International Phenomenological Society

Signs About Signs About Signs Author(s): Charles Morris Source: Philosophy and Phenomenological Research, Vol. 9, No. 1 (Sep., 1948), pp. 115-133 Published by: International Phenomenological Society Stable URL: http://www.jstor.org/stable/2103856 Accessed: 13/12/2010 09:38
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=ips. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

International Phenomenological Society is collaborating with JSTOR to digitize, preserve and extend access to Philosophy and Phenomenological Research.

http://www.jstor.org

DISCUSSION SIGNS ABOUT SIGNS ABOUT SIGNS on on that I comment some of the comments It is withsomehesitation piled on commentaries Signs,Language,and Behavior. For commentaries and theladderofsignsbecomesmore circle find ever-narrowing ofreaders an to it shakythe higher ascends. Perhapsit would be wisestmerely approare mycritics' laterwritings. But we philosophers in of priatetheinsights and not as wise as our admonitions. for gluttons words we To keep our bearings mustrecallthe purposeofSigns,Language,and of the The aim ofthe book was to further development "a comBehavior. science of signs." The book was to be a proprehensiveand fruitful of to legomenon a naturalscienceof signs,and not the actual presentation was envisagedas a of such a science. The development the scienceitself in for working manyfields manygenerations." taskfor"manyinvestigators to by could contribute thisdevelopment atIt wvas that a philosopher felt for to certainmethodological tempting clarify issues, by striving a more togethermaterialfrommany different by precise terminology, bringing might someoftheways in whichsemiotic and by suggesting fields study, of and proveto be ofhightheoretical practicalimportance. for written a mixedaudience-forthosewhowish The kookwas therefore science,forthosewho look an in to participate constructing experimental of for on for to semiotic its possiblebearing theirowndiversefields interest, of thosewho seek to advance the integration knowledge. The attemptto the nourishsuch a varietyof needs influenced languageand the contentof not the book itself. It is too analyticforsome readers, preciseenoughfor others,too dry for some, too speculativefor others. In saying that it morethan it recordsan achievement"(p. 246), the "sketchesa program
Research:the reviewby Daniel J. Bronand Phenomenological 1 In Philosophy Vol. F. stein,Vol.VII (1947),pp.643-649;"TheNew 'Semiotic,"'by Arthur Bentley, of to VIII (1947),pp. 107-131;"An Introduction thePhenomenology Signs,"byJohn Vol. Wild,Vol. VIII (1947),pp. 217-233;"Logic and Semiotic,"by Elaine Graham, IX (1948),pp. 103-114. In otherjournals: "Peirce's Theory of LinguisticSigns, Vol. XLIII Thought,and Meaning," by JohnDewey, The Journalof Philosophy, Semiotic," by Max Black, (1946), pp. 85-95; "The Limitationsof a Behavioristic Vol. LVI (1947),pp. 258-272;"The Semioticof Charles Review, The Philosophical Review Morris,"by PhilipBlair Rice,TheKenyon (1947),pp. 303-311. To thesemay and of by be added the reviews David Ryninin the Journal Aesthetics ArtCriticism (1946), of Review Literature Jr. (1947),pp. 67-70,and W. K. Wimsatt, in theQuarterly pp. 180-185.
115

116

PHILOSOPHY

AND PHENOMENOLOGICAL RESEARCH

rectified, amplified, to as book clearlyis presented something be modified, superseded. have measuredthe book by I stressthispointbecause some ofits critics claimswhichit did not profess. Bentleybanishesit as "a failure"because to attributes me,in hiswords, the it does notjustify claimwhichhe wrongly language,and thus creatinga science." On that of "buildinga scientific is it and page 247 ofSigns,Language, Behavior, is statedthat "a terminology studies "may even make desirable not . . . a science," and that further terminology.Black writesas if my statedrasticchanges"in the present ment that "technicalsemioticmust provideus withwordsthat are sharpened arrows" is a claim that the presentbook is withoutterminological a in defects. The bookis quiteexplicit expressing hope "to reduceforscien(p. distinctions tificpurposes"the "vaguenessand ambiguity"of current than of 4). That it is in Black's words"a quiverbrimful slingshot" rather sharpenedarrowsis of coursecorrect. But David used a set of perfectly effect. Precisionis a matterof sling shot on Goliath with considerable degree,as Black made quite clear in his 1937 paper on "Vagueness" in of Philosophy Science.
I. THE GENERAL PROBLEM OF TERMINOLOGY

signs,or any proposalfor It is a truismthat any inquiryabout existing is itselfcarriedon in part by signswhichare not use of signs, the future these analyzed in that particularinquiry. Let us distinguish themselves sign by two classes of signs,relativeto a giveninquiry, the termsanalyzed and unanalyzedsign. The unanalyzed signs of a given inquirymay be thoseof an establishedscienceor theymay be termsof the everydaylanguage in whichthe inquiryis being carriedon. Thus an Englishchemist use for a may,in proposing terminology chemistry, withoutanalysisterms in physicsand such wordsas 'it,' 'no,' 'thing,''become,''cause,' 'produce,' etc. The questionof the adequacy of the unanalyzedtermsdependson the one purposeforwhichtheyare used. Wherethe purposeis the scientific that the unanalyzedterms it warranted assertions is important of getting in be such that thosewho cooperatein the inquiryare in wide agreement undergivenconditions. Then on the basis of the applicationofthe terms such unanalyzedtermsthe termsof the sciencein questionare analyzed, and analyzed("defined"),as whenthenecessary sufficient either completely for conditions theirapplicationare stated,or partiallyanalyzed,as when for but sufficient not necessaryconditions theirapplicationare stated by the use of the unanalyzedterms. unIn Signs, Language,and Behaviorit is suggestedthat the scientific analyzed termsof semioticbe drawnfromthe biologicaland the physical sciences,and that such termsas 'idea,' 'mind,' 'thought,'and 'conscious-

