EXTINCTION OF CONDITIONED INHIBITION
MITCHEL CRAWFORD MORROW
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
ACKNOi, iU: DnGENTS
The author wishes to express his appreciation to the
supervisory committee, Drs. Kime], Pennypacker, "lhaw, Bunnell
and fiartlett. Dr. Runnell Irracioiisly agreed to join the committee
when the study was in its final stapes, yet it profited by his
interest and helpful suggestions.
Dr. Kimcel, Chairman of the Com:;ittee, has provided help
and encouragement, copent criticism, and considerable tine and
interest to the stidy, just as he hns done throulihut the author's
TARLI: F CCPnI:dT
AC.LiO'i L rI 'iNrS . . . . . . . .. .. .
LIST C IG . . . . . . . . . . iv
LIST OF TAfL.S . . . . . . . . . . . v
I:TP UCTIO . . . . . . . . I
II. HITilOD . . . . . . . . .. . . . 11
III. I'CSIJLTS . . . . . . . . ... . . 1.
IV. I I;CUSSI . . . . . . . . .. 32
V. SUIM'AY . . . . . . . . . . . I.
IDI.I0.';,1'Y . . . . . . . . . . . 4.
PFrnICES. . . . . . . . . . . . . 45
RIO';PAPIIIAL KETI .. . . . . . . . ... 69
LIST orF FIGcUnr
1 Averave magnitude of the unconditioned GE;O Juring
reinforcement and on test trial one 17
2 Mean magnitude of the GSR on the first two CS
Adaptation trials 19
3 Mean magnitude of the CSR on UC! Adaptation trials 20
4 'ean nagnitude of GCNf on the second four C?
Adaptation trials 21
5 Mean magnitude of SR on the last two reinforced
trials and on the first three UCS-alone trials
for Croup Cucs 24
6 tean magnitude of fSR on tho two post extinction
test trials 26
7 The relationship between the UCR Difference Score
and the number of extinction trials 28
8 c:n magnitude of the conditioned G';R on first
three extinction trials 30
I.IST OF TAPI..S
Tabl I P, -e
1 Pean and Standard Deviation for C.iR Differenco Scores
for rroup ;, [',r. L12 and C3; C6, (12 31
2 Analysis of Variance of UCR 'ifferenco 'cnrIes 47
3 Acquisitinn liCk 'mapnitrrds (lon Ac) for Croupi I3 A4
4 Acquir.itinn LUCR Hagnitudes (loi ac) for Group F.C 49
5 Acquisition UCR Magnitudes (log bc) for group p rl2 5sn
6 Acquisition UCR 'lanitudes (log Ac) for Croup C3 51
7 Acquisitlon IUCR Hagniitudes (loI Ac) for rromip C6 52
A Acqiiisition 11CR IHanlttudes (lo Ace) for Group CI2 53
9 Ac.suiisition UCRn 4anitudes (Inr Ac) for Croup Cu 54
10 UCR HManitudes (log Ac) for all 5s on Test Trinl 1 55
11 Response Hapgnitudes (log Ac) for all Ss on First
Two CS Adaptation Trials 56
12 'osponse "agpnltudes (lor Ac) for Crnup [3, 1r, and
1.2 onl UCS Adaptation Trinls 57
13 Iesponse 'lar'nitiues (log Ac) for Croup C3, C(, CI2,
and Cues on IJCS Adaptation Trials 53
14 Response Magnitudes (log &c) for Grou: rF3 and C3 on
Second Four CS Adaptation Trials 59
15 Response Magnitdles (log Ac) for Croupl F, an.l on
Second Four CS Adaptation Trials 60
16 Srsp;onse 'lagnitudos (log Ac) for Group .11 and C,, nn
Second Fnur C' Adaptation Trials 61
17 l:osponse MagnEitudes (lo] Ac) for i"rou;, Cucs nn
Second Fnor Cr Adaptation Trials 62
LIST OF TABLES (Continued)
18 Response Magnitudes (log Ac) for Group CuC on First
Three UCS-alone Trials Pollowing R~einforced Trials 63
19 UCIZ Magnitude (log Ac) Difference Scores Between Iean
of Final Two Reinforced Trials and Mean of First Three
UCS-alone Trials Following Roinforcement for Group Cues 64
20 UCR Magnitude (log Ac) Difference Scores Between Mean
of Final Two Reinforced Trials and First Test Trial
for E and C Crouns 65
21 CR Magnitudes (log Ac) for Group FP, E6, Rnd E12 on
First Three CS Extinction Trials 66
22 CR Magnitudes (log Ac) for Croup C3, C6, and C12 on
First Three CS Fxtinction Trials 67
23 CR .Magnitude (log Ac) Difference Scores Between Yean
of Last Three CS Adaptation Trials and Mean of First
Three CS Extinction Trials for E and C Groups 68
INTRODU(CT I N
During the course of classical conditioning the unconditioned
response (IICP) Crnws gradually smaller. It has beor recognized th.i
rart of this diminution nay ihe duo tn muscular fatirue of tlo respoiid-
ing orcan (Cason, 1937), may involve a cliange in sensory t' -sholdls
(Judd, 195.), ror may hu a relatively permanent reduction (h.ibitu;ationn)
which is distinguished frun tlhe ormer two (Thorpe, 19S(). In addition
to those an associative factor has also ieen assumed to attenuate tihe 'un.
The associative or learned factor was first reported 1,v I'avlov (1927).
die observed that the U1CP suffered nn inhibitory effect in the: presence
of a conditioned stimulus (CO), and a neutral sritulu', after the two
stimuli had been presented together a Punter or times without thc tin-
conditioned stimulus (UCS). Others (lillard, 1'.33; Lufnrt ard Kimhle,
19SS; Spence and Ilunnuist, 19I5), have also reported the nccurrence of
UnR attenuation in the presence of the C9.
It was not until 1961, however, that rimble anJ Cst (1:61) reported
a study specifically Josigned to investigate conditioned diminution of
trh U(CR. Tlhey assumed that part of tle I.Cq
of a conditioned Inhibitory process under the control of the C',. B)
presentir.c the I.CS sometimes with and so-'etimes without the CS, they
found that tlea ntnlitndo of the unconditioned e-ebhlink as, indeed,
attenuated in the presence of the r" during reinforced conditioning
trials, and nlan that the strength of thli conditior.ed inhibitory
process was related to the interstinulus interval in the same way
as is the conditioned excitatory process. Kitmel and Pennypacker
(1962) extended the findings of Kimble and Ost to the classically
conditioned galvanic skin response (G!RS), and, in addition, found
that the strength of the conditioned inhibitory process was a func-
tion of the number of reinforcements, i.e., there was a greater
reduction in the size of the UCRi during acquisition, and a greater
increment in its size on UCS-alone test trials following reinforce-
ment for groups that received a p.roatur number of reinforced trials.
The present study was an attempt to determine whether the con-
ditioned inhibitory process extinguishes (as does the conditioned
excitatory process) as a result of the presentation of CS-only extinc-
tion trials following reinforced trials, and, if so, whether the amount
of extinction is a function of the number of such extinction trials.
Typical extinction of excitatory CRs has been shown to he a function
of the number of extinction trials. The assumption that the conditioned
inhibitory process extinguishes involves, more basically, the assumption
that such a process is present in the first place, so that the exrorieont
was, on the one hand, an attempt to verify the presence of an inhibitory
process under the control of the CS, and, on the other hand, an attempt
to show that the conditioned inhibitory process behaves in extinction
like the conditioned excitatory process. Two groups of subjects ('s)
were given either paired or unpaired CS(liigt)-UCS(shock) acquisition
trials. At the end of acquisition training each of the two groups was
divided into three sub-groups, differentiated by whether they received
either three, six, or 12 CS-only extinction trials. The test deternining
whether extinction of the inhibitory process had occurred required the
presence ct' the IC? (GCC':!), and thus the U(7. "inc. the inhibitory
process is presa mr to be controlled bV the C', its presence too wau
re;ulred on the critical tests. Ience, to provide the a:,iroriate
testing situation, two aired f'-'IC-. trials woro prerse;arc' waiter the
The uruose of the inpairud c ntrol grouJi was to provide a
situation in which tihe sae- number of C.Ss nnJ I'C's were present Lut,
at the same time, one in which an essential con,!ition for leurnirn
(i.e., te;mporal conti-uity of the stimuli) was not present. In
addition, another control Rroup was run. This lIttcr Broatp; received
the same tr3atmrnt Juring the ac-qulition ,hase of thi exeriment as
the paired acluiition -roun, but Instead of receiving CT-only extinc-
tion trials, it received 12 UC'-alone trials follovirn reinrforceent,
thus providing a test for tl: presence of cornJitioned dliir.ution of
tie UCR in t'i3 ciustoaary tnnner, i.e., as an Incrcment in 11CP nonlitude
on the UiC-nalone trials following reinforcement over that of the final
Historical Poview and Related Studies
Returning to "avlov's report of UTC reduction, he attributed the
Jiminntion to the action of Internal inhibition hroupht on by the pair-
ing of a neutral stimulus with a conditioned stimulus, the r.pared
stiiall never being reinforced. Tihe effective CS, when piescn:ed with
the previously neutral stimulus results in a reduction (i.e., in
magnitude) of the conditioned response (CR), "nnd tl-c inhibitory effect
may extend even to the uncnn liticncd refleies themselves" (r. 75).
