The Neurobiology of Language

:HAPTEF
The Neurobiology of Language

Sheila E. B l u m s t e i n
I. I N T R O D U C T I O N
One of the most challenging issues related to human language is under-
standing how it is organized and processed in the brain. Such study is
described under the broad rubric of the neurobiology of language and de-
fines the field ofneurolinguistics. The domain of inquiry has largely focused
on investigations of aphasic patients, exploring the clinical and behavioral
manifestations of their disorder and the accompanying lesion localization.
These "experiments in nature" provide the traditional approach to the study
of language/brain relations, and without question, they have provided, the
major insights to date. Nevertheless, in more recent years, the increased
sophistication of electrophysiological and metabolic techniques has pro-
vided a new and exciting approach to the study of the neurobiology of
language. These latter studies allow investigation of neural activity in nor-
mal subjects in addition to the study of brain-damaged patients. Taken
together, these three approaches: clinical/lesion investigations using imag-
ing techniques such as computerized axial tomography (CAT scan) and
magnetic resonance imaging (MRI); electrophysiological studies using
evoked related potentials (ERP); and metabolic techniques using positron
emission tomography (PET), provide converging data on the nature of the
representation of language, the processes involved in language use, and
ultimately its underlying neural organization.
It is the object of this chapter to review the current state of knowledge of
the neurobiology of language. Particular emphasis will be on studies of
Speech, Language, and Communication

Copyright 9 1995 by Academic Press, Inc. All rights of reproduction in any form reserved.
339
340 SheilaE. Blumstein
adult aphasia. Studies involving ERP and PET are also reviewed as they
speak to issues concerning language organization and language processing.
A. Research M e t h o d o l o g i e s
All experimental methodologies have both strengths and weaknesses. Clini-
cal/lesion investigations of adult patients provide a view of language break-
down subsequent to brain damage. As such, they serve as a window into the
fractionation of language in individuals who had normal language and cog-
nitive functions prior to the damage. Charting the relation between patterns
of language breakdown and underlying neuropathology provides the closest
analogy to lesion mapping in animal investigations. Unlike lesion mapping
studies in animals, however, it is impossible to control for the site and size
of the lesion. Grouping patients by lesion site then is quite difficult and
usually involves gross localization schemas making it difficult to determine
the precise relation between language behavior and underlying lesion. In
addition, it is problematic to assume a one-to-one relation between brain
structure and language function. As John Hughlings Jackson stated " . . . to
locate the damage which destroys speech and to localise [sic] speech are two
different things" (quoted in Henry Head, 1963, p. 50). It is possible that a
lesion may affect pathways leading to a particular functional area without
damaging that area. Moreover, the processing components ultimately con-
tributing to language are complex and are usually inferred from investiga-
tions of patients' behavior on different language tasks, such as naming and
comprehension, behaviors that in themselves may require different process-
ing components. Finally, studies with aphasic patients typically rely on
behavioral measures and aphasic patients typically show increased vari-
ability compared to neurologically intact control patients on these mea-
sures. As a result, it is often difficult to titrate out statistically reliable effects.
Nonetheless, studies with aphasic patients have provided the basis for the
major hypotheses generated and tested to date in metabolic studies with
neurologically intact subjects.
The advantages of metabolic studies such as PET are that they provide a
potential means of mapping brain function in neurologically intact subjects
and of investigating specific aspects of language without the complicating
factors of lesions, which potentially affect more than one component of
language. PET is a technique that examines changes in cerebral blood flow
in localized brain areas during cognitive activities. The time frame over
which blood flow is measured is about 40 s. The assumption underlying this
methodology is that the greater the involvement of a particular area of the
brain in a particular cognitive activity, the greater will be the metabolic
activity, hence cerebral blood flow, to that area. Nevertheless, this technique
has its own weaknesses and potential problems. Perhaps the biggest chal-
10 The Neurobiology of Language 341
lenge is determining whether and how much neuronal activity relates to a

functional activity. Typically, PET studies use complex statistical algo-
rithms for determining the significance of the metabolic activity obtained.
To determine the localization of particular functions, PET studies use the
subtraction technique. This technique assumes that particular cognitive and
language tasks or functions can be broken down into discrete components
or processes contributing to that function, and that these functions are
additive. To focus on the particular function of interest, the metabolic activ-
ity contributing to the other functions or processes is subtracted out, leaving
the isolated function and its proposed neurological site. For example, in
determining the metabolic activity in understanding the meaning of a word,
the metabolic activity related to auditory perception and the perception of
the sound-shape of the word is subtracted. The neuronal activity remaining
is assumed to reflect cerebral activity specifically related to the meaning of
the word. What is not clear is whether the processes of language can be
assumed to be additive in this simple way. Nevertheless, the PET technique
provides an exciting new way of exploring the neurobiology of language.
Finally, electrophysiological studies have recently provided some inter-
esting correlates of language function. To date, however, they do not pro-
vide information on localization since the exact origin of the electrical activ-
ity cannot be determined by current techniques. ERPs do provide a means
of exploring patterns of electrical potentials that correlate with various as-
pects of language including speech or phonological processing, word mean-
ing, and syntactic processing. The ERP methodology measures brain activ-
ity at the scalp in synchrony with the presentation of a stimulus. Variations
in electrical brain potentials are assumed to serve as electrophysiological
markers of the particular activity under investigation. As with PET studies,
analytic techniques must be developed to identify significant patterns from
the underlying random electrical activity of the brain. Identifying these
patterns from background noise is indeed a major challenge. Moreover, it is
unclear whether this technique is sensitive enough to measure and distin-
guish the more subtle aspects of language and its underlying processing
components.
B. T h e o r e t i c a l F r a m e w o r k
As a preliminary step in the discussion of the neurobiology of language, it
may be useful to consider some broad theoretical issues that have guided
research in this field. Much as linguistic research has been guided by consid-
erations of the structural levels of the linguistic grammar, such as phonol-
ogy, lexicon/semantics, and syntax, so have linguistic investigations of
aphasia (see Frazier, Chapter 1, this volume). The major question has been
whether language deficits in aphasia reflect selective impairments to these
342 Sheila E. Blumstein
components and to their representations, that is, the primitive units or

structural properties of the components, or alternatively, the processes in-
volved in accessing these components. Consideration of aphasia as a deficit
in the components of the grammar, or in the processes involved in accessing
the components, speaks directly to the issue of the modularity of language.
Throughout the history of the study of aphasia, various forms of mod-
ularity theories have been postulated (for a review, see Fodor, 1983; Head,
1963).
Broadly speaking, evidence clearly supports the view that language is
modular, in that it is an autonomous system that is (within limits) func-
tionally and neurologically distinct from other higher cortical functions.
Dissociations have been found in which language is impaired and cognitive
abilities are spared and vice versa (Irigary, 1967; Tissot, Lhermitte, & Du-
carne, 1963; Zangwill, 1964). Studies of brain-damaged patients also sug-
gest that mathematical, musical, and artistic abilities may be dissociated
from language (Gardner, 1983). Nevertheless, considerably stronger claims
have been made about the modularity of language. In particular, in some
versions of the modularity theory, it has been hypothesized that language
consists of submodules such as lexical, syntactic, and semantic, which are
functionally autonomous, with a restricted domain of analysis and process-
ing (Fodor, 1983; Garrett, 1979). Evidence from the neurobiology of lan-
guage may speak to the broad claims of the theory of modularity and to the
particular form that such a modular theory may take. Consideration of the
components of the grammar (phonology, lexicon/semantics, and syntax)
provides an organizational framework from which to explore the neuro-
biology of language.
Before reviewing the major components of the grammar in relation to
the neurobiology of language, it is worthwhile to briefly review some of the
major clinical and neuroanatomical manifestations underlying the aphasias.
For approximately 90% of right-handers, the left hemisphere is dominant
for language. Brain damage within the left hemisphere has different effects
on language and speech depending on the site and extent of the lesion.
Although there have been recent challenges to the reliability of classification
schemas in aphasia, and more important, concern about whether such sche-
mas provide the bases from which language impairments may be studied
(Caramazza, 1984; Schwartz, 1984), it goes without saying that a broad
array of language behaviors is impaired in particular ways as a function of
different areas of brain damage. The clinical syndromes that characterize
these aphasias are based on a constellation of language abilities and disabil-
ities defined in terms of language faculties or functions such as speaking,
understanding, reading, and writing (Goodglass & Kaplan, 1972).
Classical approaches to the clinical/neurological basis of language disor-
ders in adult aphasics have typically characterized the aphasia syndromes in
broad anatomical (anterior and posterior) and functional (expressive and

receptive) dichotomies (see Geschwind, 1965). The two aphasia syndromes
that best characterize the anterior/posterior and expressive/receptive di-
chotomy are, respectively, Broca's and Wernicke's aphasia. Patients with
Broca's aphasia show a profound expressive deficit in the face of relatively
good auditory language comprehension. Speech output is typically non-
fluent in that it is slow, labored, and often dysarthric, and the melody
pattern seems flat. Furthermore, speech output is often agrammatic, charac-
terized by the omission of grammatical words such as the and is, as well as
the incorrect usage of grammatical endings. Syntactic structures are gener-
ally limited to simple sentences with few embeddings and dependent
clauses. As to other language abilities, naming an object on presentation is
generally fair to good, and repetition of language is usually good or better
than spontaneous speech output. The underlying neuropathology includes
the frontal operculum as well as premotor and motor regions posterior and
superior to the frontal operculum, extending to the white matter below and
including the basal ganglia and insula (Damasio, 1991).
The syndrome of Wernicke's aphasia is characterized by both language
input and language output impairments. Auditory language comprehension
is severely impaired. Speech output is fluent and well articulated. Gram-
matical structures seem relatively intact, although syntactic phrases may be
inappropriately juxtaposed. Most characteristically, the semantic content of
the output is severely compromised. Language is empty of semantic con-
tent, often difficult to understand, with the overuse of high-frequency, low-
content words such as thing, be, have, this. In addition, literal paraphasias
(sound substitution errors) and verbal paraphasias (word substitution er-
rors) occur in speech output. There are some Wernicke's aphasia patients
who also produce neologisms or "jargon." Another frequent characteristic
of this disorder is "logorrhea" or 'press' for speech, whereby patients, even
in a conversational setting, talk on and on without stopping. Other associ-
ated language impairments include a moderate to severe naming deficit and a
severe repetition disorder. The neuropathology associated with Wernicke's
aphasia involves the posterior region of the left superior temporal gyrus,
often extending to the supramarginal and angular gyrus (Damasio, 1991).
II. THE S O U N D STRUCTURE OF LANGUAGE: P H O N O L O G Y