SIGNS ABOUT SIGNS ABOUT SIGNS

117

contermsbecause of the notorious the ness' be excludedfrom unanalyzed in disagreements theirapplication. The biologicalorientation temporary and methodological primarily is of Signs,Language,and Behavior therefore as does not involvea defenseof "behaviorism" against "mentalism"conwithrats doctrines. Nor does it involvestarting sideredas psychological of or dogs as some of my criticsseem to imply;the organisms babies or woulddo just as well. philosophers thus takes as unanalyzedsuch termsas Signs, Language,and Behavior etc. 'need,' 'energy,' 'reaction,''muscle,''gland,' 'sense-organ,' 'organism,' are of coursemanyotherunanalyzedtermsin the book: 'behavior,' There 'situa'art,' 'religion,' 'produce,''possible,''condition,''cause,' 'influence,' withthe are 'reader,'etc., etc. Some oftheselatterterms combined tion,' the unanalyzedtermsdrawnfrom biologicaland physicalsciencesin order to analyze (partiallyor completely)the termsproposedforsemiotic,but someare not. Thus 'cause' and 'condition'are so used; 'behavior'is emanalysisof 'sign' (p. 7) but not in the attemptto ployedin the preliminary analysis(p. 10). give a moreprecise of as and 'response' Since students behaviordiffer to the terms'stimulus' Clark Hull). Four othertermswere termswere defined(following these, 'disposition to respond,' 'response-sedefined:'preparatory-stimulus,' in ('behavior'is not employed the definition quence,' and 'behavior-family' terminolof 'behavior-family').Withthe help of thesetermsa semiotical ogy was proposed. 'Sign' was partiallyanalyzed (not defined). Then of 'denotatum,' 'interpretant,' there were given definitions 'interpreter,' 'signal,' 'denotes,' 'signifies,''sign-vehicle,''sign-family,' 'significatum,' 'forappraiserr,' 'prescriptor,' 'symbol,''lansign,''ascriptor,''designator,' of structure Signs,Lanmator,'etc. Such is in outlinethe terminological and guage, Behavior. is used in developingthis terminology It seems to me that the method directedat the terminology correct. Some of the criticisms scientifically of are based on a confusion the analyzed and the unanalyzedsignsin my of account, or on a confusion unanalyzedsignswhichplay a role in the and of terminology thosewhichdo not. These development the semiotical makemyterminology appearto be looserthanthefactswarrant. confusions whichfurther problems and Thereare ofcoursea hostoflegitimate difficult advance in semioticmust meet. The term 'sign' itselfmay serve as an example.
II. WHEN IS- SOMETHING TO BE CALLED A SIGN?

under Our initialtask in analyzing'sign' is to state a set of conditions a whichwe will call something sign. If in doingthiswe are not arbitrary, are but keep one eye cockedon what often called signs,the analysisproves by to be verydifficult.The earliestformulations studentsofbehavior-

118

RESEARCH PHILOSOPHY AND PHENOMENOLOGICAL

which or stimulus that a signis any stimulus that a signis any substitute to a influences response some otherstimulus-werefoundto be too simple. meetssome of the diffiin The formulation Signs, Language,and Behavior new and presents rather complex but cultiesofearlierformulations, is itself problems. of Two partial analyses (not definitions) 'sign' were givenin the book, the orientation, second as an attemptto move for the first preliminary precision. towardgreater A, formulation p. 7) was as follows:"If something, controls The first (on with) identical a towards goalin a waysimilarto (butnotnecessarily behavior goal in a zith to behavior respect that else, thewaysomething B, wouldcontrol then it observed, A is a sign." in situation which were is It was held that this formulation adequate formany purposes,but was desirable purposesa more preciseformulation that forexperimental beof the in orderto clarify terms'similarity behavior'and 'goal-seeking was suggested admittedto be tentative, havior.' A second formulation, which A, (p. 10): "If anything, is a preparatory-stimulus in theabsenceof of initiatingresponse-sequences a certainbehavior-family stimulus-objects by conditions to undercertain in causesa disposition someorganisms respond then of response-sequencesthisbehavior-family, A is a sign." withClark Hull, of out months discussion worked after This formulation, does not make use of the Edward Tolman, and some of theirco-workers, reactto a signas it does that an organism term'behavior,'does not require of to what the signdenotes,does not requireidentity reactionin the situain tionin whichthesignappearsand thesituation whichit does notappear, is at at and does not require themoment whichsomething a signthatthere be any responseat all (in the sense of muscularor glandularreactions). seemed to me, and still In this,and in otherways, the new formulation advance. seemsto me,to marka substantial complexity. Criticshave feltthat But it pays forthis by an unwieldy at it the precision introduced someplaces is jeopardizedby the lack ofprecision in such termsas 'disposition'and 'need' (a termemployedin the of definition 'response-sequence'). Two years of work with the second respects. I agreewith has formulation shownme itsinadequacyin certain Philip Blair Rice that it may be betterto stressreactionsmore and rekind of reaction)less. I agreewithMax Black that sponses(as a specific the relationbetweengoal-objectsand signsis not clear in the secondforof we certainly do notwantto require a sign(suchas 'black') that mulation; does and onlygoal-objects, ifmysecondformulation thistheterm it signify to needs modification droppingout the reference by 'response-sequence' certainlydoes not requirethat signs goal-objects(the. firstformulation only goal-objects). I agree with ArthurBentleythat the second signify

SIGNS ABOUT SIGNS ABOUT SIGNS

119

becomes unwieldy. The term 'sign' clearly needs a better formulation thanI have givenit, or thanI knowhowto giveit. grounding
III. REPLY TO MR. BLACK

wastes a good arrowin the emptyair. occasionally Black, an able critic, in Thus he "rubs his eyes" at the vaguenessof the word 'influence' the to a phrase"influences response some otherstimulus"and aims thisverbal inarrow: "It would be absurd to say that the stimulusto the, morphine to a jectiondenotes rabbitwhichis yet to be presented the dog or unifies the in the presenceof food. Nor is therethe least difficulty generalizing upon subsequentbewhichhas some causal influence point: Any stimulus have to count haviorwill,ifwe insistupon the letterof Morris'sdefinition, to a as a sign." But the simplefactis that thephrase"influences response and of some otherstimulus"is partof thedefinition 'preparatory-stimulus' makesclear and of a notin itself definition 'sign'. Signs,Language, Behavior 9) and uses on page 7 the very are thatnot all preparatory-stimuli signs(p. as to responses laterstimuli examplewhichBlack uses ofa druginfluencing whichhas somecausal a reasonfornotanalyzinga sign as "any stimulus behavior." influence upon subsequent in and arbitrariness my acBlack makesmuch of an alleged circularity it is count. His argument complexand I will not attemptto reproduce in he full. In termsof the argument claims that my method is defective is whichI employ notreallyan empiriof sincethecriterion meaningfulness in cal one but dependsupon a preference thechoiceofa metalanguage. In Black's terms:
are stimuli all deIf the denotataassociatedwitha set of givenpreparatory language(orwhatmeansof a predicate availablein theEnglish scribableby ever language the semioticianprefers),the stimuliwill be "signs" and the "dispositionto renot. Since thereis no way to identify otherwise, a exceptby finding set of associated despond" or the "behavior-family" to of of notata,verification the allegedpresentation a signreduces this: A is the which wouldremove needsmotivata signifand onlyifthecircumstances the speaking, circum(or, ingcertainassociatedresponse-sequences roughly in stancesto whichthe signis applicable) can be described theEnglishlanor sense from nonsense, guage. But this leaves us helplessto discriminate in unverifiable, the case the from nonsensically meaningful the empirically of signsbelonging theEnglish languageitself. Justat the point where to semioticmightbe expectedto becomeuseful,in ways otherthan that of to of of speculativeinterpretation the signification animalresponses condius brings to a halt (pp. procedure of tionedstimuli, circularity theentire the 269-270).