Pavlnv called the action of the ralred stiEiuli conditionedd inhibition,"
and the previously y neutral stimulus a "conditioned inhibitor."
During the time between Pavlov's reference to the phenomenon and
1961, only anecdotal mention was nade of an nssociative factor produc-
ing UCR diminution. lilpard (1933), in an nyelid conditioning study,
reported that in one ; the UCR to sound was greater when the C! (light)
was absent than when it was present.
Dufort and Kimblo (1958) noted a lar.e increase in UCR magnitude
when the IKCS was presented alone following 20 paired (S-UCS trials.
The unconditioned eyeblink was larger in the former case, i.e., in
the absence of the CS. Aain in 1958, Spence and Runqiuist (1958)
observed that the UCR to an airpuff was smaller (tliou-h not sipnifi-
cntly so) in the presence of a light that had previously been paired
with an electric shock in a forward condition than it was when the
light had been paired with electric shock in a backward paradigrm,
i.e., when the puff was delivered at the time the shock had been
(i.e., 0.5 sec.) but not at 4.5 sec. It appeared that inhibition
may have been conditioned to the CS (lifht) tand tht the inhibition
generalized to other responses.
The last studies to be mentioned in which an associative inhibi-
tion operating on the UCR was apparently demonstrated but in which
it was not being investigated directly were by Calarbos, Sheatz, mnd
Vernier (1956), and ilrnandez-Peon (1955). These workers, after
implantinr electrodes in either the acoustic pathway (Galanibos,
et al.) or the visual cortex (iernandoz-Peon) of cats, noted a reduc-
tion in the electrical response to a tone, and to a lipht, respectively,
in these areas after repeated presentations of these stimuli. Both
researchers presented the stimuli at fairly constant temporal intervals.
They noted that when novel stimuli were presented the response returned to
near its original strength, and thus in their uords "becomes
delialituated." Cavaiglnni, Clanelli, .and ail ntib'no: (1959),
confirredl the results or lHernand :-Pnon. Galna bs anil 'ernanlez-
Peon ioth ascribed the results to an inhihitory nechanism acting
by way of the hrai ,-stem reticilar fnrnation.
As previously noted, the first study specifically Cesirned and
executed to investi2are conditioned diminution ir the urR has done
ly Kimble and Ust (1961.). This Inquiry attempted, first, to Jomoin-
strate the phenomenon rmpirically and, second, to slinh that tle
amount of recovery of the UCR .lhen the C9 was laittod was a func-
tion of the training interstimlluis interval, i.e., tie interval
between the onsets of the CS und UC., Four groups of Ss were given
50 paired presentations of the CS (light) and UCS (air puff) at
intervals of either 25n, 503, 1000, or 2000 msec. After the con-
ditionin, trials each group was given ton 11C;-alone trials. A
comparison of the average UCR nmanitude on the last five condition-
ing trials on which no CR occurred with the averaco UCl imgnituuio
on the first five IJC'-alone trials showed that therr was a signifi-
cant overall increase in average m.anitude of the I1CR on the UC2;-nlone
tri.ils. However, the increase did not favor cne CS-llIC. intervul over
another, tlout.h the trend or the differences was in the expected
direction. In a sub-experi.ent usinr only .s who showed the greatest
UC;R dimiiintion at any interval, cs were Inter presented 40 paired
acquisition trials, ten at cach of the fiour CS-I('W intervals used
in the main experiment. The order of the trials wns random with
the restriction that each interstinuluc interval had to appear twice
i:n each llock of eight trials. The test trials, UCS-alone, were
presented in the sat(e wny as in the main experieent. C'tatlstically
reliable differences within Ss were obtained in this part of the
experiment with the 500 msec. interval producing the greatest
recov ry and the other conditioning intervals falling off in onount
of UCR recovery on each side of this value.
The authors concluded that they had demonstrated an inhibitory
process in addition to the one comonly called adaptation or habitua-
tion, this loarneo inhibition bein' under the control of the CS. Their
basis for this reasonin- was that when the FC was omitted there was a
significant increase in UCR amplitude. Since there was, in addition,
evidence that the amount of inhibition was a function of the inter-
stimulus interval, and that the function was similar to t!.e function
obtained with other excitatory responses (k'fible, 1947; '!cAlllster,
1Q53), they concluded that the inhibition has learned, and thus were
"inclined to call this a conditioned, or associative, irhibition"
Other possible explanations of the inhibition were considered by
Kimble and Ost: (1) The likelihood that the inhibition was due to
the automatic interaction of reflex rechanisms was rejected because
only a!out one-half of the Ss showed the inhibitory phb'tnomenon, and
one would expect reflex level phenomien to be more tubiquilous, i.e.,
occur in the vast majority of the Ss; (2) The fact that there was a
graJual diminution of the UCR over trials militated a-ainst the con-
tention that the 's were voluntarily rosistin; heirn itade to blink,
for if this were so, one would expect a zore abrupt reduction of the
response; (3) That the inhibition was due to personality variables
was a possibility, but this could rnot be assessed; (4) Finally, the
possibility was entertained that the inhibition resembled Pavlov's
(1927) inhibition of dolay, a forn of internal inhiitinli, whirh
operates to delay the occurrence of the Ck until just before the
occurrence of the IJC. If the last possibility were the case one
would have to assume that the inhibition of delay was operative at
least until t1e p'resi.ntation of the UCS and that it inhibited the
response to it. IL should be noted that this last altornattivo is
itself a learned source of Inhibition and, beinp so, is not in con-
flict with their basic contention.
Kimiel aud rPonyp.icker (1962) Iperformed ;;n experi-.ent desirnicd
to extend the finlinngs of Kinble and Ost to the classically cundltioned
CSHI and to investigate the relationship between the niaher of rein-
forcements anJ the amount of diminution and recovery of the IMLH.
In tl.eir study tree prouns of F-s received either four, eirht, or
16 paired acquisition trials. A 1,000-cps tone was the CS and a 2.(0
ma. electric sl oc was thio iUC3. A delayed conditioning paradig~n as
era;loyed, with the C. Inlsting 1.0 sec. and the UI's 0.1 ser. The two
stimuli terminated together. Fnllowiin the conditionl!ip trials two
UC'-slone trials were presente.
Comparison of the UCR values on the last tun reinforced trials
anld the two IUC-alone trials shro'ed that there was a significant
overall increment of UCR apilitude on the two Ihfr-alrne trials, and
that the number of reinforcements interacted with this increment
significantly, i.e., there was a preater difference with a greater
number of reinforced trials.
Baxter (1964) classically conditioned the GC', usinp both the
trace and delayed conditioning paradi ms with different grou;.s. In
addition, an unpaired control group was employed as in the present
experiment. !o found that the conditioning procedure led to
significantly greater diminution of the UCR than did the pseudo-
conditioning procedure. Comparisons of the UCRs o' the trace and
delayed groups showed that the delayed paradigm led to greater
reduction than the trace conditioning.
In research similar to that beinp presently reported, Grinjs
(1960) and his associates (e.g., Kiamel, 1960) have investigated
what they call a "perceptual disparity" variable operating on the
amplitude of autonomic CHs and UCRs. Cringes suggests that besides
the excitatory CR, which is learned in conditioning experiments,
there may he another CR, a preparatory set for recent of the UCS.
This preparatory set may result in a depression of the amplitude
of the anticipatory excitatory CR and of the UCR. The perceptual
set which develops during conditioning aay be analogous to the
development of an expectancy, an expectancy s-ecifically of the
presentation of the UCS. When the UCS is not presented or when a
new, i.e., novel, stimulus is presented instead of the customary
UCS, it results in a disconfirmation of the expectancy and a
perceptual disparity response, an increase in CR or UCR amplitude
over that which occurs when the UC: is presented.