It is generally assumed that in the case of both language production and
language reception words or lexical entries ultimately contact a common
representation, specified in terms of segments, phonetic features, and rules
for their combination that are specific to the sound structure of language.
Nonetheless, to effect encoding and decoding of speech requires different
mechanisms relating to the speech apparatus in the former and to the audi-
tory system in the latter, and for that reason the interface of these two
systems with the lexicon for speech production and speech perception re-
quires a unique set of operations (see Fowler, Chapter 2, and Nygaard and
Pisoni, Chapter 3, this volume). For this reason, it is useful to consider the
neurobiology of speech production and speech perception separately.
A. Speech Production
Linguistic theory makes a clear distinction between phonology and phonet-
ics, and the facts of aphasia support such a distinction, particularly in the
investigation of speech production deficits. In some cases, patients may
produce a wrong sound segment, but its phonetic implementation is cor-
rect, that is, for teams the patient says keams. In other cases, patients may
produce the correct sound segments but their phonetic implementation is
distorted, that is, for teams the patient produces an i n i t i a l / t / t h a t is overly
aspirated.
Research investigations have focused on the patterns of speech produc-
tion errors of aphasic patients. A number of different methodologies have
been used to elicit productions from patients including conversational
speech, naming, and repetition. Despite the different methodologies used, a
consistent finding is that nearly all aphasic patients, regardless of the aphasia
syndrome, display impairments implicating a deficit at the phonological
level. Critically, the patterns of impairment are similar, suggesting that
despite the different underlying neuropathology, a common mechanism is
impaired. The similar patterns of performance are particularly striking giv-
en the different clinical characteristics and neuropathology of the patients
investigated. The mechanism impaired in the phonological output impair-
ment of all of these patients relates to the selection and/or organization of
the features comprising the candidate lexical entries. Selection impairments
reflect the inappropriate choice of a feature and hence the appearance of an
inappropriate segment in a particular position. Planning impairments reflect
the inappropriate ordering of the segments or alternatively the inappropriate
influence of phonetic context on the ultimate feature or segment selected.
The patterns of sound substitutions indicate that it is indeed the selection of
an inappropriate feature and not the selection of an inappropriate segment
that underlies these types of errors.
The types of errors produced by aphasic patients can be categorized in
terms of four descriptive categories. Together they suggest that the basis of
the errors reflects a selection as well as a planning deficit (Blumstein, 1973).
These errors include: phoneme substitution errors in which a segment is
substituted for another (e.g., teams ~ keams); simplification errors in which
a phoneme or syllable is deleted (e.g., green ~ geen); addition errors in
which an extra phoneme or syllable is added to a word (e.g., see ~ stee); and
environment errors in which the occurrence of a particular phoneme is

influenced by the surrounding phonetic context; these errors include meta-
thesis (e.g., degree--o gedree), and progressive and regressive assimilation
errors (e.g., Crete--o kreke, and ~ trete, respectively). The stability of these
patterns is evidenced by their occurrence across languages" French (Bouman
& Grunbaum, 1925; Lecours & Lhermitte, 1969); German (Bouman &
Grunbaum, 1925; Goldstein, 1948); English (Blumstein, 1973; Green,
1969); Turkish (Peuser & Fittschen, 1977); Russian (Luria, 1966); and Finn-
ish (Niemi, Koivuselka-Sallinen, & Hanninen, 1985).
The pattern of sound substitutions is consistent with the view that the
incorrect features have been selected or activated but have been correctly
implemented by the articulatory system. Thus, most substitution errors
involve a change in value of a single feature. For example, the production of
doy for to), reflects a change in the feature of voicing. Similarly, the produc-
tion of dut for nut represents a change in the feature of nasality. Relatively
few substitution errors produced by aphasic patients involve changes in
more than one feature.
Phonological errors also suggest that the nature of the syllable structure
(i.e., the organization of consonants and vowels) of the lexical candidate
constrains the type and extent of errors made during the selection process
(Blumstein, 1973, 1990). In particular, the occurrence of phoneme substitu-
tion errors is more likely to occur when the syllable contains a single conso-
nant than when it is part of a consonant cluster, for example, /f/ is more
likely to undergo a phoneme substitution error in the word feet than in the
word fleet. Simplication and addition errors are more likely to result in the
canonical syllable structure Consonant-Vowel, for example, consonants are
more likely to be deleted in a word beginning with two consonants, sky--o
ky, and consonants are more likely to be added in a word beginning with a
vowel, army--o jarmy. And finally, assimilation errors across word bound-
aries preserve the syllable structure relations of the lexical candidates. That
is, if the influencing phoneme is at the beginning of the target word, so is
the assimilated phoneme, for example, history books--o bistory books. If the
influencing phoneme is at the end of the target word, so is the assimilated
phoneme, for example, roast beef---~ roafbeef. These results show that infor-
mation about the syllable structure of a word is represented in the lexicon,
and this information is used in the planning buffer for sentence production.
If this were not the case, the syllable constraints shown in the assimilation
errors would not occur across word boundaries.
Despite the systematicity and regularity of the phonological errors just
described, the particular occurrence of such an error cannot be predicted.
That is, sometimes patients may make an error on a particular word, and
other times they will produce it correctly. Nor are the errors unidirectional
(Blumstein, 1973; Hatfield & Walton, 1975). A voiced stop consonant may
become voiceless (e.g., /d/ --o /t/), and a voiceless stop consonant may
become voiced ( e . g . , / t / ~ / d / ) . Thus, the patterns of errors reflect statisti-
cal tendencies. Because patients are able to produce the correct utterance and
because their errors are systematic (although not predictable), it suggests
that the underlying phonological representations are intact, but that either
the level of activation of the features has been changed or there has been a
failure to select the appropriate phonological representation from the lexi-
con. Even though virtually all aphasic patients regardless of lesion site dis-
play similar patterns of phonological impairment, they vary in the severity
of the impairment. Patients with Broca's aphasia and damage to anterior
brain structures produce many more such errors than do patients with
Wernicke's aphasia (Blumstein, 1973).
While the phonological patterns are similar across patients, phonetic defi-
cits seem to be more selective. A long-held observation is that patients with
anterior aphasia produce phonetic errors. The implied basis for these errors
is one of articulatory implementation; that is, the commands to the articula-
tors to encode the word are incorrect, poorly timed, and so forth. A num-
ber of studies have explored these phonetic patterns of speech by investigat-
ing the acoustic properties or the articulatory parameters underlying the
production of particular phonetic dimensions. The dimensions investigated
include voicing in stop consonants (e.g., /p t k / v s . /b d g/) and fricatives
(e.g., / f s / v s . /v z/), place of articulation in stop consonants (e.g., / b / v s .
/ d / v s . /g/) and fricatives (e.g., / f / v s . /s/), and the nasal and stop manner
of articulation (e.g., /n m/ vs. /d b/). Studies of speech production in
aphasia patients with damage to anterior brain structures (including patients
with Broca's aphasia) have shown that these patients have difficulty produc-
ing phonetic dimensions that require the timing of two independent articu-
lators. These findings have emerged in the analysis of two phonetic dimen-
sions, voicing and nasality. In the case of the feature voicing, the dimension
studied is voice-onset time, that is, the timing relation between the release
of a stop consonant and the onset of vocal cord vibration. The production of
nasal consonants also requires appropriate timing between two articulators,
in this case, the release of the closure in the oral cavity and the velum
opening.
Results of analyses of the production of the voicing and nasal phonetic
dimensions have shown that aphasic patients with anterior brain-damage
(including patients with Broca's aphasia) evidence significant deficits
(Blumstein, Cooper, Goodglass, Statlender, & Gottlieb, 1980; Blumstein,
Cooper, Zurif, & Caramazza, 1977; Freeman, Sands, & Harris, 1978; Gan-
dour & Dardarananda, 1984a, Itoh, Sasanuma, Hirose, Yoshioka, & Ushi-
jima, 1980, Itoh et al., 1982; Itoh, Sasanuma, & Ushijima, 1979; Shewan,
Leeper, & Booth, 1984). These same patterns emerge across different lan-
guages. They occur not only in English and Japanese, for which voice-onset
time serves to distinguish two categories of voicing, voiced and voiceless,

but also in Thai, for which voice-onset time serves to distinguish three
categories of voicing in stop consonants, prevoiced, voiced, and voiceless
aspirated.
Nevertheless, these patients also show normal patterns of production.
The constellation of spared and impaired patterns of articulation suggests
that their disorder affects impairments, in particular articulatory maneuvers
rather than the articulatory implementation of given phonetic features. The
evidence comes from studies of voicing in stop consonants. In English, the
feature voicing in stop consonants can be cued in several ways. As discussed
earlier, voice-onset time provides one measure of voicing for stop conso-
nants occurring in initial position. A second measure is the duration of the
vowel preceding a stop consonant. Vowels are short before voiceless stops
(e.g., write) and long before voiced stops (e.g. ride). If patients have a deficit
related to the implementation of the feature voicing, then they should dis-
play impairments in the production of vowel length preceding voiced and
voiceless stop consonants as well as voice-onset time. In contrast, if they
have a deficit related to particular articulatory maneuvers, such as the timing
of two independent articulators, the production of voice-onset time may be
impaired, while the production of vowel length may be normal. Results
indicate that although these patients show an impairment in the implemen-
tation of the voicing phonetic dimension via voice-onset time, they are able
to maintain the distinction between voiced and voiceless stops on the basis
of the duration of the preceding vowel (Baum, Blumstein, Naeser, & Pal-
umbo, 1990; Dully & Gawle, 1984). Moreover, there is no systematic rela-
tion within the same patient between the ability to realize the voicing di-
mension by means of voice-onset time and vowel duration (Tuller, 1984).
Thus, these patients do not have a disorder affecting the articulatory pro-
duction of the feature voicing, but a disorder affecting particular articula-
tory maneuvers, namely the timing or integration of movements of two
independent articulators.
Although patients with anterior aphasia show a disorder in temporal
coordination, their disorder does not reflect a pervasive timing impairment.
Fricative durations do not differ significantly from those of nonaphasic
patients (Harmes et al., 1984), and the patients maintain the intrinsic dura-
tion differences characteristic of fricatives varying in place of articulation;
for example, /s / and/~ / are longer in duration than / f / a n d / 0/ (Baum et
al., 1990). English-speaking patients with anterior aphasia maintain differ-
ences in the intrinsic durations of vowels; for example, tense vowels such as
/i/ and /e/ are longer than their lax vowel equivalents, /I/ and /E/. In
addition, Thai-speaking patients with anterior aphasia maintain the contrast
between short vowels and long vowels, a contrast that is phonemic in this
language (Gandour & Dardarananda, 1984b).
In addition to the impairment in the timing of independent articulators,