Black has perhapsscalped an Indian, but u less my In this argument deceiveme, my own scalp is quite intact. For on my ac ount "a fingers

120

PHILOSOPHY

AND PHENOMENOLOGICAL RESEARCH

of withouttherebeing a formulation what it sign may, of course,signify and betweena significaturn a distinguish (p. signifies" 20); 1 specifically significatum.PerhapsBlack as a logicianis so accustomedto formulated in is workwithsignswhosesignification alreadyformulated othersignsthat out of that a theory signsmustprovideus withways of finding he forgets whata signmeans. empirically the sayingthatunlesswe are able to formulate signification He is in effect the of a signthe signcannothave signification-hence supposedrelativity richenoughto formulate to for ofmycriterion signification a metalanguage and arbitrariness, the signification.And hence the supposed circularity would not have signification since the signs in the-chosenmetalanguage theirsignification someotherpreferred in unlesswe couldin turnformulate metalanguage. But on the accountgivenin Signs,Language,and Behavior withoutour beingable it is quite possiblefora sign to have signification eitherbecause our own languageis not rich the to formulate signification, of of enough or because our empiricalinvestigation the signification the outlinedgives us a way to sign has not gone farenough. The procedure of of the signification signsin a seek the signification any sign,including the to we provided do notattempt formulate significagivenmetalanguage, collapsesinto the tion of all signsat once. The allegedviciouscircularity thereare some unanalyzed harmlessfact that at any momentof inquiry and unformulated significations. playsa crucial 'condition' and Behavior It is truethatin Signs,Language, role, and is unanalyzed. It is also true that it mightbe well to analyze this term. But the fact that it is not analyzed does not mean that it is withwithoutsignification.Black in his own accountuses 'circumstance' uses out battingan eye: the Englishlanguagewhichwe both employoften or both termsundersimilar"circumstances" "conditions." We may and a or oftendo ask underwhat conditions underwhat circumstances person of explodes,or philosobecomesPresident the UnitedStates, or dynamite pherswritebooks. I see no reasonwhy we "must" analyze these terms of in talkingabout signs. Analysisand reconstruction our language inevitablyproceedspiecemeal. A finaland somewhatminorpoint. Black holds that " 'response-sewithhumansigns. quences'and 'needs' are fictions"-at least in connection in But sequencesofresponses admitted the approachesofbothTolman are recognizeonly such "general and Hull. Black talks as if psychologists human 'needs' as food, sex, play, and the rest." Actually "the rest" is will in to a largeclass,as reference Murray'sExplorations Personality show. of And whileit may be well to attempta reformulation semioticin which that it 'need' is not a basic term, is significant bothTolman and Hull stress and of variablesin psychology, use needas such theimportance intervening

SIGNS ABOUT SIGNS ABOUT SIGNS

121

a variable. Even the elementary laws of learningare stated by Hull in with the reductionof need. If response-sequences needs connection and I are fictions, at least have not been responsible theirintroduction for in the studyofhumanbehavior.
IV. THE CASE OF MR. BENTLEY

AlthoughI do not see that Black has vindicatedhis charge that my methodis at fault,his criticism at least pointed. Of BentleyI can only is is showthebasic inaccuracy say thathis criticism at least long. I willfirst of.whathe presentsas a "report" of Signs, Language,and Behavior, and in then commenton his generalremarks the second part of his paper. In his second footnote Bentley says that he will use the word 'define' "casuallyand loosely"as I am said to do. In Signs,Language, Behavior and 'define'is quite precise:it is synonymous with givingthe necessaryand sufficient conditions the applicationof a sign; 'sign' itselfis expressly for is said not to be defined 250). Bentley'scasual and loose use of 'define' (p. it reallyessentialto his strategy: obviates the necessityof distinguishing in thedefined terms thebook,the partially the analyzedterms, unanalyzed takenfrom terms and the unanalyzedterms general in biologyand physics, downthesefences has a romping use in English. By breaking he field-day but eventuation. indeed,exciting without Take the term'behavior.' In Signs, Language,and Behavior this term is part of the unanalyzedvocabularyof the Englishmetalanguage. It is not an analyzed termin the languageproposedforsemioticand it is not in employed defining ofwhat Bentleylikesto call my "official" any terms (it is used in the preliminary analysisof 'sign'but not in the secondpartial analysis of 'sign' to whichBentleymainlyrefers). Bentleyobscuresthis in factby a typicalinaccuracy hisreference Signs,Languageand Behavior. to He says 'behavior'is used in the book's definitions; on page 115 ofhis and articlehe arbitrarily replacesin the definition 'response-sequence' of the term'response-sequences' the term'behavior.' This gives the reader by theimpression that 'behavior'is a keydefined term, and thatit is restricted to "muscle-gland action." Such carelessness his part misrepresents on the book he is said to be "reporting." Or consider page 111 ofhis articlethefunhe has at myuse of 'or' and on 'um' endings(to distinguish what signifies what is signified). The exand is ample he uses to show my ineptness a straight misquoting the text: of he says thatforme a locatumis a sign,whereasthe page 67 ofSigns,Lanhe guageand Behavior quotes makes clear that only the word'locatum' is a sign. This is thesame order confusion wouldbe involvedin claiming of as that someoneheld that man is a noun because he said that the sign'man' was a noun.