Assuming that a CS-controlled associative or conditioned inhibi-
tory process, which operates to reduce the strength of the UCR during
reinforced trials, has been demonstrated in previous studies, it was
of interest to further explore the process to determine if it behaves
under similar conditions like the conditioned excitatory process.
In particular, since it has heen demonstrated that the inhibitory
process is influenced liy the interstimulus interval variable and by
the number of reinforcements variable in the same way that the conJi-
tioned excitatory process is affected, it was of interest to determine
whether the conditioned inhibitory process would extinguish upon the
presentation of CS-alone trials. The following s;'eciflc hypotthescs
1. Paired CS-'ICS reinforced trials will lead to a conditioned
inhibitory process which results, when the CS is present, in a
reduction in size of the !ICR.
2. C: extinction trials following reinforcement will produce
extinction of the conditioned inhibitory process.
3. A greater number of C' extinction trials will result in
greater extinction of the conditioned inhibitory process.
4. lCR recovery follcwinp C" extinction trials will occur at a
more rapid rate for groups that received paired acquisition trials
than for groups that received unpaired acquisition trials.
Hypothesis No. I simply asserts that an inhibitory process will
be conditionrJ to the CS, and lIIpotheses los. 2 and 3 state that the
inhibitory process will extinguish, and that the amount of extinction
will be a function of the number of extinction trials, rec:,ectively.
IhTpothesis :o. 4 Ipredicts that there will hl an interaction Iectween
type of training, (paired or unpaired) and numhrr of extir.ction trials
(three, six, or 12). This prediction was based on the assumption that the
paired reinforcement groups will have, early in extinction, two factors
serving to reduce the amplitidle of the UCR on the test trials. The
first factor is the in!iliitinn resulting fro hbnhituntion or adaptation,
and the second factor is the inhibition resulting' front the conditioned
inhibitory process SBt the unpaired reinforcement groups will be
influenced by only the first of thcse factors. Later in extinction,
it is expected that each of those factors, habituation or adaptation,
on the one hand, and conditioned inhibition, on the other, will both
have been dissipated.
One hundred and five University of I'lorlrn uLdergradulates serveJ
as volunteer Is. The tcital sample consisted of 63 men and 42 woven.
There were nine men and six wowen in each grouzi, and each 5 recoivred
one dollar for participating.1 The Ss were assigned randnlsly to the
various experimental conditions with the exception of sex lalancing as
A two X three factorial design was euprloned in the main part of
the study. The factor with two treatment catreorics was typc of
training; the treatment consisted it whether t:e CS ani. 'JC'; ere
presented paired or unpaired, during reinforce.*riit. The other factor
was the numt.or of extinction trials; the inalivildual treatments were
either three, six, or 1' of such trials. Thus six Jifferent treatrent-
coahinations were generated From such a design; three Croulps (E3, L(,
r.12) lich received paired acquisition trials ~ero differentiateI hvb
receiving either tree, six or 12 extinction trials before tne two
aired test trials, and three rrnuns (C3, r., C]2) which received
unpaired ncauisitinn trials were also diffrer.tiatre in extinction
l s were pnid out of Nil' rrant ('4-00.t0-2) to Dr. it. I. Kimel,
University of Florida.
by whether they received three, six or 12 extinction trials before the
paired test trials.
In addition to the groups discussed above, another group (Cucs)
was run. This group was treated identically to the paired CS-UCS
groups during acquisition. However, instead of receiving 12 extinction
trials with the CS only, it received 12 UCS-alone trials, and then the
two paired test trials. The purpose of running this group was to
provide for the demonstration of the conditioned inhibitory process
in the same way as it has previously been demonstrated, i.e., as an
increment in UCR amplitude occasioned by the omission of the CS.
After ho arrived at the laboratory, the S's left hand was cleaned
with acetone and then he was seated in the ventilated, dirmly lighted
experimental chamber. Next, the shock electrodes were attached to
the fingertips of his right hand and the GS1 electrodes were attached
to the palm and back of his left hand. He was asked if he was com-
fortable, and he was told that the door was to be closed but that he
could speak to the exnerirenter (F.) at any time through the intercom
system. Indicating that the instructions would be read over the
intercom system, E. pointed out the projector which presented the CS,
left the chamber and closed the door. Approximately two minutes later
the instructions were read. They indicated that the experiment dealt
with the effects of certain stimuli on the galvanic skin response,
and that the S should pay attention to these stimuli.2
2The complete instructions appear in Appendix A.
The ; was fir5t given thn presontatlins of the CS alone, followed
by three presentationss of the 0ICS .elono, in intensity increment i;' to
the intensity used during conditlnnin-. This was followed by four
additional presertations of the C' alone. duringg the conditioning
period, which occurred next, three grcups of '; (r3, r., i-12) received
16 paired CT-UCf presentations, and three Froups of Ss (Ci, C6, C12)
receiveJ 16 urnpaireJ C'-lC' l resentatlnns. After the conditioning
trials, ?roup;s r3 and C3 received three CS-only trials, .Lrou:s 1:6 and
C, receiveJ six CS-only trials, an.d groups C., and C12 received 12 Cf.-
only trials. Group rucs received, after 16 paired CS-iC'S presentations,
12 '.CS-alona trials. Followiin tiese trials all rroups received two
paired CS-UC'; test trials.
The 's were seated in an iiidustrial Acoustics Co., Inc. AiJdu-
metric Chamber, Modul No. 1200. The chair in which they sat was a
standard office chair with arrests on which the fs' kept their arms
during the experimental session. The r. could communicate with the
Ss at all times through a two-way Intercom system.
The CS was a circular, white light placed at eye level 30 Icches
directly in front of the S's face. The light wa. produceJ by .
Irason-Stadler Co. '1ttltt:le Stimulus Projector (Type r458G). The
light had a Jiameter of one Inch.
The UICS was an electric shncl delivered throuhli silver-coated
electrodes attached to the middle and index fingertips of the r's
rirlit hand, The shock was produced by a Schultz Instruiments Co.
Constant Current Shock Gienerator. The shock intensity used during
conditioning was 1.5 ma.
The durations of the CS and UCq were controlled electronically by
a Tektronix, Inc. Wavoform Generator, Type 162, and Pulse Cenerator,
Type 161. The CS had a duration of 5.1 sec. and the IFCS a duration
of 0.1 sec. On paired conditioning trials the two stimull terminated
together. The interval soparatinq trials was varied unsystematically
by E within 30-60 sec. Both stimuli were presented within a 30 sec.
interval for the unpaired control groups. These 30 sac. intervals
were separated from each other in the saie way that paired conditioning
trials were separated, i.e., varied unsystematically between 30-60 sec.
On half of the trials the CS occurred first and on half of the trials
the UCS occurred first, except that in no case did more than two
consecutive CSs or UCSs occur. In addition, the next to the last and
the last CS preceded the UCS.3
The CSR was picked up from the palm and back of the S's left hand
by 3/4 in. zinc electrodes, covered with a few drops of zinc sulphate
solution, in lucite cups filled with saline electrode jelly. It was
amplified by a Dio-Physical Research Instruments, Inc. CSR Amplifier,
Model 201-R, and recorded on a Texas Instruments Co. Recti-Riter with
a paper speed of 12 in/ain. All responses were transformed according
to the formula:
log -~ 109 + 1.0
3The CS was presented first on the last two acquisition trials
because the dependent variable, as noted in the text, was computed
as a Difference Score between those trials and the first test trial.
Following this procedure, the testing conditions were identical for
the paired and unpaired acquisition groups, except for the temporal
separation of the stimuli in the latter group.
where Ph is the resistance at the moment of initlntion of tl.o response
anid i is the resistance at the peak of the response. This log
cniiductancc-chanre unit has leen shown to be the most appropriate
measure for transforming CS Il ata (Iacey annJ Teell, 1949; !a~aJrd,
Diminution of the UCR
Since support of the major hypotheses of the experiment pre-
supposed the presence of conditioned inhibition of tho UCR, it was
first necessary to establish that inhibition of the UCR occurred and
that part of this inhibition was due to a conditioning factor.
The UCR was measured as the maximum decrease in resistance which
occurred between two and five seconds after the offset of the UCS,
and was transformed as previously noted.
Fig. 1 shows the mean magnitude of the [UCR for the combined i
(N=45), combined C (?N45), and Cues (:=15) grouis as a function of
reinforced trials.l Also indicated in Fig. I are the mean UCR wagni-
tudes of the V, C and Cucs groups on the first post extinction test
trial. Discussion of the Ccs group will be presented below.
