Primer of Verbal Behavior - Winokur.pdf
Post on 02-Jun-2018
262 Views
Preview:
Transcript
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 1/89
a•
prImer
ofverbal behavio:
an oant view
STEPHEN WINOKUR
Prentice-Hall, Inc. Englewood Clifs, New Jersy
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 2/89
ib j Cng atalging n blain ataWinokur, Stephen (dae)
A prmer of verbal behaio
Bibliography: p.
Incldes index Verba behaor 2 Operan behao Title
BF455.W54 153 76-4989
IBN 0376098
Prentice-Hall Series in Experimental Psychology
James J. Jenkins, Editor
976 by PenceH, nc, Eneoo Cff, N J rght reerved No part ofth bookma be reproduced n an form or b an me anthout permon n rtng from the publher.
rinted in the United States of America
10 5 4 3 2
renticeHa Internatioa Inc LondonrenticeHa of Astraia ty Limited Sdne
rentice-Ha of Canada Ltd Torontorentice-Ha of India rivate Limited Ne Delhrentice-Ha of aan Inc Toorentice-H f teast sia te Ld Snaor
Foreword vii
Acknowledgments
1Introduction
Aim 1Characteristics 2Status 3A nonverbal example 4Definition of operant 6
2
xi
The Interlockng Verbal Operant Paradigm
The wrong paradg 11Interockng prad 1
0
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 3/89
Cos
Size o f units of analysis 17Indices operant strength 22
3
Mands 25Definition and paradigm 25Why deprivation and aversive stimuation ecome controing 27Reminder of how reinforcers wor 27Diagnosing mands 31Supplementary strengthening and the impre mand 32Why do speaers mand? 33Why does the hearer comply? 33Deprivation and aversive stimulation as suppery vaiales 37
The self as mediator 38
Anomalous mands 38
4Tacts
Definition and paradigm 41Reinforcement for tacting 44Comparison of tact and mand 48Why do mediators pay their roe? 50
Discriminative stimuli as suppementary variaes 51
Distorted tacts: inexact correspondence 52
5Extended Tacts
Two mechanisms 55Discrimination and generalization in tacting 61Degrees of discrimination in tacting 62The size of the response in tacting 65
Abstractions 67Do animas tact 6
55
Cos
6udiences
Definition and paradigm 71Address 72
Audiiity and the auience 73Audiences as response selectors 76Multiple and concurrent audiences 81Shaping by the audience: a personality effect 82One's self as one's audience 82
7Echoics
Definition and paradigm 84Echoism in children 87Echoism in aduts 0Some odds and ends 1earing and echoics 1
8Textuals
Definition and paradigm 3Reinforcing textuals 6
9
7
84
93
Intraverbals 100
Definition and paradigm 100ow intraverals are reinforcd 101
Cains 103Clusters 107
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 4/89
Cos
10utipe Causatio
Problems arising from multiple causation 115Misspeaing 122
Verbal engineering 125
11Autocitics
Denition and paradigm 128inds of autoclitics 130The autoclitic a fudge factor"? 132Ordering as an autoclitic process 135
Reinforcement history for autocitics 137Awareness 13
12Some Impicatios
The hearer 144Meaning 146Thining 148
Suggestios For Further Readig
Refereces
Idex
115
127
1
15
155
161
wAs Professor Wiour tels us in this oo fr my ys
have taught a course on veral ehavior at th niversiy is tscestry traces bac to Sinner's earlier classic formuation of behvior in&eera, including speech Winour too this cours in 1962 d but fivyears later he sat in and isened againfrom egining to end; pp that with this cass I have nevr venured to nd ou why fearing i wsbecause they could not believe their ears the frst ime This boo ressures
me; it sows that the operant analysis of verbal behavior can and oe provoe and sustain thought What' s betterit survives continuing informedreection and a critical second listen
Winour has since taught a simiar course of his own, and I amenormously peased to see what the earlier, always incipient formultionhas become in his hands He has found elaborations an consonances andclarifications that certainly had eluded me No doubt thre are more tocome, free for the pondering to anyone who troubles to ponder This pointof view is something one can live with, and upon being lived with it wil bseen to grow as if by a vitality of its own One goes bac to it and discoversnew possibilities, new facets It gives the psychoogist an experience that israther rare in our disciplinethe excitement of init
I should thin this oo has something for everyone, in the best pos sible sense For those areadycommitted to the ways of operat anysisthere will be the rssurc that vral bhavior cn, in spit
vII
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 5/89
oewod
recent critical harassment) still e considered within the domain of operantanalysisand here's how They will e interested to se how the extensionof operant analysis to a very complicated ehavior class illuminates somefamiliar dogma without changing ittelling us what we must have eenmeaning all along For example, the veral case shows clearly that its controlling stimulus is inevitaly involved in the specification of an operant,
and therefore mut e included in its definition; the operant is really notsimply defined y its consequences " Autoclitic ehavior, a rich vein thatwill demand and then repay a reworing y the reader, teaches us somethingimportant aout the interaction etween stimulus generalization andresponse iduction, and aout how temporally concurrent variales canproduce temporally succesive responses I should thin it inevitale thatsomeday we must all understand te autoclitic process; even nonveralehavior has its grammar
For the elsewherecommittedthe psycholinguist, the humanist, thephilosophical rationalistwho want to now hat the ehaviorists position on language really iscongratulations To learn aout eaviorismyou have done est to come to the ehaviorist ere he is is oo willprovide a clear and ehavioristically sanitary accont of itself And it willno dout e a revelation to you if your previous notions are due tosecondary, nonehaviorist sources
I suspect there are few noncommitted on tis issue This is curious,since no dout noncommitment and an asolutely open attitude of inquiryare the only sensile positions for any of us to tae in these early timesAnyone who really elieves, and then announces in pulic that he or shenow has th nal answerthe ey to veral ehavior (or language, orspeech)simply cannot now what they are taling aout Speaing onlyfor the view expresse in this oo, I do thin it reasonale to conclude thatgiven what we o now know aout nonveral ehavior, most of what this
oo says aout veral ehavior nearly mt e true Th ltimate Scienceof Langage, when it is written, will have much of this in it I errors areunliely to e those of commssion, ecause of the way operant analysesproceed The sensile question is whether this truth, and truth derived asthis was, is the wole truth The ehaviorist adopts the woring hypothei(it is not a claim or a theory) that th ehavioristic consraints andeconomies are sufcient; if they are, his system will e elegant Meanwhilehe extends d purifies and corrects the system where he can, and as alwayswith honest men, e watches for a disconfirming datum Such disconfirmingdata, e assured, have not een encountered yet All those rave newproofs" of ehaviorism's insufficiencies seem hardly to have drawn anyehavioristic adrenin, much less lood
The woringhypothesis attitude is the optimum, and certainly anexcelent one for te reader to maintain ere te ehavioristic ac
orewod
count to e accurate and sufficient Taing full advantage of its apparatuswithout ending it anywhere, apply it according to its own rules to thedomain of veral ehavior See how far it will tae you Ten mae yourets ut not efore
You are in for a good time, and you may well end up at a destinationyou never expected to reach
ut first a caveat to the reader what follows is titled a Primer ecausei lays out the structure of its argument in clean, straightforward, nofatrose It is rather formal and not very chatty It is intended to e taen quiteliterally; we do not use metaphor It does not elaorate greatly; it does notrepeat itself nor underscore especially important points ecause, in fact, itaes almost no unimpotant ones All this maes for a highme sageto
of ts revty And hee is the caveat the revity and surface simplicityi h etray you into judging that the oo's ideas are as rief and simpleas ts prose ell, don't The underlying sense of the argument is rich,sbtle, and sophisticated, as you will discover for yourself if you readattentively
So, this short oo is a favor to your time and pocetoo; it pays youe compliment of supposing that you will do some of the wor yourselsince you are given all the necessary systematic apparatus You will finassertions here that do not seem to follow from what has gone efore, they do follow, and you shold pause to igure out why they o, awhence Some sentences will e counterintuitive; they are, however,proucts of the analysis They are not necessarily wrong ecause they are at
riance with common sense," and I think, should no e ligtly written off. If all this were nothing ut common sense and ideas comfortale to it,
the boo would not e worth the writing The paradigms, or diagrammatic definitions of the unctional parts of
speech should e studied and fully grasped They eep us moored to ourasic operant origins and show clearly that no new stimulus or responseariables, and no new processes, have crept into the analysis The analysis nvolves only new cominations and complexities of already familiarvariales that have een defined previously and that have ecomewelltempered in the laoratory with other species and other respones
Kennet MacCoquoale
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 6/89
kwlmOffice space an irary faciities se ring te preparatin f
the manscript were prvie y te niverity f Minnesta CaifrniaState niversity at Haywar an Texas Cristian niversity epeciayt J G. Darey N ivsn an J Rsenerg fr arranging tese
R Be T R Dixn J enins MacCrqae an S CWinur rea the manscript an mae many epf sggestins
I am especiay gratefu t B F Sinner an ennet MacCrq-
ae The rst as een my teacer via is written vera eavir wie tesecn has een a mre persna instrctr am greaty inete t tf them fr the materia in this In arge part it is their raterthan mine Hwever tey sh nt e e respnsie fr any f itserrrs
Sanra C Winr rea te manscript severa times type it fean therwise encrage te autr an refuse t e satisfie ness teaccunt was mae cear an namigs Witht er ep encragement an criticism the manscript w never ave een cmpete Tis is fr her
x
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 7/89
nodon
or some reders, h e irs pr o his oo wi e rgey re-view o wh hey redy now ou e wor o B F Sinner For oherrders, his pr o he oo wi provide concise inroducion o he odoogy nd phiosophy o Sinnerin psychoogy Boh groups oreders, however, shoud give i heir enion ecuse i wi ene hemo deveop more criic undersnding o he pproch I wi e ing ohe sujec mer o he res o he oo
AM
My im here is o inroduce you o Skinner's nurl science c-coun o n individu's ver ehvior. This ccoun seeks o expining in erms o person's ps hisory, curren circumsnces, ndnohing more Anoher wy o syig his is h we wi oo nenumerion o he vries o which ver ehvior is uncion. Th is,you will be introduced to what Skinne, and I, consider a behavioral analy-sis o alking
You my now h Sinne hs wrien oo ced Vebal -avi nd you my hve wondered why you ren' eing sed o redh oo insed o his one I woud sy h mny yers o eching ex
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 8/89
nroducon
perence have shown ha Vebal Beavio s no organze n a way hamakes suale or se as an nroducory ex I aso conans a lo moremaeral han we w e concerned wh here Aer al s a complee accoun o he vews o one o he world's mos emnen psychologss on avery large area o human ehavor Sknner's Vebal Beavio ncludes dscussons o wrng and edng as well as akng and readng I'll jus ellyou aou he las wo My approach here wll e derved rom ha developed y Kenneh MacCorquodale who or many years has augh a
course alled Vera Behavor" a he Unversy o Mnnesoa MacCoruodae egan o each hs course a aou he same me ha Sknnerwas gvng hs Wllam James Lecures (he orerunners o Vebal Beavio)a Harvard Unversy The course and he ook are very smlar; I cannosay whch conaned wha rs No maerhs ook s grealy ndeed ooh The echncal erms and e asc pan o anayss are surey Sknner's u almos all o he dagrams and many o he examples were deveoped y MacCorquodale n hs pedagogc paraphrase" o Sknnerdurng hs ecures I consder oh men o have me mporan conruons o he accoun o veral ehavor and o e excellen an nsprngeachers Ths ook s hen an aemp o share wha I have learned romhem wh you
CHARACTERSTCS
I wan o characerze hs approach rely y menonngseveral salen aspecs o
I s a STIMULUSRESPSE PSYCHLGY THIS MEANS THAT
BEHAVOR IS UNDERSTOOD TO BE T SUBJECT MATTER OF THE INQUY. THE
RE MOVEMENTS OF AN ANIMA OR ITS PARTS WHICH ARE PRODUCED BY THAT
ANAL'S MUSCLES CONSlTE WHAT IS TO BE EXPLAINED, raher han deasmeanngs wans desres nenons expecaons or any knd o hypo
hecal physologcal mechansmsA NATURAL SCIENCE ACCUNT IS WHAT IS OFFERED TH
SUBJECT MATTER CONSISTS OF, AND IS LIMITED TO, PUBLIC, EMPICAL, RE,
NTERSUBJECTIVELY OBSERVLE EVENTS. Veral ehavor s ehavor; sno somehng deren rom nonveral ehavor Makng nose wh yourmouh (alkng) s jus as much and no more or no less ehavor as s rng a cycle e wl no e concerned wh akng as s ymoc ehavor"or as a vehcle or deas" Ths approach s no a popular one Manyscholars ncludng many psychologss or ha maer do no agree hawha s o e explaned s muscle movemens or he nose ha hey produceSome see hese daa as a ass or nerence o nds cognons panssrucures grammars algorhms and so orh However we wl no e
noducon
concerned here wh explanng hose sors o nerred enes nor wll wemake use o hem n expanng ehavor
n hs accoun he varaes o whch ehavor s a uncon aresough among pulc emprcal rea oservale evens We look a presenand pas saes o he organsm (deprvaon) hs genec consuon (ahuman or a chmpanzee can acqure veral ehavor u a pgeon canno)and presen and pas saes o he envronmen (smul); all o whch arereal physcal nersujecvely oservae hngs The covaraons o
changes n hese emprcal ndependen varales and changes n ehavormay e descred (saed) DESCRTIONS OF FUCTINAL RELTINSMONG EMPIRICAL VIBLES CONSTITUTE T SORT OF CAUSA ACCOUT
HAT IS PRESENTED HEE. CAUSE CNTR AND SIMAR WORDS ME,
WHEN USED HERE, PRECISELY T OBSERVED RELATIONSP BETWEEN PUIC
DEPENDENT AND INDEPENDENT VARIABLES.
A undamenal premse o hs approach s ha he asc erms andprocesses nvoved n he scence o ehavor are general and appcale oall ehavor even veral ehavor We can do research wh pgeons andras o locae mporan processes and her eecs and hen appy heresuls o ha work o vera ehavor
STATUS
A he presen me we can gve ony a pausle reconsruc-on" or heory o he orgn manenance and conrol o veral ehavor
\ hese erms The vocaulary and he aws come rom aoraory researchon oher organsms (ras and pgeons) dong oher hngs (pressng ars andpeckng a plasc dsks) Bu hs vocauary and hese aws shoud apply oalkng Vera ehavor sreches Sknner's explanaory sysem o s lmsThe approach aken here s o see how ar we can go wh a scenc accoun whou resorng o enes or processes occurrng somewhere else a
some oher level o dscourse ha s n he ran or mndI he were asked I hnk ha Sknner would say: No only s hs aproposal or an accoun o veral ehavor i an accoun o ha sujecmaer; proaly s correc; s suppored y evdence; s no a ocor can e urer suppored y new no ye oleced evdence; he ac-coun and he evdence are emprcal pulc and oservaonal u mosyno expermenal; ha s oservaons are made and daa are collec ed undependen varales are rarely manpulaedneverheless hs s sll em-prcal scence as much as asronomy and geology are empcal scences
Sknner does no y any means have he only smulusresponse psy-chology There are many oher approaches o smuusresponse psycho-ogy due largely o men such as Hul Spence Guhre and Eses o ne
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 9/89
noducion 4
jus ew Such oher simuusresponse psychoogies re eyond he scopeo his ook Anoer pproch o he sudy o verl ehvior which islso eyond he scope o is ook is relivey new ype o psychology
This is cled psychoinguisics" (or exmpe, see he ook ySoin) nd is sed on he ssumpion h here mut e somehing moreo verl ehvior, nd h ny simulusresponse ccoun mut e indeque ecuse i leves hings ou This pproch origined in linguis-ics nd closely resemles Ges psychoogy in mny respecs, u mospriculry ecuse i hs een criicl movemen I see no specildvnge o compring nd conrsing psychoinguisics wih Skinner'spproch I you red his oo k s pr o course, you my ind h yourinsruco wi mke he comprisons r you my, once you inish hisook, mke hem yoursel My purpose here is o explin his ccoun ses I cn, displ some o he evidence or i, nd le you ech your ownconclusions
A NONVERBAL EXAMPLE
Alhough Skinner' s ccoun my e cimed o hve cerin imsnd chrcerisics, i my righully e sked wheer o n such n -proch is possile, nd i i is possi e, hs i een successul Cn one relygive cus ccoun o n individu's ehvior in erms o his ps hisorynd curren circumsnces, where erms, oh inpus nd oupus, rephysiclly speciie; nd where inpus consis o ses o he environmen(simui) nd o he orgnism (deprivion) nd oupus consis o ehvior(responses) Cn he ccoun lso e n exposiion o he reionships eween vries descripion o wh leds o whwhere l erms regener nd no resriced o ver individuls
I clim h hese quesions cn e nswered in he rmive, e-cuse mny exmples show h he Skinnerin progrm hs worked Le us
ook cse where his pproch hs een emineny successul There resever dvnced ccouns o his, s well s mny inroducins o i whichre i in gre dei (see Suggesios or Furher Reding in he ck ohis oo, especi y i wh oows is uncomory new or you) This ishe pr wih which I guessed h you migh e we cquined Even i youre, pese e me review i
Le us se e how o give nur science, psychoogicl, cusl ccouno he nonver ehvior o nonverl nim excusively in erms o hisps hisory nd curren circumsnces To do hs we olw Skinner's edy inding pigeon, nd hen puing him in sml ox Then we oservehis peckng
noducon
e my, s n side, mke mehodologicl oservion hispoin e didn' c like he Noel Prizewinning ehoogiss Lorenz ndTinergen e didn' sk wh he ird does e didn' survey is repertoire e nmed ehvior uni (he peck) nd sked, hen does i occurnd h cuses i o hppen e will do he sme sor o hing wih verehvior e will sk wh he circumsnces r under which mn wisy c," uncle, " Grce me no grce, nor uncle me no unce" This is deiere dierence in pproch Neiher we nor he ehoogiss re righ or
wrong; he ims o he wo pproches re dierenLe us eurn o he pi n e see h i he peck plsic disk onthe wa of t ox (he key) is ollowed y he presenion o wh isnormlly ood or he pigeon, he pecking my increse By mnipulinghe consequences o he ird's ehvior we cn cuse or chnge (shpe) heehvior's re or requency or is inensive properies, such s orce, dur-tion, or excursion (how r in spce he ird moves) To e sure, heoperions I hve descried work only i he ird is hungry" They woruomicly nd regulrly i he hs een deprived o ood, nd hey don'work ll i he hsn' In c, we now h i he hsn' een ood de-prived, sy, us er ig mel, he peck goes wy
As you my ememr, we cn ge ncy ou his; ncy or dumird, nywy e cn pu Chrisms ree ul ehind he key nd rrngehings so h i he ird pecs when he ligh is on, ood is presened; nd ihe ird pecks when he igh is o, no ood is deivered o h im Aer wood nd noood periods, hungry" ird erns" The pecing comes
nd goes wih he ighgive, o course, h he is deprived o od eorehnd
These seem o e he signiicn evens
Depivation he wihholding o he ood eorehndStimulu he ighed ey
Repone he peckReinfoce he oodContinency he reion h we se up mong he preceding iems
The exmple seems o show sureire mehod o conroing pecing Noonly h, we know h his mehod o condiioning works i mos evercse, irrespecive o he individu, simuus, response, reinorcer, orspecies invoved The exmpe is one o opeant conditionin There reimiions on his process, u we won' go ino hem now e us irsreview wh opern ehvor is
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 10/89
ntoduction
DEINITIN PERANT
hat follows wll e somewhat new even to those of you who arefamlar wth the wrtngs of Sknner and hs followers The llustratonbelow shows a paradgm for an operant In ths book I dsplay the modelcase or dealzed pattern for each mportant class of behavor or eachprocess n the form of a generalzed schematc dagram whch I cll a para-dgm PERANT BEHAIR IS TH NM OF A VERY LARG CLASS OR
GROP OF DIFFRNT BHAVIORS. THY ALL HAV CRTAIN COMMON PROPRTS TH MOST PORTNT OF WHICH IS THAT THY AR T RSPONSS THAT
PARTS OF OPRANTS. Ths defnton dffers from the tradtonal oneHere responses are not operants They are pat of operants It should be?oted that ths usage s a sgnfcant departure from Sknner's way of talkg about these matters and resembles the postons taken by J E R Staddon and J Fndley
DR R
Dp� AN PERANT
Because this way of talking about operants may be new to you, let usnow look at the components of the paradigmatic operant, starting from theright
side of the diagram. We beg in with the response (R)-the pigeon's
peck ou r nonverbal example. RESPONSS CA N B DSCRmD IN TWO WA YS.
(1 ) TH TOPOGRAPHIES MA Y B STA TD; W MA Y B TOLD WHA T MU SCLS MOV WHICH PA RTS OF TH BODY , A ND WITH WHICH FORCS, DU RA TIONS, XCU RSIONS, OR OTHR INTNSIV PROPRTIS. (2) W MAY B TOLDTHE EFFECTS OR CONSEQU ENCES OF THE MOVEMENTS U PON THE ENVIRON-MENT. Most operant responses are described in terms of their effects For example, the bar was pressed, the rat entered the g oal box, the key was pecked, the switch was flipped, the letter was typed A response such as the ba�ress is a c �ass
(� et). It has members which are individual barpresses.
Th Just ke d vduals who are members of a species (a man, the species man) . We w ll see that reinforcement does not reproduce the individual response that it follows, but it does lead to a changed probability of membersof that class. Our pigeon's individual pecks never return, but he does produce more responses of that sort even though each peck is somewhat different from the others .
TSD is the notation which we wil use for a d iscriminative stimulus.
DISCRIMINATIVE STIMULUS FOR A RSPONS IS A STA T OF TH N-VIRONMNT THA T IS OFTN PRSNT WHN THAT RESPONS IS FOLLOWD BY A REINFORCER. Dscrmnatve stmul are classes also. Not every occurrence of
noduction
the lght behnd the key s exactly the same as every other one Remember thatdscrmnatve stmul are not CS's (the condtoned stmul that acqure thepower to elct reflex responses n the procedure nvestgated by Pavlov) Theresponses under the control of CS 's are parts of reflexes; operant responsesare not reex responses Also remember that not every physcally specfableenergy change s a stmulus for a gven ndvdual or speces Lght s not astmulus for blnd dogs nor are Xrays stmul for humans although photographs made by Xrays are Although an anmal may hae the anatomc
and physologcal ablty to be controlled by a physcal energy change t s nota dscrmnave stmulus for hm untl some of hs behavor has been re-nforced n ts presence A pgeon can see red but he does n't do so untl a psy-chologst or some other part of the world feeds hm for peckng red thngs andextngushes (does not renforce) hs pecks at nonred thngs
hen we work wth anmals other than humans we oten work wth arenforcer that depends on epivation for ts effectveness Even to anexperenced pgeon food s not a renforcer unless he s hungry" Youmay recall that not all renforcers depend on deprvaton for ther effectve-ness; candy and other sweet stuffs work whether one s hungry or notNotce als that deprvaton s not a stmuus It s a poceue that some-tmes but not always produces some stmul EPRIATIN IS TH
WIHHOLDING OF STIMLI TO WHICH TH ORGANISM HAS PRVIOSLY HA SOGR OF ACCSS. They must be stmul; f they aren't there won't be nyeffect on behavor If you wthhold oxygen from a man ut contnue tosuppy hm wth ar" there s no effect on hs behavor but he des Therequrement that an organsm have some degree o access to the stmuus
ecomes clearer f we consder whether I have been eprved o Sophaoren or Nelson Rockefeller's money e would not normay say that
have and the reason seems to be that I have never ad regular or rreguaraccess to ether even though both mght be renforcng We may note afna pecularty of deprvaton; t s n one respect lke a dscrmnatvestmuus hen an ndvdual s deprved any responses renforced durng
ths deprvaton may come under the control of the deprvaton n that theyare more lkely to recur when the deprvaton recurs and less lkely to occurwhen the deprvaton s absent Ths phenomenon appears rrespectve ofhether or not the deprvaton produces characterstc stmul
T ARROW IN TH PARADIGM FOR T OPRANT RPRSNTS TH FACT
OF CNTINENCY OR CNITINALITY-T RLATIONSHIP BTWN
TH ISCRIMINATIV STLS OR T DPRIVATION AND TH RSPONS. Theoperant then s composed of a elation between the antecedent condtons(deprvaton or dscrmnatve stmulus) and the response class Instances othe oerant occur when a memer of the response class occurs under thecontrol of a member of the class f controllng antecedent condtons Butthe operant s a set (cass) tself wth each nstance of t beng a member o
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 11/89
nroducon 8
the set. S, ur pigen's anteceent cnitins were n f an the ightehin the key eing n. Its respnse cass was key-pecking. Eachccurrence f an perant was an instance f timesincef (eprivatin)pus ightnthekey (iscriminative stimuus) fwe y a peck at the keyrespnse (R). And ach such instance f cntre keypecks was an instance (a member) of the class of operants.
In the preceing I have mentine the re f the reinfrcing stimuussevera times. The flwing iustratin shws the reatin f reinfrcers t
perants by means f a paraigm.
SD ( SD )OR �R�S� Pr OR �R tEP EP
OPEANT
The sym S is use t inicate a reinforcin'imlus hie the symbPr ( )t indicates an increase in the prbabiity f whatever is within theparentheses. Thus, the paraigm inicates that when an perant is fwey a reinfrcing stimuus, the subsequent prabiity that peran is in
crease. REINFORCER (G MUU) Y ULU H H UBQU BBLY G H LLG UM UL). A reinfrcer strentens an perant; it increases the cntr exerte ver the respnse by the anteceent cnitins. The probaiityf emissin f the respnse in the presence f thse cnditins in the futurewi be higher. The functina reatinship represente y the arrw in theparaigm fr the perant is mae strnger. We may ntice that perants nt exist befre reinfrcers have acte n ehavir. The behavir may ccasinaly ccur with a w prbabiity ut the ccurrence f a reinfrcingstimuus fwing a respnse is what creates, strengthens an maintainsperants.
MMB H REINFORCERS AWAS INCREASE SPONSE PROAIIT. POSITIVE REINFORCERS BY BGPRESENED H ; NEATIVE REINFORCERS BY BG TRNED OFF QU H . REINORCEENT, , H G GH ay
G BY MG G .he reatin etween reinfrcements an perants is nt aways ne frne. Reinfrcements may e scheue intermittently an thus pruce verysrng perants (perant strength is ten shwn as persistent ehavir rhgh r� tes f respning). This is especiay true in vera behavir; n prfessr reinfrce each time he pens his muth. Smetmes the respnsees nt ccur at a r es nt have esre ntensve prperte ampi
noducon 9
tue, atency, frce, uratin, excursin, etc.). In this situatin, a prceure cae sapin must be use. SAPIN L U XM H . In ur eampe of the pigen if the bir in't peck the ighted ey,even thugh it was f eprive we wu have t shape eypecing. At first we wul reinfrce (y presenting f) any movemettwar the wa n which the key was munte. Then we wul wit forhea mvements twar the ey an reinfrce them ut not othr be
havirs. Finay reinfrcement wul be mae cntingent n bhaviorwhich brught the bea in cntact wit the key.These are aut a the terms an efinitins that we nee for this
accunt f vera ehavir. What flws next is an imaginary iagewhich iustrates the kin f expanatin that we give fr verbal behaviorut oes it fr the keypecking pigen.
