrase length, grammatical complexity, melodic
line, paraphasias, oral reading, repetition, dysarthria measures, and re-
112 JEFFREY L. CUMMINGS and D. FRAl'\)K BENSON
iterative disturbances. The DAT patients exhibited a characteristic pattern

comprising a fluent verbal output, impaired auditory comprehension,
anomia, and aphasic agraphia, with relative preservation of repetition,
oral reading, and motor speech. The language profile resembles that of
transcortical sensory aphasia (Benson, 1979).
Analyses revealed that only one subtest had a significant gender ef-
fect: Female DAT patients had better performance on the narrative writing
portion of the examination. There was a consistent trend for patients
whose disease began before age 65 to exhibit more severe deficits in the
middle years of the disease course (4 to 9 years' duration), suggesting that
an early onset of DAT was associated with more rapid decline 'of language
function. This trend was most obvious for information content of spon-
taneous speech, auditory comprehension, and writing to dictation. In-
creasing length of illness was reflected in decreasing phrase length, loss of
melody, impaired repetition, declining comprehension, abnormal count-
ing, disturbed speech pitch, and decreased intelligibility. Thus, as the
disease progressed, the verbal output acquired some of the characteristics
of Wernicke's aphasia along with loss of some aspects of fluency and
increasing dysarthria.
There was a consistent correlation between severity of dementia as
reflected in MMSE scores and loss of linguistic function. Stepwise multi-
ple regression analyses revealed that three subtests-comprehension, nar-
rative writing, writing to dictation-accounted for 70% of the MMSE
score variation. Correlation between disease duration and language altera-
tion was less robust, reflecting the wide range of reported disease
durations.
Six patients had family histories suggesting an inherited form of
DAT. Those with familial DAT tended to perform better on the speech and
language subtests, with significantly better scores on sentence repetition,
alphabet recitation, and oral reading, and they had fewer neologisms.
With a separate series of tests, the oral reading and reading com-
prehension abilities of 13 DAT patients were investigated (Cummings,
Houlihan, & Hill, 1986). Patients with a wide range of disease severity
(MMSE scores from 19 to 1) were studied. Oral word and letter reading
was preserved throughout the entire range of intellectual deterioration
even when incomplete or partially obscured letters were presented. Read-
ing comprehension, on the other hand, deteriorated progressively as de-
mentia severity increased. Reading aloud of irregular words was largely
preserved, and DAT patients did not recapitulate the errors typical of
children first acquiring reading skills.
Parkinson's Disease
There was no difference between the DAT and PD patient groups in
age or dementia severity (Table 3). DAT and PD could be distinguished by
TABLE 3. Demographic Features and Significant Speech and Language

Differences between Patients with Dementia of the Alzheimer Type (DAT)
and Parkinson's Disease (PD) with Overt Dementia
Bonferroni
PD score DAT score significance
(SEM) (SEM) level
Sample characteristics
N 16 10
Age 71.81 (0.96) 73.20 (3.30) n.s.
MMSE 20.25 (0.78) 17.90 (0.84) n.s.
Distinguishing features
Phrase length 2.31 (0.31) 0.40 (0.10) .001
Information content 1.18 (0.24) 2.80 (0.24) .001
Melody 2.56 (0.15) 0.40 (0.16) .001
Writing mechanics 3.62 (0.37) 0.60 (0.40J .001
Naming 0.43 (0.20) 2.10 (0.54) .05
Grammatical complexity 1.06 (0.28) 0.20 (0.13) .05
Animal naming (total/min) 10.02 (0.91) 5.50 (1.19) .01
Loudness 3.81 (0.29) 0.50 (0.22) .001
Pitch 2.18 (0.31) 0.20 (0.13) .001
Articulation 2.75 (0.32) 0.50 (0.22) .001
Rate 2.81 (0.27) 0.30 (0.15) .001
Intelligibility 2.62 (0.28) 0.30 (0.15) .001
both speech and language parameters. All five dysarthria measures, writ-
ing mechanics, and speech melody were all significantly more abnormal
in PD. Phrase length and grammatical complexity were also more dis-
turbed in PD than in DAT. Linguistic compromise, on the other hand, was
more evident in DAT patients, who had significantly poorer scores on
information content of spontaneous speech, confrontation naming, and
word list generation.
Multi-Infarct Dementia
MID and DAT patient groups did not differ significantly in age or
dementia severity (Table 4). Two language and four speech parameters
were significantly different in the two groups. Information content of
spontaneous speech and confrontation naming were more abnormal in
DAT, whereas speech pitch, melody, articulation, and rate were more
disturbed in MID.
The most abnormal language parameters in MID (scale score> 1.5/6)
were narrative writing, writing to dictation, writing mechanics, comple-
tion of nursery rhymes, and comprehension of complex auditory com-
mands.
MID patients were also compared with PD patients with overt demen-
tia. The two groups were not significantly different in age or MMSE scores
TABLE 4. Demographic Features and Distinguishing Speech

and Language Characteristics of Multi-Infarct Dementia (MID) and Dementia
of the Alzheimer Type (DAT)
Bonferroni
MID score DAT score significance
(SEM) (SEM) level
Patient characteristics
N 18 14
Age 67.38 (2.45) 73.14 (2.61) n.s.
MMSE 15.71 (1.31) 18.77 (0.95) n.s.
Information content 0.72 (0.25) 2.92 (0.24) .001
Naming 0.77 (0.27) 2.28 (0.45) .05
Pitch 1.83 (0.39) 0.35 (0.16) .01
Melody 2.67 (0.45) 0.57 (0.17) .05
Articulation 1.88 (0.40) 0.50 (0.17) .05
Rate 1.72 (0.27) 0.28 (0.12) .05
(Table 5). PD patients exhibited significantly more severe motor distur-

bances, including speech volume, speech rate, and writing mechanics. In
addition, their spontaneous speech had shorter phrase lengths and they
stuttered more. MID patients had more difficulty completing nursery
rhymes. Classical language functions such as naming, comprehension,
repetition, and reading did not distinguish these two disorders and were
relatively preserved in both conditions.
TABLE 5. Demographic Features and Distinguishing Speech

and Language Characteristics of Multi-Infarct Dementia (MID)
and Parkinson's Disease (PD) with Overt Dementia
Bonferroni
MID score PD score significance
(SEM) (SEM) level
Patient characteristics
N 18 17
Age 67.38 (2.95) 71.94 (0.91) n.s.
MMSE 18.77 (0.95) 19.70 (0.91) n.s.
Phrase length 1.05 (0.36) 2.29 (0.29) .05
Nursery rhyme completion 2.44 (0.47) 0.81 (0.39) .05
Loudness 1.77 (0.47) 3.64 (0.32) .01
Rate 1.22 (0.23) 2.76 (0.26) .001
Stuttering 0.00 (0.00) 0.70 (0.25) .05
Writing mechanics 1.83 (0.44) 3.70 (0.36) .01
DISCUSSION
These investigations demonstrate that speech and language observa-

tions can distinguish between dementia syndromes even when the patient
groups have dementia of equivalent severity. In addition, DAT patients
show significantly more disturbed language functions than either MID or
PD-dementia patients.
All DAT patients could be distinguished from normal control sub-
jects by their performance on the language tasks utilized in the study. The
language abnormalities of DAT resembled transcortical sensory aphasia
(Benson, 1979; Cummings et a1., 1985): fluent verbal output (impover-
ished information content of spontaneous speech but preserved phrase
length, grammatical complexity, and melodic line), poor auditory com-
prehension, and relative preservation of repetition abilities. Like transcor-
tical sensory aphasia, DAT also produced impairment of reading com-
prehension, while reading aloud was strikingly resistant to the effects of
declining intellectual function. DAT patients differed from the classical
transcortical sensory aphasics in their lack of completion abilities (e.g.,
completing the final line of familiar nursery rhymes read by the exam-
iner), relative absence of paraphasia and echolalia, and more impaired
automatic speech (e.g., alphabet recitation). Similar observations have
been made by others studying the verbal output of DAT, including
Appell et a1. (1982), Bayles and Tomoeda (1983), and Martin and Fedio
(1983).
Further progression of DAT is marked by evolution from the transcor-
tical sensory aphasia pattern to a more global linguistic impairment. Com-
prehension and repetition abilities are increasingly disturbed, as in Wer-
nicke's aphasia, but elements of fluency such as phrase length and speech
melody become impaired and elements of motor speech disorder appear.
Such alterations may reflect extension of the disease beyond its initial
focus to involve other brain regions (Cummings & Benson, 1983).
The pattern of verbal output in DAT indicates that the disease process
does not produce a global language loss. Syntactical and phonological
aspects of verbalization are preserved until the terminal phases, whereas
the semantic and pragmatic aspects of language are disturbed early and
are progressively more compromised as the disease progresses. Semantic
abilities appear most dependent on posterior cortical structures where
DAT has its greatest biochemical, histopathologic, and metabolic impact
(Brun & Gustafson, 1978; Cummings & Benson, 1983). In contrast, syntac-
tical elements of language may be mediated by subcortical-frontal systems
less affected in DAT.
Observations on reading also contribute to understanding DAT. Al-
though agnosia is frequently described as an early consequence of DAT,
the preservation of oral reading skills indicates that competency in graph-
eme-morpheme translation is spared until late in the disease course.

Moreover, the oral reading of DAT patients is not foiled by presenting
incomplete letters or partially obscured letters and words-techniques
that impair the reading of patients with visual agnosia (Cummings & Ben-
son, 1986; Landis, Graves, Benson, & Hebben, 1982). At most, agnosia is
confined to the late stages of DAT.
Two principal explanations have been proposed for the intellectual
loss of DAT: (1) that DAT produces instrumental deficits similar to those
observed in patients with structural damage underlying vascular aphasia,
and (2) that DAT patients undergo a progressive dedifferentiation (regres-
sion) of intellectual abilities to the preliterate stages of childhood and
infancy (Emery & Emery, 1983). The intact oral reading of irregular words
on the New Adult Reading Test (Nelson & O'Connell, 1978) by DAT pa-
tients until severe dementia supervenes and the relative rarity of phonetic
errors of the type noted during the acquisition of reading skills is more
consistent with an instrumental deficit than a dedifferentiation model of
DAT (Cummings & Benson, 1986).
The relationship between DAT and the dementia of PD is controver-
sial, and language studies of the type reported may have an impact on this
question. The observation that some PD patients with dementia have neu-
ritic plaques and neurofibrillary tangles in the cerebral cortex suggested
that DAT and the dementia of PD may represent the same process (Boller,
Mizutani, Roessmann, & Gambetti, 1980; Gaspar & Gray, 1984, Hakim &
Mathieson, 1979). In addition to histopathologic similarities, some pa-
tients with PD also have atrophy of the nucleus basalis of Meynert similar
to that found in DAT patients (Arendt, Bigl, Arendt, & Tennstedt, 1983;
Whitehouse, Hedreen, White, & Price, 1983). However, distinct neuropsy-
chologic deficits have been associated with subcortical dysfunction, and
the dementia of PD may be associated with the subcortical alterations
classically occurring in PD (Albert, 1978; Cummings & Benson, 1984;
Huber, Shuttleworth, & Paulson, 1986). Language alterations are promi-
nent in DAT, and if the dementia of PD is a product of DAT, similar
language defects should be present in PD-dementia. The findings de-
scribed here demonstrate that the two disorders produce different speech
and language changes; in fact, language skills are comparatively spared in
the dementia of PD. DAT patients had significantly less information con-
tent in their conversational speech ("empty speech"), showed more
anomia, and had greater difficulty with word list generation. They also
tended to have more problems with auditory comprehension, although
the differences did not reach statistical significance. PD patients with
equally severe dementia, on the other hand, had more profound dys-
arthria and disturbances of phrase length, melody, and grammatical com-
plexity. PD patients with dementia did not manifest an aphasic syndrome
but suffered severe motor speech disorders. These observations indicate
that PD-dementia and DAT have distinguishable speech and language

profiles, and, on a clinical basis, it is unlikely that the two dementias
represent the same disease process.
Naming and comprehension disturbances were present in some indi-
viduals with PD-dementia, and those patients with naming errors ac-
counted for most of the comprehension errors. It could be conjectured that
the subset of patients with PD-dementia and language errors may be suf-
fering from an additional disorder, such as DAT. Alternatively, severe
disturbances of subcortical processes may be sufficient to interfere with
naming and comprehension (Alexander & LoVerme, 1980; Brunner,
Kornhuber, Seemuller, Sugar, & Wellesch, 1982; Damasio, Damasio,
Rizzo, Varney, & Gersh, 1982).
PD-dementia patients had elements of a nonfluent verbal output, in-
cluding shortened phrase length, loss of speech melody, and diminished
grammatical complexity of spontaneous verbalizations. Thus, PD-demen-
tia patients had more syntactic than semantic abnormalities, a pattern
opposite that of DAT. These syntactic disturbances may be a product of
the motor abnormalities of PD, but it is also possible that subcortical
nuclei and projections have an important role in mediating syntactic
functions.
MID could be distinguished from both DAT and PD on the basis of its
speech and language profile. Compared with MID patients, those with
DAT were more anomic and had significantly less information content of
spontaneous speech. MID, on the other hand, caused more motor speech
abnormalities than DAT. These distinguishing features resemble the char-
acteristics that distinguished DAT from PD-dementia. In general, MID
patients had near-normal performances on the language subtests and did
not manifest identifiable aphasic syndromes. They tended to produce
grammatically simple and short phrases, the "Brocoid" output described
by Hier, Hagenlocker, and Schindler (1985) in patients with stroke-related
dementia. These investigators observed that, compared with DAT, MID
patients exhibited decreased syntactic complexity, shortened utterance
length, diminished total and unique words, and increased sentence frag-
mentation. They also noted that empty speech and anomia were more
prominent in DAT. Together, these observations suggest that aspects of
naming, characteristics of spontaneous speech, and dysarthria are impor-
tant features that distinguish MID from DAT.
MID is a clinically heterogeneous disorder with different syndromes
dependent upon the different brain regions involved by infarction.
Among the patients studied here, a wide range of deficit severity was
present on speech and language subtests. Most MID patients also man-
ifested elements of dysarthria; motor speech abnormalities appear com-
mon to MID patients regardless of the linguistic and neuropsychological
diversity within the syndrome. The prevalence of motor deficits may
reflect the high frequency of subcortical lacunar infarctions involving

descending motor tracts (Hachinski, Lassen, & Marshall, 1974; Hughes,
Dodgson, & MacLennan, 1954; Roman, 1985).
PD-dementia and MID could be differentiated primarily by the con-
siderably greater motor disability of the former. PD-dementia patients had
shorter phrase lengths, lower voice volume, slower speech, more stutter-
ing, and poorer writing mechanics than MID patients. PD-dementia pa-
tients, however, were better able to complete partially presented nursery
rhymes. Language functions such as naming, comprehension, reading,
narrative writing, word list generation, and repetition did not distinguish
the two disorders since they were relatively spared in both conditions
within the range of dementia severity studied. The similarity of the
speech and language profiles of these two conditions supports the sug-
gestion that both involve dysfunction of common subcortical sites.
CONCLUSIONS
Systematic assessment of speech and language abilities reveals differ-

ent profiles of preserved and impaired functions in each dementia syn-
drome investigated. Differences in the patterns cannot be attributed to
dementia severity since patients with equivalent degrees of intellectual
deterioration were studied. DAT patients manifested significantly more
language and significantly less motor speech impairment than patients
with PD-dementia or MID. DAT and PD-dementia patients had different
patterns of linguistic compromise, suggesting that DAT is not the cause of
all PD-dementia. MID patients had motor speech abnormalities in com-
mon despite the heterogeneity of their disorder. DAT had its greatest
impact on the semantic aspects of language; PD-dementia and MID had
more effect on syntactic and phonologic functions. Thus, speech and
language assessment provides valuable information for the differentiation
of dementia syndromes, reveals different patterns of abnormalities in dif-
ferent dementing conditions, and contributes to the understanding of
brain-behavior relationships within the dementias.
ACKNOWLEDGMENTS
This project was supported by the Veterans Administration, the John

Douglas French Foundation, and the Los Angeles Chapter of the Alz-
heimer's Disease and Related Disorders Association. The authors are in-
debted to the following Neurobehavior Fellows who contributed to data
collection: Sandra Horowitz, Michael Frankel, Stephen Read, Mario Men-
dez, Adam Darkins, and Artiss Powell. They also thank Mary Ann Hill,
Ph.D., for her aid with the statistical analyses, and Bonita Porch and
Wendy Kashitani for preparing the manuscript.
REFERENCES
Adams, R. D., & Victor, M. (1977). Principles of neurology. New York: McGraw-Hill.
Albert, M. L. (1978). Subcortical dementia. In R. Katzman, R. D. Terry, & K. L. Bick (Eds.),
Alzheimer's disease: Senile dementia and related disorders (pp. 173-180). New York:
Raven Press.
Alexander, M. P., LoVerme, & S. R., Jr. (1980). Aphasia after left hemisphere intracerebral
hemorrhage. Neurology, 35, 1193-1202.
American Psychiatric Association. (1980). Diagnostic and statistical manual of mental dis-
orders (3rd ed.). Washington, DC: Author.
Appell, J., Kertesz, A., & Fisman, M. (1982). A study of language functioning in Alzheimer
patients. Brain and Language, 17,73-91.
Arendt, T., Bigl. V., Arendt, A., & Tennstedt, A. (1983). Loss of neurons in the nucleus
basalis of Meynert in Alzheimer's disease, paralysis agitans, and Korsakoff's disease.
Acta Neuropathologica, 61, 101-108.
Bayles, K. A., & Tomoeda, C. K. (1983). Confrontation naming impairment in dementia.
Brain and Language, 19, 98-114.
Benson, D. F. (1978). Amnesia. Southern Medical Journal, 71, 1221-1228.
Benson, D. F. (1979). Aphasia, alexia, and agraphia. New York: Churchill Livingstone.
Benson, D. F., & Geschwind, N. (1985). The aphasias and related disturbances. In A. B. Baker
& R. J. Joynt (Eds.), Clinical neurology (Vol. 1). Philadelphia: Harper & Row.
Boller, F., Mizutani, T., Roessmann, U., & Gambetti, P. (1980). Parkinson disease, dementia,
and Alzheimer disease: Clinicopathological correlations. Annals of Neurology, 7, 329-
335.
Brun, A., & Gustafson, L. (1978). Limbic lobe involvement in presenile dementia. Arch
Psychiatr Nervenkr, 226, 79-93.
Brunner, R. J., Kornhuber, H. H., Seemuller, E., Sugar, G., & Wellesch, C. W. (1982). Basal
ganglia participation in language pathology. Brain and Language, 16, 281-299.
Caine, E. D., Ebert, M. H., & Weingartner, H. (1977). An outline for the analysis of dementia.
The memory disorder of Huntington's disease. Neurology, 27, 1087-1092.
Cummings, J. L. (1986). Subcortical dementia: Neuropsychology, neuropsychiatry, and pa-
thophysiology. British Journal of Psychiatry, 149, 682-697.
Cummings, J. L., & Benson, D. F. (1983). Dementia: A clinical approach. Boston: Butter-
worths.
Cummings, J. L., & Benson, D. F. (1986). Dementia of the Alzheimer type. An inventory of
diagnostic clinical features. Journal of the American Geriatrics Society, 34, 12-19.
Cummings, J. L., & Benson, D. F. (1984). Subcortical dementia: Review of an emerging
concept. Archives of Neurology, 41, 874-879.
Cummings, J. L., Benson, D. F., Hill, M. A., & Read, S. (1985). Aphasia in dementia of the
Alzheimer type. Neurology, 35, 394-397.
Cummings, J. L., Houlihan, J. P., & Hill, M. A. (1986). The pattern of reading deterioration in
dementia of the Alzheimer type: Observation and implications. Brain and Language, 19,
315-323.
Damasio, A. R., Damasio, H., Rizzo, M., Varney, M., & Gersh, F. (1982). Aphasia with
nonhemorrhagic lesions of the basal ganglia and internal capsule. Archives of Neu-
rology, 39, 15-20.
Damasio, A. R., Graff-Radford, N. R, Eslinger, P. J., Damasio, H., & Kassell, N. (1985).
Amnesia following basal forebrain lesions. Archives of Neurology, 42, 263-271.
Dixon, W. J., Brown, M. B., Engleman, 1., & Hill, M. A. (1985). BMDP statistical software
(2nd ed.). Los Angeles: University of California Press.
Emery, O. B., & Emery, P. E. (1983). Language in senile dementia of the Alzheimer type.
Psychiatry Journal of the University of Ottawa, 8, 169-178.
Folstein, M. F., Folstein, S. E., & McHugh, P. R (1975). Mini-mental state. Journal of Psychi-
atric Research, 12, 189-198.
Freedman, M., Rivoira, P., Butters, N., Sax, D. S., & Feldman, R G. (1984). Retrograde
amnesia in Parkinson's disease. Canadian Journal of Neurological Science, 11, 297-
301.
Gaspar, P., & Gray, F. (1984). Dementia in idiopathic Parkinson's disease. Acta Neu-
ropathologica, 64, 43-52.
Goodglass, H., & Kaplan, E. (1976). The assessment of aphasia and related disorders.
Philadelphia: Lea and Febiger.
Hachinski, V. C., Iliff, 1. D., Zilkha, E., Du Boulay, G. H., McAllister, V. 1., Marshall, J.,
Russell, R W. R, & Symon, L. (1975). Cerebral blood flow in dementia. Archives of
Neurology, 32, 632-637.
Hachinski, V. C., Lassen, N. A., & Marshall, J. (1974). Multi-infarct dementia. Lancet, 2, 207-
210.
Hakim, A. M., & Mathieson, G. (1979). Dementia in Parkinson disease: A neuropathologic
Hier, D. B., Hagenlocker, K., & Shindler, A. G. (1985). Language disintegration in dementia:
Effects of etiology and severity. Brain and Language, 25, 117-133.
Huber, S. J., Shuttleworth, E. c., & Paulson, G. W. (1986). Dementia in Parkinson's disease.
Archives of Neurology, 43, 987-990.
Hughes, W., Dodgson, M. C. H., & MacLennan, D. C. (1954). Chronic cerebral hypertensive
disease. Lancet, 2, 770-774.
Kertesz, A. (1979). Aphasia and associated disorders: Taxonomy, localization, and recov-
ery. New York: Grune & Stratton.
Landis, T., Graves, R, Benson, D. F., & Hebben, N. (1982). Visual recognition through
kinaesthetic mediation. Psychological Medicine, 12, 4515-531.
Martin, A., & Fedio, P. (1983). Word production and comprehension in Alzheimer's disease:
The breakdown of semantic knowledge. Brain and Language, 19, 124-141.
McKhann, G., Drachman, D., Folstein, M., Katzman, R, Price, D., & Stadlan, E. M. (1984).
Clinical diagnosis of Alzheimer's disease: Report of the NINCDS-ADRDA Work Group
under the auspices of Department of Health and Human Services Task Force on Alz-
heimer's Disease. Neurology, 34, 939-944.
Nelson, H. E., O'Connell, A. (1978). Dementia: The estimation of premorbid intelligence
using the new Adult Reading Test. Cortex, 14, 234-244.
Roman, G. C. (1985). Lacunar dementia. In J. T. Hutton & A. D. Kenny (Eds.), Senile dementia
of the Alzheimer type (pp. 131-151). New York: Alan R Liss.
Whitehouse, P. J., Hedreen, J. c., White, C. 1., III, & Price, D. 1. (1983). Basal forebrain
neurons in the dementia of Parkinson disease. Annals of Neurology, 13, 243-248.
Wilson, R S., Kaszniak, A. W., & Fox, J. H. (1981). Remote memory in senile dementia.
Cortex, 17,41-48.
III
Oral and Written Language

Disorders
7
Mechanisms Underlying Aphasic

Transformations
HUGH W. BUCKINGHAM
This chapter treats the various "mechanisms" that may be said to under-
lie aphasic transformations. I have quoted the word mechanisms for a
reason, since that term implies some sort of device (actually, another
metaphor) or internal agent that does things. Whatever these mechanisms
are in their neurophysiological instantiations, they must be understood
psychologically as manipulating some types of elements in some domain
of operation. Without first attempting to characterize what the mecha-
nisms might be at a psycholinguistic level of explanation and what they
might do, we will not be in any position to begin to characterize aphasic
transformations. That is, in order to start I must layout a production
model of some sort that can justifiably be said to form part of the cognitive
system for language-a model constructed from evidence external to the
data from aphasia.
The justification for the type of model I intend to use comes from
modern studies in the philosophy of psychology (e.g., Fodor, 1983),
where the claim (p. 29} is made that "contemporary cognitive theory takes
it for granted that the paradigmatic psychological process is a sequence of
transformations of mental representations and that the paradigmatic cog-
nitive system is one which effects such transformations." Like Fodor, I
will assume for the purposes of the present chapter that if I am going to
speak of psychological mechanisms, then I must assume that these mech-
anisms will be computational systems of one sort or another. Concurring
further with Fodor (1982, p. 279), I will assume that these computational
processes are both symbolic and formal. The processes are termed sym-
HUGH W. BUCKINGHAM· Program in Linguistics and Communication Disorders, Loui-

siana State University, Baton Rouge, Louisiana 70803.
123
124 HUGH W. BUCKINGHAM
bolic because they are defined over (operate upon) representations that
consist of symbolic elements. The processes are termed formal because
they apply to the various representations in virtue of the configuration or
syntax of those representations. The formal operations apply, as it were,
to various types of symbols without "understanding" anything about
what those symbols mayor may not mean. Or, to put it another way, the
operations manipulate the symbols in terms of their shapes, not their
meanings.
Once we spell out the levels of representation and the various types
of computations that map one level onto another, we will be in a position
to consider what types of derailments of the computations could occur to
account for the production of aphasic transformations.
THE GARRETT MODEL
Over the past 10 years or so, Merrill Garrett (1975, 1976, 1980) has
been developing a model for sentence production based on data he and
others have collected from the slips of the tongue of normal, non-brain-
damaged subjects. More recently Garrett (1982, 1984) has shown how his
model may be used to characterize aphasic syndromes. Schwartz (1987)
and Buckingham (1980, 1985, 1986, 1987) have incorporated and ex-
tended Garrett's model in explaining their aphasic data. Garrett's publica-
tions (1982, 1984), however, have ensured that his model will playa
major role in psycho linguistic accounts of the aphasias for some time to
come. That there is, indeed, a clear relation between slips of the tongue
and many types of aphasic errors has been indicated in Cutler and Fay's
(1978) introduction to the reissue of the landmark study of Meringer and
Mayer (1895) and in Buckingham (1980). It is therefore not without prece-
dent that I utilize a model such as Garrett's to characterize the computa-
tional and representational breakdowns seen in aphasic language.
The overall model of Garrett (Garrett, 1984, p. 174) consists of in-
ferential processes that map conceptual structures onto a message level
representation, which is viewed by Garrett as a real-time construct that
determines all subsequent sentence level production. It is in this sense
that all subsequent mental processes are "executive driven" (Fodor, 1986,
p. 81). At these higher levels, which, parenthetically, have not been ar-
rived at through data from slips of the tongue but rather through philo-
sophical argument (e.g., Searle, 1983), intentionality is the key, and since
"executive control" is established at the message level, it may be said that
"intentionality is the key to the mental" (Fodor, 1986, p. 81). Again, the
message level is based upon the speaker's current perceptual and affective
states of mind and on his general knowledge of how the world is. So-
called scripts and plans as well as other schemata are crucial in the me-
APHASIC TRANSFORMATIONS 125
morial structures pertaining to knowledge of the world, and are the sine
qua non for characterizing message level representations. In addition,
representations of the message level must be composed of a basic sym-
bology of simple concepts and some sort of syntax. That the representa-
tions are composed of symbols in an arrangement follows from the re-
quirement that computational processes be both symbolic and formal. It is
also important to emphasize that there is a syntax at every level of
representation.
Mapping from the message level to the functional level in Garrett's
model (e.g., Garrett, 1984, p. 176) is realized through what he calls logical
and syntactic processes. There are three principal computations here.
First, there is a determination of functional level structures. Although
they are not the same, as Garrett points out, these functional level struc-
tures are strikingly similar to the "deep case" structures of Fillmore
(1968), and they are couched in f(x) notation consisting of atomic predi-
cates and arguments. The second computation is that of a meaning-based
lexical identification. The computations that operate here access the lex-
icon solely in terms of meaning,l but a "linking address is set up here that
establishes an association with the shape of the word in question. The
shape here, however, may be very abstract (see Caplan, 1987, pp. 246-
247). Moreover, it should be pointed out that these computations are very
fast and may initially access all the meanings of some word or perhaps
initially all meaning-related words. The meaning "sphere" or semantic
field may be what is pointed to, and initially the linking addresses of all
members of the sphere apparently present themselves in a rapid, and
almost reflexlike, fashion (e.g., Swinney, 1982). Derailments here would
ultimately cause incorrect selections of the shapes of words that were
nevertheless in the semantic sphere of the target, and consequently it is
within these computations that one would locate the mechanism of se-
mantic paraphasia. Meaning-related, lexical slips of the tongue are lo-
cated precisely here by Garrett. The third computation in the message to
functional mapping is the assignment of the semantically selected lexical
items to functional structures. The Functional Level structure is thereby
computed and the representation is a syntactic one but, again, a very
IGiven Fodor's stipulation that computations manipulate symbols only in terms of their
shapes, we may question the status of the operations involved in the first, or meaning-
based, lexical lookup at the functional level. Is there any essential computational difference
between "appreciating" the meaning of a word and "selecting" that word from the mental
lexicon on the basis of its phonological shape? It is certainly logical that a word can be
"selected" on the basis of its meaning, but how is this carried off as a computation if that
selection does not involve symbol manipulation. "Symbol" implies form as well as mean-
ing. So, it would appear that even at the functional level where the selection processes do
not have access to the phonological shape of a word, those processes do have access to the
word in terms of its semantic symbolic form. Phonological form is connected to the first
lexical lookup computations through the "linking address."
abstract syntax cast in the atomic predicate-argument form of deep case

structure. It still must be viewed, however, as a real-time construct, ren-
dering it quite unlike the deep case structural level in Fillmore's theory.
Syntactic and phonological processes then work off of the functional
level representation and map it onto Garrett's positional level representa-
tion (e.g., Garrett, 1984, p. 178). These processes are exceedingly impor-
tant, since evidence to date indicates that many aphasic transformations
are the result of derailments that would appear to involve them. Accord-
ing to Garrett, there are four processes here. Later on I will augment the
list of computations in this domain, bringing to bear mechanisms pro-
posed by Butterworth (1979) and by Shattuck-Hufnagel (1979', 1983).
The first of the positional level computations 2 construct phrasal
frames or matrices, which carry superficial constituent structure informa-
tion, markings for functor morphemes, and grammatically positioned
empty slots for the contentives (nouns, verbs, adjectives, and adverbs).
Phrasal intonational contours are also coded with the frames.
The second computation feeding into the positional level is ex-
tremely important and involves the retrieval of lexical items in terms of
the form or shape of those items. Thus, retrieval here is distinguished
from functional level retrieval, which, as we saw, involved selection ex-
clusively on the basis of meaning. The "linking address," recall, will have
previously been established at the functional level retrieval in order to
assure that in the normal situation, the form-based retrieval will match
the meaning-based retrieval. That is, if book has been retrieved initially in
terms of its meaning, we want to ensure that that item's formal counter-
part will subsequently be accessed in terms of its phonological shape.
Ultimately, both retrieval computations must operate correctly to produce
the proper target. Data from slips of the tongue and aphasia demonstrate
that these two retrieval processes may be dissociated-one operating, the
other failing. Words may be appreciated in terms of their meaning, while
their forms are completely inaccessible. Forms of words may be accessed
at times when there is absolutely no appreciation of their meaning.
The third computation is the actual assignment of the form-retrieved
lexical items to their respective frame slots. It is important to point out
that the computations here are also manipulating the segmental units of
the words as well and are readying them for utterance order. That is, at
some point, the representative shapes of words as they exist in the lexicon
must be, in a sense, reordered for real-time production. I view this reor-
dering as taking place in parallel with the assignment of the whole word
to its phrasal frame slot; however, in order to discuss these additional
2We call them positional level computations, but in a strict sense only representations are
understood to exist at the levels in Garrett's model. Computational processes map levels
onto levels and they exist between the levels, not at them. Often in the wording of the
operations in this model, by x level processes we mean that map onto level x.
computations, I will borrow from Shattuck-Hufnagel, since Garrett in his

original work did not consider segmental ordering manipulation in any
detail. Later on, I will further augment the computational structure at this
location by adding a mechanism that will supply phonological syllabic
forms when no regular target forms can be accessed, and subsequently
these "de novo" forms will be assigned to the empty contentive slots. The
notion of a separate segmental "generator" comes from Butterworth
(1979).
The fourth computation leading to the positional level of reprEisenta-
tion is one that assigns the frame elements (the functors, both bound and
free) to their respective positions in the lexically interpreted phrasal
frames. These computations place items such as affixal morphemes, de-
terminers, and the like onto their positions. Again, data from slips of the
tongue as well as aphasia demonstrate that the processes that assign con-
tentives to their slots and the processes that assign functor morphemes to
their slots can be separately disrupted, and thus those computations are
functionally dissociated.
After these computations have operated, we arrive at the positional
level of representation, where the utterance order is more directly re-
flected. The representation here is phonological, and, although it is a level
in a real-time model of sentence production, it closely parallels the "sys-
tematic phonetic" level in the sense of Chomsky and Halle (1968).
The basic computations that map from the positional level to the
phonetic level are akin to the so-called regular phonological processes,
and these essentially establish the allophonic and allomorphic manifesta-
tions of the phonemic and morphemic units as they are represented at the
positional level. The computations that map the positional onto the pho-
netic are clearly dissociated from the positional level process just dis-
cussed. There is a wealth of data from slips of the tongue and from aphasia
that demonstrates derailments with the later and none with the former.
That is, the literature on aphasia or slips of the tongue is replete with
examples of errors where, despite what has occurred above, all the al-
lophonics and allomorphics are computed correctly. If, for example, a Ipl
slips out of a nonaspirated position into a syllable initial position in a
stressed syllable, its allophonic realization will be aspirated. In the error
papple, the unaspirated Ipl in the target, will pick up its aspirated al-
lophone in the initial position of the error. If the past-tense morpheme
moves to a new slot, such as in he added ten spoken as he add tened, the
past-tense allomorph will be I-d/, instead of the I-:-Jdl after add.
A lower set of computations will then map the phonetic level of
representation onto the articulatory level, off of which are computed the
motor commands to the articulators. The phonetic-to-articulatory pro-
cesses compute ever finer phonetic detail (below that of the allophones
and allomorphs) and construct units that will more naturally feed into
actual motor commands. The computations that run off of the phonetic
level take into consideration all other nonlinguistic contextual factors of
articulation that exist in each speaking situation, factors such as tempo-
rary obstructions in speech cavities (such as bite blocks and chewing
gum), attentional states, and other factors that do not playa role in the
underlying linguistic system but which must be accounted for on each
speaking occasion, since the information is nevertheless relevant for
motor production.
For the purposes of this chapter, however, we need only focus upon
the functional and positional levels of representations and the computa-
tions that construct them. I will now turn to the various additions to the
Garrett model that I feel are necessary for a more complete picture of the
computations between the functional and positional levels of that model.
THE AUGMENTATIONS OF THE GARRETT MODEL