The figure shows that the mean UCRs of both the F and C groups
diminished as trials increased, that they diminished at about the
same rate, and that the UCRs of the F groups were not smaller than
the UCRs of the C groups except on trials 11 through 14. Since it
was expected that the UCRs of the E groups would show greater diminu-
tion during acquisition than the C groups due to the combined effects
IThe data from which this and all other Firs. were plotted
appear in Appendix B.
a a 0
o o _)
0 9 < I? ',< o
i I \ u -
( > z> -
, I I. -
<0 fl 0 0 L 0
0 0mm. a N
I o L 1 I I i I i n
0 0 cn cn o i -.
(3v E0O-) 3anl]N NvlAI a13n rItIV3
of habituation of the response and a conditioned inhibitory factor,
but that the UCRs of the C groups would diminish only because of
habituation, it may be that there already were proup differences
at the start of the conditioning trials, and that the F groups were
initially more responsive than were the C groups. If this were so,
it would be indicated by the respective groups' responses to the iCS
and CS adaptation trials, prior to conditioning. Figs. 2, 3, and 4,
showing the average responses to the CS and UCS before conditioning,
support the contention that the groups made larrer initial responses
to the UCS and CS than did the C groups. Again, the discussion of the
Cues group, whose data also appear in these fii're,., will be delayed.
In Fig. 2 it can be seen that the E groups produced a larger mean
GSR to the first two adaptation CSs than the C groups, on both presen-
tations of the CS. The overall difference between these two groups was
highly significant (t = 3.54 with 88 df, n <.001). Fig. 3 shows that
the E groups made greater moan GSRs to all three initial UCSs than did
the C groups. The figure slows also that the mean GCR increased over
trials. It will be recalled that the intensity of the UCS was increased
on each trial so that on the final UCS trial the intensity was equal to
that used during reinforced trials. The increasing' intensity of the
UCS during these trials presumably accounts for the fact that the
response became, on the average, larger instead of smaller as it did
on the two sets of CS adaptation trials (with a constant CS). The
overall difference between the mean GSI; magnitudes of the E and C
groups on these trials was also statistically simrficant (t 2.62
with 88 df, p <.02). fig. 4 indicates that the mean CSR to the second
four CS adaptation trials was larger for the E1 groups than it was for
I I I I I I I
O l O inO 0n 0 0 0
r- r-. (D in 1 T
(07 901) 3anilN9VLAJ 8SO NV31N
o_ a- D
\ I I Q
\ \ \\ \
I I I I
0\ 0 0 |000
(OV 9O0) 3GnlIlNgV t$9 NV3h
I I I I I I
0 0 0 0 0 0 0 'n 0
r-. (.0 InO It rn
(0-7 90) 30fiiN9O3VN S9O NV3VA
the C groups although th1e groups were virtually equal on the last
trial. The overall difference between the means of these two
groups also approached significance (t = 1.92 wit! R8 df, p < .10).
In addition to these data, the acquisition performance of the
Cucs group also sig rests that conditioned diminution of the UC!i occurred
because of pairing of the CS and UCS. It will be recalled t!at this
group was treated Identically to the rF roups but, instead of receiv-
ing; CS extinction trials, it received 12 UCS-alone presentations during
the "extinction" phase of the experiment. As rigs. 2, 3, and 4 show, the
Cucs gronu had a Sean GSCR nagnitude intermediate to that of the ? and C
groups on the initial CS and UCS adaptation trials, but showed, us is
indicated in Fig. I, greater atte.uation of the UCPR Jurlng acquisition
than either the I or C groups. In Fig. 1, showing the mean UCRs of the
Cucs grrup on the 16 reinforced trials, there was a rapid drop in magni-
tude on the third trial, after a slight initial increase on the second
trial. Following this there was a slight, more pradual reduction as
trials continued, with a rise toward the end of the trials. These data
further support the contention that the i groups fallen to show a smaller
UCR on the reinforced trials because of their increased responsiveness
at the beginninR of the experiment, i.e., snnplinG differences. It
supports the contention because in this one instance where the hyper-
responsiveness was absent greater UCR diminution was demonstrated.
The Cucs group also permits the Jeionstration of conditioned
inhibition as a difference (i.e., incrmennt) in UCR mragnitude between
the nean UCP on the final two reinforced trials and the mean UCR on tlo
first three LCS-alone trials ia~ediately thereafter. In general, this
is the way in which it was demonstrated in the three previous studies
mentioned love (Ei.ihle aun Ost, 1961; mnnmel ani PennypacLer, 1962;
Baxter, 1964). difference Scores "were compuuted for the rC., Criup
between t:e mean ICR on the f'iral two reinforced trials and the first
three liS-alnne trials. Fig. 5 depicts the differences hetwe.n the
mean UCT's at these two points. It shors that ttcre uns an increase
in the UCR on the first three I'CS-ulone trials over the LICI: rn the
final two reinforced trials. The TeAil of theo "ifferenco cvres was
.155, with the UICR showing a greater value on tle IUC;-alone trials
for all but four Ss. A t test indicateJ t!ht this increase in tihe
IUCi on tile UlCS-alon trials was hiphly significant (t 3.15 with 14
df, p <.OOS). It isouild bh noted spain that the Differenco Scores
cormputed Ifr this rroup were between the final two reinforced trials
and first three iiCS-alone trials. Since the Iroan UCR on the last t-o
reinforced trials happened to he larger than the ICils on any uneiihl:lr-
iny trials after the third, as shnwn in Fir. I, it lend; even reater
confidence in the signifiiunce level referred to above, and, thus, to
its impilicati nns.
rxtinrction of liCR Inhibition
Since th' above data were taken to support the pr.nsu.positinn that
a condiitioned inhibitory phenomenon occurrcd in the prnoups that received
pairdJ Cs-ics trials hilt not in tlose Prouns that received unpaire.i
trials, it was of interest nrxt to test t!:e hypothesis that prrsenta-
tion of the r(; without the LIUC (i.e., an extirction pirnc'LIOre) r.t;ulJ
result in extinction of the inhibition. To evaluate this h.potlheir.,
another l:CR Difference core was crnployel as the dolpender't tariahle.
It was cot;uted by subtractinl th ean ICr rn the fin-.l two reinforced
trials frno the UCIh on the first post-uxtinction p;ired. test trial for
\ 0 0
\ .d 041
I ,- O
I- c" N
(OV 901) 3anlNsv~J aS9 Nv]iAJ
esch S. Since ;, carefl cop;'arison of thle t11:Rs n the tuo test trials
indicated that relearninZ hnd occurred eniote rapidly, because the lC;t
was considerahlv smaller on the second of the two post-extinction
trials, especially on the part of the r ground, it was decideJ to use
only the response on the first test trial in computinr the differencee
Score described above. Fir. 6 shows the nean UCR magnitudes for tih
E nnd for the C groups as well as for the Cucs ~ group on the two ;ost-
extinction test trials. Ai is indicated, there was a considerable
drop in moan 11CR nagnitudo in the I groups on the second test trial
hut not in the other two grous. Incidentallv, this finding also is
in conformity with the contention th.t a conditioned inhibition occurred
in the r. groups and not in the C groups. (It does not relate to another
such an inriibitioln occurred in the Cues group I.ccause this group dild
not receive CS extincticn trials, did riot extinguish the in!iibitlin,
and thus could not releani the inihibitic.n.)
;t can be seen in riEs. I and 6 that the P group i prodJcedJ a larger
mean UCR on the first post-extinction test trial than did either the C
or the C,,, groups. Indeed, the size of tle average UCR of the I"
groups was comparable to that on the very first reinforced trial, whi;l
the C and Ccs groups showed little If any increment following extinction.
On the contrary, the C1cs grOLIp shoe-d a lower mear II'CR on the post-
extinction test trial than it did on the final two reinforced trials.
The analysis of variance of the LCR Difference Scores indicated that
the greater overall increment in IJCR ma-nitude on theo ost-extiniction
test trial by the E groups in com.parrion to the C groiius was statis-
tically significant (p <.05). Table 1, in ArpenJix II, presents a
suamarv of the Analysis of Variance which evaluated these data. Tho
0 ] ^0
W7 CO -
S I I I I I I
o o. I 0
(3V 901) 30NllN9VN ?1S9 NV:IIJ
greater increment Ib the r groups when comp.lre,1 to the C groa1lps was
nttributahle solely to the types of training received by the different
prou;'s because they received identical trentment in all other phases
of the experiment.