difficulties for aphasic patients with anterior brain-damage (including
Broca's aphasia) have also emerged with laryngeal control. They have
shown impairments in voicing in the production of voiced fricatives (Baum
et al., 1990; Harmes et al., 1984; Kent & Rosenbek, 1983), and impairments
in the production of intonation (Cooper, Soares, Nicol, Michelow, &
Goloskie, 1984; Ryalls, 1982).
Finally, recent investigations of coarticulation effects in patients with
anterior brain-damage show that they produce relatively normal anticipa-
tory coarticulation. For example, in producing the syllable/su/, they antic-
ipate the rounded v o w e l / u / i n the production of the preceding/s/ (Katz,
1988). Nevertheless, they seem to show a delay in the time it takes to
produce these effects (Ziegler & v o n Cramon, 1985, 1986), and they may
show some deficiencies in their productions (Tuller & Story, 1986; but see
Katz, 1987, for discussion). What these results suggest is that phonological
planning is relatively intact, but it is the ultimate timing or coordination of
the implementation of the articulatory movements that is impaired (cf. Kent
& Rosenbek, 1983).
Of interest is the point that, although clearly distinguished from patients
with anterior aphasia, posterior aphasia patients also display a subtle pho-
netic deficit. Most typically, they show increased variability in the imple-
mentation of a number of phonetic parameters (Kent & McNeill, 1987;
Ryalls, 1986), including vowel formant frequencies (Ryalls, 1986) and vowel
duration (Gandour et al., 1992; Ryalls, 1986; Tuller, 1984). Because these
phonetic impairments are not clinically perceptible but emerge only upon
acoustic analysis, they are considered to be subclinical. Nevertheless, they
indicate that both anterior and posterior brain structures ultimately contrib-
ute to the speech production process.
On the basis of the findings just reviewed, several conclusions can be
made concerning the nature of the phonetic disorders for patients with
anterior aphasia and their ultimate underlying mechanisms. In particular,
the impairment is not a linguistic one, in the sense that the patient is unable
to implement a particular phonetic feature. Moreover, the patients have not
lost the representation for implementation nor the knowledge base for how
to implement sounds in context. Rather, particular maneuvers relating to
the timing of articulators seem to be impaired, ultimately affecting the
phonetic realization of some sound segments and some aspects of speech
prosody.
Investigations exploring the potential neuroanatomical structures con-
tributing to the phonetic implementation of speech suggest that there are
specific neuroanatomical substrates relating to such phonetic implementa-
tion patterns. These investigations include neuroimaging (CAT scan) stud-
ies of aphasic patients, attempting to relate the behavioral characteristics
outlined above with particular areas of brain damage, and PET and ERP
data with neurologically intact subjects. In particular, CAT scan correla-
tions with patterns of speech production deficits suggest the involvement of
Broca's area, lower motor cortex regions for larynx, tongue, and face, and
some white matter structures as well in the phonetic implementation of
speech (Baum et al., 1990). PET studies with neurologically intact subjects
determining regions of cerebral blood flow activity during speech produc-
tion also show the importance of these areas as well as the precentral gyrus
and premotor areas surrounding Broca's area (Petersen, Fox, Posner, Mint-
un, & Raichle, 1989). It is interesting that ERP studies with these subjects
suggest that these areas seem to be involved in the production of articula-
tory maneuvers related specifically to the production of speech sounds and
not more generally to vocal tract maneuvers that do not relate to speech
(McAdam & Whitaker, 1971).
B. Speech Perception
Similar to production studies with aphasic patients, most studies exploring
the role of speech perception deficits in auditory comprehension impair-
ments have focused on the ability of aphasic patients to perceive phonemic
or segmental contrasts. Studies on segmental perception have indeed shown
that aphasic patients evidence deficits in processing segmental contrasts.
These studies have explored patients' abilities to discriminate pairs of words
or nonwords (e.g. pear vs. bear, pa vs. ba) or they have asked subjects to
point to the appropriate word or consonant from an array of names or
syllables that are phonologically confusable. Results show that patients with
both anterior and posterior aphasia display some problems in discriminating
phonological contrasts (Blumstein, Baker, & Goodglass, 1977; Jauhiainen &
Nuutila, 1977; Miceli, Calatgirone, Gainotti, & Payer-Rego, 1978; Miceli,
Gainotti, Caltagirone, & Masullo, 1980) or in labeling or identifying conso-
nants presented in a consonant-vowel context (Basso, Casati, & Vignolo,
1977; Blumstein, Cooper et al., 1977; cf. also Boatman, Lesser, & Gordon,
1994). These problems emerge in the perception of both real words and
nonsense syllables. Although there are more errors in the perception of
nonsense syllables than in the perception of real words, the overall patterns
of performance are similar, and essentially mirror the patterns found in the
analysis of phonological errors in speech production. Namely, subjects are
more likely to make speech perception errors when the test stimuli contrast
by a single feature than when they contrast by two or more features (Baker,
Blumstein, & Goodglass, 1981; Blumstein, Baker, & Goodglass, 1977; Mic-
eli et al., 1978; Sasanuma, Tatsumi, & Fujisaki, 1976). Among the various
types of feature contrasts, the perception of place of articulation is partic-
ularly vulnerable (Baker et al., 1981; Blumstein, Baker, & Goodglass, 1977;
Miceli et al., 1978). It is interesting that similar patterns emerge in non-

brain-damaged subjects when perceiving speech under difficult listening
conditions (cf. Miller & Nicely, 1955).
What is not clear from many of the studies exploring the perception of
segmental contrasts is whether the failure to perceive such contrasts reflects
an impairment in the perception of abstract phonetic features or alter-
natively an impairment relating to the extraction of the acoustic patterns
associated with these features. To explore this issue, several studies have
investigated categorical perception of the acoustic parameters associated
with phonetic features. Categorical perception relates to the way neuro-
logically intact subjects perceive the phonetic categories of speech. In partic-
ular, continuous changes in an acoustic parameter such as voice-onset time
associated with a phonetic contrast, in this case voicing, give rise to discon-
tinuous changes in perception; when asked to label or identify the stimuli on
an acoustic continuum, listeners perceive them as belonging to discrete
categories corresponding to the end-point, exemplar stimuli, and they show
a sharp change in the identification of the categories usually at a particular
stimulus along the continuum; when asked to discriminate the stimuli, they
accurately discriminate only those stimuli that they labeled as belonging to
two different categories, and fail to discriminate those stimuli that they
labeled as the same phonetic category, even though all of the discrimination
pairs vary along the same physical dimension. The studies exploring cate-
gorical perception in aphasia have investigated two phonetic dimensions,
voicing (Basso et al., 1977; Blumstein, Cooper, et al., 1977; Gandour &
Dardarananda, 1982) and place of articulation in stop consonants (Blums-
tein, Tartter, Nigro, & Statlender, 1984). For voicing, the acoustic dimen-
sion varied was voice-onset time, and for place of articulation, the dimen-
sion varied was the frequency of the formant transitions appropriate f o r / b d
g/, and the presence or absence of a burst preceding the transitions. Results
showed that if aphasic patients could perform only one of the two tasks
(labeling or discrimination), it was the discrimination task. Most important,
the discrimination functions were generally similar in shape and the locus of
the phonetic boundary was comparable to that of neurologically intact sub-
jects, even in those patients who could only discriminate the stimuli.
The fact that no perceptual shifts in boundary location were obtained for
the discrimination and labeling functions for aphasic patients, that the dis-
crimination functions remained stable even in those patients who could not
label the stimuli, and that the patients perceived the acoustic dimensions
relating to phonetic categories in a fashion similar to neurologically intact
subjects suggest that aphasic patients do not have a deficit specific to the
extraction of the spectral or temporal patterns corresponding to the phonet-
ic categories of speech. Rather, their deficit seems to relate to the activation
of the phonetic/phonological representation itself or to its ultimate contact
with the lexicon. Consistent with this view is the finding that although
patients may show speech perception impairments, their performance is
variable; they do not show selective impairments relating to a particular
phonetic feature; and the pattern of errors is bidirectional, for example,
voiced consonants may be perceived as voiceless, and voiceless consonants
may be perceived as voiced. This pattern of results mirrors that found in
speech production studies with aphasic patients.
In contrast to the segmental features of speech, the prosodic cues (i.e.,
intonation and stress) are consistently less affected in aphasia. Severely im-
paired aphasic patients have been shown to retain some ability to recognize
and distinguish the syntactic forms of commands, yes-no questions, and
information questions when marked only by intonation cues (Green &
Boller, 1974), even when they are unable to do so when syntactic forms are
marked by lexical and syntactic cues. Nonetheless, as with intonation cues,
patients' performance is not completely normal. A number of studies have
revealed impairments in the comprehension of lexical/phrasal stress con-
trasts (e.g., h6tdog vs. hotd6g) (Baum, Kelsch Daniloff, & Daniloff, 1982;
Emmorey, 1987), as well as sentential contrasts (e.g., he fed her d6g biscuits
vs. he fed her dog biscuits) (Baum et al., 1982). Similar findings emerged for
the perception of tone contrasts serving as lexical cues in Thai (Gandour &
Daradarananda, 1983) and Chinese (Naeser &Chan, 1980). No differences
have emerged in any studies between the performance of patients with
anterior and posterior aphasia, an important finding consistent with the
results for the perception of phonemic contrasts.
That speech perception deficits emerge in patients with both anterior and
posterior aphasia suggests that the neural basis for speech perception and the
auditory processing of speech is neurally complex, and includes far greater
neural involvement than simply the primary auditory areas and auditory
association areas in the temporal lobe. While the number of neuro-
physiological and electrophysiological studies focusing particularly on
speech reception are few, these results provide converging evidence consis-
tent with this view. PET studies have shown that the primary auditory
cortex is activated in the processing of simple auditory stimuli (Lauter,
Herscovitch, Formby, & Raichle, 1985), and both the primary auditory
cortex and the superior temporal gyrus are activated in passive word recog-
nition (Petersen, Fox, Posner, Mintun, & Raichle, 1988, 1989). These re-
sults are of no surprise. What is of interest, however, is that other cortical
areas seem to be activated as well. For example, Zatorre, Evans, Meyer, and
Gjedde (1992) showed increased activity in Broca's area near the junction
with premotor cortex as well as the superior parietal area near the supra-
marginal gyrus when subjects were required to make a phonetic decision or
phonetic judgment about auditorily presented CVC stimuli. Further,
Ojemann (1983) has shown impairments in the ability of patients to identify
auditorily presented sound segments embedded in a phonetic context

/a__ma/ during electrical stimulation to a wide range of cortical areas in-
cluding the inferior frontal cortex.
C. Speech Perception and Lexical Access