As additions to the syntactic and phonological processes that map
onto the positional level, I will propose two mechanisms and a buffer
from the world of Stephanie Shattuck-Hufnagel (1979, 1983). On Shat-
tuck-Hufnagel's view, phonological forms are retrieved from the lexicon
and placed into a buffer that can hold up to a clause. There is some sort of
interlocking between this buffer and the initially constructed phrasal
frames of Garrett, since the content words operated upon segmentally are
destined for the empty slots designated by the frames. The words are
placed in the buffer in their representative shapes, whereupon a device
termed the scan copier scans the segments by syllable position and copies
those segments into their productive orders. In the normal case, the pro-
ductive order will be quite similar to the order of the lexical segments as
they are stored in their representative shapes. Obviously, a somewhat less
abstract phonology will work better here, but where the dictates from
underlying systems force a more abstract analysis, production models
such as Garrett's will have to settle for more recondite representative
forms and proceed accordingly.
Once a segment is copied onto its productive order position, it must
be eliminated from the buffer in order to avoid repetitive copying, which
would lead to copious reiteration (or perseveration). Accordingly, Shat-
tuck-Hufnagel has suggested a mechanism she calls the "checkoff
monitor." Once the segment is copied, the monitor "checks it off." The
scan copier operates in accordance with the syntax of the phonological
material, which means that it operates over the syllabic structure of repre-
sentative forms, and it copies segments onto a syllabically coded produc-
tive order structure.
To fully understand this, we must take a closer look at the constituent
structure of the syllable. The model I incorporate goes back to the work of
Pike and Pike (1947) and is further discussed in Bell and Hooper (1978).
To begin with, the syllable bifurcates into an "onset" and a "core," as
illustrated in Figure 1. The onset consists of any initial consonants of the
syllable. The core bifurcates into a "peak" and a "coda." The peak is the
syllabic element that may carry stress, and is thus usually the slot for the
vowels, although certain of the sonorant consonants in English may func-
tion as syllabics. The core consists of any terminal consonants assigned to
the syllable. Like the onset, the core may be empty. The only unit of the
syllable that must be filled is, of course, the peak. On this view, the syntax
of syllables is essentially binary. Coda consonants are more strongly glued
to the peak than are onset consonants, and there is much data to support
this (e.g., Shattuck-Hufnagel, 1987). We can now return to the scan copier.
It scans the segmental elements of the phonological forms in the
buffer by syllable position. That is, it scans onsets and cores, but within
the core there is some evidence that the scanner is sensitive to whether it
is manipulating a peak item or a coda. Slips of the tongue have instructed
us here. The literature on them is replete with examples where the slip
involves the movement of a segment to some other location in another
syllable. However, it has been noted that in practically every case, the
segment that switches syllables ends up in the same position as it oc-
cupied in its original syllable. That is, onsets move to onset positions,
S (yllable)
o (nset) C (ore)
c
~
012
c c .......... c
n
~
P (eak) C (ada)
(+ syllabic)
(+ stress) ~
C C C ...... C
o 1 2 n
FIGURE 1. A hierarchical representation of the major syllable constituents. This is essen-

tially a binary view of the structure of the syllable. Onsets and codas may be empty or may
consist of one or more consonants. depending upon the language. The specific arrangements
of the consonants that make up the onsets and codas are first conditioned by the universal
principle of sonority. Language-specific phonotactic constraints then account for slightly
different possibilities for consonantal sequences in the onsets and codas for different lan-
guages. The range of phonotactic differences among languages is. however. highly con-
strained by the overriding universal principle of sonority.
peaks go only to positions for peaks, and codas move to coda positions.
The only recent finding is that ambisyllabic consonants have a greater
range of choice as to where they may end up, since by their very nature
they are ambivalent as to whether they are the coda of one syllable or the
onset of another (Buckingham, 1980; Stemberger, 1982). Consonants that
are unambiguously assigned to a syllable are called "tautosyllabic." The
scanner, then, scans and copies segments by syllable position. If it mis-
scans, then some onset that occurs later on in the form will be copied in
an earlier onset position. At times there will be an exchange of segments,
and at other times either the original segment will be deleted by the
checkoff monitor or it will remain in its original position-the checkoff
monitor failing to check the misscanned segment off from its original slot.
In this latter case, the so-called doublet error will be the result (see
Lecours & Lhermitte, 1969). In the end, all of the processes suggested by
Shattuck-Hufnagel will have to interact with the computations suggested
by Garrett at the positional level, since phonological shape and utterance
order are crucial at this point in the production process.
The next augmentation to the positional level processes in the Garrett
model is the so-called random generator suggested by Butterworth (1979).
Butterworth suggested this device on the basis of data from neologistic
jargonaphasia. That is, there appeared to be a need for getting some sort of
phonological material into the buffer in those cases where it seemed that
the speaker could not retrieve target words on the basis of their form. The
neologisms Butterworth analyzed consisted of phoneme aggregates that
did not follow the normal distributional characteristics of words in En-
glish. In fact, his statistical analysis showed them to be randomly dis-
tributed, not in terms of phonotactic arrangement but in terms of expec-
tancy. He therefore posited a "random generator" that would go into
operation at the point of blocked access to phonological shapes. Again, by
"random" Butterworth did not mean haphazard ordering of phonemes
but rather that the frequency-based expectancies for phoneme use were
not followed in the construction of the neologisms. The ability assumed
by this random generator is not so strange and should not be taken to be
some recondite capacity that results from brain damage (Buckingham,
1987). Merrill Garrett (1982, p. 46) discusses Butterworth's proposal in
the following manner: "Butterworth hypothesizes that the neologistic
forms are generated by jargon aphasics in response to a word retrieval
difficulty-rather than halt, the speaker generates a phonetically accept-
able slot filler using the phonological and morphological structure build-
ing systems normally at one's disposal."
In a footnote supporting the assumption that this random generator is
not some de novo creation by the brain lesion but rather part of the normal
cognitive system, Garrett (1982, p. 72) writes: "Any speaker can generate
nonsense forms on request, and everyone on occasion produces nonce
forms .... investigations of voluntary glossolalia are convincing in the

former regard."
Butterworth suggested that this device was a segment generator, but it
is more reasonable to suppose that it produces syllable-sized chunks of
phonological material. It is also likely that the syllable stock that this
device has recourse to abides by the phonotactic patterns of the language,
since, no matter how bizarre they are, neologisms do not in general dis-
obey the constraints that condition permissible syllable patterns of the
aphasic's language. Figure 2 illustrates how the scan copier, the checkoff
monitor, and the random generator fit in with Garrett's positional level
processes.
Having now characterized a model for normal sentence production,
we are in a position of analyzing the computational derailments that
produce aphasic transformations. I will begin with lexical phenomena.
LEXICAL SUBSTITUTIONS
Full lexical substitutions are triggered either on a meaning-similarity

base or on a form-similarity base. The point of the present chapter is not to
present a wealth of new data for these kinds of paraphasias but rather to
suggest the type of computational breakdown involved and the location of
the breakdown. As should be obvious by now, meaning-based lexical
errors and form-based lexical errors would take place at different points in
~EANING
LEXICON
FORM
SCAN COPIER
} ~SITIONAL ::YEL REP.
RANDOM
GENERATOR
PRODUCTIVE ORDER
SYLLABIC SLOTS I PHONETIC LEVEL REPRESEN)
FIGURE 2. An augmented schematic diagram of the Garrett model for sentence production
demonstrating the additions of the scan copier, the checkoff monitor, the random generator,
the working memory buffer for the serial ordering of segments, and the productive-order
syllabic slots, The diagram highlights the fact that these processes enter into the computa-
tions that operate between the functional level and the positional level, and as such they
feed into the positional leveL (From Buckingham, 1987, p, 382.)
the production process, which is to say that they occur in different com-
putational domains in a model such as Garrett's.
Meaning-based lexical errors take place owing to derailments of the
functional level computations of meaning-based selection (or apprecia-
tion). These so-called semantic paraphasias have been catalogued and
described in numerous publications, such as Buckingham (1981), Buck-
ingham and Rekart (1979), and Rinnert and Whitaker (1973). Recently,
disturbances of lexical semantic representations have been summarized
in Caplan (1987, chap. 12). The semantically related substitutive errors all
indicate that the subject nevertheless correctly arrives at the semantic
sphere of the word sought, so, in a sense, there really is no semahtic field
disturbance-at the level of the sphere itself. These spheres are many and
are discussed fully in Buckingham and Rekart (1979), in Rinnert and
Whitaker (1973), and in the references found in those works. For these
kinds of lexical substitutions, the linking address serves only to hold onto
the semantic field of the target word, to ensure that at the point of form-
based retrieval, the phonological shape found will be from the necessary
semantic area. Form-based retrieval need not be disrupted, and in fact
cannot be, if there is to be a semantic substitute for the target, since in an
obvious sense the error must nevertheless be accessed in terms of its
shape. Consequently, there may be a disruption with the linking address.
Anomias that are characterized by numerous semantic substitutions are
quite different in nature from anomias where no forms are accessed what-
soever. In addition, the semantic paraphasia does not necessarily take into
account phonological structure, since a typical error would be to sub-
stitute Thanksgiving for Easter, the two forms having little in common in
terms of their shapes. Caplan (1987, p. 212), however, observes that there
is often formal similarity between target words and their semantically
related substitutes. This need not be the case, however.
The network of connections in the lexicon is a web-shaped structure
that stretches in the horizontal as well as in the vertical plane. There are
vertical (paradigmatic, hyponymic relations as well as horizontal (syntag-
matic) modificationallattributive relations. These are semantic relations;
but there are form relations as well, since in some obvious sense words
with similar phonological makeup must be linked (see Fay & Cutler,
1977). And this in turn allows for the form-based selection errors. Form-
based selection derailments, in the Garrett scheme, take place at the posi-
tionallevel, well after meaning-based selection has been computed. This
does not mean, of course, that it would not be possible in some instances
to claim that some failure to appreciate semantic connectiveness could
give rise to form-based selection difficulties. It only means that form-
based selection disturbances can be dissociated from meaning-based ap-
preciation of words. There is ample evidence of this type of dissociation
(e.g., Buckingham, 1985).
The so-called malapropism is indicative of a form-based lexical sub-

stitution, but at least two types of malapropism must be distinguished.
The "classical" malapropism is a product of learning and is not consid-
ered to be an "on-line" slip or an aphasic derailment (Zwicky, 1982). The
sociolinguistically relevant malapropism is the classical one described in
Sheridan's novel The Rivals, where through lack of familiarity and use,
some word is misperceived (a sort of slip of the ear) and initially stored in
memory incorrectly. For example, the word discretion may be misper-
ceived (or perceived, but subsequently misanalyzed) and stored in the
lexicon as discrepancy. As a result, the speaker mayan some occasion say
of a boss that he often leaves decisions up to the discrepancy of his
employees. The classical malapropism is an advertent error. The "on-
line" malapropism, however (Fay & Cutler, 1977) is of a very different
nature and truly represents a selection error based upon shape similarity;
it is inadvertent. The phonologically based errors, therefore, will not be
meaning-related, and those who commit them will quite often catch them-
selves in the act and correct the error. Aphasics tend to produce fewer on-
line malapropisms than semantic paraphasias.
LEXICAL BLENDS
It can occur that two lexical items will be selected simultaneously,

the result being a blended form. Blends can take place both at the lexical
and at the phrasal level (Fay, 1981). In fact, slips of the tongue that had
been classified as derailed transformational processing turned out, upon
close inspection, to look more like the simultaneous blending of two-
semantically related-phrases. Lexical blends also involve two words
that are related semantically. For example, a blend of slightest and least
would result in the error form sleast (Ellis, 1980). Most often, the blended
error is not a word in the language, but interestingly enough it will con-
form to the phonotactics. Another example would be the blend of
gripping and grasping to yield grisping (Fay & Cutler, 1977). In some rare
cases, the error will match a word in the dictionary, but that word will not
bear the meaning relation that the two input items shared. Fay and Cutler
(1977) provide an example where the semantically related words heritage
and legacy blend to form heresy, most certainly unrelated to the seman-
tics of the blended items.
Phrasal blends would be characterized according to the Garrett model
as an example of a derailed positional level computation of phrasal frame
construction with concurrent lexical selections. For example, the error
when you boil down to it, would come from the blend of when you come
down to it and what it boils down to (Fay, 1981). Note that the two
blended phrases share very similar semantic properties. However, since
meaning is crucially involved, the similarity in meaning between the

blended forms must implicate the functional level as well. Many of the
fluent aphasics exhibit blended forms of this nature (Buckingham & Ker-
tesz, 1976, pp. 60-61).
Pure lexical blends most certainly argue for a functional level derail-
ment because those forms must share meaning, but the blending itself
involves the simultaneous form-based selection of two words. Blended
errors clearly argue for a more parallel processing view of the functional
and positional levels, which of course does not mean that the computa-
tions themselves are not encapsulated and cognitively impenetrable in
the sense of Fodor (1983). As indicated above, the linking addresses serve
as crucial mediating processes between functional and positional lexical
manipulation. Consequently, the lexical blend may be viewed as some
sort of coalescence of different linking addresses eminating from the same
semantic domain. In other words, there is very close interaction between
these two levels in the production of blends. They are coalescences of
shapes that share meaning.
Before leaving blends, I should point out that there are other sorts of
blends that have been discussed in the aphasia literature, but these result
from the perseveration of one of the forms to be blended. For example,
Buckingham, Whitaker, and Whitaker (1978) describe many instances
where some perseverated form will carryover into the production of some
subsequent response, blending in with that response. For example, as
discussed in that paper, a patient who was perseverating on the word
drink would contaminate other forms, such as cross and crack, such that
they were produced as dross and drack. In these cases, of course, the
blended forms will not share meaning. Perseverated blends, therefore, are
more likely to be derailments at the positional level and to involve items
that somehow get "stuck" in the buffer through computational derail-
ments of the checkoff monitor, thereby interfering with newly selected
items. So much for the whole-word derailments. I now turn my attention
to the disruption of computations that manipulate the segments of the
words.
PHONEMIC SUBSTITUTION
The nature of the phonemic substitution is not transparent, although

at the outset it would appear to be. The important question to ask is from
where does the substituting phoneme originate? For there to be a strict
phonemic substitution, some phoneme must be substituted for some other
phoneme-or so it would seem. However, even this apparently obvious
point is deceiving, since the substitution may not be of one phonemic unit
for another but rather simply a switch of one distinctive feature of some
phonemic unit, the result being the appearance of some other phonemic
unit that differs from the original by that one feature. In that event, the
substitution is really simply the reversal of some feature + or - marking.
Shattuck-Hufnagel and Klatt (1979) have argued that, at least for slips of
the tongue, accounting for phonemic substitutions in terms of distinctive
feature switches buys one very little, and that the proper way of viewing
these kinds of transformations is in terms of whole phonemic units sub-
stituting for whole phonemic units. In any event, usually only phonemes
that are similar in terms of their distinctive feature makeup substitute for
one another. Practically every investigator of phonemic substitutions has
pointed this out (e.g., Blumstein, 1973). The ambiguity in analysis (feature
vs. whole phoneme unit) is still a serious problem for the substitution of
phonemes that differ minimally in their feature array; however, it repre-
sents less of a problem for the substitution where the two phonemes
involved differ in several features, which is apparently the case for many
of the so-called conduction aphasics (see Caplan, 1987, p. 220). For sub-
stitutions involving two or three features, it would seem more likely that
the transformation involves whole phonemic units. The issue, however, is
still unresolved, and in many instances we simply cannot be sure whether
the substitution involves whole segments or switches in feature markings.
We now ask ourselves where in the production process could
phonemic substitutions take place. In terms of the combined model pre-
sented in this chapter, errors in the selection of unit phonemes would
appear minimally to occur somewhere during the manipulation of the
phonological structure of words and the copying of representative forms
onto levels more in line with articulatory production. It must be admitted,
however, that the investigator who proposed the scan-copier model of
phonological production (Le., Shattuck-Hufnagel) did not observe many
exclusively paradigmatic phonemic substitutions in her slip of the tongue
data and, accordingly, did not set up her model to readily account for
them. In fact, Shattuck-Hufnagel's findings lead her to the conclusion that
few if any strict phonemic substitutions occur-that is, phonemic sub-
stitutions that do not implicate linear ordering disruptions. Be that as it
may, the scanner could on occasion somehow select a closely related
segment (or perhaps not so closely related) and copy it onto the productive
level. Just how this may come about is not well defined from the analyses
provided by Shattuck-Hufnagel, since, again, she does not think that pure
phonemic substitutions account for much of her error data. This finding
from slip data is rather puzzling, given the fact that one of the most
detailed studies of phonemic paraphasias to date (Blumstein, 1973) as-
sumed that most of the phonemic paraphasias across syndromes (Broca,
Conduction, and Wernicke) were, in fact, phonemic substitutions.
To get around the puzzling "no-source" (purely paradigmatic)
phonemic substitution, Shattuck-Hufnagel (1979, pp. 317-318) offers sev-
eral suggestions. First, it may be the case that the source of the substitut-
ing phoneme was in part of the utterance not recorded or written down by
the investigator. Here we must recall that Shattuck-Hufnagel is talking
about data collection for slips of the tongue, not aphasia. Consequently,
this suggestion is bound to the research methodology. Most investigators,
given the very nature of the data they are collecting, do not accumulate
slip of the tongue samples through lengthy tape recordings, but rather in
one setting or another they overhear a slip and quickly try to write it down
with paper and pencil. They may, of course, fail to write down enough of
the utterance that surrounded the slip to spot a possible linear source of a
slip recorded as a pure phonemic substitution originally. For this reason,
Shattuck-Hufnagel's first alternative is a real possibility for slips of the
tongue. The aphasia researcher, on the other hand, is more likely to at
least audiotape-record the speech output of the patient. That being the
case, a fairly distant linear source for a phonemic substitution may be
somewhat easier to ascertain.
Second, it may happen that the information specifying a target seg-
ment is incompletely or incorrectly copied onto its syllabic slot by the
scanning mechanism. How this might occur is left undescribed by Shat-
tuck-Hufnagel. This account may simply reduce to the claim that the
phonemic units that get to the buffer are underspecified in their featural
makeup, and the error occurs during the computations that fill in the rest
of the feature matrix, as suggested by Beland and Nespoulous (1985).
A third possibility suggested by Shattuck-Hufnagel is that some com-
ponent of a segment (Le., a feature) in an analogous syllable slot down line
interferes with (contaminates) the segmental array currently being copied.
Admittedly, though, Shattuck-Hufnagel pays little attention to the feature
level of explanation for her error corpus.
A fourth alternative account considered by Shattuck-Hufnagel is a
more intriguing one. It is that the substituting segment's source is in a
word that somehow mistakenly presents itself as a competing or alter-
native element into the set of ongoing planning units. Competing plan
errors are discussed in detail in Baars (1980). This type of lexical account
for phonemic substitution errors is also considered by MacNeilage,
Hutchinson, and Lasater (1981).
Although I cannot go into much detail in this chapter, an additional
possible confounding factor involved in the observation of so many sup-
posedly nonsource phonemic substitutions could be that subtle artic-
ulatory asynchronies, which are not generally perceived without tech-
nical measurement, give rise to shifts in the acoustic spectrum. These
acoustic shifts may in some cases be affecting crucial perceptual cues for
the hearers. The altered acoustic cue may be such that the perceiver
"hears" another phonological unit. In that case, the hearer would make
the claim that the speaker substituted the phonemic unit that was actually
APHASIC TRANSFORMA nONS 137
constructed in the perceiver's own mind. This process has been referred
to as "phonemic false evaluation," and the clinical and theoretical reper-
cussions of it have been discussed in Buckingham and Yule (1987). Suf-
fice it to say that, consequently, in many cases, what have been tagged as
phonemic substitutions may in reality be substitutions in the minds of
hearers only-the speaker having selected the correct target phoneme
only to produce it with some sort of articulatory aberration, thereby shift-
ing the acoustic quality such that the hearer constructs another phoneme.
In this case, it would be incorrect to claim that the speaker selected the
wrong segment, and thus it would not be a case of phonemic substitution.
To the extent that Shattuck-Hufnagel's alternative accounts have some
feasibility and in fact could represent what is actually going on, then
those processes, together with presumed instances of phonemic false
evaluation, may be robust enough to rule out pure phonemic substitutions
altogether, for aphasia and for slips of the tongue.
PHONEMIC TRANSPOSITION
Phonemic transpositions, on the other hand, constitute a majority of

phonological errors in aphasia, and they mayor may not condition
phoneme substitutions. Transpositions of this sort involve linear ordering
derailments and are seen in great quantities in both slips of the tongue and
aphasia. Fromkin (1971), Garrett (1980), and Shattuck-Hufnagel (1979)
provide ample cases of linear segmental slips of the tongue, while Blum-
stein (1973), Buckingham and Kertesz (1976), Buckingham et al. (1978),
Kohn (1984, 1985), and Lecours and Lhermitte (1969) are but a few of the
many studies that have documented linear switches in aphasic patients.
Transpositional errors among segments are either anticipatory or per-
severative, or in some cases there may be complete exchanges, although
the latter type are seen much more in slips of the tongue than in aphasia.
Anticipatory errors may result in the addition of the anticipated segment
to the total stock of segments in the original word. For example, the error
papple for the word apple creates an additional Ipl to the total word
string. In other words, a doublet is created. The error non don for the word
London does not create an additional segment, but rather the anticipated
Inl substitutes for the III of the target. Note that in both these cases the
anticipated segment remains in its original slot. Not all anticipated seg-
ments will necessarily remain in their original positions. Whether they
will or will not remain in their original positions is usually conditioned
by syllable structure constraints. That is, if by its not appearing in its
original slot an unpermissible sequence of two segments would result, the
anticipated segment will most often remain in that position. If upon its
disappearing from that slot a permissible sequence still obtains, the antic-
ipated segment may delete from its original position in a word. Buck-
ingham (1987, 1989) details the constraints involved in whether doublets
are produced or not. We will briefly consider the nature of these con-
straints later on in this chapter.
Now it should be clear that the above transformations would most
likely occur as derailments with the scan copier and checkoff mecha-
nisms postulated earlier in this chapter. The scanner would erroneously
scan ahead a syllable or two down line and copy, for example, some
syllable onset too early. That onset could then be added to the stock of
segments of the word by filling an empty slot (see Beland & Nespoulous,
1985) or it could substitute for the original consonant onset. Moreover, the
anticipating segment mayor may not be deleted from its original position
by the checkoff monitor. That is, if there is a disruption of the checkoff
monitor, it will fail and the anticipated item will not be checked off. So, in
a sense, the checkoff monitor's operation appears to be constrained by
principles of syllable structure (again, see Buckingham, 1987, 1989, for
more details).
Virtually everything that has been said about anticipatory transposi-
tions can be said of perserverative transpositions, except in this case the
process works from left to right, whereas the anticipatory shifts went from
right to left. Perseverated segments may be added to the total stock of
segments of some word or they may substitute for some segment that is
down line. Furthermore, the perseverated segment mayor may not delete
from its original slot, again depending upon the sequence that would
remain. Here again, the scan copier would copy, for instance, an initial
onset consonant onto an onset position somewhere down line in the
word, or phrase, either adding another segment by filling an empty slot
(see Beland & Nespoulous, 1985) or by substituting for some other conso-
nant originally in that position. Once again, it appears that the derail-
ments that result in perseverative phonemic transpositions take place
during the operation of the scan copier mechanism.
Since practically all of the phonemic transpositions take place in the
segmental structuring of content words, it would appear that these errors
occur during the manipulation of the contentives. Furthermore, the ma-
nipulation must involve the form of the content words, since form is
crucially involved with phonemic errors. The forms of content words are
being uniquely manipulated as they are inserted into the phrasal frame
slots at the positional level in the Garrett model. It may very well be
during this phrasal frame placement that the scan copier operates as well
(perhaps in parallel) on the segments of the contentives. Consequently, I
would assign the location of phonemic paraphasia to the stage at which
lexical phonological representations of individual content words are in-
serted into the developing phrasal matrices at the positional level.
APHASIC TRANSFORMA nONS 139
CONSTRAINTS ON ERRORS CONDITIONED BY

REPRESENTATIVE/COMPUTATIONAL KNOWLEDGE
Up to this point, we have characterized a model of normal language

production, specified representative levels and computational mappings
between the levels, considered different types of aphasic transformations,
and located those transformations in terms of breakdowns in the com-
putations at certain points in the production process according to the
model. In this final section, we must consider some overriding constraints
that are built into the structure of the model and which serve to ensure
that, although errors are produced, they practically always abide bygener-
al structural principles of the language in question-in our case, English.
Or, if the constraints are universal, they will work to condition the error
typology regardless of which language is being spoken.
If we consider, first, lexical substitutions, we first note that they in-
volve content words almost exclusively. Lexical substitutions based on
meaning similarity most likely take place prior to the establishment of
phrasal frames. Accordingly, they would occur at the functional level,
where the sentential arguments (together with their predicates) are estab-
lished on a meaning base. Derailments here will predictably involve se-
mantic associates, as opposed to phonological associates, and the seman-
tic paraphasia is simply placed into the functional level structure. Its
phonological form is normally retrieved at the positional level, and it is
produced in normal fashion and in the phrasal slot specified for the target
word. This picture explains why semantic paraphasias always agree in
grammatical category with the target. Not only are the phrasal frames
evoked normally in the production of semantic paraphasias, but the al-
lophonic and allomorphic computations are produced in normal fashion
through the processes that map the positional level onto the phonetic
level. This predicts that phonetic processes will await the production of
the semantic error in order to ensure that phonetic realization rules will
apply to the error and not to the target. Further, since the evocation of
phrasal frame matrices is not involved in these errors, the syntax of the
speech output containing semantic paraphasias will remain in the normal
confines of English word order.
Phonemic substitutions and transpositions are highly constrained by
the general patterns of segmental and syllabic structure in the language in
question. Three patterns must be considered here: syllable position, sylla-
ble phonotactics, and syllable sonority. All phonemic paraphasias seem
to abide by these conditions, no matter how severe the aphasic transfor-
mations seem to be.
We have indicated that phonemic substitutions and transpositions in
aphasia are constrained by the syllable slots from which the original target
segments come in their representative forms. In substitutions of
phonemes, the resulting error must be allowed to occur in the position it

occupies. For example, the bilabial nasal Iml may substitute for the Inl in
a word like nut, since Iml may occur in syllable onset position. However,
the velar nasal may never substitute for some other initial nasal (or any
other consonant for that matter) onset in English, since English does not
permit velar nasals in onset position.
Phonemic transposition paraphasias always respect syllable position.
That is, onset segments move only to other onset positions of other sylla-
bles, peaks move only to other peak positions, while codas move to other
coda positions. This has been observed quite often in slips of the tongue
and has been recently observed to operate in paraphasia (e.g., Buck-
ingham, 1980). This constraint holds for all tautosyllabic elements. Am-
bisyllabic consonants, however, appear to have freer mobility, which
makes sense since, phonologically, they may be viewed as either onsets or
codas. By the same token, any tautosyllabic consonant may move to an
ambisyllabic position. Obviously, what we are considering here are con-
straints under which mechanisms like the scan copier operate. That is,
scan copier errors are very highly constrained by language-specific sylla-
ble structure patterns.
Language-specific phonotactic patterns are also observed by phone-
mic substitutions and transpositions. For example, It I may substitute for
lsi in a word like sip, since It! is a permissible singleton onset consonant
in English. However, It! may not substitute for lsi in a word such as slip,
since English-based phonotactic constraints rule out onset clusters * Itl-I.
Similarly, no segment can move to a position whereby a nonpermissible
sequence would be set up. For example, the III in slip of the tongue
cannot move to the second position of the onset in tongue, since, again,
* Itl-I is not permitted.
The so-called sonority principle of syllable structure (Buckingham,
1987,1989) also appears to be an overriding constraint in the production
of many phonemic transposition errors. One very apparent type of error
that seems to be conditioned by this principle is the so-called doublet
creating error, to which I alluded above. First, let me briefly describe the
sonority hierarchy. This universal phonetic principle extends over the
domain of the syllable and involves a continual increase (crescendo) in
vocal apperture and degree of relative prominence through spontaneous
voicing from the initiation of the onset to the syllable peak (the initial
demisyllable). From the peak, there is a gradual decrease (decrescendo) in
sonority out to the last segment of the coda (the final demisyllable). The
order of phonetic segments from least sonorous to most sonorous is
obstruents, nasals, liquids, glides, and vowels. Within the obstruents are
the oral stops, the affricates, and the fricatives, the oral stops being the
least sonorous. The only problem here is that it is often the case that lsi
will appear before oral stops in onset clusters; lsi seems to be the only
recalcitrant segment on the sonority scale. The liquids Irl and III are
between the nasals and glides in terms of sonority. The glides Iyl and Iwl
are quite close to the vowels in sonority, but note that glides may never be
the peaks since they are never stressed. The liquids, on the other hand,
may serve as syllabic elements in English. A typically acceptable syllable
on this scale would be the word gland, where the outermost elements are
oral stops Igl and Idl (the least sonorous), the internal consonants III and
Inl being more sonorous, and the peak a being the most sonorous. Conse-
quently, a syllable is composed of an increasing degree of sonority ipward
toward the stressed elements and from there a decreasing degree of
sonority to the final item of the coda. A syllable such as put would,consist
of a steep incline and decline in terms of sonority, whereas a syllable such
as ground would have a more gradual incline and decline: oral stop,
liquid, vowel, glide, nasal, oral stop. It is a universal fact of syllable
structure (Clements, 1988) that the most preferred, least complex, least
marked, or simplest type of CVC syllable involves sharp inclines and
gradual declines in sonority (very sharp crescendos and less sharp decres-
cendos). Syllables such as Ibal are simpler or more preferred than sylla-
bles such as Iwa/. The unmarked final VC syllable has a more gradual
decrescendo. In this way, we can set up a complexity metric for syllable
types and in a sense for sequences of two segments. Simple, two-segment
sequences are those pairs where the segments involved are far apart on
sonority hierarchy. Accordingly, there would be an enhancement of a
contrast in prominence, or what Donnegan and Stampe (1978) refer to as
an "attraction of opposites." This is why geminate consonant clusters and
sequences of two vowels are so rare in languages. It also explains why
children's earliest syllables involve segments far apart on the sonority
scale (Le., CVs, where the Cs are oral stops). Onset consonant clusters
composed of two oral stops are rare as well, as are onset clusters with
sequences of segments that are adjacent on the sonority hierarchy (e.g.,
Harris, 1983, p. 21, for Spanish). Again, the attraction of opposites holds
more strongly in initial demisyllables, where sharp crescendos are
preferred.
If we turn our attention back to doublet creating phonemic para-
phasias, we note that in order to produce the doublet, the checkoff
monitor must not work, leaving the moved segment (having been either
anticipated or perseverated) in its original position as well. In Buck-
ingham (1989), several examples are cited that demonstrate the operation
of the sonority principle. When some segment is erroneously transposed
by the scan copier, it is generally the case that it will not be checked off
from its original position, if by so doing a highly marked sequence of
segments would be set up. I should note that, heretofore, doublets have
been described but not explained. The sonority principle attempts to offer
an explanation of why we get doublets in some instances and not in
others. Take the doublet creation of two /p/'s in the paraphasia papple for
the word apple. If the doublet had not been created-that is, if the antici-
pated /p/ had been correctly checked off by the monitor-that would have
left a highly marked cross-syllabic sequence of two vowels: lre/ and /:}/.
Most linear transpositions of the doublet-creating sort come from syllable
peripheries and cross syllable boundaries. Consequently, the so-called
"syllable contact law," which follows from sonority principles (Clements,
1988), provides a partial account here. In the syllable contact sequence
X$Y (where $ indicates a syllable boundary), X is usually more sonorous
than Y. Therefore, a situation in which X and Y are both vowels would be
highly undesirable. It turns out that on close inspection, doublet errors are
most often created when the misscanned consonant that crosses syllable
boundaries comes from an intervocalic position. The principle of sonority
and the syllable contact law neatly explain why this should be the case.
Again, the computations carried out by the scan copier and by the check-
off monitor must be sensitive to the syllabic structure.
SUMMARY
The mechanisms underlying aphasic transformations have distinct

characteristics and are located at different junctures in the overall produc-
tion of sentences. In this chapter I have outlined a production model set
up for normalcy, a model based on errors that all normal speakers make
(Le., slips of the tongue). To that model I added other components sug-
gested from related research on language production. The underlying as-
sumption of this chapter has been that without some notion of how the
normal linguistic productive system operates, we have little hope of fully
understanding aphasic errors. Rather than viewing lesions as creating
new mechanisms that produce errors, it is the present author's view that
aphasic breakdowns arise through derailments of computations that are
carried out in normal language production, except, of course, where one
commits a slip of the tongue. And for this reason, I am in general agree-
ment with Eric Lenneberg (1968, p. 345), who argued that certain aphasic
errors are simply augmented and sustained productions, which in more
transient conditions are not uncommon in persons without demonstrable
pathology. Furthermore, these computations operate at different levels in
the overall system, and as a consequence, one can tease out the location of
the aphasic error, while at the same time demonstrating which representa-
tive levels and which computations remain intact. Overall constraints on
the operation of the mechanisms also play an important role, and the
intriguing aspect here is that brain damage practically never disrupts
these hard-wired structural principles of pattern. In sum, it is my firm
conviction that only through psycholinguistically sound models of nor-

mal production can we hope to characterize the various mechanisms un-
derlying aphasic transformations. This chapter has been written in such a
spirit.
REFERENCES
Baars, B. J. (1980). On eliciting predictable speech errors in the laboratory. In V. A. Fromkin
(Ed.), Errors in linguistic performance: Slips of the tongue, ear, pen, and hand. New
Beland, R., & Nespoulous, J. L. (1985). Recent phonological models and the study of aphasic
errors. Paper presented at the Academy of Aphasia, Pittsburgh.
Bell, A., & Hooper, J. B. (Eds.). (1978). Syllables and segments. Amsterdam: North-Holland.
Blumstein, S. E. (1973). A phonological investigation of aphasic speech. The Hague:
Mouton.
Buckingham, H. W. (1980). On correlating aphasic errors with slips-of-the-tongue. Applied
Psycholinguistics, I, 199-220.
Buckingham, H. W. (1981). Lexical and semantic aspects of aphasia. In M. T. Sarno (Ed.),
Acquired aphasia. New York: Academic Press.
Buckingham, H. W. (1985). Perseveration in aphasia. In S. Newman & R. Epstein (Eds.),
Current perspectives in dysphasia. Edinburgh: Churchill Livingstone.
Buckingham, H. W. (1986). The scan-copier mechanism and the positional level of language
production: Evidence from aphasia. Cognitive Science, 10, 195-217.
Buckingham, H. W. (1987). Phonemic paraphasias and psycholinguistic production models
for neologistic jargon. Aphasiology, I, 381-400.
Buckingham, H. W. (1989). The principle of sonority, doublet creation, and the checkoff
monitor. In J. L. Nespoulous & P. Villiard (Eds.), Morphology, phonology and aphasia.
New York: Springer-Verlag. (in press).
Buckingham, H. W., & Kertesz, A. (1976). Neologistic jargon aphasia. Amsterdam: Swets &
Zeitlinger.
Buckingham, H. W., & Rekart, D. M. (1979). Semantic paraphasia. Journal of Communication
Disorders, 12, 197-209.
Buckingham, H. W., Whitaker, H., & Whitaker, H. A. (1978). Alliteration and assonance in
neologistic jargon aphasia. Cortex, 14, 365-380.
Buckingham, H. W., & Yule, G. (1987). Phonemic false evaluation: Theoretical and clinical
aspects. Clinical Linguistics and Phonetics, I, 113-125.
Butterworth, B. (1979). Hesitation and the production of verbal paraphasias and neologisms
in jargon aphasia. Brain and Language, 18, 133-161.
Caplan, D. (1987). Neurolinguistics and linguistic aphasiology: An introduction. Cambridge:
Cambridge University Press.
Chomsky, N., & Halle, M. (1968). The sound pattern of English. New York: Harper & Row.
Clements, G. N. (1988). The role of the sonority cycle in core syllabification. Working Papers
of the Cornell Phonetics Laboratory, No.2, April.
Cutler, A., & Fay, D. (1978). Introduction. In R. Meringer & C. Mayer (Eds.), Versprechen und
Verlesen. Amsterdam: John Benjamins.
Donegan, P. J., & Stampe, D. (1978). The syllable in phonological and prosodic structure. In
A. Bell & J. B. Hooper (Eds.), Syllables and segments. Amsterdam: North-Holland.
Ellis, A. W. (1980). On the Freudian theory of speech errors. In V. A. Fromkin (Ed.), Errors in
linguistic performance: Slips of the tongue, ear, pen, and hand. New York: Academic
Press.
Fay, D. (1981). Substitutions and splices: A study of sentence blends. In A. Cutler (Ed.), Slips
of the tongue and language production. Amsterdam: Mouton.
Fay, D., & Cutler, A. (1977). Malapropisms and the structure of the mental lexicon. Linguistic
Inquiry, 8, 505-520.
Fillmore, C. J. (1968). The case for case. In E. Bach & R. T. Harms (Eds.)' Universals in
linguistic theory. New York: Holt, Rinehart & Winston.
Fodor, J. A. (1982). Methodological solipsism considered as a research strategy in cognitive
psychology. In H. 1. Dreyfus (Ed., with H. Hall), Husserl, intentionality, and cognitive
science (a Bradford book). Cambridge, MA: M.LT. Press.
Fodor, J. A. (1983). The modularity of mind: An essay on faculty psychology (a Bradford
book). Cambridge, MA: M.LT. Press.
Fodor, J. A. (1986). Information and association. In M. Brand & R. M. Harnish (Eds.)' The
representation of knowledge and belief. Tucson: University of Arizona Press.
Fromkin, V. A. (1971). The non-anomalous nature of anomalous utterances. Language, 47,
27-52.
Garrett, M. F. (1975). The analysis of sentence production. In B. Gordon (Ed.), The psychol-
ogy of learning and motivation: Advances in research and theory. New York: Academic
Press.
Garrett, M. F. (1976). Syntactic processes in sentence production. In R. J. Wales & E. Walker
(Eds.), New approaches to language mechanisms. Amsterdam: North-Holland.
Garrett, M. F. (1980). Levels of processing in sentence production. In B. Butterworth (Ed.),
Sentence production 1: Speech and talk. London: Academic Press.
Garrett, M. F. (1982). Production of speech: Observations from normal and pathological
language use. In A. W. Ellis (Ed.), Normality and pathology in cognitive function.
London: Academic Press.
Garrett, M. F. (1984). The organization of processing structure for language production:
Applications to aphasic speech. In D. Caplan, A. R. Lecours, & A. Smith (Eds.),
Biological perspectives on language. Cambridge, MA: M.LT. Press.
Harris, J. W. (1983). Syllable structure and stress in Spanish: A nonlinear analysis.
Cambridge, MA: M.LT. Press.
Kohn, S. E. (1984). The nature of the phonological disorder in conduction aphasia. Brain and
Language, 23, 97-115.
Kohn, S. E. (1985). Phonological breakdown in aphasia. Unpublished doctoral dissertation.
Boston: Tufts University.
Lecours, A. R., & Lhermitte, F. (1969). Phonemic paraphasias: Linguistic structures and
tentative hypotheses. Cortex, 5, 193-228.
Lenneberg, E. H. (1968). Review of Speech and brain mechanisms by W. Penfield & 1.
Roberts. In R C. Oldfield & J. C. Marshall (Eds.), Language. Middlesex, England:
Penguin.
MacNeilage, P. F., Hutchinson, J. A., & Lasater, S. A. (1981). The production of speech:
Development and dissolution of motoric and premotoric processes. In J. Long & A.
Baddeley (Eds.), Attention and performance (Vol. 9). Hillsdale, NJ: Erlbaum.
Meringer, R, & Mayer, C. (1895). Versprechen und verlesen, Eine psychologisch-lin-
guistische Studie. Stuttgart: Goschense Verlagsbuchhandlung.
Pike, K. 1., & Pike, E. (1947). Immediate constituents of Mazateco syllables. International
Journal of American Linguistics, 13, 78-91.
Rinnert, C., & Whitaker, H. A. (1973). Semantic confusions by aphasic patients. Cortex, 9,
56-81.
Schwartz, M. F. (1987). Patterns of speech production deficit within and across aphasia
syndromes: Application of a psycholinguistic model. In M. Coltheart, G. Sartori, & R. Job
(Eds.), The cognitive neuropsychology of language. London: Erlbaum.
Searle, J. (1983). Intentionality: An essay in the philosophy of mind. Cambridge: Cambridge
University Press.
Shattuck-Hufnagel, S. (1979). Speech errors as evidence for a serial ordering mechanism in

speech production. In W. E. Cooper & E. C. T. Walker (Eds.), Sentence processing:
Psycho linguistic studies presented to Merrill Garrett. Hillsdale, NJ: Erlbaum.
Shattuck-Hufnagel, S. (1983). Sub lexical units and suprasegmental structure in speech pro-
duction planning. In P. F. MacNeil age (Ed.), The production of speech. New York:
Springer-Verlag.
Shattuck-Hufnagel, S. (1987). The role of word-onset consonants in speech production plan-
ning: New evidence from speech error patterns. In E. Keller & M. Gopnik (Eds.], Motor
and sensory processes of language. Hillsdale, NJ: Erlbaum.
Shattuck-Hufnagel, S., & Klatt, D. (1979). Minimal use of features and markedness in speech
production. Journal of Verbal Learning and Verbal Behavior, 18,41-55.
Sternberger, J. P. (1982). The nature of segments in the lexicon: Evidence from speech errors.
Lingua, 56, 235-259.
Swinney, D. (1982). The structure and time-course of information interaction during speech
comprehension: Lexical segmentation, access, and interpretation. In J. Mehler, E. C. T.
Walker, & M. F. Garrett (Eds.), Perspectives on mental representation: Experimental and
theoretical studies of cognitive processes and capacities. Hillsdale, NJ: Erlbaum.
Zwicky, A. (1982). Classical malapropisms and the creation of a mental lexicon. In 1. K.
Obler & 1. Menn (Eds.), Exceptional language and linguistics. New York: Academic
Press.
8
Alexia and Agraphia in Spanish

Speakers
CAT Correlations and Interlinguistic
Analysis
ALFREDO ARDILA, MONICA ROSSELLI, and
OSCAR PINZON
Interest in reading alterations arose toward of the last century when De-
jerine described two types of alexia: alexia with agraphia and alexia with-
out agraphia. The first type was correlated with a lesion in the angular
region and the second with a lesion of the occipital lobe in which the
corpus callosum was involved. The relationship between the angular re-
gion and alterations in both reading and writing has been repeatedly
confirmed in the literature (e.g., Benson, 1979; Geschwind, 1965; Hecaen,
1962). The topography of alexia without agraphia has, however, been
debated by several authors (Damasio & Damasio, 1983; Greenblatt, 1983),
and the presence of this reading disturbance has been accepted in cases of
exclusive occipital lesions (Damasio & Damasio, 1983).
Almost a century after Dejerine described the two cases of alexia
classically known in the literature, Benson in 1977 proposed the term
frontal alexia to describe the reading alterations observed in patients with
Broca's aphasia.
Although reading seems to be a strongly lateralized function, since
alterations are usually a product of left-hemisphere lesions, right-hem i-
ALFREDO ARDILA and MONICA ROSSELLI· Konrad Lorenz Foundation University,