Q: Why is that pigeon pecking the plastic disk?
A. : Because the light is on (if I turn it off, see, he stops), a nd he has been de-
prived of food
Q : But why does he peck when the light is on and he is hungry?
A. Because in the past, if he pecked when the light was o, I gave hm foo
when he was "hungry.
Q Is that all?
A.: Yes
This ex a mple ispa ys a functina view of cause or casl reationships.
\, Wha t is invve here is a specif icatin of "wha t lea ds to hat. That is'
wha evens in he e virmen lea d o cha g i avior? Fthr qs
tins, such as "why does he peck if in the pst wh h pckd a d was f ood
eprive a n the light was o yo f d him?, e for othr dis
cipine-physigy By agreement, we stp here. There is plenty of wor
f r us t . The physilgists wil ex plain smeay what it is inside the
bir tha t makes f a reinfrcer t him wh en h e is "hungry, a nd wha t itis insie the bir that makes reinfrcers ha ve the effects upon operant
srengh ha hey . We wi wa n nw a bu such ma ers, a nd for-
unaey here a re pepe wrking n geing his sr f inf rmaion. B,
in the mea ntime, we can g ahead and fnd ut what the psychogica laws
a re. This wi be helpfu in f rming a cmpee accun f he way he word
wrks, beca use these psychgica aws a re wha t the physiologist is tg
t expain.
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 12/89
2h
nrlokngvrbal
opan parad
Wh we now hve o do is pply he generl opern prdigmo he verl cse We hve he prdigm, on he one hnd, nd speing,on he oer We hve o see i we cn i hem ogeher wihou doingdmge o eiher To do so we wil e he ccompished speer, oservewh he is doing, enumere he wys in which discriminive simuli, depri-vions, nd reinorcers pper o conrol his speech, nd giv pusilereconsrucion o how ps hisory o reinorcemen hs creed hisconrol
Noice h in his cse we re sring wih n ccomplished speer
This is ecuse engineering lerning o l nd minining linghroughou lieime re wo dieren reinorcemen scheduing proems,nd hence require somewh dieren nlyses ur emphsis wil reminon he speer's ehvior, no on his words or wh he mens o sy"There re sever resons or ignoring wods, s such, righ now A word is response y he speer u is simuus o he herer Sying words iscused y dieren hings hn hering words; one doesn' simpy run helngugegenering mchinery in reverse ger when one is isening Buconcenring ou enion on words s such woud end o cuse us o y-pss hese imporn issues; hence our dismiss o he prolems o wordsnd menings or he presen
he n eockn g ebal opean paadgm
We re restricing ourselves to n ccount of vocal vbal baviott s lkng There is verl vocl ehvior
SnoringspngMing sounds such s oo" or ghWhslingHumming unes
The lst wo re opern s you cn see when someone is clling xi,whislng ixie, or owin you how he melody goes he ohers re noopernt ecuse hey do no depend upon ps hsory o reinorcemenfor her exsence They re no condiioned the wy he ls wo re
There is nonvocl verl ehvior
Wriingt hs diferen voculry re se o conrolling simui,set o reinorcersjus ou everyhing
esureFcil expression
TypingI Typeseing
h specil lnguges" of owers, gems, nd smps
l graPhY
\jl o hese re ineresing nd worhy o urher sudy, u hey re no
germne o our min purpose here
THE WRONG PARADGM
I s unliely h nyone oher hn he Pvovin psychologissof e Sove Union now ls ou verl ehvior in he olowing wy,but some people used o e ugh he wrong prdigm" or verl e-hvor in her psychology courses We now now h verl ehvior isopern ehvor u one me when opern ehvior ws no welundersood verl ehvior ws sid o e cquired y he process illus-tred in he following prdigm
C ATR C - CR R
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 13/89
e nerloc ng erbal operan paradgm
As you know, ths s called Pavlovan or classcal or res ondent condtonng t conssts of the parng of some stmulus that doe snt elct a reflex response (the CS) wth one that does (the US) After a numer of suchparngs, sometmes as few as one, a response, the condtoned response(CR), whch s the same as or smlar to the uncondtoned response (UR),wll come to e elcted y the presentaton of the CS alone In Pavlovsclasscal example, the U S was meat and the CS was the rngng of a ell TheUR elcted y the US was savaton, and the parng of the ell (CS) and
food (US) was suffcent to estalsh a CR of salvaton to the CS (ell)aloneWhen the Pavlovan paradgm s appled to the veral ehavor stua
ton, somethng lke what s depcted n the followng llustraton s supposed to happen
FRT
� (PT
"G"
( C F AT
LATR
G
C F H
�
' G"C F CH
t was clamed that the chld learns to say dog" when he sees the anmal,ecause the sght of the anmal s pared wth the sound of the adult sayngdog " That s, the voce of the adult sayng d og" s a US whch elcts thechlds U R of sayng d og, " and the mere parng of the CS, the sght of thedog, wth the US wll cause the chld to later emt the CR dog" to the CS,the sght o f the dog Well, yo u can see where followng that road would takeus The Palovan vew of veral ehavor suggests several very duousnotons
One It was sad, at one tme, that o s a sgn for," symol-zes, sgnals," stands for, or represents" the mssng anmal Thats, t was sad that we respond to the name as we would respond to the thngThat pont of vew runs nto severe prolems, for we nether feed nor pet thesound, nor does t scratch us or make us sneeze Ths prolem s not lmtedto dogs, for we do not respond to the sound as f t were the thng tselfwhen someone says skunk," ce," ee," uncorn," square root oftwo," red, ed," free eer," or orange"
he n erlocn g erbal operan paradgm 3
Two The Pavlovan paradgm requres that the adults veral response ( use the symol Rv for a veral response) functon as a stmulus (Iadd another symol to ndcate the stmulus functon of a veral response,that s, RvS) and licit the Rv dog " It s clamed that we learn to name"ths wy The illicit term n the Pavlovan account s elicit og" can serveether as a Rv or as a dscrmnatve stmulus, ut t s not an elctngstmulus (US or CS) for talkng The Pavlovan paradgm s not applcalehere ecause veral ehavor s not elcted ehavor It s not made up of
the responses of reflexes (USUR relatonshps) There are no real, dentfale, empral, ologcally determned uncondtoned elctng stmul(USs) for talkng Hence, t wll e wser for us to gnore the Pavlovan ap-proach and us the operant condtonng paradgm nstead
INTERLOCKING PARADIGM
We may now look at what Sknner has caled the inte/ockinvebal beavio paaim Ths allows us to see n a most general andsmplfed way how the varales (S, epn, Rv, S) arrnge themseves na veral epsode The justcaton of the cam that the tems n the reworld are functonng n the way that we say they do s most gener ndsketchy at ths pont Ths s a promssory note of sorts Much more wcome later; I l pay off ths telectual det and tghten thngs up But, letus look at the next llustraton to egn
FR A "M AY HA A HT G"
NXT A CERTANLY" (HAN N T A
NXT A liT HANK (TA K HT AND TE
NXT A AN TM
It s not g reat lterature, ut you mg ht well hear somethng ke t n
flm of the reast stye, ecause t s veral ehavor No deas, nformaton,
or symos are dsplayed, ut nnetheless t s veral ehavorur prolem here s to app the smpe operant paradgm shown n
page 8 to ech response dspayed n the precedng usrton To do th,
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 14/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 15/89
e interocki n erba operant paradm 6
t s clamed that n some specal sense I do, then t must e clamed that mylver knows not onl chemstry ut algera, calculus, and quantum physcsIt then seems to me that sayng that there exsts some specal knowin o thssort s just usng the word n such a pecular way so as to turn the whole dscusson nto a deate aout the meanngs o wo rds, rather than an nqurynto what people can or cannot do
A retracton o sorts (not really): ur example s not asolutelygeneral It s a nce easy one, and thus a good place to egn And, t has
that one, good, ovos renorcer, the hot dog Some examples are harder:A: Hw are u tda sr?
B: ne tans.
We wll have to get through several chapters eore we have dscussedenough o the analyss so that we can handle that on
Here are some mplcatons o the nterlocg vra paradgm T eydon't sound lke what we have een consderng tus ar, ut thy ollowrom t
Vebal beavio ocial beavio Snner has dened SCIALBEHAVIR AS THAT EHAVIOR WHOSE REFRCEMENT IS IATE Y
OTHER ORGANISM. course, most o what socal psychologsts study s alosocal ehavor n ths sense But a lot o other thngs are also socalehavor n ths sense, ncludng veral ehavor The contngency (thearrow), etween the Rv and the renorcer s always mantaned y anotherorgansm The other anmal may e a pet dog or cat, ut there must eanother organsm That's ecause nature or machnery does not renorceveral ehavor ranges don't jump o the tree nto your mouth when yousay Please" or I'm hungry" Never mnd aout the yodeler who startsavalanches y warlng, or the opera snger who cracs hotel wndows asast as terrorst oms Frst o all, t's not very mportant Second, andmore mportant, although these are vocal ehavors, they are not vocal
veral ehavorsIn veral ehavor, as n all cases o socal ehavor, a person can ncrease the lelhood that he wll get renorced He does ths smply y renorcng the ehavor o the REINFRCEMENT MEIATR (THE FELLOW
WHO ACTS AS A REINFORCER ISPENSER) whch produces renorcements orhmsel I I have apples and you have ananas, and ananas are renorcng to me at the moment ut apples are not, and, apples are renorcng toyou, the I can renorce you or renorcng me I smply gve you an appleeach tme you gve me a anana Sounds lke Economcs I, doesn't t Thesocologst and socal psychologst George Homans thought so and haswrtten a very good ook aout t
But wat aout the Thans" ad the Any tme" hey ae not the
e nterockn erba operant paradim 17
same as the ananas o the rut exchane o the mmedatey precedng example Rememer that these are conitione einfoce and, as we wll seeater, they have the addtonal unctonal role o settng thngs up or theuture Each speaer, y dspensng such condtoned renorcers, ncreasesthe proalty that the other wll comply n some uture exchange
Even the renorcers work, an mportant eature o socal ehavors that renorcement medators are not perectly relale The herer may ormay not have any coee, cgarettes, m l, e er, car s, or money People are
less relale than machnes (Come now, you know that s true How manytmes has what's hs name who sts next to you n class een out o cgarettesor candy ars or stcs o gu when you wanted one, and how many tmeshas the machne een out or not worked Have you really counted) So,veral ehavor, lke all socal ehavor, s ntermttently renorced And,lke all ntermttently renorced ehavor, t ecomes very persstent I renorcements aren't mmedately orthcomng, the ehavor contnues onand on (the naggng we) When the machne s the renorcementmedator, renorcements are not delvered, emotonal ehavr ensus;people ck t, ht t, scream at t They ehave just le the rat who s TIGUISHE (NOT REFORCE) ATER HAVING EEN CONTINUOUSY RE
FORCE (REINFORCE EACH TIME)
SE OF U NTS OF ANALYSS
A proem whch I have een avodng out now to e squary: aced. To make ths sort o analyss wor , we have to isolate sngle vents orems n veral ehavor We av to determne unts o anaysis We have
to decde when somethng has happened The pigeon's pck is easy It as a
dscrete unty, and t recycles and repeats. But lsten to My I have a hot
dog" when you say t to yoursel It ows-there are no stops or dscret
unts, there s no recycng , there aren't any pauses or an spaces It seems
to e just a stream Spoken speech has no pauses etwen words or doulespaces aer colons What then s the utterance? Is t one thng, an ntact
sentence? Is t sx thngs, words? Is t a whole lot o thngs, phones (spech
sounds)? Here s the gst o the answer to these sorts o questons
Dstngush etween the response (Rv) and the veral operant Rv's exst at
many levels o complexty; we wll examne some Operants ar unctonal
unts, unamguous and unequvocal We'll see as ollowsWe can egn y concentratng on responses and loong at several o
the possle ways o decomposng any utterance (Rv) nto ts parts I wetake nto account only the structural propertes o veral ehavor, we willnd that the work has already een done Ths was orgnally tan to ethe provnce o structura lngustcs (another dscplne concerned with
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 16/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 17/89
he n elocking erba opean paadigm 0
Wd: Other causes seem to roduce resonses at the level of words.n the followng examles, let me resent the cause, and you do theresondng
Questons sometmes oerate at ths levelWho wrote the Gettys burg Address?
Who s the resdent of the Unted States?What s the catal of France?But other varables, not questons, do tooAs wse as an . . .He nvented the electrc lght bulb.Father ChrstmasMama Bear, Paa Bear, and Baby . . .
Phrae: These are easy; agan, I'll sul a conroll varable adnotce how your resonses are multword hrases
Snow hte and the . . .A enny saved s a . . .He who hestates . . .Hon sot qu mal . . .
That last one sn't an effectve varable for the same knd of verbal behavoras the others, unless you have had some secal ast hstory of renforcement called French lessons.
Bi ne: Some causal varables are ecular. The class bell roducesffty mnutes of talkng by me. There are some eole for whom Do youknow the Gettysburg Address? s a causal varable such that they say
Fourscore and seven years ago . . . and don't ause or sto untl . . .and that government of the eole, by the eole, for the eole, shall notersh from the earth. One 's audence has certan controlln effects. Thevocabulary I use n my classes s qute dfferent, as s my seed of talkng,from that occurrng when I dscuss the same subject matter wth the otherrofessors at the lunch table. The language the mnster uses n the churchservce s very dfferent from that used n hs bedroom, even though hessues under dscusson may be the same.
We may conclude, then, that snce causal analyss s our goal, we cannot fx the sze of the resonse unt; we must let t vary from small to large.t wll change from the smallest and smlest unt to the largest and mostcomlex, deendng uon the controllng varables resent. Ths leaves usfree to dscover whch varables are oeratng at all levels , snce they exst at
he nero cking eba operan paradim 1
all levels. Ths lets the data decde. Put another way, the sze of the unt ofthe deendent varable for verbal behavor s no t a structural queston; t sa functonal one.
Ths all tells s that the arorate unt s not smly a resonse but afunctonal unt, a causeeffect relatonsh, namely, the verbal opeant
A vebal operant s merely a dsoston (tendency, lkelhood) to
resond n a certan way to a certan state o f affars because of a ast hs-tory of renforcement. Ths s just lke our nonverbal examle of the -geon's keyeck. Look at the frst llustraton n Chater 1 agan.emember tat an oerant has three comonents (1) antecedent causalvarables (dervaton or dscrmnatve stmul), (2) a resonse, and (3) acontrollng relatonsh between the frst and the second (the arrow). Themembers of each end of an oerant are classes, and a certan amount ofvarablty s ermtted among members of the antecedent or the resonseclass. So, varablty n sze, color, fur, and so on, occurs among the ds-crmnatve stmul (anmals) for the Rv dog. But f one dscrmnavestmulus s an anmal, and another s rntng or wrtng on aer, we havethe stuaton dected n the followng llustraton. Here we have two oerants, wth only one R v, whch s shared by both of them.
DI <W DOG
R
D G v OG
We can have some, but not much, vartn n h sonse sde. fthe resonse vares a great deal, as n th hr css deced n the next
llusraton, we get nto several dffern oerns (tre shon n the llus-taton) under the control of the sm cusl vrabl. E ach oernt n theseast two llustratons has a dfferen rnforcement hstory; ach s a df-ferent tendency to resond and s dnt srength.
D f , I I'
OG
v RE
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 18/89
e ne ockng eba opean parad gm
Although the antecedents (dervaton or dscrmnatve stmul) andthe consequent v of an oerant can vary, ther varaton wthn a gvenoerant s not great, and any changes n them tend to occur slowy. herefore, what vares rncally n oerants, and over a wde range wth surrsng radty, s the strength of the dsoston (tendency) to resond.That s, what changes s the lkelhood of te v iven the controllng var
able. Ths lkelhood (robablty) can vary from 0 to 1 t exsts due to aast hstory of renforcement and vares (s greater or lesser) deendnguon the number, sze, qualty, and schedule of renforcements that havebeen obtaned n the ast. f a resonse has been renforced n the resenceof a dscrmnatve stmulus many tmes n the ast, ts robablty of recurrence (strength of the oerant) wll be close to 1 . For examle, consder thedscrmnatve stmulus What s your name? f the resonse has occurredonly rarely n the resence of the dscrmnatve stmulus, and hence hasbeen renforced only rarely, the strength may be close t Consder the dscrmnatve stuus What was Mrs. Wahgn's maden ame?
INDCES O OPERANT STRENGTH
Sknner argues that the verbal oerant s what we need, what ewant to know about. f we knew a seaker's ast hstory of renforcement,we would know hs verbal oerants, and when he would seak an what hewould say, and relatvely how lkely he would be to say t n those crcumstances. n rncle the ast hstory of renforcement s knowable; n ractce t sn't feasble. That s, n rncle we could follow a erson all thetme and record everythng that haened to hm snce hs brth, but nractce we haen't the resources. In ractce, t s hard to measure terobablty of a v n a gven set of crcumstances. ate of reetton, the
deendent varable for the geon, won't do. We are shaed by our renforcng communtes to seak only once. The frst tme s nterestng; thesecond and thrd tmes are borng, and unshment ensues.
n ractce, then, the best that we can do s to resent a susected con-trollng varable, and record the resultng verbal behavor, f there s any.Whatever gets sad s the v of the strongest oerant havng that controllngvarable. There are a number of ways t do ths. The wd aciatin testgves us the mos t robable v when the dscrmnatve stmulus s a sokenor wrtten word. The ppite test and the varous vcabulay tests gve ussmlar data wth dfferent sorts of dscrmnatve stmul. By usng a tachstoscoe, we can resent wrtten words brey, and measure robabltyby countng the frequency of bef exosures requred before a v s gven . Or
else we can rogressvely lengthen the exosure tme, and measure robablty b the recrocal of the tme elased before a v occurs.
e ineock ng ebal operan paadgm 3
Above all, there s no ntensve roerty of the v alone, excet tsoccurrence, that s a relable ndcator of oerant strength. We mght havesuosed that some roerty of the v tself would tell, but the followngexamles show that the desrable degree of relablty s far from resent.
epetitin: o, no, a thousand tmes no seems stronger thanWell, noo. But, a v emtted just once s no necessarly weak: Please
dear, just ths once, or am delghted to accet the nomnaton to byour canddate for resdent, or Take the money; just don't shoot me.Ludne: Strong reflexes have large resonse magntudes, and a
screamed Rv may be art of a strong oerant. But, one does not shout hsame when ntroduced. Sweet nothngs are usually whsered, but maybe very strongly determned. Prsoners lottng an escae do not yel thelans to the guards. Mothers, when vstng the lbrary wth ther chldren,hss nstructons to be quet.
apidity: (Ths s not the same as rate. ATE IS THE REQUECY O
RPETITION O A RESPOSE PER UIT TIME; APDITY IS T RECROCAL O
HE DRATO O THE RESPOSE .E. HOW AST EACH RESPOSE IS Badnews s delvered, unavodably, but oten n a haltng manner.
Combnng all these onts nto one grotesque examle, we se thateven though hs behavor s strongly determned, the olceman eakquetly and slowly, rather than runnng u s creamng at two hundd woda mnute: YOU WFE S DEAD; YOU HOUSE AD KIDS AREALL BUT UP; ALL BUT UP, ALL, AL; EVEY , EVERY-THG; ALL BUT UP, ALL BUT UP; ALL, ALL, ALL; DED,DAD, DEAD.
EPETTO, LOUDESS, AD APIDITY RD-TO EFFECTS, SUALLY The oeant gv th pe omthng tosay. f the controllng relaton between the antcednt nd the R strong, t wll be sad. But sepaate vaable whch ct n a uppl-
mentary way determne the se oducton efct. A thee apect oseakng are cause, but by dffeent thng We wll dcu th n uchmore detal later.
Sknner's nventon of the cumulat ecodr made recordng the
Occurrence of the g eon' keyeck nd ht pceded and ollowed t vey
easy. n recordng verbal bhavo, e h to decde what tm ar to be
used n the descron of the event tht ntnd to explan. Her rec-
son and fdelty, whch, thanks to tchnolog, ey come by, may b
mstaken for "more scentc. Rcodng ca b dtaled as you lke.
We can reroduce the acoustc efct th mot complete fdty by
usng mag netc tae recordng We c m cod of hch mucle
move by the rocess called ectromyog py W cn me a phontc
tanscrton, usng the nternatona phonc phbt o one o t equvalents. We can wrte down drect quotto "Yo, , ae a aca,
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 19/89
he n erlockin g erbal operan paadigm 4
the court reorter does. Or we can use ndrect quotton (He sad thatJones was a bad eson), whch would be a wrtten descrton of themort of the seaker's utterance.
Drect quotaton s suffcent for ou uroses. I names the behavoremtted and descrbes t, because the name actually resembles the thngnamed (erhas ts s the ony case n whch t does) . Indet quotaton s
too far from secfyng behavor. A seaker whose behavor s ndrectlyquoted may have soken n another language, or gestured, or used dffeent words. The other methods of recodng generaly rovde moe detalthan we need.
Descbng the antecedents, causa varables, n oerants s ess of aroblem. The causal varables are ony dervatons, aversve stmulaton,and varous sorts of dscrmnatve stmu. But they fal nto a number ofsubclasses whch deend uon dffeent knds of renfocement contngen-ces We wl look at how each of these antecedents affects Rv' s, and how,n terms of ast hstores of renforcement, go control. To use kner'stermnology, we wl look at
Dervaton and aversve stmuaton as arts of mand;Dscrmnatve stmul for tact;
The audience as a dscrmnatve stmulus;Dscrmnatve stmul fo tetual;
Dscrnatve stmul for echoic behavor;Dscrmnatve stmul for intraverbal; and, fnaly,Multiple cauation and autoc/itic behavior or how the vaables con
verge to roduce real seech.
mands
Because dervaton and aversve stmulaton dffe fom thother tyes of controllng varables for verba oeants (whch e dcmnatve stmu of one sot or another), t s wse to ook at thm patSo n ths chater we wll look at deraton and avesve stmuatonSAV and the mand as a class of verba oerants. We may begn wth adefnton.
DENTON AND PARADGM
A MAND I S A VERBAL · OPERA NT WHOSE, CONTROLLG VLE I S
A DEPRI VAION OR A N A VERSIVE STULUS, A ND WHOSE R SPECIFS I TS -
FORCER. Secfes here does not ml meanng n any metahysca
way. Secfes here s equvalent to s the conventona name of, but,
moe recsely, accordng to Chater 4, secfes ts renfocer s equ-
valent to would tact ts renforcer, f the renfocer were resent
The word mand s a neologim (.e . , a new word made u for the occa
son) and s related to such words as command, demand, an mandatoy.
Mand tact autoc/itic echoic; intraverbal and textual ae al neoog sm
that Sknner made u and used n hs book Verbal Behavior. But, and w should be efectly clear on ths, our use o f such tems n ths book df
somewhat from knne'.
6
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 20/89
ans 6
Our defnton of ncludes seces ts renforcer. Ths meansthat the v of a mand tells the renfocement medator (the ndvdual whofunctons lke the geon's food dsenser and gves the seaker the renforcer exactly what he must do to renforce the mand. enforcementmedators are, by vrtue of beng members of a vrbal communty, condtoned to resond to some v 's n hghly secfc (conventonal ways whch
are renfocng to the seaker. The next llustraton shows a aradgm of thenteracton between the seaker and the renforcement medator n the caseof mands. n ths nstance there s a charactestc conventonal covaratonbetween the form of the v and the form of the that the medator emts.
AR � R �PAKR
RNFRMNT
Pt
-i- -
-
MATR O R
Our defnton asserts that whenever ths s tre, the controlng varable for the v wll be ether a dervaton or an aversve stmulus.
et us look at some smle examles. The followng could be mands.But, beware! We never know from observng the v what the controllng varable s. o v tself, or any of ts ntensve roertes, sdagnostc of ts contrllng varables. We wll go nto ths n detal after afew ages, but now the examles
Respse May I have a hot dog
ced tea, lease WaterQut tPut away your rubbes Chase your talGet off my foot
RefrcerHot dog
ced teaWater comesAnnoyance stosubbers goneDog runs n crclesGets off
f you know what the renforcer was, you know what the renforcementmedator dd. And, wthn broad lmts, f you know what the renforcer s,you know what the mand term for t s. Thus the renforcement medatorwho hears these v forms enforce. If he does, he makes a dfferent
resonse n each case. Hence, the seaker's v secfes the renforcement;the renforcement characterstcs covay wth the v.
ans
DEPR VATON AND AVERSVE
STMULATON BECOME CONTROLLNG
7
Ths may, at fst, seem mysterous because normally th nforcement medator doesnt sto to fnd out f an arorate state of der-vaton exsts or f an aversve stmulus s affectng the seaker. H maks hs
comlyng resonse solely on the bass of the seaker's v, as shown nthe aradgm for the mand. The renforcement medato may not complyon a gven occason, but that doesn't destroy the relatonsh. Fo whn hdoes, he s usually ndffeent to or unaware of the state of the saker. Onof the occasonal excetons occurs when the seaker s a chld who says,Hot dog, lease. Hs mother may nterject, He's not hungry.
So, how does the arorate antecedent condton get control? Theanswer has two ats whch are related, but a dgresson on the nature ofrenforcers must come between the two arts.
enforcers secfed by mands may be of two sots One sort deendson dervaton for ts momentary effectveness; the other sort deends onthe resence of an aversve stmulus for ts momentary effectveness.
Deendence on dervaton s true of some ostve uncondtoned n-forcers (what w denote by S. The next llustraton shows two rototys.n both of these cases the seaker must be devd o else th unco-dtoned renforcer wll not work, that s, renforcemnt doesn't occu vnf the renforcng stmulus s resented to the seakr. He won't ngg nconsummatoy behavor (eatng or dnkn unlss dpvd, nd h dosso only f derved. And, the medator's behavor s nfocng to thseaker only f the latter engages n consummato bvo th.urthermore, th e occurrence of that dervaton w, n th utu, ncrsthe robablty of the emsson of that v. Ths s ht Snn nd I manb controllng behavo.