While results of speech perception experiments suggest that aphasic patients
do not seem to have a selective impairment in processing the segmental
structure underlying the auditory properties of speech, several recent stud-
ies have suggested that aphasic patients display impairments in the intersec-
tion of the sound properties of speech with lexical access. These studies have
shown some interesting dichotomies in the performance of patients with
Broca's and Wernicke's aphasia, which suggest that patients' impairments
do not reflect speech perception deficits per se but rather deficits in the
intersection of sound structure with the lexicon (cf. also Baker et al., 1981;
Martin, Wasserman, Gilden, & West, 1975).
Milberg, Blumstein, and Dworetzky (1988) explored the extent to which
phonological distortions affect semantic facilitation in a lexical decision
task. They investigated whether a phonologically distorted prime word
such as gat or war would affect the amount of semantic facilitation for target
words semantically related to the undistorted prime, for example, cat. Sub-
jects were asked to make a lexical decision on the second item of stimulus
pairs in which the first stimulus was semantically related to the second (e.g.,
cat-dog), or alternatively, it was systematically changed by one or more
features (e.g., gat-dog, war-dog). Neurologically intact control subjects
showed a monotonic function with less semantic facilitation as phonetic
distortion increased, that is, less semantic facilitation for war-dog than for
gat-dog. In contrast, Wernicke's aphasia patients showed priming in all pho-
nologically distorted conditions (e.g., gat-dog, wat-dog), suggesting a re-
duced threshold compared to the control subjects for lexical access, and
patients with anterior aphasia showed priming only in the undistorted se-
mantically related condition (e.g., cat-dog), suggesting an increased thresh-
old for lexical access. These results suggest that impairments in the use of
phonological information to access the lexicon can manifest themselves in
different ways in aphasic patients in the absence of a deficit in processing the
phonological properties of speech themselves.
Similarly, the lexical status of a word affects differentially how aphasic
patients perform phonetic categorization. Neurologically intact subjects
typically show a lexical effect. For example, the locus of a phonetic boundary
on an acoustic continuum varying in voice-onset time changes as a function
of the lexical status of the end-point stimuli. When the end-point / d / s t i m u -
lus is a word (e.g., dash), and the end-point/t/stimulus is a nonword (e.g.,
rash), there are more d responses along the continuum; in contrast, when the
e n d - p o i n t / t / s t i m u l u s is a word (e.g., task), and the e n d - p o i n t / d / s t i m u l u s

is a nonword (e.g., dask), there are more t responses along the continuum
(Ganong, 1980). Patients with Broca's aphasia show a larger than normal
lexical effect, placing a greater reliance on the lexical status of the stimulus
in making their phonetic decisions than on the perceptual information in the
stimulus. In contrast, patients with Wernicke's aphasia do not show a lexical
effect at all, suggesting that top-down information does not influence pho-
netic categorization, and perhaps such top-down processing may even fail to
guide their language performance (Blumstein, Burton, Baum, Waldstein, &
Katz, 1994).
III. T H E L E X I C O N
One of the most c o m m o n clinical attributes in aphasia is a naming deficit.
Such a deficit is operationally defined as an inability or inconsistency in
giving the appropriate name for an object, the depiction of an action, and so
on. This deficit may also manifest itself in word-finding difficulty (i.e.,
groping for the appropriate word), semantic paraphasias (i.e., word substi-
tutions), or empty speech (i.e., the use ofnonspecific words, as thing, be) in
the patient's spontaneous speech output. A number of attributes contribute
to the evocation of a name and can influence a subject's overall performance,
including word frequency (Howes, 1964; Rochford & Williams, 1962) and
its concreteness or picturability (Goodglass, Hyde, & Blumstein, 1969). In
fact, some patients have been identified with selective impairments in nam-
ing particular semantic categories or classes of words (cf. Hart, Berndt, &
Caramazza, 1985). Nonetheless, most aphasic patients do not have such
selective deficits, but rather have a naming deficit cutting across semantic
categories and word classes. The question remains as to what is the basis for
these nonspecific and pervasive naming deficits. Two possibilities have been
proposed. One possibility is that the patients have a lexical access impair-
ment. In this case, the underlying lexical representation and organizational
properties are intact, but the processes required to access this information
are impaired. A second possibility is that the patients have an impairment in
the lexical representation or semantic structure of the words themselves.
Whether the deficit is one of access or representation, such lexical impair-
ments would not only influence naming but could also affect the compre-
hension of language (see Cutler, Chapter 4, and Seidenberg, Chapter 5, this
volume).
There is a fair amount of clinical evidence suggesting that the basis of
naming impairments reflects a retrieval or access problem. A patient who is
unable to name a particular object in one context or at one time may be able
to do so at another. This inconsistency in object naming suggests that the
representation of the lexical item is intact, but the patient may be unable to
evoke its name. In support of this interpretation is the fact that the patient
who has failed to name or has misnamed an object may recognize the
appropriate name of the object when it is offered by the examiner. Further,
having failed to name an object, patients can often say what it is used for
(e.g., for cup they may say drink). In addition, they may be able to provide
the generic properties of the word such as the letter it begins with and the
number of syllables it contains (Goodglass, Kaplan, Weintraub, & Acker-
man, 1976). Presumably, the representation of the word must be in the
internal lexicon in order for the patient to be able to recall these properties.
Further insight into naming deficits can be provided by analyzing the
types of misnaming errors a patient may make. When patients cannot name
a particular stimulus, they may often substitute a word in the same semantic
field or produce a word that is a semantic associate to the target (Rinnert &
Whitaker, 1973) (e.g., for chair, they may say sofa; for mother, father). Nev-
ertheless, although consistent with an access deficit, these results could also
reflect a representation impairment. That is, if patients were operating with
a restricted semantic field or a modified conceptual framework for lexical
items, they could well access the appropriate semantic category without
having a full sense of the word and its appropriate name. Semantically
associated verbal paraphasias would then be a natural consequence of such a
deficit.
Consistent with the view that patients' deficit reflects a representation
impairment are results from a study by Zurif, Caramazza, Myerson, and
Galvin (1974) who investigated patients' judgments about semantic related-
ness. In particular, they asked subjects to group the two words of three that
best went together. The words (mother, father, partner, knight, husband,
shark, trout, dog, tiger, turtle, crocodile)varied along several semantic
dimensions (human-nonhuman, gender, functional attribute). Results
showed that neurologically intact control subjects grouped the words along
these semantic dimensions; patients with Broca's aphasia did so, too, with
only one minor exception. In contrast, patients with Wernicke's aphasia
showed no specific pattern--they were as likely to group mother with father
as mother with trout. These results were used as evidence supporting the view
that patients with Wernicke's aphasia have an impairment in the underlying
organization of the lexicon, hence a deficit in lexical representation (cf.
Goodglass & Baker, 1976).
More recently, however, a series of studies have been conducted that
challenge this view. These studies have explored on-line lexical processing
by exploring semantic priming in a lexical decision task. Using an adapta-
tion of a procedure developed by Meyer and Schvaneveldt (1971), the ex-
perimenters asked aphasic patients to make a lexical decision about words
and nonwords presented either visually (Milberg & Blumstein, 1981) or
auditorily (Blumstein, Milberg, & Shrier, 1982). The target word to be
judged (e.g., dog) was preceded by a prime that "was semantically related
(e.g., cat), semantically unrelated (e.g., table), or neutral that is, a nonword
(e.g., glub) with respect to the prime. In both studies, patients with Wer-
nicke's aphasia showed, as do control subjects, semantic facilitation, that is,
they responded faster to a real-word target preceded by a semantically
related prime than to one preceded by a semantically unrelated or nonword
prime. Surprisingly, patients with Broca's aphasia did not show this seman-
tic facilitation as consistently as did patients with Wernicke's aphasia. When
the prime-target pairs included highly associated words, patients with
Broca's aphasia showed evidence of priming (Blumstein et al., 1982). When
the semantic relationships were more subtle, that is, ambiguous words
(Milberg, Blumstein, & Dworetzky, 1987), or when the lexical decision
target word was not temporally paired with the preceding prime (i.e., the
subject was required to make a lexical decision response to each stimulus
presented) (Milberg & Blumstein, 1981), they did not as a group show
evidence of semantic priming. In contrast, when the same patients were
asked to make judgments about the semantic relatedness of the word pairs,
patients with Wernicke's aphasia performed randomly, and patients with
Broca's aphasia performed well.
To account for the dichotomy between the performance of patients with
Broca's and Wernicke's aphasia on these tasks, it was proposed that lexical
deficits in aphasia are better characterized in terms of the processing opera-
tions required for lexical access, rather than on the integrity of the stored
lexical knowledge base. In particular, it has been suggested that there are
two distinct processes involved in lexical access, one called automatic process-
ing that is fast acting, of short duration, not under voluntary control, and
thus largely automatic, and a second called controlled processing that is slower
acting, under a subject's voluntary control, and thus influenced by a sub-
ject's expectancies or attentional demands (Posner & Snyder, 1975; Shif-
frin & Schneider, 1977). Subjects with Wernicke's aphasia show consistent
evidence of semantic facilitation in the lexical decision tasks, but do not
appear to be able to analyze word meaning in a simple semantic judgment
task. Therefore, they seem to be able to automatically access words, and yet
are unable to use the knowledge about semantic features or semantic rela-
tions in making decisions about the meanings of words. For subjects with
Broca's aphasia, results are less clear-cut. The lack of consistent priming
effects suggests that the automatic activation of lexical candidates may be
impaired. However, these patients appear to have little difficulty analyzing
word meaning using controlled processing. In fact, recent debate in the
literature has centered on whether the impairment of subjects with Broca's
aphasia reflects slowed access to lexical representations (Hagoort, 1993), or
alternatively, a diminished activation of word candidates (Milberg, Blums-
tein, Katz, & Gershberg, 1995). Whatever the correct interpretation, it is
worth noting that in both cases the explanation turns on the processing
routines used for lexical access.
Similar results emerged in studying the time course of lexical access in
sentences. Swinney, Zurif, and Nicol (1989) explored the extent to which
aphasic patients use sentence context in accessing the meanings of ambigu-
ous words. In a cross-modal lexical decision task in which the sentence is
presented auditorily and the lexical target is presented visually, neuro-
logically intact subjects show access to all meanings of an ambiguous word
at its initial presentation even with a preceding biasing context toward one
interpretation. Only further downstream is the meaning appropriate to the
context uniquely retained. For example, in the sentence, the man saw several
spiders, roaches, and other bugs (1) in the corner (2) of the room, subjects show
semantic facilitation to all meanings of the word bugs (i.e., ant and spy) at
time (1), but at time (2) show semantic facilitation only to ant (Onifer &
Swinney, 1981). Swinney and his colleagues (1989) explored semantic facili-
tation in aphasic patients investigating whether they showed semantic facili-
tation to both meanings of an ambiguous word at time (1) [they did not
investigate semantic facilitation downstream at time (2)]. They showed that
similar to neurologically intact subjects, Wernicke's aphasia patients showed
priming for both meanings of the ambiguous words at time (1), irrespective
of the biasing context. In contrast, patients with Broca's aphasia showed
semantic facilitation only to the most frequent meaning of the ambiguous
word regardless of the biasing context. Thus, while patients with Broca's
aphasia clearly show an impairment in accessing word meanings from the
lexicon, similar to control subjects and Wernicke's aphasia patients, such
access is not affected by the sentence context in which the word appears.
Evidence consistent with the observation that the emergence of word
meaning is influenced by contextual factors and is not all-or-none comes
from ERP studies, beginning with the seminal work of Kutas and Hillyard
(1980). In a series of studies, they showed that while there are no ERP
"signatures" tied to either a specific word such as dog or a semantic category
such as fruit, there is a marker, the N400, which is sensitive to semantic
anomaly and semantic expectations (e.g., I take my coffee with cream and dog).
The N400 is a time-locked component that occurs between 350 and 800 ms
after stimulus presentation. The N400 emerges whether the anomaly occurs
in sentence-final or sentence-medial position, and whether the anomaly is in
a sentence or in a more natural prose context (Kutas & Hillyard, 1980,
1983).
O f critical importance, the N400 "marker" relates specifically to seman-
tics; it does not emerge for syntactic anomalies (i. e., tense or number errors)
nor for physical changes to the stimulus (i.e., unexpected differences such as
change of letter size) (Kutas & Hillyard, 1983). Moreover, lexical processing
is dependent on the activation level of candidate lexical items. In particular,
ERP studies have shown that the amplitude of the N400 varies as a function
of the likelihood that a word will appear in a context. The N400 signifi-
cantly decreases if an anomalous word is semantically related to an appro-
priate completion; for example, in the 2ame was called off because it started to
umbrella, umbrella is semantically related to the word rain, the natural com-
pletion of the sentence, and thus shows a significantly smaller N400 than for
coffee, which bears no semantic relationships between coffee and the natural
completion of the sentence. These results suggest that lexical access is
graded, changing as a function of the level of activation of a word candidate
(Kutas & Hillyard, 1984) and the extent to which the word candidate acti-
vates a semantic network (Kutas, Lindamood, & Hillyard, 1984). These
priming effects also emerge for single words; the N400 is larger to weakly
primed versus strongly primed lexical associates and is larger to unrelated
than to related words (Kutas & Hillyard, 1988).
PET studies also provide important converging evidence concerning the
processing components involved in lexical access (Petersen, et al., 1988,
1989; Posner, Petersen, Fox, & Raichle, 1988). These studies have shown
the involvement of anterior brain structures in the semantic processing
of single words in non-brain-damaged subjects, a result consistent with
the impairments shown in lexical priming for Broca's aphasia subjects de-
scribed above. They also showed the role of different brain areas in the
perception of visually and auditorily presented words, even though each
share access to a common output (articulatory) and meaning (semantic)
system. Thus, the PET studies show that the performance of cognitive tasks
requires the integration of a large number of component computations and
in this sense such tasks seem not to be performed in a single area of the
brain.
Taken together, the results of studies with aphasic patients and with
neurologically intact subjects using both ERP and PET suggest that the
neural basis of lexical processing is indeed complex. Both anterior and
posterior brain structures seem to be involved in lexical access, although to
date it is not clear what the contributions of these various structures are to
lexical access and lexical processing. What does seem to be the case is that
there are neurophysiological markers related to access of meaning. More-
over, lexical impairments seem to reflect deficits in lexical access and not
lexical representation. A point of interest is that the processing components
contributing to lexical access may be affected differently as a function of the
type of aphasia and presumably as a function of the different neural struc-
tures involved.
IV. SYNTAX
Nearly all aphasic patients display syntactic impairments in their speech
output. Nevertheless, the hallmark of syntactic production deficits is char-
acterized by patients with Broca's aphasia, who clinically are often de-
scribed as agrammatic or telegrammaticin their speech production. The speech