Bogota, Colombia, and Miami Institute of Psychology, Miami, Florida 33166-6612.
OSCAR PINZON • San Juan de Dios Hospital, National University of Colombia,
Bogota, Colombia.
147
148 ALFREDO ARDILA et al.
sphere lesions can produce reading alterations that cannot be correlated

with the ability to code or decode written language, but rather with spatial
organization abilities (Ardila & Ostrosky, 1984; Hecaen & Albert, 1978).
Patients with right-hemisphere lesions can present what is known as spa-
tial alexia, but very little is known about its anatomical correlates.
In an attempt to find reading models that comply with the knowledge
we have about cerebral organization, Marshall and Newcombe (1973) de-
veloped a classification system for alexias depending on the type of para-
lexias presented by the patient: visual, semantic, or phonemic. The differ-
ential presence of these types of paralexias corresponds to a particular
type of alexic disturbance: phonological dyslexia, deep dyslexia, and sur-
face dyslexia (Marshall & Newcombe, 1980). A very important limitation
to this classification of alexias is that in the case of paralexias the descrip-
tion has been made in languages, such as English, whose writing system is
not strictly phonological. In English, the phoneme-grapheme conversion
system plays an indirect role, while the semantic channel (understanding
of the read word) and recall of orthographic rules playa principal role
(Miceli, Silveri, & Caramazza, 1985). In languages such as Spanish and
Italian, there is an almost one-to-one relationship between pronunciation
and orthography; the letters are pronounced in almost always the same
way independently of the word they are found in. In spite of some excep-
tions, shown in Table 1, a person with phonological knowledge of the
graphemes can read and write in Spanish even though he may not know
the meaning of the words.
These considerations would therefore suggest that the use of a pho-
nological and/or semantic system will depend on the particular linguistic
system and will not be an intrinsic part of the cognitive system. The
characteristics of reading alterations observed in patients with cerebral
lesions have not yet been determined in languages such as Spanish.
Errors in writing can become apparent at different levels: (1) in graph-
ic quality (calligraphy), (2) in letter sequence (orthography), (3) in the
correct choice of words, lexical and syntactic organization, and mor-
phological composition, and (4) in spatial organization (e.g., distribution
of spaces, direction of the line). However, with regard to writing altera-
tions, no classification exists that has been completely accepted by ap-
hasiologists. Many authors use the term agraphia for all types and degrees
of writing alterations, and agraphia has thus become a confused term that
has not been sufficiently studied. Different attempts at classification have
been made (e.g., Benson & Cummings, 1985; Hecaen & Albert, 1978; Luria,
1977). Most authors agree that agraphias can be divided into two large
groups: agraphias that accompany aphasic language disturbances and
agraphias that do not. The first group comprises those aphasic agraphias
that include fluent and nonfluent agraphias and agraphia with alexia. The
second group includes agraphias such as motor, apraxic, visuospatial, and
hemiagraphic agraphias.
ALEXIA AND AGRAPHIA IN SPANISH SPEAKERS 149
TABLE 1. Phoneme and Grapheme Correspondence

in Spanish
Phoneme Grapheme
Ibl b,v
lsi c (before e, i), s, z
ItSI ch
Idl d
If I f
Igl g (before a, 0, u) and gu (before e, i)
Iii i, y (as a conjunction and in diphthongs)
Ihl j, g (before e, i)
Ikl c (before a, 0, u) and qu (before e, i)
III I
IAI or Ijl II
Iml m
Inl n
IJlI fl
Ipl p
Irl r
Ii:I rr, r (at the beginning of a word)
It I t
Ikl + lsi x
Ijl y
h
lal a
lei e
101 o
lui u
In 1982 Bub and Kertesz suggested the term deep dysgraphia to de-
scribe the profile characterized by the semantic substitutions in writing
that one of their patients presented. Similarly, Roeltgen, Sevush, and
Heilman (1983) gave the name phonological agraphia to the inability
another patient presented to write nonsense words to dictation in spite of
a good ability to write known words. Roeltgen, Rothi-Gonzalez, and
Heilman (1986) have proposed the existence of a semantic agraphia in
five patients who could spell irregular words but had difficulties in giving
them their meaning.
Little is known about the relationship between writing alterations
and cerebral damage in reading-writing systems like the one used in
Spanish. The classifications used are very vague and do not sufficiently
satisfy observations made in clinical practice. Furthermore, no studies
have been carried out to correlate reading-writing alterations in Spanish-
speaking patients with lesions in particular areas of the brain.
One of the methods that has probably allowed us to advance furthest
in our knowledge about cerebral lesion topography is the brain scan or
150 ALFREDO ARDILA et a1.
computerized axial tomography. By means of this radiological method it

has been possible to visualize and exactly determine the location of cere-
bral pathologies of different etiologies and, as a result, establish correla-
tions between clinical manifestations and structural damage to the brain
(Ardila, Montanes, Bernal, & Serpa, 1986; Damasio & Damasio, 1983;
Greenblatt, 1983; Kertesz, 1983, Chapter 3 this volume; Naeser, 1983;
Novoa & Ardila, 1987; Poeck, Bleser, & Von Keyserlingk, 1984; Rosselli,
Rosselli, Vergara, & Ardila, 1985).
In this work, 62 Spanish-speaking patients with cerebral focal lesions
were studied and their alterations in reading and writing were correlated
with the topography of the damage found by the scanning method.
METHOD
Subjects
Sixty-two patients with cerebral damage were studied (27 women, 35
men; average age = 41.43, SD = 13.92, age range = 16-65). These subjects
presented various etiologies (vascular = 44, tumoral = 14, traumatic = 4).
Sixty-two normal subjects matched for age, sex, and schooling with the
first group were also studied. The cerebral damage had evolved in a peri-
od varying from 1 to 4 months. Patients had no background of previous
neurological or psychiatric illnesses. Average schooling was 8.14 years
(range = 4-19, SD = 4.54). Table 2 shows the general characteristics of
the population studied. All lesions were confirmed by means of comput-
erized axial tomography, and the analysis of damage topography was
carried out by a neuroradiologist.
Procedure
The patients included in the sample were taken from the neuropsy-
chological services of the San Juan de Dios Hospital and the Colombian
Institute of Neurology. Each patient was administered a neuropsycholo-
gical battery consisting of the following tests: (1) Boston Diagnostic Ap-
hasia Examination (Goodglass & Kaplan, 1979), (2) Token Test (De Renzi &
Faglioni, 1978), (3) a reading and writing test specifically designed for
Spanish-speaking subjects and standardized in a population of normal
subjects (Rosselli & Ardila, 1987), (4) the Rey-Osterrieth Complex Figure
(Osterrieth, 1944). The specific language tests (Boston Diagnostic Aphasia
Examination and the Token Test) were applied only to patients with left-
cerebral damage. Normal subjects were given only the reading and writing
test.
Patients with left-hemisphere lesions were divided into seven groups,

and the following criteria were jointly considered: results on the Boston
Diagnostic Aphasia Examination, topography of the damage as shown on
scans, and the general neuropsychological and neurological examination.
The following groups were formed: (1) prefrontal (PF-6 patients), (2)
Broca's aphasia (BA-5 patients), (3) conduction aphasia (CA-6 patients),
(4) Wernicke's aphasia (WA-13 patients), (5) anomic or amnesic aphasia
(AA-4 patients), (6) alexia without agraphia (AWA-3 patients), and (7)
global aphasia (GA-4 patients). As well as being classified in this way,
patients with left-hemisphere lesions were divided into two large groups:
prerolandic and retrorolandic, depending on whether the scans revealed
their lesions to be in front of or behind the fissure of Rolando.
Patients with right-hemispheric lesions were divided into two groups
according to damage topography: (1) prerolandic (6 patients) and (2) retro-
rolandic (15 patients).
The results obtained by the patients on the reading and writing tests
were grouped into tables in order to carry out the pertinent comparisons.
Their scores were compared with those obtained by the matched sample
from the normal population. Comparisons were also made between the
means obtained by patients with right and left lesions. The means ob-
tained by prerolandic patients were compared with those obtained by
patients from the retrorolandic group, in the case of patients with right-
hemisphere lesions. The means of patients from the different left hemi-
sphere lesion groups were compared one with the other. A t test (Ed-
wards, 1968) was used to carry out the comparisons between means. After
carrying out the pertinent comparisons between the means obtained by
the different groups in the reading and writing test, an analysis was made
of the types of error made for each item. The means of some types of error
were subsequently compared for patients with right and left lesions, as
were the means of the types of error for the right subgroups and for the left
subgroups. The results obtained on the Boston Diagnostic Aphasia Exam-
ination and the Token Test were used as classificatory criteria for the
patients in the different groups. Similarly, the Rey-Osterrieth Complex
Figure test was used to relate spatial alterations in writing with the pres-
ence or absence of constructional apraxia. The classification criteria used
in the Rey complex figure were taken from Lezak (1983).
In order to transcribe the patients' lesions, a template was designed
with 10 standard scanner cuts from the base of the brain to the highest
region. Each lesion identified by the scan was transcribed to the standard
template. The lesions corresponding to the different patients in the same
category or group were then superimposed onto one template, following
the method described by Kertesz (1983) and Naeser (1983) in order to
identify the critical compromised zone in the appearance of the deficit
analyzed.
TABLE 2. General Characteristics of the Sample
Educational
Sex Age level Topography Hemisphere Etiology Token Category
Pat. 1 F 34 5 Frontal Right Tumor Prerolandic (1)

Pat. 2 F 53 5 Frontal Right Vascular Prerolandic (2)
Pat. 3 M 19 8 Frontal Right Trauma Prerolandic (3)
Pat. 4 M 39 5 Frontal Right Vascular Prerolandic (4)
Pat. 5 M 63 10' Frontal Right Tumor Prerolandic (5)
Pat. 6 M 32 15 Frontal Right Tumor Prerolandic (6)
Pat. 7 M 50' 11 Occipital Right Vascular Retrorolandic (1)
Pat. 8 F 44 4 Par-occ Right Vascular Retrorolandic (2)
Pat. 9 F 65 11 Par-occ Right Vascular Retrorolandic (3)
Pat. 10' F 33 5 Par-tern Right Vascular Retrorolandic (4)
Pat. 11 M 55 5 Par-occ Right Vascular Retrorolandic (5)
Pat. 12 M 26 5 Tern-par Right Vascular Retrorolandic (6)
Pat. 13 F 44 11 Tern-par Right Tumor Retrorolandic (7)
Pat. 14 F 16 7 Parietal Right Vascular Retrorolandic (8)
Pat. 15 M 45 5 Parietal Right Vascular Retrorolandic (9)
Pat. 16 F 49 5 Tern-ins Right Vascular Retrorolandic (10')
Pat. 17 M 38 5 Temporal Right Tumor Retrorolandic (11)
Pat. 18 F 19 11 Tern-par Right Vascular Retrorolandic (12)
Pat. 19 M 55 5 Tern-par Right Vascular Retrorolandic (13)
Pat. 20' F 29 5 Tern-par Right Vascular Retrorolandic (14)
Pat. 21 M 60' 5 Tern-par Right Vascular Retrorolandic (15)
Pat. 22 M 55 6 Frontal Left Vascular Prefrontal (1)
Pat. 23 M 20' 11 Frontal Left Vascular 26 Prefrontal (2)
Pat. 24 F 47 5 Frontal Left Tumor 33 Prefrontal (3)
Pat. 25 F 56 16 Frontal Left Tumor 20' Prefrontal (4)
Pat. 26 M 45 4 Frontal Left Vascular 24 Prefrontal (5)
Pat. 27 F 23 11 Frontal Left Tumor 26 Prefrontal (6)
Pat. 28 F 43 16 Frontal-tern Left Vascular 28 Broca (1)
Pat. 29 M 26 16 Frontal Left Vascular 31 Broca (2)
Pat. 30 F 30 5 Frontal-par Left Vascular 18 Broca (3)
Pat. 31 F 64 4 Frontal-par Left Vascular 29 Broca (4)
Pat. 32 M 36 16 Frontal-par Left Vascular 30 Broca (5)
Pat. 33 M 35 5 Frontal-par Left Tumor 26 Conduction (1)
Pat. 34 F 61 8 Parietal Left Vascular 22 Conduction (2)
Pat. 35 F 39 5 Insular Left Vascular 31 Conduction (3)
Pat. 36 M 33 8 Parietal Left Vascular 32 Conduction (4)
Pat. 37 F 43 5 Parietal Left Tumor 31 Conduction (5)
Pat. 38 F 26 11 Parietal Left Vascular 27 Conduction (6)
Pat. 39 M 63 6 Tern-par Left Vascular 17 Wernicke (1)
Pat. 40 M 34 11 Tern-par Left Tumor 18 Wernicke (2)
Pat. 41 M 31 8 Temporal Left Vascular 14 Wernicke (3)
Pat. 42 F 48 5 Temporal Left Vascular 18 Wernicke (4)
Pat. 43 M 18 9 Tem-occ Left Trauma 15 Wernicke (5)
Pat. 44 F 53 4 Temporal Left Vascular 5 Wernicke (6)
Pat. 45 M 28 5 Temporal Left Trauma 22 Wernicke (7)
Pat. 48 M 46 16 Tern-par Left Tumor 8 Wernicke (10)
Pat. 49 M 28 5 Temporal Left Trauma 18 Wernicke (11)
Pat. 50 F 37 16 Tern-par Left Vascular 4 Wernicke (12)
Pat. 52 F 43 7 Temporal Left Tumor 30 Amnesic (1)
Pat. 53 M 36 5 Par-occ Left Tumor 17 Amnesic (2)
Pat. 54 M 65 11 Angular Left Vascular 27 Amnesic (3)
Pat. 55 M 30 11 Tern-par Left Vascular 24 Amnesic (4)
Pat. 56 M 54 5 Occipital Left Malfor 27 Alexia wlo agraphia (1)
Pat. 57 M 47 5 Occipital Left Vascular 30 Alexia wlo agraphia (2)
Pat. 58 M 65 16 Occipital Left Vascular 30 Alexia wlo agraphia (3)
Pat. 59 F 25 5 FPT Left Vascular 8 Global (1)
Pat. 60 M 31 5 FPT Left Vascular 11 Global (2)
154 ALFREDO ARDILA et al.
Instrument
A reading and writing test designed for Spanish-speaking subjects
was used (Rosselli & Ardila, 1987). The test consisted of the following
subtests:
Reading
Reading of letters (20; 15 consonants, 5 vowels)
Reading of syllables (12; e.g., po, li, elus, trans)
Reading of logo tomes (11; e.g., tala, fasaja)
Reading of words (13; e.g., coso, libra, ventana, bicieleta)
Reading of sentences (5; e.g., La cantina es de Juan)
Understanding orders (9; e.g., close your eyes)
Reading and comprehension of texts (9; 108-word text, 4 questions;
and 185-word text, 5 comprehension questions)
Ideographic reading (7; e.g., Coca-Cola)
Writing
Writing of letters (dictation and copy) (14; 9 consonants, 5 vowels)
Writing of syllables (dictation and copy) (10; e.g., po, ela)
Writing of words (dictation and copy) (6; e.g., coso, caballo)
Writing of sentences (dictation and copy) (4; e.g., La gente se reune)
Written description of a picture (cookie theft)
The test uses a scoring system based on the number of errors; the
possible number of errors is indicated in brackets. The written description
of a picture was taken from the Boston Diagnostic Aphasia Examination
(Goodglass & Kaplan, 1979) and the proposed scoring method was used.
RESULTS
Seven patients of the 62 with cerebral damage were excluded from

the statistical analysis (patients 11, 12, 21, 59, 60, 61, and 62 in Table 2)
because their lesions were considered to be too extensive, and for that
reason, the very large number of errors they made on all the subtests could
have contaminated the results. Of these 7 patients, 4 had an infarct of the
whole left-middle cerebral artery and presented global aphasia, and 3 had
widespread lesions of the right hemisphere as a result of an extensive
infarct of the right-middle cerebral artery. The results obtained by the
patients with global aphasia were analyzed separately.
Population with Cerebral Damage and Normal Population

On comparing the means of the scores obtained by patients with left-
hemisphere lesions with those obtained by the matched normal controls,
a significantly greater number of errors were observed in the patients in
all the subtests of the reading and writing sections.
Significant differences were observed in the results produced by the t
test when comparing the means of the scores obtained in the reading and
writing subtests in patients with right-hemisphere lesions with those ob-
tained by the matched normal subjects only in the reading of syllables,
logotomes, words, sentences, and understanding-of-orders subtests. In the
writing subtests the average number of errors was significantly higher in
patients with right-hemisphere lesions than in normal subjects.
Patients with Right-Hemisphere Lesions and Patients with Left-

Hemisphere Lesions
The results obtained on comparing the means corresponding to the
reading subtests of patients with right-hemisphere lesions and those with
left-hemisphere lesions are shown in Table 3. The average number of
errors (scores) obtained by patients with left lesions were significantly
greater in all the subtests except for the reading of sentences, in which the
difference in scores did not reach significance level.
The results of the t test used to compare the performance of patients
with right- and left-hemisphere lesions on the writing subtests showed
significant differences for the writing of syllables and sentences subtests
but not for the writing of letters and words (Table 4). The average number
of total errors obtained by the two groups of patients (with right and left
lesions) show that the patients with left-hemisphere lesions presented a
greater number of errors in both reading and writing.
Intrahemispheric Comparisons
After comparing the results obtained by patients with right and left
lesions, intrahemispheric comparisons were then carried out. For this
purpose, the patients in each group were divided into two subgroups:
prerolandic and retrorolandic, depending on whether the scanner image
was found in front of or behind the central fissure.
The left retrorolandic group showed a significantly greater number of
errors in reading [X = 41.65 (43% errors)], and in writing [X = 15.81
(47%)], than the prerolandic group [reading: X = 23.55 (25%); writing: X
= 9.18 (27%)]. The group of patients with right retrorolandic lesions pre-
sented an almost equivalent number of errors to the right prerolandic
TABLE 3. t Test of Means Obtained by Patients with Right and Left Cerebral Lesions in the Reading Subtests a
Subtest
group LTRb SLB LGT WOR SEN ORD TXT IDG
Right hemi- X = 0.5 X = 1.00 X = 2.72 X = 2.00 X = 1.44 X = 1.21 X= 3.05 X = 0.22
sphere 2.5% 8% 25% 15% 29% 13% 34% 3%
n = 18
Left hemi- X= 2.8 X= 3.64 X = 5.89 X = 5.10 X = 2.24 X= 3.91 X = 5.78 X = 1.29
sphere 14% 30% 54% 39% 45% 43% 64% 18%
n = 37
df 53 53 53 53 53 53 53 53
2.27 2.11 2.006 2.15 1.08 2.33 1.71 2.18
P .02 .02 .05 .02 .15 .02 .05 .02
opercentage of errors of each group is shown in relation to the total possible.

bReading of letters (LTR). reading of syllables (SLB). reading of logotomes (LGT). reading of words (WaR). reading of sentences (SEN). comprehension of written orders
(ORD). comprehension of text (TXT). ideographic reading (IDG).
TABLE 4. t Test of Means Obtained by Patients with Right and Left Cerebral
Lesions on the Writing Subtests Q
Subtests
groups Letters Syllables Words Sentences
Right hemisphere X= 1.44 X=1.72 X= 1.44 X= 1.83

n = 18 10% 17% 24% 46%
Left hemisphere X= 3.08 X= 4.56 X= 2.81 X= 3.10
n = 37 22% 46% 47% '78%
df 53 53 53 53
1.44 2.11 1.61 1.73
P .10 .02 .10 .05
aPercentage of errors of each group is shown in relation to the total possible.
group in both reading [retrorolandic: X = 18.67 (19%); prerolandic: X =

10.67 (11%)], and in writing [retrorolandic: X = 7.17 (21%); prerolandic:
X = 5.00 (15%)].
On comparing the means of each subtest for the left prerolandic group
with those of the left retrorolandic group, the latter group was found to
present a significantly greater average number of errors on all the subtests
with the exception of the ideographic reading subtest. In the writing sec-
tion, the left retrorolandic group obtained significantly greater means than
the left prerolandic group on the writing of letters and words subtests.
The statistical differences between the means of the right pre- and
postrolandic groups can be seen in the reading section, in which dif-
ferences were observed in the means obtained on the reading of log-
otomes, words, and text comprehension subtests. In these subtests, the
retrorolandic group had a greater number of errors. In the writing section,
no differences in the performance by the two groups could be appreciated.
As explained in the procedure, the patients with left-hemisphere le-
sions were also divided into six groups according to lesion site and to the
neuropsychological alterations presented in the Token Test and the
Boston Diagnostic Aphasia Examination. The means and standard devia-
tions obtained by each group on the two sections of the reading and
writing test were compared. Maximum number of errors in reading was
found in A WA group; minimum number of errors in reading and writing
tasks was found in PF patients.
Table 5 shows the percentage of errors obtained by each of the sub-
groups with cerebral lesions (excluding the global aphasics) on each of the
158 ALFREDO ARDILA et 01.
TABLE 5. Percentage of Errors Obtained by Patients from Different Groups

in the Reading Subtests
LTRa SYL LGT WOR SEN ORD TXT IDE
Prefrontal
n=6 0 7 15 14 2 8 4 0
Broca
n=5 15 38 49 29 60 47 47 37
Conduction
n=6 29 28 64 48 40 42 48 14
Wernicke
n = 13 17 28 55 42 48 57 84 14
Anomic
n=4 13 25 73 42 50 50 89 16
Alexia without agraphia
n = 3 30 50 82 69 86 100 100 51
Right prerolandic
n=6 7 4 11 6 6 11 12 3
Right retrorolandic
n = 12 25 18 36 25 43 22 50 6
aReading of letters (LTR). reading of syllables (SYL). reading of logotomes (LGT). reading of words (WaR).
reading of sentences (SEN). comprehension of written orders (ORD). comprehension of a text (TXT).
ideographic reading (IDE).
reading subtests. The greatest number of errors was made in the reading of
letters and words by the CA and A WA groups. The AW A, AA, and CA
group showed the greatest number of errors in the reading of logotomes; in
reading comprehension A WA, AA, and WA were most impaired. In the
case of ideographic reading, the greatest numbers of errors was observed
in patients with AW A. Patients with PF presented the lowest mean for
errors. Table 6 shows the percentage of patients presenting errors in each
reading subtest.
In Table 7, we can appreciate the percentage of errors obtained by
each subgroup of patients in each subtest of the writing section. After the
GA group (79% errors), the patients in the WA (36%), CA (30%), and AWA
(30%) groups presented the largest number of errors in the writing of
letters. In the writing of syllables, the greatest number of errors was made
by groups AA, A WA, CA, and WA, and in the writing of words, the groups
A WA, W A, and CA presented a higher average number of errors in rela-
tion to the other groups. The AA and AWA groups presented the greatest
number of errors in the writing of sentences in spite of the fact that the
number of errors was very similar in all groups except for the groups with
right lesions and the PF left group, for which the average of errors was
lower. Table 8 shows the percentage of patients presenting errors in each
writing subtest.
TABLE 6. Percentage of Patients Presenting Errors in Each Reading Subtest
LTRa SYL LGT WOR SEN ORD TXT IDE
Prefrontal 17 33 83 67 17 33 67 0
Broca 80 80 100 100 80 100 80 60
Conduction 50 83 100 100 83 100 83 67
Wernicke 69 69 100 92 62 100 100 38
Anomie 75 75 100 100 100 75 100 75
Alexia without agraphia 100 100 100 100 100 100 100 100
Right prerolandic 17 33 50 50 33 66 33 17
Right retrorolandic 33 41 83 91 66 83 109 25
"Letters (LTR), syllables (SYL), logotomes (LGT) , words (WOR), sentences (SEN), comprehension of
orders (ORD), comprehension of a text (TXT), ideographic reading (IDE).
TABLE 7. Percentage of Errors in Different Groups for Writing Subtests
Letters Syllables Words Sentences
Prefrontal 4 26 22 58
Broca 20 42 43 75
Conduction 30 48 55 83
Wernicke 36 48 55 75
Anomie 18 57 33 88
Alexia without agraphia 30 53 60 100
Right prerolandic 7 13 18 38
Right retrorolandic 14 24 27 55
TABLE 8. Percentage of Patients Presenting Errors in Each Writing Subtest
Letters Syllables Words Sentences
Prefrontal 33 50 66 83
Broca 40 80 60 100
Conduction 83 83 83 100
Wernicke 77 92 100 100
Anomie 75 75 50 100
Alexia without agraphia 100 100 100 100
Right prerolandic 33 50 50 50
Right retrorolandic 75 75 41 75
Quantification of the Type of Errors

Table 9 shows different types of errors in some reading subtests.
Literal substitutions on reading letters, syllables, logotomes, words, and
sentences were more evident in the AW A, CA, and WA groups. Of these
literal substitutions, changes in consonants were more frequent than
changes in vowels. Neologisms were observed in some patients with W A,
AA, and A WA. Morphological substitutions were recorded in the group of
patients with A WA. A tendency to literal reading was apparent in the
AWA group, and defects in naming a letter by means of words (e.g., for the
letter D, the patient would say "that's the one for donkey") that would
give the desired letter were observed only in the AA group. Anticipations
were found in patients with BA and CA. Omissions of letters within
words were principally observed in the BA group. In the AWA and CA
groups, a tendency was observed to present fewer of errors in the left
portion of the syllables or words read. The left PF group showed the
smallest number of errors.
Patients with right-hemisphere retrorolandic lesions mainly present-
ed errors due to left hemi-inattention, with a tendency to confabulate the
left half of the word. In some patients with left lesions, particularly in the
AWA and WA groups, a phenomenon similar to invention was observed,
which consisted in meaningfully reading syllables and logotomes-e.g.,
instead of reading sligo (nonsense), they read Ie digo ("I tell him"). Final-
ly, patients with right retrorolandic lesions had difficulty in following the
lines of the text on reading, and this prevented them from following the
content of the text. Patients with right prerolandic lesions presented mini-
mal errors as a result of neglect in the reading of logotomes; in the other
subtests they presented no errors.
Analysis of the types of error in the writing subtests showed the
presence of literal paragraphias in all the subgroups with left lesions; the
greatest number, however, were made by patients with WA. Persevera-
tions were registered in the groups with left PF, BA, W A, and A W A.
Omissions were frequent errors in the CA, W A, and A W A groups. Errors
due to anticipation were more evident in the BA group and neologisms
were observed in the groups with AA, WA, and CA. Paragrammatism in
spontaneous writing was observed in W A. The group with CA was the
only one to show evident self-correction. Errors due to agrammatism were
found in the writing of sentences in the CA and BA groups. In the writing
of sentences a tendency was observed to present fewer of errors on the left
side of the sentence in groups A WA, WA, and left PF. The A WA subgroup
increased the number of errors when copying (Table 10).
In patients with right-hemisphere lesions (in particular in the retro-
rolandic group), omissions of letters and omissions of the strokes of a
letter were observed. Similarly, a tendency was observed to fragment
TABLE 9. Type of Errors in Five Reading Subtests (Letters, Syllables, Logotomes, Words, Sentences)a
LSU NED VMS ANT PER AND LRE SLM 10M NOM LAD NEG CON
Prefrontal 1.0 0.0 0.0 0.0 0.0 0.0 0.0 4.0 2.0 0.0 0.5 0.0 0.0
Broca 5.0 0.0 0.0 5.0 6.5 0.0 0.0 0.5 8.0 0.0 2.0 0.0 0.0
Conduction 11.5 1.5 2.0 3.5 1.5 0.0 3.5 3.5 3.5 0.0 3.5 0.0 0.0
Wernicke 11.5 8.0 2.5 1.0 1.0 0.0 2.5 3.0 5.0 2.0 2.0 0.0 0.0
Anomie 9.0 6.0 1.0 0.0 0.5 6.5 6.5 1.5 4.0 10.0 0.0 0.0 0.0
Alexia without agraphia 15.0 5.0 11.0 0.0 9.0 0.0 15.0 2.5 1.5 1.5 4.0 0.0 0.0
Right prerolandic 0.5 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.5 0.5 3.0
Right retrorolandic 1.5 0.0 0.0 0.0 0.0 0.0 0.0 3.0 3.0 0.0 1.5 5.5 8.5
oEach cell represents the average number of errors. Literal substitutions (LSU). neologisms (NED). verbal morphological substitutions (VMS). anticipations (ANT).
perseverations (PER). anomie errors (AND). literal reading (LRE). substitution of logotomes for meaningful words (SLM). letter omissions (LOM). noun omissions (NOM).
letter additions (LAD). neglect (NEG). confabulation (CON).
162 ALFREDO ARDILA et at
TABLE 10. Type of Errors in Writing Testa
PER LIT NEO LAD LOM FAD FOM ANT GEL COP
Prefrontal 3.2 2.0 0.0 0.0 1.6 0.0 0.0 0.4 0.0 0.0
Broca 3.2 5.6 0.0 1.2 0.8 0.0 0.0 5.2 0.0 8.8
Conduction 1.6 7.2 6.0 1.6 6.4 0.0 0.0 2.0 0.0 4.0
Wernicke 3.2 8.8 9.6 2.8 6.0 0.0 0.0 0.0 0.0 0.4
Anomie 0.0 5.2 12.0 2.0 1.2 0.0 0.0 2.4 0.0 2.0
Alexia without agraphia 3.2 7.6 4.0 3.6 6.4 0.0 0.0 1.2 0.0 37.2
Right prerolandic 0.0 1.2 0.0 2.8 2.4 5.2 0.8 0.0 1.2 0.0
Right retrorolandic 0.0 1.2 0.0 2.4 8.0 3.2 2.4 0.0 12.0 0.0
aEach cell represents the average of errors. Perseverations (PER). literal substitutions (LIT), neologisms
(NEO),letter additions (LAD), letter omissions (LaM). feature additions (FAD). feature omissions (FOM),
anticipations (ANT), grouping of elements (GEL), copying errors (COP).
words by separating the syllables of one word in an inadequate way, and

grouping of words was clear. They wrote obliquely at an angle of 13
degrees to the line on average. An increase or decrease of approximately 3
cm of the left-hand margin was observed as the patient wrote words and
sentences and in the written description of the picture (spontaneous
writing).
Reading was impossible for the patients with global aphasia since
they had no language (or their language had been reduced to a verbal
stereotype) and identification was purely by chance. Some patients spo-
radically recognized syllables or words by the number of letters they
contained and pointed to a syllable with the equivalent number of ele-
ments to those asked for. It was of note, however, that these patients had
on average a better performance in ideographic reading than patients from
the A WA group. In writing, patients from this group presented a severe
impossibility owing to hemiparesis or apraxia. Those who managed to
write with the left hand presented stereotype writing in a perseverative
way; they were even less able to copy correctly, and none of them could
present spontaneous writing.
Scores on the Rey-Osterrieth Complex Figure

The Rey-Osterrieth Complex Figure could be applied only to 19 pa-
tients with left-hemisphere lesions (14 with retrorolandic lesions) and to
12 patients with right-hemisphere lesions (1 with prerolandic lesions and
11 with retrorolandic lesions). Scoring criteria for performance on the
copy of the Rey-Osterrieth Complex Figure were taken from Lezak (1983).
According to this scoring system, 2 points are given for a perfect copy of
each of the 18 parts into which the figure is divided; omISSIOns or a

performance different from the figure copied score zero, half a point, or 1
point, depending on the severity of the error. The maximum score possi-
ble is 36 points.
Patients with left-hemisphere lesions had a mean score of 28.39 and a
standard deviation of 6.93, and patients with right hemisphere lesions a
mean of 13.09 and a standard deviation of 6.93. The Student t test showed
the significance level for this difference in the means to be .005 (t = 5.83,
df = 29).
The errors observed in patients with left-hemisphere lesions were
due to their simplifying the figure and omissions of details, particularly
those on the right-hand side; disintegration or fragmentation of the figure
were never observed. In all cases of left lesion, the drawn figures could be
identified by an observer as figures from the Rey. In the performance by
patients with right-hemisphere lesions, disintegration of the figure was
evident, as were disarticulation of its elements, the tendency to persevera-
tion of the strokes, and the omission of elements in the left part of the
figure, and, on occasion, it was impossible to identify the figure drawn as
the Rey figure. None of the patients with right lesions showed attempts at
self-correction or made comments indicating their concern over their bad
performance.
Superimposition of Scans
The scans of the patients from each group were superimposed in
accordance with the steps explained in procedure. In this way, templates
were obtained corresponding to the groups studied, seven of which are
shown jn Figures 1 to 7. Owing to the small sample in the AWA and GA
groups, the superimposition of their scans did not produce important
data. The templates of the figures indicate the areas that had a certain
level of superimposition: 20-40%, 50-60%, 70-90%, and 100%.
Figure 1 shows the regions of the left hemisphere that were involved
in the PF group (n = 6). The anterior prefrontal cortex and the paramedian
and parasagittal region of the left frontal lobe are those principally in-
volved. Figure 2 shows how in patients with BA there is an extensive
damage of the motor and premotor cortex and also of the anterior portion
of the insular cortex. In the CA group (n = 6), a damage principally of the
anterior portion of the insula appeared, involving a small portion of the
sensitive cortex in the middle part and extending to the beginning of the
supramarginal region (Fig. 3).
The topographical findings obtained from the superimposition of
scans of patients with WA (n = 13) are shown in Figure 4. The posterior
third of the insular cortex was the site of the lesions in all the patients.
164 ALFREDO ARDILA et 01 .
FIGURE 1. Scan superimposition: Left prefrontal (n = 6).
SQ(Dw ~
\ /... ~
~
_
I!lIID
~
100\
70\ - 90~
SO~ - 60~
--...... CJ 20~ - 40~
FIGURE 2. Scan superimposition: Broca's aphasia (n = 5).

FIGURE 3. Scan superimposition: Conduction aphasia (n = 6).
. 40~
FIGURE 4. Scan superimposition: Wernicke's aphasia (n = 13).

The posterior temporal cortex, particularly the first gyrus, was also ob-
served to be involved and, to a lesser degree, the posterior parietal cortex
(angular gyrus). The scanner results of the patients with AA (n = 4) can be
seen in Figure 5. Deep involvement of the posterior temporal lobe and
angular gyri predominated. There were very few patients with AWA (n =
3), and it was therefore not possible to produce an adequate superimposi-
tion of the lesions; the posterior portion of the occipital lobe was observed
to be involved in all the patients, and 1 patient also presented a certain
posterior parietal damage. The patients with global aphasia (n = 4) pre-
sented an extensive cortical and subcortical damage that extended from
the premotor regions of the frontal lobe to the parieto-temporo-occipital
junction.
The superimposition of scans of the lesions presented by the patients
with right-hemisphere lesions is shown in Figures 6 and 7. The patients
with prerolandic lesions (n = 6) had involved the superior region of the
right prefrontal area, and, on occasions, this damage reached the premo tor
area. The cerebral compromise observed in the patients from the retro-
rolandic group was located principally in the deep cortical portion of the
parietal lobe. In some patients, however, a deep posterior temporal and
anterior occipital damage was also observed.
so~
:!o~
FIGURE 5. Scan superimposition: Amnesic aphasia (n = 4).

FIGURE 6. Scan superimposition: Right prerolandic (n = 6).
FIGURE 7. Scan superimposition: Right retrorolandic (n = 12).