122

PHILOSOPHY

AND PHENOMENOLOGICAL RESEARCH

Such carelessreporting occurs throughout article. On page 125 he his are says thatformators signsabout signs;thisis specifically deniedon page 157 of Signs, Language, and Behavior. On page 113 he gives a supposed of definition 'reaction,' referring page 355 ofmybook; but there'stimulus' to and is defined not 'reaction.' On page 116he says thatmuscle-gland action in itself held to be purposive;Signs,Language,and Behavior is says nothing of the sort. On page 117 he statesthat 'behavior-family' itself"defined is withrespectto behavior"; a glance at pages 10 and 346 of the book will showthatthisis not so. On page 119 he says that the distinction between is between'meaning'and 'sign' and 'sign-vehicle' "the ancient difference but still before us in all its ancient unexplained 'word,>rechristened crudity";p. 251 of Signs,Language,and Behaviorpointsout that the distinctionintendedis that made by Peirce in termsof 'type' and 'token.' Bentleyhas two otherfavorite devices. One is to list a numberof unanalyzed wordsused in the book ('substitute for,''because of,' 'initiates,' in 'controls,''affects some-way,''determines,' etc.), and then say without inexplanation(as on page 122) that theseare instancesof the "linguistic because theyare security"of Signs,Language,and Behavior-presumably not analyzed. Is the Englishlanguagereallyso insecure? The second and more commondevice is to restatethe language of the book in such a way that it seemsto involvean endlesscongeries disconof entities. Signs,Language, Behavior and defines response a nectedand reified as an action of a muscleor gland; because the word 'action' ratherthan of 'reaction'is used, Bentleysays this "has the effect pushingit offto a distanceand presenting rather"on its own" thanas a phase ofsemiosical it are process" (p. 114). Because organisms said to be disposed tovarious and behaviorsare thus types of behaviorwe are told that "dispositions set over against each other" (p. 144). I am said to make behavior "a of to compartment the organism," make sign "a glistenly transmogrified denoteror signifier," make "musclesand glands all on theirown" purto ,are as posive (p. 123). Such distortions passed off honestreporting. Evmade in Signs, Language,and Behavior seen by Bentley is ery distinction an entities. as a reification, unnecessary of multiplication disconnected By thesevariousstrategies Bentleygivesthe readerthe impression that Signs, Language,and Behavioris a terminological chaos. In the second part of his article,thus prepared, moves in forthe kill. The book is he dismissedas an unholy alliance of "goal-seekingpsychology the rat at level," set up "withlittlechange,"and "the cobwebsofthe olderlogicsand a philosophies," mixture the outworn of traditions Newtonianmechanics of withphilosophic formalism dualism. and As forthe first memberof this supposed monstrosity, is hard to see it how an attemptto use and extendthe behavioralanalyses of Mead, Tol-

SIGNS ABOUT SIGNS ABOUT SIGNS

123

all man, and Hull is to employa rat-levelpsychology: of these men have endeavoredto develop a general theoryof behaviorwhichwill embrace the behaviorofrats and menand yet accountfortheirdifferences.As for has an explicitdiscussionof 'mechanism,' Signs, Language,and Behavior freedom (pp. 197, 275). PerhapsBentley'sreadingof the book is blocked 'stimulus'and 'response';perhapshe by the mereuse of the sign-vehicles thinks "mechanically" calls out a response. But thisis not everystimulus without the positionofthe book. On its terminology therecan be stimuli thisresponse maybe and when thereis a responseto a stimulus responses, partlydue to the way in which the stimulushas been signified. There can thus occur responses-sequences whichhave never beforeappeared in the behaviorof the responding organism. Dewey's positionin the article as on thereflex-arc to conceptin psychology whichBentleyrefers out-modby ing my view is nowherecontradicted the book-nor by the work of Tolman or Hull. symbolic logic As for"the cobwebsof the olderlogicsand philosophies," an later is presumably "older" logic,a strangeuse of terms;I will return to the relationof symboliclogic and a behavioralsemiotic. -The book is view of thought"by making supposed to "maintainthe older mentalistic. "between" the organismand the a sign a third "thing" that intervenes the world,and thus "to bringback sub-rosa verythingthat Peirce-and to James and Dewey as well-spent a good part of theirlives trying get objectsand organisms of ridof." Now thisclaimthatI am "fearful getting intodirectcontact,"and put theghostof "ideas" betweenthem,is without is respondto something somefoundation. The fact that how organisms in of does not imply is times part a function how that something signified, and world is ever lost, or that signs that the directcontact of organism always occur, or that when signs do occur some mental or quasi-mental of of "thing"has intervened. Whenbehavioris a function a number variables, no one of these variables need be made into a mental substance whatbetweenthe organism and the world. Thereis no warrant standing as soeverforinterpreting Signs,Language,and Behavior an attenuatedCartesian dualism. My own cosmology-not defended the book itself-is in in a versionof objectiverelativism. A briefsketchof it is presented the article,"Multiple Self and Multiple Society," in Freedom and Experience: to Press, 1947,pp. Essays Presented Horace M. Kallen (CornellUniversity 70-77). is Semiotic so undeveloped of todaythata plurality variousconstructions are desirable,even withina behavioralframework. I wish Mr. Bentley well on his own enterprise. He and Dewey have chosen to begin their terminological analysisat a lower level than did Signs,Language,and Behavior. But in theirconstruction have to developan altertheywilleither