PEAKER R "M:Y R AOHO
EP
SPE AKE R: Rv OPORN LEAE"� SRPOPCORN
P
REM NDER O HOW RENORCR WOKWe have to embark n a dg sson bot sp
ther classfcaton and mechansm of oeraton.Som people m to
stll qute confused about ths. We may dvd ths to
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 21/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 22/89
ads
A MA E FF S ON OOT - Rv MY FOO"- S TERMINATES
S EAR - v " VE UP - ERMAES
-'DRA IN- v
"OPEN THE-S
R-TERMNATES UMBREA
S
r-'D SCK OU
R
" Sr
UE - SP EMAESSr
- 'D SIGHT OFR
" Sr-
HYPODERMIC - V NO NO, NO � TERMNATES
SYRE
Sr UBBES PU AAY r- F v YU UBBER"
�
TEMINATES
30
We may thu conclude the frt art of our anwer to the queton ofhow dervaton and averve tmulaton get control a follow. They do owithut the medator' arrangng for them to do o. He doen't have toworry about t. If the medator attemt to enfce at any othe t, tdoen't work. Such other tme get no control. We may ay that the law ofychology take care of t. Et her dervaton or averve tmulaton mutbe reent for the attemted renforcement oeraton to work. d, anythng routnely reent at the tme of renforcement ge control over tereone.
ow, here a ecularty of mand, and the econd art of theanwer. Nthin ele i utinely peent Dervaton and averve tmula-ton have eentally no cometton from envronmental (dcrmnatvetmulu) control. The next llutraton hel u comare mand wthnonverbal dervaton deendent behavor. The v May I have a hot do gdoe not covary wth the envronment. You can do t utar, owntar,n the wmmng ool, nde, outde, and even ude down n an arlane.There may or may not be a hot dog n vew. he mand work n a wde var
ety of uaton. Verbal behavor very free; the eaker can eak at anytme and need no ecal envronmental ro. Even though an audtor lkely to be reent when eech occur, h charactertc do not determnewhch v wll be emtted. or doe the v or mand that occur revealanythng about the eaker' external envronment; t doen 't vary wth t .How unlke the nonverbal alternatve to the mand th . The hot do g n therefrgerator govern the robablty, the locaton n ace and tme, and theform of the reone to t. By contrat, n the uret cae of the mand there abolutely no outde nuence, uch a n the olated deertdwellnghermt' cry, Water.
et me ue a warnng. Don't be temted to argue tht mand are incecretly controlled by tmul arng from the abence of the tem. That
, don't try to clam that the dcrmnatve tmulu of no hot dog controlthe v hot dog. f you were t make uch a clam, you would ut yourelf
ads
VBA CAS
P
+ tk S
PU PAS VSBLE
VBAL CAS
P
R
MAY AVE S
- A T
31
n an moble tuaton. emember that you agreed that a hot dog adcrmnatve tmulu for the v hot dog, and now you want to y thatthe v hot dog alo trong n the abence of the dcrmnatve tmuluof the hot dog. If both controled the Rv hot dog, a eron would be nthe oton of ayng hot dog at al tme, whether the hot dog werethere or not. Alo, many dcrmnatve tmul are abent at all tme, ytwe only mand ome abent tem at ome tme. So, abence a uch cannotbe the ontroler of the v hot dog when no hot dog there.
Sometme when there are no gant anda, nor Sherman tank, noquare root of two around u, ether you or I may eak of them. The
roblem of talkng about, or namng, abent tmul are very erou. Weagreed that the mere abence of the tmulu n't uffcent to roduce thev. Some other varable or varable mut be exercng control. We wlhave to look at th roblem n more detal later , but fo now we can ee thtdervaton the ource of control n ome ce, uch a when the art of a mand.
Mand are mot lke nonverbal oerant, whch why e tate wthtem. (1) They are controlled by dervaton or averve tmulton; mt-vaton mortant. (2) They have narrowly defned conequence; the Rvecfe mlk, mlk what roduced, not beer. Mt nonvebal eantare th way too, n the ene that a narrowly demted et of conequence(a food ellet, thrty econd wthout electc hock, etc. ) are correlated wthreone of moerately delmted effect and toograhe (keyeck, bar-ree). (3) But unlke the nonverbal oerant, mand are frmally nde-endent of the eaker' dcrmnatve tmu.
DIAGNOSING MANS
Can we ay whch v' are mand and whch are not? Tech-ncally, and trctly eakng, a mand a verbal oerant, and a uch re-qure both a controllng varable and a v. So, no Rv n telf a mand.But we can do better than jut th. Some v' can be mmedately ruled outa not beng art of mand. For examle, however, 3 1417 " andtock market crahe are not mand. But there no v whoe ocur-rence unquely and relably ndcate the occurrence of and. Th be
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 23/89
Mands 32
cause all 's that occur in mands are also parts of verba operants that are
not mands. Just look at our preious examples of mands and figure out
cases where the Rv' s wouldn't be parts of mands. Here is a hint: If you readthem aloud, the 's are not mands; they are parts of textuals (to be defined
and explained later). But, for any R that is a part of more than oneoperant, an inference can be made from the context in which it occurs. We
all do this, because we are all reinforcement mediators. Thus, if we hear the hot dog," we may ask, Is it a mand? It could be, and to judge, weexamine the context in which it occurs. If said in the presence of a endor at
a baseball park, it probably is; if said just after the discriminative stiulus
What did you hae to eat," it probably isn't part of a mand.
SUPPLEMENTARY STRNGTHENIN G
AND THE IMPUR MAND
Because a Rv, for example the Rv hot dog," may be part of amand on one ocasion and not on another, diagnosis is difficult. Further
more, it seems that the Rv hot dog" may occur on one occasion or two
reasons: (1) depriation may be in effect, and (2 a disriinative stimulu,the hot dog itself, may be present. This is called SUPPLEMENTARY
STRENGTHENING. IT OCCUS I TWO O MO CONTOLLG VAIABLS
TWO O MOR OPANTS THAT SH TH SAM RA PSNT AT TH SA
T; TH FFCT ON T POBABITY OF TH Rv IS ADDITIV. Therefore a
may occur for seeral reasons , one o f which is deprivation, and theothers of which are likely to be discriminative stimuli. Of course, the two ormore controlling variables could be all depriations, or they could be all dis-
criminatie stimuli.TO T XTNT THT SOM VAIBL OT THAN DPIVATION O
AVSIVE STULATION IS XTING CONTOL OV TH Rv (CONTIBUTING TOITS POBABITY OF OCCUNC), TH MN IS SAID TO B IMPURE. Wemy look at some examples of mands of differing degrees of purity. I' ldescribe them; you point out the depriations and the discriminativestimuli.
Pure mand: A man who hasn't eaten for a while lying in bed alonesays aloud, Pepperoni pizza with extra cheese" ; or a dehydrated hermit alone in the desert says, Water. "
Impure mand: Some food deprivation, and a person who usualy hasfood-"Do you, by chance, Christopher Rbin, hapen to have ajar of honey?" Aer a large meal-That looks g ood; I' ll have a
large piece of the strawberry cream pie, please. Ths next one sn 't a mand at al: D Pooh, wha hae you been eat
ng?" + Dthree emy qua�ars Rv (n a si oi Honey."
Mands 33
Nearly all mands controled by aersie stimulation are impure be-cause an unconditioned or a conditioned negatie reinforcer is lso ikely tofunction as a discrimiatie stimulus. The elephant that stands on your footis a isual discrminative stimulus as well as an unconditioned ng aive ein-forcer. There is just that ery large, public discriminative stimlus whichcannot be gotten rid of. Pure mands are, no doubt, very rare. Usl som-
thing else shows up. That's OK; pure mands still exist occasionl. Wedefined the pure case and showed that it was possible. If we find ohcausal ariables at work, so much the better. We are in good shap if b-haior has several causes rather than none at all.
Something like the typical impure mand can be shown in he pigeon'sbehaior. W e may keep him i n the pecking apparatus all the time an feedhim any time he peck s. The keypeck is, then, functionally equivalent to the
human's Food, please." But notice, the keypeck must be performedn a special place. The key is the bird's reinforcement mediator, and o a
certain extent goerns the form and locus of his response. eal verbabehaior acts at a distance and is free of environmental control over is
form.
WHY D SPEAKRS MAN D?
For all operant classes, such as mands, we will have o explain
why the speaker and the hearer do what they do. We have to exhibi he wosets of reinforcing contingencies. The reinforcement coningencies fo hpeaker are obvious and familiar in their dynamics. Manding is disincly
.interestedbehavior. The speaker gets something out of it. We now what hegets by seeing which Rv he emitted. Th einforcement per nit of physicawork is amazingly large. The magic of words was een moe imprssive be-
fore the adent of modern machiney, but even so, the advantages ae still
there. Compare the Rv Give me a loaf of bead, please" with all theresponses required to grow the whea, mi it, mik the cow, mix the dough,and bake it. We can get reinforcers via mands tha would otherwise e
unobtainable. We merey mand bread," toohpaste," colo TV,"
Mercedesenz. Some of the specialization of our society i s madepossible, in part, by the ability to mand.
Probably, the more important question is, why don't people mand
more than they do?" Bossy people and tyrans mand a lot; so did God,according to the Old Testament. The answer forms our next section.
WHY DO AR R COMPLY?
There are limits to the hearer's probabity of complying (ds-
pensing reinfoes). Eeryone cannot mand everything becaus hn n
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 24/89
ands 34
one would do anything. The roles of mander and reinforcement mediator
must reerse eery once in a while. Al the mands don't go into a few hands,because the complier must be reinforced for complying. As Skinner has put
it, the mand is an imposition on the hearer. He spends more energy than the
mander does; he oten oses property. There is ony one way to keep himdoing a thatreinforce hm for doing it.
Complying by the hearer increases the probability that the speaker will pay him back" when each has the other's job to do. Not every speaker willdo this eery tme. But, overa, compying increases this probabiity.Howeer, the role reersa s likey to be delayed, and thus would be a veryweak reinforcer. Therefore, the mander g ies immediate reinforcement in the form of grattude or promises or g eneral acknowedg ment of indebted-ness. Tese, of course, are conditioned reinforcers which are correlatedwith other unconditioned reinforcers. The next ilustration depicts thisinterchange agan, and names some o f the possibe conditioned reinforcers.Obviously, one of the best conditioned reinf es i money. We know hatit can be traded in at any time that its recipient wants to mand. Our onlysupposition here is that " Thank you, smiles, pats, nods, kisses, and so
forth, are functonaly somewhat ike money. If you think the have no simiarity, I sugg est that you consider whom you reifc with Tankyou" -friends or tota strang ers? And I arg ue that one of the characteristicsof friends is that you will be wiing to reverse the roes with them. Thus a Than you from a friend is more valuable than a Thank you from astranger.
MANDE:
R:
"THANK YOUSMLE
D KISSRv MAY I HAV�
,- Sr· HOT DOG R I 'LL PAY YOU BACK"
T DOG IN HAND
I
'L
�T THE N EXT
U
U
�ON Y
Another reason for the hearer's complying is that refusal may g enerateaversive stimulation . The nxt illustration depicts this paradigmaticaly and shows seera examples of the sorts of aersie stimuli that may ensue when the hearer refuses to comply. Compliance is a form of aoidance behavior;its reinforcement consists in stopping the mand (terminatng a conditionedneg ative reinforcer) and preventing the aversive consequences of noncom-
pliance.
ands
PEAT LOUDE
MANDE Y I AV�_ - D_R ;DNVELA OT DOG HINE, AFTE ALL.I
I
I'V DON FOR yOU" - u
- -
PHYS
IC L ATTACK
D Rv"NO MDIATOR
35
In extreme cases the mand may hae explicit colateral aersie stimu-lus properties, as shown in the next illustration. Al mand situations are
probaby combinations of the eements shown in the next to ast and nextilustrations, that is, both positie and negative conditioned reinforcement
are operatie. I suspect that al mands are in some degree aersie to the
hearer, although sometimes the subsequent reinforcement for compyng
may be positie as well as negatie.
MAND
R
R "YOUR MONY - Sr ·D M
NEY
LEAVE v O YOU L
-i D- · D RG� r NDS
SPA A QUAEMANDE. R
"I AN HAD NYHING TO EAT'�r·D R GOESv "Y K DS AWAY
R
"Y FE . • -n- _
1
- u - -
-D -D DIRTY + SMELLY
-
A fourth rason for compying is that mands sometimes consst of
adice. " The hearer may be himse reinforced rather immediatey anddirectly by his own behaior in copying. Skinner has said: Adice is f
of promises o reinforcers . " These R' s impy that the hearer w be re-
forced for complying, as diagramed in the next ustration. When I askforadice, I mand you to mand to me so tat I am reinforced for compyg
MAND
R
STOP"TUN HE
v JUMP""GET ON HS LNEEAD S BOOK
i - - - -
-
O r
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 25/89
ads 36
with your mand. Initations follow the same paradim, such as in the R Coe to dinne. A hearer who gets a lot of these discriminatie stimulimay deelop a disposition to comply with mands kinner has suggestedthat a good mander slips one of these sorts of mands in from time totime." uch mands, along with adice, produce an indiidual who isgenerally likely to comply with all sorts of mands. He does this because ofthose occasional reinforcements that follow complying with the specialadice and initation mand forms
Finally, the hearer has usually been, as kinner has put it, sotenedup." Mands, nowadays, seldom coe in raw form; we request, and do notorder. Bryant and Aiken hae gien many examples of this, and we listsome here:
Please gie me a hot dog"o gie me a hot dog"Would you mind giing me a hot do"Will you please gie me a hot dog" Could you g ie me a hot
dog , pleaseWould you like to gie e a hot dog"
Won't you gie me a hot dog"o you want to gie me a hot dog"Would you care to gie me a hot dog" Might I ask you to gie me a hot dogLet me beg of you to gie me a hot dog"May I ask you to gie me a hot dog"I should be glad if you would gie me a hot dog"Please do me the faor of giing me a hot dog"
May I trouble you to gie me a hot dog"Let me request you to gie me a hot dog" Gie me a hot dog, if you pleaseI would like you to gie me a hot dog" May suggest that you gie me a hot dogWould it be too much to ask you to g ie me a hot dog"
Would yo u be good enoug h to g ie me a hot dog
" Please to g ive me a hot d og
Would you please gie me a hot dog"Wouldn't you like to gie me a hot dog"
I would be happy if you would gie me a hot dog"I shoul lke you to gie me a hot dog"
ads
Wuld you be kind enough to gie me a hot dog"I would appreciate your giing me a hot dog" I should appreciate your giing me a hot dog I would be grateful if you would gie me a hot dogI should be grateful if you would gie me a hot dog"I would consider it a faor if you would gie me a ho o"
I would be eternally grateful if you would gie me a ho o"
3
I cannot tell you how much I would appreciate it if yo wol ivme a hot dog"
I would be greatly relieed if you would gie me a hot dog"
These seem to increase oerall compliance as a k ind of generalized disposi-tion (tendency) They are also full of promise; they gie the complier anopportunity to do some ery reinforceable thing
DPR IVATION AN D AVRSV
STMULATION AS SUPPLMNTARY
VARABLS
We hae already seen that pure mands are few and fr wMost mands are impure because any R may be eoked y mor n o
ariable. So, an ordinarily weak depriation, which exerts so w con
trol oer a R, but which by itself wouldn't eoke the R, my prodc
if some stiulus that also controls the same R is concurrently prn
Additionally, depriation and aersie stimulaion may help ek a R in a c
nonmand situation This is not a case of impure mands Here we strt ot
with depriation or aersie stimulation without any control oer te Rv in
question. (Of course, if reinforcement then ensues, mands may e crad
and end to recur.) Ambiguous stimuli, such as those presented in the vai-
os projectie tests (eg. , the TAT or the Rorschach), produce Rs , for ex-
ample, fried chicken " These R's are usually parts of tacts (soon to
discussed), not mands . Intensifying depriation or aersie stimulation in-
creases the probability of these R's . But, such R's arent parts of mands
because they do not specify their reinforcers and thy are not the bsis for
conentional compliance by a reinforcing cmmunity. They are not undr
the control of depriation or aersie stimulation here. They would not
occur without the discriminatie stimuli present. All that the depriation or
aersie stimulation does is amplify" the controlling power of te is-
criminatie stimuli, that is , increase momentarily and temporaiy t
strength of he arrow etween the SD and the R .
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 26/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 27/89
ads 40
which allows the quesioner, the mander, to act so as to be reinforced bydeprivation depenent stimuli or the eliminaion of aversive stimulation, asin Where is the men's room?" Cursing is also a form of manding underthe control of aversive stimulation: Go to the devil." Gay Wilson Allensaid that in he William James family a favorie mand was, May you al-ways have lumps in your mashed potatoes."
tacts
DNTON AND PARADGM
A TACT A RBAL OPRANT HO ANTCNT N-
' RONMNTAL NT OR TAT OF AFFAR (CRNT TL N
HOE RNFORCNT HA BN CONTGNT PON A CONTONAL COR
REPONDNC BTN TH CRNAT TL N T R, THT
T PRNC OF THE CRMNAT TML N CHRCTRTC CON
VENTONAL R FORM (TOPOGRAPHY .This definition says ha here is a endency fo spec o chng a
objects and events in te speaker's envionmn cange. Cerain Rv' aconventional for cerain stats of t environmn; if bot are prsen, ehe reinforcement mediator reinfocs Noce tat re rnforcemmediaor as been conditioned to eiforc sm verbal beavior not n basis ofth Rv alone, as e did fo mads, no n e a of dicimative stimulus, but simply because tre is a match or correponenceaccording to th verbal commniy's conenions bwn Rv an maspect of he peake's environmen. f tere is o sc corspodc, mediator dosn' reinforce. T product of te dffeial ifcmcotingency is a tact Tacs are opants ude dsciminav conrl; ar sad o e obcve." T conrol is almot toally b scmastimul; the beavo may be sad to be disnerese" o nslf " T
41
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 28/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 29/89
acts 44
RINORCMNT OR TACTING
A reinforcing community consists o f seeral people who tend tobehae in the same way They will engage in reinforcing behaior whensome sounds occur in a certin enironment That is what we meant by aconentional match between a discriminatie stimulus and a R That
interaction is paradigmatically shown in the illustration on page41.
To theextent that someone reinforces, a tact is created or strengthened Becauseany sound can be made in any enironment, the precision of the match between the R and the discriminatie stimulus is us ually required to be fairlystrict But some deiance is tolerated in shaping up the behaior of less thanfully conditioned speakers, as shown in the next two illustrations The first
) UTOMOILE . CR"DTRCK . CRDUS - CR"
SPAK FIRE ENINE . "CRMULNCE . CR
DMOTOR HOME . "CR.
- AT ?
° R"2 ) SPAK CR - v "FR"K "CH"
- RHT"
Sr
- � -YES" OOD
TRUE"HTS RHT""OOD OY"
-
AT
illustration shows that when the speaker is a child or a foreignera person learning to talk any one of seeral discriminatie stimuli will sere in acnventnal match with the R car " That is, the reinforcement mediatorwill reinforce when this speaker makes that one R in the presence of any ofthese discriminatie stimuli The second illustration indicates that someariation in the R's form may be allowe That is, R's that do not meetthe erbal community's criteria for conentional correspondence between
the discriminatie stimulus and the R may be followed by reinforcement ifthe speaker is discriminated by the reinforcement ediator as not being afuly conditioned speaker a person who is ust learn
acts 4
Especially in the case of children, the reinforcement commuity willaccept less than exact correspondences as occasions for reinforcemet Fur-thermore, the R form will be allowed some ariation, initia The com-munity acts in an pportunistic way; when an approximation t te esiredR form occurs, or the R occurs in the presence of somethi ike te di-criminatie stimulus, the reinforcement is deliered Thin are lee p
later Once you get the speaker going, it is easier to sharpen up either te Rvor the correspondence That's the essence of shaping"These processes are not limited to the interactions of chidren an
adults Learning to talk" is not exclusiely child's work Tacts are acquired throughout life Educated people learn many tacts aer they aresixteen, and students in medical schools and graduate schools spend most ofthir time in their twenties learning tacts No r is adult acquisition of tactinglimited to the professional experts The man from New York City has tolearn some new tacts when he gets out West, as shown in the next illustra-tion Many of us in the oerthehill generation learned to tact sch thins
CW
CW v Hv HNE v HEI
as grommet, gantry, apor trail, command modle, transistor, rille, laser,
and computer long after people had stopped teaching us to tal and we were
Qer our deelopment and into our senescence.
By definition the R in a tact is controlled, in both its form ad its
probability, by a discriminatie stimulus t is objective. Pre tacts do exist;
other ariables hae been kept out TO SY THT A TACT IS PURE MENS
THAT THE REFORCG COMMUNITY HS BEEN BLE TO REINFORCE TCTNG
WITHOUT GIVG T SPEKER'S NTECEDENT MOTIVTION (DEPRIVON OR
AVERSIVE STULTION) CONTROL. Wolly disinterested tacting does occr A
horse is tacted as such whether the speaker is hgry or thirsty o r cold or
wet or, as N R Hanson said, s sitting there, fat, dumb, and happy "
Objectie, disinterested tacting i s created by aoiding the use of an
reinforcers that depend on depriation for their effectieness (food, water,
air etc) f one aoids such reinforcers, depriation won' t be present when
he R is reinforced and thus won't ain some contrl oer it Also, one
aoids any stimuli that hae been exclusiely established as means to soe
depriationdependent unconditioned reinforcer, that is, any simple con-
ditioned reinforcers The reason is that simple condiioned reinforcers
depend on depriation for their ffectieness, too, and they pose the same
problems as do the unconditioned reinforcers. And, nally, one has to stay entirely away from atie reinforces of any sort, because they hae the
conspicuous discrimiatie stimui compounded with them.
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 30/89
acts 46
What are left, and what are used, are te CONITIONEGENERLIZE REINFORCER TH AR TMUL THAT HA BN
PAR WTH, OR TABLHD A MAN TO, Y OTHR RFORCR, BOTH
CONTON AN UNCONTOND The generalzed condtoned renforcersnether reduce nor depend upon any partcular deprvaton condton forther effectveness Nor do they vary wth deprvaton or aversve
stmulaton The next llustraton shows some examples n the paradgmr "CCT
Sr "IT
Sr FIE"
Sr "
Sr S
SD
S
r
" A"
SD _ v
r "TAKS"
Sr "EAL
Sr
l
i
,
"
Sr
Sr "TU"
Another good generalzed condtoned renforcer s moey It seems torenforce at any tme, totally rrespectve of deprvaton, because t s a meansto all other renforcers, postve or negatve, condtoned or uncondtonedWe know why the verbal condtoed generalzed renforcers lsted n theprecedng lustraton work We suppose that they have been correlated wtmany other renforcers, the way money has Furthermore, some nonverbageneralzed condtoned renforcers may be used n the same way, for the
same reasons, such as nods, smles, pats on the ead, uprased thumbsAnd, nay, some punshment may be used for nexact correspondences,such as rdcule, contempt: No, no; not a horse, a pg, on't you knowanythng?
Many experments have shown that condtoned renforcers may beused to ncrease the probablty of selected Rv's The strengthenng of arb-tarly selected operants s the one area of expermental research n verbalbehavor that s based on knner's analyss Holz and Azrn revew andsummarze these studes n ther chapter n te book edted by Hong, whcs lsted n the References Here are bref descrptons of two classc studeswhch have shown the effectveness of condtoned renforcers on verbal behavor Greenspoon found that f he asked college students to ust talk (say
words) and followed selected Rv's wth the sound mmmhmm" made byhm, the frequency of those Rv 's ncreased n subsequent speech by te stu
acts 4
dents. This occurred despite the fact that Greenspoon did notig else weneach student was speaking . Hildum and Brown found that if tey made theRv g ood" over the telephone, tey could bas people's resposs to qus-tons on atttudes toward g eneral educaton courses at Harvard UnvrstyWe should, however, be aware that some dubety has been expresse wt regard to certan studes that have demonstrated the renforcn power o
stmul such as mmmhmm" Most of the reservatons and controvrsesare about whether or not the speaker may have to be aware" of te co-tngences of te experment n order that the stmulus ncreas the re-quency of te selected Rv There s lttle need for the knneran to worryThe necessty for awarnss does not dsqualfy stmul such asmmmhmm" from beng placed n the category of renforcers They do,after all, ncrease Rv frequency when they are dspensed followng theselected Rv's Furthermore, awarnss s only requred f the response classs badly chosen
The earlest experment I know of that used operant condtonng pn-cples on verbal behavor was reported n the unpublshed Ph tess ofr ay W Estes Rather than rely upon any verbal renforcers, r Estes
constructed some specfcally for her experment They conssted of temovement of the sngle hand of a specally made cuckoo clock Wen tehand moved, a bell rang, and each tme the and moved to the twelve0'clock poston a door on the clock opened and a small toy anmal maeof color�d ppe cleaners was pushed out. Prevous work sowed tat
veyearold chldren, who were the subects of Dr. Estess exprmntsfound the toys renforcng. The cldren were not renforced by t mov-
\ ments of the clock hand before they were pared wt toy presentatonsThus, f clockhand movements would ncrease the rate of esson ofselected Rv's, t would be safe to conclude tat tey dd so on the bass ofbeng made renforcers by beng pared wth or servng as a means to the toys
To generate so m verbal behavor upon wch te renforcers could btested, r Estes had to get the chldren t talk he dd ths by showngeach chld a scrapbook n whch were pasted pctures Two or more peoplewere shown n twentyone of the pctures; two or more anmals n sx pctures;people telephonng or talkng n nne pctures; people and anmals n four pc-tures; and colored nkblots only n eght pctures Each chld was asked,What s the man (the dog, the lady, the boy, etc , as approprate) sayng?"If n hs response to r Estes 's queston, the cld ncluded the Rv please, "the clock hand was advanced The chldren, o fcourse, were not told that nythng that they dd would advance the clock hand, although they were n-formed that they would be allowed to take the toys whenever the clock gavethem one
r Estes found that te frequences of R vs o all sort s were ncreasedby her expermental proceure Whe she analyzed the chldren's responses
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 31/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 32/89
acts 0
negatie reinforcer and emit the The door' s open. " his is a mand. Ifyou replied Oh yes so it is; hanks for telling me I might change theform of my mand for example Close the door.
Motiation for tacting has been postulated by oter accounts as if itwere logically impossible for behaior to occur without some underlyingdrie or motiating state. It has been said that we hae a need for
selfexpression and that tacting reduces that need. If this theory were truesaying anything would satisfy the need and it would be easy to learn anddo. But lies and nonsense aren't common 's of people who are exposed todiscriminatie stimuli politicians excepted. Also because 's do coarywith their enironmental circumstances they are not selfexpressions; theyexpress the enironment. And to the extent that tacting is a result of theactiities of reinforcement mediators it is an expression of conentions notthe speaker's selfexpession.
WY DO M EDIATORS PLAY THEI R ROLE?