DISCUSSION
In general, patients with left lesions showed a considerably greater

number of errors than those with right lesions. However, the fact that
patients with right lesions showed differences between their means and
those of normal subjects suggests some type of relation between this hemi-
sphere and written linguistic processes. It is also interesting to observe the
similarity between the performance of patients with right lesions and
normal subjects on some of the reading subtests (e.g., letters and ide-
ographic reading).
Patients with Right and Left Lesions

The patients with right lesions presented better reading abilities than
the left patients with the exception of the reading of sentences, in which
the two groups of patients had an equivalent number of errors. These
results would suggest that the items used in the reading of sentences
could have had an equivalent level of difficulty for the patients because
the items had a semantic content.
The results of the writing tests show that the writing of syllables and
sentences is more sensitive to left lesions than to right while the writing of
letters and words presented a similar level of difficulty for the two groups
of patients. However, for both groups of patients, the level of difficulty
increased with the number of written elements: Writing letters is easier
than writing syllables, but syllables are easier than words, and so on. The
patients with left lesions presented defects in verbal memory and there-
fore had a reduced memory volume. As a consequence, they could more
easily retain items with a few elements.
Intrahemispheric Comparisons
If the left prerolandic and retrorolandic groups are compared, a great-
er number of errors can be observed in the retrorolandic group on both the
reading and writing tests, while no important differences were observed
in the two groups of patients with right-hemisphere lesions. Several as-
pects could be related to these findings: (1) The posterior areas of the left
cerebral cortex have classically been considered to be principally respon-
sible for the symbolic aspects of reading and writing, while the anterior
areas have been related to the motor aspects of the same. (2) Several
authors (e.g., Bradshaw & Nettleton, 1981) have spoken of the homoge-
neousness of the right-hemisphere areas, with a supposedly more diffuse,
less localized organization of functions, in contrast to the heterogeneity of
the left hemisphere areas, with a supposedly more defined organization of
its functions.
In a more specific analysis, by subtests, it was observed that ide-

ographic reading was conserved without difficulty in left (except AWA)
retrorolandic patients, probably indicating that this type of reading is
performed in a gestaltic way, with the right hemisphere playing a very
large part. In the writing section, the patients with posterior lesions made
a larger number of errors in all subtests, thus showing that these patients
were the ones responsible for the low scores the left group had on these
subtests in relation to the right group.
On analyzing the means obtained on the reading tests by the groups
with left lesions, it was observed that the PF patients who presented no
aphasic compromise of language had the smallest number of errors, while
the GA patients, whose language involvement was maximum, had the
greatest number of errors, showing the close connection between altera-
tions in oral language and written language. The other groups each had a
similar average number of errors, with the AA group obtaining the lowest
mean. The WA, AWA, and CA groups obtained a fairly similar number of
errors and shared some types of mistake. However, the types of error
could be defined for some groups: The WA group presented literal para-
lexias and neologisms, errors very similar to those presented in their oral
language; the CA group, as well as presenting literal substitutions, also
showed anticipations and self-corrections. Omissions of letters were ob-
served in the BA group, but the omission of grammatical connectors de-
scribed in these patients by Benson (1977) were not necessarily observed.
A tendency to literal reading was observed in the A WA group, as was the
tendency to invent the right half of the word, thus producing mor-
phological paralexias. These patients usually recognized and read the
initial letter in a word adequately but invented the rest of it. This defect,
which has already been defined in these patients (Benson, 1979), could be
explained by their greater verbal alexia and lesser literal alexia. It was of
note that none of the three patients in this group was aware of the alexical
defect presented and constantly tried to spell the words they were pre-
sented with as if it were the usual reading procedure they used. Naming
errors of letters were evident only in the AA group; the circumlocutions
these patients used when reading a letter were very similar to those used
in spontaneous language. None of the patients presented semantic verbal
substitutions, while morphological verbal substitutions were observed in
the AWA group.
It is important to emphasize that none of the patients presented se-
mantic paralexias. The nearest to semantic paralexias were the mor-
phological verbal paralexias principally presented by patients in the
AWA group and the tendency in the WA group to make the logotomes
meaningful. It would seem that semantic paralexias are not produced in
Spanish speakers as a result of cerebral lesions. A further point could be
added here: The reading of logotomes is equivalent in Spanish to the
reading of low-frequency words. Perhaps the lack of meaning would give

the logotomes a greater level of difficulty, but the reading strategies used
(grapheme-phoneme decoding) would probably be the same.
Analysis of the results obtained in the writing subtests of each of the
left groups showed that the largest number of errors were presented by the
WA, CA, and AWA groups and smallest number by the PF group. In spite
of the fact that the patients with AWA presented a greater defect in read-
ing than in writing, they made numerous writing errors due to omissions.
Moreover, this was the only group whose copy writing showed more
errors than writing to dictation. The patients from the WA group present-
ed paragraphias due to literal substitutions, perseverations, omissions,
and even neologisms. Paragrammatism in spontaneous writing was ob-
served in this group.
Agrammatism and, to a lesser extent, neologisms and omissions were
characteristic of the CA group. Awareness of their errors was evident in
this group since they presented numerous attempts at correction; these
patients seemed to have forgotten the graphemic representation of
phonemes since they repeated the word they wanted to write several
times to themselves as if they were trying to recover the written memory
trace of the words. Difficulty to write spontaneously was also frequent in
these patients, as it was in some patients with BA, and they made com-
ments such as "I can't think how to write." In the BA group anticipations,
literal paragraphias, and agrammatism were observed.
Although the patients with right-hemisphere lesions (pre- and
postrolandic) showed very similar results in the overall scores on the
reading and writing tests, differences were observed on some reading
subtests, such as the reading of logotomes and words and the comprehen-
sion of texts. These tests were particularly compromised as a result of
errors due to the left neglect, which frequently made them invent the
omitted half of the word in order to produce a semantic content (e.g., on
reading caballo ("horse"), the patient initially read allo but in order to
give it a meaning placed a t in front to obtain the word tallo ("stem"). As
well as this type of error, the retrorolandic patients presented defects in
following the lines. Although the patients with right lesions did not pre-
sent defects in the symbolic recognition of letters, the spatial defects
mentioned above together with a complex anosognosia resulted in low
scores on some of the subtests and a great difficulty in text comprehen-
sion.
In the writing tests, patients with right retrorolandic lesions present-
ed omissions and additions of letters in words, and omissions and addi-
tions of strokes in letters. The additions of strokes and letters can be
explained in two ways: On the one hand, they could be the result of the
tendency to perseverate and not inhibit an initiated movement-this ten-
dency was clearly observed in their copy of the Rey figure. On the other
hand, it could be explained also by their left hemi-inattention-the pa-

tients forget the letter they have just written since it is in the left visual
field, which is neglected, and as a consequence the patient repeats the
letter; e.g., on writing the word libra ("book"), the patient writes libbrao.
Other errors observed in this group of patients were related to an inade-
quate separation of the words, since the patients forgot the spaces be-
tween them and sometimes even grouped words together or put in spaces
in the middle of a word, thus fragmenting it. Furthermore, the wide left-
hand margins due to hemi-inattention were very clear, as was the pres-
ence of the "cascade phenomenon"-that is, an increasingly greater in-
crease or decrease of the left-hand margin (Figure 8). The right patients
began to write at one point on the page, and as they went on writing, they
would begin the following lines at points increasingly apart (or closer),
thus forming an oblique line traced with the first letter of each line. Great
difficulty in conserving the written line in a horizontal position was also
observed. These spatial difficulties in writing were related to the greater
or lesser difficulty the patients had in copying the Rey figure. It could
then be supposed that spatial agraphia is secondary to constructional
apraxia.
FIGURE 8. Sample of writing of a patient with right hemisphere lesion. "Cascade" phe-
nomenon: Margins increase each time a new line is begun.
Topography of the Clinical Findings

If we correlate the clinical findings with the scans we could conclude
the following:
1. Lesions of the insula could alter oral and written language in a
differential way. The most anterior portion of the insula associated with
left motor lesions (the opercular area) was related to Broca-type aphasias,
agrammatism, and substitutions due to anticipations in writing; in read-
ing, omissions of letters were observed. The middle portion of the insula
seems to be correlated with conduction aphasias and hence with defects
in repetitive language. This insular region with the anterior portion of the
parietal lobe was correlated with a certain written agrammatism, pho-
nological paragraphias, and omissions in writing. Finally, lesions of the
posterior portion of the insula associated with lesions of the upper tem-
poral cortex (first gyrus) were correlated with Wernicke-type aphasias,
with phonological substitutions and neologisms in reading and with pho-
nological paragraphias, perseverations, omissions, neologisms, and para-
grammatism in writing. It is important to remember the fact that some
patients with W A and AA presented a certain degree of involvement of
the angular region, and perhaps for this reason similarities could be ob-
served in the alterations in written language in these two groups.
2. The posterior region of the parietal and temporal lobes (the parieto-
temp oro-occipital junction) was correlated with defects in the naming of
objects (anomic aphasia), with difficulties in naming letters and literal
substitutions in writing.
3. Damage of the occipital lobe was correlated with alexia without
agraphia. Although there were few cases in this group, it was observed
that in all the patients there was a tendency to spell written words, but
performance in reading letters was much better than in reading words.
Confabulation in the three patients, secondary to their tendency to guess
the written word, constituted morphological paralexias. The three pa-
tients presented a dissociation between their ability to read and their
ability to write. However, they all presented some paragraphias charac-
terized by the omissions and substitutions of letters.
4. The deep cortical parietal region of the right hemisphere on the
side of the deep posterior temporal region was associated with left hemi-
inattention, a tendency to invent the left side of the words and difficulties
in following the line in reading. In writing it was correlated with omis-
sions and additions of strokes, cascade phenomenon, fragmentation, and
grouping of elements. These alterations were correlated with construc-
tional apraxia in copying the Rey figure.
5. The prefrontal regions of the two hemispheres were only slightly
correlated with alterations in written language.
An Interlinguistic Consideration
Finally, it is worthy to mention that the psycholinguistic models of
the alexias and agraphias developed in other languages, particularly in
English, do not seem to be applicable to Spanish. In this work (and in all
the authors' personal experience) no semantic paralexia appears in Span-
ish-speaking aphasic patients. Morphological verbal paralexias do, how-
ever, frequently appear: One part of the word is read correctly (usually the
first morpheme) and the rest is deduced (e.g., Ikonsiderandol - .lkon-
siderasyonl; "considering", "consideration"). Moreover, given the writ-
ing system of Spanish, there are no appreciable differences between the
reading of logotomes and the reading of low-frequency words. The writing
system of the Spanish allows the correct reading of even completely un-
known words. Spelling is a task with a low level of difficulty.
The existence of a double system of reading (phonological and se-
mantic) would not then seem to be acceptable, as has been proposed for
English. This seems to be a fundamental point of view: while Spanish
presents an almost complete phonological writing system, the English
writing system is partially phonological and partially logographic (Samp-
son, 1985).
For Spanish speakers, the underlying cognitive operation during
reading is to convert graphemes into phonemes. This is not totally true in
English: Reading is achieved at a more morphological level. For Spanish
speakers it is just surprising to approach the English writing system, as it
is perhaps for English speakers to approach an ideographic system. A
Latin-American woman living in the United States for many years ex-
pressed this feeling in a very descriptive way: "The problem with English
is that you read one thing, but you have to say another." For her, of course,
reading could be thought of only as converting graphemes into phonemes.
Our point, simply, is that reading in English and reading in Spanish
represent quite different cognitive activities. Consequently, brain repre-
sentation of written language and models for alexias and agraphias have
to be somehow different. Many more studies are needed in this regard.
ACKNOWLEDGMENTS
This research was possible thanks to a grant received from Colcien-

cias (Fondo Colombiano de Investigaciones Cientificas y Proyectos Es-
peciales "Francisco Jose de Caldas") and the support received from the
Faculty of Psychology of the Fundacion Universitaria Konrad Lorenz
(Bogota, Colombia). All the patients were taken from the Departament of
Neurology of the Hospital San Juan de Dios (National University of Co-
lombia) and the Neurological Institute of Colombia. Beatriz Penagos, An-
gela Flores, Catalina Ramirez, and Lissy Sperber helped in the collection
of the data. The English version was prepared by Jenny Lewis. To all of
them our gratitude.
REFERENCES
Ardila, A., Montanes, P., Bernal, B., & Serpa, A. (1986). Partial psychic seizures and brain
organization. International Journal of Neuroscience, 30, 23-32.
Ardila, A., & Ostrosky, F. (1984). The right hemisphere: Neurology and neuropsychology.
London: Gordon and Breach.
Benson, D. F. (1977). The third alexia. Archives of Neurology, 34, 327-331.
Benson, D. F. (1979). Aphasia, alexia and agraphio. New York: Churchill Livingstone.
Benson, D. F., & Cummings, J. L. (1985). Agraphia. In J. A. M. Frederiks (Ed.), Handbook of
clinical neurology, Vol. 45: Clinical neuropsychology. Amsterdam: Elsevier.
Bradshaw, J. L., & Nettleton, N. C. (1981). The nature of the hemispheric specialization.
Bub, D., & Kertesz, A. (1982). Deep agraphia. Brain and Language, 17, 146-165.
Damasio, A., & Damasio, H. (1983). The anatomical basis of pure alexia. Neurology, 33,
1573-1583.
De Renzi, E., & Faglioni, P. (1978). Normative data and screening power of a shortened
version of the Token Test. Cortex, 13, 424-433.
Edwards, A. L. (1968). Experimental design in psychological research. New York: Holt,
Rinehart & Winston.
Geschwind, N. (1965). Disconnection syndromes in animals and man. Brain, BB, 237-294.
Goodglass, H., & Kaplan, E. (1979). Evaluacion de las afasias y de transtornos similares.
Buenos Aires: Editorial Medica Panamericana.
Greenblatt, S. H. (1983). Localization of lesions in alexia. In A. Kertesz (Ed.), Localization in
neuropsychology. New York: Academic Press.
Hecaen, H. (1962). Clinical symptomatology in right and left hemisphere lesions. In V. B.
Mountcastle (Ed.), Interhemispheric relations and cerebral dominance. Baltimore:
Johns Hopkins.
Hecaen, H., & Albert, M. (1978). Human neuropsychology. New York: Wiley.
Kertesz, A. (1983). Localization of lesions in Wernicke's aphasia. In A. Kertesz (Ed.),
Localization of neuropsychology. New York: Academic Press.
Lezak, M. D. (1983). Neuropsychological assessment. New York: Oxford University Press.
Luria, A. R. (1977). Las funciones corti cales superiores en el hombre. La Habana: Editorial
Orbe.
Marshall, J. C., & Newcombe, F. (1973). Patterns of paralexia: A psycholinguistic approach.
Journal of Psycholinguistics, 2, 175-199.
Marshall, J. C., & Newcombe, F. (1980). The conceptual status of deep dyslexia. In M.
Colheart, K. Patterson, & J. V. Marshall (Eds.), Deep dyslexia. London: Routledge and
Kegan Paul.
Miceli, G., Silveri, M. C., & Caramazza, A. (1985). Cognitive analysis of a case of pure
dysgraphia. Brain and Language, 25, 187-212.
Naeser, M. A. (1983). CT scan lesion size and lesion locus in cortical and subcortical ap-
hasias. In A. Kertesz (Ed.), Localization in neuropsychology, New York: Academic
Press.
Novoa, O. P., & Ardila, A. (1987). Linguistic abilities in patients with prefrontal damage.
Brain and Language, 30, 206-225.
Osterrieth, P. A. (1944). Le test de copie d'une figure complexe. Revue de Psychologie, 30,
206-256.
Poeck, K., Bleser, R., & Keyserlingk, T. (1984). Computed tomography localization and
standard aphasic syndromes. In J. J. C. Rosen (Ed.), Advances in neurology: Progress in
aphasiology (Vol. 42). New York: Raven Press.
Roeltgen, D. P., Rothi-Gonzalez, E., & Heilman, K. M. (1986). Linguistic semantic agraphia: A
dissociation of the lexical spelling system from semantics. Brain and Language, 27,
257-280.
Roeltgen, D. P., Sevush, S., & Heilman, K. M. (1983). Phonological agraphia, writing by the
lexical-semantic route. Neurology, 33, 755-765.
Rosselli, M., & Ardila, A. (1987). Alteraciones de la lectura, la escritura y el calculo. In J.
Bustamante & F. Lopera (Eds.), E1 lenguaje: Principios de neurolinguistica. Medellin:
Prensa Creativa.
Rosselli, M., Rosselli, A., Vergara, I., & Ardila, A. (1985). Topography ofthe hemi-inattention
syndrome. International Journal of Neuroscience, 27, 165-172.
Sampson, G. (1985). Writing systems. Stanford, CA: Stanford University Press.
9
Semantic Aphasia Reconsidered

ALFREDO ARDILA, MARIA VICTORIA LOPEZ, and
EUGENIA SOLANO
In 1920 Head described a language alteration that he defined as an in-

ability to recognize simultaneously the elements within a sentence, and
he called this semantic aphasia. Only a few authors referred to this ap-
hasia during the following years (Conrad, 1932; Goldstein, 1948; Zucker,
1934). Luria (1966, 1970, 1973, 1976) took up the concept again and
analyzed it extensively. With the exception of a few references to it in
Western literature over the last few years (Ardila, 1981, 1984; Brown,
1972; Kertesz, 1979) and the report of three cases elaborated by Hier,
Mogil, Rubin, and Komros in 1980, few authors have shown special in-
terest in studying semantic aphasia. However, an analysis of this aphasia
is of great help in the clarification of the organization of spatial concepts,
the mechanisms underlying acalculia and certain types of apraxia and
agnosia, the understanding of the role of the left posterior parietal region
in the organization of cognitive activity and of its interhemispheric asym-
metry. Probably the term itself, semantic aphasia, is not very appropriate,
but neither are other terms such as transcortical or conduction aphasia,
broadly used in aphasiology.
Luria (1970, 1976) considers that difficulties in the following aspects
of language appear in semantic aphasia:
1. Sentences that include a complex system of successive subordi-
nate clauses, particularly forms that include the conjunction
kotory ("which" or "that"), prepositions and conjunctions of the
type nesmatria ("in spite of"), vsledstvie chego ("as a result of").
ALFREDO ARDILA • Konrad Lorenz Foundation University, Bogota, Colombia, and Miami
Institute of Psychology, Miami, Florida 33166-6612. MARIA VICTORIA LOPEZ. Co-
lombian Association of Neuropsychology, Bogota, Colombia. EUGENIA SOLANO'
Neurologic Institute of Colombia, Bogota, Colombia.
177
2. Reversible constructions, particularly of the temporal and spatial

type (krug pod kvadratom-"the circle underneath the square").
3. Constructions with a double negative (Ia ne privyk ne podchiniat-
sa preavilam- "I am not used to not obeying the norms").
4. Comparative sentences (sian volshe muhi- "the elephant is big-
ger than the fly").
5. Passive constructions (zemlia tsvetlaetsa sointsem- "the earth is
illuminated by the sun").
6. Constructions with transitive verbs (prosit- "to ask").
7. Constructions with attributive relations (brat otsa, otets brata-
"the father's brother, the brother's father").
According to Luria (1976), left posterior parietal or parietal-temporal-

occipital lesions produce components of spatial apraxia and agnosia, se-
mantic aphasia, and acalculia. Patients do not show alterations in
phonemic discrimination or word articulation, nor do they exhibit defects
in the comprehension of words and simple sentences. However, these
patients do experience difficulty in understanding grammatical struc-
tures, especially the relationships between words within a sentence.
Therefore, although word comprehension does not suffer, this is not true
of comprehension of the sentence structure. The patient is unable to han-
dle prepositions and spatial or quasi-spatial structures (Luria, 1973).
On occasion, it has been shown that patients with difficulties in
understanding syntax suffer from spatial disorders characterized by some
constructional apraxia and elements of the Gerstmann syndrome. Luria
(1973) suggested that these spatial disorders do not just incidentally ac-
company semantic aphasia, but the semantic aphasia is a defect in the
perception of simultaneous spatial structures only transferred to a higher
symbolic level. Consistent with Luria's hypothesis, patients with seman-
tic aphasia have difficulty in understanding the meaning of words tinged
with spatial or quasi-spatial meaning.
Thus, semantic aphasia is not merely a deterioration in the under-
standing of words tinged with spatial meaning, but the comprehension of
logicogrammatical structures has also deteriorated. These patients have
lost their ability to integrate the elements of a sentence into a whole. As a
result, they are unable to grasp the meaning of relationships. The patient
is aware of the fact that certain words in the sentence should be combined
for him to find the meaning he is looking for but is unable to understand
the interrelations. In grammatical constructions there is a common factor:
All of them in one way or another are a verbal expression of logicogram-
matical relations of a quasi-spatial kind that become apparent, for exam-
ple, in the use of prepositions.
We shall analyze two cases of left posterior parietal damage with
SEMANTIC APHASIA RECONSIDERED 179
semantic aphasia in an attempt to compare its characteristics with those

of patients who have been systematically reported in the literature (Figure
1). And we shall analyze the basic cognitive defects underlying the defi-
cits found in these patients.
~
--- ....... ~ ~
CASE I ~ LUR IA 1~101 CASE :t (I.UR IA 1970 ] CASE 1 (LUR IA "70)
~ -;::"' ....
~ ~....... ~
---- .......
CASE.... I L.URIA 1970) CA SE 50 (LUR I A 197(]l H I ER el il l. IU:801
~ ~ ~:'." ~
~""
HIEA. ct al . 11980) HIER et .at. ~ "801 ARO I L A {UNPUBLISHED)
~ ~ ~
PATIENT 1 PATIENT 2 SU PER IMPOS I T ON
I
FIGURE 1. Topography of damage in 11 patients reported to have semantic aphasia. The

existence of a posterior parietal superimposition can be observed. The last two cases (patient
1 and patient 2) correspond to the two cases analyzed in this chapter.
CASE REPORTS
Patient 1
This is a 21-year-old right-handed woman with 12 years of schooling,
monolingual (Spanish), who in November 1983 presented involuntary
contraction of her right arm with subsequent loss of consciousness.
The only important fact in her history is a bilateral fronto-parieto-
temporal pulsatile headache from the age of 13. Her examination on
admittance to the Neurological Institute of Colombia proved to be
normal except for a discrete hypoesthesia in the right hemibody. As a
result of the scans she was taken to the operating room, where a left
posterior par~etal glioma was removed. She subsequently received co-
balt therapy. During the following years, she remained apparently
symptom-free, suffering occasionally from paresthesias; on one occa-
sion she had a right hemibody focal seizure. However, she stopped
studying and worked sporadically as a manicurist. In June 1986 a
control scan was taken (Figure 2) that recorded the presence of a
mixed-density lesion. This clearly showed the contrasting medium
with well-defined borders and an irregular shape with no great con-
tralateral deviation from the structures of the middle line.
Her neurological examination showed her to be an alert, cooper-
ative patient, with adequate affect, fluent conversational language, and
no evidence of agrammatism or dysprosody. She exhibited occasional
verbal paraphasias and word-finding difficulties. No comprehension
defect was evident in casual conversation. There was extinction in the
presence of simultaneous double tactile stimulation, moderate diffi-
FIGURE 2. Patient 1-brain scan.

culty in the localization of right hemibody stimuli together with

agraphesthesia, astereognosis, finger agnosia, strange feelings in her
hand, and ideomotor and ideational apraxia. The rest of the neu-
rological examination was found to be within normal limits.
Patient 2
This is a 28-year-old right-handed man with 14 years of schooling,
monolingual (Spanish). He was diagnosed as having a left parietal
astrocytoma, which was removed in 1981 after an episode of right
hemibody partial motor seizure with subsequent generalization. The
patient subsequently returned to his work in the government, but only
on a part-time basis, owing to what he called his "tiredness and slow-
ness." He still had a four-fifths right hemiparesis and moderate hypo-
esthesia. Since then he has had frequent partial somatosensory and
motor seizures on his right side, some of them with subsequent gener-
alization. In May 1986 a control scan was taken (Figure 3) that showed
a zone of lesser density in the operative region, multiple calcifications
that do not exert any apparent effect, and dilatation of the left occipital
horns.
His neurological examination showed him to be an alert, cooper-
ative patient, depressed in affect, with fluent language, no evidence of
agrammatism, dysprosody, or articulatory defects, and with occasion-
al verbal paraphasias and word-finding difficulties. A right hemi-
hypoesthesia and a four-fifths hemiparesis was observed. There was
tactile extinction, right agraphesthesia and astereognosia, finger ag-
nosia, and apraxia. The rest of the examination proved to be within
normal limits.
FIGURE 3. Patient 2-brain scan.

Testing Procedure
Each of the patients was given a battery of tests that lasted approx-
imately 10 hours and was divided into l-hour sessions. The battery in-
cluded the following:
1. Wechsler Adult Intelligence Scale (WAIS; Wechsler, 1955).
2. Boston Diagnostic Aphasia Examination (BDAE; Goodglass & Ka-
plan, 1972).
3. Token Test-shortened version (De Renzi & Faglioni, 1978).
4. Rey-Osterrieth Complex Figure.
5. Right-left orientation: The patient indicates on his body and on the
examiner's body simple and crossed instructions, recognition of right and
left on a drawing, reproduction of the examiner's movements with the
appropriate hand (Head's test).
6. Finger gnosis: The patient names/shows the fingers on his and the
examiner's hand; shows, on his hand, the finger that corresponds to the
one indicated by the examiner; indicates that one or two of his fingers are
being touched; says how many fingers there are between the two touched
by the examiner.
In addition to these tests, the following were designed for the purpose
of this research:
7. Reading: letters, syllables, meaningless sequences, words of differ-
ent levels of difficulty, sentences, comprehension of written instructions,
reading of text aloud, reading of ideograms (e.g., Coca-Cola), recognition
of symbols (e.g., danger).
8. Writing: letters, syllables, words with different levels of difficulty,
copying words, dictation and copying sentences, changes in the type of
writing.
9. Calculation: reading of numerals, writing of numbers, transcription
of numbers to letters and vice versa, pointing out the largest number in a
pair, arithmetical operations (mental and written), reading of arithmetical
signs, successive operations (100 - 13; 1,4,7 ... ), counting forward and
backward, ordering numbers in columns, arithmetical problems, appre-
ciation of quantities (e.g., how much does an egg weigh?) and of time (e.g.,
how long does it take a person to walk around a block?).
10. Grammar: conversational language, completion of sentences
(verbs, adverbs, prepositions), changing the tense, article-noun agree-
ment (Spanish has masculine/feminine, singular/plural forms of nouns
and articles), indicating the verb that corresponds to a noun (e.g.,
thought-to think) and vice versa, antonyms (with the use of affixes and
with a change in the stem), saying what action is being performed in a
drawing, ordering the parts of a sentence, understanding of passive sen-
tences, understanding of comparative sentences, understanding of coordi-
nation (e.g., the secretary you sent with your cousin is a friend of Peter),
spatial relationships (with reference to a drawing, say yes/no to a series of

statements (e.g., the square is above the circle).
RESULTS
Intelligence. Table 1 shows the two patients' scores on the WAIS.
Patient l's IQ was 95 and Patient 2's was 91. In both cases, the lower
scores were obtained on the Arithmetic, Digit Span, and Digit Symbol
subtests. Both had higher Verbal than Performance IQs.
BDAE. Patient 1 had errors in auditory and reading comprehension.

Performance was impaired in naming tasks, especially with regard to
body parts. Writing proved to be involved with mistakes due to literal
omission and some verbal paragraphias. Praxic mistakes were also ob-
served in tasks such as blowing out a match, greeting someone, lighting a
cigarette, putting a piece of paper in an envelope. In Patient 2, greater
involvement appeared in the items of auditory and reading comprehen-
sion, which became more accentuated with the increase in the volume of
verbal material. Performance on naming tasks was deficient especially
when they referred to parts of the body. In the case of writing, there were
omissions of letters and syllables, and literal paragraphias. Some praxic
errors to verbal instructions were also evident in tasks such as greeting
someone, teeth cleaning, and using screwdrivers. Figure 4 shows the as-
sessment scale for the language characteristics of both patients. As ex-
TABLE 1. Scaled Scores of the Wechsler Adult

Intelligence Scale
Patient 1 Patient 2
Full scale IQ 93 91
Verbal IQ 95 93
Performance IQ 91 90
Information 10 10
Comprehension 10 13
Arithmetic 7 4
Similarities 10 9
Vocabulary 10 11
Digit span 7 6
Digit symbol 7 5
Picture completion 10 8
Block design 8 10
Picture arrangement 10 8
Object assembly 8 9
....00
2 3 4 5 6 7 "'"
MELODIC LINE I I I 1 I,
intonational contour Absent limited to runs through
short phrases and entire sentence
stereotyped
expressions
1
PHRASE LENGTH
longest occasional (1 /10) word 4 words 7 words
uninterrupted word runs
---
ARTICULATORY AGILITY
facility at phonemic and syllable always impaired normal only in " never impaired
level or impossible familiar words
and phrases ""
GRAMMATICAL FORM
"
variety of grammatical none available limited to simple normal range
'"
construction (even if incomplete) declaratives and
./
stereotypes /'
./
PARAPHASIA IN _ J
RUNNING SPEECH present in every once per minute of ~ absent
utterance conversation~
WORD FINDING t~
informational content in fluent without ~information speech exclusively
relation to fluency
::>
information proportional content words t""
"rj
to fluency
..... G=l
~_ .......... 1 t:l
AUDITORY COMPREHENSION o
converted from objective absent (z= -1.5) (z= -1) (z= -.5) (z= 0) (z= +.5) normal ::>
z-score mean (z= +1) §
F
FIGURE 4. Rating scale profile of speech characteristics (continuous line = patient 1, dotted line = patient 2). ::>
~
2..
pected, the profile is close to that for anomic aphasia (and, indeed, both
patients presented associated anomic defects). However, auditory com-
prehension has been fairly well retained, being almost normal in Patient
1, with a low quantity of paraphasias in discursive speech.
Token Test. The comprehension deficit was slight in Patient 1 (28/36)

but greater in Patient 2 (15/36), who found it impossible to carry out
instructions in which two tokens were involved simultaneously.
Rey-Osterrieth Complex Figure. For both patients, minor errors ap-

peared in their copies of the figure. In particular, there was confusion of
spaces and the omission of details.
Right-Left Orientation. Patient 1 correctly showed right-left on her

and the examiner's body and performed crossed instructions. The only
mistakes were on Head's test. Patient 2 was unable to perform any of the
tasks correctly, and all the answers were at random. He also showed
particularly long latencies.
Finger Gnosis: Neither of the patients was able to perform all the
tests of recognition and naming of fingers. A certain difference was ob-
served in Patient 1, where performance with the left hand was discretely
better.
Reading. In letter reading, both patients revealed spatial confusion

(e.g., p - q, b - d). The reading of words, sentences, and texts showed some
literal paralexias, but comprehension of written instructions and texts
was relatively good. The patient recognized ideographic words and sym-
bols satisfactorily.
Writing. Both patients used their right hand to write. Their writing
was slow, hesitant, and poorly legible, with incorrect spacing, distortions
of letters, and some literal paragraphias (Figure 5). In the case of the 10-
word dictation in the BDAE, correct oral spelling scores were higher than
in the written spelling scores, suggesting that the writing defect is apraxic
or spatial rather than linguistic. The same difficulties were found in
copying.
Calculation. The reading of simple numbers was correct. However,