output of these patients is typically nonfluent and syntactically sparse in
terms of both the array of grammatical structures produced and the absence
of grammatical endings such as plurals, tense markers, and free-standing
grammatical words such as the and have. One critical issue is determining
the basis of this agrammatic deficit. Three hypotheses have been proposed.
The first suggests that the deficit reflects an economizing effort on the part
of the patient to compensate for expressive difficulties and effortful speech
production (Lenneberg, 1973). The second proposes that the grammatical
deficit is secondary to a phonological impairment related to phonological
stress (Kean, 1977). The third considers the deficit to reflect a fundamental
impairment in syntax related to either the representation or processing of
syntactic structures (Caplan, Baker, & Dehaut, 1985; Grodzinsky, 1990; and
see Bock, Chapter 6, and Tanenhaus and Trueswell, Chapter 7, this vol-
ume). Let us explore the evidence that bears on this issue.
Analysis of data from agrammatic patients across different languages
provides a first window into the possible basis of the syntactic deficit.
Languages of the world implement syntactic structure in quite different
ways; English depends highly on word order to implement syntactic func-
tions such as subject or object of the verb (e.g., John kissed Mary vs. Mary
kissed John); other languages are highly inflected, such as Russian or Ger-
man, and use inflectional endings to mark the functions of subject and direct
object; languages such as Japanese are agglutinative and mark such syntactic
functions by using particles appended to word stems; still others such as
Hebrew or Arabic are infixing languages and mark syntactic structures by
changes on vowels within the word stem itself. It is interesting that agram-
matic aphasia patients display syntactic deficits regardless of the language
they speak and the way the language instantiates syntax (Menn & Obler,
1990). Thus, all of these patients display impairments with syntactic struc-
tures whether they are realized in terms of word order, inflectional endings,
particles, or infixes. These results suggest that the agrammatic deficit may
more likely reflect either a compensatory production mechanism or alter-
natively a syntactic deficit. The explanation that the deficit is a phonological
one is less likely because of the very different phonetic/phonological mani-
festations of the syntactic properties relating to phonological stress across
these different languages.
Perhaps more important, these results suggest that there are particular
neural structures that are dedicated to the instantiation of a linguistic func-
tion, namely syntax, and these neural structures lie in Broca's area and the
surrounding cortical and subcortical neural structures. Thus, syntactic defi-
cits emerge with damage to similar neural structures irrespective of the
different ways syntax may be realized in language, whether by word order,
inflection, agglutination, or infixing. In this sense, syntax is marked neuro-
logically. Additional support for this view comes from the language pat-
terns of congenitally deaf signers who sustain brain damage to these same
neural structures. These patients also display agrammatic deficits, in this
case relating to the syntactic properties of sign (Corina, Vaid, & Bellugi,
1992). In American Sign Language (ASL), syntax is marked by certain
hand-shape and movement configurations. The only common attribute be-
tween hand-shape and movement configurations in ASL and word order,
infixing, and agglutination in aural/oral language is the syntactic function
these attributes play in language.
Further analysis of productive agrammatism also reveals that it is charac-
terized by the substitution of syntactic particles or endings for each other
rather than the "loss" of these endings. For example, in English the singular
form for a word is characterized phonologically by a zero marker [i.e., boy
+ 0 (singular)], whereas the plural is determined by the addition of the
plural f o r m / s / t o the word stem [i.e., boy +/s/(plural)]; thus, for English
it is impossible to ascertain whether the production of the singular boy for
the plural boys reflects the loss of the plural form or alternatively the substitu-
tion of the singular (zero) morphological ending for the plural. However, in
other languages such as Russian, German, Japanese, Hebrew, and Arabic,
the stem itself does not appear alone and all words must be marked with a
morphological ending. Thus, it is possible to determine whether the agram-
matic patient's production reflects the loss or substitution of a morphologi-
cal marker. Results show that in these languages, agrammatic patients al-
ways use a syntactic marker, albeit from time to time, the wrong marker
(Menn & Obler, 1990); thus, the syntactic markers are not lost but rather are
substituted for each other. These results lend credence to the view that the
basis of the agrammatic deficit is primarily a syntactic one and not an
economizing effort by the patient to reduce the amount of language output.
Additional evidence supports the view that agrammatism is primarily a
syntactic deficit. In addition to the deficit with morphology described
above, agrammatic patients also display a reduction in the variety of syntac-
tic structures that they use, as well as a failure to concatenate phrases that
constitute a sentence. Thus, patients have great difficulty in juxtaposing the
noun phrase (NP) and verb phrase (VP) of a sentence, while they may still
be able to maintain the internal cohesion of the NP itself by using adjectives
and the VP itself by concatenating the verb with a following object. More-
over, the occurrence of more complex sentences and phrases containing
embeddings and compound nouns and verbs is rare. The failure to produce
these syntactic structures, however, does not seem to reflect a loss of the
"knowledge" of these structures. Goodglass, Gleason, Bernholtz, and Hyde
(1972) have shown that patients may use alternative strategies to convey a
meaning that they seem to be unable to express using a given syntactic
structure. For example, given a structured story completion test designed to
3(~ SheilaE. Blumstein
elicit a broad array of syntactic structures, patients may use reduplication to

connote very (i. e., hot hot for very hot), use an adverbial to connote verb tense
(i.e., yesterday he go for he went), or use descriptive adjectives instead of
comparatives (i.e., Mary weak, Bob strong for Bob is stronger than Mary)
(Gleason, Goodglass, Green, Ackerman, & Hyde, 1975; Goodglass et al.,
1972).
Perhaps one of the strongest pieces of evidence suggesting that agramma-
tic deficits reflect an underlying syntactic impairment comes from a study
by Zurif, Caramazza, and Myerson (1972) who explored whether agram-
matic patients could use syntactic structures properly if given the oppor-
tunity and time. They provided the first evidence that patients with Broca's
aphasia have a syntactic comprehension impairment as well as a syntactic
production deficit. Specifically, they asked patients to judge how words in a
written sentence "go best together." Subjects were required to cluster or put
together the two words of an array of three taken from a presented sentence.
Neurologically intact subjects group words very much as linguists, organiz-
ing them in terms of the syntactic relations that they hold in the sentence;
thus, given the sentence, the dog chases a cat, a control subject clusters the
determiner the with the noun dog, and clusters the verb chases with the direct
object a cat. In contrast, patients with agrammatic aphasia do not seem to
know what to do with the "little words." They are as likely to cluster the
with a or chases as they are with dog. Thus, they show not only a failure to
produce and use the grammatical words in speech production, but they also
fail to use them appropriately when asked to identify the relationships
among the words in a sentence.
Further studies have supported this observation that agrammatic patients
have difficulty in using syntactic structures in language comprehension as
well as production. In particular, these studies have examined the process-
ing and comprehension of grammatical words and syntactic structures by
patients with Broca's aphasia. One study explored the processing of the
definite and indefinite articles a and the in English. If neurologically intact
are presented with a square and two circles and are required to press either a
square or round button, they show a shorter reaction time to the command,
"press a round one" than to the command "press the round one," presum-
n
ably because reference to one of the two circles requires the use of an
indefinite article. In contrast, agrammatic aphasia patients show no differ-
ences in reaction time to the two commands "press a round one" and "press
the round one," suggesting that they do not appropriately process the
meaning difference between the two articles (Goodenough, Zurif, &
Weintraub, 1977). In other research (Caramazza & Zurif, 1976), patients
with Broca's aphasia showed comprehension deficits in understanding
subject-object relations in embedded sentences in a picture-pointing task. It
is interesting that these deficits emerged only when the test sentences were
stripped of other potential semantic cues that could be used by the subjects.
That is, their deficit emerged in the context of semantically reversible sen-
tences such as the clown chases the midget, and not when the meaning of the
sentence was constrained by semantic and real-world probabilities such as
the man patted the dog.
The fact that most agrammatic aphasia patients display a syntactic com-
prehension deficit as well as a syntactic production deficit (but cf. Miceli,
Mazzucchi, Menn, & Goodglass, 1983) suggests that a single impairment,
namely, a central syntactic impairment, underlies their language difficulty.
The deficit is considered "central" because it affects syntax independent of
its language modality~speaking, comprehending, reading, and writing.
As to the specific nature of this syntactic deficit, recent research has sug-
gested that the impairment is more likely a deficit in syntactic processing
than it is a deficit in syntactic representation. Several lines of research con-
tribute to this view (although cf. Caplan et al., 1985, and Caplan & Futter,
1986, Grodzinsky, 1984, 1986, for alternative points of view). Patients who
are agrammatic in both production and comprehension, nonetheless are
able to make grammaticality judgments (Linebarger, Schwartz, & Saffran,
1983). Thus, these patients are able to judge the grammaticality of sentences
that violate different types of syntactic structures, for example, *he shut the
window that the door was open (gapless relatives), *the girls laughed the clown
(subcategorization), and *my father knew they would give the job to (empty
elements). This relatively spared ability to make syntactic judgments in the
face of grammatical comprehension and production deficits for the same
type of syntactic structures suggests that syntactic representations are intact.
The comprehension and production impairments probably reflect the fact
that these patients have a processing deficit affecting the "mapping" of
syntactic structures onto their appropriate semantic representations.
Consistent with this view are the results of several on-line syntactic
processing experiments. These studies have shown that agrammatic aphasia
patients do not show a sensitivity to morphological constraints that non-
brain-damaged control subjects show. In particular, unlike control subjects,
they fail to show faster reaction times in a lexical decision task to verbs
when preceded by a grammatically appropriate modal (e.g., is-going) than
when preceded by a grammatically inappropriate modal (e.g., could-going)
(Blumstein, Milberg, Dworetzky, Rosen, & Gershberg, 1991). Nonethe-
less, these same patients show an ability to make correct syntactic judg-
ments of these same verb phrases in an off-line grammaticality judgment
task. Similarly, Friederici and Kilborn (1989) showed a delay in accessing
syntactic information specified by morphological endings in subjects with
Broca's aphasia compared to syntactic information provided by individual
lexical items.
That the syntactic processing impairment extends beyond the processing
of morphological constraints has also been shown. In particular, Baum