'umber of extinction Trials
it was of interest next to evaluate the hypothesis that t:.e amount
of extinction of conditioned UCP inhlibition wouldd he a positive function
of the nunlier of extinction trials. Fie. 7 shows the ealn lCR Difference
Score for the I groups that received either three, six, er 12 CS extinc-
tion trials, amid for the C groups that received either three, six, or 12
C, extinction trials. Ne-ative values on the ordin.te indicate that thu
average UCR on the test trial was lower than on the flial two reinforced
trials, and positive values indlic.te that the average UCPC was higher on
the test trial. It appears from tle figure that extinction of the
inhibition was essentially complete hy the their. extinction trial and
that further presentation of extinction trials had little, if any,
effect on the nvorage magnitude of the I:CR on the test trial. Con-
trary to the hypothesis, a -reater nueher or extinction trials did
not lead to greater extir.ctino7 of the conditioned (Ci' inhibition .as
shown by the fact that the F ratio t'fr tie effect of pusher of cx-
tinction trinis lailcd to reach' a significant value. A gin, Table I
in Ap1pendix B surazrizes the Analysis of Vorinnce of t..ese data.
The fourth major experimental hyipothcsis was that tire would be
an interaction between typL of truiring (whetl.er paired or unpaired
CS-I:tC prosontations) and nuaor of C; rxtincticr. trials (whether
threr, six, or 12). This interaction was iredicteI because ir was
N0 - o -
s-)m Q ) *
0 LO O 3a uI_ 0 -I 0 0I
0 0 0
expected that the r gprnups WoulJ lose the Inhibition as well bs the
habituation acquired turint reinforcdaent, but tl-t the C rrou:-s,
since they 'ad not been condJtioned to inhibit the UCk, uoiild lose
only wlint habituation hadt been acquired. Pef'rrin; to lig. 7 at-in,
it can be seen th:.t except for Croup C3 the curves; of tl'e liirferences
Scores are essentially parallel. '.y the C3 Croi-i should. show a nopa-
tivn moan UClH Difference Score value (i.e., why it should produce
smaller average IC1s on the post-extinction test trial) is ont readily
ap;arcnt. An irisportion of the res-onses of the individual ';s in this
frniup Indicatedr that fiir S; were primarily resronsihle for the loier
favcrace 11CR Dirference ccore. Three of the r'for 's did not repo.nd
on the test trial after producing relatively lnrre mean L.CPII on
tho final two reinforced trinls. Analysis of lr risnce r' the Piffer-
enco scores for t:'o interaction effect did not reach szinificarce,
and thus the hypothesis was not surorted. P-efer to T.able I in
Appendix !I for the Analysis of Variance summary.
rxritatnry (I') ConlJitionin"
It was :lso of interest to examine t!o relationship I etwoen tile
type of training nnl C."R nignitudo (ir rcsi-nse to the CS) Jurirv CS
extinction trials. 1his is, of course, the iusial excitatory (CI:)
measure of conditioning, with which nost classical condirtining
stuJies :,re ccncernel. The CR wa.'s easiired -s the emximur Jecrease
in resistance which occurred between three nnd seven seconds after
the onset of tLe C ', nn;d wns transforne,' as previousl' noted.
Pit;. S iindic.tes tile avernge GCin (CR) mannituJe for the colrbied
L and cnlbined C groups on each of the first three C, extinction trial%.
It Is clear that tle F groiuip prn!iiced greater avcrapao Cr's on each of
I In I
tho three L.tinctlior triuls rlthn the C groui;'s. It is .-lsn cleir that
little, if" nny, CH extinction occurreJ on the three trl.1s. CR
Liffercncoe corcs for each S were co.'iutei by ihtracting t!:e lverapF
oaTritude nF Cr,'!' on the anst thicu CiS-nlono presentations -nrior to
crnditioninu from t-e- average na2niltude f r'.r! on the three extinction
trials. TIie iean of tlhe C-' differencee Sc'ors lindrctc.I that the ivecr:re
C-IR of the cE groi;-s increased fron before to after the r'einforcolnt
period, nn.l the avcrage C'R of the C groups 1ecreaocd durir,. the s:\ac
period. TIe difference botr~een the rieans of the fDfference carries was
statistically siriificant Ct = 2.25 with A. 9 O <.-05), .urrnrtinii
thel conclusion, tlhat conventional cxcitntory (1r) conditioning nccurrer.
T.bhle I below shows the reans and stnnardJ dviationr, no tce iffercence
scores for each of tlh I1 and C rroups.
Ieanr anJ Srtandard Deviation for CR Difference Scores for Croiij L.3, L.
El, and C3, C6 C12
rS 1(,17 C; C., C1
Mean .186 .045 -.Old -. 12< -.055 -.147
.D .290 .033 .27. .120 .34 .176
Evidence of a Conditioned Inhibitory Process
The results of this study support the hypothesis that an inhibitory
process under the control of the C; developed during reinforced trials,
and that the inhibitory process reduced the magnitude of the UCR.
During reinforced trials the UCR diminished gradually, and when the
UCS was presented alone following reinforcement the size of the IICR
recovered immediately to near its pre-acquisition value. The recovery
of the ICR in this manner was demonstrated with Ss group p Cucs) who re-
ceived paired CS-UCS acquisition trials followed by a series of IJCS-
Other evidence that a conditioned inhibitory process was present
was sunplied by Ss who received paired acquisition trials hut were not
tested for the conditioned inhibitory process in the usual manner.
These Ss showed a greater relative reduction in their CSR1s from, on
the one hand, the CS and UCS adaptation trials to, on the other, the
final few reinforced trials than did groups that received unpaired
acquisition trials. It is assumed, regarding this greater relative
drop by the paired acquisition groups, that the conditioned inhibitory
process acquired during reinforced trials was not great enough to over-
come the initial group differences. As a result, smaller UCRs were not
produced on the reinforced trials. Supporting this argument was the
f.ct that the UC.s of Cropu'- C a a rnoup whons CSlis to the adaptation
trials were Intermediate to those of the F and C groups, were sallor
th.-.n the I!CRs of the C ;roups on every reinforced trial after the second
except two, the eighth and sixteenth. Aii.d, finally, although! the pres-
ence of the usual cnnditioneJ excitatory process did not cup 'ort the
notion that a ccIlditiored iiihibitory princess 6as present in the paired
acquisition groups, it was comfortin to the argument exprcssedr that
excitatory contlitionins was "resent in these groups, and not in the un-
paired acquisition prou.s, since its aesencr weight very well be taken as
harmful to the argument.
The data presented in this .*xpurlaent nre iin accord with the finJ-
inr.s of Kimllea and Ost (1961), Kimel andl Pennypacker (1962), and
Baxter (1064), nnd, toethicr with these stuJies, represent -nother
study in which a conditioned nliihitory process under the control of the
C; has been deimnstrated.
However, the results of these experiments might 1ie viewed Iy some
as showing the effects of "surprise" on the ma.nitiuJe of the UCH. The
argument that the increase in !UCR streoisth on I'CS-alone trials folioan.g
reinforced trinls was the result of the q's surprise at the omission of
the CS is dlfficllt to answer hecnusa it has not boen hyponthesized
unamabiuously Just uhat the effects of a siirrise factor or process are
on Iaohavior in general, let alone the unconditioned CrR in particular.
Green (l95A), however, suvgests th;it surprises" 'iay lie defined in at
least tho uwys. First, it is "roposod only that the term be "defined anJ
mtasorcd in rarts of autenonic respfrTisr's such as the ('R" (p. 340).
Alternatively, it may be related to tihe foresecl.bility of al event
accordin- to a predictionn ade onn an inductivo hasis." (p. 3.10). For
example, a verbal itea preceded by a lon" series of numerical itoas has
more surprise value than the same item preceded by a shorter series.
Green hypothesized that the surprise engendered by such a procedure more
adequately explains the so-called Von "estorff effect. A search of the
literature failed to turn up other suggestcins regarding the effects of
surprise on behavior. However, Kieal and Pennypacker (1962), antici-
pating the "surprise" argument presented above, noted that even though
the S might very well be surprised at the omission of the CS, the
"reduction-of-surprise function" eccruinp to the CS during reinforced
trials is itself an inhibitory function.