124

PHILOSOPHY AND PHENOMENOLOGICAL RESEARCH

nativescienceofbehaviorthat will convincescientists, theywillhave to or show how the basic termsof existingbehaviortheorycan be definedin theirown terms. No mereinsistence "transactions" on will be sufficient. That what happens happens only in contextsdoes not freeus fromthe to obligation speak precisely about what does happen in variouscontexts. When they have elaborated theirnew semioticsufficiently deal with to behaviortheory, withsignsin animals,withsignsin the sciences, existing and logic,arts,and religions, withsignsin personality disturbances, will it to no be instructive assess theirresults. In the meantime usefulpurpose is sustainedby a wholesaledistortion Signs,Language,and Behavior. of V. INTERPRETANT SIGNIFICATUM AND In footnotes 60, anid61 of his article,Bentley summarizes 57, Dewey thatmyversion semiotic "a complete is of to theeffect inversion Peirce." of however, claimsto be "an attemptto carry Signs,Language,and Behavior, the insightof Charles Peirce that a sign gives rise to an out resolutely and is interpretant that an interpretant in the last analysis'a modification towardaction.'" Something wrong is somewhere. ofa person'stendencies Dewey's articlein whichhe makes his charge("Peirce's Theory of LinVol. XLIII and Meaning,"Journalof Philosophy, guisticSigns,Thought, (1946), pp. 85-95) dealt only with my earliermonograph, Foundations of the Theory Signs. For this and otherreasons it did not seem fair to of of replyto Dewey's analysisin termsof the fullertreatment Signs, Language,and Behavior. But since Bentley'snote showsthat Dewey has not and since changedhis positionbecause of Signs, Language,and Behavior, he (Bentley)claimsthat I have "evaded" the issue Deweyhas raised, some somewhat'indiscussionof Peirce is in order. This may serve to clarify and 'significatum'-terms whichhave botheredmany critics, terpretant,' and understandably. I shouldliketo makeclearthattheposition developedin Signs,Language, did and Behavior not startfrom Peirce. GeorgeH. Mead first stimulated me to thingabout signs behaviorally. Signs, Language,and Behavioris. of in many ways a further development Mead's Mind, Self, and Society. refer Peirce. Only later to I neverheardMead in lectureor conversation did I workearnestly Peirce,Ogdenand Richards,Russell,and Carniap, at in and stilllaterat Tolmanand Hull. All ofthesepersons influenced variof ous ways theformulation- Signs,Language,and Behavior. Nevertheless, in historical it perspective, seeds to me that the positionof Signs, Lanwas not derivedfromPeirce, guage,and Behavior, thoughits orientation "an his to is in effect attempt carryout resolutely" approachto semiotic. that I have invertedPeirce's posiBut Dewey says no. His argument and 'interpretant.' tion centersaround the use of the terms'interpreter'

SIGNS ABOUT SIGNS ABOUT SIGNS

125

of Dewey states that the crucial erroris my "gratuitousintroduction an of into a perinterpreter" 86), and thenthe "translation 'interpretant' (p. to sonal user" (p. 88), that is, into an interpreter.According Dewey the interpretant Peirce is a sign which in a process "supplies meaningto for the earlierones" (p. 89). It is perhapswell to quote Dewey's positionat morelength:
Interpretant, The misrepresentation question consists in converting in as used by Peirce, into a personaluser or interpreter.To Peirce, 'interif which thereby giving preter,' he used theword, wouldmeanthat interprets, meaning a linguistic to sign. I do notbelievethatit is possibleto exaggerate a the scornwithwhichPeircQwould treat the notionthat whatinterprets whimor capriceofthosewhohappen givenlinguistic signcan be leftto -the this factthat Peirce holds that the to use it. But it does not followfrom sign,is an "object" in the interpretant, that whichinterprets linguistic a in the interpretant, sense of an existential"thing.' On the contrary, Peirce's usage, is always and necessarily another linguistic sign-or, better, "On set of such signs. The following passage is strictlyrepresentative: we the whole,then,if by the meaning a term,proposition, argument of or of the interpretant, themeaning an then understand entiregeneralintended is argument explicit. It is its conclusion;whilethe meaningof a proposito or tion or ter-m all that. is thatproposition termcould contribute the don(p. clusionofa demonstrative argument" 87).

In Signs,Language,and Behavior interpreter an organism which is for an in to somethingis a sign,an interpretant a disposition an interpreter is respond in certainways because of a sign (p. 17), and a dispositionto respondis "a state ofan organism a giventimewhichis .uch that under at certainadditionalconditions responsein questiontakes place" (p. 9). the the to Finally,thoseconditions'which would permit organism respondas it is disposed to respondbecause of the sign are the signification signi(the ficatum)of the sign. The questionnow is whetherthe positionof Signs, Language,and Beis withPeirce'sdoctrine whether is, as or it havior in substantial agreement of Dewey contends,a falsification Peirce's doctrine. To answerthis we mustbringPeirce'swordsintothe argument. does employthe term'interpreter.' The interpretantPeirce himself of whether or emotional, energetic, logical-is an effect a signon an interpreter. Thus Peirce inquiresinto "the nature.., of-theessentialeffect upon the interpreter, broughtabout by the semiosisof tne sign, which the constitutes logicalinterpretant" Papers,V, ?484). (Collected whichhe gives The answeras to the nature of the logical interpretant is as follows may be a thought, (ibid., ?476): "Shall we say that thiseffect thatis to say, a mentalsign? No doubt,it may be so; only,ifthissignbe kind-as it would have to be-it mustitself have a logical of an intellectual

126

PHILOSOPHY AND PHENOMENOLOGICAL RESEARCH

so of interpretant; that it cannot be the ultimate logical interpretant the concept. It can be provedthat the onlymentaleffect that can be so produced and that is not a signbut is of generalapplicationis a habit-change, a of toward meaningby a habit-change modification a person'stendencies bothered the mentalistic terms this acin action." Peirceis somewhat by countand "to clear the signof its mentalassociations"he pointsout that a "habitis by no meansexclusively mentalfact" (ibid.,?492). A habitis a and when readiness"to act in a certainway undergiven circumstances actuated by a givenmotive" (ibid., ?480). A habit is "a tendency . . to . in behave in a similarway undersimilarcircumstances the future." He even speaksofa habitas a "disposition respond a givenkindof stimuto to of lus in a given kind of way." Man by "fanciedreiterations" a desired in kind of conductcan "producehabits, just as do reiterations the outer will world;and these habits havepower influence to actualbehavior theouter in world"(ibid., ?487). In termsof these quotationsI can only concludethat the positionof is Signs,Language,and Behavior verycloseindeedto thatofPeirceand not of an inversion his position. Both talk of effects signson interpreters, of in both describeinterpretants termsof such effects, both regardthe inon as tendsto such that the interpreter terpretant an effect an interpreter act in a certainway undergivencircumstances whenactuated by a given need Signs,Language,and Behavior to uses the sign-vehicles 'disposition 'tenrespond,''respond,''conditions,'need'; Peirce uses the sign-vehicles dencyto act,' 'act,' 'circumstances,' 'motive.' Both use the terms Bentley findsso objectionable:'person,''behave,' 'produce,' 'influence,' 'modify,' 'constitutes.' Whereigtheessential in difference thetwoaccounts? 'effect,' Wheredoes Signs,Language, and Behavior invert Peirce? It is stillnecessary explainhow Dewey was led to his interpretation to of Peirce. I thinkan ambiguity Peirce's term'interpretant' in gives the explanation:'interpretant' sometimes, -above, as namesa certain kindofeffect ofa signon an interpreter tendency action),but often signifies what (a to it I have called the signification (significatum) the sign (the circumstances of underwhichone could respond he is disposedto respondbecause of the as that Dewey refers whenhe to sign). It is thisseconduse of 'interpretant' says that forPeircethe interpretant not something is existential but a set ofsigns. A better instance perhaps thissentence is from Peirce:"thelogical is in interpretant general its possibilities reference of (i.e., refers is related or to whatever theremaybe of a certaindescription)" (ibid.,?486). To formulatethis"whatever"in othersignsis to givewhat Peircecalls the meaning of a particular"logical interpretant"-and this process can go on indefinitely.When Dewey singlesout -only this aspect of Peirce's terminologyhe writesthat "signsas such are connectedonlywithothersigns."