there were confusions in the reading of compound numbers (e.g., 391
instead of 3,091) and in the reading of Roman numerals. The transcription
of numbers into letters (e.g., 7-seven) and vice versa was correct. The
selection of the larger or smaller of two numbers showed two mistakes
(out of six) in Patient 1 and three in Patient 2. Mental arithmetic opera-
~~~cJo~ ~~ <;;e. -\'("Q.'n~fOf't"t"Q ~'""" ~"ff'05~

(\ e <:,<!' s;1ct 7~'1/fj3
a./JV??t7 5, I'll 5-- ?'O S ~ ~&4-.n-:5;-U!??77~??

e 77 /J77 a... -;p, Fe/of e:t-
r ~
Ylj've! Jt/~Cee,'7i ~~~705

FIGURE 5. Samples of writing of patient 1 (above) and patient 2 (below). The two sentences
dictated are Miguel necesita zapatos ("Miguel needs shoes") and Algunos gusanos se con-
vierten en mariposas ("Some worms become butterflies"). Patient 2 needed approximately 5
min to write these two sentences.
tions with just one figure (e.g., 3 + 5) were possible but impossible with
two digits (e.g., 93 - 13). The same results were found with written
operations. Recognition of arithmetical symbols (+, -, x, :) was correct.
The patients could count adequately forward and backward but could not
perform successive calculations (e.g., 1, 4, 7 ... ). The ordering of num-
bers in columns was adequate. Both made mistakes in solving simple
arithmetic problems and in the appreciation of quantities and time.
Grammar. Conversational speech in both patients was grammatically

correct, and on answering simple questions they always used grammatical
connectors. Table 2 shows their performance on the designed grammar
test. No defects were observed in the use of verbs or tenses, the naming of
actions, and organizations of the parts of a sentence. Errors (or inability)
appeared in the use of adverbs, prepositions, transformations (noun-verb,
verb-noun, adjective-verb), antonyms (particularly with prefixes), pas-
sive sentences, comparative sentences and coordination, and the han-
dling of spatial relationships.
DISCUSSION
Our two patients and all the other patients reported in the literature
presented acalculia in association with the difficulties in the comprehen-
sion of different elements within language. Also, to a certain extent, spa-
tial, praxic, and gnosic defects would point to the proposal that they form
a unified syndrome (Luria, 1976). According to the results of the BDAE,
both patients exhibited mistakes in auditory comprehension (minimal in
case 1, moderate in case 2), some naming defects (particularly with regard
TABLE 2. Performance on the Grammar Test
Part Example Patient 1 Patient 2
Adverb of place The dog hid the door. 2/3 2/3

Adverb of man- Clara took the drug. Now she feels 2/5 3/5
ner
Appropriate verb Yesterday the secretary a 6/6 4/6
letter.
Past Now I'm going to say your name. Yes- 3/3 3/3
terday, your name.
Future Today Susana came to say hello to you. 3/3 3/3
Tomorrow to say hello to
you.
Present Yesterday I listened to a concert. Now 3/3 3/3
to a concert.
Verb to go in its Next week to your house. 5/5 5/5
appropriate
form
Prepositions Most people write their right 22/27 14/27
hands.
Change verb into To think: thought 3/5 4/5
noun
Change noun Entrance: enter 3/5 3/5
into verb
Change adjective Obedient: obey 4/5 3/5
into verb
Antonym (with a Fat: thin 5/6 5/6
different stem)
Antonym (prefix) Real: unreal 0/6 3/6
Naming actions A picture showing a person eating. Cor- 5/5 5/5
rect answer: "to eat"
Organizing the in/fishes/the/river/the/boy 5/5 5/5
parts of a
sentence
Passive sen- A drawing of a doctor examining a pa- 1/4 3/4
tences tient. Say yes or no: "The doctor is
examined by the patient."
Comparative Planes are faster than cars. 9/12 7/12
sentences
Coordinations "The secretary you sent with your cous- 5/10 7/10
in is Peter's friend." Who is Peter's
friend?
Spatial relation- "The square is above the circle." Say 5/10 6/10
ships yes or no.
to parts of the body), some paraphasias in conversational language, and

moderate deficits in reading and writing. The Token Test shows that
comprehension of simple orders is adequate, but comprehension of com-
plex orders, particularly if they include elements of a spatial nature, is
found to be compromised. The agraphia is more of the apraxic than the
188 ALFREDO ARDILA et ai.
aphasic type, and certain spatial confusions appear in reading. Both pa-
tients present finger agnosia-moderate left-right disorientation in Pa-
tient 1 and severe disorientation in Patient 2; severe acalculia; and diffi-
culties in the use and comprehension of adverbs, prepositions, passive
sentences, comparative sentences, and verbal elements with a spatial
content.
The similarity between our cases and those reported by Hier et al.
(1980) is striking. The WAIS, for example, shows, in these authors' results
and in our own, a Verbal IQ higher than a Performance IQ, with decreases
in Arithmetic, Digit Span, and Digit Symbol. The associated disorders
(right-left disorientation, finger agnosia, acalculia, and apraxic agraphia)
are the same. However, despite the apparently similar lesion location and
lesion volume in our patients, the performance in our two patients is
somehow different in some tasks. For example, Patient 1 has a near-
normal language comprehension but Patient 2 has moderately severe
comprehension deficits, based on the Token Test; right-left disorientation
and agraphia were more evident in Patient 2; and the performance on the
grammar test shows some differences, especially in the use of preposi-
tions.
The defects found in our patients typifies the left posterior parietal
syndrome and illustrates the semantic aphasia described by Head (1920)
and analyzed by Luria (1970, 1976). Conversational language is adequate
and fluent, with only a discrete tendency to word forgetfulness. There are
no defects in articulation, phonological recognition, prosody, or the com-
prehension of isolated words, and agrammatism was not observed. How-
ever, language comprehension does not go beyond the limits of isolated
elements or simple sentences, and comprehension of logicogrammatical
structures is seriously compromised. Comprehension of the existing rela-
tionship between the different parts of a sentence is difficult or impossi-
ble, as is the simultaneous synthesis of the different parts of a sentence.
Thus, there is, to say, a "relationship agrammatism" or "impressive
agrammatism" (Luria, 1970). Aspects of a spatial or quasi-spatial nature
(preposition-e.g., to, from; passive sentences-e.g., from whom and to
whom the action is directed), in particular, show alterations.
It is interesting to compare the effects of posterior parietal lesions and
temporal-parietal-occipital lesions in both hemispheres. Above all, it
should be pointed out that in subhuman primates damage in this region in
either of the two hemispheres produces similar deficits: defects in the
perception of simultaneous stimuli, disorders in the visual control of
movements and in occulomotor control in general, errors in spatial orien-
tation and in going through labyrinths and such (Lynch, 1980). In man,
the effects of equivalent lesions (as happens in the case of damage to any
area of cortical association) produce asymmetric defects that are in some
way parallel. Right posterior parietal lesions imply constructional aprax-
SEtv1ANTIC APHASIA RECONSIDERED 189
ia, difficulty in the performance of spatial tests, hemispatial neglect,

hemiasomatognosia, topographic and spatial agnosia, and such (Ardila &
Ostrosky-Solis, 1984). Lesions in the left homologous area produce se-
mantic aphasia, which on occasion is associated with autotopagnosia,
agraphia, anomic aphasia, disorders in visual exploration, and simul-
taneous agnosia (Botez, 1985; Hecaen & Albert, 1978; Luria, 1966).
LeDoux (1982, 1984), basing his observation on studies of patients
with a divided brain, emphasizes that mechanisms of visuospatial percep-
tion are represented in both hemispheres. Also, the specialization of the
posterior parietal region of the right hemisphere refers not to visuospatial
perception per se but to the regulation of spatially guided behavior that
can be observed in activities such as orientation in the medium according
to spatial signals from the environment. It can be supposed that the ac-
quisition of articulate language by man in some way modified the relative
basic interhemispheric symmetry found in subhuman primates. Although
the fundamental function continued to be the same (knowledge of space
and one's own body with respect to external space), it began to suffer the
intervention of language (spatial knowledge with the intervention of lan-
guage), such as the use, for example, of logicogrammatical relationships of
a spatial kind and arithmetical abilities.
Angular acalculia is a primary acalculia since it is considered not to
be a consequence of other cerebral disturbances (e.g., attentional disor-
ders, amnesia, aphasia; Boller & Grafman, 1985). On occasion it has been
included within the so-called Gerstmann syndrome (1940), together with
finger agnosia, agraphia, and right-left disorientation, a syndrome that
Gerstmann considered to be a consequence of a disorder in the "body
scheme."l
An extensive study of Hecaen, Angelergues, and Houiller (1961) rep-
resents a key point in the neuropsychological analysis of the acalculia.
The authors distinguish three groups of acalculia: (1) alexia and agraphia
for numbers (left temporal-occipital damage), (2) spatial acalculia (right
parietal-temp oral-occipital lesions), and (3) anarithmia (left retrolandic
lesions). Other forms of acalculia have been mentioned in the literature;
for example, Grewel (1960) speaks of frontal acalculia and is corroborated
by Luria (1973). The acalculia that accompanies Gerstmann syndrome
would be anarithmia (Boller & Grafman, 1985). Critchley (1953) empha-
sizes the spatial deficit in acalculia, which is especially apparent in writ-
lWe should remember that the word digit comes from digitus. the Latin for "finger." Digits
can be taken to be the fingers of the hand. and we can see that the act of calculating stems
from counting the fingers of the hand. This association between fingers and counting seems
to go back to the very origin of calculation (Cauty. 1984). The association between finger
agnosia and acalculia would not then be fortuitous as far as cerebral organization is con-
cerned. nor would the fact that both can be included simultaneously within the same
syndrome.
190 ALFREDO ARDILA et oJ.
ten calculations, and related it to the parietal damage. Kinsbourne and

Warrington (1962) indicate the simultaneous presence of computational,
spatial, and symbolic errors in acalculia. Luria (1966, 1973) observes that
the difficulties with calculation in patients with left parietal damage are
related to the deficits in grammatical structures of complex numbers.
The distinction between alexia and agraphia for numbers and ana-
rithmia, although conceptually valid, is, in practice, sometimes difficult
to establish. Damage topography is similar (Levin & Spiers, 1985), and if
the existence of Gerstmann syndrome is accepted, it must be supposed
that the anarithmia is associated with at least certain agraphia. Deloche
and Seron (1984) emphasize the existence of a double code (alphabetic
and numerical) in calculation. Using trans coding tasks between codes,
they observed the presence of errors due to position in a series (130-140;
lexical errors) in patients with Wernicke's aphasia (the opposite of what
happens with Broca's aphasia). In Broca's aphasia, the errors are stack
errors (140-104)' and as a result, syntactic errors. In patients with Wer-
nicke's aphasia, the errors would be more of a semantic type, and the
deficits in calculation can be related to lexical-semantic errors in the
reading and writing of numbers.
Boller and Grafman (1983, 1985) consider that ability in calculation
can be altered in various ways as a result of (1) the inability to appreciate
the meaning of the names of the numbers, (2) visuospatial deficits that
interfere with the spatial arrangement of the numbers and also with the
mechanical aspects of the operations, (3) inability to remember mathe-
matical facts and to use them appropriately, and (4) defects in mathe-
matical thinking and in the comprehension of the underlying operations.
Perhaps, we could add the inability to conceptualize quantities (quan-
tification) and reverse operations (e.g., add-subtract). Angular acalculia
would correspond more to defects in the second and fourth points fre-
quently associated with defects of the first type (alexic and agraphic
acalculia). The spatial defect (point two) would be a consequence not of
errors in the handling of the external space on performing calculation
operations (spatial acalculia due to right lesions) but of a defect in the
handling of the spatial concepts underlying the numbers and their per-
mutations (like "carry" in arithmetical operations), as has been stated by
some authors as previously observed.
Dahmen, Hartje, Bussing, and Sturm (1982) studied calculation disor-
ders in patients with Broca's and Wernicke's aphasia. Using factor analy-
sis, they were able to identify two different factors: numericosymbolic
and visual-spatial. The slighter calculation defects found in patients with
Broca's aphasia are derived from their linguistic alterations, while with
Wernicke's aphasia defects in visual-spatial processing contribute to cal-
culation difficulties.
The existence of Gerstmann syndrome has been widely debated and

even questioned in the literature (Benton, 1977; Botez, 1985; Poeck &
Orgass, 1966; Strub & Geschwind, 1983). Some authors have reported the
presence of Gerstmann syndrome without aphasia (Roeltgen, Sevush, &
Heilman, 1983; Strub & Geschwind, 1974; Varney, 1984), but the presence
of a possible semantic aphasia has not been specifically explored in the
evaluations used. According to Strub and Geschwind (1983), the localiza-
tion would be angular, with the lesion extending not toward the occipital
lobe (as Gerstmann proposed) but toward the supramarginal gyrus and
inferior parietal. The associated agraphia would be an apraxic agraphia
(and not aphasic), from which it is supposed that it need not necessarily
be related to alexia (Benson & Cummings, 1985). The report of Morris,
Luders, Lesser, Dimmer, and Hahn (1984) on the appearance of a
Gerstmann syndrome with electrical stimulation of the cerebral cortex
would affirm its angular localization. Finger agnosia could be interpreted
as a restricted form of autotopagnosia, and the right-left disorientation
implies difficulties in the application of spatial concepts in the body's
lateral orientation.
Agrammatism in aphasia has classically been considered to be a com-
ponent of Broca's aphasia (Benson, 1979; Hecaen & Albert, 1978; Kertesz,
1985). However, Luria (1970,1976) refers to defects in the comprehension
of the relationships between parts of a sentence found in semantic aphasia
as a "relationship agrammatism" or "impressive agrammatism," thus
opening the door to the existence of a second type of agrammatism, but
this time at a different level (also different from the paragrammatism that,
on occasion, was associated with the Wernicke's aphasias). This would
correspond to a defect in the comprehension of the grammatical struc'ture
of the sentence or an error in the comprehension of the syntax (Hier et aI.,
1980).
Schwartz, Saffran, and Marin (1980) used tasks similar to those used
in the assessment of semantic aphasia on agrammatical patients with
Broca's aphasia: comprehension of active and passive structures, locative
statements, and transitive verbs. The aphasics made mistakes on all of
these tasks, which, from the point of view of the performance, would
make them equivalent to patients with semantic aphasia. However, the
authors find that their defects are the result of a different underlying
factor: errors in the comprehension of word order within a sentence.
Thus, they conclude that agrammatic patients have a syntactical mapping
defect such that they are unable to utilize a fixed and principal set of
procedures to recover the relational structure of spoken sentences. Broca's
aphasics are capable of processing the semantic features of spatial prepo-
sitions (Goodglass, Gleason, & Hyde, 1970), but they fail to decode the
syntax of words even in simple, active declarative sentences. In semantic
192 ALFREDO ARDILA et aJ.
aphasia the difficulty would depend more exactly on the semantics of the
spatial relationships underlying passive sentences (from whom and to
whom the action is directed), locative statements, and transitive verbs.
It would seem, then, that it is possible to distinguish basic cognitive
defects that in some way could underlie the diversity of neuropsycholo-
gical alterations found in the case of left posterior parietal and parietal-
temporal-occipital damage: defects in spatial conceptualization, spatial
through which language intervenes, knowledge of the external spatial
environment (and the spatial dimensions of one's own body and in the
spatial organization and sequencing of its movements) by means of verbal
symbols and such. In other words, Gerstmann syndrome (left angular
syndrome) and the so-called semantic aphasia are manifestations of the
same underlying cognitive deficits. They would conform a unified neuro-
psychological syndrome.
REFERENCES
Ardila, A. (1981). Las afasias. Bogota: Instituto Neurologico de Colombia.
Ardila, A. (1984). Neurolinguistica. Mexico: Trillas.
Ardila, A., & Ostrosky-Solis, F. (1984). The right hemisphere and behavior. In A. Ardila & F.
Ostrosky-Solis (Eds.), The right hemisphere: Neurology and neuropsychology pp. 3-
50). New York: Gordon and Breach.
Benson, B. F. (1979). Aphasia, alexia and agraphia. New York: Churchill Livingstone.
Benson, B. F., & Cummings, J. 1. (1985). Agraphia. In J. A. M. Frederiks (Ed.), Handbook of
clinical neurology, Vol. 45: Clinical Neuropsychology (pp. 457-472). Amsterdam:
Elsevier.
Benton, A. 1. (1977). Reflection on the Gerstmann syndrome. Brain and Language, 4, 45-62.
Boller, F., & Grafman, J. (1983). Acalculia: Historical development and current significance.
Brain and Cognition, 2, 205-223.
Boller, F., & Grafman, J. (1985). Acalculia. In J. A. M. Frederiks (Ed.), Handbook of clinical
neurology, Vol. 45: Clinical neuropsychology (pp. 473-482). Amsterdam: Elsevier.
Botez, M. I. (1985). Parietal lobe syndromes. In J. A. M. Frederiks (Ed.), Handbook of clinical
neurology, Vol. 45: Clinical neuropsychology (pp. 63-86). Amsterdam: Elsevier.
Brown, J. W. (1972). Aphasia, apraxia and agnosia. Springfield, IL: Thomas.
Cauty, A. (1984). Taxonomie, syntaxe et economie des numerations parlees. Amerindia, 9,
111-146.
Conrad, K. (1932). Versuch einer psychologischen Analyse des parietal syndromes.
Monatschrift fiir Psychiatrie und Neurologie, 34.
Critchley, M. (1953). The parietal lobes. London: Arnold.
Dahmen, W., Hartje, W., Bussing, A., & Sturm, W. (1982). Disorders of calculation in aphasic
patients-Spatial and verbal components. Neuropsychologia, 20, 145-153.
Deloche, G., & Seron, X. (1984). Some linguistic components of acalculia. In F. C. Rose (Ed.),
Advances in neurology, Vol. 42: Progress in aphasiology (pp. 215-222). New York:
Raven Press.
De Renzi, E., & Faglioni, P. (1978). Normative data and screening power of a shortened
version of the Token Test. Cortex, 14,41-49.
Gerstmann, J. (1940). The syndrome of finger agnosia, disorientation for right and left,
agraphia and acalculia. Archives of Neurology, Neurosurgery and Psychiatry, 44, 398-
408.
Goldstein, K. (1948). Language and language disorders. New York: Grune & Stratton.
Goodglass, H., Gleason, J. G., & Hyde, M. R (1970). Some dimensions of language com-
prehension in aphasia. Journal of Speech and Hearing Research, 13, 595-606.
Goodglass, H., & Kaplan, E. (1972). Assessment of aphasia and related disorders.
Philadelphia: Lea Febiger.
Grewel, F. (1969). The acalculia. In P. J. Vinken & G. W. Bruyn (Eds.), Handbook of clinical
neurology (pp. 181-196). (Vol. 4). Amsterdam: North-Holland.
Head, H. (1920). Aphasia and kindred disorders of speech. Brain, 43, 87-165.
Hecaen, H., & Albert, M. L. (1978). Human neuropsychology. New York: Wiley.
Hecaen, H., Angelergues, T., & Houillier, S. (1961). Les varietes cliniques des acalculies au
cours des lesions retrorolandiques: Approche statistique du probleme. Revue Neurologi-
que, 105, 85-103.
Hier, D. B., Mogil, S. I., Rubin, N. P., & Kornros, G. R (1980). Semantic aphasia: A neglected
entity. Brain and Language, 10, 120-131.
Kertesz, A. (1979). Aphasia and associated disorders. New York: Grune and Stratton.
Kertesz, A. (1985). Aphasia. In J. A. M. Frederiks (Ed.), Handbook of clinical neurology, Vol.
45: Clinical neuropsychology (pp. 287-332). Amsterdam: Elsevier.
Kinsbourne, M., & Warrington, E. K. (1962). A study of finger agnosia. Brain, 85, 47-66.
LeDoux, J. E. (1982). Neuroevolutionary mechanism of several asymmetry. Brain, Behavior
and Evolution, 20, 197-213.
LeDoux, J. E. (1984). Cognitive evolution: Clues from brain asymmetry. In A. Ardila & F.
Ostrosky-Solis (Eds.), The right hemisphere: Neurology and neuropsychology pp. 51-
60). New York: Gordon and Breach.
Levin, H., & Spiers, P. A. (1985). Acalculia. In K. M. Heilman & E. Valenstein (Eds.), Clinical
neuropsychology (2nd ed., pp. 97-115). New York: Oxford University Press.
Luria, A. R (1966). Higher cortical functions in man. New York: Basic Books.
Luria, A. R (1970). Traumatic aphasia. The Hague: Mouton.
Luria, A. R. (1973). The working brain. New York: Basic Books.
Luria, A. R (1976). Basic problems of neurolinguistics. The Hague: Mouton.
Lynch, J. (1980). The functional organization of posterior parietal association cortex.
Morris, H. H., Luders, H., Lesser, R P., Dinner, D. S., & Hahn, J. (1984). Transient neuropsy-
chological abnormalities (including Gerstmann's syndrome) during cortical stimulation.
Neurology, 34, 877-883.
Poeck, K., & Orgass, B. (1966). Gerstmann's syndrome and aphasia. Cortex, 2, 421-437.
Roeltgen, D. P., Sevush, S., & Heilman, K. M. (1983). Pure Gerstmann's syndrome from a
focal lesion. Archives of Neurology, 40, 46-47.
Schwartz, M. F., Saffran, E. M., & Marin, O. S. M. (1980). The word order problem in
agrammatism: Comprehension. Brain and Language, 10, 249-262.
Strub, R, & Geschwind, N. (1974). Gerstmann's syndrome without aphasia. Cortex, 10, 378-
387.
Strub, R, & Geschwind, N. (1983). Localization in Gerstmann syndrome. In A. Kertesz (Ed.),
L~calization in neuropsychology (pp. 295-322). New York: Academic Press.
Varney, N. R. (1984). Gerstmann syndrome without aphasia: A longitudinal study. Brain and
Cognition, 3, 1-9.
Wechsler, D. (1955). Wechsler Adult Intelligence Scale: Manual. New York: Psychological
Corporation.
Zucker, K. (1934). An analysis of disturbed function in aphasia. Brain, 57.
IV
Basal Ganglia and Cognitive

Activity
10
Cognitive Effects of Adrenal

Autografting in Parkinson's Disease
FEGGY OSTROSKY-SOLIS, IGNACIO MADRAZO,
RENE DRUCKER-COLIN, and LUIS QUINTANAR
The microsurgical autografting of adrenal medullary tissue to the caudate

nucleus has recently been implemented to treat patients with Parkinson's
disease (PD) who are no longer satisfactorily responding to pharmacolog-
ical treatment. Initial results show significant improvement in the motor
symptomatology of these patients (Drucker-Colin et aI., 1988; Madrazo et
aI., 1987).
Clinical studies have indicated that some degree of cognitive impair-
ment is as much a feature of PD as the classical motor symptoms. In fact, a
spectrum of cognitive disorders have been reported in PD patients. In
some patients, specific deficits in areas such as memory (Pirozzolo, Hans-
ch, & Mortimer, 1982), language (Matison, Mayeux, Rosen, & Fahn, 1982),
visual perception (Villardita, Smirni, Le Pera, Zappala, & Nicoletti, 1982),
visuospatial processing (Mortimer, Pirozzolo, Hansch, & Webster, 1982),
and behavioral programming under novel conditions (Taylor, Saint-Cyr, &
Lang, 1986, 1987) have been reported, whereas in others, global dementia
has been found (Boller et aI., 1980; Elizan, Sroka, Maker, Smith, & Yahr,
1986; Gaspar & Gray, 1984; Lieberman et a1., 1979). Dementia or the irre-
versible deterioration of intellectual functions, including memory, cogni-
FEGGY OSTROSKY-SOLIS· Department of Psychophysiology, Faculty of Psychology, Na-

tional University of Mexico, Mexico D. F. 4510, Mexico. IGNACIO MADRAZO· Depart-
ment of Neurosurgery, Specialties Hospital, Centro Medico "La Raza," IMSS, Mexico D.F.
02990, Mexico. RENE DRUCKER-COLIN· Department of Neurosciences, Institute of
Cellular Physiology, National University of Mexico, Mexico D.F. 4510, Mexico. LUIS
QUINTANAR. Department of Psychophysiology, Faculty of Psychology, National Univer-
sity of Mexico, Mexico D.F. 4510, Mexico.
197
198 FEGGY OSTROSKY-SOLIS et al.
tion, and perception, have been observed in approximately 30% of the PD

patients (Elizan et aI., 1986; Lieberman et aI., 1979).
PD is a chronic neurodegenerative disease in which neural death is
progressive and so motor signs and cognitive deterioration are a slow,
inexorable process. There is at present no pharmacological treatment for
the progressive deterioration of these patients.
PATHOLOGY AND PATHOGENESIS
In 1817, Parkinson described the clinical characteristics of the illness

nowadays known as idiopathic PD, but it was not until 1912 that Lewy
described a pathological abnormality known as the "Lewy body"
(eosinophilic cytoplasmic inclusion bodies), which is necessary to con-
firm the PD diagnosis. In 1919 Tretiakoff showed that the cell loss is
located in the substantia nigra, thus giving rise to the anatomoclinical
concept of PD that distinguishes idiopathic PD from other parkinsonic
syndromes (see Agid, Javoy-Agid, & Ruberg, 1987, for a review).
In 1960 the analysis of the biochemistry underlying PD was initiated.
Carlsson showed that the administration of reserpine to rats caused cata-
tonia associated with a decrease in the concentration of dopamine, which
could be reversed by reestablishing the normal concentration of
dopamine with L-dopa, a precursor of dopamine (Carlsson, Lindquest, &
Magnusson, 1957). Ehringer and Hornykiewicz (Ehringer, Hornykiewicz,
& Verteilung, 1960) showed that the concentration of dopamine had de-
creased in the basal ganglia of PD patients. These findings led to the
administration of a precursor of dopamine (Cotzias, Papavasiliov, &
Gellene, 1969).
It has been shown that patients with PD show loss of nerve cells and
depigmentation in the substantia nigra and in other pigmented subcor-
tical nuclei (e.g., the locus coeruleus). The severity of the changes in the
substantia nigra parallels the reduction of dopamine in the striatum.
Given that the pars compacta of the substantia nigra contains most of the
brain's dopaminergic cell bodies, these, observations suggested that the
nigrostriatal dopaminergic pathway is involved in this disease.
Dopamine normally is synthesized in the striatum, in the nerve end-
ings of the dopaminergic neurons whose cellular bodies lie in the substan-
tia nigra; at these nerve endings, the neurotransmitter is taken up into the
vesicles and released in the synaptic cleft when the cells fire.
Neuroanatomically, the denervation of the striatum affects the output
from the striatum to the cortex via the striato-pallido-thalamo-cortical
(motor cortex) system and the nigro (pars reticulata) thalamo-cortical (pre-
motor and prefrontal cortex) system. The concept of "motor" and "com-
plex" loops has been proposed in the relationship between basal ganglia
and frontal lobes (Delong, 1974; Delong, Georgopulos, & Crutcher, 1983).
ADRENAL AUTOGRAFTING IN PARKINSON'S DISEASE 199
In this model, the motor loop is devoted to the control of motor

parameters and involves the agranular sensorimotor and premotor cor-
tical areas, the putamen, the caudal portions of the basal ganglia efferent
system, and a diencephalic relay through the ventral lateral nucleus to the
supplementary motor area. The complex loop has a topographically orga-
nized input from all cortical association areas to the caudate nucleus; it
transmits information to the rostral portion of the basal ganglia efferent
system with diencephalic relays via the ventral anterior nucleus and cen-
tromedian nucleus to the frontal eye fields and frontal association areas
that are involved in cognitive operations (see Taylor et al., 1986, for a
review). Figure 1 shows a schematic representation of the motor and
complex loops.
Functionally, the basal ganglia show a segregation between motor
and nonmotor functions. Thus, the differences in connectivity between
the putamen and the caudate nucleus have led to the notion that the
decrease in dopaminergic activity in the putamen causes the motor symp-
tomatology (tremor, rigidity, and akinesia). Within the motor loop there is
also segregation of functions; for example, it has been speculated that the
loss of pallido-cortical fibers is important in the genesis of the tremor and
that rigidity is related to the loss of putamino-pallidal fibers. The symp-
toms of akinesia and the postural and equilibrium defects are symptoms
derived from the degeneration of the cellular bodies in the substantia
nigra (Selby, 1967). The caudate nucleus seems to be involved in complex
cognitive functions. Symptoms reminiscent of frontal lobe dysfunction
have been reported among PD patients. These frontal lobe type symptoms
are expressed by an impaired ability to order and maintain cognitive
(goal-directed) programs and by frontal type motor signs, such as the
inability to maintain repetitive gestual sequences. In visuoperceptual and
visuospatial tasks, segmentation and loss of figure-ground perspective are
observed (Agid, Ruberg, & Dubois, 1986; Ostrosky-Solis et al., 1988; Tay-
lor et al., 1986).
As Taylor et al. (1987) point out, the normal distribution of dopamine
within the caudate nucleus becomes critical to cognitive functions in
terms of the ability to affect fronto-caudate circuits within the complex
loop that ultimately return information processed in the caudate nucleus
to the prefrontal cortex.
It has also been found that the ventral area adjacent to the pars com-
pacta of the substantia nigra shows a loss of dopaminergic cells in PD
(Javoy-Agid et al., 1984; Uhl, Hedreen, & Price, 1985). This area gives
origin to the meso-limbic-cortical pathway that projects primarily to the
medial frontal area and the limbic areas (nucleus accumbens, amygdala,
cingulate cortex, hippocampus, paraolfactory gyrus, and septum) (Javoy-
Agid & Agid, 1980; Javoy-Agid et al., 1984). Reduction of dopamine in the
frontal cortex areas of PD patients (Scatton, Rouquier, Javoy-Agid, Agid,
1982) seems to be related to the cognitive disorders observed (Agid et al.,
MOTOR LOOP
COMPLEX LOOP
NOTOR
SuPLENENTAR'I" COllTEx
MOTOR AREA
""(MOTOR
CORTEx
TENPO~A L
COllTEX
, ,
,, .
, ,
,, ,,
. ,
,
,, ,,, ,,
,
, .
//
..
VENTRAL VENTRAL
ANTERIOR ANTERIO R
rotuCLEuS NUCL EUS ,,/
VENTt:'lAL
L ATERAL
l<UCLEUS
-----_ ......,>
CENTRONE OIA,... "
THAL ANUS ,, THALAMUS , ~Cl..EUS "
,, , , , "
,
, , ,
",
SUBSTANTIA
fIIlGRA
·· "'t!) ~~:::Ar<TIA
...... - ...
··
SUPERIOR
..
COLL ICULUS
FIGURE 1. Schematic diagram of the motor and complex anatomical circuits. Motor loop: cortico-putamino-pallido-thalamo-cortical (motor
cortex). Complex loop: cortico-caudato-nigro-thalamo-cortical (premotor and prefrontal cortex). (Adapted from DeLong, Georgopulos, & Crutcher,
1983.)
1986; Brozoski, Brown, Rosvold, & Goldman, 1979; Lees & Smith, 1983;
Stern & Langston, 1985; Taylor et 01., 1986, 1987).
In spite of the fact that the primary pathology of PD is the degenera-
tion of the dopaminergic projection to the striatum, not all these patients'
symptoms can be attributed to loss of nigrostriatal dopamine. There are
other neurochemical systems that are found to be affected in PD, such as
the noradrenergic cells in the locus coeruleus, serotoninergic neurons in
the dorsal raphe nucleus, and acethylcholine due to lesions in the sep-
tohippocampal system and the substantia innominata. At the cortical
level, a reduction in somatostatine has been reported (see review in Agid
et 01., 1987).
The reduction of acethylcholine and its enzymes in the nucleus
basalis of Meynert has been associated with demential disturbances (Gas-
par & Gray, 1984). As for the other neurotransmitters, no clear relation has
yet been established between the biochemical changes and the clinical
symptomatology. However, in light of data obtained from animal re-
search, it has been hypothesized that the selective alteration in the nor-
adrenergic systems could cause attentional disturbances; the reduction of
serotoninergic metabolism has been associated with depression, and the
decrease in somatostatine at the cortical level has been correlated with
intellectual deterioration (Agid et 01., 1987).
It would appear that the lesions in the different neuronal systems do
not evolve in parallel but may be additives or potentiate one another in
terms of functional expression. From the biochemical point of view, it has
been found that the cholinergic activity in striatum and cortical areas is
related to the quantity of dopamine receptors. In PD patients with a low
number of dopamine receptors, labeling of cholinergic muscarinic recep-
tors revealed decreased numbers in both the caudate nucleus and the
cortex. Individuals with an increased number of dopamine receptors dis-
played evidence of a corresponding increase in cholinergic receptors
(Rinne, 1982). The variety in the extension and the degree of lesions that
has been found among PD patients could be the pathological substratum
for the wide variety of motor and cognitive symptoms that have been
observed.
TREATMENT
On the basis of the physiopathological description of PD, multiple

pharmacological and surgical treatments have been tried.
The motor disturbances can be grouped in two types of deficits: pri-
mary functional deficits directly attributed to the loss of function sub-
served by specific neurons, and secondary deficits that may be caused by
the appearance of an abnormal pattern of action in neurons when part of
202 FEGGY OSTROSKY-SOLIS et 01.
this controlling input (usually inhibitory) is destroyed as a result of the

illness.
Neurosurgical Treatment
The tremor and rigidity of Parkinson's disease have been attributed to
a loss of an inhibitory influence within the basal ganglia, which leads to
the release of the inhibition and to an abnormal outflow of the internal
portion of the globus pallidus to the ventral anterior and lateral nuclei of
the thalamus and finally to the motor cortex. Neurosurgical treatment of
PD began in 1930 (Selby, 1967) and involves stereotaxic lesions in the
globus pallidus or in the ventro-lateral thalamus contralateral to the side
of the body that is most affected. This technique apparently decreases the
abnormal activity (disinhibited) and relieves the tremor and rigidity. The
bradykinesia of the patients does not improve, however (possibly because
it is a primary deficit), and for this reason the performance of daily ac-
tivities remains affected. It has been reported that the tremor and rigidity
reappear 1 to 3 years after surgery. In the cognitive area, different distur-
bances appear as a result of the surgery (especially when performed bilat-
erally), such as deterioration in language and in visuoperceptual func-
tions, and severe conceptual and emotional disturbances (Darley, Brown,
& Swenson, 1975; Riklan & Levita, 1970). The success of pharmacological
therapy has decreased the need for this type of surgery.
Pharmacological Treatment
The cornerstone of pharmacological treatment in PD is the use of the
precursor of dopamine, levodopa, a substance that is able to cross the
hematoencephalic barrier and be transformed into dopamine by the
dopaminergic system and, more recently, the use of dopaminergic agonist
agents.
When PD patients are treated with levodopa, an initial improvement
is observed, and then gradually, over the years, the improvement de-
creases. A large number of patients present secondary effects to the action
of the drug, including dyskinesias or abnormal involuntary movements,
"on-off" phenomena characterized by the presence of severe akinesia and
periods of relative mobility, "wearing-off" or "end of dose" phenomena
characterized by an accelerated deterioration in the beneficial effects, and
psychiatric disorders. These side effects are on occasion more severe than
the disease itself.
The conventional treatments used do not alter the course of the ill-
ness since they produce only a temporal symptomatic relief. Currently,
the need exists to develop new therapeutic approaches that will produce
the medication or appropriate procedure that can supply dopamine in
situ and in an adequate concentration for the individual physiological

demands.
Tissue Transplants
It is in this context that a line of research arose within the basic
neurosciences in which the treatment of experimental models of Parkin-
son's disease with transplanted dopaminergic neurons has been explored.
That is, an attempt has been made to fulfill all the theoretical assumptions
of the ideal drug, infusing dopamine through the neurons or cells that
synthesize it (Le., the substantia nigra or adrenal medulla) into portions of
the brain deficient in this neurotransmitter.
The capacity of several tissues to survive and grow when they are
transplanted within the central nervous system has been known since
Ram6n y Cajal (1938). The central nervous system is an immunologically
privileged organ with a low capacity for rejection and an ideal place to
transplant autologous or fetal tissue (Barker & Billingham, 1977). It is
thought that in the brain the access to the immune system to foreign
tissues is limited for two reasons: (1) The brain lacks lymphatic vessels
and lymph nodes from which many of the cells of the immune system are
deployed, and (2) the walls of the blood vessels in the central nervous
system are specialized in creating a "blood-brain barrier." The absence of
rejection in the neuronal transplants can also reflect the suitable charac-
teristics possessed by the nervous cells. Many cells on their surface bear
large molecules known as Class I major histocompatibility antigens. Dif-
ferent in each animal, the antigens are molecules that the immune system
recognizes as foreign when it rejects grafted tissue. Few of these antigens
are found in neurons.
The development of experimental models of Parkinson's disease and
the verification of the viability of the monoaminergic neurons on being
transplanted within the central nervous system produced a wide field for
research in neurobiology. In 1971 Ungerstedt (1971a, b) described an ex-
perimental model for Parkinson's disease that consists in the destruc-
tion of neurons from the nigrostriatal system with a selective-6-hydrox-
idopamine neurotoxin. This substance selectively destroys the neurons,
fibers, and terminals that contain catecholamines (for example, dopa-
mine). When the dopamine pathways on both sides of the brain are de-
stroyed, a generalized akinetic syndrome is produced that is reminiscent
of the hypokinesia of Parkinsonism and includes interruption of eating
and drinking habits and death of the animal if it is not given intensive
care. Unilateral lesions of the nigrostriatal pathway result in turning
movements away from the side of the lesions that can be measured quan-
titatively by a rotometer. Following unilateral lesions of the substantia
nigra, the ipsilateral striatum responds to denervation by becoming super-
204 FEGGY OSTROSKY-SOLIS et aI.
sensitive to dopamine. If the lesions spare more than 5% of the dopamine-

containing fibers, the animal can recover within 1 or 2 weeks, and rats do
not exhibit abnormal turning unless they are challenged with drugs.
Amphetamine releases dopamine from the intact side, causing ipsilateral
rotation. Apomorphine, which is a postsynaptic dopamine agonist, pro-
duces a greater effect on the denervated striatum, resulting in rotation of
the animal away from the side of the lesion.
Another experimental model of PD is provoked using a neurotoxic
chemical (MPTP). In the course of a few days, the animal develops some
of the characteristics of PD, especially hypokinesia, rigidity, and postural
instability (Forno, Langston, Delanney, Irwin, & Ricaurte, 1986; Langston,
Ballard, Tetrud, & Irwin, 1983). MPTP damages the dopaminergic neurons
that are found in the pars compacta of the substantia nigra and results in a
degeneration of the fibers of the nigrostriatal system and in a loss of
striatal dopamine and its metabolites.
Recent studies employing the above animal models of PD have shown
that grafting of fetal substantia nigra cells to the lateral ventricle adjacent
to the denervated neostratum reverses most of the behavioral, bio-
chemical, and anatomical abnormalities induced by the dopamine-dener-
vating lesions. Biochemically, the transplants restored up to 50% of the
lost caudate dopamine. Anatomically, the transplanted dopaminergic
cells appeared normal and sprouted in a limited manner into the sub-
stance of the caudate, and, behaviorally, all the abnormalities except adip-
sia and aphagia could be reversed (Bjorklund, Dunnett, Stenevi, Lewis, &
Iversen, 1980; Dunnett, Bjorklund, & Stenevi, 1983; Dunnett, Bjorklund,
Stenevi, & Iversen, 1981a,b; Freed, 1983; Perlow, 1987; Perlow et 01.,
1979).
Since transplanting embryonic human tissue for clinical trials raises
ethical difficulties, investigators began to examine alternative sources of
catecholaminergic tissue for grafting to the brain, specifically adrenal
medullary cells. These cells have the same embryonic origin as neurons;
that is, both develop in the ectoderm. They contain large amounts of
catecholamines: adrenaline, noradrenaline, and dopamine. Moreover,
chromaffin cells possess a high degree of phenotypic plasticity. When
they are surrounded by adrenal cortex, they become rounded and absorb
large amounts of epinephrine, but when they are removed from the gland,
placed in culture, and incubated with nerve growth factor, they change
morphologically and biochemically, simulating catecholaminergic cells
(Unsicker, Rieffert, & Ziegler, 1980; Wurtman, Pohorecky, & Baliga, 1972).
In 1981 Freed et 01. began a series of studies in which adrenal medull-
ary allografts were transplanted in the lateral ventricle adjacent to the
denerved striatum and found that apomorphine-induced rotation was in-
duced by the grafts. These and further studies (Freed, 1983) suggested that
behavioral recovery was related to the release of catecholamines, adrena-
line, noradrenaline, and dopamine and their diffusion to hypersensitive

dopamine receptors in the striatum.
A different mechanism of action of the transplant has been suggested
by recent experiments carried out by Bohn, Marciano, Cupit, and Gash
(1987) in rats and by Gash et 01. (1987) in cebus monkeys. Both groups
have found that adrenal medullary grafts into the caudate nucleus pro-
moted rapid recovery of tyrosine hydroxylase immunoreactive fibers that
appear to be from the host rather than from the graft itself. The authors
postulated that adrenal medullary exert a neurotrophic action in the host
brain, promoting recovery of dopaminergic neurons damaged by MPTP.
ADRENAL MEDULLARY GRAFTS IN PD PATIENTS

In 1982, researchers in Sweden made the first attempt to transplant
adrenal medullary tissue into the striatum of PD patients. Using a ster-
eotaxic technique, they transplanted suspensions of autologous adrenal
medullary tissue into four PD patients. The tissue was contained in a steel
spiral to mark its position in the brain. The implants were placed in the
parenchyma of the caudate nucleus in two patients (Backlund et 01.,
1985), and in two others the tissue was placed in the putamen (Lindvall et
01., 1987). With this technique, the patients presented a modest clinical
improvement during short periods of time. As reviewed in Drucker et 01.
(1988), the question arises as to whether placement of adrenal medullary
tissue in the parenchyma is an appropriate procedure, or whether better
survival of this tissue is obtained when it is placed within the ventricle.
Animal work has strongly suggested that the latter is more appropriate
since it appears that survival of adrenal medulla grafts placed within the
striatum is limited, regardless of whether they are introduced as solid
blocks or as dissociated cells. As a result, such grafts do not effectively
induce recovery of apomorphine rotational behavior, although, with the
injection of nerve growth factor at the site of transplantation, the graft
then becomes much more effective. On the other hand, when grafts are
placed within the lateral ventricle, rotational behavior is significantly
reduced without need of nerve growth factor. The fact that the cerebral
ventricles provide a fluid-filled cavity that may act as a nourishing medi-
um for the maintenance of grafted tissue prior to vascularization, and also
provide a medium for transport of neuroactive substances released from
grafts, may explain in part the better results obtained in counteracting