In addition, it should be pointed out that in the present study the
unpaired acquisition groups should have, following the second definition
proposed by Green, produced! greater rSRs on the test trial than the
paired acquisition groups since the "foreseeability" of the UCS following
the CS on the test trial had not been increased by the training procedure
as it had for the paired acquisition groups. And because the unpaired
groups could not have foreseen the occurrence of the UCS so presented,
they should have been more surprised and thus should have produced great-
er UCls, assuming that surprise leads to larger autonomic responses such
as the GSR. The data, as has been pointed out, did not confirm this
1 xtinctiCnh or th; Con, it-l ri. d Ilohibitory I'roc.s
The results of the experiment indicate that the paired acquisition
groups produced a significantly greater overall increase in UCR on the
paired test trial following the CS-only extinction trials than did the
unpaired acquisition groups. This finding confirms the hypothesis that
the conditioned inhibitory process extinguishes following unreinforced
presentation of the C'. I'ollowin, the extinction trials the paired
acquisition groups produced UCRs on the test trial similar in magnitiude
to those produced on tlie initial acquisition trials, indicating thit the
inhibitory effects of both the conditioning process nind those of the
adaptation process had teen lost during the CS-onlv presentations.
Tic (ICs of the unpaired acquisition r.roujns on the test trial were,
overall, considerably smaller titan those they produced on the Initial
acquisition trials. Indeed the value of their test trial UCRs uas
strikinrly similar to the average value of the final few acquisition
trials. lTis finding was somewhat unoxnacted in vieu of the fact that
thel adaptation to the lIC.; aciqired during the acquisition trials was
exl-ectoJ to dissipate durin= the C.-only trials. Fig. 7 indicates that
the greatest amount of UCR recovery occurred in Cron'ps C6 and C12, and
Ground C3 resrondCd, on the average, wit:i smaller UCIR on the test trial
thin it did oni the last toh acquisition trials. As mentioned previously,
the reduction of the UCR on tlhe post extinction test trial by Croup C3
war primarily due to the response of four Ss. three or wihom failed to
Iffect of Niumber of rxtinction Trials
Tl.cnuih t!ere uas a p.reater overall recovery of the IfCP on the test
trial ly the promis tliat received paired acquisition trials over that nF
the groups tlh.t received unpaired acquisition trials, the data indicate
that a greater number of extinction trials did not ie:a to .a greter
amount of recovery. Stated in another %ay, the hypothesis ed positive
relationship Ihetween nu.L-,er of extinction trials anA extinction of tile
conditioned inhibitory [rocmess was not confirm udl by the data. Ns Fi... 7
demonstrates, extinction of the conditioned inhibitory process, as
indicated in part by amount of recovery of the P'CR on the test trial,
was virtually complete after the third extinction trial. Further
extinction trials led to only slightly greaterr recovery on the test
It may be that the course of extinction of the conditioned
inhibitory process does not nrogress in the same fashion as does
extinction of the conditioned excitatory process. That is, perhaps
the relationship depicted in Fig. 7 represents the "true" relationship
governing these variables. Reasoning analogically from CR extinction
curves is not especially helpful since so many different curves have
begun observed with that response process. For example, the most
cowonly found curve of extinction of Ci's is essentially negatively
accelerated (Kleitman and Crisler, 1927; liilyard, 1933; Ilovland, 1937;
and others). Other curves have been observed P(ilgard and Marquis,
1935; liovland, 1936), hut perhaTs rpost relevant with regard to the
present experiment is the rapid initial drop in the extinction curve
of CRs reported by Razran (1956), and Spence, (1963). In this latter
curve there is a further gradual diminution of the CR, following the
abrupt drop, very similar to that in the present study. The abrupt
initial drop has been explained as due to generalization decre~snt,
discrimination decrement, automatic deconditioning, loss of motivation
due to the omission of the UCS, and to an inhibitory set resulting from
recognition by the S that the conditions of the experiment have changed
from those in acquisition to those in extinction. It may he that one
of these factors, or a combination of them, in conjunction with a factor
to be discussed below, was responsible for the rapid extinction of the
conditioned inhibitory process indicated by the data. Incidentally,
the factors contribution' to the rdevelo-,'cnt of thr Inhibitory set
could have been respr.isible, in = similar way, for the ra,,id relearning
of the conditioneJ inhibitory process on the, paired test trials follo.-
In extinction by the r. proupps, oxciept that in this ra-c the sot would
have to be virewd as facilitative. 'pcncc, Ilonzir, and RutledIle (1964)
hive sho.,n that when the de-ree of discrininnhility of the procedural
changes that occur -ith shift froi acii isition to extinction is reduced,
thl initial drop in CR strength on the first few extinction trials largely
disappearnr. That facilitative or In'iihitory sets 'ave an effect on a.tuint
of con:ditonnin. has also been dcnonstrateJ in stuJics that have varied
instruction tn '. prior to cnnlitinnin- (':orris an.i Gr:int, 1349; Nicholls
and Fi ble, Q164).
rIelevance to Pavlovian and I illian Theory
An important consiilration in unJrrstniilin2 the extinction of tho
inhibitory process in the present experilmont is the wJv in which it
may interact with tlihe growing exttnctiv- inhibit in resulting from the
CS-only trials, Accrdinp to Pavlov (1327), not only do C'-alono
trials result in the development of internal inlhiition, hut reinforced
trials do also. Indeed, Pavlov viewed the function of reinforcement
after the establishment of the excitatory process as serving principally
to retard or delay the development of Inhibition. Various forms of
inhibition, besides the one reported in the present experiment, obrerveJ
during reinforced trials, include Inhilition of eelay (Pavlov, 1927) and
inhibition of reinforcement (Hlovland, 193f).
Pavlov viewed extinctive inhibition, conditioned in'ihition (a form
of inhibition paradiftatcnally different fro- the subject matter of this
experlimnt), arnd inhibition of delay as being all instances of internal
inhiition. Internal inhibition was a Dore thstract notion, bein-
conceptualized as one of two major cortical processes (the other
was cortical excitation) presumed to mediate overt behavior.
Different C's were all viewed as being instances of the unitary cortical
excitation. Though little work has been directed toward the problem,
it is well known (Kimble, 1961) that other responses than tFe one
recorded and reported by the experimenter are conditioned during rein-
forced trials, and that these other responses may interfere with or
facilitate the response observed.
It is clear that the method used in this experiment, i.e., CS-
only trials, to overcome the conditicneJ inhibitory process was not
the method reconnonded by Pavlov to reduce inhibition. Tnstead,
Iavlov is unequivocal in asserting that such a procedure leads to
the development of inhibition.
It may be that the development of inhibition resulting fro, the
CS-only trials was responsible for the reduction of the inhibitory
process conditioned by the reinforced trials. 6ut since the develop-
rent of extinctive inhibition usually appears to be much slower than
what would have to be the case in the present experiment, perhaps some
other factor was also involved. It should he noted that Lull's (1943)
explanation of extinction, embodying the notions of reactive inhibition
(an inhibition similar to Pavlov's internal inhibition), and conditioned
inhibition (a learned tendency not to respond), likouise has difficulty
handling the present data. The theory implies that extinction will
occur gradually, unlike the abrupt extinction indicated by the present
The Disiiihibition 'lypotiiesis
It is possible that the factor responslhle for the ra.'id loss of
the conditioned inhibitory process is similar to the disruption of
inhthition, i.e., dlisinihihtion, de-onstrited in other studies (I'Fvlov,
1927; !lovlanJ, 1936; tinnick ~iJ [lint, 1951) by tihe introduction cf
novel or extraneous stiluli during extinction. If this is so, it
could be that the change in the stimulus situation occasioned bIv tl.e
omission of the UCr: during thfl cxtinctinn trials was oi' s-ufTicirnt
novelty to produce the disinlibiting effect. :ovlan.l (l 3.') 'xplainedi
tho rise in tie CR extinction curve followin:r '.as-ed practice in a
similar way. He olberved that massing of reinforced trials produced
nn inhilitory effect ihlch resulted in a depression of the CR. i:e
labeled the phenomenon "inhibition of riinforcement," and hypothesized
that the early extinction trials would result in a disruption of this
inhihition. Ilovland concluded from the data that the increase in CR
strcnrth on the first few extinction trials waq due to the disinhibl-
tion of the inhibitiun of reinforcement. If the relatively rnpid
extinction in this study (i.e., only three trials needed to roacn
almost complete UIl:-recovery) was ,Ju to disitliihition, it is not
Isurprising tl.nt further trials had little a!Jditienal influence. This
,possibility, of course, siiuposts immediately an ema;irical test, n. rmly
n experiment on extinction of UCR-Jiminution ini which only one
extinction trial is piven. If the UCR in the Cr-1ufS test trial
following a single extinct, n trial is as Inrre as t1'ea NR carl)
in condlitior.ing, the disinhlibltion exrlrinarinn would be greatly
The purpose of the present study was to demonstrate the presence of
a conditioned inhibitory process in classical conditioning, and to deter-
mine whether it extinguishes in a manner similar to the extinction of
the conditioned excitatory (CR) process. To do this, two props of Ss
(N=45) were given 16 paired or unpaired CS-UCS (Lipht-shock) acquisition
trials. Following this, each of the two groups was divided into three
sub-groups differentiated by roccivin: either throo, six, or 12 Cr-only
extinction trials. In addition, another group was given paired acquisi-
tion trials, but during the "extinction" phase of the experiment it rec-
eived 12 UCS-alone trials. Following" the extinction trials, each group
received two paired test trials. It was hypothesized that:
1. An inhibitory process under the control of the CS would be
conditioned during classical conditioning trials;
2. The conditioned inhibitory process would extinguish during
C-only extinction trials;
3. The amount of extinction of the conditioned inhibitory process
would be a positive function of the number of CS-only extinc-
4. Croups which received paired acquisition trials would show a
more rapid rate of recovery of the UCR on the test trials thnn
p.roups which received unpaired acquisition trials.