At first blush it is hard to see why mediators do what they do inreinforcing tacting. Anyone who plays the role of reinforcement mediatorfor tacting doesn't know any more after the speaker has spoken than he ion the basis of the discriminatie stimulus alone. Yet he reinorces on thebasis of the exact correspondence between the speaker's and the discriminatie stimulus. oing all this seems to be a waste of his time andenergy. Howeer if the members of the reinforcing community do their jobproperly in most (ideal) cases the tacter
1 . Tacts exactly on the side i.e. the has a specic form;. Will be completely under the control of discriminatie stimuli and
discriminatie stimuli alone; he will speak if they are present andbe silent if they are absent;3 Behaes disinterestedly from the standpoint of his own depria-
tions and aersie stimulations; and4 oes all of the preceding whether the reinforcement mediator has
access to the discriminatie stimulus or not.
Therefore by preparing exact tacters the mediator can astly extendthe functional range of his own senses. The tacter can often see hear orsmell things that the mediator cannot. His tacts are not those stimuli buthis erbal behaior is a ery direct exact known function of them. The nextillustration shows the reinforcement accruing to the mediator in a para
acts
SPEAKER
MEDIATOR
S� R v � Sr
u -D TNKSS �Rv
� R OR R
digmatic form. Here are some examples of 's from tacts; you should beable to figure out the nature of the discriminatie simulus for the tactingthe mediator's response and his reinforcement:
Lunch is ready.The suit you ordered is here.There's water in the cellar." The garbage man is here.I can let you hae two hundred shares for a thousand dollars."
In short as kinner has said the mand i s an imposition on the hearer; thetact is a faor to him. " Anyway that is usually the case. ifferent speakerskeep us on different schedules. Babies are boring. They neer tell us any-thing that we don't already know but reinforcing their tacting is a good inestment. Experts can see or hear discriminatie stimuli that we cannot.They tact them for us. There are special subcommunities of other expertshich reinforce the experts' tacting; that is how the tacting was generated."
Th other experts taught them how. ccurate tacting of real thing s that can
not be seen is part of science. The physicist does it when he tacts atoms and
electrons. This kind of tacting is harder to acquire and the distances
between responses and reinforcements are longer. This does not mean that
no discriminatie stimuli control the physicist's behaior but the relation-ships between the discriminatie stimuli that control his tacts and his re
sponses that refer to" unobserables are ery complicated. We need to
know about textuals intraerbals and autoc1itics before we can complete
that story.
DISCRIMINATIVE STIMULI AS
SUPPLEMENTARY VARIABLES
Tacts show up in erbal behaior that is principally goerned bysomething else. xamples of these are extremely common and sportswriters
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 33/89
acts 52
and announcers speialize in producing things where one tact seems to intrude ino another:
The Bengals clawed . . .
The Pirates stole . . .
The Cowboys roe over . . . "
The Broncos stampeed . . . The Inians scalped . . . "
Here are two not very funy ones that I manufacture:
His report hammere at the ontrast between the strikers' and the
managers' behavior."
His pithing mae me want to throw up."
In utteranes like these the speeh i s controll by someting else; it is not
about vomitig or hitting thigs. Other Rv's migt have occurre, such as
oncentrated on," emphasized," poite," quit," give up," be
sick," go home." With all of these available, there is a proble of theresolutio of the competition of the various possible responses with one an-
other. What determines which Rv appears? A supplementary variable may
help to solve this question, but more about that later.
DISORD ACTS: INXAC
CORRSPODNC
In the ieal tact, the meiator reinfores the speaker o n the basis
of the conventionlity and exatness o f the math between the latter's Rv
and the discriminative stimulus. The meiator is inifferent to the specifi
discriminative stimuli ad responses involvewhat he ares about iswhether they match accoding to the conventios of their reinforcing com-
munity. If this is all that happens, then all iscriminative stimuli will be
ated with equal and great exactness and precis ion. Ho wever, the mediator
does not always play his role properly. His principal temptation, and on-
sequent defect, is to reinforce on the basis of the Rv alone, irrespective of its
correspondence to the disriminative stimulus. The result is inexactness, un
conventionality, and some distortion of the speaker's tacts. The purity of
the tact is diluted, and it becomes a tact plus something else. There are
several possibilities hat produe some familiar products. Examples follow,
grouped by the properties of the distorted tat's Rv.Some Rv' s are aversive and are not followed by reinforement of the
speaker. I f the v is aversive to the mediator, he isn't likely to reinforce, nomater how good the corresondence etween the v and the disriminatie
acts 53
stimulus. The next illustration shows a paradigm for this Sometmes the ex-
tinction of the tacter is permanent. In 2 Samul 4: 9- 10, we e tol that1 SPEAKERMEDIAT
OR SPEAKEEDIATR
o D D rS
. NEWS � Rv
NEWS" � NO S
- - - - �-'
LEVES r-S � � S TERMINATS
O R WY" S TERNTES
Davi answered Rehab an Baanah: As the Lor lives, who has re-
deemed y life out of every aversity, whe oe tol me, Behold, Saul is
dead,' an thought he was brigig me goo ews, I eize him slew
him at Ziklag, whih was the reward I gave him for his ews. "Not ie
boy, that David.
Other kinds of aversive Rv's are those that tat ex or exretory e-havior, may tatless tats: You are too fat," Y ou hve f,"
I hear that your wife let you." These may be perftly true ext,
but they remain weak. Euphemism and unerstatemet may the our
be reinforce: It appears that you have a weight proem," There eem
to be some ust o your oat, " Perhaps you ought to kow tht tee e
rumors that thigs are not going well at home."
Palmer has tabulate my iterestig euphemss or hep or -
�ual foos; here are some examples. The ommo Rv s o the le d the
euphemisti Rv is on the right:
Fried sh
Bag pudigBread crust an garli
Cow heel
Sheep's head stewed withonions
Red herrings
Red herrings
Liver and potatoes
Shrimp
Potatoes
CodfishSturgeon
Bombay uk
Leiestershire ploverCapo
Cobbler's lobster
German duk
Norfolk capons
Gourok hams
Poor man's goose
Gravesend sweetmeats
Irish apricots
Cape Cod turkeyAlany eef
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 34/89
acts 54
Some 's are more useful, important, or interesting to the mediatorthan others are This means that the probability, or magnitude, of the rein-forcement, which results from the mediator's own behaior, for which thespeaker's is a discriminatie stimulus, is greater Such 's get preferential reinforcement from the mediator If, in addition, the mediator has noaccess to the controlling ariables for the speaker's behaior, inexactness islikely Exaggatn may ensue: an eighteeninch mushroom," a billion-aire," an absolutely perfect figure" Speech that produces special effects such as laugh getting or tear jerking occurs when the same discriminatiestimulus may be tacted in seeral ways But distortion is usually intr ducedby the special contingency these have with respect to the mediator's elicited,respondent, behaviors. They begin to specify their consequences, and become mandlike, even though they are still tacts. The ultimate case of dis-tortion occurs when there is no discriminative stimulus for the Rv, accordingto the community's conentions, that is, lyng. The man who cries wolf, ordenies committing an action that he did prfor, is manding He andsgeneralized reinforcers in the first case (attention) an termintion of aersie stimulation in the second
Criteria of conentionality ad exactness may differ in different rein-forcing communities ifferent mediators may reinforce different ' s tthe same set of discriminatie stimuli These iffering cmmunities, o u diences, hae different conentions A person who is a member of one audi-ence may discriminate tacts that are exact, according to the conentions ofthe other, as distortions Psychologists, for example, may be members ofboth the literary and the scientic audience, whereas English professors arlikely to be members of the literary audence only So, when the discrimina-tie stimulus is a person, he can be tacted to the psychologists (scientific au-dience) in terms of MMPI, WAIS, CPI, orschach, and TAT scores, forexample Or he might be tacted with the He 's a regular Scrooge " Theliterary audience says that the rst set of tacts is lifeless jargon; it complainsbecause tacts like these analyze to death, are far too complete, leae noth-ing for the hearer to do" he scientific audience says that the Scrooge"tact is untestable, inexact, and incomplete The point is that the existence ofdistortion oftn depends upon the accepted conentionality of one's wnerbal community, and there are many erbal communities within the largerone
extended as
We hae a parallel here with te mand Namey, we fi i-
stances of tacts without, strictly speaking, a past history of reiforcemt
{or emitting that in the presence of that disciminatie stimuls W
we encounter a new discriminatie stimulus we ar ot alws
speechlesswhat we do is emit an ol R, or part of n ol R W will is-
cuss, in this chapter, cass of old responses to new or noel stimuli My gol
is , of course, to show you that such occurrences are still comprehensible
within the restrictions we he placed upon our explanatory system
WO MCHANISMS
We can easily see that the erbal problem that we hae been dis-cussing in a preliminary way, just now, is a subclass o f stmulus gnalzatn. et me repeat our definition of this: STIMS GENEAIATION T TNDNCY FOR TU OTHR THAN TH ON NOLD T
PAT HTORY OF CONDTONNG TO OK T RPON HN T HAPPN
N TH TACT CA, PAK OF EXTENE TACTS We will consier twoways in wich stimulus generaliation occurs
Simpl stimulus generalization should be looked at first, in prt
because you already know about it. he next illustration gies a pri
for simple st imulus eerlization with the pig eon as the subect If t pi-
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 35/89
xteded tacts
PIGEON D YELLOWREEN PECK
TACT D LUE v "LUED NEXT R"YER'S CRS v C
D THE SIDE - V FRED" VIEW OF RED
D FEDS
v "FRED"OTHE S
56
geon s contone to peck at a yellowsgreen spot of lgt on te key, econtnues to peck, but at lower rates as te lgt gets bluer (green, bluegreen, blue, volet) or as t gets reer (yellow, yelloworange, orange, re)Ts sort of effect may be observe, n general, for any stmul on te samepyscal contnuum, suc as louness, ptc, ue, or brgtness Te sameprocess also appears wen te response s a Rv nstea of a keypeck Telower part of te preceng llustraton so ts lso e wll say blue"to a ban of colors, some of wc we ave neer experence before mlarly, n te ays of te annual moel canges, next year' s cars n't lea
us speecless, an a vew of goo ol Fre from a new angle oesn't leavus wt notng to sayTML ALO CLTER, A TO OR MOR COMPLX TMULi
MOR PROPRLY CLUTR OF TMUL ON HA COMMON ELEMENT. Te next llustraton sows te processes tat go on ere n para
TACTER
MEAOR
UT T THE SE TIE
gmatc form . T e renforcement meator scrmnates te correspon-ence between scrmnatve stmulus an Rv \ an renforces accornly
xteded tacts 5
But, at te same time te total situation as oter iscriminativ stimuli int, an some of tes may g an cotrol ov er Rv\ . If som of te oter s-crmnatve stmul (numbers two troug fve) ave a farly frequencyof accompanyng scrmnatve stmulus 1 , tey may et lo of controlover Rv \ because t gets renforce n ter presence too, a t satonepcte at te bottom of te llustraton may result. Te next lsraton
gves an example of extene tacts base onwhrt cluter
Ts s
THESE PRTS CLUSTERWIT THE WHOLE NDRE PRESENT DUINEINFOCEENT N
D ERSON
D EYE
v "EYE" D
R C SS D
D
OV
THUS GAIN SOME D TCONTOL OE THE
LTER WHEN SOE OFHE PRTS OCCUR SPRT OF NOTHEWHOLE THE v TENS
TO OCCU
D
SURROUNED FLESHD CONTINS FLUID
D NEEDLE
RECESSED
OVL
PT
"EE
SUOUNEY STEEL CONTIN THE
mtaphr, a fgure of speec" wc s supose to b e creatve But you
see, te speaker oes not crate, nsea stmul controlby eneralzaton
Furtermore, e s actn ju st lke a peon Many metapors nvolve oy
parts , for example If te scrmnate st mulus s a projecton from t
boy, te Rv arm" es renforc, r s t arm of a man or e
of a cross
Simil s ynamcally muc e sae s mapor We may say a
man s gentle as a lamb or tat e s m as ox Te next llusrao sows te ynamcs 6f te smle entle a lb." Smle an mp
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 36/89
xtended tacts
ESE OCCUR OGETERAN REINFORCEMENT OCCURSN TE PRESENCE OF ALL OFTEM, AN EACH ONE GANS
SOME CONTROL OVER THE v
0, LATE WHEN ONE OFHOSE OCCURS IN AIFFERENT CLUSTER HEv OVER WIC T ASCONTROL 15 LELY TOOCCR
:LAMB�:o �
R
V"LAMB
FL C
G NTL
BAAA"
MAN
plNK
LARGE "AB
SIN - /
D
G N
/
SWEETE
58
are quite similar, but as the next illustration shows, in simile the common
elemet is tacted twice·.
V GENTLE GENTE �
. v LAMB
RV "HAD" AR
v "NALS
When the name of a part is evoked by the whole ensemble, we have case of what is termed synecdoche. The next illustration shows this for acommon folk expression, I just hired two new hands. " The processe
illustrated here are a plausible reconstruction of what the original reinforce-ment history probably was. We do not learn these tacts in this way. For u
the following Rv's are learned as echoics or textuals, but each was yn-ecdoche for someone first:
This art of the symphony is played by the strings."
A herd of eighteen hundred head."
Hs bulding has seven oors." A twohunredvoce chor.
xtended tacts
A OF ESE S CLUSER;EY OCCUR OGEER, ANWEN A v GETS REINFORCEN THE PRESENCE OF ONE,THE OERS GAIN SOMECONTROL
SO, ATER WHEN SOME OF TEOER IN TE CLUSTER
OCCUR, THE v ENS TOOCCUR
SOMEONE SAYS "ANS FORSOME REASON EG , WE NEE
° AOTER PAIR OF HANDS ERE" WORK NG D . r
HANDS -.v "HANDS
-if HAS-�I
1 JEANS/ // I SPURS /
/ BOOS
COWBOY/
WORKING
SPURS �
O HATS�
V
"HANDSII
O
J N
S
S
O
eo
T
COWBOYS
59
Metonymy is really just a special case of common elements clustering.
It ' oc cu r
s when stimuli have common accompaniments. We can look at
an
th
r of Skinner' examples. The president and the White House are
common accompaniments; seeing one oen leads u to speak of (tact) the
other for very g ood reaons. We therefore have an exceptionally strong in-
stance of metonymy in The White House says . . . " Interesting ly, it works
only one way. When the preidential residence was being repaired during
the Truman administration, no one said that they were "repainting the
president, even thoug h the Republicans claimed that the newspapers had
whitewashed him during the freezer and mink coat scandals. There are
plenty of other examples where metonymical extension is institutionaized:
The Chancellor's office ays . . . , " Wall Street is waiting for . . . ,"
The Pentagon reports . . . ," Seventh Avenue is looing for . . . "
Metonymy is not merely a literary curiosity and source of jo�e�. In
large part it enables us to dispose of what is said to be a very dffcult
problem in verbal behavior. It allows us to explain tacting (speaking o a
missing discrminatve stmulus One explanation for the tacting of aben
objects is that the dynamcs of the metonymicl situation are operatng. T
�
next illutrtion g ve a parag matic example. When we wer bou to SIt
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 37/89
xteded tacts 60
down, my wif said (), Thr s no watr on t tabl." Why did sh say() watr ? t is not part of a tact to th discriminati stimulus watr,nor is t a tact to th dscrmati stimulus of o watr, or was it a md(sh wnt to gt th watr). mmbr, it wot do to say that absc ofwater is also a stat of th ironmnt, so it srd as a discrimiatistimulus for th o watr." t s, idd, a stat of th rom,but may, may othr bscs wr thr too, ad sh didt tct thm,for xampl, o spidrs, no trucks, o chang, no Gorg WashigtonBridg. Th llustratio shows how watr" gts sad in this situtio. t is
IN TE PAS TE v WAS ENFOCED N EPESENCE OF A CUSTE OF Ds EEN TOUG
° E EDATO CONCENTAEDS TABLE
ON JUST ONE. BUT OTES COT, GOT SOME CONTO KNFE
'
° . ' FOK-_ '
.
WATE-
:V "WATER Sr
SPOON / / / GASS /
SAT // PEPPE/
TABE
COT
KNFE
FOK
SPOON
GASS
SAT
PEPPE
SO ATE WEN TE OTESOCCU TEE S A ENDNCYFO E v TO OCCU OO
�V WATE
still th of a tact, but t is okd by watrs commo accompanimts.Th s thr is " and no" ar still to b accoutd for, bcus thy arundr diffrt sourcs of control and parts of diffrnt oprants (autocli -
tics). tymly is oftn a history of how a of a tact got assd alogfrom scrimiati stimulu to discrimti stimuus by t rocsss of
xteded tacts 61
gnralization which w h xamind. Th nxt illustrato s -othr of kinrs xampls. All ths procsss combi os,
OGNA
KNG VSU
BUT ALSO KNIGHT - R
"_ v SPU KCK OSE
SO KCK OSE VSPU
TEN KCK TNG VSPU
AND AE
MOTATE VSPU
EFFECT VSPU
KCK AWAY v IISPUN SPU SAPED NG V SU OF AOAD
SPU OF BONESPU OF FOWE
SPA OF BOAso th tactr is not spchlss. But h i s ot crt ithr Nor is hs rbhior radom. Whichr stmuli ar ibl cotriut rh otacts or parts of thm so tht parts of old tcts occur.
SCRMTO RLZTO
TCT
f thr wr o procsss opposg or mg g rzo
(xtnso) could prhaps go o forr. Hr s xmp of wh cou
happ. Th nos" starts udr th cotro of th dscrmt stm-
ulus os of th man ad th gts trasfrrd y g rlzto o th fo
lowing dscrmati stimul: os of rpl (mthorl of h
airplan (sycdoch)all closurs (gnrliztocr mo
thigs (mor g nralizato)rythg . S o, t would b possbl to w�d
up with a on-word tact ocabulary, that i, os. " W do ha somt
lik that with th thig ." Althoug h, i th cas of th th," th ·
dscrimiati stimuli that control t lso control othr R's wich h
sam tim allow for much mor spcific tactig. In th hypothetic o x
ampl, nos" is assumd to b th oly R okd by th stimu
sort o xtndd g nraliztion would, howr, mak tactg us o�
harr. Such tacting woud ot b a basis for ffcti (rnforc) behaVor
on th harrs part. magin th plight of th poor harr who sam from th spkr irrspcti of whthr th discim
�
.
,
\
uus s dinr o th b or fir n th iquor cbt.
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 38/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 39/89
xtee tacts 6
strained by discriminaton training, e c an now tact a new, strange purplepig
Hearers can receve te functonal equvents of proper namestroug te speaer's complete or prolonged tacting of bject cass namesor property names f te speaer names enoug subcasses or properties, eacieves a specfcty equivalent to tat of proper namng Of course, edoesn't ave to go all te way to tat extreme He can stop at some ntermedate eve tat is a suffcent dscrmnative stmulus for te earer's pur-poses, one tat w alow te earer to respond and be renforced Here san example of suc proonged tactng Remember, te dscrmnatve stmuus remans te same for al of tese Rv's: srt," Fred's srt,"Fred's new srt," Fred's new wte srt," Fred's new wte dresssirt," Fred's new wte dress srt wt te monogram," Fred's newwte dress srt wt te monogram wic e s wearng now Tat lastutterance s (practcaly or functonay) a proper name Ts process sats-es te speaer's nerest (e can generaz_) an te earer's (e as towait, but e gets te specfcty of a proper name) Ts s a powerfu pro-cess It generates precse speec for al occasons, weter tey be new or
od We may notc tat speca problems of ambguous tactng, wc arele generalzaton, are generated n anguages tat ae any synonys orhomonyms Te next llustraton paradgmatcally depcs tese two effects
SYNONYMY
HOMONYMY
RV"
S SOC EET<RV"DCE"
S°
SOC ET"R
V
"BA
L
S TH USED /
ES
ynonymy s te result of an extended renforcement story, bu t t does n'tncrease te precson of te discrmnatons of eter te speaer or teearer sng a synonym clarfes matters by tappng te earer's renforce-ment istory But, te earer and te speaer must ave simlar reinforcement stories if synonyms are to be elpful Te speaer wo uses synonyms as ad severa reinforcement stores and goes troug teir
products (tacts) unti e finds one tat serves as a dscrmnatve stmuusfor is earer Homonymy just eads to ess dscrmnao on te earer's
xtee tacts 6
part Prolonged adtona tactng s needed, and reor o ostraigten t out ome verbal communtes are not wg to p p aof tis, and terefore pans speaers ave reatvey few ooy
TH SZ OF TH SPOS TCT
Earer we sad tat te size of te response ut m be e oor fexble ecause different varables control responses of dee e We can exmne varous examples of tacts t o see ts varaby
Muc tactng occurs wt sngle unts tat correspod to te lexwords For example, objectRv srt"; SD propertyRv wte" f wecange some part of te dscrmnative stmulus, t teds to sow up as asngle word cange
ome seemngy unbreaabe word cas occur n tacts:
D
� §
S
IC gredent
DsorderCorrect
RClean as a wstleNttygrtty, wa ad oofHeterseer applepe order
Here are some proper names tat f te bl: George Wao Be
cag o Transt Autorty, Harvard Unverty Ad ere re ome -
word Rv's of tacts tat seem to be dosycrac For expe, a a o
says, as busted as te Ten Commandmets" as as a g rese
pig" ; and anoter man I now says, te ote g sce sced bread "
Eac s a unt because t aways apper ts way d ever appers ay
oter wayCertan tacts occur wt Rv' s at a eve beow tat of words Tese
frag ments of words never occur wtout sometng ese on e response
side Generay tese frag ments are wat ave tradtonay been caled
roots For exampe, f te dscrmnatve stmulus s part of e toa stm-
ulus stuaton, super" s ely to be part of te response as super-
structur, survisor, suprordinate. f te stmuus stuaton ncdes a
ong ong acton, te Rv ng " s liely to sow up, as n aking, walkng,
sitting. If te discrminatve stmuus ncludes sometng below or eer
tan someting ese, te Rv sub" s ley to appear, as n su?ore,
subbasmnt, subthrshold. Tese are al famliar from junor g oo
Englis class However, may very fragmentary roos re ee
sylables Te etymooges at te subsyabe fragmet leve re oeunexpected and are ot we estabsed n tradtona accouns of
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 40/89
xteded tacts 66
Here is an example from kinner. If the discriminative stimulus containsemanation from a point source, the Rv is ikey to be sp," as in srng,slash, sew, slay, slat, srnkle, sarkle, sray, sangle, seak, sutter,slutter, sut ome other examples, which come from Bloomfield andMacCorquodale, are the following: If the discriminative stimuus is a nosethe Rv is likely to be sn," as in snut, snre, snze, snuf sn sne:sneeze, sneer
If the discriminative stimulus i s smooth and wet, the R v islikely to be s, as in slme, slush, sl, slbber, sl, slde, sther If thediscriminative stimulus is diminutiveness, th Rv is likey to be the short "sound, as in the foowing: slot-slt, -bng, sn, -tng, snuffsni -y You can do this too. f you go through your dictionary, Ithk that you wil be convinced that if the discriminative stimulus is ong,sender, and twisting, the Rv sk" is likely to appear; that if the stimulushas eements of attack, destruction, or reduction, the Rv sm " is likely; andf the stimulus situation has elements that are smooth and at, the R v paIS probable.
This can lead to a sorts of fun word games, but there is a serious point to it. Namely, the speaker in certain situations is likey to make certain
kinds of sounds because the discriminative stimuli present may stengthen acertain sound fragment on the basis of tact operants that occurred pre-viously in ld situations and were reinforced. Furthermore, the hearr islikely to reinforce these fragmentary response tacts because they soundright" to him; they do so for the same reasons that the speaker is inclied toemit them. ometimes an argument is made that these sub sylable soundsare onomatopoeic (sound like the discriminative stimui). But, cross-cultura investigation dispatches such a claim easily; s" does not sounsmooth and wet to a Frenchman. Here are some English words and theirFrench translations:
lake Etancherleek Lisse, uisat, polilimelushlobloplobberliplidelitherleetlosh
Vase, baveBoue, fangeVase, limonLavasseBaverGlisserGlisserGlisser, ramper (what a snake does), onduerNeige moitifonduelanquer un coup
xteded tacts
lugluicelicklinkloglough
ludge
LimaceEcluse, bondeAdroit, en bon ordre'esquiver, se droberCogner durFondrire
Boue, vase, cambouis
67
Fragmentary tacting may sometimes be seen occurring in the utteranceof a kind of neoogism called rtmanteau wrds The next ilustrationshows two examples of this. The first comes from Lewis Carrol, who wasan expert at making them up. He was the one who gave them the name
SO OUS Rv FUMOUS"
SO MNG
SO MPLCATONS Rv "MPFCAONS
SO AMFCATONSportmanteau wordsbecause, as he said, they are like a fexibe suitcase; you
can g et it al into one. Here we have new words expicity composed o frag-
mnts of responses o old situations; the component Rv s and their discrim-
ative stimuli are, according to Carroll, those depicted here. n mny cses,
the maker of a neoog ism knows what t he component tacts are; in a few
cases, he doesn't . What this does sugg est, however, is that just about al
new words are composed of components of old ones.
ABSTRACTONS
This is somewhat eusive to some people because it is so contrary
to our usual view on these matters, but i f you bear with me and g o over it
ag ain, I t hink you wil see that it is coherent, and almost necessarily true.
Earlier we considered tacts that named discriminative stimuli that are sine
isolated properties of objects and said that we woul call these abstrctions
Several aspects of that kind o f behavior are of considerable importance
If we want to condition an abstraction of tacting Rv red" to redness
as such, we have an interesting engineering problem, because wheneer t
Rv red" is folowed by reinforcement, something ese other than redns is
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 41/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 42/89
xteded tacts 70
the work of Wenner, Johnson, and Frisch, all that stomping and twistingaout which tells" the other ees where the nectar is located is proalynot operant (see the recent note y aenport). The same is true of irdsongs. These are emotional cries; they are respondent and hardly disinterested. They act, if anything, as unconditioned elicitors for the ehaiorof other irds. olphins mght tact, though, ecause they are smart socialanimals. Howeer, no one has cracked their code as yet, so we cannot esure whether they really hae a linguistic community or are like the rest ofthe wild animals with primarily unconditioned emotional cries. These latteranimals don't tact ecause they hae no conditioned generalized reinforcerswith which to put operant ehaiors solely under stimulus control.
Can animals e made to talk? Two major experiments attempting toanswer this question hae een recently done with chimpanzees, ut neitherhas een reported in sufficient detail to proide definitie answers to thequestion. Washoe, the chimpanzee trained y R. Allan Gardner andBeatrice Gardner, mands in American sign lanuage. Eac tme se made asign for an oject, they tended to gie it to her. In he Nw Yrkr EmilyHahn wrote that Washoe's then current owner, W. B. emmon, called
Washoe the greediest chimpanzee that he had eer come across. aidPremack's arah used pieces of colore plastic as R's and does seem tohae some astractions that are transferale from situation o situaion. utit is hard to see whether these will remain as disinterested R's, since she wasfed for correct responses. I would enture to say that mands hae definitelyeen taught to chimps, and that tacting may e teachale. It's ery in-triguing, ut we need much more information efore we can come to anyfirm conclusions as to whether the animals can e taught a full repertoire ofnonocal eral ehaior.