rotational behavior of lesioned animals.
In view of this difference between intra parenchymal and intra-
ventricular placements of grafts in animals, and in view of the very mod-
est improvement reported by the Swedish team, Madrazo et 01. (1987)
modified the procedure and transplanted adrenal medullary tissue in two
young (35- and 39-year-old) PD patients. The adrenal medullary frag-

ments were grafted within the lateral ventricle with partial inclusion
within the head of the caudate nucleus.
Surgical Technique
The surgical procedure involved simultaneous adrenalectomy and
frontal craniotomy. Upon extraction of the adrenal gland under a dissect-
ing microscope, six to eight fragments of adrenal tissue were obtained (0.8
g in total approximately) and placed on a wet surface. Simultaneously, the
caudate nucleus was approached, with the aid of a surgical microscope,
through the lateral ventricle by means of a nontraumatic transcortical
(second frontal gyrus F2). In the head of the caudate nucleus, a 3 x 3 x 3-
mm bed was constructed and the adrenal medullary fragments were
placed within the cavity. Initially, six small fragments were inserted in
the cavity, and then the last two fragments, which were slightly larger,
were anchored to the ependyma of the caudate nucleus with a couple of
stainless-steel miniature staples. In this manner, the inner fragments can-
not dislodge themselves from the cavity, but the cerebrospinal fluid can
bathe all the grafted tissue. In both patients, clinical improvement was
noted at 15 and 6 days (respectively) after implantation and has been
maintained 24 months in the first patient and 18 months in the second.
Rigidity and akinesia were significantly reduced, and functionally both
patients are completely independent in their daily activities and are
working, the first as a farmer and the second as a civil engineer.
Since these results appeared, the autologous graft of adrenal medulla
to the caudate nucleus has been carried out in a larger number of PD
patients who presented severe symptoms of rigidity, tremor, akinesia, and
pronounced on-off effects. Patients have been studied with video, spec-
trophotography, electromyography, neurophysiological, and neuropsy-
chological studies.
NEUROPSYCHOLOGICAL FINDINGS
The technique of autologous graft of adrenal medulla implanted in

the ventricle surface of the caudate nucleus is performed by means of a
right frontal craniotomy, and in order to implant the fragments of adrenal
medulla a cavity is made in the caudate nucleus; for this reason one of the
objectives of the neuropsychological evaluation of the patients subjected
to this surgery is to monitor if any negative effects appear as a result of the
surgery.
The chromaffin cells of the adrenal medulla secrete a large quantity of
substances such as encephalines, somatostatine, neuropeptides, epi-
nephrine, and norepinephrine, as well as dopamine. It has been reported

that different cognitive disorders associated with deficiencies in
dopamine and other neurotransmitters are present in the PD population.
For this reason, another objective of the neuropsychological evaluation
has been to study whether positive effects exist in the cognitive area as a
result of the transplant. A group of operated and unoperated patients with
PD (diagnosed at the Parkinson's Clinic at the Centro Medico "La Raza")
were studied. The operated PD group comprised 18 patients who under-
went autografting of adrenal medullary tissue to the caudate nucleus to
treat PD as described above. They were 12 males and 6 females, whose
ages ranged from 34 to 60 years (mean = 47.4 years). The duration of
illness ranged from 3 to 16 years (mean = 9.5 years). Education ranged
from 1 to 17 years (mean = 9.8 years). The unoperated PD group included
10 patients (6 males and 4 females) who were candidates for surgery.
These patients were matched with the operated PD patients with respect
to illness, severity of symptoms, and duration of treatment. The severity
of the disease was evaluated using the Unified Parkinson Rating Scales,
the Hoehn and Yahr Classification, and the Schwab England Scale, which
measure disability expressed in the activities performed during daily liv-
ing. Given the scores obtained, 5 patients were classified as mild to mod-
eratelyaffected, 7 as moderately to severely, and 6 as severely affected.
Before surgery, the patients had received L-dopa alone or in combina-
tion with other "antiparkinsonian" drugs. The initial response to the
medication had been positive, but there was a gradual decrease of the
beneficial effects, and all patients developed drug-related complications
that included an "end of dose" deterioration phenomenon and disabling
side effects (dyskinesias and/or on-off phenomena).
A series of neuropsychological tests based on Luria's diagnostic pro-
cedures and adapted for the Mexican population were also used (Os-
trosky-Solis et 01., 1985, 1986). These tests explore nine different areas,
which are described in Table 1.
To obtain additional information on behavioral and cognitive func-
tions reported to be at risk among PD patients, such as attention, memory,
verbal fluency, depression, and global dementia, several other psycholog-
ical instruments were applied. Details of the tests and procedure can be
found in the paper by Ostrosky-Solis et 01. (1988).
The preoperative neuropsychological testing was carried out 2 to 4
weeks prior to surgery. In the immediate postoperative period, behavior
was evaluated on a day-to-day basis. Formal neuropsychological testing
was carried out at 3 months and at 12 months after surgery. The control
group was evaluated with the same battery and at the same time intervals
as the operated PD patients.
The preoperative neuropsychological testing revealed specific cog-
nitive deficits, which varied in degree. The patients showed frontal lobe
TABLE 1. Neuropsychological Scheme a
Motor functions Includes tasks that require the coordination, repro-

duction, and repetition of simple and complex
movements with the hand, the arm, and bucofacial
movements. Series of alternating motor activities.
Somatosensory Includes the discrimination of tactile stimuli, recogni-
knowledge tion of shapes, reproduction of hand positions, and
tactile memory.
Visuoperceptual and vi- Explores recognition of simple and complex draw-
suospatial recognition ings, figure-ground discrimination, visual closure,
visual analysis and synthesis, reprodu~tion of draw-
ings and designs, object assembly, and block de-
signs.
Auditory knowledge Assesses the detection and discrimination of pho-
and language nemes, tap-out asymmetrical rhythms, retention and
evocation of a list of 5 meaningless syllables, repeti-
tion of verbal sequences, and recognition of natural
sounds.
Cognitive processes Includes logical reasoning, classification of objects,
understanding of analogies, picture completion, and
picture arrangement.
Oral language Explores the production of simple and complex
words, comprehension of language, verbal learning
curves for bysyllabic words, immediate and delayed
memory for sentences, naming of objects and body
parts, complex grammatical relations and passive
constructions.
Reading Includes recognition of letters, syllables, and words,
oral and silent reading.
Writing Assess automatic writing, copy, and dictation.
Basic calculus Explores mathematical notion and basic arithmetic
operations.
aInciudes 95 items from which 195 scores can be obtained, and emphasizes two aspects: (1) quality of the
mistakes-each item is scored according to one or several criteria and not simply according to whether
the subject performed the task or not, and (2) a simple quantification is carried out under three categories
for each criterion-namely, normal performance; regular performance, moderately anomalous; and im-
possible performance. Hence, the poorer the subject's performance, the higher his score.
type deficits with alterations in behavioral programming, leading to diffi-

culties in the organization of motor sequences and alternating programs.
Visuoperceptual and visuospatial tasks were also affected, reflecting a
pattern of performance commonly observed after lesions of the frontal
lobe, such as segmentation and loss of figure-ground perspective (see
Figure 2). Verbal functions were better preserved, with slight reduction in
fluency, but no aphasic symptomatology was found. Immediate memory
was diminished, with marked difficulties in delay memory. Speech was
hypophonic, dysarthric, and aprosodic. Facial expression was hypo-
ADRENAL AUTOGRAFTING IN PARKlNSON'S DISEASE 209
Sample
Tests
FIGURE 2. Preoperative perfor-

~Q)" G.H. L C E R
mance on a block design task by
PD patients. Patients showed
segmentation and loss of figure-
ground perspective. Even in the
simplest design, they broke the 2 ~ ®
x 2 configuration and attended CD
only to the salient features of the
design. G.H. E.R MM. N M TC
mnnIc. No limb or bucofacial apraxia was observed. There was formal

perseveration of reading and writing, although motor problems and bra-
dykinesia affected the quality of the written product, there was no
dyscalculia.
Figure 3 compares the neuropsychological test-retest evaluation pro-
files obtained from the unoperated PD with the pre- and post-operative
evaluation profiles obtained from the operated PD groups. The test-retest
profiles of the unoperated PD control group overlapped, whereas the pro-
files of the operated PD group did not. A median difference analysis
(using the Mann-Whitney U test), which compared the scores obtained in
each section of the neuropsychological diagnostic scheme in the test-
retests condition by the unoperated control group and in the pre/postop-
erative condition by the operated PD group, revealed a significant (p <
.05) postoperative improvement in I (motor functions) and III (visuoper-
ceptual and visuospatial recognition).
Individual analysis revealed that preoperatively 7 (39%) of the pa-
tients showed a normal cognitive profile and 11 (61%) patients presented
specific cognitive deficits. In 7 of the 11 patients with abnormal pre-
operative profiles, the postoperative evaluation revealed a significant
amelioration of the frontal lobe type symptoms, and visuospatial deficits
(see Figure 4) as well as an improvement in memory tasks that require an
active organization of the response. Immediate and delay memory diffi-
culties remained unchanged. The improvements were unrelated to im-
proved increased alertness or sustained attention. Three patients with
abnormal preoperative profiles did not show a significant amelioration of
the frontal lobe signs. Patients with a normal preoperative neuropsycholo-
gical profile showed no postoperative deficiencies, and their profiles
proved to be very similar to their preoperative performance. Changes were
shown to be sustained in the second evaluation carried out at 12 months.
In two patients there were neurological complications as a result of
N
....
o
AGE (YEARS) 50.78± 7.77 47.44 ± 9.38
EDUCATION (YEARS) 8.22 ±5.65 9.83 ± 4.54
DURATION OF 9.67 ± 3.84 9.56 ± 3.90
ILLNESS (YEARS)
DURATION OF L-DOPA 7.44 ± 2.74 7. 25± 3.92
THERAPY (YEARS)
a 90 b ------------------90
80------------------ --------------------80
U) -. .~.
1,&.1 70~ C>. l'"
~ '<7-...... ~ ~A\\
~ 60 7_", (
-
~....<;",.-d-- 3--
I
~ 50----------------------
40~'~--~~--~~~--~~ , , 40
I II III IV V VI VII VIII IX II III IV V \,1 VII VIII IX
--TEST - PREOPERATIVE
•••••• RETEST •••••• POSTOPERATIVE 'Tl
trl
C"l
* P < 0.05 C"l
>-<:
FIGURE 3. Mean scores of the neuropsychological diagnostic scheme obtained by (a) the control unoperated PD oen
group (n = 10) in the test-retest condition and by (b) the operated PO group (n = 16) in the pre- and postopera-
tive evaluation. The dark lines show the mean and the limits of 2 standard deviations. Higher scores reflect ~
en
greater number of errors than average and lower scores reflect fewer numbers of errors. Level of significance of 7<:
>-<:
postoperative improvement is indicated, as well as mean values and standard deviation"s for the descriptive cJ:,
characteristics for the two groups. I, Motor functions; II, somatosensory knowledge; III, visuoperceptual and o
visuospatial recognition; IV, auditory knowledge and language; V, cognitive processes; VI, oral language; VII, C
en
reading; VIII, writing; IX, calculus. ~
e..
X
SAMPLE
m1EB]
mm
TESTS
V ISU AL
MEMORY
TASK
t \ '\ I
I7
PRE -OP
? I
<::
POST - OP
(91 days)
FIGURE 4. Pre- and postoperative (3 months) performance on an immediate visual memory

task by a 50-year-old writer with 10 years of evolution of PD. In the preoperative evaluation
he was able to copy designs but presented fragmentation and perseveration on visual memo-
ry tasks. Three months after surgery he shows integration and lack of perseveration. Before
surgery the patient was dependent on others for most of his daily living activities. He was
being treated with 750 mg of levodopa. Three months after surgery the patient is completely
independent in most chores, although it takes him twice as long and he is aware of his
difficulties. He has returned to work. He is on 500 mg of levodopa.
the surgery. One of the patients had permanent damage in the septum,
causing an accentuation of cognitive difficulties, and the patient's disori-
entation, inattention inertia, and lack of initiative increased. In another
case, after several postsurgical complications, the patient developed an
encephalopathy that severely affected all cognitive functions .
In the immediate postoperative period, six patients presented visual
and one auditory hallucinations that cleared spontaneously within 72
hours after the operation. Patients were aware of these hallucinations, in
general were not frightened by them, and could speak about them without
anxiety. The content of the hallucination varied in each patient. For ex-
ample, one patient with hallucinations saw insects on the wall and an-
other often heard the telephone ringing. Mental confusion was observed
in five cases and perseveration in motor and verbal tasks in six. This
behavior disappeared gradually within 2 to 4 weeks after surgery.
GENERAL CONSIDERATIONS
Neuropsychological evaluation has revealed that in a high percentage

of the PD patients with an abnormal preoperative neuropsychological
profile, autografting of adrenal medullary tissue to the caudate nucleus
has positive effects on specific cognitive symptoms.
The neuroanatomical basis of the cognitive alterations in PD is still a
matter of controversy. Some authors have suggested that cognitive
changes are secondary to a dysfunction of the basal ganglia (Albert, Fel-
dman, & Willis, 1974; Freeman & Albert, 1985; Mortimer et al., 1982).
Others believe that a cortical abnormality is primary and that the coexis-
tence of an Alzheimer-type dementia in PD is responsible for the mental
deterioration (Boller et al., 1980; Hakim & Mathieson, 1979; Lieberman et
aI., 1979). Several studies have reported that the frontal-type symptoms
appear early in the evolution of the disease and are observed in all pa-
tients who deteriorate progressively (Agid et al., 1986; Lees & Smith,
1983; Taylor et aI., 1986). Our results agree with this since they show that
the tasks sensitive to frontal deficits are precisely the ones that are most
affected and that show an important improvement after transplantation.
Thus, apparently in our patients, the frontal cortex is not damaged but
only hypoactive, owing to the dysfunction of afferent fibers arising from
the subcortical nuclei. Since immediate memory and retrieval processes
remained unchanged after the autograft, these disorders might be due to
the involvement of different cortico-subcortical structures, such as lesion
of the septo-hippocampal cholinergic systems. Selective destruction of
cholinergic neurons in the substantia innominata has been shown to in-
duce complex mnemonic and cognitive disturbances (Dubois et aI., 1983),
and administration of subthreshold doses of anticholinergics in nonde-
mented PD patients impairs performance in visual memory tasks (Dubois
et al., 1987). In our study, memory tasks that require an active organiza-
tion of the response, such as the recall of logic passages (W AIS Memory
Scale subtest), did show an improvement. This is probably related to a
frontal lobe involvement in the performance of these tasks (Luria, 1977).
Many of the cognitive and motor signs that improved after the auto-
graft seem to be dopaminergic-dependent. Two of the main dopaminergic
pathways are the nigrostriatal and the meso-limbic-cortical pathways
(Bjorklund & Lindwall, 1978; Farley et aI., 1978; Javoy-Agid et aI., 1984).
Current data seem to indicate that both dopaminergic pathways are in-
volved for adequate cognitive processing.
The selective degeneration of the dopaminergic nigrostriatal path-
way, as observed in patients intoxicated with MPTP, induced Parkin-
sonism with frontal-type cognitive impairment (Stern & Langston, 1985).
As previously reviewed, anatomically, neural connections between the
ADRENAL AUTO GRAFTING IN PARKINSON'S DISEASE 213
striatum and premotor and prefrontal cortex (complex loop) have been
described.
A variable involvement of the dopaminergic system could be the
basis of the heterogeneity in the motor signs and the cognitive profile
obtained during the preoperative evaluation, as well as in the differential
response to the autografting procedure.
Several investigators have reported a higher frequency of cognitive
impairment in patients with more bradykinesia and rigidity. Mortimer et
al. (1982) proposed two clinical forms of idiopathic PD-one with pre-
dominant bradykinesia and cognitive impairment, and the other with
predominant tremor and relatively intact function. Lieberman (1974) re-
ported less prominent tremor in PD patients with dementia, and Zetusky,
Jankovic, and Pirozzolo (1985) reported a significant association between
deterioration in mental status with bradykinesia, postural instability, and
gait difficulty, whereas tremor was associated with a relative preservation
of mental status and less functional impairment. Bernheimer, Birkmayer,
Hornykiewiez, Jellinger, and Seitelberger (1973) found that regional
chemical changes were correlated with the type and degree of clinical
manifestations. For example, severity of akinesia correlates best with
dopamine and HVA deficiency in the caudate nucleus, whereas the de-
gree of tremor paralleled severity of HVA deficiency in the pallidum, and
rigidity was not related to any specific regional distribution of dopamine
or HVA deficiency.
Within the caudate, different cognitive functions are subserved by
different regions. In PD, the maximal dopamine reduction is observed in
the anterodorsal head of the caudate, which is the area that has greatest
connection with the frontal cortex (Rosvold, 1972). In subhuman pri-
mates, experimental lesion in the anterodorsal and ventral part of the
head of the caudate causes difficulties in tasks that require response-
inhibition and produce perseveration, difficulty in shifting responses,
and deficits in behavioral programming (Cools, 1980; Johnston, Rosvold,
& Mishkin, 1968; Rosvold, 1972; Teuber & Proctor, 1964), and, in contrast,
lesions in the tail of the caudate produced deficits in visual discrimina-
tion (Divac, Rosvold, & Szwarcbart, 1967). It could be that in our study, in
the patients who show abnormal preoperative profiles, the disease af-
fected the caudate nucleus, and thus the dopamine concentration in this
structure is decreased. It is worth noting that the adrenal medullary frag-
ments are grafted (within the lateral ventricle with partial inclusion) with-
in the head of the caudate, and therefore the supply of dopamine to this
region could explain the positive cognitive effects after the autograft. The
possibility also exists that behavioral recovery is induced through the
regeneration of fibers and dopamine-containing cells of the meso-cortical
dopaminergic system.
214 FEGGY OSTROSKY -SOLIS et 01.
One of the most fundamental questions to date in the study of trans-

plants to the central nervous system is the identification of the basic
mechanisms that are related to the clinical improvement observed in
some of the patients affected by PD.
The transplant could exert functional effects through several mecha-
nisms: (1) The grafted tissue could release dopamine. It has recently been
demonstrated that the adrenal gland possesses a powerful dopamine-re-
leasing factor that seems to be glycoprotein, whose activity is capable of
inducing the release of high levels of dopamine from striatal tissue (Chang
& Ramirez, personal communication). (2) The transplant and the lesion
release specific neurotrophic factors that provoke the fragments of the
adrenal medulla to regenerate and reinnervate the striatum. (3) The trans-
planted tissue enhances the recovery of the host dopaminergic neurons;
that is, the improvement observed is due to the growth of the patient's
own nigrostriatal and/or meso-cortico-limbic system.
It is possible, then, that a multitude of trophic, neurochemical, and
synaptic mechanisms are involved in the recovery observed after trans-
plant. These mechanisms permit the transplanted tissue to promote the
functioning of the host brain and its recovery.
A large number of questions and perspectives have now arisen. Re-
search into the interaction that exists between grafts and the host nervous
system have suggested that the transplants may act as mediator to pro-
mote the production of growth factors or growth neuromechanisms lead-
ing to plasticity and recovery of damaged cells and pathways in the host,
and this opens possibilities for the treatment of illnesses that have been
incurable to date such as Huntington's corea or Alzheimer's disease.
A new, fascinating horizon full of questions and possibilities has
opened up for the field of clinical and basic neurosciences.
REFERENCES
Agid, Y., Ruberg, M., Dubois, B., Pillon, 8., Cusimano, G., Raisman, R, Cash, R, Lhermitte,
F., & Javoy-Agid, F. (1986). Parkinson's disease and dementia. Clinical Neurophar-
macology, 9, 522-536.
Agid, Y., Javoy-Agid, F., & Ruberg, M. (1987). Biochemistry of neurotransmitters in Parkin-
son's disease. In C. D. Marsden & S. Fhan (Eds.), Movement disorders (Vol. 2, pp. 166-
230). London: Butterworths.
Albert, M. 1., Feldman, R G., & Willis, A. L. (1974). The subcortical dementia of progressive
supranuclear palsy. Journal of Neurology, Neurosurgery and Psychiatry, 371, 121-130.
Backlund, E., Granberg, P., Hamberger, B., Knutson, E., Martenson, A., Sedvall, G., Seiger,
A., & Olson, 1. (1985). Transplantation of adrenal medullary tissue to the striatum in
Parkinsonism. Journal of Neurosurgery, 62, 169-173.
Barker, C. F., & Billingham, R E. (1977). Immunologically privileged sites. Advances in
Immunology, 25, 1-54.
Bernheimer, H., Birkmayer, W., Hornykiewicz, 0., Jellinger, K., & Seitelberger, F. (1973).
Brain dopamine and the syndromes of Parkinson and Huntington: Clinical, mor-
phological and neurochemical correlations. Journal of Neurological Science, 20, 415-
455.
Bjorklund, A., Dunnett, S. 8., Stenevi, U., Lewis, M. E., & Iversen, S. D. (1980). Reinnervation
of the denervated striatum by substantia nigra transplants: Functional consequences as
revealed by pharmacological and sensoriomotor testing. Brain Research, 199,307-333.
Bjorklund, A., & Lindwall, O. (1978). The meso-telencephalic dopamine system: A review of
its anatomy. In K. Livingston & O. Hornykiewicz (Eds.), Limbic mechanisms (pp. 307-
331). New York: Plenum Press.
Bohn, M. C., Marciano, F., Cupit, L., & Gash, D. M. (1987). Adrenal medulla grafts promote
recovery of striatal dopaminergic fibers in MPTP treated mice. Science, 247, 913-915.
Boller, F., Passafiume, D., Keefe, N. C., Rogers, K., Morrow, 1., & Kim, Y. (1980). Visuospatial
impairment in Parkinson's disease. Archives of Neurology, 41, 485-490.
Brozoski, T. J., Brown, R. M., Rosvold, H. E., & Goldman, P. S. (1979). Cognitive deficit
caused by regional depletion of dopamina in prefrontal cortex of rhesus monkey.
Science, 205, 929-932.
Carlsson, A., Lindquest, M., & Magnusson, T. (1957). 3-4-dihydroxyphenylalanine and 5-
hydroxytroptophan as reserpine antagonists. Nature, 180, 1200.
Cools, A. (1980). Role of the neostriatal dopaminergic activity in sequencing and selecting
behavioral strategies: Facilitation of processes involved in selecting the best strategy in a
stressful situation. Behavioral Brain Research, 1, 361-378.
Cotzias, G. C., Papavasiliov, P. S., & Gellene, R. (1969). Modification of Parkinson chronic
treatment with L-dopa. New England Journal of Medicine, 280, 337-345.
Darley, F., Brown, J., & Swenson, W. (1975). Language changes after neurosurgery for Parkin-
sonism. Brain and Language, 2, 65-69.
DeLong, M. (1974). Motor functions of the basal ganglia single unit activity during move-
ment. In F. D. Schmitt & F. G. Worden (Eds.), The neurosciences (pp. 319-325). Third
study program. Cambridge, MA: M.LT. Press.
DeLong, M. R., Georgopulos, A. P., & Crutcher, M. D. (1983). Corti co-basal ganglia relations
and coding of motor performance. In Neural coding of motor performance. Experimen-
tal brain research (pp. 30-40). Berlin: Springer.
Divac, L, Rosvold, H., & Szwarcbart, M. (1967). Behavioral effects of selective ablation of the
caudate nucleus. Journal of Comparative Physiology and Psychology, 63, 184-190.
Drucker-Colin, R., Madrazo, L, Ostrosky-Solis, F., Shkurovich, M., Franco, R., & Torres, C.
(1988). Adrenal medullary tissue transplants in the caudate nucleus of Parkinson pa-
tients. In J. Sladek & D. Gash (Eds.), Transplants in the central nervous system. Program
in brain research. Amsterdam: Elsevier.
Dubois, B., Danze, F., Pillon, B., Cusimano, J., Lhermitte, F., Agid, Y. (1987). Cholinergic
dependent cognitive deficits in Parkinson's disease. Annals of Neurology, 22, 26-30.
Dubois, B., Ruberg, M., JavoY-Agid, F., Pluska, A., & Agid, Y. (1983). A subcortical cho-
linergic system is affected in Parkinson's disease. Brain Research, 288, 213-218.
Dunnett, S. B., Bjorklund, A., & Stenevi, U. (1983). Dopamine-rich transplants in experimen-
tal parkinsonism. Trends in Neuroscience, 6, 266-270.
Dunnett, S. B., Bjorklund, A., Stenevi, U., & Iversen, S. D. (1981a). Behavioral recovery
following transplantation of substantia nigra in rats subjected to 6-0HDA lesions of the
nigrostriatal pathway. Brain Research, 215, 147-161.
Dunnett, S. B., Bjorklund, A., Stenevi, U., & Iversen, S. D. (1981b). Grafts of embryonic
substantia nigra reinnervating the ventrolateral striatum ameliorate sensoriomotor im-
pairments and akinesia in rats with 6-0HDA lesions of the nigrostriatal pathway. Brain
Research, 229, 209-207.
Ehringer, H., Hornykiewicz, O. (1960). Verteilung, Von. Noradrenalin und Dopamin (3-
Hydroxytyramin) im Gehirn des Menschen und ihr Verhalten bei Erkrankungen des
Extrapyramidalen Systems. Klinische Wochenschrift, 38, 1236-1239.
Elizan, T. S., Sroka, H., Maker, H., Smith, H., & Yahr, M. D. (1986). Dementia in idiophatic
Parkinson's disease. Journal of Neural Transmission, 65, 285-302.
Farley, L J., Price, K. S., & Hornykiewicz, O. (1978). Monominergic systems in the human
limbic brain. In K. Livingston & O. Hornykiewicz (Eds.)' Limbic mechanisms (pp. 333-
349). New York: Plenum Press.
Forno, L. S., Langston, J. W., Delanney, 1. E., Irwin, L, & Ricaurte, G. A. (1986). Locus
ceruleus lesions and eosinophilic inclusions in MPTP-treated monkeys. Annals of Neu-
rology, 20, 449-455.
Freed, W. J. (1983). Functional brain tissue transplantation: Reversal of lesion-induced rota-
tion by intraventricular substantia nigra and adrenal medulla grafts with a note on
intracranial retinal grafts. Biology and Psychiatry, 18, 1205-1267.
Freed, W. J., Morihisa, J. M., Spoor, E., Hoffer, B. J., Olson, L., Seiger, A, & Wyatt, R J. (1981).
Transplanted adrenal chromaffin cells in rat brain reduce lesion-indm:ed rotational
behavior. Nature, 292, 351-352.
Freeman, M., & Albert, M. 1. (1985). Subcortical dementia. In J. A. M. Frederiks (Ed.),
Handbook of clinical neurology. Vol. 46: Clinical neuropsychology (pp. 1049-1052).
Amsterdam: Elsevier.
Gash, D., Bohn, M., Jiao, S., Fiandaca, M., Okawara, S., Kordower, J., Hansen, J., Notter, M.,
Snyder, J., Marciano, F., Schwarz, H., & Shoulson, L (1987, October 18). Adrenal
medullary implantation promotes recovery of tyrosine hydroxylase immunoreactivity
in host striatum of MPTP animal models of parkinsonism. Paper presented at the Sym-
posium on Etiology, Pathogenesis and Prevention of Parkinson's Disease, San Francisco.
Gasar, P., & Gray, F. (1984). Dementia in idiopathic Parkinson's disease. Acta Neu-
ropathologica, 64, 43-54.
Hakim, A H., & Mathieson, G. (1979). Dementia in Parkinson's disease: A neuropathological
Javoy-Agid, F., & Agid, Y. (1980). Is the mesocortical dopaminergic system involved in
Parkinson disease? Neurology, 30, 1326-1330.
Javoy-Agid, F., Ruberg, M., Taquet, H., Bukobza, B., Agid, Y., & Gaspar, P. (1984). Bio-
chemical neuropathology of Parkinson's disease. Advances in Neurology, 40,189-198.
Johnston, T. N., Rosvold, H. E., & Mishkin, M. (1968). Proyections from behaviorally defined
sectors of the prefrontal cortex to the basal ganglia, septum and diencephalon of the
monkeys. Experimental Neurology, 21, 20-30.
Langston, J. W., Ballard, P., Tetrud, J. W., & Irwin, L (1983). Chronic parkinsonism in human
due to a product of meperidine-analog synthesis. Science, 219, 979-980.
Lees, A. J., & Smith, E. (1983). Cognitive deficits in the early stages of Parkinson's disease.
Brain, 106,257-270.
Lieberman, A. N. (1974). Parkinson's disease: A clinical review. American Journal of Medi-
cal Science, 267, 66-80.
Lieberman, A., Dziatolowski, M., Kupersmith, M., Serby, M., Goodgold, A., Korein, J., &
Goldstein, M. (1979). Dementia in Parkinson's disease. Annals of Neurology, 6, 355-
359.
Lindvall, 0., Backlund, E., Farde, L., Freedman, R, Hoffer, B., & Seiger, A. (1987). Transplan-
tation in Parkinson's disease: Two cases of adrenal medullary grafts to the putamen.
Annals of Neurology, 22, 457-468.
Luria, A. R. (1977). Las funciones corticales superiores en el hombre. La Habana: Orbe.
Madrazo, L, Drucker-Colin, R, Diaz, V., Martinez-Mata, J., Torres, C., & Becerril, J. (1987).
Open microsurgical autograft of adrenal medulla to the right caudate nucleus in two
patients with intractable Parkinson's disease. New England Journal of Medicine, 316,
831-834.
Matison, R, Mayeux, R, Rosen, J., & Fahn, S. (1982). Tip of the tongue phenomenon in
Parkinson's disease. Neurology, 32, 567-570.
Mortimer, J. A., Pirozzolo, F. J., Hansch, E. C., & Webster, D. D. (1982). Relationship of motor
symptoms to intellectual deficits in Parkinson's disease. Neurology, 32, 133-137.
Ostrosky-Solis, F., Canseco, E., Quintanar, L., Navarro, E., Meneses, S., & Ardila, A. (1985).
Sociocultural effects in neuropsychological assessment. International Journal of Neu-
roscience, 27, 53-66.
Ostrosky-Solis, F., Quintanar, L., Madrazo, I., Drucker-Colin, R, Franco-Bourland, R, &
Leon-Meza, V. (1988). Neuropsychological effects of brain autograft of adrenal medull-
ary tissue for the treatment of Parkinson's disease. Neurology, 38, 1442-1450.
Ostrosky-Solis, F., Quintanar, L., Meneses, S., Canseco, E., Navarro, E., & Ardila, A. (1986).
Actividad cognoscitiva y nivel sociocultural. Revista de Investigacion Clinica, 38, 37-
42.
Perlow, M. J. (1987). Brain grafting as a treatment for Parkinson's disease. Neurosurgery, 20,
335-342.
Perlow, M. J., Freed, W. J., Haffer, B. J., Seiger, A., Olson, L., & Wyatt, R J. (1979). Brain grafts
reduce motor abnormalities produced by destruction of nigrostriatal dopamine system.
Science, 204, 635-645.
Pirozzolo, F. J., Hansch, C., & Mortimer, J. A. (1982). Dementia in Parkinson's disease:
Neuropsychological analysis. Brain and Cognition, 1, 71-83.
Ramon y CajaI, S. (1938). Degeneration and regeneration of the nervous system. London:
Oxford University Press.
Riklan, M., & Levita, E. (1970). Psychological studies of thalamic lesions in humans. Journal
of Nervous and Mental Disease, 150, 251-265.
Rinne, U. K. (1982). Brain neurotransmitter receptors in Parkinson's disease. In C. D. Mar-
sden & S. Pahn (Eds.), Movement disorders (pp. 59-74). London: Butterworths.
Rosvold, H. E. (1972). The frontal lobe system: Cortical subcortical interrelationships. Acta
Neurobiologica Experimentalis, Warsaw, 32, 439-460.
Scatton, B., Rouquier, L., Javoy-Agid, F., & Agid, Y. (1982). Dopamine deficiency in the
cerebral cortex in Parkinson disease. Neurology, 32, 1039-1040.
Selby, G. (1967). Stereotactic surgery for the relief of Parkinson's disease. Part I. A critical
review. Journal of the Neurological Sciences, 5, 315-342.
Stern, Y., & Langston, J. W. (1985). Intellectual changes in patients with MPTP induced
Parkinsonism. Neurology, 35, 1506-1509.
Taylor, A. E., Saint-Cyr, J. A., & Lang, A. E. (1986). Frontal lobe dysfunction in Parkinson's
disease. The cortical focus of neostriatal outflow. Brain, 109, 845-883.
Taylor, A. E., Saint-Cyr, J. A., & Lang, A. E. (1987). Parkinson's disease: Cognitive changes in
relation to treatment response. Brain, 110,35-51.
Teuber, H., & Proctor, F. (1964). Some effects of basal ganglia lesions in subhuman primates
and man. Neuropsychologia, 2, 85-93.
UhI, G. R, Hedreen, J. C., & Price, D. L. (1985). Parkinson's disease: Loss of neurons from the
ventral tegmental area contralateral to therapeutic surgical lesions. Neurology, 35,
1215-1218.
Ungerstedt, U. (1971a). Adipsia and aphagia after 6-hydroxydopamine induced degeneration
of the nigro-striatal dopamine system. Acta Physiologica Scandinavica Supplementum,
367, 95-122.
Ungerstedt, U. (1971b). Striatal dopamine release after amphetamine or nerve degeneration
revealed by rotational behavior. Acta Physiologica Scandinavica Supplementum, 367,
49-68.
Unsicker, K., Rieffert, B., & Ziegler, W. (1980). Effects of cell culture condition, nerve growth
factor, dexamethasone and cyclic AMP on adrenal chromaffin cells in vitro. In O. Eran-
ko, S. Soinila, & H. Paivarento (Eds.), Histochemistry and cell biology of autonomic
neurons, SIF cells, and paraneurons (pp. 51-59). New York: Raven Press.
Villardita, T., Smirni, P., Le Pera, F., Zappala, G., & Nicoletti. F. (1982). Mental deterioration,
218 FEGGY OSTROSKY-SOLIS et a1.
vi suo perceptive disability and constructional apraxia in Parkinson's disease. Acta Neu-
ro1ogica Scandinavica, 66, 112-120.
Wurtman, R. J., Pohorecky, 1. A., & Baliga, B. S. (1972). Adrenocortical control of the
biosynthesis of epinephrine and protein in the adrenal medulla. Pharmacological Re-
view, 24, 411-426.
Zetusky, W., Jankovic, J., & Pirozzolo, F. (1985). The heterogeneity of Parkinson's disease:
Clinical and prognostic implications. Neurology, 35, 522-526.
11
The Striatum as a Temporary

Memory Store
ROBERTO A. PRADO-ALCALA
The search for the mechanisms involved in memory storage has yielded
an enormous amount of experimental data. To date, however, we are far
from understanding how the nervous system integrates the relevant infor-
mation derived from experience, and how that information is channeled
to effector systems when the same, or a similar, experience occurs.
Recent work has provided evidence strongly suggesting that memory
storage depends not upon the workings of a single neuroanatomicallneu-
rochemical system but, rather, upon sequential activation of different sys-
tems (for a review, see Prado-Alcala, 1985). What follows is a brief ac-
count of the experimental data that have given support to this hypothesis.
INVOLVEMENT OF THE NEOSTRIATUM IN MEMORY
A wealth of information concerning the involvement of the caudate

nucleus (eN) or neostriatum in memory processes has accumulated.
Strong support to the hypothesis that the caudate is critically involved in
associative processes (Divac & Oberg, 1979) has been given by the demon-
stration that almost every type of conditioned behavior that has been
studied is impaired after disrupting the functional integrity of this struc-
ture, with the use of a variety of lesioning techniques (Dunnet & Iversen,
1985; Glick & Greenstein, 1973; Glick, Marsanico, & Greenstein, 1974;
Kirkby & Kimble, 1968; Mitcham & Thomas, 1972; Prado-Alcala et al.,
1975; Sanberg, Lehmann, & Fibiger, 1978; Sanberg, Pisa, & Fibiger, 1979;
ROBERTO A. PRADO-ALCALA· Department of Physiology, Faculty of Medicine, National

University of Mexico, Mexico D. F. 4510, Mexico.
219
220 ROBERTO A. PRADO-ALCALA
Winocur, 1974), as well as other methods that interfere with its neuronal
activity, such as electric stimulation and topical application of potassium
chloride (Le Piane & Phillips, 1978; Prado-Alcala & Cobos-Zapiain, 1979;
Prado-Alcala, Grinberg, Alvarez-Leefmans, & Brust-Cammona, 1973;
Prado-Alcala et a1., 1975; Prado-Alcala, Kaufmann, & Moscona, 1980;
Wyers & Deadwyler, 1971; Wyers, Deadwyler, Hirasuna, & Montgomery,
1973; Wyers, Peeke, Elliston, & Herz, 1968).
Although those studies indicated that normal functioning of the CN is
important for the establishment of memory, they did not provide clues
about the nature of the mechanisms implicated in such function, but were
instrumental in the designing of experiments aimed at defining the neu-
rochemical events that may occur during learning and performance of
conditioned behaviors. Thus, it was shown that alterations in the activity
of striatal dopamine (Kim & Routtenberg, 1976a; Phillips & Clouston,
1978; Stabuli & Huston, 1978), GABA (Salado-Castillo & Prado-Alcala,
1987), and acetylcholine (Haycock, Deadwyler, Sideroff, & McGaugh,
1973; Neill & Grossman, 1970; Prado-Alcala et a1., 1972) also produce
significant memory impairments.
Since dopamine, GABA, and acetylcholine are the main neurotrans-
mitters regulating the activity of the nigro-neostriatal system, the data
suggested that the nigro-neostriatal system is critically involved in memo-
ry functions. Several lines of evidence give support to this idea: Blockade
of GABAergic activity of the substantia nigra (Cobos-Zapiain & Prado-
Alcala, 1986; Kim & Routtenberg, 1976b) and combined treatment of a
dopaminergic blocker with intrastriatal atropine (Rivas-Arancibia &
Prado-Alcala, 1986) result in amnesia.
THE OVERTRAINING EFFECT
The effects of injections of ACh-receptor blockers into the anterodor-

sal aspect of the CN have been tested on the acquisition and maintenance
of positively rewarded bar-pressing and alley-running tasks (Bermudez-
Rattoni, Mujica-Gonzalez, & Prado-Alcala, 1986; Prado-Alcala & Cobos-
Zapiain, 1977; Prado-Alcala et a1., 1972; Prado-Alcala, Kaufmann &
Moscona, 1980). In all cases a strong amnesic state was induced. These
deficits were seen using two different muscarinic blockers (atropine and
scopolamine) and two animal species (cats and rats). When the same
treatments that induced amnesia were applied to other cerebral regions,
such as the lateral ventricles (Prado-Alcala et a1., 1972), amygdala (Prado-
Alcala & Cobos-Zapiain, 1977), and parietal cortex (Bermudez-Rattoni et
a1., 1986), no significant deficits were found.
The involvement of the CN in passive avoidance has been extensively
studied (Polgar, Sanberg, & Kirkby, 1981) and, with few exceptions
THE STRIA TUM AS A MEMORY STORE 221
(Olmstead & Villablanca, 1980), it is accepted that this behavior is depen-

dent upon the functional integrity of this structure. Likewise, it has been
consistently found that the application of atropine or scopolamine into
the anterodorsal striatum shortly after training of passive avoidance pro-
duces a marked amnesic state when retention is tested 24 hours later
(Giordano & Prado-Alcala, 1986; Haycock et 01.,1973; Prado-Alcala, Cruz-
Morales, & Lopez-Miro, 1980; Prado-Alcala, Fernandez-Samblancat, &
Solodkin-Herrera, 1985; Prado-Alcala, Signoret, & Figueroa, 1981; Prado-
Alcala, Signoret-Edward, Figueroa, & Barrientos, 1984).
In contrast, blockade of cholinergic activity of the posterior CN
(Prado-Alcala, Cruz-Morales, & L6pez-Miro, 1980) or hippocampus (Hay-
cock et 01., 1973) does not produce such effect. The application of anti-
cholinergic drugs to the cerebral cortex produces a retention deficit that is
significantly smaller than that produced by the same treatments applied
to the CN (Prado-Alcala et 01., 1985).
The amnesic effect produced by atropine injections into the CN is
both dose- and time-dependent. In two related studies it was found that,
as expected, when the dose of the anticholinergic is larger, so is the
retention· deficit (Giordano & Prado-Alcala, 1986; Prado-Alcala et 01.,
1985). In another experiment it was shown that when posttrial injections
of atropine into the CN were closer in time to the time of training (2 min),
a greater amnesic state was produced; no retention deficits were observed
when the injections were made 15 or 30 minutes after training (Prado-
Alcala et 01., 1981).
An unexpected finding, which seemed to contradict the reported det-
rimental effects of cholinergic blockade of the caudate nucleus on the
maintenance of positively motivated behaviors, gave origin to a new hy-
pothesis about the way in the caudate nucleus is involved in memory
processes (Prado-Alcala & Cobos-Zapiain, 1977). The injections of atro-
pine into the CN did not alter the performance of cats that had been
trained to press a bar in order to obtain milk. A previous study showed
that this treatment produced an amnesic state in the same species trained
in the same task (Prado-Alcala et 01., 1972). The only procedural dif-
ference was that the animals of the former study had been overtrained;
i.e., they had been trained for 30 sessions instead of the usual 10 to 15
sessions.
Subsequent experiments have confirmed the protective effect of over-
training. When independent groups of rats are trained to bar-press during
5, 15, or 25 sessions and are then injected with scopolamine into the
striatum, a significant deficit in retention is found in the 5- and 15-session
groups, but no deterioration of the learned response in the 25-session
group (Prado-Alcala, Kaufmann, & Moscona, 1980). Equivalent results
were found when a more complex task (spatial alternation) was studied
(Prado-Alcala, Bermudez-Rattoni, Velazquez-Martinez, & Bacha, 1978).
In related studies it has been shown that when rats and cats are
trained to bar-press until they reach asymptotic performance and then
injected into the CN with a high concentration of potassium chloride, they
show a significant retention deficit. Again, when training is extended, the
same treatment does not produce retention deficits (Prado-Alcala &
Cobos-Zapiain, 1979; Prado-Alcala, Kaufmann, & Moscona, 1980).
The results reported above strongly indicated that the CN is critically
involved in the acquisition and early maintenance stages of positively
reinforced instrumental conditioning, and that this structure is not en-
gaged in memory after overtraining. It was reasoned that if this represents
a general way of functioning of the striatum, equivalent effects should be
produced when animals are trained in other types of tasks.
As stated above, the striatum plays a very important role in the con-
solidation of memory of one-trial passive avoidance, a negatively rein-
forced task. This task was used to determine whether the overtraining
effect could be generalized further.
Overtraining, as studied in prior experiments, involved multiple
training sessions, a high number of positive reinforcers, and a prolonged
exposure to the experimental situation. In the case of one-trial passive
avoidance there is only one training session and application of only one
reinforcer, and the duration of the trial is brief. For the designing of the
experiments to be described below, it was decided to vary the magnitude
of only one of these parameters (the reinforcer), testing different inten-
sities of the foot shock. This manipulation is equivalent to having differ-
ent amounts of positive reinforcers.
In one experiment, the effects of posttraining injection of atropine
into the anterodorsal aspect of the striatum on retention of passive avoid-
ance were assessed. Independent groups of rats were given different levels
of foot shock (0.25, 0.50, or 1.00 mAl during training. As expected, retro-
grade amnesia was seen only in the 0.25-mA group (Giordano & Prado-
Alcala, 1986).
In order to determine whether some striatal neurochemical system,
other than cholinergic, became engaged in consolidating memory after
overtraining, groups of rats were trained, using also low, medium, and
high foot shock intensities, and then given intra striatal injections of Xilo-
caine, thus arresting all normal neural actiyity of the injected area. It was
reasoned that if consolidation was dependent upon the integrative func-
tions of the striatum, an amnesic state would be produced, regardless of
the intensity of the foot shock-Le., regardless of overtraining. In agree-
ment with the studies reported above, those rats that were submitted to
overtraining had as good retention scores as intact animals (Perez-Ruiz &
Prado-Alcala, 1986).
With regard to active avoidance, there are only a few relevant studies