The results of the experiment supported, albeit indirectly, the hypo-
thesis that a conditioned inhibitory process occurred in the r groups, and
not in the C rro:ps. InJircct evidence .aF roauireJ because the UCR.
of the P groups became only slightly smaller than those of the C crotup
Analysis of variance of I'ifference Scores hctwecn the I'CVn of tile
first test trial and tl-e final t,.; acquisition trials indicated that
tlero was a sinlific;anrly greater overall recovery of the 'JCR on tlh
test trial by the I rrou:'s than hy the C groiiis. This finding, lends
support to the hypothesis that the conditinnel inlibitorv process
extinruishes on presentation of C'-only trials. I'owever, the analysis
diJ not support the third and fourth hylpotheses.
The failure of the numbor of extinction trials variable to he
positively related to 'he amount of extinction was interpreted in rela-
tion to Pavlov's internal Inhibition concept. In addition, it 3as pointeJ
out that iprha3fs Srence's notion o" the JlevulorIrent of an inhibitory set
during early extinction trials could possibly account for tie abrupt
extinction. Alternatively, it was suggested that the findlnrs could have
Ieon die to disinhibition of the conditioned uithlbitory process.
Baxter, R. Conditioned diminution of the unconditioned response.
thUublished doctor's dissertation, University of Florida, 1964.
Cason, H. The organic nature of fatigue. Aner. 1. Psychol., 1037,
Cavapgioni, A., Gianelli, G., 6 Santibanez, E'. r. Effects of repe-
titive photic stimulation on response evoked in the lateral
geniculate body and the visual cortex. Arch. Vital. Biol.,
1959, 97, 266-275. Cited by Cranit, R. =i VTs''al Cortex.
In The jye, Vol. 2, Edited by Davson, I. 'Nw York I!d lor>in:
Ac, Tr.i ~ Tress, 1962.
Pufort, R. H., & Kimble, G. A. Ready signals and the effect of UCS
presentations in eyelid conditioning. J. exp. Psychol., 1958,
Calambos, R., Sheatz, G., Vernier, V. G. tlectro-physiolo7ical
correlates of a conditioned response in cats. Science, 1956,
123, 376-377. Cited by Grnit, R. The Visual Cortcx. In
The Eye, Vol. 2, Edited by Davson, ii. te\' York- and London:
Aca Ie-ic Press, 1962.
Creen, R. T. Surprise as a factor in the Von ;estorff effect.
J. exp. Psychol., 1956, 52, 340-345.
Grings, li. W. Preparatory set variables in the classical conditioning
of autonomic variables. Psvchol. Rev., 1960, 67, 243-252.
11aggard, E. A. On the application of analysis of variance to CSR
data: I. The selection of an appropriate measure. .1. ex.
Psychol., 1949, 39, 378-392.
Hornandez-Peon, R. Central mechanisas controlling conduction alone
central sensory pathways. Acta neurol. ,atinoater., 1955, 1,
256-264. Cited by Cranit, 1- "europysTTopy' of the retina.
In The Eye, Vol. 2, Edited by Pavson, !I. New York and London:
Acad oeic ress, 1962.
Hilgard, E. R. Reinforcement and inhibition of eyelid reflexes.
J. gen. Psychol., 1933, ', 85-113.
liilgard, C. R., & Marquis, D. C. Acquisition, extinction, and reten-
tion of conditioned lid responses to light in dogs. J. comp.
Psychol., 1935, 19, 29-58.
!lilgard, r.. I., C '4arquis, '. C. C(ndJticnit and lear.lnp.
:iew York: Appleton-Century-Croits, 1J4tl.
.lovlind, C. I. "Inhiblition tI' reinforceaont" and pheno.nenn of ex;,cri-
mental extinction. IPrc. nat. Ac.d. ecl., 1936, 22, 47,0-433.
linvland, C. I. The generalization of conditioned responses. III.
E1tinctrin, spontaneous recovery, nnd disinhibition of conditioned
and of generalized responses. J. ex,. Prychol,, 1937, 21, 47-U2.
Pull, '. I.. Princlnles of Behavior. New York: Arpleton-Crntury-Crofts,
.ludd, P. i. I'nsic correlates of the visual stvl ilus. In 5. S. Stevens
(id.) handbook of xrerrimental Psvcholonv. New York: 'l lcy, 1951.
Kimabl, ';. A. Coiiditionina as a furictioi of the tire between cor.ditinned
and unconditioned stimuli. J. exr. Psychol., 1947, 37, 1-15.
l:ir.hle, C. A. Iilgard and ':lrqiis' Coiiditionin! and I.earnin.. ::ew York:
A'pletlon-Century'-JCrn'ts, Inc., 1'61.
Kimble, C. A., C DGuort, R. II. ThiC aRsoci:tive factor in eyelid con-
ditionin-. .1. exr. .sychol., 1956, 52, 3P6-3n1.
Kmable, r. A., :'ani, L. 1., 6 Dufort, R. II. Classical and initrimental
eyelid conditioning. .7. cx. oPsychnl., 1'15S, 4Q, 407-417.
:iUble, C. A., 5 fst, 1. W. '. conditioned iihabi:ory prr'cess in
eyelid conditioning. J. car. Pslchol., 1961, (61, I.O- SC.
inmeol, II. P. The rnlatinnship ibetacen direction and onrount of stirulus
ch.inpe .nd arnunt of norcoptual disparity re~"ons .. .. ep. Psvehnl.,
1960, SO, 68-72.
immeel, I. P., F Pennyracler, H. S. Conrditined diminution of the uncon-
dltioited response a-' a function of the inirber of reiinorceamets.
J. ex!'. Psycehol., 1962, 64, 2C-23.
Kloitmrn, t'., Crisler, C. A qigntitative study of a salivary condi-
tinnned refl(cx. maer. .1. Pt.vsirl., 19)27, 79, 471-1-1.
Lacey, ". I.., 5 Siegel, P. 1. An analysis lf' the unit of ecafurement
of the nglvanic skin response. J. exr. Psychol., 1I94, 39, 22-27.
?FcAllistor, K. y. Tyclid conditioning as n function of the C!'-UI:S
interval. J. exp. Psychol., 1.13), 45, 417-422.
:iicholls, t. r., i'ible, G. A. Irfect of instructions upon eyelid
conditloninR. J. exp. Psvchol., 1964, 67, 400-402.
Norris, F. B. r Crant, D. A. Eyelid conditioning as affected by verbally
induced inhibitory sot and counter reinforceL-ent. Amer. J. Psychol.,
1948, 61, 37-40.
Pnvlov, I. P. Conditioned Refloxes. (Trans. by r,. V. Anre-.). London:
Oxford Univesity 'ress, 1-27.
Pazran, G. II. S. Extinction re-examined and re-analyzed: A new theory.
Psychol. Rev., 1956, 63, 39-52.
Spence, K. W. (Lecture presented at University on Florida, April, 1963.)
Spence, K. W., 4 RLunquist, i '. Ten,;poal effects of conditioned fear on
the eyelid reflex. J. exp. Psychol., 1958, 55, 613-616.
Spence, K. W., llomzie, !i. J., f rutledpe, E. F. Extinction of the human
eyelid CR as a function of the discriminability of the change from
acquisition to extinction. J. exp. Psychol., 1964, 67, 545-552.
Thorpe, W. ii. Learning and Instinct in Animals. London: ?ethuen and Co.,
Winnicl, W. A., F Hlunt, .. IcV. The effect of an extra sti-ulus upon
strength of response during acquisition and extinction. J. exp.