If it takes a man or a woman to teach a chimp or a child how to engagen eral ehaior, who taught the rst man? Who was the first reinforcement mediator? I don't know; I wasn't there. It isn't necessary for us to
know how men first learned to talk. Our prolem is how they get to talknow and what keeps them going. Being so far remoed from the necessaryoserations, any hypotheses I might now make woul e merest speculations, ut for te sake of completeness, here goes one. I suppose that tal-ing got started y accidental reinforcement of operantleel aling inchildren, who then reinforced their parents and thus got the whole thinggoing. Once talking started, generalization and discrimination could akeeffect and extend repertoires.
As I said, how talking got started isn't really the prolem we areaddressing, so I won't take my speculations further. The ariales let forus to consider are all discriminatie stimuli, ut the operants they control
are not called tacts for one reason or another, which will ecome clear in theremainin chapters.
audiences
In this chapter we concentrate on the other indiida in th in-
terlocking eral operant paradigm. So far we hae cocraed o h
speaker and to the extent that we hae discussed he other perso t has
een wih respect to what he has done aer the speaker has said sotin
We now consider the functions of the second person prior to the onset of a
speech episode.
DINITION AND ARADIGM
We seem to hae seeral names for indiidual B in the following
paradig m. We g ie him a different name for each different funcion that he
performs. Generally we call him he hearer, ut w.hen he comes n afer er-
al ehaior and reinforces A's operants, he S called a reforcement
mediator or memer of the reinforcing community. Because h mus
normally e present efore eral ehaior occrs in ordr to co�e. ae�
it with a reinforcement he g ets a certain amount of behavIOr condIOned t ,�ST
himself. E IS A STULUS, AND HE, A A STIMUL, IS PRESENT WN OR BEFORE REFORCED RESPONDING OCCURS. HE BECOMES THE,
ANTECEDENT CONTROLLING DSCRIMATIVETIMULUS AND I IN ROL
CALLED AN AUDENCE . He seems to exert three principal kinds of c ntol:
71
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 43/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 44/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 45/89
udences 76
epive subect with an auience who has in the past f him an see if thesubect mans foo.
econ the verbal behavior may be COVERT. OME, OR A, OF THE
AME MUCE MAY MOVE BUT MAY DO O WITH VERY M AMPITUDE. THE
AMPITUDE OF T MOVENT MAY BE O MA THAT UCH MOVEMENT CAN
BE DETECTED ONY BY EECTRONIC AMPICATION An some movementssuch as those involve in phonation may dop out entiely. Muscle ecor
ings have been mae by many people (see the books by McGuigan anchonove an by McGuigan) an the muscleprouce movements arefoun where they ae preicte to be. Thus ou account is in goo shape.Both processes probably happen an neither is inconsistent with ounatural science estrictions.
NCS S SPONS SCTOS
Given the situation shown in the upper part of the next illustration, whee a primary variable controls seveal different Rv's with aboutqual stengths, only one Rv occurs. Sometimes, surpisingly, the weakestof the lot is the one that appears. Why is this? The answer seems to ·be givenparadigmatic ally in the lower part of the illustration. The uience may aas a supplementary vaiable to strengthen momentarily one of the alterna-tive operants available.
RBLEM:
PY E
LUI
PY E RV3
RV4 r ( Rv4) t ND Rv GETS SD
O UDECE
We can cay this analysis forwa in somewhat greate etail. In thepast ifferent subaudiences have ifferentially einforce the speaker for
emitting ifferent Rv's to the same situation.THEE AUDNCE, A DI
CRIMINATIVE TIMUI, BY BEING DERENTIALY PREEN ON DFERENT RE-
udieces
FORCEMENT OCCAION GET OME CONTRO OVER THE R' NTIAYRINFORCED BY THEM THU THEY BECOME IRI INCE AND CONTRO DIFFERENT REPONE. Remembe ce intacting fo the scientific auience an tacting fo the lter ewants to be tol that a man is a cooge and the other w iscoes on a pesonality test. The scientist can be a meme ences an tact to both. Which operants ae momentarily r
etemine by which auience is present at the momen. i hoes not say or imply that the speaker chooses his words" to u ence. He rarely ha time to pick them out" an he oes't oing anything othe than talking. He oesn't seem to be ch reciing or picking his words. The audience determines the wods hi.He is ust a place where variables come together an interact to proucesponses; he is not a variable himself.
ometimes the response selection powers of the audience may be sufi-cient to etermine whether anything at all gets sai. The next illustratinepicts several examples of this. In the first case the presence o an udiencewill a to the Pr (Rv) so that something gets sai such as May I hav
hot og." If there were no auience there isn't much chance that the Rvwoul occu. Also one auience the faternity brother may supplement
DE Rv "Y H HO DO
UD(COOK) (Rv WHOU O S EY OW
OWENDOWDG Rv ww!
UD(Y OH)
O SEG
NO O RvUD (SES WE)
the stength of Rv's that tact the speaker's appreciation of the girl whe
the othe auience the ministe's wife may reuce the r (Rv) to z. say colloquially that iffeent auiences have different taoos d -
feent ethusiasms. With some if you make the wrong response
o you not get einforce you get punishe. Some auiece r
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 46/89
diences 80 dences
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 47/89
tantly, the male form re ued only by male while talking to maes, whereas female form are ued by female while talking to either mae orfemale, and by mle while talking to females owever, if a femae quotea male who ued the mle form, then the female will ue the male form inher quotation
Sapir lso tudied he lnguage of the Nootka tribe of Alberni Canal,Vancouver land The ooka peaker add pecial ufxe and/or meaningle cononnt form o verb They will add a paricular et of formdepending upon he referent of the peech One et i added for children,another for fat people, nother for very hort adults, nother for thoe uffering fro oe defect of he eye, nother for hunchback, another forthoe th re le, noher for lefthanded peron, and another for circucied le
Special udience exit for ll of u Cnt and jrgn re the pecialvocbulaie of profeiol rde Prctitioner peak to one another differenly han they do when pking wih cien One clinicl pychologitmy y to nother, Joe i even, ine, " thereby tctng Jone' i-mary pek on peronlity et and hence cerin pecs of hi per-onaliy Jargon i perfecly good nd ueful, no ter wh the frehmncompoition intrucor y They didn' realize i, bu hey were jut rying
to get you to ue thr jargon mot of the time rgon re pecy in-vented, no out of oppine, but becue of experie they reect thepeaker' abiliy to mke ne dicrimintion nd to tc hing that he ly-man cannot f you ca learn them, epecially Skiner', good for youArgt ued o b e he thieve' cnt or jargon ore bout ht very hortly
Slng i the Rv reperoire of one' peer group hif rpidly nd ivery faddih t i the language of the now" genertion, even hough i recycle t i lwy uxiliry to the ain repertoire, nd it ake everyoe bilingual and, to certin exten, pr of the peer group The ttl lngugs have techniclly been clled hycrsm One lk baby tlk obabie and to very intimate dult ypocori i filled with mngled pro-nunciation, and much of the tndrd reperoire i left ou Vriou twn
lngugs (ecre language ued by ibling twin belong in thi cl, doany sort ofrvt lngugs (pecial language ued by ocil crowd or pir of lover
Audience conrol scl rductn cts We pek loudly o theaudience tht wer hering id, nd lowly and oeie oo oudly tothe audience that look foreign Lieutenan Buh, in C S Foreer' Hrnblwr erie, ued to hve pecil repertoire for Spihpekig udi-ences: e ended ech Rv with oh" in the belief tht it would ke hiunderstood hi i carried furher by o of u who hve pecil reper-toire for anima We ay hoo" when the dicriitive tiulu i y,
bu ct" when it i a cat Frederick he re prov he betexample e claimed to peak French o the diplo o eclergy, talian to the ladie, nd erman to hi dog
MULTIPLE AND CONCRRENT
AUDIENCES
Becaue we peak differently to differen udience to
audience are concurrently preent, ome pecil effec re pole
lof verbal ply may occur Scrt lngugs may occur Thee lo epeake to bypa one audience completely Argt w he tieve' orrgon during the seventeenth and eighteenth cenurie t w y ofplannig a crie in the preece of the victim wihout lering hi We dothi oday we pell when children are preent y grndparen opeak Ruian when dicuing wheher o allow y fher nd hi broherto go o the movie
n rny verbl behvior i biguou both udience repod o but do o differenly t deped on oe dipoitio of oe of heaudience o repond one wy, while he oher, hving oeh diffee
reinforceen hiory, i likely o repond oherwie Te folo eillutrte wht i going on For he eker, eiher of he o srvtimuli i ufficient a conroling vrible Becue of h eo
a reinforcement ediaor, he i ble o dicriine he corollig rel-
tion It become pecial fun whe one pr of the udiece icriieonly one of the conrolling relation, nd he other pr di criie boh.
An apiring auhor ubitted hi new novel le eny in fou prz
competition nd received a leter fro one of the judge, who e o
to the oher, which red: Thank you for your enry. I hl wte o tie
in reading your new book . I overherd one profeor top oher
was carrying a copy of the book review mgazine Contempora Psy-
chlgy, and y, "I'm looking for good book on cogntive pychoo
he carrier replied, Look no further. " When both udience re i e
ame kin, these are called dubl ntndrs
udences 82 dences 8
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 48/89
SAPING B TH AUDINC:
A PRSONLIT FCT
I have argued that the audience exerts fairy strong but oensubte effects upon the immediate momentary behavior of a speaker. It canaso shape a lot of verba behavior that is create new Rv's or increase theprobability of some Rv's. It does this surprisingy rapidy considering thatit is ony a suppementary variabe. This effect has been overooed ordenied. Some psychotherapists caim that they serve as audiences in thecatayst sense only when they are performing their therapeutic roes. It iscaimed that the job of the therapist is to become a p erson in whose presenceany verba behavior may audiby occur. But therapists probably do a ot ofhaping (differentia reinforcement of successive approximation to thedesired response) on the spot. There is a ong history of finding evidence forthe interviewer's point of view in protocos (reca the experiment by Hil-dum and Brown). Remember that Freud' hyteical adies a caimed tohave been sexuay assaulted when they were children. It was part of hstheory at the time. How coud this have happened A few nods smiesoh's" realy's" can you te me more about that's" and umhmm's" at accidentay appropriate paces could do the tick. r. Estes
whose experiments I described on pages 47-48 did much the same thing withher cock. And of course reenspoon shaped pural nouns with his mmmhmm." There is a huge iterature on this which I don't want to go intohere. If you are interested you might ook at the chapter by Hoz and Azrinin the book edited by Honig which is isted in the References. RobertRosenthal's book on experimenter effects in psychology aso contains someinformation on these proceses.
ON'S SLF AS ON'S AUDIC
We have considered the speaker as a sefmander and a sef-tacter. We aren't very durabe einforcement mediators for ourseves but ifwe have externa reinforcement mediators to keep our verba behaviortrong we are an exceent audience for ourseves. We are the ones to whomwe do the most aking. Ater a I am aways avalabl to me. I have privieged access to covert verba behavior. I may however address mysefoverty if there is a ot of background noise or the probem is a hard one orthere is no one ese about. I further have the advantage that I can be spokento coverty thus avoiding aversive stimuation which might arise from myvert behavior. I as stener have the same response repertoire as speaker; Iknow a the words." It is a way of hearing interesting things sometimes.ccasionay verba behavior that never appeared efore appears in taking
to one's sef I I want t find out what I know I don't inspect my ideas Itar taking and isten to mysef. If I want to in a new soution to a probem I tart taking about it and sometimes i I tak ong enough I hearometig new hic may serve as a discriminative stimuu for other beavior which wi be reinforced. Finay and this is ony introspective"te sef is a very tolerant hearer. Rv forms seem different. Much is omittedmosty what wi ater be discussed as autoclitic verba behavior. Speech toone's sef is oen very grammatica in the interests of speed. We mustrood about this for a whie but we wi have to save it for a ater chapter.
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 49/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 50/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 51/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 52/89
Echoics 92
t t f t li t ( t d ng t k k
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 53/89
actvty of te earer as a listener (wat e s dong as te speaker speaks,bt not as a renforcement medator or as a dscrmnatve stms) s toengage n covert echoics (sma amptde). Ts proposton s a conseqence of or genera stmusresponse poston wt respect to verba be-avor. In ts vew, stenng, ke any oter perceptua actvty s notjst passve absorpon. Hearng s not someting tat ends wenmessages" or bts of nformaon" go from te ears to te bran. Hear-ng s an actvty; condtoned operant beavor wc s probaby eavyecoc; te earer taks aong wt te speaker. Jams J. Jenkns once sad
tat ts was te teory tat camed tat yo ear wt yor onge. This is sometimes called the motor theory of speech perception. It hasseveral variants and proponents. We will see where ours takes us later on ints book.
textuals
DEFNITION AN PARADGM
A TEXTUAL IS A VERBAL OPERANT WHOSE CONTROLLG VA-
ABLE IS WRITNG OR PRINTING, ADWHOSE Rv IS FUNCTIONALLY EQUVALENT
HE PROPER NAME OF THE STULUS. Te folloing illuraon inroucsome exampes of texuas and some paradgmac convenion. Wen estms s a text (wrtng or prinng), e put t on a piece of paper, a is,draw te box around t, n our paradigmac notaon. Suc hings as esemcoon, ypen, das, comma, an aposrope are not iscrimnaive
S D
HAI
� Rv "HAT
SDI� Rv "e
SD[� RV AND
sDc� Rv 10 [PAUE]
SD
w i
CFO EX UAL O SD/
stmuli of textuals in this sense. The do not control any of h rr'
Rv's . I 'l back off a te on h s; te dash may b a extual iscr!a
93
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 54/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 55/89
extuas 9 extuas 99
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 56/89
extuas 9
contingencies must be effective in maintaining the strength of these oper-ants There are at east four major sources of added strength.
Our next illustration depicts the proposition that many texts tact; theycovary with stimuli and thus serve as discriminative stimuli that make other
BUS SOP-Rv : BUS SOP � R STAND
@l-Rv AFE � R N � Rv FOOD-
NANE
OANS
ONE WAY
GE HE MONEY O HE MAN N H BUESU AND HEN SEF DESU
reinforced behavior possible People read f he same · reaso that heylisten t accuate tacting on the part of others Tacts and textuas producediscriinative stimuli which are useful to the hearer of them. They let him
respond in such a way that he gets reinforced.Second, texts mand, and in so doing give advice. The last two texts inthe illustration do this gain, generally, reinforcement for the reader isguaranteed if his subseuent nonverbal behavior conventionally conformsto the discriinative stimulus provided by he speaker under the control ofthe text.
Third, texts are a source of intraverbals, that is, verbal operants underthe control f the speaker's own voice (We will treat these in detail in thenext chapter) Sometimes it is necessary to talk in a certain way; in myclasses I have to talk like Skinner. That is easy; all I have to do is get at�xt It is the easiest, quickest, and surest way. Studying the text sets upverbal chains which may be used (needed) later, that is, reinforced by a
mediator These chains are subspecies of intraverbals. If, as in that situa-tion, reinforcement is dependent on a string of Rv's of a certain kind, theseresponses may be set up by a text. So, if you are going to take a course ortake a test on Skinner, a text can set up the Rv chains that may be availabelater and be reinforced when they occur. s Skinner has said, notebooksand libraries are not filled with facts and definitions; they contain marks onpieces of paper which are discriminative stimuli for talking With a text evenI can talk about physics the way Einstein did nd I don't have to do whathe did to get that verbal behavior; I couldn't. Textuals give us the behaviorof people long dead and far away. They may properly be said to be a way ofstoring behavior, rather than information. The point is this. Reading gives
us verbal behavior which someone else may reinforce. You don't get rein-orcement or speaing unes you spea, and importan sources of strengthfor speaking are the discriinative tiui o a text.
extuas 99
Even if someone else will not reinforce me for verbal responses set upby texts, it is often reinforcing to read This may be because i is reinforcingjust to be able to speak, perhaps in order to prevent the enviroment frombecoming too quiet lso, this is reinforcing in the same way tha going to aplay is, or watching TV is It permits me to talk about experiences that Ihae neer had. Very importantly, it may provide me with behavior thatcompetes with and crowds out other behavior, where this other behavior
has aversive properties It is a means of escaping, not from my unpleasantthoughts, but from my aversive behavior I talk to myself a great deal,almost all the time, in fact. nd some of the things I hear are negative con-ditioned reinforcers. eading is a good way of preventing these from occur-ring. nd it also is an easy way of acquiring some positive conditionedreinforcers.
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 57/89
ntavebals 102 ntavebals 103
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 58/89
ntavebals 102
e pesece of the sounds and fees of his ealie speaking and has beenefoced. Those sounds and fees of the eaie speaking ae stimui in thepesence of which speaking was einfoced. hey thus become disciminaive stimui fo the Rvs that foow them.
It is aso geneally tue that opeans of a sots, both erbal and nonerbal ae usualy ony einfoced intemittenty. The intemittency ofeinfocement fo opeants is itself caused. Intemittent einfocement fo
veba behavio is geneally not due to expicit pogamming on the pat ofthe einfocement mediato. Rathe, because einfocement mediatossedom pefom thei oes in an explicitly instuctiona way, they may beunawae that they einfoce the speake o fail to do so on an particuaoccasion. The mediato talks o acts in such ways that bing einfocementto hm. That the speake is aso einfoced is oen incidenta to themediato. Sometimes both may have to wait a while. The fist pat of theext illustation shows soe SD' fo tacts. If the speake just tacted
D
MAN
,
AWAKE PERSON AWAKE
EARLY FARMER v EARL"
A _ _ _
D GOES OBED EARLY
� - ?
HOWEER,-
v "EARLY TO BED
HEALH EARLY O RSE ESON WHO
A D GES UP EARLYAND GOES OBED EARLY
D MAKES A MAN + S WEALTHY HEATH , - S
WEALHY AND
�ISE
_
u
�_
D
SO, LAER HEALH] EARLY O BED D PERSON WHO GES ° EARL O RISE UP EARLY AND GOES + WEALTHY - v MAKES A MAN : - v "WISE
T BE EARLY 0 HEALTHY, WEALTHY WISE AND
eay," the heae doest get muc of a usefu discimiative stimuus. Ingeea, fagmentay tactig is useless to te heae because it does ot po
ntavebals 103
vide him with sufficient disciminative stimuli fo hi to behave in such away as to be einfoced. So, if the speake, as in the secod pt of the illustation, tacts in an extended fashion, that behavio is einfoceable (it povides the heae with an adequate disciminative stimulus) ad it is soeimes einfoced. Note that not all of the Rvs wee immediately eifoce.A Rv may occu seveal times in an extended utteance befoe eifocement occus. Thus the einfocement fo opeants with that Rv is ite
mittent.Now see what happens if we pick any Rv in an utteance ad chrc
teize its einfocement histoy. In this case ou Rv is wise." Fist, we ynotice that everything prior to that R v can be considered a a stimulus,
beit geneated by the speake himsef. The owest pat of the illustatio
shows this. The Rvs encased by the backet ae disciinative stimui,
auditoy ones if the veba behavio is ovet, as it sometime is, ad po
pioceptive if the veba behavio is covet or ovet. Remembe, self
poduced sma stimui ae egitimate. The speake can see, feel, and hea
his musclepoduced body movements. Thus the next tie the speake
Rv:SD "eay to bed and ealy to ise, makes a man healthy, wealthy,
and . . . , the R v w ise is ikey to occu. The subsequet ecuece of
the Rv wise" in the pesence of such stimui, that is, the stength of the
intaveba, depends upon how oten wise" has occued and been ei-
foced in that situation. The umbe of einfocements, the schedule of
einfocemet, the magitude of the einfoceents, and so foth, wil a
have thei usua effects on these intaveba opeants. Tese othe vaibes,
in tun, wi depend in lage pat upon the natue of the speke's evio-
ent in the past. Fo most speakes in the United States, the Rv:SD ealy
to bed and eay to ise" has been vey common in the pas, nd associated
with consideable einfocement; hence, the Rv: SD early to be d d eay
to ise, makes a man" contos stong intavebals. So also does the Rv:SD
chai"; table" is the ost ikey Rv to folow. The Rv:SD Shake-
speae" is less ikey to conto much, and the Rv:S D eleemosynay" hasextemey weak conto ove any Rv s fo most of us. We wil now look in
geate detai at the fist of the two casses of intavebals, namey, chains.
CHAINS
Muc nonverbal beavior cosists of chains. For example, the
bapess o keypeck is actuay a eatively fixed seies of esposes.
Because it is fixed, t is conveniet to view it as a single unitay thin, ct.
But actuay competion of te ct depends upon some stimuli geete by
t beavioitsef. ee is ow Kele and Schoenfeld analyzed the b-
pess in 9
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 59/89
raverbals 106 traverbas 107
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 60/89
raverba s
he lls ehells by the hoe"
A lue ox of by iscuits"
APLOLOGY REFLECTS T TENDENCY FOR SOME OF T REPETTIVE MATE-
RIAL, OR SOMETS SO OF TE NONREPETTVE PARTS O OMOGENEOUS
CHANS, TO DROP OT The pocess poduces fom of defective chai. Thishppens fo good esos though n such chins some discimintivestimuli occu sevel times, s shon in the next illusttion The Rv's tht
Rv S� - Rv S Rv S Rv S . Rv sg Rv S
WHERE
R v S Rv S = Rv sg
follo the epetitive Rv SD's , then, e liely o ocu en ny one o heepetitive RvSD's occs Fo exmple, RSD6 is liely to jump oveRSD4 nd RvSD d occu ight te RSD3 becse R SD3 nd RvSD
e the sme Becse the epetitive RvSD' e t umiguus, thespee's Rv's my come futhe ude the contl o he nonepetitve
RSD' hich my cod out te ptitive oes Ths is eected i sometymologies
cissusism cissism�cism
Pcicism �pcifsm
BbBb
O else, sluing o elision my occu, o n ttempt to diffeentite thepetitious RvSD' vi iection my be mde
Ou ext illusttion shos ht hppens i the development of spe-cies of vebl time bomb clled the overworked chai As the top pt of theillusttion shos, t fist stimulus cotol beteen djcet RvSD pis is
Rv: SO - Rv: S
O Rv SO
- Rv SO - Sr
Rv : S O - Rv SO � Rv S
O � Rv SO � sr
built up, but hen such linges hve eched mximm stength, moeemote nodjcet ones c go, s shon in the bottom pt of the ills-ttion These contolling eltions beten odjcent RvS D's c go
in stength thogh continued einfocemet nd becom
stog s thosbeteen the djcet vSD's TE EECT O T TT T
107
OVEWORKE CA IN, FOR ANY SD EAC R TE CAN S ABOUT
QULLY LLY. The links e stong, but they e not tightly ept in plcee esult hs tditionlly been clled spooerism. t s nmed e nOfod clegymn who ws especilly pone to do it, nd ho is believed tove been te pesQn who fist sid
We ill no sing Kinkeing Kongs
It is hde fo ich mn to get into heven thn fo cmel to psstough the knee of n ido
In the time of ou quee old en Victoi"
Spoone wsnt the only one who sid thing like tht A membe of theUS ouse of epesenttives Committee on the Judiciy sid, duingte eings on the impechment of Richd Nixon Thee e those howould fie the fuels of emotion bout the o of this committee" Themost fmous of ll spooneisms, believe, s y Von Zells mssceof ebet oove, " hee in the couse of tying to coect himself, hemust hve done evey possible vition of tht chin (oobet Heeve,"
eebe oovet," ) ht s becuse ebet oove" s t thttime both oveoed nd homogeneous All of this is misspeakig, but itisnt uncused" o ccidentl" o Feudin" We ill o discuss tesecond ind of intvebl, lustes
CLUSTRS
ICLUSTERS, ORDER AND G A SENT TCALLY, N
ARRAY O RSD's EVOKE ONE ANOTR WT NO PRTCU ORDRG OR
GRAMMATCL RELATIONS AMONG TEM TT S, RSD' just fom cluste, membes of hc e elted to ec othe simply by thei poe toevoe ech othe
Wodssocition expeiments usully poduce dt tht shoclusteing he next illusttion displys to exmples of the clustes thtmy be found in odssocition expeiments; the ist is ten fomGoodenough nd the second fom Plemo nd Jenins
We must be careful in interpreting these two examples. In c lustering,
te stimulus is a member of each cluster. There are six clusters shown for
te RvSD ing" nd eight fo the RvSD blue. Ring and are them-
selves members of each one of those clusters. This gives us a hint as to te
criterion for separating chains from clusters i our aayss.
CLUSTERS ARE BIDIRECTIONAL, WHEREAS CHAS ARE -
IECTIOAL So, ing" wom" isn't elly cluste; it s tverbal chain which was set off by the RvSD rig." A Rv:SD a
ntavebals 108 ntavebals 109
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 61/89
v GD DIAMND' HAND' FINGER':) "WEDDING': ENGAGEMENT
BXING
v S RINGe v NIS' CLANG ': BELL' DR
Rv "NIB
.
ELUNGEN
\ v WRM SING
v BLAC, BRW AQUA AZUR E, GD,GREEN, GRAY, NAY, PIN , PURPLE RE,IET, WHTE, YELW
v BNNE, CLHES, CTHI NG, CA,DRESS H A, CA GWN, JACET, PANTS
/
v LAE, CEAN, SEA , SY, WAER, MI SV
RIERv :
BLU v DEPRESSED, DEPRESSN, DREARY DUL,
LNELY, GLM, SA, UNHAP NSMv CLD, C, ICE, NRH WNDv BIRD, EAGLE, BLUEBIRD JAY, SPARRW
v DANUBEv RECR
blue is a member of several clusters, but notice that other members ofany one of these clusters are not members of the others, usually. To tell, seeif you can move back and forth along the arrow. Blue" is a discriminativestimulus for the v lake" and ocean" is a discriminative stimulus for thev blue," but ocean" is not a discriminative stimulus for the vpink," even though blue" is.
What all this means, in practical terms, is that a speaker who saysblue" for any reason (and the only reason that you would have for sayingit now is a textual one) then immediately has available a great deal of verbal
behavior; he is able to go on and say something on his own." He can movealong, within and between the clusters controlled by the vS D blue . " Onetopic remnd him of another, or he makes aluon and so forth. Theseoen reect the effects of two kinds of clusters formal clusters and the-matic clusters.