(1988) showed that subjects with Broca's aphasia did not show sensitivity to
the types of syntactic violations explored by Linebarger et al. in an on-line
processing lexical decision task even though they were still able to make
correct grammaticality judgments (cf. also Tyler, 1985). An interesting
point is that the processing impairment does not seem to affect all of syntac-
tic processing. Some syntactic processes seem to be normal, in particular,
the processing of filler-gap constructions. In particular, it has been shown in
neurologically intact subjects that processing and ultimately the compre-
hension of syntactically complex sentences such as relative clauses require
that the syntactic processor identify the underlying syntactic structure and
syntactic relations of the words in the sentence. In the case of relative clause
sentences such as this is the girl that .John hit, girl has been moved from its
underlying syntactic position as object of the verb hit, leaving a gap or
empty element after the verb hit. To process the sentence and to determine
the correct syntactic relations, the listener must identify the gap, and the
filler, in this case girl, associated with it, and assign it to its correct syntactic
position (as object of the verb). For example, in the sentence the car that the
old lady was selling ~ needed many repairs, the car has been moved from its
position as object of the verb leaving a gap (_) following selling. If subjects
are reactivating the filler car at the gap site, then they should show semantic
facilitation for the filler car even if the semantically related word is presented
in the position of the gap site (after selling), some eight syllables down-
stream from the actual occurrence of car. Both neurologically intact subjects
(Tanenhaus, Carlson, & Seidenberg, 1985) and subjects with Broca's aph-
asia (Blumstein, Byma, Hourihan, & Brown, in preparation) show evidence
of reactivation of the filler at the gap site. Thus, impairment in compre-
hending Wh-constructions is more likely due to an access deficit than an
impairment in the syntactic representations themselves.
While the major focus on syntactic deficits in aphasia has been on agram-
matism in Broca's aphasia, grammatical deficits in both production and
comprehension are also common in Wernicke's aphasia. Investigations of
the speech production of patients with Wernicke's aphasia indicate that these
patients often show paragrammatic speech exemplified by the juxtaposition
of inappropriate grammatical phrases (e.g., I saw him to the boy). Moreover,
they also show a restricted range of syntactic structure in their speech output
(Gleason et al., 1980). With respect to auditory language comprehension,
patients with Wernicke's aphasia also show syntactic impairments, and in
fact, show similar patterns of performance to that of patients with Broca's
aphasia in their comprehension of a number of critical dimensions including
morphological endings, syntactic structures, syntactic complexity, and real-
world plausibility (Caplan et al., 1985; Goodglass, 1968; Goodglass &
Menn, 1985; Lesser, 1978; Parisi and Pizzamiglio, 1970).
What remains unclear is whether these similar patterns of performance

between patients with Broca's and Wernicke's aphasia reflect different un-
derlying impairments. Several researchers (Caplan et al., 1985; Goodglass &
Menn, 1985) have suggested that the syntactic deficit of Wernicke's aphasia
is similar to that of Broca's aphasia. Nevertheless, these conclusions may be
premature. Several researchers have shown dissociations in patterns of per-
formance between patients with Broca's and Wernicke's aphasia (Blum-
stein, Goodglass, Statlender, & Biber, 1983; Goodglass et al., 1979; von
Stockert, 1972; Zurif & Caramazza, 1976), and more recent on-line experi-
ments suggest that the nature of the processing impairments between these
two groups may indeed by different (cf. Blumstein et al., 1991). Nonethe-
less, on-line syntactic processing research in aphasia is in its infancy and
much more research will be needed before the nature of these processing
deficits can be elaborated.
That there are particular processing operations related to filler-gap con-
structions is supported by recent ERP investigations. Garnsey, Tanenhaus,
and Chapman (1989) used the emergence of an N400 for semantic anomalies
to explore syntactic processing of filler-gap constructions. They showed
that the filler in Wh-constructions is attached at the first possible gap site
immediately after the verb. Thus, syntactic processing operations also leave
electrophysiological markers, although they are not the same as those
shown in lexical/semantic p.rocessing (Rosleer, Putz, Friederici, & Hahne,
1993).
The exploration of such syntactic markers has just begun. There is some
preliminary evidence to suggest that there may be different electro-
physiological patterns to types of syntactic anomalies relating to violations
of phrase structure, wh-extraction, and subjacency rules (Neville, Nicol,
Barss, Forster, & Garrett, 1991). Whether there is a particular marker for
syntactic anomalies as was shown in the N400 for semantic anomalies is less
clear. A potential component termed P600 has recently been identified,
which emerges in the processing of certain types of syntactic anomalies
(Osterhut & Holcomb, 1994). Clearly, more work needs to be done, but
this research direction shows great potential for providing important in-
sights into the processing mechanisms underlying syntax.
In sum, results from studies with aphasic patients suggest that syntactic
representations are intact and that these patients display deficits in accessing
these structures. Although most attention has been payed to the syntactic
deficit accompanying agrammatism in patients with Broca's aphasia, pa-
tients with Wernicke's aphasia also display syntactic production and com-
prehension deficits. With respect to agrammatic aphasia patients, syntactic
output disorders emerge with damage to anterior brain structures and these
deficits occur irrespective of the grammatical structure of the particular
language. Nonetheless, for both Broca's and Wernicke's aphasia patients,
syntactic impairments typically affect language production as well as lan-

guage comprehension. In the case of language production, patients often
produce agrammatic or paragrammatic speech. In the case of language com-
prehension, the impairment seems to reflect a processing deficit affecting
the ability to map syntactic structures onto semantic representations, pre-
sumably contributing to the language comprehension deficit of these pa-
tients. Consistent with this view are the results of recent ERP studies show-
ing electrophysiological markers for syntactic processing and perhaps also
for particular syntactic structures.
V. SUMMARY
Investigation of the neurobiology of language provides support for the view
that the organization of the language system is indeed complex. All aphasic
patients display phonological, lexical, and syntactic deficits regardless of
aphasia type and lesion localization, suggesting that the neural instantiation
of these linguistic components requires complex neural structures that are
represented throughout the left hemisphere. Consideration of the neuro-
biology of the sound structure of language and, especially, the studies of
aphasic patients suggest that the organization and ultimate implementation
of the phonological sound structure is broadly represented in the left domi-
nant hemisphere and deficits relate to selection and planning in speech pro-
duction and access in speech perception. In contrast, as shown by studies of
patients with aphasia and results of PET and ERP investigations, the pho-
netic implementation of speech seems to be subserved by specific neural
substrates, and phonetic deficits reflect impairments in particular articula-
tory movements rather than in underlying linguistic features. With respect
to the neurobiology of the lexicon and syntax, ERP and PET studies pro-
vide support for electrophysiological markers for lexical meaning and syn-
tactic structures, and, in conjunction with data from aphasic patients, sug-
gest that with the possible exception of productive agrammatism, the neural
basis of the lexicon and syntax is broadly represented. Studies of both
lexical and syntactic impairments show that these deficits emerge in both
production and comprehension, and they reflect impairments to the pro-
cessing mechanisms accessing lexical and/or syntactic structures rather than
deficits to the representations or structures themselves.
Although this chapter has used a modular approach to language as a
theoretical framework by studying separately phonology, the lexicon, and
syntax, the results of this research suggest that there is little evidence to
support the view that the organization of the language system itself is
modular. In particular, language impairments in aphasia reflect processing
disorders that cut across the linguistic components of the grammar includ-
ing phonology, the lexicon, and syntax. Language impairments in aphasia
are n o t selective w i t h respect t o a particular c o m p o n e n t o f the g r a m m a r n o r