SIGNS ABOUT SIGNS ABOUT SIGNS

127

is to even thenthatreference an interpreter "gratuitous," It does notfollow on forif one asks-as Peirce also does-what is the effect an interpreter one signification, can holdwithout thatis nota signwhena signhas a certain with Peirce-and with Signs, Language,and Behavior-that contradiction of is thiseffect a tendency actual personsto act in a certainway undercertain conditions. Dewey himselfhas writtenin a similarvein. In his treatise,"so faras Logic,pages 48-49, he says of the readingof a scientific is definite groundedagreementor disagreement reached, an attitude is readinessto act in a responsive way when whichis a preparatory formed in the conditions questionor otherssimilarto themactuallypresentthemselves." is betweenPeirce's two uses of 'interpretant' made in The distinction and 'sig'interpretant' Signs, Language, and Behaviorby distinguishing nificatum.' I think the above account explains-and explains awaythat I have foistedan Ersatz versionof Peirce on an Dewey's contention public. unsuspecting also I may add that thisdistinction cuts the nerveofJohnWild's somewhat similarchargethat the positionof Signs, Language,and Behavioris any"subjectivistic." Wild says that on its analysis "we cannot signify our own subjectiveresponsesand thingwhichis not a complexinvolving and interpretant significatum.Signs, Lanpurposes." But thisconfuses admitseach "involves"the other(p. 18); Wildwrongly and Behavior guage, quotes this as sayingeach "belongs" to the other. The fact that a sign for to whenthereis a disposition respondby the interpreter only signifies (the interpretant) whomit is a signdoes not mean that it is the disposition on whichis signified. The book-as on page 19-is quiteexplicit thispoint.
VI. THE STATUS OF DISPOSITIONS TO RESPOND

of center aroundthe term Many criticisms Signs,Language,and Behavior I 'disposition.' In previousformulations did not use this term. It was to in introduced order makeit possibleto say thata givenitemis a signand are of even whenno responses the organism made. Suphas signification asks me thewhereabouts the Pennsylvania of pose forinstancea passer-by Station, and I say to him 'straightahead to 32nd Street and then two blocksto the left.' He does not at the timeof hearingthe wordsturnto the left,yet we would ordinarily that the words are "signs" to him. say expressesthis situationby sayingthat the Signs, Language,and Behavior is passer-by "disposedto respond"because of mywordsin a certainway, in the wordschangethe state ofhis organism such a mannerthat i.e., that ahead a need to get to the Pennsylvania Stationhe willwalk straight given and turn leftwhen-he sees the streetsign with the letters'32nd Street' uponit.

128

RESEARCH ANDPHENOMENOLOGICAL PHILOSOPHY

are The commoncriticism twofold:one set centersaround the notion we the otherset around the questionas to whether itself, of 'disposition' of can get evidence as to the existenceor non-existence a dispositionto respond. are Disposition-predicates of course very commonin everydayspeech. to makespersonsimmune somediseases, innoculation Paper is inflammable, the rubberband on the table is elastic, the sugar in the bowl is soluble, a is dynamite explosive,a certainpersonis phlegmatic, certainbar of iron is and so on. In each case something said to be at a. given is magnetic, react conditions in moment such a state that it will underotherstipulated to 'disposition in distinguishing in suchand sucha way. I see no difficulty 'inflamresponsex' and 'responsex' that is not presentin distinguishing out throwing all mable' and 'burning.' I thinknone of my criticssuggest and I do not see that any of themhave givenan arguterms, dispositional to term'disposition respond'.isobdispositional mentwhy the particular are terms not. If sameofmyreadershave if dispositional jectionable other Carnap's theyconsider I terms, suggest a generaldoubtabout dispositional of and Meaning"'(Philosophy Scsence, in ofsuchterms "Testability analysis VolumesIII and IV, 1946 and 1947), Stevenson'sanalysisof 'disposition' variof in Ethicsand Language,and the discussion the place ofintervening and Hull. givenby Tolman, ables in psychology that I have not told whatthe state of There is an occasionalcomplaint because of a is the organism when it is disposedto respondto something is sign,i.e., what the interpretant as an organicstate. The fact that I made no attemptto do so was deliberate. I do not know. The problem as of is an empiricalone and its solutionawaits the development semiotic semithat a scientific science. This does not mean,however, an empirical of is untilthisproblem solved. Centuries fruitful oticcannotmakeprogress explanationof the magnetic studyof magnetswent on beforea molecular state was given. The case may well be the same withsemiotic. of Since I did not give a biologicaldescription the state of an organism is as forwhichsomething a sign,it is not correct, Rice does, to assumethat static. Whateverit is, it is unI must mean by this state something complicatedand dynamic. Rice thinksit may be doubtedlyimmensely a primarily circularneural process,a positionwhich has been suggested whichis beingexploredin an in by a numberof theorists recentyearsandDr. and psychiatrist WarrenS. eagerlyawaited work by the neurologist to McCullough. I am inclined believethat thisis a richvein ofore. But science,and it would have been out the problemis one forexperimental of keepingwith the characterof my book to anticipatethe answer. In to this fieldwe are just beginning roll balls down inclinedplanes.