INTA VEBAL THEMATIC CLUSTES AE DEFINED NEGA-
IVEY HA IS BY EXCUSIN HEE IS N ACUSIC SILAIY BEEEN
vSD D HE v HA I CNLS This leads to the name thematc.Even though they don' t sound alike, two v's in any one thematic clusterare about the same thing, for example, lake, ocean, sky. They still aren'tchains, because the bidirectionality criterion is met. So, in the precedingillustration all the remaining intraverbals controlled by the vSD ring "are thematic clusters ecept for the intraverbal vSDv sing."
Psychologists have collected a huge amount of t itraverblthematic clusters. The wordassociation studies stated with the o ltonand continue to the present day. This is where mst of the experientl dataon verbal behavior come from. All the collections of wor nr romthose of Kent and osanoff to those of today (eg., Palero nd Jekis)show the existence of intraverbals and their commonality among peakers .Hee are some examples. I will not describe these experime rm which
they come in detail, nor will I cite them with full academic rigor they remerely illustrative.
Bouseld had subects read mixed lists that contained animal name,proper names, names of professions, and names of vegetables, allscrambled in a random arrangement. When asked to recall the words on thelist, the subects' v's appeared in thematic clusters.
Jenkins and ussell read aloud to subects disarranged stimulusesponse pairs from the earlier Kent and osanoff study. Subect werelater asked to recall the wrds that they had heard If a subect recalled aword, he usually recalled its associate immediately aterward.
Howes and Osgoo explored the method of aturatng tmul. Theygave their subects stimuli composed of a string of four words and told themto associate to the last word. For example, some subects received thestimulus 9, , , dark," while others received devil, eat, basic,dark," and still others heard devil, fearful, sinister, ark." By puttigmore of a cluster into the stimulus, the probability of a v from the rest ofte cluster rose. So that in the previous examples the Pr (v hel") sgeater for each of the stimuli as the number of neutral" stimulus rdsdeclined.
Howes and Osgood also investigated the promty effect. This mybest be defined by an example. The discriminative stimulu bull elongsto several clusters, each of which has other members, such as sexcow, "agecalf," clumsy in a china closet. " So if the discriminative stimulus
is feminine, clumsy, young, bull" and the subect is instructed to respond tothe last word, the v calf" is lkely. But if young" is moved away frombull" in the stimulus, the Pr (v calf") goes down.
Judson and Cofer examined what may be called the pror entry effect.Subects were presented multiword stimuli, such as skyscraper, temple,cathedral, prayer, and were instructed to indicate whichever one didn'tbelong. Now, cathedral" and temple" each belong to at least two the-matic groupsreligion and edifices. So, most subects will indcate ttprayer" does not belong, because the initial skyscraper" taps the eificecluster. However, if subects are devout churchgoers, skyscraper" will bechosen as not belonging. This is understandable, for such subject rerenforcements have occurred for v's in the religious cluster.
Foley and Macmillan looked at the pecal renforcement htory e
traverbals 110 nt raverbal s 1 1 1
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 62/89
fet Tey ooed a e differen associaions of aw scoo, medicascoo, and psyology sudens o su c words as adminiser , " compain," ce" (se"). As you mig imagine, given e SD adminiser"e medica sudens came u wi Rv's suc as aid," dose," drug, "rs aid," medicine," syringe," and rea"; wie e aw sudensgave esae," govern," judge," law," money," and oa." ifferen reinforcemen isories (bacgrounds) ad differeniay sreng
ened differen clusters in e wo groups.Lieray ureds of wordassociaion sudies ave been repored.
Tere were so many studies a H. C. Warren, wose Hstory of thessocaton Psycho[ogy publised in 9 , coud no even is em al.He pointed ou ta in e cassical wordassociaion lieraure, cerai raional classificatory scemes ad bee proposed. For eample, aperedwas said o sow efining; roseflower, superordinaion; owerrose, subordinaion; mawoman, common accompanimen; and darlig,opposies. But Warren was forced o concde, s w now a mus,a is sor of ierarcical cassificaory enerprise ed nowre. Tere isno funciona or dynamic difference among ese cluser pairs. Tir fi, or
ac of , ino various classificaory scemes doesn' e us anyingusefu.
Once we now abou emaic clusers, we are in a posion o eplaie penomenon a as variousy been erme medated generalzaton orsemantc generazaton Te ne ilusraion, wic sows is penomenon paradigmaicaly, is aken from Mednick. Par I of e iusraionrepresens e presening of e word lght wrien on a piece of paper anan eecric soc to a uen speaker of Englis. In Par II we see represened e resu of presening te wor dark wic ad never been pairedwi e sock. Mediaed generaizaion was supposed o be a errible probem for simuusresponse psycoogy and a grea embarrassmen to i. Remember a SIMPLE STIMULUS GENERALIZATION IS THE TENDENC
F STULI THE THAN THE NE VLVED IG CNDITING TOEVKE T SA ESPNSE Te common cases occur wen e two (ormore) simuli fal aong e same pysica dimension, or ave common elemens or accompany eac oer as par of cusers. Well, cleary, if we avean Engisspeaing subjec, and e oes wa is diagrammed in Parts aII of e iustratio, it is geeralization But i certily isnt sipe stimulus generalizaion, and i wouldn' wor wi a pigeon or a speaker ofFrenc. A lso, don' be deceive ere; am no rying to snea in a blurringof e operan versus responden disicion. Te GSR (galvaic skiresponse) is a responden penomenon, a cange in e eecrical resistanceof e sin . Bu if we ake everying into accut, as sown i Prts I aII of e ilustraion, en we see tat tis effet turns out to be a cove
ient ndicaor ta sometn else, perat beavr, s goi on. So, a
PART!cs L iGHTI
GSR SHOCK
PAR I DARK
�
GSR
PART
IGHT
EXUAL
o \
NRAERBAL
•
� v SD LI GHT� v : SD DARK"
U LIGH DARK-- GSR
PART I' cs I
D
RK
EXUA INRAERBAL
DARK v "DARK � v "IGH"
Y
�
___
_
�\
GSR
oug e measured effec is iself no verba, is occurrence eens upona special verbal condiioning isory.
Wa does e meiating are e Rv:SD i te braces. Tese re codioned cover operans, ose simuus properies serve as s for terespondens. So we see, in Part I of te ilustratio, tat wat apeare toe jus a CS is also a discriminaive stimulus for a textual, ta is, tespeaer is disposed o say ig. By virue of e past istory of reiforcemen wic as made im a uen speaker of Englis, wen e is isposeo say lig" , e is aso liely o say e Rv s wi wic ligt" clusters
dar," eavy," and so for. So ese Rv's, suc as dark," occur before e soc and ge e GSR condiioned o eir propriocepive smulusconsequences. Ten wen dar is presened on e paper, a etual occursand aso ses off an iraverbal, bo of wic ep to elicit te GSR, assown in Par II of e illusration. Finally, we can sae a LA F MIATE GENERALIZATION: IF A NE ESPNSE IS CNDITINED T NE
MEMBE F A THEMATIC CLUSTE, IT MAY BE EVOKED B THE OT ES
F THAT CLUSTE
I AN INTRA VERBAL FORMAL CLUSTER THE CONTRLLED A
CONTROLLING Rv's SOUND LE EACH OTHER. BUT THE ACOUSTIC SAY IS
NLY P ATIAL. TH ESPNSES D NOT ACUSTICALL EPRDUCET M-
ULUS, AS IN THE TRUE ECHOIC. Aso, ere te dscriminative sims is s-
generated.
ntavebas 1 1 2 ntavebals 1 1 3
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 63/89
ent and osanoff caed responses of tis sort cangs" wen teyoccurred in wordassociation eperiments. or eampe, if te v:SD isring," te foowing forma custer memers occur: sing," ing,"ting," ping, " ming," wing," zing," spring," sing," and sofort. orma custering is not restricted to te wordassociation eperiment. yming poetry is vera eavior were some responses of formacusters appear in certain specified paces.
Aimpication of te definition of forma custers is tat if a sound
occurs in one's own speech, it tends to recur soon. Suc sounds tend to cluster. Skinner reported his investigation of tis in is paper A QuantitativeEstimate of Certain Types of SoundPatterning in Poetry." Putting tis another way, the intervals etween te same ponemes, words, or prases arenot randomy distriuted. Tey are too sort. Here is a eurstic eampe: Iftere were twenty /p/'s in a fourundredponeme sampe, ten te meannumer of ponemes etween /p/'s would e twenty, and te distriution
NUMBER O PHONEME BEWEEN EAH
of te numer of oter ponemes etween successive occurrences of/p/ woud ook ike tat sown on te eft side of t e aove iustration, if
it were governed y a random process. But, if in fact the process were notrandom, te actua distriution woud ook ike tat pictured on the rigtside of te iustration. Tat is, te numer of oter ponemes occurringetween successive occurrences of /p/ woud e muc smaer, and tesesma gaps etween /p/'s woud occur more often than woud e epectedy cance. Te /p/'s woud unc up. We, te rea resut of tis custeringis tat casua, spontaneous," noniterary, nonoratorica speec is aitera-tive, ryming, assonant. Te interesting case is tat of ponemic or po-netic custering. Sometimes te resuts are rater painfu. Te foowingquotation originay appeared in a serious ook review y McNei: Sincete intentions of te autors are more modest tan te dustjacket 's en-comium, they shoud e spared te odium of eing te erads of psyco-ogy's milennium. " A professor of psycoogy was oveeard saying: Iave no rapport wit appoport, ut e is wrapped up in it . . ."
kinner studied forma custering eperimentay. He invented a de-vice tat he caed te ebal summato wic is a kind of vera projective
test. It consisted of a phonograph record composed oy of owe sounds:a e a o 00. hese were arrange in intevals of 3 , 5 with two a
1 1 3
ways accented, for eampe, a o o 00 e suject in the eperimentistens to te record unti e can ear wat it is saying, although, i fact,noting is eing said. Te foowing is a sampe protoco, wich Skinnercoected and reported in ebal Beao
ee n'est partie"do not say your part"take eave of t"o, are you"got your visa"ee ne sait pas"pppartie"are you going"wo are you"visvis"
Serock Homes woud say: Elementary, my dear Watson; the subject
speaks renc and s aout to go on a ourney. However, we see that muchof te suject's vera eavior is simpy ecoic, ut not sefechoic. Thevowes of te responses tend to ecoe., matctose of the samplesequence on te record; their numer and stress matc too.
But importanty for our purposes here, the veral summator has o
consonants . Rememer that intraverbas can occur etween prts o f wors.
o a consonants in the suject' s responses are intraverally cotrolled.
he stimuli tat evoke te consonants are te vowel sounds of the record
wic te speaker echoes, and te consonants that he has just emitted be-
fore as part of teing what is on the record. They occur entirey too oten
to ave been ited from te record or standard Engish. Also, the following
v's occur two or tree times: part," visa," aitpas." These facts show
te extreme constriction of v forms; just a few tings get said, one or two
sma forma custers, and tat is aout it.Tematic custering sows up too. We can see it in tis protoco; ack-
ing anyting on te stimuus side to narrow it tematicay, it narrowed it-sef. It a seems to e aout a trip to rance. Tis is ecause the v's areweaky controed y te discriminative stimui from te record. So, a-toug it doesn' t aways appen, here at east we got formal custering within a tematic custer.
Bruce found a simiar tematic custering effect in a different expri-ment. He payed recorded ists of words with a ackground of white oise.He found tat is sujects misidetified te words o te record but put t
errors in te same thematic groups as te words they correcty recog.Tat is, parts of the boy or foods wee identified as separat clusters
traverbals 114
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 64/89
t
each custer incue wors that the subjects sai they hear but whichweren't on the recor Extracuster errors were rare
Expicit reprouctio of rhyme an aiteration requires in the irstpace, that the speaker have an itraverba forma repertoire Skinner hasshown that Shakespeare in his sonnets in't aiterate any more thanmight have bee expecte on the basis of norma speech He might wehave raw his wors out of a hat Swinburne on the other han was
accoring to Skiner's research very much an aiteratorThere is an expaation for forma custerig t is a form of response
inuction This process has been frequenty observe in nonverba behavior RESPOSE TON IS T TENDENC F ESPNSES SIMILA
FM T T NE THAT AS EFCED T CCU T SA SITUATIN
HICH EINFCEMENT AS DELIVEED Response inuction is case bythe genetic costitution of mammas t occurs in barpressing keypeckingan aso taking So if a rat was reinforce for pressing a bar with a forceof grams he is ikey to press it with a for of 20 r or 9 or grmsSimiary if a ma is reinforce for saying ring he is key to say sing "
mlipl asaon
An that sai Jack is that We have r trogh ltheprimary variabes that contro verba beavior We ooke at each oe as ifit existe by itsef in a onevariabe universe an acte a aoe We tr toarrange situations of that sort in the aboratory where or contro o otherariabes ets exacty one variabe work at ay oe time The wor of theaboratory is just as rea as the wor otsie it; te wor isie teaboratory is not fictiona But in the extraaboratory wor many varbesare present at any one time an they interact Like nonverba behavior ver-
ba behavior is muticontroe Our previos chapters have cosiere eachanteceent controer of behavior as the soe operative variabe This wasfor ease of efiition an exposition ow we have to pt the variabes together an show how they interact to prouce observabe noaboratoryverba behavior
PROBLMS ARISING ROM MULTPL
CAUSATION
A given speech intera contais responses contribute b svfferent cotroing variabes I ' present some xampes, whc fa
from a te possibiities; workig out a the combinatios wo mbusywork , ' , ;'
ut ip e causation 16 ultip le causation 1 7
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 65/89
Mutipe tacts My new white si shirt"Mand Audience: John, oo "Tact + Echoic: You said helo Textua + Intraverba: Your verbal behavior ater you read Eary tobed and early to rise, maes a man "
eading aoud, which is mosty pure textuas, can have mistaes Theseare usuay due to the intrusion of intraverbas which are stronger than thetextuals, so these mistaes are Textuals + Intraverbals
These exampes are not merely an exercise They support the pointthat at one evel of anaysis each component of speech is nothing but"conditioned operants Thus it is ie eypecs and barpresses At thiseve, verbal operants are nothing new, and nothing that was neverpreviousy reinforced ever happens That is supposed to be the big defect ofthe whoe approach It seems to mae no provision for creative verba be-havior, for spontaneous speech owevern hs is the importanthoweer tha some peope seem to have sippedthe utterance as a wholemay be insightfu, creative, new, never said before by either the speaer or
anyone ese Even though whoe utterances may have ever preiosy occurred and been reinforced, their components have Th next iustratingives a simpe example If I can tact the moon and green cheese, as in the
D MN v HE MN"
D C v GEEN CHEESE"
D MOND v "HE MN GEEN CHEESE"
C
upper part of the iustration, then when the two discriminative stimulioccur together, I can tact them both, as shown in the ower part of thefigure emember, if at some time, somewhere else, a variable acquires con-tro of a response, if the variable occurs again, even if in a new combinationwith other variabes, the response is liey to occur too It drops into asorts of behavior, a at once"
So, if the discriminative stimuus yeow-green book is presented to aspeaer, and his v chartreuse" is reinforced in its presence, wheneveranything yellowgreen next occurs in the speaers environment, he is ielyto say chartreuse," as in chartreuse liqueur" or the dress is chartreuse" These examples are simiar to the case of the man who learned tosay moon" to one discriminative stimuus and green cheese" to another,
and then said the moon is green cheese when the two disriminative stimui occurred at the same time ets oo at another more ompex exampe
of this process If a child who is learning to tal earns one intraverbal suchas ex" as part of express," it appears a at once in every other word thathe either aready nows or wi earn Thus it may replace es, sy, in ex-ain," expore," extra," and so forh Previousy, he migh hve saidespain" or esplore" or estra" The discriminative stiml tht cmeto control the v ex" in the context of express" was pren ll longwhen es" was previously said in espain," espress," an so forh So
when the v ex" got conditioned to that discriminative imulus icrowded out es" and was said whenever the discriminative simulus oc-curred and poduced the chids nowing how to say explore" andextra" without any obvious history of reinforcement for pronouncingthese words correcty
urthermore, any one varibe contros more than one v The nextilustration shows how this applies to al the primary variabes and their corresponding operant casses that we have examined
AC
MAND
v "
v HEEFD"
D v "FEMAEv "EDBE"
FD PEASE"N "GIE ME SMEHING O EA"
"v AXI"
AUDIENCE / OCABUAYDAUD
,
AUDIBIITY" v SEECION
CHOIC D "HA v "HA"EXUALS
SD TEXT � Rv "TEXT"
INAEBA / GOD"v D "ING v "BXNG"
v "NOISE"
At first, this seems to be a to our benefit Because eople tl somuch, it is reassuring to see that an apparenty constant envionme
control generate) a great dea o verba behavior We dont nee bror progression o onsantly varying contoing variabe; a st
ul ip le causaion 1 1 8 t i p le causat on 1 1 9
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 66/89
omet is potet enough. Futhemoe, as systematic theoists we ca lookgood. We ca usally give a plausible ecostuctio of ay actual v. If wekow what is said, we can be cofidet that we ca fid some cotolligvaiable o elation pesent
But i the log run we have a embaassmet of iches. We havegive the speake too much to say. We must explai what detemies whichv occus, because not all of those that ae likely at ay oe time do appea.
Some pocess smoothly ad quickly detemies which will get said.Taditionally i was said that the speake chooses his wods," but
thee is hardly time fo that. An accomplished speake does't stop tochoose befoe each word o setece. Fuhemoe, that theoy eects abad scietific logic. If thee is a choose, he must be located ad ·his choos-ig behavio must be accounted fo. So, acually choosing is ot a solutio.It is meely pushing the poblem backwad into some ukow o uknow-able domai
Suppose we ae able t fid out what aiale s opeative t the timethe speake speaks. We the have to predict what he will say. Aother tadi tio, that of stimulusesponse psychology, advised pedictig the stogestopeat, namely, the opeat that had te most favoable (in tems of the
schedule ad umbe of eiforces ad thei magitude) eifocemet histoy. If a opeat has had the most favoable einfocement histoy, it Rvshould be the most pobable v. This solutio woks sometimes; if youstick to it, you will be ight moe ofte tha ot i the log u. But, weako ae v's do occu, ad thei appeaace has to be explaied. Hull'ssolutio made this a spontaeous pocess, which is ot very good scieceeithe. Afte caefully piing espose pobability dow to seveal decimalplaces, he put i a adom oscillatoy vaiable. Whee it was i the wold,o oe kew. But, it poduced stage esults, accodig to the theoyesponse pobability would swoop up ad dow appaetly on its own!Therefore, according to this theory, a rare response occurs when its oscilla-
tory inhibition is momentarily low and the oscillatory inhibition of all theothes is mometaily high.
We can do bette and stay within the confines of the natual sciencelimitations that we have imposed upon ouselves. Help is at hand, becaseany esponse, including Rv's, may be contolled by moe than one vaiableAnothe way of saying this is: All Rv's are membes of more than one ver-bal opeat.
Hee is anothe example which makes the poit. Conside a man whosays oil." O some occasios it may be a mand, as i Oil that door; isqueaks." On other occasions it may be a tact, as when the man from the drilling rig ran t o the ranch house yelling Oil, oiL" It may be to a greadegree uder audiece cotrol. This is the hardest to see. But if you were in-
troduced to an auience coposed o Mr. Getty, Mr. Rockeeller, the king
of Saudi Aabia, and the sheik of Kuwait, soone or aer o would belikely to say oil. " That oil " is pat of textuals o echoics is obvios Andfially hee is a itavebal example (ead it aloud and lisen o ose:The sultan of Tkey has a teible tempe. If you make a mike, e wlboil you i . . . "
How all this helps us out with the pedictio poblem c b e i wst ecall what we know about supplemenay stengthening he irs w
pats of the next illustratio show this paadigmatically he sm put ude the contol of two sepaate vaiables o two sepaae occio
V. l< 2
v 3
� V 1. v 2
+
3py 2 v4
v 5
MPURE TACT
GREEN LGHTEP
�
-
IMPURE MAN v
/ PV 2, 4
5
N v
DEP-- v "CHATEAUBAND ,
+ EXT. MU A'·
�PLEASE EAVE
AUDENCE� GO AWAY
AVERSIVEDAUD(FY)- SHOO"
the it may occu oce whe the two pimy vaiables ae concurrenlyactive. That is, if two cotrollig vaiables fo the same Rv ae presen athe same time, the effect o Rv pobability is additive. So, one response fom among the many that might be said, give one vaiable, ay e uplemetary stregtheig fom some othe vaiable. Thus a ohes eweak esponse to ly oe contolling vaiable may occur beca
supplemenay sources o stegh fom other cocurenly presen
vaiables I so, he problems o espoe competiion, occuec � v's, and pedicion o wha the speake will ay re
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 67/89
u l iple causa o n 122 u p e causaon 123
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 68/89
MISSPEAKING
f you liste caefully to wat someoe says especially by liste-ing to a tape ecodig of is speakig you may ea te esult of a supplemetay vaiable tat cotibuted a Rv it cotolled eve we it was'teeded. Sometimes espose competitio is a poblem; failues to esolve itquickly ote lead to misspeakig. Tis is to be expected. f supplemetay
vaials cotibute to espose pobability tey sould eveal temselvesclealy fo tey ae ot weak cotolles. So supplemetay vaiables caote be see to stegte vebal beavio tat is ecogized as beig wog if it is oticed at all. Tese occueces ae lawful as Feud waste st to poit out. But claim tat is explaatio o wy tey occuedwas te tig tat was wog. Te followig classes of examples ae pob-ably ot omogeeous o eite te cause o te effect side but tat's watyou'd expect wit mltiple causatio.
NTRUSONS CC HEN A Rv THAT S LL BECSE F SE
MENTAY VAABLE'S PESENCE DSPLACES A EAKE NE cogessma atte Nixo impeacmet eaigs said: Few ay oe of te tigs doe
by te special uit wee atioal secuity. " ote said: Too ay membes ae begiig to tik tat tey ae castig the fial decisi ee. da woma edecoatig e ouse said: We ca put a walke mea a little·
ue i te all."STORTONS CC HEN A TMATCALLY STNG Rv DISPLACES A
FMALLY SMA NE Te decoatig woma was also a citic of medicalpactice ad said: You do ot subscibe a dug just because you do'tknow wy someoe is avig a seizue e edecoatig se got a ewstove fo te old stove blew up because te pilot ligt was jimmy-igged. d some of te cogessme foud te testimoy ad to believe;oe said: t seems to me ate icedulous . . . "
BLENS MAKE S SSPECT ABT EAL STENGTHENG AND MPE
FECT RESOLUTON OF RESPONSE COMPETITON. Lewis Caoll was a maste blede ad was esposible fo fumig" + fuious"-fumious";bight" + glitterig" billig" ; ad mery" + whimsey" mim-sey. A radio commetato explaied some of the activities at the impeac-et heaigs as follows : N ow she is talkig about the tasfomatio fomthe Keedy admiistratio to the Johso admiistatio" (trasitio +tasfe). A woman was offeed a policewoma's puse ad said: "If I buy that puse, I'll look like a uderclothes woma (udecove aget +
plainclothes woma). This is't limited to portmateau wodsit alsooccurs as pase blends suc as emmig ad awig about te bus" for that matter of fact it i s easier for a rich man to pass though a camelthan get ino heaven " when the gum wi grow back I have no way ofsaying," "I am so mule braied. "
FOLK ETYMOLOGES AE DSTTNS F EXTC DFCLT
DS wmCH SALLY SH A SPPLENTAY NFLENCE THEY ASALLY CMBNATNS F PATAL ECHCS, PLS FMAL THEATC FAG
MENTS A BLEND • Palme compiled a etie dictioay of tem ad was ableto indicate some of tei souces. Hee ae some of tem :
alabasteyellow plasteawdeead ioapoplexyappy plexte sip sailed oud te atipaties (atipodes"te aquaium (equiem sevice fo te deceased pope"my usbad is a egula sipo (cipe"
e tells atidotes (aecdotes of is yout"e cild died ad se wet ito asteisks (ysteics"tey ad te bads (bas put up"blue as azueblue as a azo
Cout of Commo Pleascommo placecoqulles Sait JacquesJack' s cookiescota dasecouty daceduplexdouble xedemausflity mousetesoldfes wooda widow by gace (a uwed motegace widowgass widow (a
woma wose usbad is goe to te golf couse most o tetime
gaga fisguad fisemigateammegate
apooapig ioysteicsig stikesalteoldeomilyumbleiotaooteoot (so te folks ow say: do' t give a oot"
jaudicejandesjasmiJessie's owegedamesJoy dabiesjay bidjoy bidcankekangao
quelque cose (Fenc fo someing"kick soes
ul ip le causaio 124 ul iple causaio
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 69/89
lecternant ornenten lleantorn llellaclly oakmermadmerry madnoncalancenne llngacornoak cor
panterpanterpentouepant ouepeony roepano roeplace (a knd of f)pla ukepomegranatepound garnetcuonquetonlampreyramper eelrendeerranged deerreynard (renc for fox")eynoldceleryalary
extonaxonconcence
cet autre coe (renc for t ometng ele") anoter par ofoe
currltyqurrltySt. Vtu danceSt. Vper dancegoamerummer gooetelegraptalley gra
delerum trementrangletwarttougt
tolettetwlgtwatercrewater grawater craewater creae
Even Sakepeare dd t: Te bet courter of tem all could never avbrougt er to uc a canary" (quandary) n Merry Wes of Wndsor;cvl a an orange" (Sevlle) n Much do bout Nothng
Finally, if words are fairly strongly related intraverbally, they soundwell together (are reinforce able) even i f they are complete nonsense, as inTroops are beng flown into Vietnam by the truckload. " Or as the movie actress said, "I made twice as much money as Calvin Coolidge put to-
geter.
VRBAL NG INRING
Sometimes verbal behavior must be delivered acco to soe
pror pecfcatn. Th preent a problem for u n our rol engs. for we have no direct access to, and cannot directly manipult speck i�elf. We olve t problem n a way tat reveal a very mportt prncione tat underle te wole Sknneran approac. We manplt abe of wc te dered beavor a functon. Tere are two
gBclae of manpulatonte prompt and te probe.
POMPTS CC HEN HE FM F HE DESIED VE�
IS SPECIED ADVNCE IN PMPG, ENGINEE SUP
CNLLING VAIABLES AND ADDS ME AND ME F HEM Dl
DESED SPEECH CCS If I want you to ay ranklin, I myp
o
you by upplyng you wt dcrmnatve tmul tat I belev cool thv. I do t on te ba of my knowledge or etmte of yourp �of renforcement for peakng. or example, I mgt ay to yo:
Thoma Jefferon ad of hm: 'I ucceed hm; no o c �
m. He gned bot te eclaraton of Independence n t U S C�ttuton. "He opened te Pot Offce."He nvented a tove."Wo nvented te rockng car?"Wo nvente te lgtnng rod?"Wo wrote Poor Rchard's lmanac?"