do they reflect i m p a i r m e n t s to particular linguistic representations. M o r e
specifically they reflect i m p a i r m e n t s to those processing routines used to
access this i n f o r m a t i o n . This v i e w is consistent w i t h clinical and e x p e r i m e n -
tal findings s h o w i n g similar patterns o f linguistic i m p a i r m e n t across differ-
ent aphasia groups, and also accounts for the fact that nearly all aphasic
patients s h o w s o m e degree o f deficit in phonological, lexical, and syntactic
processing, a l t h o u g h the constellation o f abilities and disabilities that these
patients display is nevertheless quite different.
F u t u r e w o r k requires greater integration o f clinical and e x p e r i m e n t a l
findings w i t h results f r o m electrophysiological, metabolic, and n e u r o i m a g -
ing studies w i t h b o t h neurologically intact and b r a i n - d a m a g e d patients.
T o g e t h e r , they will p r o v i d e a w i n d o w into one o f the m o s t challenging
areas in cognitive n e u r o s c i e n c e - - t h e u n d e r s t a n d i n g o f the n e u r o b i o l o g y o f
h u m a n language.
Acknowledgment
This research was supported in part by NIH Grant DC00314 to Brown University.
References
Baker, E., Blumstein, S. E., & Goodglass, H. (1981). Interaction between phonological and
semantic factors in auditory comprehension. Neuropsychologia, 19, 1-16.
Basso, A., Casati, G., & Vignolo, L. A. (1977). Phonemic identification defects in aphasia.
Cortex, 13, 84-95.
Baum, S. R. (1988). Syntactic processing in agrammatism: Evidence from lexical decision and
grammaticality judgement tasks. Aphasiology, 2, 117-135.
Baum, S. R., Blumstein, S. E., Naeser, M. A., & Palumbo, C. L. (1990). Temporal dimen-
sions of consonant and vowel production: An acoustic and CT scan analysis of aphasic
speech. Brain and Language, 39, 33-56.
Baum, S. R., Kelsch Daniloff, J., & Daniloff, R. (1982). Sentence comprehension by Broca's
aphasics: Effects of some suprasegmental variables. Brain and Language, 17, 261-271.
Blumstein, S. E. (1973). A phonological investigation of aphasic speech. The Hague: Mouton.
Blumstein, S. E. (1990). Phonological deficits in aphasia: Theoretical perspectives. In A. Caramazza
(Ed.), Cognitive neuropsychology and neurolinguistics: Advances in models of cognitivefunction and
impairment. Hillsdale: Erlbaum.
Blumstein, S. E., Baker, E., & Goodglass, H. (1977). Phonological factors in auditory com-
prehension in aphasia. Neuropsychologia, 15, 19-30.
Blumstein, S. E., Burton, M., Baum, S., Waldstein, R., & Katz, D. (1994). The role oflexical
status on the phonetic categorization of speech in aphasia. Brain and Language, 46, 181-
197.
Blumstein, S. E., Byma, G., Hourihan, J., & Brown, T. (1994). On-line processing of filler-
gap constructions in aphasia, manuscript in preparation, Brown University.
Blumstein, S. E., Cooper, W. E., Goodglass, H., Statlender, S., & Gottlieb, J. (1980). Produc-
tion deficits in aphasia: A voice-onset time analysis. Brain and Language, 9, 153-170.
Blumstein, S. E., Cooper, W. E., Zurif, E. B., & Caramazza, A. (1977). The perception and
production of voice-onset time in aphasia. Neuropsychologia, 15, 371-383.
Blumstein, S. E., Goodglass, H., Statlender, S., & Biber, C. (1983). Comprehension strategies
determining reference in aphasia: A study ofreflexivization. Brain and Language, 18, 115-
127.
Blumstein, S. E., Milberg, W. P., Dworetzky, B., Rosen, A., & Gershberg, F. (1991). Syntac-
tic priming effects in aphasia: An investigation of local syntactic dependencies. Brain and
Language, 40, 393-421.
Blumstein, S., Milberg, W., & Shrier, R. (1982). Semantic processing in aphasia: Evidence
from an auditory lexical decision task. Brain and Language, 17, 301-315.
Blumstein, S. E., Tartter, V. C., Nigro, G., & Statlender, S. (1984). Acoustic cues for the
perception of place of articulation in aphasia. Brain and Language, 22, 128-149.
Boatman, D., Lesser, P. R., & Gordon, B. (1994). The functional organization of auditory
speech processing as revealed by direct cortical electrical interference. Journal of the Acousti-
cal Society of America, in press.
Bouman, L., & Grunbaum, A. (1925). Experimentell-psychologische Untersuchungen sur
Aphasie und Paraphasie. Zeitschri[ifur die Gesamte Neurologie und Psychiatrie, 96, 481-538.
Caplan, D., Baker, C., & Dehaut, F. (1985). Syntactic determinants of sentence comprehen-
sion in aphasia. Cognition, 21, 117-175.
Caplan, D., & Futter, C. (1986). Assignment of thematic roles to nouns in sentence compre-
hension in an agrammatic patient. Brain and Language, 27, 117-134.
Caramazza, A. (1984). The logic of neuropsychological research and the problem of patient
classification in aphasia. Brain and Language, 21, 9-20.
Caramazza, A., & Zurif, E. B. (1976). Dissociation of algorithmic and heuristic processes in
language comprehension: Evidence from aphasia. Brain and Language, 3, 572-582.
Cooper, W. E., Soares, C., Nicol, J., Michelow, D., & Goloskie, S. (1984). Clausal intonation
after unilateral brain damage. Language & Speech, 27, 17-24.
Corina, D. P., Vaid, J., & Bellugi, U. (1992). The linguistic basis of left hemisphere specializa-
tion. Science, 255, 1258.
Damasio, H. (1991). Neuroanatomical correlates of the aphasia. In M. T. Sarno (Ed.), Acquired
aphasia (2nd ed.). San Diego, CA: Academic Press.
Duffy, J., & Gawle, C. (1984). Apraxic speakers' vowel duration in consonant-vowel-
consonant syllables. In J. Rosenbek, M. McNeil, & A. Aronson (Eds.), Apraxia of speech.
San Diego, CA: College-Hill Press.
Emmorey, K. D. (1987). The neurological substrates for prosodic aspects of speech. Brain and
Language, 30, 305-320.
Fodor, J. A. (1983). The modularity of mind. Cambridge, MA: MIT Press.
Freeman, F.J., Sands, E. S., & Harris, K. S. (1978). Temporal coordination of phonation and
articulation in a case of verbal apraxia: A voice-onset time analysis. Brain and Language, 6,
106-111.
Friederici, A. D., & Kilborn, K. (1989). Temporal constraints on language processing: Syntac-
tic priming in Broca's aphasia. Journal of Cognitive Neuroscience, 1, 262-272.
Gandour, J., & Dardarananda, R. (1982). Voice onset time in aphasia: Thai. I. Perception. Brain
and Language, 17, 24-33.
Gandour, J., & Dardarananda, R. (1983). Identification of tonal contrasts in Thai aphasic
patients. Brain and Language, 17, 24-33.
Gandour, J., & Dardarananda, R. (1984a). Voice-onset time in aphasia: Thai. II: Production.
Brain and Language, 18, 389-410.
Gandour, J., & Dardarananda, R. (1984b). Prosodic disturbances in aphasia: Vowel length in
Thai. Brain and Language, 23, 177-205.
Gandour, J., Ponglorpisit, S., Khunadorn, F., Dechongkit, S., Boongird, P., & Boonklam, R.
(1992). Timing characteristics of speech after brain damage: Vowel length in Thai. Brain
10 The N e u r o b i o l o g y of Language 367
Ganong, W. F. (1980). Phonetic categorization in auditory word perception. Journal of Experi-