that could bear on this problem; however, these studies were not designed
to study the overtraining effect, and interpretation of these results should
THE STRIATUM AS A MEMORY STORE 223
be made cautiously. Briefly, when acetylcholine receptor-blockers are ap-

plied to the anterior aspect of the striatum of animals that are still learning
the task (Neill & Grossman, 1970), have just learned it (Prado-Alcala,
Cepeda, Verduzco, Jimenez, & Vargas-Ortega, 1984), or are "overtrained"
(Prado-Alcala, Cruz-Morales, & Lopez-Miro, 1980), a performance deficit
is observed only in the former group.
In a recent study (Cobos-Zapiain & Prado-Alcala, 1986), it was shown
that injections of picrotoxin into the substantia nigra produced a pro-
found amnesic state of passive avoidance; again, when the same treatment
was given to overtrained rats, no deficits in learning were observed.
CONCLUSIONS
From the studies reviewed here it can be concluded that cholinergic

activity of the caudate nucleus plays a major role in the acquisition and
early maintenance stages of instrumental learning. The experimental evi-
dence also points to the conclusion that striatal cholinergic activity is not
involved in the performance of tasks that have become overtrained. The
results obtained after the application of picrotoxin into the nigra of over-
trained animals also point to the same conclusion. Taken together, the
studies here reviewed suggest that activity of the acetylcholine-containing
interneurons of the caudate nucleus and the striatonigral GABA neurons
are critically involved in the processes underlying recent memory, while
long-term storage of information is mediated by a different neurochemical
system, which very probably is located outside the nigro-neostriatal
system.
Along these lines, Miller (1981) has recently proposed a two-stage
storage model of memory. According to this model, the caudate nucleus
would constitute the first memory store for operant tasks, which then
would be transferred to the cerebral cortex.
To date there has not been a single published study demonstrating
unequivocally that the activity of a particular brain structure in the mam-
mal is necessary for the maintenance (long-term memory) of instrumental
behaviors. It seems that the failure to find such structure, or set of struc-
tures, simply reflects a misleading theoretical approach to this problem-
i.e., trying to find an all-purpose single memory store (supposedly located
in one structure or a system of structures). It is suggested that the search
for the "engram" should be guided by a search for neuroanatomical and
neurochemical systems that may be sequentially involved in such com-
plex function. Very probably we will find multiple systems, each subserv-
ing different types of learned behaviors, which become operative, one
after the other, depending on the relative age and degree of mastery of the
learned response.
ACKNOWLEDGMENTS
This work was supported by Fundacion Miguel Aleman, A. C. and

Fondo Ricardo J. Zevada.
REFERENCES
Bermudez-Rattoni, F., Mujica-Gonzalez, M., & Prado-Alcala, R. A. (1986). Is cholinergic

activity of the striatum involved in the acquisition of positively-motivated behaviors?
Pharmacology, Biochemistry and Behavior, 24, 715-719.
Cobos Zapiain, G., & Prado-Alcala, R. A. (1986). Aplicaci6n de picrotoxina en la substancia
nigra reticulada: Etectos sobre la memoria de largo plazo, en una tarea sobreentrenada,
XXIX Congreso Nacional de Ciencias Fisiol6gicas, Mexico.
Divac, I., & Oberg, R. G. E. (1979). The neostriatum. Oxford: Pergamon Press.
Dunnett, S. B., & Iversen, S. D. (1981). Learning impairments following selective kainic acid-
induced lesions within the neostriatum of rats. Behavioral Brain Research, 2, 189-209.
Giordano, M., & Prado-Alcala, R. A. (1986). Retrograde amnesia induced by post-trial injec-
tion of atropine into the caudate-putamen. Protective effect of the negative reinforcer.
Glick, S. D., & Greenstein, S. (1973). Comparative learnin'g and memory d-eficits following
hippocampal and caudate lesions in mice. Journal of Comparative and Physiological
Psychology, 82, 188-194.
Glick, S. D., Marsanico, R. G., & Greenstein, S. (1974). Differential recovery of function
following caudate, hippocampal, and septal lesions in mice. Journal of Comparative
and Physiological Psychology, 86, 787-792.
Haycock, J. W., Deadwyler, S. A., Sideroff, S. I., & McGaugh, J. L. (1973). Retrograde amnesia
and cholinergic systems in the caudate-putamen complex and dorsal hippocampus of
the rat. Experimental Neurology, 41, 201-213.
Kim, H.-J., & Routtenberg, A. (1976a). Retention deficits following post-trial dopamine injec-
tion in rat neostriatum. Society tor Neuroscience Abstracts, 2, 631.
Kim, H.-J., & Routtenberg, A. (1976b). Retention disruption following post-trial picrotoxin
injection into the substantia nigra. Brain Research, 113,620-625.
Kirkby, R. J., & Kimble, D. P. (1968). Avoidance and escape behavior following striatal
lesions in the rat. Experimental Neurology, 20, 215-227.
Le Piane, F., & Phillips, A. G. (1978). Differential effects of electrical stimulation of amyg-
dala, caudate-putamen or substantia nigra pars compacta on taste aversion and passive
avoidance in rats. Physiology and Behavior, 21, 979-985.
Miller, R. (1981). Meaning and purpose in the intact brain. Oxford: Oxford University Press.
Mitcham, J. C., & Thomas, R. K. (1972). Effects of substantia nigra and caudate nucleus
lesions on avoidance learning in rats. Journal of Comparative and Physiological Psy-
chology, 81, 101-107.
Neill, D. B., & Grossman, P. S. (1970). Behavioral effects of lesions or cholinergic blockade of
dorsal and ventral caudate of rats. Journal of Comparative and Physiological Psychol-
ogy, 71, 311-317.
Olmstead, C. E., & Villablanca, J. R. (1980). Effects of caudate or frontal cortex ablations in
cats and kittens: Passive avoidance. Experimental Neurology, 68, 335-345.
Perez-Ruiz, C., & Prado-Alcala, R. A. (1986). Differential effects of lidocaine injections into
the striatum on short- and long-term retention of passive avoidance in overtrained rats.
Society tor Neuroscience Abstracts, 12, 714.
THE STRIATUM AS A MEMORY STORE 225
Phillips, A. G., & Clouston, R (1978). Disruption of one-trial appetitive learning and passive
avoidance following stimulation of the substantia nigra pars compacta. Behavioral Biol-
ogy, 23, 388-394.
Polgar, S., Sanberg, P. R, & Kirkby, R J. (1981). Is the striatum involved in passive avoidance
behavior? A commentary. Physiological Psychology, 9, 354-358.
Prado-Alcala, R A. (1985). Is cholinergic activity of the caudate nucleus involved in memo-
ry? Life Sciences, 37, 2135-2142.
Prado-Alcala, R A., Bermudez-RaUoni, F., Velazquez-Martinez, D., & Bacha, M. G. (1978).
Cholinergic blockade of the caudate nucleus and spatial alternation performance in rats:
Overtraining-induced protection against behavioral deficits. Life Sciences, 23, 889-896.
Prado-Alcala, R A., Cepeda- G., Verduzco, L., Jimenez, A., & Vargas-Ortega, E. (1984). Effects
of cholinergic stimulation of the caudate nucleus on active avoidance. Neuroscience
Letters, 51, 31-36.
Prado-Alcala, R A., & Cobos-Zapiain, G. G. (1977). Learning deficits induced by cholinergic
blockade of the caudate nudeus as a function of experience. Brain Research, 138, 190-
196.
Prado-Alcala, R A., & Cobos Zapiain, G. G. (1979). Interference with caudate nucleus ac-
tivity by potassium chloride. Evidence for a "moving" engram. Brain Research, 172,
577-583.
Prado-Alcala, R A., Cruz-Morales, S. E., & L6pez-Miro, F. A. (1980). Differential effects of
cholinergic blockade of anterior and posterior caudate nucleus on avoidance behaviors.
Neuroscience Letters 18, 339-345.
Prado-Alcala, R A., Fernandez-Samblancat, M., & Solodkin-Herrera, M. (1985). Injections of
atropine into the caudate nucleus impair the acquisition and the maintenance of passive
avoidance. Pharmacology, Biochemistry and Behavior, 22, 243-247.
Prado-Alcala, R A., Grinberg, Z. J., Alvarez-Leefmans, F. J., & Brust-Carmona, H. (1973).
Suppression of motor conditioning by the injection of 3M KCl in the caudate nuclei of
cats. Physiology and Behavior, 10, 59-64.
Prado-Alcala, R A., Grinberg, Z. J., Alvarez-Leefmans, F. J., G6mez, A., Singer, S., & Brust-
Carmona, H. (1972). A possible caudate-cholinergic mechanism in two instrumental
conditioned responses. Psychopharmacologia (Berlin), 25, 339-346.
Prado-Alcala, R A., Grinberg, Z. J., Arditii, Z. L., Garcia, M. M., Prieto, H. G., & Brust-
Carmona, H. (1975). Learning deficits produced by chronic and reversible lesions of the
corpus striatum in rats. Physiology and Behavior, 15, 283-287.
Prado-Alcala, R A., Kaufmann, P., & Moscona, R. (1980). Scopolamine and KCl injections
into the caudate-putamen. Overtraining-induced protection against deficits of learning.
Prado-Alcala, R A., Signoret, L., & Figueroa, M. (1981). Time-dependent retention deficits
induced by post-training injections of atropine into the caudate nucleus. Pharmacology,
Biochemistry and Behavior, 15, 633-636.
Prado-Alcala, R A., Signoret-Edward, L., Figueroa, M., & Barrientos, M. A. (1984). Post-trial
injection of atropine into the caudate nucleus interferes with long-term, but not with
short-term retention of passive avoidance. Behavioral and Neural Biology, 42, 81-84.
Rivas-Arancibia, S., & Prado-Alcala, R A. (1986). Nucleo caudado y aprendizaje. XXN.
Interacci6n entre los sistemas dopaminergico y colinergico en procesos de memoria.
XxIX Congreso Nacional de Ciencias Fisiol6gicas, Mexico.
Salado-Castillo, R., & Prado-Alcala, R A. (1987). Effects of picrotoxin injections into differ-
ent regions of the striatum on retention of passive avoidance. Society for Neuroscience
Abstracts, 13, 657.
Sanberg, P. R, Lehmann, ]., & Fibiger, H. C. (1978). Impaired learning and memory after
kainic acid lesions of the striatum: A behavioral model of Huntington's disease. Brain
Research, 149, 546-551.
Sanberg, P. R., Pisa, M., & Fibiger, H. C. (1979). Avoidance, operant and locomotor behavior
in rats with neostriatal injections of kainic acid. Pharmacology, Biochemistry and Be-
havior, 10, 137-144.
Stabuli, U., & Huston, J. P. (1978). Effects of post-trial reinforcing vs. subreinforcing stimula-
tion of the substantia nigra on passive avoidance learning. Brain Research Bulletin, 3,
519-524.
Winocur, G. J. (1974). Functional dissociation within the caudate nucleus of rats. Journal ot
Comparative and Physiological Psychology, 86, 432-439.
Wyers, E. J., & Deadwyler, S. A. (1971). Duration and nature of retrograde amnesia produced
by stimulation of caudate nucleus. Physiology and Behavior, 6, 97-103.
Wyers, E. J., Deadwyler, S. A., Hirasuna, N., & Montgomery, D. (1973). Passive avoidance
retention and caudate stimulation. Physiology and Behavior, 11,809-819.
Wyers, E., J., Peeke, H. V. S., Elliston, J. S., & Herz, M. J. (1968). Retroactive impairment of
passive avoidance learning by stimulation of the caudate nucleus. Experimental Neu-
rology, 22, 350-366.
v
Recovery from Brain Damage

12
Development and Plasticity in the

Central Nervous System
Organismic and Environmental Influences
DONALD G. STEIN
Throughout this chapter I will be talking about plasticity in the central

nervous system. Although the concept is central to both biological and
psychological development, the term plasticity has been used in so many
different ways that it has lost much of its meaning. For example, almost
any change in behavior or in an organism's response to stimuli can be
taken as an example of plasticity. Here, I will try to define the term in a
way that makes sense for people in the neurosciences as well as in psy-
chology, then I will give some specific examples of the kinds of plasticity
that can be seen in both the developing and mature nervous system. Over
the last decade, despite the problem of how to define the term adequately,
there has been a major revolution in ideas about how the nervous system
functions, how it develops, and what actually represents "plasticity."
Kaplan (1983) has provided a broad definition of plasticity in devel-
opment that fits very well with events taking place in the central nervous
system. In general terms Kaplan describes plasticity as "an ability to
modify organic systems and patterns of behavior. In this context different
means can be used to obtain specific goals. Plasticity will also refer to the
wide individual differences in response to external or internal environ-
mental demands." Kaplan also characterizes development as "a special
wayof regarding, analyzing, describing, and ordering phenomena of sys-
temic change." "Developmental analysis is oriented principally to the
relations between forms and function, parts and wholes, means and ends.
DONALD G. STEIN· Dean of the Graduate School and Associate Provost for Research.
Rutgers University, Newark, New Jersey 07102.
229
230 DONALD C. STEIN
Development is also defined in terms of differentiation and hierarchic

integration. "
Kaplan stresses that plasticity cannot be defined without reference to
the means that are used to achieve certain ends. Thus, "a developmental
analysis of plastic phenomena will also require careful study of means-
ends relationships and how such means change in a given person, or
organism, or group of subjects." In this context plasticity in development
is seen as a form of creativity or flexibility.
In another definition, GoUin (1981) refers to plasticity as "the possi-
ble range of variations that can occur in individual development or to
systematic structural or functional changes in a process and may involve
variations that lie on a continuum of variation around some hypothesized
average value" (p. 231). For Gollin, plasticity is the potential for change,
the capacity to modify one's behavior or function and to adapt to the
demands of a particular context. Again, the role of context or environment
will be given considerable attention in my discussion about recovery from
brain injuries.
From my perspective, GoUin's key phrase is adaptation to a "particu-
lar context" because what might be seen as a constancy-for example, a
specific syndrome or deficit following damage to a specific CNS region-
might only be relatively constant-that is, the development of an out-
come occurring in one very specific context but not in another.
In one sense, the topic of this essay is "What is reality?" In terms of
prevailing scientific paradigms, we will be talking about different "real-
ities," different ways of looking at central nervous system function, asking
what is "constant" and what is contextual. I will argue, for example, that
much of what we know about cerebral localization is based on the evalua-
tion of brain-damaged patients and animal subjects who are studied in
very limited and circumscribed situations that are specifically designed
to yield a desired outcome.
To highlight this point, I would like to quote from the introductory
chapter in what is thought to be the most widely read textbook in the
neurosciences (Kandel & Schwartz, 1981). In the section titled "Cognitive
functions can be localized within the cerebral cortex," the authors state
that "much of what we know about the localization of normal language
has come from the study of aphasia" (p. 7). In the same chapter, Kandel
and Schwartz also argue that "affective and character traits also are ana-
tomically localizable" (p. 10). They also state that "clinical studies and
their counterparts in experimental animals suggest that all behavior
(emphasis added), including higher (cognitive as well as affective) mental
functions, is localizable to specific regions or constellations of regions
within the brain. The role of descriptive neuroanatomy, therefore, is to
provide us with a functional guide to localization within the three-dimen-
sional neural space-a map for behavior" (p. 11) (emphasis added).
DEVELOPMENT AND PLASTICITY IN THE CNS 231
Given this set of beliefs and attitudes, it makes sense that, in many
laboratories, experimental lesions are created, employed, and controlled
in such a way as to guarantee that an observable deficit will occur. In fact,
this would be the most reasonable approach for those concerned with the
mapping of behavioral functions onto increasingly discrete brain regions,
zones, or even specific "command or control" neurons. But what if ex-
perimental conditions can be manipulated in such a way that the same
extent of injury, created in a different way, leads to behavioral sparing or
marked reduction in the impairment? Thus, if such outcomes were possi-
ble, the "constancy" of the lesion effect would be challenged and a new
set of "contextual" factors would have to be considered in determining
the development of events that will follow the injury.
Workers with a heavily vested interest in cerebral mapping and neu-
ronal specificity are not happy with such a conceptual (or contextual)
approach to the study of cerebral organization, and one can readily under-
stand why. If, as we shall try to show, the cerebral circuitry underlying
behavioral events can shift as a function of environmental, hormonal, and
other organismic events, map making becomes very difficult at best. Pic-
ture what it might be like to have a state map where the locations of cities
and towns (and the highways to and from them) would shift from day to
day as a function of economic conditions, the weather, population shifts,
famine, and other factors. Yet, in geography, such realities do occur, albeit
perhaps more slowly. Nonetheless, following a catastrophe (earthquake,
war, flooding), rapid and massive change could occur as new "centers"
and paths develop to replace those that were lost. Can such contextually
induced plasticity occur at the level of the individual central nervous
system following the equivalent catastrophe of stroke, trauma, or pen-
etrating injury? Is the hard-wired nature of the nervous system an inher-
ent characteristic of the brain itself, or is it more a characteristic of the
particular beliefs and attitudes we have developed to study it? We will
need to return to this point many times as we proceed with this essay.
When we talk about brain injury or plasticity, we must keep in mind
that these events (or processes) do not occur in a static system. Injury and
subsequent change can be seen as a dynamic series of events in which
many different variables playa role in determining the outcome of injury.
Table 1, taken from a recent paper by Geschwind (1985), shows a number
of specific and nonspecific changes that occur after lesions of the nervous
system. Anyone of these parameters could be the subject of a lifetime of
research. It is easy to understand why, in the face of so many variables,
one would be tempted to focus upon one or two factors and ignore the
others.
Most research in neuroscience takes the approach that given parts of
the central nervous system have highly specific structure-function rela-
tionships. In this view, each neuron, each organ of the brain is presumed
232 DONALD G. STEIN
TABLE 1. Changes after Lesions of the Nervous System
Mechanisms (listings are not necessarily exclusive)

Biochemical
Nonspecific
Edema and disappearance of edema
Ischemia without destruction
Immune responses
Scarring
Pressure effects
Specific
Increase or decrease in number of receptors
Induction or inactivation of enzymes
Viral alterations of specific cell function
Immune responses to specific neural components
Alterations in transmitter release
Transmitter reuptake by inactive neurons (Zigmond effect)
Alterations in membrane properties
Structural
Disrupted organization
Sprouting-colateral and terminal
Anomalous connections
Demyelination
Transynaptic and trans neuronal degeneration
Diminished cell death in other areas (in fetal and possible early life)
Physiological
Changes in facilitation, inhibition, and occlusion
Loss of feedback effects
Pacemaker shifts
Slowed conduction
Changes in synchrony
Opening of ineffective synapses
Changes of spatial or temporal summation
Kindling and epilepsy
"Psychological"
Perception of change by intact regions
Alternate strategies
Factors influencing late change
Age
Species
Previous learning
Individual differences (type and size of structures, connections, bio-
chemistry, individual talents, handedness)
Previous exposure to drugs, hormones, etc. (pre- and postnatal)
Serial lesions
Epilepsy
Stimulation (such as skin lesions or deliberate stimulation)
Emotional effects-motivation, hormones, circulating transmitters, etc.
Susceptibility to immune attack of areas formed abnormally in prenatal
life
to have a genetically determined, specific role to play. It is not uncommon

to see this mechanistic view translated to the molecular level in the neu-
rosciences in talk about command neurons, analyzers, perceptors, and so
forth. In the context of such a "hard-wired" nervous system, there is
apparently little weight given to the environmental and organismic fac-
tors that can modify and shape eNS function in temporary as well as
permanent fashion.
Here is a good example of "contextual plasticity" from our own labo-
ratory (Attella, Nattinville, & Stein, 1988) examining the question of
whether hormonal factors determine the outcome of brain injury. It is well
documented that females have fluctuating levels of estrogen or pro-
gesterone in the central nervous system depending upon the particular
phase of their estrous cycle. When estrogen is high, progesterone is low.
When progesterone is high, estrogen is low. What possible role could
these alterations in ovarian hormones play in determining whether cog-
nitive impairment would occur following injury to the frontal cortex? In
the traditional view of static structure-function relationships in the eNS,
one would hardly expect ovarian hormones to affect the outcome of brain
injury- yet this is exactly what we were able to observe.
Studies show that when fully intact adult female rats were examined
on a delayed spatial alternation task, alterations in hormone levels had no
effect on learning ability. On errors and perseverations in the T-maze,
whether an intact animal was pseudopregnant (low estrogen, high pro-
gesterone) or was cycling did not change cognitive performance. In rats, at
least, "daily learning skills" remained intact despite significant shifts in
hormonal cycling.
Further study has demonstrated what happens to normal cycling or
pseudopregnant rats after they have been subjected to bilateral removal of
the frontal cortex. In the standard brain-injury paradigm, this lesion al-
most always produces severe impairments of spatial learning ability in
rats, cats, and monkeys. The sham-operated animals (no cortical damage),
pseudopregnant or normal cycling, showed completely normal retention
of the learned alternation; there are few, if any, errors during this phase of
the training. There exist some interesting data showing what happens to
rats that were brain-injured during normal cycling and then compared
with pseudo-pregnant conspecifics. These data provide a dramatic dem-
onstration that normally cycling animals with relatively high eNS levels
of estrogen at the time of brain injury will have severely impaired reten-
tion performance in comparison to rats who were first made pseudopreg-
nant at the time of injury.
This finding provides clear evidence that the outcome of brain injury
is very much dependent on the female's hormonal state, not just on the
site of the injury itself. The performance deficits we observed were dra-
matic and persistent.
234 DONALD G. STEIN
Hormonal factors may also be important in regulating the extent of

injury-induced, synaptic sprouting in the brain. Recently, Scheff, Hoff,
and Anderson (1986) used quantitative electron microscopy to measure
synapse formation in adult rats given partial lesions of the hippocampus
followed by treatment with corticosterone. These workers found that if
the rats were maintained on high levels of corticosterones (often em-
ployed as an anti-inflammatory agent), there was a decrease both in the
rate of removal of degenerating synapses and in the rate of synapse re-
placement. If regeneration of this type were involved in the mediation of
functional recovery, then one might expect a negative prognosis if the
subject happened to have high steroid levels at the time of injury.
If nothing else, the results of these two experiments reveal that there
is a very complex interaction between organismic (i.e., hormonal) state
and outcome of brain injury. It is not just a question of damaging a given
"circuit" and expecting to observe a fixed outcome at a given time. The
context in which the injury has occurred has rather dramatically altered
the behavioral sequelae. How, then, do we draw the detailed "map of
behavior" referred to earlier?
PLASTICITY AND LEVELS OF ANALYSIS OF CNS FUNCTIONS
One important and difficult problem for many of us working in con-

temporary neuroscience is that often there is a confusion about the "most
appropriate" level of analysis required to define function in the central
nervous system. With respect to "plasticity," we can expect that different
levels of analysis and different ways of measuring "important" events
will lead to a different construct about what is "really important" to
understand the eNS. Each area or subdiscipline of the neurosciences
looks at these different outcomes and then attempts to derive its own
working definition of plasticity. For example, a neurochemist might look
at "plastic" phenomena in terms of alterations in the flux of ions at the
membrane level of an injured neuron. Here, at the site of the injury or at
the synaptic end-foot, stable or changing rates of calcium inflow or im-
paired neuropeptide uptake, or the incorporation of neurotrophic factors
might be defined as an example of plasticity (or lack of it). In other words,
changes in neurotransmitter levels or ionic influx or changes in mem-
brane structure can become the definition of dynamic, neuronal plas-
ticity. One step up, at another level of analysis, one might look at how a
given region of the brain releases or induces growth promoting (trophic)
factors in response to injury. The specific circuits or pathways are not
evaluated. For biochemical assay, the tissue itself is ground up and de-
stroyed so that morphological detail is forever lost. At an anatomical
level, researchers have discussed neuronal and synaptic plasticity in
terms of growth or regeneration of neurons, and they carefully map the

connections and pathways that are formed. And at yet another level there
are investigators who focus upon alterations or changes in specific struc-
tures of the CNS in response to injury.
In this context, the analysis turns from neurochemical assay and fix-
ed morphology to the evaluation of ongoing activity in the living tissue in
situ. Such assays often employ electrophysiological recordings from brain
areas as subjects perform various tasks (e.g., maze running in rats, prob-
lem solving in patients). More recently, functional analyses of cerebral
blood flow or other measures of ongoing metabolic activity employing
various types of auto radiographic assays (e.g., NMR, CAT scan, uptake of
labeled glucose analogues) have been used to study what different brain
regions might be doing (e.g., active, inhibited) during task performance.
Following brain damage, for example, an investigation could employ
these "functional" techniques to ask whether there is substitution of func-
tion, wherein one area of the brain is thought to take over the function of a
damaged neighbor (Slavin, Laurence, & Stein, 1988). Finally, one can
examine changes in the cognitive, emotional, or perceptual aspects of
behavior after injury, bringing us to the psychological level of definition.
Here, there are sometimes attempts to correlate behavioral deficits with
the locus of the injury, but there is not much interest in specifying the
underlying mechanism causing the actual deficits or the recovery that can
sometimes follow."l
Is there one level of analysis that is more appropriate for understand-
ing CNS plasticity than another? I believe that if we are to understand
recovery of function, we will need to apply a number of different levels of
analysis. It is very doubtful that we will find a simple, monolithic expla-
nation for "recovery." In this context, we should not fall victim to the idea
that our approach is the best or that only the latest in new technology (e.g.,
molecular biology) can provide the answer. Given the complexity of the
problem and the need for understanding CNS function at multiple levels,
I believe that the best approach is to develop cross-disciplinary collab-
orative efforts.
In my laboratory we emphasize the importance of behavioral change,
but we try to relate those behavioral changes to the different structural,
anatomical, and biochemical alterations that could account for the "plas-
tic" events we study. Roger Sperry in his Nobel Prize-winning address
1psychologists often criticize their colleagues in the neurosciences for what sometimes
seems to be a highly demonstrable lack of interest in what the nervous system does with
respect to behavior. However, in all fairness to those in the neurosciences, there are many
psychologists who feel there is nothing to be learned by developing a better understanding
of how the nervous system works. This "black box" approach to the brain is still very much
with us in academic and clinical psychology. Neither level of parochialism can ultimately
benefit our understanding and treatment of neural and behavioral pathologies.
236 DONALD G. STEIN
commented about whether or not reductionistic or molar approaches are

appropriate and said, "All approaches to the analysis of 'plasticity' are
legitimate and useful. Analysis at one level does not obviate the useful-
ness of analysis at another level."
It is also worth mentioning the obvious, and that is that change at any
one level of function has the potential to alter plasticity or determine it at
another level. For example, there is now a growing body of evidence (see
Nieto-Sampedro, 1988) that growth-promoting and survival-promoting
factors are endogenously induced by brain injury in adult subjeds. These
trophic substances that are produced at the site of injury increase the
survival of neurons that would ordinarily die after injury. The surviving
cells then maintain their projections (contacts) with the damaged zone
and thus permit sparing or more rapid recovery of behavior. After a CNS
injury there is a cascade of effects, over time, represented by alterations in
brain biochemistry on up to significant behavioral changes. Thus, chem-
ical factors that promote the survival of neurons are critical for mor-
phological survival and reinnervation, which, in turn, might be essential
to maintain behavior.
ENVIRONMENTAL AND BEHAVIORAL FACTORS "SHAPE"

CNS MORPHOLOGY
It is also important to point out that behavioral activity can be used to
promote neuronal activity that stimulates the growth of dendrites and
synapses. In other words, how an organism responds to its environment
can have structural (morphological/biochemical) effects on the central
nervous system.
In one program of research, Greenough and Chang (1985) exposed
rats to either isolated environments (in which animals live alone), social
environments (in which there are several animals in a cage), or complex
"condominiums" in which the animals are exposed to "enriched" stim-
ulation, where they have an opportunity to interact vocally, play, and be
exposed to many objects that are changed daily. After the rats were chron-
ically exposed to the different environments, Greenough and his students
employed the Golgi technique of staining entirely a select number of
neurons, which could then be examined in three dimensions for the ex-
tent of dendritic branching and the number of spines (the processes along
the dendrite where synaptic end-feet form the synapse and maintain con-
tact with the afferent neuron). The number of dendritic branches and the
number of spines are painstakingly counted to determine if exposure to
different environments can "feed back" on the brain to produce structural
alterations in the neurons themselves.
It was shown that chronic exposure to a "complex" environment can
increase the length of dendritic fields in several classes of neurons found

in the visual cortex. Exposure to complex environments and maze-learn-
ing experience also contributed to increasing the number of synaptic vesi-
cles (the structures within the synaptic end-feet that contain neurotrans-
mitters). Greenough and Chang (1985) have repeated and reviewed many
other structural changes in brain morphology that are related to specific
experience. In summary, the environment-induced changes in mor-
phology and ultrastructure of neurons is important because these changes
are directly related to the ability of the animal to process, store, and
retrieve information.
The incredible amount of time required to perform both light- and
electron-microscopic evaluations of the qualitative characteristics of den-
drites and synapses in response to environmental stimulation forces in-
vestigators to focus on only a few brain regions. But if behavioral manip-
ulation can, in fact, alter CNS morphology, we cannot assume that only
the limited areas under study are changed. Can one assume that all other
nuclei and their connections (pathways) remain fixed in function and
activity, that the pathways are immutable? Or is it likely that many struc-
tures are in a state of structural flux in response to environmental pres-
sures? If so, then what is the appropriate "time" to measure a constant
aspect of CNS morphology?
The belief in fixed elements in the CNS can be considered as a philo-
sophical statement that derives from the kinds of methods that are em-
ployed in the study of brain hodology. Even in the face of the newer,
metabolic (functional) techniques, most neuroanatomical studies of the
brain are made in dead tissue that is prepared at a single time, using
highly toxic preservatives (or deep freezing) to permit cutting and mount-
ing on slides. Geoffrey Raisman (1978), an electron-microscopist, known
for his seminal work demonstrating injury-induced growth in the adult
CNS, said that "it is this very feature of anatomical methodology that has
tended to give anatomical observations the appearance of permanence
and rigidity. The fact that a synapse exists and can be photographed in a
form in which it appeared at the moment when the animal was killed
gives no indication of how long that particular synapse had existed before
the moment of death or how long it would have remained had the animal
survived" (p. 102).
Raisman's observations are consistent with the results of an elaborate
electron-microscopic investigation of serotinergic (5-HT) nerve terminals
in the frontal cortex of adult rats conducted by Descarries, Baudet, and
Watkins (1975). They used 3H-5HT labeling to mark the nerve terminals
and then sampled 50,000 photomicrographs to determine that only about
5% of the nerve terminals had formed "synaptic contacts" with "postsyn-
aptic" membranes. Yet the presynaptic "knobs" had all the charac-
teristics and morphology of normal terminals. The authors suggested that
238 DONALD G. STEIN
the presence of so many "free" terminals are not compatible with the
view that the frontal circuitry is fixed and static. Instead they suggested
that in living tissue there is an ongoing process of synaptic shifting and
relocation that cannot be seen when tissue is examined postmortem.
If this type of plasticity does occur in the adult eNS, then it is possi-
ble to think that synaptic morphology and function will be altered by the
organism's experience. 2
If we can agree that behavioral manipulation can have a direct effect
on eNS morphology, can such manipulations alter the outcome of brain
injury? If, for example, environmental "enrichment" can enhance recov-
ery or sparing of function after injury, under what conditions, 'then, is it
appropriate to "specify" structure-function relationships in the central
nervous system? Here one might see environment or training as part of the
developmental process contributing to the dynamic reorganization of
eNS morphology and function. Anatomical measurements taken from
animals exposed to different environments have shown that enriched
conditions can result in greater cortical thickness and more complex syn-
aptic profiles in comparison with those raised under isolated conditions.
Kelche and Will (1982), in a follow-up to their observations of behavioral
recovery, demonstrated that hippocampal lesions can produce a reduc-
tion in the number of dendritic spines of pyramidal neurons in the oc-
cipital cortex. However, the number of spines was increased if the oper-
ated rats were maintained in complex environments. A thorough review
of the issues surrounding the effects of differential housing and outcome
of brain injury has been prepared by Dalyrimple-Alford and Kelche
(1985).
Here again, the summary point to be made is that the context in
which injury (or any other ongoing event) occurs cannot be overlooked in
developing a theory of nervous system specificity of plasticity.
TEMPORAL FACTORS ALSO IMPORTANT IN PREDICTING

THE OUTCOME OF BRAIN DAMAGE
Now we have seen that organismic (hormonal) as well as environ-
mental (differential housing) effects can alter the "specific" outcome of
brain injury. But even these important variables are alone, not always
sufficient to account for defining structure and function in the eNS. When
we look at the effects or the outcome of brain injury in order to define
recovery, plasticity, or deficit, very often we look at relatively restricted
periods of time following the injury (or treatment). Yet investigators in
clinical neuropsychology have long recognized that it is not sufficient to
2Indeed, some investigators concerned with the anatomical substrates of learning (e.g., Bau-
dry & Lynch, 1984) are reevaluating the earlier notions of Hebb (1949) in this very context,
but this discussion is beyond the scope of this chapter.
look at one fixed period of time to examine (and predict) the outcome of
brain injury. In one of the last papers of Norman Geschwind (1985) dis-
cussing the mechanisms of change after brain injury, he said that "there
must be many cases in which the capacity for recovery is latent and
revealed only by some further manipulation, but experimenters have only
rarely been zealous in their search for the right maneuver" (emphasis
added). Geschwind argued that the time needed for recovery in adults
may be much longer than in infants. He went on to say that "most neu-
rologists are gloomy about the prognosis of severe adult aphasia after a
few weeks, and pessimism is reinforced by lack of prolonged follow-up in
most cases." He then said: "I have, however, seen patients severely ap-
hasic for over a year who then made excellent recoveries: one patient
returning to work as a salesman, the other as a psychiatrist. Furthermore,
there are patients who continue to improve over many years. For example,
a patient whose aphasia was still quite evident six years after onset clear-
ed substantially by eighteen years. Change after damage may show great
individual variation" (p. 3). The concept of delayed recovery is yet an-
other example of the "plasticity" that needs to be explained in developing
an appropriate conceptual model of nervous system organization. There is
no rule saying that, if recovery is to occur, it has to be immediate! One of
the tasks of our laboratory and of the many students who have worked in
it with me is to find a way to promote recovery as rapidly as possible, to
get it to occur in a shorter period of time than the 6 or 18 years mentioned
by Geschwind. Very often, as Geschwind indicated, we tend not to look
for recovery if it has not occurred for a very long period of time. Thus, we
can often miss the fact that the processes that underlie central nervous
system plasticity may take far longer than we have previously assumed.
We should also not overlook the fact that the "potential for plasticity"
may be present but blocked by ongoing processes inimical to its man-
ifestation. The focus in research, then, is to unlock those neural events
that can lead to functional recovery even though the time course may be
far longer than we would like to imagine.
ADAPTIVE AND MALADAPTIVE ANOMALOUS GROWTH IN

RESPONSE TO eNS INJURY
In summary, and with respect to the inherent potential for CNS re-
organization, it should be kept in mind that whether or not recovery
occurs depends upon the various interactions taking place around the
time of the injury. When we try to make definitive statements, particularly
about the determinative aspects of structure-function relationships, we
have to recognize that they are primarily probabilistic statements. In other
words, the outcome (i.e., a specific deficit), like the outcome of epigenesis
or development itself, is not inevitable because one cannot specify all of
240 DONALD G. STEIN
the factors that determine or influence the organism in its environment.

With respect to brain injury, the extent to which the processes we refer to
as "plasticity" are successful in mediating functional recovery depends
upon the coming together of the appropriate variables at the appropriate
time.
In this same context we can also point to examples of "negative"
plasticity. Here one can think about the anomalous growth of neurons in
response to injury that sometimes results in maladaptive behavior.
About 15 years ago, Gerald Schneider and his students at M.LT. be-
gan a series of studies designed to study the anatomical basis for restora-
tion of visual function following lesions of the optic tectum iN. neonatal
hamsters (e.g., Schneider & Jhaveri, 1974). The hamster is a good model
for this type of research because at birth the cortex has not yet grown over
the superior colliculus, so the structure is easy to visualize and remove
surgically. In response to this injury, Schneider found that fibers coming
from the optic nerve do not simply die off. Instead, there is considerable,
redirected growth that could be systematically studied and evaluated for
its contributions to functional recovery-in this case restoration of vision.
If one were to examine the normal pathway of neurons from the eye
crossing over at the optic chiasm and entering the superior colliculus
(SC), one would see that most of the fibers terminate in the superficial
layer of the Sc. If, however, the SC is removed, the growing fibers end up
by entering into the posterior portion of the lateral posterior nucleus,
which does not, in normal animals, receive a heavy innervation from the
optic nerve. Some of the fibers even terminate in the medial geniculate
body, a part of the auditory system that does not receive such projections
in normal animals. In addition, if one eye that normally has its termina-
tions in the opposite hemisphere in the superior colliculus is removed,
the fibers from the intact eye cross to the opposite hemisphere, but in the
absence of an intact superior colliculus, they grow back across the mid-
line of the brain to reinnervate the contralateral tectum, where they
should not be.
What do these findings show? First, Schneider's data demonstrate a
clear example of injury-induced neuronal plasticity, unequivocal evi-
dence of neural regeneration and regrowth in the developing hamster.
What of the behavioral correlates?
First, it does appear that the anomalous growth forms "structurally
normal-looking synapses in abnormal places after neonatal lesions of the
se" (Kalil & Schneider, 1975), and single-cell electrophysiological re-
cordings have been used to demonstrate that the new synapses are func-
tionally active.
Second, when sunflower seeds were presented in various parts of the
visual field, the hamsters would turn in the direction opposite of (away
from) the presentation. In these animals, it was later confirmed that the
developing retinal projections grew to the "wrong" side of the brain as a

result of the neonatal injury (see Schneider, 1979, for a detailed review).
Schneider (1979) pursued the role of the abnormal projections one
step further by asking what would happen if the abnormally recrossing
fibers were to be eliminated in the adult animal. First, he mapped the
visual fields (with perimetry studies) where both correct turning and in-
correct turning to the sight of a sunflower seed could be elicited. Next,
microsurgery was employed to transect the recrossing bundle of optic
fibers. In the animals whose anomalous fibers were surgically prevented
from recrossing the midline, turning in the wrong direction was com-
pletely eliminated, while turning in the appropriate direction to obtain
reinforcement remained completely intact!
Schneider's clever experiments clearly demonstrate that, at least in
early life, there is considerable neuronal reorganization in response to
brain injury. However, it is sometimes the case that the redirected growth
ofaxons can result in significant behavioral abnormality as well as poten-
tial sparing of function. The task of experimenters working in this area is