SIGNS ABOUT SIGNS ABOUT SIGNS

129

It is nevertheless possibleto obtain evidenceconcerning occurrence the ofinterpretants even ifwe cannotdescribethemneurologically physiand was studiedbeforethe magneticstate was ologically-just as magnetism described. In Signs,Language, and Behavior number typesofevidence a of are discussed(pp. 13-15), and sincenone of my critics specifically has consideredthe analysis theregiven,I feel no obligationhere to reopen the matter. Perhapsone pointcan be made. Peircewrites: "How otherwise can a habit be describedthan by a description the kind'of action to of whichit gives rise,withthe specification the conditions of and of the motive?" (Collected Papers, V, ?491). Insofaras this is so, evidenceforthe presenceof a habit is foundby observing whether underthe given conditions and motivethe kind of action in questionis performed. This is in fact the first kind of evidencediscussedin Signs, Language,and Behavior. But we can go much further. Once we have this kindofevidencewe can begin to study the state of the organismwhich does so act under such and circumstances, as we get such knowledge gives us a new source of it evidenceforthe presenceof habits even when no overt behavioris performed. This has been the courseofdevelopment the studyofmagnetin ism, solubility, and immunity. A semioticwhich starts with overt reremainat that descriptivelevel. But for the sponses need not forever is present macroscopic.descriptionour necessary anchorage.
VII. LOGIC AND SEMIOTIC

Elaine Graham's paper, "Logic and Semiotic,"will serve as the basis fora fewremarks one of the mostdifficult on (and mostinteresting) problemsthata biologically grounded semiotic mustface: Can it includewithin itselfformalor symbolic logic? ChapterVI of Signs, Language,and Behavior,in an admittedly tentativeand explorativemanner,suggesteda positiveanswer. This means that logical constantsand logical sentences set up dispositions respond do non-logical to as signs. It does not,ofcourse, as Bronstein'scriticism this positionassumes, requirethat an actual of responsealways be made when one interprets logical (or a non-logical) a. sign. Graham'spaperfirst the considers relation between behavioralanalythe sis of certainformators Signs, Language,and Behavior in and the analysis oftheirsignification of givenby thetruth-tables thesymbolic logician. She aims to show,and I thinkdoes show,that when some ambiguities the in book are resolved,the "two formulations in a real sense equivalent." are My only questionat this pointis whether interpretation formators her of in terms "patterning denotation" "patterning interpretants") of of of can (or be extended apply to all formators to (such as 'not' and 'some') or whether it holdsonlyforsuchconnectives 'and' and 'or.' as

130

RESEARCH AND PHIgNOMENOLOGICAL PHILOSOPHY

proposedin Signs, the criterion The second part of her paper concerns (analyticand contradictory ascriptors for and Behavior formative Language, is ascriptor a was sentences). This criterion statedas follows:"A formative such that the denotationof one or moreof the comcompoundascriptor is, ascriptors) because ofthe signifi(called the antecedent ponentascriptors for a condition the denotationor lack of cation of the ascriptor, sufficient or (called the ascriptor ascriptors component denotationof the remaining of or and ascriptors), henceforthedenotation lack ofdenotation consequent itself" (pp. 162, 163). the compoundascriptor is Graham'sanalysisshowsthat thisformulation unsatisfactory.Some only one possiblereadingof the cricome from she of the difficulties finds terion,namely,where 'antecedent'is taken to mean 'before'and 'consequent' to mean 'after.' Thus she says that 'Johnismyson and I am John's since if ascriptor, an father'would be on this criterion analyticformative for are 'Johnis my son' denotesthenthe conditions fulfilled 'I am John's for father'to denote,and hence the conditions the compoundsentenceto meantby is denote-and yet thiscompoundsentence not what is normally an analyticsentence. forced, While this readingof 'antecedent'and 'consequent'is somewhat that in defective not specifying was certainly as the criterion formulated implication(see page 22 of Signs, Lanthe relationintendedwas that of for guage,and Behavior the term'implicate'and page 165 forthe view that involve an implicate relation). The examples given analytic sentences was in thatthedefect formulation notaccidental however, in thebookshow, but the resultofinadequateanalysison my part. So thisarrowof Elaine spot. Grahamhitsa vulnerable thismight (though criterion Ratherthanattemptto sharpentheprevious way out would be to take I be done successfully), now thinkthe simplest by sentencesformulated symbolic criteriaof analytic'and contradictory theirtermson the basis of the biologically and attemptto define logicians, logic of terms semiotic. If thisweredone,thewholeofsymbolic grounded a within "behavioral"semiotic. And I see no theoretical wouldbe brought in difficulty doing this. Carnap, forinstance,in Meaning and Necessity, an page 10, defines 'L-true sentencein languageSI' in termsof 'holds in is in everystate-description SI.' A state-description "a class of sentences eitherthissentenceor its in Si which containsforevery atomicsentence negation,but not both, and no other sentences" (p. 9). He then gives whethera given sentence"holds" in a (on p. 10) rules for determining state-description.If we could define'sentence, 'atomic sentence,''negaand tion,' and 'holds' in the vocabularyof Signs,Language,and Behavior, the whole of Carnap's ascriptor,' equate 'L-true' and 'analyticformative as and terminology methodof analysiswould becomepart of semiotic developed in the book.

SIGNS ABOUT SIGNS ABOUT SIGNS

131

us This brings to Elaine Graham'sthird question, whether analyticformative ascriptorsnecessarily denote (and so are necessarily"true"). She answersyes; Signs, Language,and Behavioransweredno on the general is groundthat as semiotic thereconstructed signification such neverinas volvesdenotation. What is nowclearis thattheproblem terminological, is and dependsupon how 'denote' is defined. In the Graham construction denote' is, as she admits,itselfan analytic 'analyticformative ascriptors formative ascriptor, just as in Carnap's terminology L-truesentence 'an is true' is itselfL-true. In my terminology, however,to say 'X denotes' implies'thereis or was or will be some y such that . . .'; hence I argued does not imply'x denotes.' Grathat-'xis an analyticformative ascriptor' ham questionswhetherI am rightbecause of the unique "patterning of whichoccursin analyticformative interpretants" ascriptors. She may be on correct thispoint. The moralto be drawnis thatwe mustbe cautious of variousconstructions semiotic, not to confuse and in any givenconstruction not to confuseformative ascriptorswith otherkinds of ascriptors. of That is whythe axiomatization semiotic mustbe carriedout withmuch care than is done in Signs,Language, Behavior. Here semiotic and greater needs the help offormal logic. A finalpointon the relationofformallogic and semiotic. Formal logic of but is not a theory signification an analysisofgivenor proposedsignifications. (This does not implyof coursethat a formal logicianmay not also have a-theoryof signification.)As such,formal logic is compatiblewith a various theoriesof signification-with Platonic as well as a behavioral has analysis. The formal logician,therefore, nothing fearin attempts to the to define basic termsof formal logic on the basis of termsin a biologiwas the heir of a long tradition cally groundedsemiotic. Peirce himself whichconceivedof logicas part of semiotic. He himself, rightly, and saw betweenhis work as formallogicianand his analysis of signs no conflict to tendencies behavior. And semiotic, developedin Signs, as involving as Language,and Behavior providesa place withinone systemforthe logical of ofanalysisofthe significations signsand forthe investigation what hapfor pens interpreters whomsignshave signification.
* * *