mental Psychology: Human Perception and Performance, 6, 110-125.
Gardner, H. (1983). Frames of mind: The theory of multiple intelligences. New York: Basic Books.
Garnsey, S. M., Tanenhaus, M., & Chapman, R. (1989). Evoked potentials and the study of
sentence comprehension. Journal of Psycholinguistic Research, 18, 51-60.
Garrett, M. (1979). Levels of processing in sentence production. In B. Butterworth (Ed.),
Language production. London: Academic Press.
Geschwind, N. (1965). Disconnexion syndromes in animals and man. Brain, 88, 237-294,
585-644.
Gleason, J. B., Goodglass, H., Green, E., Ackerman, N., & Hyde, M. (1975). The retrieval of
syntax in Broca's aphasia. Brain and Language, 2, 451-471.
Gleason, J. B., Goodglass, H., Obler, L., Green, E., Hyde, M., & Weintraub, S. (1980).
Narrative strategies of aphasic and normal speaking subjects. Journal of Speech and Hearing
Research, 23, 370-383.
Goldstein, K. (1948). Language and language disturbances. New York: Grune & Stratton.
Goodenough, C., Zurif, E., & Weintraub, S. (1977). Aphasics' attention to grammatical
morphemes. Language & Speech, 20, 11-19.
Goodglass, H. (1968). Studies on the grammar of aphasics. In S. Rosenberg & J. H. Koplin
(Eds.), Developments in applied psycholinguistic research. New York: Macmillan.
Goodglass, H., & Baker, E. (1976). Semantic field, naming, and auditory comprehension in
aphasia. Brain and Language, 3, 359-374.
Goodglass, H., Blumstein, S. E., Gleason, J. B., Hyde, M. R., Green, E., & Statlender, S.
(1979). The effect of syntactic encoding on sentence comprehension in aphasia. Brain and
Language, 7, 201-209.
Goodglass, H., Gleason, J., Bernholtz, N. A., & Hyde, M. R. (1972). Some linguistic struc-
tures in the speech of a Broca's aphasic. Cortex, 8, 191-212.
Goodglass, H., Hyde, M. R., & Blumstein, S. (1969). Frequency, picturability and the avail-
ability of nouns in aphasia. Cortex, 5, 104-119.
Goodglass, H., & Kaplan, E. (1972). The assessment of aphasia and related disorders. Philadelphia:
Lea & Febiger.
Goodglass, H., Kaplan, E., Weintraub, S., & Ackerman, N. (1976). The 'tip-of-the-tongue'
phenomenon in aphasia. Cortex, 12, 145-153.
Goodglass, H., & Menn, L. (1985). Is agrammatism a unitary phenomenon? In M. L. Kean
(Ed.), Agrammatism. New York: Academic Press.
Green, E. (1969). Phonological and grammatical aspects of jargon in an aphasic patient: A case
study. Language & Speech, 12, 103-118.
Green, E., & Boller, F. (1974). Features of auditory comprehension in severely impaired
aphasics. Cortex, 10, 133-145.
Grodzinsky, Y. (1984). The syntactic characterization of agrammatism. Cognition, 16, 99-120.
Grodzinsky, Y. (1986). Language deficits and the theory of syntax. Brain and Language, 27,
135-159.
Grodzinsky, Y. (1990). Theoreticalperspectives on language deficits. Cambridge, MA: MIT Press.
Hagoort, P. (1993). Impairments oflexical-semantic processing in aphasia: Evidence from the
processing of lexical ambiguities. Brain and Language, 45, 189-232.
Harmes, S., Daniloff, R., Hoffman, P., Lewis, J., Kramer, M., & Absher, R. (1984). Tempo-
ral and articulatory control of fricative articulation by speakers with Broca's aphasia.
Journal of Phonetics, 12, 367-385.
Hart, J., Berndt, R. S., & Caramazza, A. (1985). Category-specific naming deficit following
cerebral infraction. Nature (London), 316, 439-440.
Hatfield, F. M., & Walton, K. (1975). Phonological patterns in a case of aphasia. Language &
Speech, 18, 341-357.
Head, H. (1963). Aphasia and kindred disorders of speech (Vol. 1). New York: Hafner.
Howes, D. (1964). Application of the word frequency concept to aphasia. In A. V. S. de Reuck
& M. O'Connor (Eds.), Disorders of language. London: Churchill.
Irigary, L. (1967). Approche psycholinguistique de langage des dements. Neuropsychologia, 5,
25-32.
Itoh, M., Sasanuma, S., Hirose, H., Yoshioka, H., & Ushijima, T. (1980). Abnormal articula-
tory dynamics in a patient with apraxia of speech. Brain and Language, 11, 66-75.
Itoh, M., Sasanuma, S., Tatsumi, I., Murakami, S., Fukusako, Y., & Suzuki, T. (1982). Voice
onset time characteristics in apraxia of speech. Brain and Language, 17, 193-210.
Itoh, M., Sasanuma, S., & Ushijima, T. (1979). Velar movements during speech in a patient
with apraxia of speech. Brain and Language, 7, 227-239.
Jauhiainen, T., & Nuutila, A. (1977). Auditory perception of speech and speech sounds in
recent and recovered aphasia. Brain and Language, 4, 572-579.
Katz, W. F. (1987). Anticipatory labial and lingual coarticulation in aphasia. In J. Ryalls (Ed.),
Phonetic approaches in speech production in aphasia and related disorders. Boston: College-Hill
Press.
Katz, W. F. (1988). Anticipatory coarticulatory in aphasia: Acoustic and perceptual data. Brain
Kean, M. L. (1977). The linguistic interpretation of aphasia syndromes: Agrammatism in
Broca's aphasia. Cognition, 5, 9-46.
Kent, R., & McNeill, M. (1987). Relative timing of sentence repetition in apraxia of speech and
conduction aphasia. In J. Ryalls (Ed.), Phonetic approaches to speechproduction in aphasia and
related disorders. Boston: College-Hill Press.
Kent, R., & Rosenbek, J. (1983). Acoustic patterns of apraxia of speech. Journal of Speech and
Hearing Research, 26, 231-248.
Kutas, M., & Hillyard, S. A. (1980). Reading senseless sentences: Brain potentials reflect
semantic incongruity. Science, 207, 203-205.
Kutas, M., & Hillyard, S. A. (1983). Event-related potentials to grammatical errors and
semantic anomalies. Memory & Cognition, 11, 539-550.
Kutas, M., & Hillyard, S. A. (1984). Brain potentials during reading reflect word expectancy
and semantic association. Nature (London), 307, 161-163.
Kutas, M., & Hillyard, S. (1988). An electrophysiological probe of incidental semantic associa-
tion. Journal of Cognitive Neuroscience, 1, 38-49.
Kutas, M., Lindamood, T., & Hillyard, S. (1984). Word expectancy and event-related brain
potentials during sentence processing. In S. Kornblum & J. Requin (Eds.), Preparatory
states and processes. Hillsdale, NJ: Erlbaum.
Lauter, J., Herscovitch, P., Formby, C., & Raichle, M. E. (1985). Tonotopic organization in
human auditory cortex revealed by positron emission tomography. Hearing Research, 20,
199-205.
Lecours, A. R., & Lhermitte, F. (1969). Phonemic paraphasias: Linguistic structures and
tentative hypotheses. Cortex, 5, 193-228.
Lenneberg, E. (1973). The neurology of language. Daedalus, 102, 115-133.
Lesser, R. (1978). Linguistic investigations of aphasia. London: Arnold.
Linebarger, M., Schwartz, M., & Saffran, E. (1983). Sensitivity to grammatical structure in
so-called agrammatic aphasics. Cognition, 13, 361-392.
Luria, A. R. (1966). Higher corticalfunctions in man. New York: Basic Books.
Martin, A. D., Wasserman, N. H., Gilden, L., & West, J. (1975). A process model of repetition
in aphasia: An investigation of phonological and morphological interactions in aphasic
error performance. Brain and Language, 2, 434-450.
McAdam, D. W., & Whitaker, H. A. (1971). Language production: Electroencephalographic
localization in the normal human brain. Science, 172, 499-502.
10 The N e u r o b i o l o g y of Language 369
Menn, L., & Obler, L. (1990). Agrammatic aphasia (Vols. 1-3). Amsterdam: John Benjamins.
Meyer, D. E., & Schvaneveldt, R. W. (1971). Facilitation in recognizing pairs of words;
Evidence of a dependence between retrieval operations. Journal of Experimental Psychology,
90, 227-234.
Miceli, G., Caltagirone, C., Gainotti, G., & Payer-Rigo, P. (1978). Discrimination of voice
versus place contrasts in aphasia. Brain and Language, 2, 434-450.
Miceli, G., Gainotti, G., Caltagirone, C., & Masullo, C. (1980). Some aspects of phonological
impairment in aphasia. Brain and Language, 11, 159-169.
Miceli, G., Mazzucchi, A., Menn, L., & Goodglass, H. (1983). Contrasting cases of Italian
agrammatic aphasia without comprehension disorder. Brain and Language, 19, 65-97.
Milberg, W., & Blumstein, S. (1981). Lexical decision and aphasia: Evidence for semantic
processing. Brain and Language, 14, 371-385.
Milberg, W., Blumstein, S., & Dworetzky, B. (1987). Processing of lexical ambiguities in
aphasia. Brain and Language, 31, 138-150.
Milberg, W., Blumstein, S. E., & Dworetzky, B. (1988). Phonological processing and lexical
access in aphasia. Brain and Language, 34, 279-293.
Milberg, W., Blumstein, S. E., Katz, D., & Gershberg, F. (1995). Semantic facilitation in
aphasia: Effects of time and expectancy. Journal of Cognitive Neuroscience, in press.
Miller, G. A., & Nicely, P. E. (1955). An analysis of perceptual confusions among some
English consonants. Journal of the Acoustical Society of America, 27, 338-352.
Naeser, M. A., & Chan, S. W.-C. (1980). Case study of a Chinese aphasic with the Boston
Diagnostic Aphasia Examination. Neuropsychologia, 18, 389-410.
Neville, H., Nicol, J., Barss, A., Forster, K. I., & Garrett, M. F. (1991). Syntactically based
sentence processing classes: Evidence from event-related brain potentials. Journal of Cogni-
tive Neuroscience, 3, 151-165.
Niemi, J., Koivuselka-Sallinen, P., & Hanninen, R. (1985). Phoneme errors in Broca's aphasia:
Three Finnish cases. Brain and Language, 26, 28-48.
Ojemann, G. A. (1983). Brain organization for language from the perspective of electrical
stimulation mapping. Behavioral and Brain Sciences, 6, 189-230.
Onifer, W., & Swinney, D. A. (1981). Accessing lexical ambiguities during sentence compre-
hension: Effects of frequency of meaning and contextual bias. Memory & Cognition, 9,
225-236.
Osterhut, L., & Holcomb, P. (1994). Event-related brain potentials elicited by syntactic anom-
aly. Journal of Memory and Language, 31, 785-806.
Parisi, D., & Pizzamiglio, L. (1970). Syntactic comprehension in aphasia. Cortex, 6, 204-215.
Petersen, S. E., Fox, P. T., Posner, M. I., Mintun, M., & Raichle, M. E. (1988). Positron
emission tomographic studies of the cortical anatomy of single-word processing. Nature
(London), 331,585-589.
Petersen, S. E., Fox, P. T., Posner, M. I., Mintun, M., & Raichle, M. E. (1989). Positron
emission tomographic studies of the processing of single words. Journal of Cognitive
Neuroscience, 1, 153-170.
Peuser, G., & Fittschen, M. (1977). On the universality of language dissolution: The case of a
Turkish aphasic. Brain and Language, 4, 196-207.
Posner, M., & Snyder, C. (1975). Facilitation and inhibition in the processing of signals. In
P. M. A. Rabbitt & S. Dornic (Eds.), Attention and performance V. New York: Academic
Press.
Posner, M. I., Petersen, S. E., Fox, P. T., & Raichle, M. (1988). Localization of cognitive
operations in the human brain. Science, 240, 1627-1631.
Rinnert, C., & Whitaker, H. A. (1973). Semantic confusions by aphasic patients. Cortex, 9, 56-
81.
Rochford, G., & Williams, M. (1962). Studies in the development and breakdown in the use of
names. IV. The effects of word frequency. Journal of Neurology, Neurosurgery, and Psychia-
try, 28, 407-413.
Rosler, F. Putz, P., Friederici, A., & Hahne, A. (1993). Event-related brain potentials while
encountering semantic and syntactic constraint violations. Journal of Cognitive Neuros-
cience, 5, 345-362.
Ryalls, J. (1982). Intonation in Broca's aphasia. Neuropsychologia, 20, 355-360.
Ryalls, J. (1986). An acoustic study of vowel production in aphasia. Brain and Language, 29,
48-67.
Sasanuma, S., Tatsumi, I. F., & Fujisaki, H. (1976). Discrimination of phonemes and word
accent types in Japanese aphasic patients. International Congress of Logopedics and Phoniatrics,
16th, pp. 403-408.
Schwartz, M. (1984). What the classical aphasia categories can't do for us, and why. Brain and
Language, 21, 3-8.
Shewan, C. M., Leeper, H., & Booth, J. (1984). An analysis of voice onset time (VOT) in
aphasic and normal subjects. In J. Rosenbek, M. McNeill, & A. Aronson (Eds.), Apraxia
of speech. San Diego, CA: College-Hill Press.
Shiffrin, R. M., & Schneider, W. (1977). Controlled and automatic human information pro-
cessing. II. Perceptual learning, automatic attending, and a general theory. Psychological
Review, 84, 127-190.
Swinney, D. A., Zurif, E. B., & Nicol, J. (1989). The effects of focal brain damage in sentence
processing: An examination of the neurobiological organization of a mental module.
Journal of Cognitive Neuroscience, 1, 25-37.
Tanenhaus, M. K., Carlson, G., & Seidenberg, M. S. (1985). Do listeners compute linguistic
representations? In D. Dowty, L. Kartunen, & A. M. Zwicky (Eds.), Natural language
processing: Psychological, computational and theoretical perspectives. Cambridge, UK: Cam-
bridge University Press.
Tissot, R., Lhermitte, F., & Ducarne, B. (1963). Etat intellectuel des aphasiques. Encephale, 52,
285-320.
Tuller, B. (1984). On categorizing aphasic speech errors. Neuropsychologia, 22, 547-557.
Tuller, B., & Story, R. S. (1986). Co-articulation in aphasic speech. Journal of the Acoustical
Society of America, 80(Suppl. 1), MM17.
Tyler, L. (1985). Real-time comprehension processes in agrammatism: A case study. Brain and
Language, 26, 259-275.
von Stockert, T. R. (1972). Recognition of syntactic structure in aphasic patients. Cortex, 8,
323-334.
Zangwill, O. L. (1964). Intelligence in aphasia. In A. V. S. de Reuck & M. O'Connor (Eds.),
Disorders of language. London: Churchill.
Zatorre, R. J., Evans, A. C., Meyer, E., & Gjedde, A. (1992). Lateralization of phonetic and
pitch discrimination in speech processing. Science, 256, 846-849.
Ziegler, W., &von Cramon, D. (1985). Anticipatory coarticulation in a patient with apraxia of
speech. Brain and Language, 26, 117-130.
Ziegler, W., &von Cramon, D. (1986). Disturbed coarticulation in apraxia of speech: Acoustic
evidence. Brain and Language, 29, 34-47.
Zurif, E. B., & Caramazza, A. (1976). Psycholinguistic structures in aphasia: Studies in syntax
and semantics. In H. Whitaker & H. Whitaker (Eds.), Studies in neurolinguistics (Vol. 1).
New York: Academic Press.
Zurif, E. B., Caramazza, A., & Myerson, R. (1972). Grammatical judgements ofagrammatic
aphasics. Neuropsychologia, 10, 405-417.
Zurif, E. B., Caramazza, A., Myerson, R., & Galvin, J. (1974). Semantic feature representa-
tions for normal and aphasic language. Brain and Language, 1, 167-187.

The Neurobiology of Language

Uploaded by

Copyright:

Available Formats

The Neurobiology of Language

Uploaded by

Document Information

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

The Neurobiology of Language

Uploaded by

Copyright:

Available Formats

:HAPTEF

The Neurobiology of Language

Speech, Language, and Communication

lenge is determining whether and how much neuronal activity relates to a

components and to their representations, that is, the primitive units or

broad anatomical (anterior and posterior) and functional (expressive and

II. THE S O U N D STRUCTURE OF LANGUAGE: P H O N O L O G Y

environment errors in which the occurrence of a particular phoneme is

time serves to distinguish two categories of voicing, voiced and voiceless,

In addition to the impairment in the timing of independent articulators,

Miceli et al., 1978). It is interesting that similar patterns emerge in non-

auditorily presented sound segments embedded in a phonetic context

C. Speech Perception and Lexical Access

e n d - p o i n t / t / s t i m u l u s is a word (e.g., task), and the e n d - p o i n t / d / s t i m u l u s

scribed as agrammatic or telegrammaticin their speech production. The speech

elicit a broad array of syntactic structures, patients may use reduplication to

of morphological constraints has also been shown. In particular, Baum

What remains unclear is whether these similar patterns of performance

syntactic impairments typically affect language production as well as lan-

are n o t selective w i t h respect t o a particular c o m p o n e n t o f the g r a m m a r n o r

Ganong, W. F. (1980). Phonetic categorization in auditory word perception. Journal of Experi-

You might also like