He flew a kte n a tundertorm. "It begn wt an F"
POBES CC N HE CASAL VAIABLE IS SPECIFD I YAND HE QESIN IS HA VEBAL BEHAVI IS CNLLED BY I tubect repertore to be explored, rater tan extende. Aga, ti iuually accompled by prentng dcrmnatve tmul of varou or,a n Wat te dfference between forma and tematc cluterg?" Wat do you tnk about abolng te deat penalty for knp
All of tat fne, obou, and agreed to be true by mot popl jua long a te peaker omeone oter tan temelve. ut te radcal taementte procedure eem to be te ame w w mengneer our on verbal beavor. Even an accomple ekr does�manpulate own verbal beavor drectly. He can only u .
varable of wc peec a functon. n oter wor, te I
merely and only place wer varable act, and not a vr.
*
ltip le causaton 126 1
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 70/89
as an entity ontributes nothing. Horrors ! First Gaileo took the earth fromthe enter of the osmologial stage. Then arwin put man among theanimas. Now Skinner is ompeting esartes's program and is puttingman among the mahines. It seems, however, to be a neessary onsequeneof a strit adherene to sientifi materialism.
Let's briey look at a few examples ilustrating selfprompts and sef-probes. If you must give a speeh, or write a term paper, or pay a part, or
take a test, you don't ust pik out al the relevant words from your voabu-ary even though they are all there. You do, however, prompt yourself,usualy with textuas, or oasionay with ehois that ontain this be-havior, that is, you onsut books or experts. In ating, one uses intraverbas that are started off by ehois from the person aled the prompter.
Traditionally we laim that we hoose our words to express our viewsor feeings" about apital punishment, or display our knowledge" aboutformal and themati lustering But what we atually do is inspet ourverbal behavior itself, in order to know or feelings. Whe presntedwith a disriminative stimulus suh as What do you think about apitalpunishment, we probe ourselves. We may start taking aoud, as in di-
tating a rough drat, or else we may do it overtly. That is, in orr to findout what we have to say, we may first listen to what we have to say to ourseves when we talk overtly to ourseves before we speak aloud. We may beable to do this to some extent whie speaking aloud, espeially if the overtspeeh is slow and has pauses. We listen to what we have to say aboutapita punishment, or any other subet matter. One woman wasamazingy aurate, but somewhat ingenuous, when she said, don'tknow what I think unti start talking; that's how I find out . I just open mymouth and throw myself into it.
1
autoclii
The topi of this hapter, autoitis, is the mot ubtle p the theory. It i sn't ompliated, or hard to understn the wy teis, but the proesses desribed here are very busy. t is iffl o ee ll their impliations at one. Moreover, it is here esily th kiertheory runs ounter to our traditionl views. This is the pr the nlthat has seemed most obetionable to the traditional nlys o sekinand to the new mentalists.A impliation of the aim of the analysis that we have been exinin
in this book is that all the variables that are neessary and sufient forgenerating speaking will be disussed and their effetive ontrollinreationships desribed. So, we shoud inuire whether the variables th wehave onsidered and their respetive operant asses are suffiient taount for al of speeh. n other word, has any known variable been eout, or are there any parts of speeh not aounted for? t seems tht in extended utteranes exept textuals, ehois, and intraverbal hains, thereare two response harateristis whose ontroling variables hve eenomited. Consider the folowing examples, whih l go into in de
at man is fat."That girl is ot at. " veryone i here exept ohn.
1
utocitics 128 utocitics 129
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 71/89
All swans ae white"opeully, the plane will be on time"
Upon looking at these examples, we see ou ist poblem t appeasthat some Rv's ae not accounted o by the opeant classes that we havediscussed Tacts would account o that man" and at," but not is"We have to nd ou what contols the Rv is" t isn't pat o a tact o
existence; eveything exists, so ne is wholly nondisciminable; you can'teinoce in the pesence o existenceper e and extinguish in its absence nthe next utteanceThat gil is not at"a tact may account o thatgil, and fat, in this utteance, may possibly be unde echoic o intavebal contol But, what contols the Rv is not"? t can't be abence inand o itsel that contols is not" Fo whethe she is o is not at, she always is not a e engine Then thee is the poblem o what contols theexcept" in Eveyone is hee except John" All swans ae white,"whiteness can be the disciminative stimulus o hite, but all wans canot be the disciminative stimulus o all swans The hopeully is sup ·posed to be an adveb modiying on time" Well, it may be, but that is no
cientiic explanation o ow it got said n all o ou examples o poblemse st detemine what the contolling vaiables o hese Rv's ae, andhat thei physical loci and compositions ae Ae all, such Rv's as is ,"" ot," ll," and except" ae pats o opeants; have been condi-ed; ad must have causal antecedents Theeoe, pat o the job e-nngto be inished in this chapte is to identiy the causal antecedents oe Rv's and give a plausible account o thei conditining histoies
ontolling vaiables aely occu in any special ode, again except tose vaiables that contol echoics, textuals, and intavebal chainsOn contolling vaiables ae concuent, and i they ccu ad eratum ode does not paallel that o the Rv' s Thus, ou second majo poblm is to account o the odeing o Rv's in speech Fo example, the vai-
les that cntol the Rv that man" an the Rv at" and the Rv is"ou all a once, but the Rv's do not
These sots o wods and the odeing o Rv's ae said to be gam-tical" So, we have a popety Rv's that may be called gamma But do not intoduce a gammatical vaiable ee gamma is the name o fct, and not a cause
FTO PRGM
ammatical wods and the odeing o Rv's ae what Skinne
cae auolt behavios I'll give the denition shotly, but ist let's io it b ooing at a paadigm Autoclitic behavi is a esut o a
twostage pocess, which is shown paadigmatically in the ollowing illus-tation n the uppe pat o the illustation the st stage o the pocess is1LRV
' ' . . ? RvN N
�vD E P D E P , . �
Rvepesented HEE PIMY CNLLG VIBLES (P V '
S) F HE SS
H E HVE DISCUSSED E PESEN ND CIVE ND INCESE HE CU
EN MMENY PBBILIY F HE Rv's F SEVEL PENS. HS IS
CLLED PRMARY VERBAL BEAVOR SICLY SPEG IS OL
LECIN F PRMARY OPERATS These opeants may be mands, tacts,intavebal clustes, small chains, echoics, audiececntolled Rvs, anything we have examined so a, all jumbled up in no aticula odeThe speake now has the ability to speakhe has smething to say example, a hungy man pesented with an apple by his ministe
is dispsed
h frt ed to say many thgs e mg t say app e, p ease, , ,h k " I d God povides, " sauce," t an you, an app e a ay
When, and only when, that ist stage has happened, ca te speakeespnd to his alteed cicumstances and his own dispositi t sa healteed cicumstances and vaious dispositions to emit Rvs seve as scm-inative stimuli o him as shown in the lowe pat o the peceding illusta-tion e can discimi
ate and talk about the vaiables affecting his dis-
position t speak and the Rv's that they contol Fo example, e might say: was so hungy " o This eminds me " o was ust abt topay " This talking about talking is autoclitic behavio t is smewhatsedependent and includes such Rv's as is," is not," except," an
d
all " And, it includes odeing We will look at many examples o thpocess in geat detail in what llows
ow, omally and oicially, N AUTOLT IS C HSE D-CIMINIVE SIMULUS IS CUENLY SNG PEN pY
VEBL BEHVI ND HSE Rv IS DISCINIVE (VIES) WH SE
H PEN'S CNLLNG VIBLE SENGH Rv FO
We may say, with somewhat less igo, that the pi vea
behavio makes liely te vebal ehavio (i e, autcitic) tha upon it, and a detines the oe of its emissio. S, W n
utocl it ics 130 utoc it ics 131
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 72/89
do two thngs. We respond to the prmary variables and to our verbal oper-ants themselves and we play t all out as a running commentary on what weare dong.
Notce that autoclitc operants are a special subspecies of tacts. Therdscrmnative stmul are operants not ust Rvs and their Rvs may serveas discrmnatve stimuli for ntraverbals. An autoclitic s a comment oncompettion among operants or the resolution of the competiton or the
strength of the controllng variables or the past history that led t thoseoperants n the primary repertore and much more whch we wll soon examne in some detail. But it all works like tactng and so it snt really verynew. irst however we should analyze some examples in detail.
KNDS OF AUTOCLTCS
Some autoitics tact the controling ariales fo primar erbl behavior. Suppose my current priary variables lead me to say She is n
Toledo. This tells my hearer nothing about why I would say this. However
t may be important for my hearer to know the causes of m behavior .
Hs reinforcement may depend on his subsequent behavior. That subseuent
behavor may be very dfferent from what it would be f my behavior
was part of a textual as opposed to a tact. So tell him by tacting the con-
trolling variable.
see by the paper (that she s in Toledo)."A textual My brother says . . . n echoic predct that . . . "Not a smple tact possbly an ntraverbal guess that . . . "Not a tact probably an intraverbal
Thus the primary verbal behavior remains the same but the effect on thmedator s considerably dfferent.Some autoitics tact operant strength f the variables controllng He is
in Chicago" are functioning the ollowing ta their strength
am sure that . . . Very strong know that . . . " suppose that . . . "Not so strong beleve that . . . " guess that . . . "Not very strong magne that . . . "
I hate to tell you this but . . . " Rather weakbecause o ctitin
hesitate to say so but . . . "
So we can see that Hopefully the plane wll be on tme" tells the hearerthat the tact the plane will be on tme was not quite stron nuh t besad by tself and to wait because some other tact may follw.
Hopefully is related to those autocltcs that tact Rv competition inthe primary verbal behavior. or example put anothe way" st"
or " however " although " are all Rv s that tact that anothe piytact is coming up or could have been emtted and that if or whn idths next tact competes with the ne that ust was sad.
Some Rvs have been called quantiers and this terminology s silstrong n phlosophcal crcles. Remember that in the standrdphilosophers example All swans are white" there is no discrinatistimulus for the regular tact: S D Rv all swans. " t is ipossible to see llthe swans that there are much less all those that ever were or eer wil b.Therefore emsson of the Rv all swans" in their presence is impossibbut one can respond to hs verbal behavior about swans. The next ilus-ton shows the dynamics of this tact. The Rv all functionally tells thehearer that n th prmary verbal behaior of the speaker i a swan is thedscriminatve stimulus whte" s going o be a Rv bcause of a hih-strength tact. Remember all" s part of an autoclitic; swan" is pa o a
SO SWAN • v SWAN- v WHTE
;
O
-v ALL
SP N
. PIGEON
.
GRAY
WHTE
S• SOME
S PRFESSOR� v ROFESSOR R IC
�
y
I
SO
v NO
132 utocit i cs 133
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 73/89
t, t qantifiers tat the numerial properties of the relation between arog variable and a R v in primary behavior.
TH 'AUTOCLTIC A "FUDG E FACTOR?
To some people, this is the plae where we have gone out the
or nd limbed up a metaphysial tree This whole business seems too$ge to them, or strains their redulity too muh. It does not seem, to pople, that a operant ould be a disriminative stimulus. It does n otsem to the to be a disrimiable thing, some real entity that an be re-d to. They argue that operants are only abstrations, and bad ones at.
el, I don't think hat things are as bad as all that. Operants are real,pubc vnts. They are abstract, to be sure, but their members are observ-
, d so is the relationship beween ther members. Therefore, uch
ios a serve as the dscriminative stimuli for tacts If your operants
ae o secret to an outside observer, a scientist, they are no secret to you
ouslf. After all, you have the best access o the controlling varibles that
ft you and the responses that you make. So you can observe these con-
rolng variables, tact them and tact the relations between them. And thes
l tats are autolitis.Actully, we tat our behavior and our operants all the time, both
hn the behavior is verbal and when it is nonverbal. Somehow we are lessrpid when a speaker tats his nonverbal operants in a desriptive way.The t part of the nxt illustration shows an example of this. We often tat
SO CLOCK
)SO COA ON
� GO HOME
SOFAIGUE
DEPN
�
___
.
v " WAS JUSLEANG"
SO PAS B
+ SO OR PN� v
SO
-v AWAYS SAY
SO MMINEN B+ p\
�
v "I WAS ABOU SA
the behavior of nonverbal animals in a similar way For example, we maysay of a pigeon: He is about to pek the key. When the bhavior is veralthe proess is exatly the same, but it appears to be somewha diffrn be-ause many of the Rv's of the autoliti voabularyeg., is," is no,"exept," all," o," some,"do not name or desri hviorOther autoliti Rv's do desribe: I always say . . . ," rad ht ," I was going to say . . " The first kind of autolitis (eg, al," x-
ept ) us ovary with features of primary verbal behavior in ovioways. That is all that is required; remember he nondesriptiv nonnigtats that ontain the Rv's hello," good morning," phooey" Thetoo are tats that ovary with the disriminative stimuli but don't na odesribe.
We may agree that autolitis are tats, but then to be good sietiss,we mus show why they our at all. But we know why they our. Thhearer needs that information (disriminative stimulus) Any Rv may aused by more han one variable, that is, e a member of more than oneoperant. So the hearer must know not only what is said, but why it is sai,an oten he must know how strong the ause of its bing said is H n
to know this beause his behavior may differ depending on what th situ-tion is, and his subsequent einforement depends on the bhavior h ifollowing the speakr's Rv. Reall our example of the man who ays oi"His hearer' s future ehavior is grealy determined by whethr tha Rv i pof a mand, a tat, an intraverbal, an ehoi, or whav hu h hrreinfores the speaker for tating his ontrolling varibls d h srhf their ontrol. When the speaker and the hearr are in th s sin, h
harer has ust as muh aess to th primary variables s h spkr othus, and this is an empirial, testable hypothsis, hen he spk is hhearer the autoliti ommentary probal drops ou. eithr is' hror, if it is, it ours with very redued amplitde or in an inomplt form
Let's look at some more examples of auoiti ommentary. The next
llustration shows the dynamis of assertion and negation As soon as, bunot before, the Rv's tha man" and fat" are probable, then is" oursas part of a tat about the relationship between the speaker's verblbehavior and its soure. It tells the hearer what the anteednt of the vralbehavior is You say is" whn you have something to say that has a ertairelationship with its anteedent, not ust when something, anythng, othe hearer is serves as a disriminatie stimulus that funtions to ell hhat the same disriminative stimulus is ontrolling oth tats, tht "and fat. " So i s" tats that both tats have the same isriiistimulus.
One that girl" and fat" ae likely, for some reason, if h Rvfa" is due o o vaiable other han the fatness of th dii
stimulusperhp an ehoi, or a textual, or even an intvr � -
mentar oure of hth spaker sai ' t .
e
uociics 134 uocliics 13
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 74/89
v "HA MAN"SO
v A
�
- IS
O HA GIRL � v "AT GIRL
O OER VB - v "Ai,
" ISNT"
O EVERYONE � v "EVERYON"
(MEONYMCAL0 v "JON
S
v "XCP
isn t" until he is in a position to say that which is not That is, we say isnt" when we have something to say, not when notness is pesent The
v isnt" seves as a disciminative stimulus to the eae It tells him thatthe v at" is due to some othe vaiable than the at o he discimintivestimulus It is unctionally equivalent to sayig hat at" is not pat o atact o that gil "
Theeoe, i eveyone, that is, i all the membes o a goup have inthe past been common accompaniments o John, then John" is likely tobe said in thei pesence, even i John isnt thee When this does occu,ecept" may be emitted to seve as a disciminative stimulus to the heaethat John" is not in this instance pat o a tact but is being contolled bysomething else That is, it tells him that John" is likely to be due to thatom o genealization that we called metonymical etension
the autoc1itic tacts occu I havent given a complee list, no will I,
but hee ae some moe eamples The v and" usually occus as pat oan autoclitic that tells the heae hat thee is moe to come, that the speakewill speak again unde the inluence o the same contolling vaiables Thev o" is a disciminative stimulus to the heae o two o moe opeantsin the speakes pimay vebal behavio which ae both at high stengthand competing ten these opeants shae the same di sciminative stiuluso he same v A lot o grammatcal taggng is autoc1itic Technically theseae simple tacts, such as walk," walk," ng" These seve asdisciminative stimuli o the heae to inom him o the tempal elationships between the contolling vaiables and the speaking So it goes Alhough this ist is not complete, it shows the geneal eatues o the auto-
iti voauay Autocitics compise veba behavio that oesnt happenunt h is somhin ese o say the somehin ese ovens it The autoc pee s akn aou takn
At this point it might be appopiate to conside a g. Rv mybe pat o an autoclitic opeant in one utteance, but o be tclitic inanothe The v is" may be autoc1itic when the disciminive iml ia at peson, suh as in the Rv that man is at" oweve, he me Rvis," cannot be autoclitic when the disciminative stimuls conig h;speake s behavio is () anothe peson s v in which case the v i p an echoic), () witing o pinting (in which case the v is pat o a exl),
o () the speakes own pio vebal behavio (in which cae the v is pato an intavebal) Eamples o these ollow:
( ) ECOC A v "SAY E CAKE IS GOO"-u- B Ov "E CAKE S GOO"
() TEXUAL 1 S S T -
v "TIS S
INRAVERBA
I you say the ollowing alo u, yo iitil vebl ehvi wil textual but you will keep going ne itavebl cotl ee e meadages:
A stitch in time saves nineLook beoe you leape who hesitates A penny saved
ORDRG AS A AUTOCLTC ROCSS
Let me emind you o the second pat o the poblem beoe The vaiables contolling pimay vebal behavio oen occ simle-ously, o in an ode dieent om that o the iished speech Now, whollows is not vey intuitively acceptable o many people We ont oi-naily tend to speak o the seial position o one v in a sting o R a oneo its dimensions o behavioal popeties Nonetheless, elative emplposition (ie, when, with espect to othe vs, a given v occs) i s mcho an obsevable dimension as v om (ie, which wo, phoneme, phe,etc, is emitted)
ODEING I ALSO A AUOCLIIC PROCESS: IT S SY
ALTRNATIVE TO USING THE AUTOCLITIC WORDS (' SUCH AS IS," "AO,BU") . U AUOCIIC ORDERING IS NO A EXCLUSIVE ALTERNAT TO
utoc itics 136 utocl it ics 13
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 75/89
HE ACLIIC DS I IS SED PALY IN PLACE F HEM, AND PALY
IH M. OERIG IS ISELF A FM F ACING CNLLED BY HE
DISCIMAIVE SLI IN PIMAY BEVI. It deped upo oper-at i primary verbal behavior, jut a the autolitic word do. Youdo't order omethig ule you have omethig to put i orderHow verbal operat are ordered deped upo themelve, upo what i tobe aid ad why.
Remember, a autoclitic i a operat. Like all operat, it ha acotrollig variable ad a repoe. Oe cla of autoclitic ha ecialword a the repoe. The other cla of autoclitic ha order of word athe repoe. Eve though a particular order eem like a odd ort ofthig to be a repoe, it ca be. Ad if it i cotrolled by the right ort ofdicrimiative timulu , it ca be part of a autoclitic operat We haveparticular order a repoe i otalkig behavior, a i typig fromcopy. Strikig the key of the typewriter i the order H I T S i a differetrepoe tha trikig them i the order T H S , if a particulararragemet of the v' of the operat o primary verbal behaviorcovarie accordig to a covetioal correpodece wit either the co-trollig variable or the tregth of the operat, the thi aragemetfulfill the defiitio of autoclitic that we gave o page 9 Particulrarragemet covary with particular cotrollig variable or tregt ooperat, or competitio amog operat, jut he ay autoclitic Rv' do.Therefore, the arragemet or orderig are autoclitic too Becaue theycovay with evet i primary verbal behavior, orderig tell the heareromethig about the circumtace cotrollig the emiio of the primaryverbal Rv'. The et illutratio how the differece ad great imilaritybetwee the autoclitic word olutio ad the autoclitic orderig olutio to
SO� v "RED" } PRIMARY VERBAL " " HAOR. v B
K
, AUTOCLITIC WORD v "TH BOOK 15 R D"
OR
v "TH R OOK"•AUTOCLTC ORR
the problem of tellig the hearer that the ame dicrimiative timulu ibeig tacted twice. Thi how that the order of the Rv' ha the amefuctioal role ad i cotrolled by the ame variable a the Rv i.
It might eem omewhat wateful to have two way of doig thi Juthik of all the additioal time ad reiforcer that were e eded However,
the ext ilutratio how how orderig ca be a much more efficiet meth-od of tactig ropertie of the primary verbal behavior . If a red book i o a
"R" "BOOK
GRN"
v TABL
Iv TH OOK 5 R"
"TH TAL 5 GR" v"TH BOOK 5 O TH
TAL"
OR "TH R BOOK 5 ON
TH GRN TABL"
gree table, the peaker ha available the followig Rv' at oce: "red,
green," b ook," table. " Thus he can sa y, The book is red nd he
table i gree H e ca alo dicrimiate the locatio of the book ad ay,he book i o the table ." But, itead of givig al thoe Rv' , clu
all the autoclitic oe, he ue order of Rv' , that i, The re book o h
gree table. He make a dicrimiatio. If red ad ook re boh
aid (that' the dicriiative timulu), the ook mut be afe r
(that's the autoclitic response) Of course, this doesn't alwas wok u so
nicely. Sometimes the variable controlling book is very strong, and the-
fore the conventional order breaks down somewhat and the speaer says;
The book, the red one, is on the green table, the one in the hl."
RFORCMT HISORY FO
AUTOCTICS
What follow may be difficult for ome people. do't wa ooud defeive or patroizig here. However, everal people who read hbook i maucript form uggeted tha you be wared that thi may be aplace where you will have to do a lot of the work to udertad. P!eaedo't imi the difculty of what follow a jut lack of claty or le
utocl it ics 13 utoclitics 139
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 76/89
tioning a uent speaker must have had some reinforcement history forautoclitics.
Such a reinforcement history probably started in the way shown in thenext illusraion n he eariest insances of he speaker's verba ehavior
D2 CAR
GOO O
GERALZED
� v RE CA �
� v GOO O
v OBC CLASS AM v PROP �LAS AME
S
� vPROP�RY-OJ C
he was presented with discriminative simui tat had multiple properties aof which were tacable So he learned to tact them usually as sound nswith builtin order. For example some chidren learn all gone as a singleunitary Rv and it takes a while before it is separable into all" and gone"with each under separate discriminative stimuus ontrol Sometimes theseparation requires considerable explicit tuition The dispos ition the oper-ant is very strong and t generalzes, as shwn in the bottom part of the pre-c�ing ilustration
Now this is extremey important especially in light of some of thecriticism raised against our approach Once the child learns a particuarorder and this order generalizes; it has become autoclitic Furthermore all
situations (discriminative stimulus combinations) need not be separateylearned. Some people seem to think that autoclitics require separateconditioning for each paricular Rv combination that is a separateeinforcement history for each of the following combinations: good b oy "red book" nice doggie" and so forth That isn't so; it would takeforever. We know that generalization takes care of it Generaization isinstantaneous; one reinforcement may be sufficient to get going withwidespread effects So i f the child is effectivey reinforced for red book "or he goes" or if then " these orderings generalize A morphemic ordering example is provided by the child who who learned the orderup" er" and generalized it to downer. " Autocitic orderings likeother conitioned responses drop in al over the place once they have been
attaced o a iscriminaive stimulus. Tha's ecause he iscriminative
stimuli were all over the place all along and once the responses are putunder their control the responses occur whenever the simui do.
Furthermore response nducton helps this process aong Rememerthat response nducton s the heghtened probablty of occuence ofresponses that are smlar n form to the response that was renforce Sostimulus generalization allows the response of propertyojec order otransfer to other discriminative stimuli (propertyobect pairs). And the v
form of propertyobect tact order tends to recur by induction. The seemindifficulty in comprehending this is due to the fact that generaization andinduction are aspects of the same process in verbal behavior
Such speed and extent of generaization and inducion may seemimpossible to some people. But those who do research on generalizaion andinduction tell us that they are both very broad wide spreading proces-sesunless they are specifically contained by discrimination raining andeven that wil not wipe them out The human organism is an especialy goodgeneralizer and inducer. This may be part of the uniqueness of primaegenetic constitution. Men and monkeys are generalizers The probem facedby educators is to teach discriminaions Pigeons and horses are gr dis-criminators but poor generalizers.
Other orderings are similarly acquired Here are a few moe eapesIf the discriminative stimuus is an actor performing some acion he v'sthat tact the actor occur before those hat tac the acin as in he boyruns" the girl cooks" we study." You can draw the pardigm for his; just use the lower part of the preceding ilusration s a ode ere re a
couple of examples of open frames. If (i rains; I study; you say you wil)
then (I won't go; I get an A; wi too) So (good righ epesive (
couldn't resist; had o shut my eyes; I couldn't buy it).