to determine under what conditions such growth might lead to enhanced
recovery and what leads to impairment. Unfortunately, most investigators
interested in neuromorphology rarely examine the behavioral
consequences of the anatomical "plasticity" that they observe, so one
does not often know whether the changes seen are beneficial, maladap-
tive, or of no consequence to the functioning organism.
Schneider's work highlights the importance of careful behavioral fol-
low-up following neonatal brain injury. Even more important, his work
shows that, in some instances, maladaptive behavior caused by anoma-
lous growth can be eliminated by later surgery.
Recently, Earle (1987) reviewed some of the literature on lesion size
and recovery following damage to a number of different neural systems
and concluded that, under the right conditions, secondary lesions or ex-
pansion of lesion size could actually lead to better recovery than with
more circumscribed injury. One might speculate that the more extensive
injuries could remove some of the "neural noise" (e.g., inhibition, spas-
ticity) created by anomalous growth resulting from the initial damage.
This is an area of research that clearly needs more attention, despite the
fact that it falls outside of the more "traditional" approaches to structure-
function analyses in the CNS.
BEHAVIORAL CONSEQUENCES OF LESIONS EARLY IN LIFE

VIS-A-VIS INJURIES AT MATURITY
To demonstrate further how "context" plays such an important role

in determining the outcome of CNS injury, let me now turn to studies of
242 DONALD G. STEIN
functional recovery after spinal cord damage in newborn and adult cats.
Over the last several years, Barbara Bregman and Michael Goldberger
(1983a,b) have been examining both the behavioral and anatomical corre-
lates of recovery in cats with spinal cord injuries inflicted early in life or
at maturity. First, Bregman and Goldberger carefully analyzed motor pat-
tern recovery after hemisections of the spinal cord in newborn or adult
animals. The cats' movements were filmed and quantified, and, although
both groups showed recovery of function, the results indicated that the
recovery was not "uniform"; some motor behaviors were better after adult
injury, while some recovered better when the damage was inflicted early
in life (Bregman & Goldberger, 1983b).
For instance, tactile placing was completely abolished after spinal
cord hemisection in adults, but no loss of hind limb placing occurred
when the lesions were made in newborn animals. In contrast, accurate
foot placement during locomotion was much better in adult operated
animals, as was hopping with the foot opposite to the side of the lesion.
Bregman and Goldberger suggested that recovery or sparing of function
occur when the lesions are made prior to the time that the motor reflexes
begin to develop in normal animals. The better recovery in cats damaged
as adults is thought to be due to the possibility that there may be more
(neural) sources (spinal and supraspinal) that could participate in the
adjustments needed for good locomotion.
To support these speculations, Bregman and Goldberger (1982,
1983a,b) carefully examined the anatomical changes that might account
for the sparing or failure of motor function they observed in their injured
cats. As in the visual system, lesions in the immature spinal cord lead to
the development of aberrant neuronal fiber paths, which could mediate
some degree of recovery (or maladaptive behavior).
Using retrograde neuronal marker techniques (HRP), the authors ex-
amined histopathology in cats with spinal lesions at 1 day of birth or as
adults. The anatomical results supported the complexity of their behav-
ioral data. First, whether lesions were made in neonates or adults had no
effect on sparing of brainstem-spinal fibers-there was no "plasticity"
here. In addition, neonatal lesions led to much greater retrograde degener-
ation in this path than did similar lesions created in adult cats. In con-
trast, corticospinal projections were dramatically spared, but "only in the
neonatal group." There were also anomalous projections in the latter
group that were not seen in the former (corticospinal neurons reached
lumbar segments of the cord by an alternate course). Bregman and Gold-
berger emphasize that tactile placing is completely abolished after adult
lesions, but completely spared after neonatal injury. One can infer, then,
that the functional sparing is the result of the "anatomical plasticity of
(this new) corticospinal projection." The sparing and development of the
anomalous path is likely due to the relatively late arrival of this pathway
during the normal course of development. Thus, since the fibers are not
really present at the site at the time of the injury, they can grow past the
site and reroute themselves to their appropriate targets (Bregman & Gold-
berger, 1982).
In summary, the work on early versus late injury and anomalous
growth highlights yet again the importance of the context; in this case,
stage of maturation can determine the outcome of traumatic brain injury.
The factors that lead to survival and growth on the one hand, or loss of
neurons on the other, are not well understood. Nonetheless, there is con-
siderable progress being made to unlock the appropriate conditions or
factors that lead to consistent recovery of function where none is ex-
pected.
Especially when attempts are made to facilitate recovery with drugs
or experimental agents, the specific context, experience, or "background"
of the subject will interact with the "treatment" to determine the outcome
of the brain injury. In adult organisms, the potential for plasticity in the
form of reorganization (or deblocking) of function is also present, but
different rehabilitative strategies may be required to activate the mecha-
nisms underlying the recovery itself.
One last example taken from the work of Dennis Feeney and his
students will suffice. Feeney, Gonzalez, and Law (1982) found that injec-
tions of amphetamine can enhance the levels of the neurotransmitter nor-
epinephrine in adult animals who have been depleted of this substance by
lesions of the catecholaminergic system. Injuries in the motor cortex often
block or eliminate visual, postural, proprioceptive, righting, and tactile
placing reflexes. Amphetamine administration can apparently restore
these reflexes and restore motor and visual performance, even after acute
bilateral injury. In the face of a considerable amount of literature, it would
seem that such a finding would lead to the development of an effective
clinical treatment for brain injury, but the question is more complex.
First, in reviewing the literature, Feeney and Sutton (1987) found that
recovery could not be maintained if animals did not receive repeated
testing (general visual experience in the home cage was not sufficient). In
cats, many tests of the specific placing response had to be administered
for the drug to be effective in inducing recovery. To test the importance of
experience in amphetamine-induced recovery from brain injury, Feeney
and his students restrained rats with motor cortex lesions in a small box
during amphetamine intoxication and compared them with counterparts
allowed to roam freely or who were given beam-walking experience.
Blocking ballistic movements eliminated the beneficial amphetamine ef-
fects entirely (Hovda & Feeney, 1984).
In reviewing their work, Feeney and Sutton (1987) suggest that "drug
therapy alone may not be efficacious for inducing recovery. Rather a com-
bination of drug plus appropriate behavioral training and testing
244 DONALD G. STEIN
(emphasis added) is required for induction of recovery. Importantly, these

studies suggest that the capabilities (and neural pathways) for behavioral
recovery may be present in brain-injured patients, but may never function
or be expressed unless appropriate treatments are provided" (p. 142).
IMPORTANT INDIVIDUAL DIFFERENCES IN RESPONSE TO

BRAIN INJURY THAT ARE NOT ADEQUATELY STUDIED
For the most part, the prognosis for recovery from brain damage in
humans is poor. When it does occur, it is often characterized as an in-
teresting anomaly, and there is little in the way of systematic inquiry into
either the historical or present conditions that led to the patients' unique
outcome in comparison with those who seem to be less fortunate. Yet
individual differences in response to the same type or extent of injury
could provide important clues as to some of the mechanisms underlying
functional recovery. As early as 1902, two developmental neurologists,
Baldwin and Polton, said that "apart from plasticity in general, the facts
of individual accommodation make the nature and limits of brain plas-
ticity a matter of great interest." They may have been among the first to
talk about individual differences in response to brain injury.
Unfortunately, in the laboratory setting, our own empiricist orienta-
tions and the social conditions in neuroscience do not often permit the
thorough study of individual accommodation to trauma. We are forced to
employ statistical averaging methods to seek the mean case, and we are
taught that deviation from the average is due to error in measurement. In
the final analysis, we will need to address the question of whether there
are effective tools to pursue our understanding of why there is so much
individual variability in response to brain injury and what that variability
can tell us about mechanisms of eNS plasticity.
Let me illustrate this point further by referring to work by R.
Thompson (1978). In A Behavioral Atlas of the Rat Brain Thompson
plotted all of the different areas reported by researchers to affect different
behavioral functions. The ostensible purpose of the atlas was to provide
the reader with a visually easy way to determine where one could create a
specific lesion to produce a specific deficit. In illustrations, Thompson
showed where damage to the brain results in maze-learning deficits. In
examining all of the areas that appear to produce deficits in learning
ability, it is hard for me to conceive of the fact that behavior as complex as
maze learning could be localized in a single given structure or pathway in
the central nervous system. Yet this atlas represents the results of many
studies that have attempted to do just this. For the sake of argument one
could say that with respect to a phenomenon as amorphous as learning,
the processes involved should be relatively distributed throughout "high-
er" regions in the central nervous system. In contrast, consummatory or

regulatory behaviors would be expected to be controlled by more "pre-
mature" subcortical structures, and there would be much less plasticity
when the "centers" controlling innate behaviors such as eating and drink-
ing are destroyed.
Thompson's work contains another map showing where damage re-
sults in aphasia and adipsia. Damage to the areas of the brain demarcated
there results in failure to eat and drink. Animals who suffer lesions of this
type will usually die if there is no forced feeding or drinking. Earlier
research on consummatory behavior had localized these functions to the
lateral and ventromedial hypothalamus. However, it can be demonstrated
that even behavior as basic as eating and drinking can be affected by
damage to a wide variety of cerebral "centers." Conversely, one may look
first for one type of behavior while ignoring many other changes because
that is what the person is interested in studying. The point that I wish to
make here is that any given lesion will have widespread effects in the
CNS. These effects (anatomical, biochemical, electrophysiological, behav-
ioral) will be tempered or modulated by the individual history of the
organism as well as the contemporary context in which the injury has
occurred.
For example, Valenstein, Cox, and Kakolewski (1968) implanted bi-
lateral electrodes into the lateral hypothalamus of adult rats. Electrical
stimulation of this area was thought to activate the "excitatory" areas
controlling eating (while lesions resulted in aphagia and adipsia). In the
presence of food, Valenstein and colleagues stimulated the rats through
the electrodes, and they observed that even though the rats were sated,
they would begin to eat. This would certainly seem to confirm the hypoth-
esis of an excitatory center for eating, except that the researchers carried
their experiment several steps further.
The food was removed and either water or wood shavings were
placed in the animals' cages. When the electrical stimulation was turned
on in the presence of these "stimuli," the rats drank or gnawed on the
chips, depending on what was available. When nothing was present,
some of the rats simply picked up their tails and carried them around. It
would seem from these data that the "context" or environment plays a
major role in determining the behavioral outcome during activation of a
CNS region. The different behaviors could have resulted from the ever-
changing interplay of CNS regions rather than the specific "output" of a
singular region.
In a similar experiment, Gazzaniga (1974) made rats severely adipsic
by creating large lateral hypothalamic lesions. Such animals will not
spontaneously drink and will eventually die unless forcibly fed water.
Rats with these lesions will, however, try to run in an activity wheel.
Taking advantage of this fact, Gazzaniga made running in the wheel
246 DONALD G. STEIN
(through behavioral shaping) contingent upon drinking first from a water

spout for increasing amounts of time. The adipsic rats quickly learned to
drink in order to be able to run.
Had Gazzaniga not manipulated the environment appropriately, he
would have obtained the "classic" syndrome and confirmed the "princi-
ple" that the lateral hypothalamus controls drinking and eating. Instead,
he, too, was able to show that the outcome of injury depends on the
context in which the damage occurs rather than just the site itself.
It is appropriate that in experimental situations we try to control, as
much as possible, the history and current context of the organism so that
we can explain how our various manipulations might work. Nonetheless,
in the clinic situation, carefully examining, rather than ignoring or down-
playing variability, may be the key to understanding why some indi-
viduals seem to miraculously escape the effects of CNS injury while oth-
ers are so devastated.
Work has been done showing what happens when a lesion is made in
the rat's motor cortex. Results have shown the extent of change in the CNS
that can occur from a focal lesion. The degeneration does not take into
account trans neuronal degeneration or retrograde degeneration where
neurons that project to the lesion site also die off over time. Even though
the degeneration represents monosynaptic changes, one can see how
many different structures in the central nervous system, including visual
areas, motor areas, and nuclei that are involved in consummatory behav-
ior, are altered by loss of afferent inputs as a result of this lesion. The
organism as a whole will change as the "outputs" of the different struc-
tures begin to adapt (or fail to adapt) to the brain injury.
Given the greater complexity of the human brain, would we expect
less, or even more, variability in people? A composite map taken from
Kertesz (1983) shows that direct electrical stimulation of the brain tested
prior to surgery for excision of epileptogenic foci produces alterations in
naming and sentence reading throughout a wide area of the cortical man-
tle, and not just in Wernicke's or Broca's area.
Once again the clash of concepts of plasticity, adaptability, and indi-
vidual variability in the central nervous system and that of the concept of
trying to localize functions by ever finer units of analysis (e.g., command
neurons, pontifical cells) comes from the fact that there are really deep-
rooted preempirical ideas about the nature of "neural" reality. Basic para-
digms that scientists hold determine the methodologies they select and
what interpretations they impose on the data that they collect. Stephen
Jay Gould called this social phenomenon the "presupposition of limits."
These presuppositions of limits, spoken or unspoken, certainly affect our
views of how we conceive of the development of plasticity in both adult
and young animals and what should or should not be studied.
An eloquent example of an explicit presupposition of limits was

stated by a distinguished visitor to the United States early in this century.
That visitor, a Nobel Prize winner in neuroanatomy, was Santiago Ramon
y Cajal, who said: "Once development is ended, the fonts of growth and
regeneration of the axons and dendrites dry up irrevocably. In adult cen-
ters the nerve paths are something fixed and immutable. Everything may
die; nothing may be regenerated" (Cajal, 1928, p. 750). That is one very
good example of a "presupposition of limits."
Another, more modern example of the presupposition is taken from
the work of R. F. Thompson (1986), who made the following statement
about memory, "All evidence to date indicates that the mechanisms of
memory storage are local and do not involve the formation of new projec-
tion pathways. Furthermore, to the extent that they have been identified,
essential memory trace circuits in the vertebrate brain are localized" (p.
944). Thompson's statement clearly implies that there is a very limited
path (or set of methodologies that one should follow in defining the mem-
ory process or, by inference, in trying to develop a proper scientific mea-
sure of brain injury outcome or the process of recovery.
Contrast Thompson's statement with another that was made a month
later in the same journal by John, Tang, Brill, Young, and Ono (1986).
John, who is also interested in the physiological basis of learning, first
trained cats to perform a maze-learning task. He then used radioactive
uptake of a glucose analogue to measure changes in metabolic activity in a
large number of brain structures as the animals tried to learn visual dis-
crimination tasks. In light of his results, John concluded (and remember
this follows 1 month after Richard Thompson's remarks):
Our results do not fit well with the general computer-like model of the brain
with information stored in discrete registers, no matter how many in number. A
radically different model is necessary. Our data, like data from diverse elec-
trophysiological and anatomical studies, better support notions of cooperative
processes in which the nonrandom behavior of huge ensembles of neural ele-
ments mediate the integration and processing of information and the retrieval
of memory. In view of the large number of neurons involved the question of
how information represented in these neurons can be evaluated and appreci-
ated by the brain becomes of critical theoretical importance. No conceivable
neuron or set of neurons, no matter how diffuse its synaptic input, can evaluate
the enormous amount of neural activity here shown to be involved in retrieval
of even a simple form discrimination. Memory and awareness in complex
neural systems may depend upon presently unrecognized properties of the
system as a whole and not upon any of the elements that constitute the system.
(p.1174)
What we have here is a clear clash of paradigms. The question be-

comes what is the most appropriate way of analyzing and looking at brain
function? (What is the best paradigm?) John uses the same scientifically
248 DONALD G. STEIN
accepted methodologies employed by Thompson (single-cell elec-

trophysiology, brain biochemistry, behavioral analysis) but comes to very
different conclusions and interpretations about how the brain works.
Kuhn, writing in The Essential Tension (1977), says:
A paradigm is a set of beliefs and attitudes shared by a group of people that are
committed to the same rules and standards of scientific practice. This commit-
ment is a prerequisite for normal science, that is, the genesis and continuation
of a particular research tradition. Without this commitment, any set of beliefs,
attitudes, and techniques could serve as acceptable approaches to reseEjrch.
And that research would basically be a random activity with no unifying
theme. Almost anything would be equally relevant. (p. 294)
I believe that basically what we need to do is to establish a different

paradigm relevant to the study of recovery from brain injury. Kuhn (1970)
also says the following:
Normal science, the activity in which most scientists invariably spend almost
all of their time, is predicated on the assumption that the scientific community
knows what the world is like. Much of the success of the enterprise derives
from the community's willingness to defend that assumption, if necessary, at
considerable cost. Normal science, for example, often suppresses fundamental
novelty because they are necessarily subversive of its basic commitment.
Nevertheless, so long as those commitments retain an element of the arbitrary,
the very nature of normal research ensures that novelty shall not be suppressed
for very long. (p. 5)
In summary, the examples that I have provided show that there is still
turmoil in neuroscience that is being generated by dramatically different
ways of looking at nervous system functions. It is an important conflict
that spills over even into the more clinical areas of human neuropsychol-
ogy. Let me again repeat the quote I gave earlier from the neuroscience
texts of Kandel and Schwartz (1981), here describing the clinical psycho-
pathology of temporal lobe function. They claim that both cognitive func-
tions and character traits are anatomically localizable. Thus, "their
clinical studies and their counterparts in experimental animals suggest
that all behavior, including higher mental functions, is localizable to
specific regions or constellations of regions within the brain. The role of
descriptive neuroanatomy is, therefore, to provide us with a functional
guide to the localization within the three-dimensional neurospace, a map
for behavior" (p. 11). Given the data presented so far, perhaps the reader
should draw his or her own conclusions about which of the competing
paradigms might be most appropriate for the study of physiological and
organismic plasticity.
Finally, I present here the summary statements taken from a recent
book on cortical localization in clinical neuropsychology (Kertesz, 1983).
Kertesz argues that "the function of an area is related to the lesion if: (1)
The same functional deficit always follows the lesion. (2) The same deficit
is not produced by other independent lesions. (3) The deficit is measured

according to standardized and meaningful methods. (4) The lesion loca-
tion is determined objectively and accurately. (5) Biological variables,
such as time from onset, age, etiology and so forth, are controlled" (p.
294).
If one takes the contextual and organismic view that I have presented
in this chapter, it seems to me that none of the criteria described by
Kertesz can be used to infer "the function" of a given area. If the paradigm
that I am proposing has merit, then one must begin to question whether
traditional neuropsychological testing can provide meaningful answers to
the questions about the physiological substrates of behavior in normal
and brain-damaged subjects.
As Nottebohm (1985) recently stated in the preface to his book on
CNS plasticity, new frontiers in the neurosciences are constantly being
discovered. "However, even as we bow to the powers of reductionism and
to the promise that molecular approaches hold for brain repair, we should
remember that brains are also more than vast arrays of molecules" (p. ii).
SUMMARY AND CONCLUDING REMARKS
As we come to the end of this essay, I think it is safe to say that the
neurosciences are undergoing a dramatic shift in ideas about the "limits
of plasticity" in the central nervous system. We have seen that there are a
variety of developmental processes leading to recovery or change that
must be considered in describing or predicting the outcome of brain inju-
ry. It is no longer possible to describe or report symptoms simply in terms
of damage to a specific path or circuit because those paths or circuits may
not be constant. The conditions under which damage occurs at a given
moment in the cumulative history of the organism must now be taken into
account in our attempts to define structure-function relationships in the
CNS. In this context, we must also be more aware of the fact that environ-
mental influences can also modulate and sculpt neuronal morphology,
such that the functional demands made on morphology may determine
the output of a given neural area; thus, function (environment) may play
some role in determining structure.
Mapping of neuronal circuits in fixed tissue is an important and
useful endeavor in that it may give us the conditions by which we can
begin to explore the more dynamic aspects of nervous system function.
However, when such mapping becomes an end in itself, or when it be-
comes merely the demonstration of technique without any reference to
behavior, then the science of understanding the function of the nervous
system will not really progress. The fashion now is toward an ever-in-
creasing, mechanical and ideological "reductionism" where behavior is
250 DONALD G. STEIN
often seen as an almost unwanted by-product that complicates the pro-

cess(es) of studying neuronal tissue just for its own sake.
It may be time to move away from the concept of the brain as a colony
of neurons that, when all the maps are done and all the circuits plotted,
will give us a complete "understanding: of complex phenomena like
learning, memory, and emotion. This view has been with us for almost a
century and is probably one of the reasons why so little has been done to
help the victims of brain and spinal cord injury (Le., it is obvious that
removal or damage of fixed and hard-wired circuits must lead tQ perma-
nent impairment once the cortical component is gone). Now there is be-
ginning to emerge a hope for a new neurology, as Nottebohm claimed in
the title of his recent volume. There is a beginning (and maybe even
grudging) acceptance of the multitude of factors that can shape and modu-
late nervous system function in normal organisms and alter the outcome
of recovery in brain-damaged patients.
We have made great strides in our understanding of many operations
of the nervous system at the molecular level through the employment of
reductionistic research strategies. Now we have to begin to ask whether
the collective properties of the nervous system may display novel features
that are not discernible in the detailed observation of its component parts.
Paul Weiss recognized this when he repeatedly emphasized that the dy-
namic interaction between neurons may not be more than the sum of the
parts, but, surely, it could be very different!
ACKNOWLEDGMENTS
A fuller version of this chapter was published in the Heinz Werner

Lecture Series-Clark University Press. This chapter is published with per-
mission from Clark University Press.
REFERENCES
Attella, M., Nattinville, A., & Stein, D. G. (1987). Hormonal state affects recovery from frontal
cortex lesions in adults females rats. Behavioral and Neural Biology, 48, 352-367.
Baudry, M., & Lynch, G. W. (1984). Glutamate receptor regulation and the substrates of
memory. In G. Lynch, J. L. McGaugh, & N. M. Weinberger (Eds.), Neurobiology of
learning and memory (pp. 431-450). New York: Guilford Press.
Bregman, B., & Goldberger, M. (1982). Anatomical plasticity and sparing of function after
spinal cord damage in neonatal cats. Science, 237, 553-555.
Bregman, E., & Goldberger, M. (1983a). Infant lesion effect: II. Sparing and recovery of
function after spinal cord damage in newborn and adult cats. Developmental Brain
Research, 8, 119-135.
Bregman, B. S., & Goldberger, M. E. (1983b). Infant lesion effect: III. Anatomical correlates of
sparing and recovery of function after spinal cord damage in newborn and adult cats.
Developmental Brain Research, 9, 137-154.
Cajal, S. (1928). Degeneration and regeneration of the central nervous system (R. M. May,
Trans.). London: Oxford University Press.
Dalyrimple-Alford, J. D., & Kelche, C. (1985). Behavioral effects of pre-operative-post-
operative differential housing in rats with brain lesions: A review. In B. E. Will. P.
Schmitt, D. C. Dalyrimple-Alford (Eds.), Brain plasticity, learning and memory. Ad-
vances in behavioral biology (Vol. 28, pp. 441-458). New York: Plenum Press.
Descarries, L., Baudet, A., & Watkins, K. C. (1975). Serotonin nerve terminals in adult rat
neocortex. Brain Research, 100, 563-588.
Earle, E. (1987). Lesion size and recovery of function: Some new perspectives. Brain Re-
search Reviews, 12, 307-320.
Feeney, D. M., Gonzalez, A., & Law, W. A. (1982). Amphetamine, haloperidal and experience
interact to affect rate of recovery after motor cortex injury. Science, 217, 885-857.
Feeney, D. M., & Sutton, R. (1987). Pharmacotherapy for recovery of function after brain
injury. Critical Reviews in Neurobiology, 3, 135-197.
Gazzaniga, M. (1974). Determinants of cerebral recovery. In D. G. Stein, J. J. Rosen, & N.
Butters (Eds.), Plasticity and recovery offunction in the eNS (pp. 203-216). New York:
Academic Press.
Geschwind, N. (1985). Mechanisms of change after brain lesions. In R. Nottebohm (Ed.),
Hope for a new neurology, Annals of the New York Academy of Sciences, 457, 1-11.
Gollin, E. S. (1981). Development and plasticity. In B. S. Gollin (Ed.), Developmental plas-
ticity: Behavioral and biological aspects of variation in development. New York: Aca-
demic Press.
Greenough, W., & Chang, P. F. (1985). Synaptic structural correlates of information storage in
mammalian nervous system. In C. W. Cotman (Ed.), Synaptic plasticity (pp. 335-374).
New York: Guilford Press.
Hebb, D. O. (1949). The organization of Behavior. New York: Wiley Press.
Hovda, D. A., & Feeney D. M. (1984). Amphetamine with experience promotes recovery of
locomotor function after unilateral frontal cortex injury in the cat. Brain Research, 298,
358.
John, E. R., Tang, Y., Brill, A. B., Young, R., & Ono, H. (1986). Double-labeled metabolic maps
of memory. Science, 233, 1167-1175.
Kalil, R. E., & Schneider, G. E. (1975). Abnormal synaptic connections of the optic tract in the
thalamus after midbrain lesions in newborn hamsters. Brain Research, 100, 690-698.
Kandel. E. R., & Schwartz, J. H. (1981). Principles of neuroscience. Amsterdam: Elsevier
North-Holland.
Kaplan, B. (1983). A trio of trials. In R. M. Lerner (Ed.), Developmental psychology: Histor-
ical and philosophical perspectives (pp. 185-228). Hillsdale, NJ: Erlbaum.
Kelche, C., & Will, B. (1982). Effects of postoperative environments following dorsal hippo-
campallesion on dendritic branching and spines in rat occipital cortex. Brain Research,
245, 107-115.
Kertesz, A. (1983). Localization in neuropsychology (p. 18). New York: Academic Press.
Kuhn, T. (1970). The structure of scientific revolutions. Chicago: University of Chicago
Press.
Kuhn, T. (1977). The essential tension. Chicago: University of Chicago Press.
Nieto-Sampedro, M. (1988). Growth factor induction and order of events in CNS repair. In D.
G. Stein & B. A. Sabel (Eds.)' Pharmacological approaches to the treatment of brain and
spinal cord injury (pp. 301-338). New York: Plenum Press.
Nottebohm, F. (1985). Neuronal replacement in adulthood. In F. Nottebohm (Ed.), Hope for a
new neurology (pp. 143-162). Annals of the New York Academy of Sciences, 457.
Raisman, G. (1978). What hope for repair of the brain? Annals of Neurology, 3, 101-106.
Scheff, S. W., Hoff, S. F., & Anderson, N. J. (1986). Altered regulation of lesion-induced
synaptogenesis by adrenalectomy and corticosterone in young adult rats. Experimental
Neurology, 93, 456-470.
252 DONALD G. STEIN
Schneider, G. (1979). Is it really better to have your brain damage early? A revision of the
"Kennard Principle." Neuropsychologia, 17, 557-583.
Schneider, G., & ]haveri, S. (1974). Neuroanatomical correlates of spared or altered function
after brain lesions in the newborn hamster. In D. G. Stein,].]. Rosen, & N. Butters (Eds.)'
Plasticity and recovery of function in the eNS (pp. 65-110). New York: Academic Press.
Slavin, M. D., Laurence, S., & Stein, D. G. (1988). Another look at vicariation. In S. Finger, T.
E. LeVere, C. R. Almli, & D. G. Stein (Eds.), Brain injury and recovery: Theoretical and
controversial issues (pp. 165-180). New York: Plenum Press.
Thompson, R. F. (1986). The neurobiology of learning and memory. Science, 223, 941-947.
Thompson, R. (1978). A behavioral atlas of the rat brain. New York: Oxford University
Press.
Valenstein, E. S., Cox, V. c., & Kakolewski,]. W. (1968). Modification of motivated behavior
elicited by electrical stimulation of the hypothalamus. Science, 158, 1118-1121.
Index
Acalculia, 177-178, 186, 189, 190 Anarithmia, 189-190; see also Acalculia
Accommodation, 244 Anarthria, 101
Accumbens (nucleus), 199 Anarthric retardation, 87, 100
Acetylcholine, 201, 220, 222, 223 Angular gyrus, 41, 147, 153, 189
Acoustic cue, 136 Anomia, 102, 112, 132; see also Aphasia
Adipsia, 245, 246 Anosognosia, 47, 170
Adrenalectomy, 206 Anterograde degeneration, 38
Adrenaline, 204 Anticipatory errors, 137
After-image, 11 Anticipatory transposition, 138
Agnosia, 7, 8, 56, 86, 115, 178, 189 Aphasia, 6-8, 17, 39-41, 45, 51, 56, 71,
auditive, 87 76, 85-86, 88, 102, 107-108, 123-124,
visual, 166 126-127, 134, 136-137, 142-143, 148,
Agraphesthesia, 181 239, 245
Agraphia, 112, 147-149, 173, 187, 189- anomic, 15, 45, 153, 158-162, 166, 169,
190 172, 189
aphasic, 148 Broca's, 46, 135, 147, 151-153, 159-164,
apraxic, 188 169, 170, 190, 191
deep, 149 conduction, 46, 135, 151, 153, 158-163,
phonological, 149 165, 169, 170, 172, 177
semantic, 149 congenital, 85
spatial, 171 cortical motor, 46
Agrammatism, 52, 160, 170, 172, 180, 181, developmental, 86
188, 191 global, 141, 153, 163-166, 169
Akinesia, 199, 202, 206 motor, 46, 87
Alexia, 73, 147-148, 190 semantic, 177-192
frontal, 147 transcortical motor, 44-45
literal, 169 transcortical sensory, 45-47, 54, 112,
spatial, 148 115
verbal, 169 Wernicke's, 44, 54, 112, 115, 135, 151,
with agraphia, 147 153, 158-163, 165, 169, 170, 172, 190
without agraphia, 147, 151-153, 157- Aphasic retardation, 87
163, 169-170 Apperceptive process, 5
Alpha rhythm, 43 Apractagnosia, 87
Alzheimer's disease, 32, 50, 54, 108, 214; Apraxia, 7, 56, 177-178, 181, 183, 187
see also Dementia constructional, 18,56, 151, 171, 178, 188
Amnesia, 56, 107-108 expression, 86, 162
Amphetamine, 204, 243 orofacial, 97
Amygdala, 199 verbal, 46
Analyzers, 91-92, 94, 98-100 Astereognosia, 181
253
254 INDEX
Atropine, 220-222 Dyslexia, 72, 76; see also Alexia

Auditory attention task, 68 Dysphasia, 86, 112; see also Aphasia
Auditory imperception, 101 Dysprosody, 180-181
congenital, 85
Autotopagnosia, 189 Echolalia, 96, 109, 111
Electroencephalogram (EEG), 43, 55-56,
Basal ganglia, 49, 198, 202, 212 67, 109
Bilingualism, 75 Embryogenesis, 89-91
Boston Diagnostic Aphasia Examination, Empty speech, 116
108, 150-151, 154, 157, 182, 184-185 Encephaline, 206
Boston Naming Test, 32 "End of dose" phenomenon, 202
Brain development, 10 Engram, 223
Broca's area, 41-42, 44-46, 246 Epilepsy, 40-42, 56, 88, 180-181
Epinephrine, 204, 206
Cascade phenomenon, 171 Estrogen, 233
Catecholaminergic system, 243 Event-related potentials (ERP), 43, 55-56,
Caudate nucleus, 50, 107, 201, 204-207, 61-62, 64, 66-77
212-213, 219, 221, 223 Evoked potentials: see Event-related
CBF technique, 48, 55-56 potentials
Cerebral blood flow, 42
Cerebrospinal fluid (CSF), 51, 206 Face perception, 27, 56
Circumlocutions, 169 Frontal lobe, 42, 56, 71, 151-153, 158-
Cognition, 3, 5, 7, 19, 30, 39, 48, 52, 197 160, 162, 166, 169-170, 172, 198-199,
Cognitive development, 12 208-209, 212-213, 237
Commissurotomy, 20, 23 operculum, 39, 48, 68
Complex loop, 198-199, 213 Functional recovery, 240
Consciousness, 9-10, 13
Constructional disability: see Gamma-aminobutyric acid (GAB A), 220
Constructional apraxia Garrett model, 124, 128, 130-133, 138
Contingent Negative Variation (CNV), 43 Gerstmann syndrome, 189-192
Corpus callosum, 41 Gestalt movement, 5
Cyclotron, 49 Globus pallidus, 201
Glossolalia, 131
Defense mechanism test, 12 Glycoprotein, 214
Delayed recovery, 239 Grapheme-phoneme correspondence, 71,
Delusions, 10 74, 148-149, 170, 173
Dementia, 7,9, 107-108, 197 Growth-promoting factor, 236
of Alzheimer type, 108, 110-118, 212;
see also Alzheimer's disease Hallucinations, 10, 211
multi-infarction, 108, 110, 113-114, Hematoencephalic barrier, 202-203
117-118 Hemi-inattention: see Spatial neglect
in Parkinson's disease, 108, 110, 112- Hemispherectomy, 41
118; see also Parkinson's disease Hemiasomatognosia, 189
Diaschisis, 54 Hippocampus, 199, 221, 238
Dichotic listening, 62 Hologenetic techniques, 11
Diencephalon, 90 Huntington's disease, 50, 214
Dopamine, 198-199, 202, 204, 207, 220 Hypothalamus, 245
Double tactile stimulation, 180
Down syndrome, 30-31 Ideographic reading, 162
Dream state, 10 Imperception, 18
Dysarthria, 86, 108-109, 111, 117, 208 Insula, 39, 47, 153, 163
Dysgraphia: see Agraphia Interior language, 96-97
Dyskinesia, 202, 207 Inversion recovery, 51
INDEX 255
Ischemic Scale Score, 110 Palilalia, 109, 111

Isopropyl amphetamine (IMP), 50 Paradigmatic cognitive system, 123
Isotope scanning, 44, 55-56 Paradigmatic relations, 132
Paragrammatism, 52, 170, 172
Jacksonian principle, 4, 7 Paragraphias, 172, 183
Paralexias, 148, 169
Paraphasias, 8, 52, 102, 109, 111, 125,
Leukodystrophies, 91
131-132, 134, 135, 139-141, 180, 187
Lexical blends, 133-134
Parietal lobe, 19, 29, 42-43, 46-47,152-
Lexical semantics, 69-70, 75
153, 166, 172, 178, 192, 220
Lexical substitutions, 131, 139
Parietal-temporal-occipital junction, 166,
Locus coeruleus, 198, 201
172, 189, 192
Logoclonia, 109, 111
Parkinson's disease, 197-214; see also
Luxury perfusion, 49
Dementia
Perceptgenesis, 8, 10, 12
Magnetic resonance imaging (MRI), 45, 50- Perceptual closure, 18
51, 55-56 Perceptual disorders: see Agnosia
Magnetoencephalographic studies, 12 Perseverative transpositions, 138
Malapropism, 133 Phonemic substitutions: see Paraphasias
Map of behavior, 230-234 Plasticity, 229-250
Maze learning, 28 Phonemic paraphasias: see Paraphasias
Mental chronometry, 65-66 Phonemic false evaluation, 137
Mesocortic limbic system, 214 Phonemic transposition, 137
Mesocortical dopaminergic system, 213 Phonological processes, 126, 127, 130, 148
Mesolimbic cortical pathway, 212 Phonotactic patterns, 140
Meta-contrast technique, 12 Positron emission tomography (PET), 42,
Meynert (nucleus basalis), 201 49-50, 55-56
Microgenesis, 3-13 Prefrontal cortex: see Frontal lobe
Mini-Mental State Examination, 108, 133- Premo tor cortex, 163-166
114 Progesterone, 233
Mooney's visual closure task, 28 Proprioception, 93-94, 97
Motor cortex, 42, 52, 163, 198; see also Prosody, 95
Frontal lobe Psychosis, 9
Motor impersistence, 47 Putamen, 50
Motor loop, 198-199
Multi-infarction dementia: see Dementia
Muscarine blockers, 220 Quadrilateral space, 39
Myelination, 91
Radioisotope scans, 45
Neologisms, 111-112, 130, 161-162, 172 Raphe nucleus, 201
Neologistic jargon, 46 Reaction time, 62, 71
Neural noise, 241 Reticular activating system, 52
Neuritic plaques, 116 Retrograde degeneration, 38, 242
Neurofibrillary tangles, 116 Retrograde neural marker technique, 242
Neuropeptides, 206 Rey-Osterrieth Complex Figure, 150-151,
Nigrostriatal system, 203, 212, 214, 226; 162-163, 170-171, 182, 185
see also Substantia nigra; Striatum rCBF measures, 12
Noradrenaline, 204-205 Right hemisphere, 18, 22-24, 27-29, 39,
41,47,52-54,65-66,69-70,72,74,
Occipital lobe, 29-30, 41, 147, 152-153, 76, 148; 155-156, 163, 166-169, 172,
166, 172, 189 189
"On-off" phenomenon, 202, 206-207 prerolandic area, 151, 155, 159-160,
Organa dynamic theory, 9 162, 167
256 INDEX
Right hemisphere (cant.) Supramarginal gyrus, 41, 163

retrorolandic area, 151, 157, 159-160, Survival-promoting factor, 236
162, 167, 170 Syllable contact law, 142
Syntactic processes, 126
Schizophrenia, 6, 11 Syntagmatic relations, 132
Schwab England Scale, 207
Scopolamine, 220-221 Tautosyllabic, 130
Semantic field, 132 Telencephalon, 90
Semantic system, 148 Temporal lobe, 28,41-44,68, 166
Semantic paraphasias: see Paraphasias Thalamus, 90, 202
Septum, 199, 201 Token test, 150-153, 157, 182, 185, 187-
Serotonin, 237 188
Single-photon computerized tomography
(SPECT), 50, 55-56 Unified Parkinson Rating Scale, 207
Slips of the tongue, 126-127, 129, 133,
135-137 Visual attention, 19-20
Somatostatin, 201, 206 Visuospatial perception, 17, 21, 27,43,48,
Spatial alternation task, 233 52
Spatial neglect, 56, 160-161, 171, 189 deficits, 17-20, 25, 27, 56
Spin echo, 51 Verbal humor, 20
Split-brain, 20-21 Verbal stereotypes, 94
Striatum, 198, 203-205, 214, 219-222
Stuttering, 109, 111, 114 Wearing-off phenomenon, 202
Subcortical syndromes, 87, 117-118, 166; Wernicke's area, 45, 246
see also Dementia Western Aphasia Battery, 108
Substantia ignominata, 201, 212 William syndrome, 30-31
Substantia nigra, 198, 203-204, 220, 223 Wurzburg group, 4-5
Superior colliculus, 240
Supplementary motor area, 44-45 Xenon inhalation technique, 49-50

Ardila - Ostrosky Solis - (Ed) - 1989 PDF

Uploaded by

Copyright:

Available Formats

Ardila - Ostrosky Solis - (Ed) - 1989 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ardila - Ostrosky Solis - (Ed) - 1989 PDF

Uploaded by

Copyright:

Available Formats

Brain Organization

Editorial Advisory Board:

ASSESSMENT ISSUES IN CHILD NEUROPSYCHOLOGY

BRAIN ORGANIZATION OF LANGUAGE AND COGNITIVE PROCESSES

HANDBOOK OF CLINICAL CHILD NEUROPSYCHOLOGY

MEDICAL NEUROPSYCHOLOGY: The Impact of Disease on Behavior

NEUROPSYCHOLOGICAL FUNCTION AND BRAIN IMAGING

NEUROPSYCHOLOGY, NEUROPSYCHIATRY, AND BEHAVIORAL

RELIABILITY AND VALIDITY IN NEUROPSYCHOLOGICAL

Plenum Press • New York and London

© 1989 Plenum Press, New York

ALFREDO ARDILA, Konrad Lorenz Foundation University, Bogota, Co-

ANDREW KERTESZ, University of Western Ontario, St. Joseph's Health

Neuropsychology has presented a particularly formidable array of devel-

scale study of a brain-damaged population with regard to alterations in

analysis and differential characteristics of dementia syndromes; the ap-

Part I: THEORETICAL AND TECHNICAL APPROACHES

Early Studies ................................................ 3

Historical Perspective ........................................ 18

Overview of ERP Methodology .. , .................... '" . ..... 61

Part II: DEVELOPMENT AND DISSOLUTION OF LANGUAGE

Organic and Functional Syndromes ........................... 88

Methodology... .. . . . . . . . . . . .... ... . . . ... . . . .. . . .. . . .. .. . . .. . 108

Part III: ORAL AND WRITTEN LANGUAGE DISORDERS

The Garrett Model ........................................... 124

Case Reports ................................................ 180

Part IV: BASAL GANGLIA AND COGNITIVE ACTIVITY

Pathology and Pathogenesis ",.".".,.,",.,",.",.,',."" 198

Involvement of the Neostriatum in Memory,. ,.,. , ........ , , ,., 219

Part V: RECOVERY FROM BRAIN DAMAGE

Plasticity and Levels of Analysis of CNS Functions. . . . . . . . . . . . . 234

Theoretical and Technical

Currently, there is renewed interest in process-oriented approaches to the

Approaches to the analysis of the microgenesis of thought predomi-

ROBERT E. HANLON· Department of Psychology, City College of the City University of

ries of conceptual stages that unfold, following the presentation of a

lage, during which there occurred only the mere apprehension of a

Until this time, process-oriented studies and theories were concerned

perimental techniques to demonstrate microgenetic phenomena. These

Building primarily on the findings of Sander, Conrad (1947) present-

behavior. Second, any human activity such as perceiving, thinking, act-

normal subjects with tachistoscopic methods. He briefly presented scenes

Ey elaborated three levels of disorganization of consciousness in psy-

tional process breaks down, the subjective and objective components of

Returning to experimental psychology and the classic microgenetic

there has been a modest renewal of micro genetic research (Draguns,

states. Two tachistoscopic procedures, the Meta-Contrast Technique

This is a brief overview of some clinical and experimental work from

De Bleser, R. (1987). From agrammatism to paragrammatism: German aphasiological tradi-

Schilder, P. (1953). Medical psychology. New York: International Universities Press.

Fractionation of Spatial Cognition

Visuospatial dysfunction denotes a constellation of perceptual and con-

into one category and referred to as "visuospatial dysfunction," "con-

Visuospatial difficulties in neurological patients were first noted by

Until recently, few neuropsychological studies utilized constructs

Visual attention, one aspect of visuospatial processing, has been the

disrupted in a unitary manner, a result that would have been of little

PARTS AND WHOLES

plicitly operationalizing these constructs. Furthermore, investigations