I will not attemptnow to climb still higheron the semioticalladder. about esthetics the lightof comments in I had intendedto say something made by Philip Blair Rice, W. K. Wimsatt,and David Rynin. And I had intendedto discuss JohnWild's Platonic "phenomenology" signs, of to and the relationof pragmatism semiotic. But thesedispositions reto spondhad bestbe checkedforthemoment. Our discussionhas been limitedto foundational problems. My critics, have quite properly as philosophers, concentrated such problems, on and they have given Signs, Language,and Behaviora very throughanalysis

132

RESEARCH PHILOSOPHY AND PHENOMENOLOGICAL

I indeed. I thankthemforthis. Undoubtedly have not done fulljustice of the in arguments myattemptto defend generalsoundness Signs, to their we arrows have drawnin attack and Behavior. The manyverbal Language, are and defense evidencethat we at least agreeupon the basic importance thought. of in of semiotic the reconstruction contemporary APPENDIX of I the Since writing above comments, have seen the manuscript L. 0. Kattsoff's"What is Behavior?" (this journal, Vol. IX, (1948), pp. 98paper, GeorgeGentry's 102), and readtwoanalyseswhichI had overlooked: Vol. XLIV (JournalofPhilosophy, and "Signs,Interpretants, Significata," (1947), pp. 318-324), and VirgilC. Aldrich'sreviewin the same issue (pp. 324-329). is inside the interpreter" holds that to admit "something Mr. Kattsoff whathe meansby 'idea,' to introduce "ideas." Sincehe does notexpliciate as it is notclearwhathe wishesto add by thistermto 'interpretant' defined in Signs, Language,and Behavior. His claim that we can only findout or is for whether a humanbeingthere an interpretant idea, and the purposes to him and listening his forwhichsomeoneuses a given sign,"by asking; as reply,"seemsto me wrong. We do ofcourseuse verbalreports evidence of of the signification signsand the purposesforwhichsignsare produced. But personsdo lie and do have signswhose signification theycannot preand ciselyformulate do use signsforpurposesof whichtheyare unaware; to recognize thisiTh admitthatthereis evidenceotherthanverbalreports to of by which the reliability verbal reportsthemselvescan be controlled. discourseis lexicativeand not thirdclaim-that metaphysical Kattsoff's formative-cannotbe discussedsince he did not state what he means by discourse.' 'metaphysical The precedingpages have made clear that I am not, as Mr. Gentry in to suggests,"committed a motortheory"of the interpretant contrast workin the that scientific to a corticaltheory. I do not believe,however, what is the state of the until one determines study of signs is impossible for is organismwhen something a sign; much good work in linguistics; example, has been accomplishedwithoutthat question being answered. of His secondmain thesis-that the significatum a sign is itselfa cortical process-arises fromhis attemptto answerthe questions "What sort of to?" I believe that entityis a significatum?What does the termrefer is of thisformulation the problem misleading. It is truethatI statedthat a and that "a signsignifies signifieatum." But these "a signmustsignify" formative and Behavior are ascriptors in the language of Signs, Language, and "a and not lexicative aseriptors; signhas signification" "a signsignifies ascriptors. Hence to searchfor a significatum" equivalentformative are a "the signx signifies to" an "entity"whichis "referred in the ascriptor

SIGNS ABOUT SIGNS ABOUT SIGNS

133

the be processor not-is to runthe significatum"-whether entity a cortical withinsemioticitselflexicativeand formative discourse. riskof confusing the is thatofhowto describe what If we avoid thisdanger, problem merely othersigns. I do not see that thisdescripa givensign from differentiates tion must be given solelyin terms of the corticalprocesseswhichoccur. the signification 'spiral nebula' with of Hence I see no need to identify or occurwhenthe termis a signto someone, to whatever corticalprocesses say that the term"refers"to such processes. It is forsuch reasonsthat and ('signification'). Per'interpretant' 'significatum' I have notidentified basic terms semiotic whichfewer in has haps Mr. Gentry in minda simpler are involved. Such a semioticwould be welcome;I have not been able to one distinctions. to construct that did not seem' blurimportant Mr. Aldrichsinglesout in his reviewthe problemof the differentiation of the modes of signifying.I agree withAldrichthat my accountleaves but I do not thinkthat he has put his finger the on much to be desired, nor between maindifficulties thathis own suggestion a basic distinction for evocativeand cognitive signsprovidesthe way out. His chargethat I do things not actually analyse the ways of signifying only the different but the signified-sothat I am leftwithonlyone mode of signifying, designaare on tive-is unconvincing two points: the modes of signifying distinof kindsof interpretants thatwe may (so guishedin factin terms different the etc.), appraisively, ways,designatively, signify sameobject in different Aldrichuses 'designates'in a widerway than I do, perhaps and secondly, as equivalent to my term'signifies.' I would not say in 'x is good' that x, to good' designates even whenx is good, since 'designates'is restricted of characteristics objects. 'X is casually efficacious signs which signify begood' is appraisive only if someoneis disposedto accord preferential havior to x because of the sign 'good'; to say that "y likes x" is not apof praisiveof x but designative y in relationto x. So in my terminology and does modeofsignifying not theappraisive(and prescriptive formative) reduceto thedesignative. Thereis on myaccount,to use Aldrich'sterms, an "evocative" and "cognitive"aspect to signsin all modes of signifying; to pull theseapart and make themthe basis fora classification modesof of is signifying in my opinionto returnto the state of analysis represented of meaning. My account in theold dichotomy "emotive"and "referential" needsto be improved, I think indicatesa direction which in but it certainly have a "cognitive"characterwhich is conappraisals and prescriptions and trollableby evidenceand yet remaindistinctin signification in emfrom designativestatements. If thislead can be developed piricalcontrol we to satisfactorily can eat the cakes offered us both by the pragmatists and the logical empiricists. I forone am hungryenoughto want both. CHARLES MORRIS.
UNIVERSITY OF CHICAGO.