AWARNSS
Skinner's autoclitic hypothesis is that we discriminate our own
operants However this does not mean that we are aware that we are doing
so. That is we have no veral behavior for which the autocitic processes
themseles are the controlling discriminaive stimui f we did we woud be
able to tact our autoclitic processes as they were occurring. Presuaby
such tacting could be taught (reinforced) but i probably woud have o
make use of nonvocal Rv's ecause the speechproducing musces re usy
This approach to the probem of awareness says tha we my ehvediscriminatively with respect to a se of discriminaive simui woknowing that we are doing o or even what the relevant discrinivetmuli are. We can quie ofen specify the conroling variales fo o
behavior as in I' m having hree superb urgers ecause haven' hd
tocitics 10 toclitics 11
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 77/89
breakfast or lunch"; said it's raining' because it is, that's why" Quiteoften we cannot, though You can see some nonveral examples when youtry to specify the variables that control your walking, or riding a bicyceYou cannot tel someone how to do it Another somewhat more verbalexample is provided by my wife To claim that she tals without awarenessof what she is saying woud e too risky or me However, she was able toengage in a highy sied discriminated orm o verbal ehavior without
knowing how she did it or what the discriminative stimuli or her ehaviorwere Furthermore she was taught this ehavior without having thediscriminative stimui named or descried or her his all came about inthe ollowing manner
Ater her ather had quit the rodeo circuit as a perormer, he was ased
to judg e severa events. With the passag e o time and many more rodeos, he
gained a reutation as a good judge o cutting horses His aility to judgethese anmals came in part rom his experience as a rider o them in rodeosand in part from his experiece as a workig rnher . . Cuting horse are
specially trained animals used y cowboys in seecting a cow or a steer fro
a group and cutting it out o the herd, therey isolating it rom the rest o·the cattl so that it may be tied or branded or whatever Because o theimportance of these operations in practial ranc work, cutting horse eventsare o great interest to serious ranchers and roeo attenders When my wiwas about ten years old she egan to sit with her ather while he oiciatedas a judge of cutting horse contests, and soon she began to eep score, abeitunofcially At the end o each horse and rider's perormance, she wouldcompare her estimate o the perormance with that o the ocial judge, herfather Aer a ew years and a numer o rodeos, she was ale t o developjudgments that agreed with his within one or two points Generaly oth sheand her ather woud assign the same score to a contestant
So, here we have an example o veral behavior, the assigning o a scoreto a horse and rider, which is based uon a highly complex series o
discriminative stimuli, the perormance o the man and the animal, whicwas learned without explicit tuition, and which cannot e descried by thebehaver hersel That is, even to this day my wie can judge cutting horseperformances, but cannot say what cues she uses, nor can she tell me hatto look for She can also tel me from what sort o animal the armer toothe natural ertilizer as we drive past the ield, but she cannot tel me whtdiscriminative stimuli dierentiate te odor o horse versus cow dun Youand don't speciaize in discriminating odors, ut erume chemists dohey can be trained to speak in response to the discriminative stimuli thatgovern their behavior and to tact them with some degree o precision, justas the soil expert can learn to tact the discriminative stimuli that govern hisbehavior in identiying ertilizers Once this is done, the door is open to the
explicit verbal tuition o others in the same sils
Sli of the tongue are caused ut oen ass unnoticed As hve saidefore, a good way to find these is to listen to recordings Aother goodsource is a classroom lecturer f, inste ad of trying to write it o you ayattention to what the proessor is saying and how he is sayig it y o wlhear all sorts of intrusions and distortions You can oten dis thirprobable causes But notice, the speaker usually discrimintes t scontrolling variales nor his Rv's, because he does not recognize tt is
misspeaking Sometimes he does and corrects himsel," but this sting occurs much less requently than one might imagine i one as tstudied tae recordings n general, then, awareness presents no diictisor the autoclitic explanation of verbal behavior
ther more highly verbal examples abound, too A somewhat esotericon is automatic iting he writer's hand, pen and paper may econcealed from his eyes y an eevated board He answers questions ywriting but may read an interesting story or carry on a conversation wit theexperimenter on some other topic ten the suject doesn't now that hehas written nything, mch ess wt e hs writte or y Ts hproduces verbal behavior, albeit written vera ehavior, witot yawareness o the discriminative stimuli by which his ehvior is controdor the Rv's that he emits
Leon Soomons and Gertrude Stein pubished one o te earist son this topic in the Pychological Revie in 9, and B . Sinr rtean article in the tlantic Monthly in 9 ointing out the gret siiitieso style and content in Miss Stein's 9 pper and her ater ity orsespecialy Tende Button He claimed that these similarities indict ttshe had written her novels by the automatic writing rocess r tethree samples of Miss Stein's automatic writing that were ined in te9 paper:
1 ence tere s no possle w o vong wt I ve spoen o n ts not eeve te people o wom ou ve spoen ten t s not pos
sbe to prevent te peope o wom ou ve spoen so gll 2. Wen e coul not e te longest n tus to e n tus to e testrongest
3 Ts long tme wen e d ts est tme nd e coul tus ve eenbond and n ths ong tme wen e could e ts to rst use o ts longtme . .
Miss Stein became so good at automatic writing that distractio yreading was almost unnecessary. Miss Stein found it suiient distrtio
oten to simply read what her arm wrote, but folowing three or or orsbehind her penci
Let us now oo at some sampes from Tende utton At this t aside is in order Sinner's argument is much more comex tha t
parison of te sames o Miss Sein's automatic writing
utoc it ics 142 utoclitics 13
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 78/89
1 A tabl mans dos it not my dar it mans a who stadinss Is it likythat a changA tabl mans mor than a gass vn a looking gass is tall A tabl mansncssary pacs and a rvision a rvision o a litt thing it mans it dosman that thr as bn a stand a stand whr it did shak.
2 China is not down whn thr ar pats lights ar not pondros and incaclabl
3 Rs BeefIn th insid thr is slping, in th otsid thr is rddning in th
morning tr is maning in th vning thr is ling. In t vning thris ling. In ling anything is rsting in ling anything is monting inling tr is rsignation in ling thr is rcognition in ing thr is rcrrnc and ntirly mistakn thr is pinching Al th standards havstramrs and a th crtains hav bd inn and all th yow has discrimination and al t circl has circling. is maks sand. [Paragraph on o a 36paragraph pom by iss Stin]
M Sten dened Sknner' clam and h ltear peo oectedvgoroul to Sknner' artcle. Perhap the were worred that the explana ton took the art" ou o the product. Judth Thurman recentl contnued
the lterat' protet agant Sknner' anal. And he reprntd RoatBee ! I have to agree wth Sknner who wrote that t wa unortunate thatM Sten pulhed th tu or t detracted rom her other morevaluale work. Automatc wrtng not art. It no more art than ampleo the varou order o approxmaton to Englh. Here are ome ampleo thoe collected Mller and Selrdge:
1 . thy saw th play Satrday and sat down bsid him
2. road in th contry was insan spcially in drary rooms whr ty havsom books to by or stdying Grk
3 go it wil b pasant to yo whn I am nar t tab in th dining roomwas crowdd with popl it crashd into wr scraming that thy had bn
4 hos to ask or is to arn or living by working towards a goal or his tamin od wYork was a wondrl plac wasnt it vn plasant to talkabot and lagh hard whn h tls is h shold not tll m th rason whyyo ar is vidnt
In hi Creating the Creative Artit, " Skinner oerve tat the crea-tive artit i an expert in variation and election. The variation (original idea,themes, materials, teatments) come from many sources. The experimental analyst of behavior has no access to these, so he cannot tell us abot them.But the creatve artt uuall ha no veral repertore for the ource ethe.He han't been reinforced for tacting them, o he i unaware of them too.he book Anna Livia Plurabee: The Making of a Chapter, edted b Hg genon how that Joce dd jut what Sknner clamed. Finnegans Wakeeai ha the compexit and in part the redundanc and amgut of Mi
Sten' automatc wrtn t a and t ucce ul (noc o otthe writer and the reader).
Th ha een a on dreon ut I have tro ehoaout" all t. Incdenal t nteretng to ee that Skn noject o on to the hmant even at the egnnng o rr Bmore mportant or the act that M Sten' atomac r wthat gncant veral ehaor ma be produced dcrmt m
o whch the ehaer not aware. And the eave ma lo e aweo h rpone ntl he examne them at a later te
1Some i mpl icat ions 145
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 79/89
1
some
implcatons
We are in the homestretch now. All that remains is to apply whatwe know to some problems that have greatly concerned the traditionalanalysts of speech and verbal behavior
THE HEARER
First, let us pull the hearer together, so to speak We have dis-
cussed him at various times in the preceding chapters, but we have largelyconcentrated on the speaker. It is now appropriate to put the hearer side of
the fluent speakerhearer in perspective This is mostly ju st a resume o f whatwe have said before The hearer has two major roles
His first major mode of functioning is as stimulus generator. He doesthis in two ways WHEN PRODUCES STULI AFTER A SPEAKER RESPONDS;
CALL HIM A REINFORCEMENT MEDIATOR. Without him, the response
reinforcer relation does not exist; and without that, operants cannot be crea-ted or maintained In his role of mediator, he behaves slightly diffeently foreach class of controlling variables for the speaker's operants He must dis-criminate among controlling variable Rv relationships affecting the speakerin order to reinforce appropriately WN THE ARER PROVES STIMULI
BEFOE TH SEAKE BEGS TO SEAK,W CALL HA AUIENCE Here
44
he functions as a discriminative stimulus and shows up in address and as a
variable which helps select among Rv's and audibility in a supplementaryway
The hearer's other major role is as a stimulus receiver AS A HEARER
E SE HIS BEHAVIOR IS PRCIPALLY ECHOIC; HE LISTENS THAT IS, SE
ING IS COMPOSED OF ECHOIC BEHAVIOR. Usually this echoic behaior i
covert, but it can be, and sometimes is, overt Listening, or hearing, i not
just stimulus absorption. It is not merely a passive transaction between teear an the brain via the organ of Corti and the tympanic nerve The hear
of speech does things as he listens, and his doing things is part of his hear
and necessary to itSpeakers have their behavior modified by the heaer's actiitie, a
their behavior reects this Effective speakers talk at rates that are
conducive to echoism: not too fast, or there wouldn't be enough time or
the hearer to echo; and not too slow so as to prevent the hearer frostarting on intraverbals which might interfere with his echoing the speaerEspecially in dramatic speaking or emotional exhortations, the pace ospech must be varied so that some words (Rv's) may elicit emoionarespondents, and the elicited responses may then subside Remebe, a
stimulus may serve as a discrimiative stimuus for sea oeatresponses, and also serve as a CS fo other respodent behai o
example, the rat's visual food pelet may serve a a disciati tfor operant s
eizing and chewing as well as a CS for respodn aiation
Appropriate rates of speaking ar e selectively reinfoe y t h haThis is generally done by termination of negative cnditioned reifoesSometimes punishment is appied when the speaker responds at rates that
are not conducive to the hearer's echoing For example
Hurry up; don't take all day"
Get on with it, man."
Slow down you talk too fast"
Take it easy
Relax, I won't go away"
The hearer does this so that h e can echo and·in some cases produe
intraverbals If the stimuli prouced by the speaker change too ast, th
hearer cannot do that. If the stimulus ow is too slow, it is aversive o the
earer. In the latter case he has to mark time with behavior that t o
chains or clusters of intraverbals which are competitos for what i to
echoed next. This is, of course, related to U NDERSTADIN G, W I
OUR ABILITY TO ECHO A USE OUR ECHOIC Rv's AS DISCRINATV s
FOR OU INTRAVERBAL BEHAVIOR WCH LEADS ULTATELY TO O U"
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 80/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 81/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 82/89
Some implications 12
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 83/89
CENRALSM / \{ BP- BP . BP - BP
SD RvSr
1 /BP\
BP\ r 0 r
PRI PHERALSM S R :S R S Rv�S
SDR: S
DR SD Rv �s
response. Moreoer, centriss sy tht the rin is spontneousy ctie,tht it cn initite trin o proses iniio
Osere creuy tht the PERHEALISTS' POSITION DOES N DENY
T EXISTENCE OR NECESSITY OF CENTRAL NERVOUS SYSTEM ACTION
TNKING IT, AS DEPICTED IN T WER PART OF T nLUTRATION, ARGUES
THAT BRAIN PROCESSES A BY THEMSELVES N ICENT OR TK-
G. THE PERHERALISTS CONTEND THAT ADDITION NCESRFACTOR IS PERIPHERAL EVENTS, THAT IS, STULI ARISING FROM MUSCLE-
PODUCED MOVEMENTS
Unortuntey, the wy to decide which o these two posit ons is corrcis conceptuy cer, ut technicy impossie. we coud show tht competey pryzed person or rin in jr cou thin, then we woudknow tht peripherism is wrong. Suppose we were e to eep the rino mn ie in jr, s in the science iction stories Then we coud gieit n rithmetic or geometry proem to do. It woud he to e proem we were sure it didnt now the soution o ut did he theinormtion to soe. Then, i we remoed extern stimution sourcesor whie, sy h n hour, nd then reconnected the communictionmchinery, we coud see i it hd the soution I it did, nd the process
w repete with new proems, nd it coud te us wht it did whie itws disconnected, then we woud now tht peripherism is wrong ndcentrism is correct Becuse tht experiment hsnt een done, peripher-ism is sti ie terntie, nd it is consistent with dt we het this time
We he competed our exercise We see now tht compete scientiicccount o why mn sys wht he sys when he sys it cn e gien Thisccount mde no ppe to eents, entities, or processes somewhereoutside the domin o stimui or responses. Minds nd rins, grmmrsnd intentions, ides nd cognitions, he een shown unnecessry or compete, cus, nturistic ccount. Mn himse hs een eiminted s
cus rie; he is ust pce where cus ries interct toproduce tin
sugetionfor her readin
or a introduction to reserch y Sinner n his oower,see the oos y
Hond nd SinnerHonigKeler nd choenedReeseReynods
s see Sinners oos
T Bvior of Orgnim
Scinc nd Humn Bvior
T Tcnolog of Tcing
Coningnci of Rinforcmn
Vrl Bvior
Skinner' s views on scientific method are stated in the above ad i
Beyond Freedom and Dignty
1
Suggetio n for further readi ng
Cumulive Record
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 84/89
Cumulive Record
Aou Beviorim
Advncd w n Snnn ch cntnu nd ptd nbo by Ft nd Snn, Hong, nd Hndy, xmpl ThJournl of e xperimenl Anlyi of Bevior nd th Journl ofApplied Bevior Anlyi gully publh vl thund pg
pt ch n th Snnn ttn ch y
reerence
An G.W William James; a biography. w Yo Viing 1967
Boomid L. Language. w o Hot 133
Bosd W. A. h on o sting in th o noaangd assoiats Joual of General Psychology, 1953 49 229-
B D J Ets o ontxt pon intigibiit o had sph. In .hy (d) Information theo papers read at a symposium o <ioraotheory' held at the Royal Institution, Londo, September 12th to 16t h 1955, pp24552 London Bttwoths 1956
Byant and Ain J. R. Psychology of Englh. w Yo
Combia Unisity Pss 140Cao L. psdonym o L Dodgon Alice's adventures in wonderlad.
London amian 1865
Danpot D B angag. Science, 174 186 95
D Lagna G Speech: Itsfunction and development. Boomington Unisity o Indiana Pss 12 163
Ebbinghas H Uer das gedchtnis. Lipzig Dn Hmbot 1885issd as Memory A contribution to perimental psychology, tns H ARg and C E. Bssnis. w Yo Do 1
Ein S . Langag and A ontnt in biingas. Joual oAbran ocial Psychology 164 68 5-507
Ests K W. Som ts o inomnt pon b bo ohidn Unpbshd dotoa disstation nisity o innsoa 5
Ests W. K. Laning nnual evew ofPsycholoy, 1956 7 -8
1
References 156
Ferster C B and Skinner B F Schedules of reinorcement New York:
References 157
Howes D ad Osd On the combination of associatie prbabili
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 85/89
Ferster, C. B., and Skinner, B. F. Schedules of reinorcement. New York:Aeton-CenturyCros, 195 7.
Findley, J. D. An exerimenta outline for building and exloring muti-oerant behavior reertoires Joual ofthe Experimental Analysis ofBehavior, 1 96 2, 5,11366.
Foey, J. P., Jr, and Macmillan, Z. Mediated generalization and theinterretation of verbal beaior: V. 'Free association as related to differences inrofessiona training Joual ofExperimental Psychology, 1943, 33, 299-310.
Frisch, . . Dance langage and orientation of bees. Cambridge, Mass:Harvard University Press, 1967.
Frisch, v., enner, A. M., and Jonson, D. L. oneybees: Do they usedirection and distance information rovided by their dancers? Science, 1967, 158,107277.
Galton, F. Inquiries int o huma n faculty and its development. London:Macmillan, 1883.
Gardner, R. A ., and Gardner, B. T Teacing sin language to a cimpanzeeScience, 1969, 165, 672.
Goodenough, F. L. Semantic choice and pesonlit strutre Science, 46,104, 45156.
Greensoon, J. The reinforcing effect of two spoken sounds on the frequencyof two resonses. American Joual of Psychology, 68, 409-16.
Guthrie, E. R. The psychology of leaing. New York: arer, 1935.Hahn, E. A reorter at arge: ashoese New Yorker, 1 December 197 1, 47�
5498.
ale, . Ethnography and philolgy. Piladelpia: Lea and Blanchard,1846 Ridgewood, N.J: Gregg Press, 1968
anson, N R Discussion. In M. Radner and Sinokur (eds.), Analyses oftheo ries and meth ods ofphysics and psychology; Minnesota studies in the philosophy ofscience, Vol. IV, . 243. Minneaois: Uniersity of Minnesota Press, 1970
Hayman, D. (ed.) A first-draft version ofFinnegans Wake. Austin: University of Texas Press, 1963.
Hendry, D. . Conditioned reinforcement. Homewood, Ill.: Dorsey, 1969.
Higginson, F. H. (ed.). Anna Livia Plurabee: The making of a chapter.Minneaois: University of Minnesota Press, 1960.
Hidum, D C., and Brown, R. . Verbal reinforcement and interiewerbias. Joual ofA bnormal and Soci al Psychology, 1956, 53, 108-11 .
Hoijer, H. The Sair-horf hyothesis In H. oijer (ed.), Language inculture, . 92105 Chicago: University of Chicago Press, 1954.
Holand, J. G., and Skinner, B. F. The analysis of behavior. New York:McGraw-Hill, 1961 .
olz, C., and Azrin, N. H Conditioning human verbal behavior. In W
K. Honig (ed.), Operant behavior: Areas of research and application, . 79826.Englewood Cliffs, NJ: Prentice-al, Inc., 1966
Homans, G. C. Soial behavior: Its elementary forms (rev. ed). New York:Harcort, Brace & Jovanovich, 1974.
Honig, W. K. (ed). Operant behavior: Areas o research and appication.Engeood Ciffs, NJ renticeall, Inc., 966.
Howes, D ., ad Osd, . On the combination of associatie prbabilities in linguistic contexts American Joual of Psycho logy, 1954, 67, 24158
Hull, . L riciples ofbehavior. nglewood liffs, NJ: Prenticel, Inc ,1943.
Irwin, O. C. Infat speech Scientc American, 1949, 181, 2224.
J akobson, R Stdies on child language and aphasia. Te ague: Mouton,1 9 7 1 .
Jakobson, R, Fant, C. G. M., and Halle, M. Pre/�minaries to speech
analysis: The dtinctive features and their correlates Cambdge, Mass.: MT.Press, 1951, 1952, 1961, 1963, 1967
Jenkins, J J , and Russell, W. A. Associative clustering during recall.
Joual of Abnormal and Social Psychology, 195 2, 47, 81 8-2 1 .
Jesperson, O . Language: Its nature, development and origi. Lndon: Gere
Allen Unwin, 1922Johnson, D L Honeybees: Do tey use the direction information contained
in thei dance maneuver? Science, 1 967, 15 5 , 8-47.
Joyce, J . Finnegans wake. New York: Viking, 1939
Judson, A. J . , and Cofer, N . Reasoning as an associative process. L
Direction as a smle verbal problem Psychological Reports, 19 56, 2, 46976.
Kagan, J , and Haveman, E Psychology: An introduction. 2nd ed. New
Yrk: Harcout, Brace & Jovanovich, 1972.
Kaplan, E. L , and Kalan, G. Th e prelinguistic child In J. Elit (ed.),Human development and cognitive processes, . 35 88 1 New York: Holt,
Rinehart & Winston, 197 1
Kelle; , F . S . , and Schoenfeld, W. N. Principles ofpsycholog : A systematic
text in the science of behavior. Englewood Cliffs, NJ : Prentice-al, Inc., 19 5 .
Kent, G. H , and Rosanoff, A. J A study of associatin in insity
A merican Joual of Insanity, 1 91 0, 67, 37-96, 3 17.
Lorenz, K . Stdies in animal and human behavior, 2 ols Cambridge, Mass :
Harard University Press, 1970, 19 7 1 .
McGuigan, F. J . Thinking: Studies of covert language processes. Englewod
Cliffs, NJ: Prentice-Hal, 196 6
McGuigan, F. J . , and Schoonover, R. A. (eds ) The psychophysiology of
thinking: Studies of covert processes. New York: AcademIc, 1 973
McNeil, E B. Dependency: Te egg ardens when te butter melts. Cotem
porary Psychology, 1960, 5, 390-92.
Mednick, M. T. Mediated generalizatin and te in�ubation efctas a
function of manifest anxiety. Joual of Abnormal and SOCal Psychology, 195 7,
55, 3 15 -21 Miller, G. A. Langage and communication. New York: McGraw-HIll, 195 1 .
Miller, G . A . , and Selfridge, J . A Verbal context and the recall f menful
material. American Joual of Psychology, 19 50, 63 , 17 6-85.
Palermo, D. S . , and Jenkins, J J. Word-association norms: Grade shl
through colege. Minneapolis: Uniersity of Minnesota Press, 19
Palmer, A. S . Folk-etymoogy, a di�tioary of erbal corptiO w�r� erverted inform o r meaning, by false dervatIOn or mstaken analogy nn .
Bell, 1882.
Rs 158
ost E L ml Post s tqutt 12th rev ed New York Fnk & Wagnals
Rrncs 9
S K W B N H C Y
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 86/89
ost, E L ml Post s tqutt 12th rev ed New York Fnk & Wagnals,1
remack, A fnctonal anayss of angage Joual of t xrmntalna lss oj Bavor 10, 14, 1025
eese, E xrmnts n orant avor New York AletonCentryCrofts, 14
eynolds, G S A rmr oj orant condtonng Glenvew, ll Scott,Foresman, 18
osental, xrmntr ffcts n av oral rsarc New York Nabrgbshng Cor, 1
ssell, B M losocal dvlomnt New Yrk Smon & Schster,15 © 15 by George Allen & Unwn Ltd
Sar, E A norma l ts of sc n Nootka Canada, Geologca Srvey,Memor 2, Anthroologcal Seres No 5 Ottawa Government rntng Brea,115 ernted n Slctd wrtngs ojdward Sar n languag cut ur and r-sonalt 1- Berkeey Unversty of Calforna Press, 14
Sar, E Male and female forms of seech n Yana In St W J Teewen (ed),Donum Natalcum Scrnn Utrecht, Njmegen, 1 erted Sltdwrtngs ofdward Sar n languag cult ur and rsonalt 20-1 Berkeleynversty of Calforna Press, 14
Sknner, B F Has Gertrde Sten a secret? tlantc Montl 134, 153, 50-5
ernted n Cumulatv rcord: slcton oj ars 3rd ed, ; 35-Englewood Clfs, NJ Prentce-Hall, Inc, 12
Sknner, B F The verbal smmator and a metod for te stdy of latentseech Joual of Psco log 13, 2, 110
Sknner, B F T avor ojorgansms Englewood Clffs, NJ rentceHall, Inc, 138
Sknner, B F A qanttatve estmate of certan tyes of sond-atternng noetry mrcan Joual ofPscolog 141, 54, 4- ernted n umulatvrcord: slcton ofars 3rd ed, 3140 Engewood Cffs, J rentceHall, Inc, 12
Sknner, B F Scnc and uman avor New York Macmlan, 153
knner, B F Vral avor Englewood Clffs, NJ PrentceHa, Inc,15
Sknner, B F. T tcno log of tacng Englewood Clffs, NJ PrentceHall, Inc, 18
Sknner, B F Contngncs oj rnforcmnt: A tortcal analssEngewood Clffs, NJ rentce-Hal, Inc, 1
Sknner, B F Creatng the creatve artst In On t utur of art 1-5ew York Vkng, 10 ernted n Cumulatv rcord: A collcton ofars 3rded, 33344 Englewood Clffs, NJ Prentce-Hal, Inc, 12
Sknner, B F Bondfrdom and dgnt New York Knof, 11
Sknner, B F Cumulatv rcord: slcton ojars Engewood Clfs,N rentceHall, Inc, 15, 11 (enlarged ed), 12 (3rd ed)
knner, B F out avorsm New York Knof, 14
on, L Pscolnustcs Genve, Il Scott, Foresman, 1 1 ooons, L M, and Sten, G Norma motor atomatsm Pscologcal
Rvw 1, 3 42512
Sence, K W Bavor heo ad codong New Haven, Conn YeUnversty Press, 15
Staddon, J. E. R. Asympttic bevir: Th e cnet o f te opernt. Psycho-
logcal Rvw 1�, 4, 371
114
121
Sten, G Teder bo becs food oom ew r Clre e,
uran, J Gerue Sten s 1974, 2, 54-5, 35Tnbergen, Th sudy of nstnct Oxord, Eng Clarenon Press, 161
Warren, H C A hory oft assocaon sycology ew or Srner',
Watson, J Pychology from h sandont oj a behavors lelha Lnott, 119
Watson J B T was ojavorsm ew York Harer & roters, 18
Wenner, A Honeybees o tey use te stance nforaton ontne nter dane anever? Scnc 1, 155, 84-4
Whorf, L Languag tought and ral Slctd wrtngs oj BamL Whor Crge, Mass M Press, 16
INDEX
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 87/89
Absence 316 128Abstration denition 63dages 104Address, definition, 7 2 Advice, 35 36 98 10 Aiken, J. R . , 36Alliteration, 1 12 Alusions, 108Alphabet, 18 Ambiguity definition 1 47Amplification 37mplitude 8Announcements definition 43Antecedent cnditions 7, 22 24Argot 80 81Assertions definition 43 133Assonance 1 12Audi dinition 7 1 73 1Autocitic defnition 1 29Autmatic writing 141-142
Autonomy 79Aversive stimuus definition 28Aversive stimulation 15 9Avoidance 34 91Awareness 15 47Azrin N H. 46 8 2
Babbing 87-89Bvior definition 2B�aior operant 6Biinguaism 78-80Bends definitin 122Bloomfie L. 1 8 66Bousfied W A. 109Brown R 47 82Bruce D J. 1 1 3Bryant M M, 36
Cant 80 81Carrol L 67 1 22Causa account 3 4Causes 3 9Cause-effect reationships 21Centraism definition 1 5 1Chain denition 101Choosing words 77 1 18 126CircumlocutionsClangs 11 2Classes 6 7Classical conditioning 1 2 (see also
spodt ondiionn)
Clihes 1Clustr dntio 5657 101, 107-108Cor C N 1Coiios 2 1 52
Common accompaniments 59Common eements 56Competition among operants 1 36Compiance 27 34 35 36 37Conditioned reinforcer definition 28Conditioned response (CR) 12Conditioned stimuus CS) 7 12 1 1 1 145Connotation 1 46Consummatory behavior 27Contingency 7Contro 3 27 11 7 128�1 29 134 136137
148Conventions 52 54 69
conventiona way 26 41 133conventional covariation 26 41 , 136
Correspondence 50 5 56Counting 104Covert behavior 74 75 76 82 103 126
145 15015 1responses 85 149150
echoics 85 87 92Creativity 11 6 142Cumuative recorders 23Cursing4
Davenport D. 70Deaf mutes 150Defective chains 105 106De Laguna G 78Denotation 43 146Deprivation defnition 7Differentia reinforcement 9 68 82Dimensions of behavior see Intensive
properties of behavior)Direct quotations 234Discriminated audience definition 77
Discrimination defnition 62Dscriminative stimuus dfnition 6Disinterest 41 45 48 50Disposition 21 37 75 129Distortions definition 52 54 122Dophins 70 91Doube entendres 81Duration 56 9
Ebbinhaus H .• 146Echoic defnition 82Electromyography 23 150Elicit 13 54 7 145Emotional behavior 1 7 , 3 9 , 1 45English 18 1 9 ,54,6 6 , 78,79 , 9 5 1
1 1 0- 1 1 1 , 1 1 3 , 14 2Enthusiasm 77Ei S M 79E 9 9
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 88/89
8/10/2019 Primer of Verbal Behavior - Winokur.pdf
http://slidepdf.com/reader/full/primer-of-verbal-behavior-winokurpdf 89/89
top related