'sychol., 1951, 41, 205-215.
This is an experiment on the effects of certain types of stimuli
upon the galvanic skin response. Your task is very simple: all you
rust do is stay awake and pay attention to the stimuli. Please
remain motionless throughout the experiment.
AI'Pr.:I x B
T'A' Ll 2
Analysis or Variance of I'Cr Itifferonce scores
Source d F
I'airin (Paired-Unpai red) I .f(m3 i.19*
Extinction (3, 6, 12 trials) 2 .:5.3 1.77
Pairing X Extinctlion 2 .IS0 1.2?
I:rrnr (within) P4 .1132
4 DMn C -C
NC ( N O r r rN 1 ^^ Ci cc \c C E tO Oc 0
C) w cc Co' cc Ut o rC W, ge- cN 0
p-i -I -
o iM 0 'A Cm N 0 'C N c
CM *-i *C **N*m *'- cc *I- c -N
mc c ir. p^ yf M c c f-<<^
S: .-. .1 4 r!
.C CNU ,r Cm e
00 CN N I -
S; ; co a
U .N N r 0 UC 0 0 ? U, 0
S N O N NN ^ 1 r V ( p ^N
Hl -v o-CT*' ^-
C m * * *
N- M lT L O COW0- NM -' Ln
SZ N CN C C. CN C N L. L N
4 NON~CCNtuaUNa N
N U C C. Ln K C -0 ".
.P. NNic.NC N.7 c L r..
SC U NN -n -N C r
U OLUM N o-. p. N C i/0N iC-m7
ss i ri N r
Zs Ln BO '5 cc %a w. c f I J. r, M VF
C c -V .
ra z. w f^ L n c% x T%
C..7-.N DN 7C : .n C
S .. .. . .
% r C r. n fj *- r, c
N c' C C C L n .M N w
N . ic. N N C C i a
SC . . ...7 . .
-e o o S (y- C C Q
K s. * e * * * *
r -IT' L- -n Q C% V C
'3 mL9.N 0JM (Na00
t C Ct bO MNO lMN
I N*- V c'. M *-' IN Q 0 0
S- 0 T0 00 m q C. T 0 M m.
- o C! 0 C "I d C. 00io 00 l
'3 0 ti C.i s C- l I l. 0 '3 CO
=. L") m IN C4 4o y co V) M G -7m o 0
C, r- a - V vn
N N N 00 C
C V Ncf CM rM VW WW
0 0 1. L C. I ( m V C
q.. c - N C M-' 0 i, e3 C NMO
0N0 N ..trii MOIto T.
3 r rCint^<>M C1tN C O C3 tM r
i/iM l~cO~a'O sa t cr t '-*-
CT-ty ~~t/trMm^Cl^ '^Ct^ i
N- M In 'o r 0 N
i n f C r c C, t, f I- v C
0 C nC.I. .N -c.
e .. N W. CM C.. .
c O r ir- o n r- r. ... -.
St rai i Ir ri r O wi r-. w. t
P --- -. C. -
C f --M--= -c
C: r c a o e- c.
0r 0 re
*d P' o o c c' r c e'i ce. .c I
C' o-a .- 'C r) U: I- C' C' O C' C I".
a C' 'a r. e e. e Ca .
a9r. N (- r C ( C C W 1 (' C'." '
u~N aS ... ..
- a. C')C'- ('aC') C
C M m- C O 0 Cl rN V A O
LA c** ** ** ***A ***.N
Ne 0-t \0 z0 C3N L
LA NA L i i4 0 u -j N 0 a C li LA O
u1M .n 0 WD N r M M f V', N
LM * * **C * **C
N C N L 0 N L . C N
br rr f M- -
^ f- .^ -4 4 erf
fFr^T ^t -m o ivecor
.4 L t A. N O 4 0-4N L L
Sc -.-.i r!* ?- V c% .. I c ..!
SN. Z 5 'T i o -. i0 N r Z L0 r0 n'
^ < ^3 c i nI. Q- m rl i
- o ncc C C', N c
Osc N.A'7-T CC'.'7 OS^f~ oos0r r
S. ) G 30 C r' C 1- r. C. Ci
. -c CA - -.
V'. r n c. B Ln m C .,m
C'. . . . r . r .
/ C. *. .. .CC
L. -cr -
II -T f ti s^ r ir c r1% s rc 7 ; r -
rI~- ( 2 C'.oot- r r-c-tr i 7. 1- C
. . .. N .
N- N C'.r NT -s rC c. 7 0) iAN .
N -M ic i .N r I" N- I ri' rS r r.
WI r l N N I r C. N -^ - S .'.
'3 03C-00 80u01
iNO Cr 0 0 C 0 m D3-
S. 1r.t'i i I n C, o 0v C tNN0!
Nl N 0 N N LA
No -to --
LN ef^3 t'0oc &3*I.- i(MNe nC N.
CM O O OCC NSMO~ i
Cn cl 1 C' Ki L o t c t^ ^ i CIr-
O Li I. -
SN V C -C'0 N N4 r MiO t)* 0
c 0 Ur n ( 1 cT -I m m N N
0 00 0 N NC
~ I- CC 0. N N N 0' 0 N N .
- -- --- -r
c- No 0 j I N r 0 -* NC 0* U*
i ccI I en3l^T o 3~-T ~ a
t- I 1 w 3r't^ Ci iA r^SOC1 O ~r~Ijd LT Ir ;
C, c I'. r' -- w C ** N 1"- ti C,, ,. -
C V. Er. Q CC M cs l c S
r. C[ E Er c w r II n r M c W EN C
rF C rj r -. C C'. C CED Q. c! -T r
X ZMeILI -v InS e m
-7 1- ". C LI C Cl K c. C, I 0 N 3
-< t0 0 csooM -LwocM m
OA 'OD 'o 0Nm 'o L
CM01. m Or, v *****)C '
wa I-0 c00"W V0)0
D ID q-I C
- m r CD c
oi II ma m
oUw O O-cCl" o c
.3 n o -C o r> ci\3o\o e iT.
-- -o cr O M mre i u O
5 U 0* U)TrC)
- r- 4 -i
(-* XC rC T- M IOS3r
U! I I I f~~ Mr- -e M M ~ a 4fi O riC
C; I M * *r *~ *C\ , ) * I *
3: ,-, I- _|
M SI I f 3 T1r
-el ti) %A) D Cr, J0 o- C Cl) m v U.
1 r'0 o nI C. 02 C :- r. Cc I. '
0 r. C. 01 I C ri 1C r- N r. C.
N r om. Lt m s r, r .vwr
Cr oe o. C C c LC
-- . . . . .
V. r- t .-. c Z N m r. C c %-
ie* c- o a -e e eew
N- CO CC C C C!r0 I
S oo C. C. eoo % C
ooeme c.-.e -Ccl.* c-
Sr c e m .x 0 C - .
c C .: N N C C c C N* : : Ln-
O C; aC c C c N c' .
M- NV mt i =C N 0 N -m Z.
C nc- f Ctrrfi-'co w-
.- --- -- --
0C '0 Q N.) C4
OC 0 C 0 vC NM 0 S C t-
m V. C c c m C. . O
c U c c m u", :
U CO 00Cl-4 tl
0 CO C C C, A
0iV MDO MeeN r --
C K 1 C.0 l O. C &Oe'M C C 0'3 0 0
S C rN N C-
0 ui r^^ab^^uC'.^cm' C i~ro
rt a- I' fi- i a I fl ocs I r VNojo
^ ^ s * * *
T- N Mr, = Crm -n CIA m3 OLfM
-' C CI In - i -. -a l o C C
Go C Ll F'- z 7 .r 0 r c.
Sc M C: 7d L. C7 i c .IC
C C- r" Cw K v r rt C i OZ L
- B. C. C .N C c N I N V.
;CC CCo w, C C- cc. I
-v eo c c. r,4 co i. c rr -,
C C 7 C 0 rC c C C C C.
7 C. C C" N C C C tL r i/. r L".
Ir- K I c ao N C 0 V ^ L C C.
v. fI i LA r N ri ri ir C c ,- N c oaC
*9 N N N c G: C. p N C c rM cc
ti N CN I
C CD 7 Ch 0 Ce 00 a C o o M
Cr C)M m 0' n N C = = ou v c '5 m
OO-csrC N b C NCO N
o * * * * * * *
(M * *
I- rOCsr M t^ lyS -
-(VIM rinci^ I CT S O .'r^~f
'-'^M ~ /i^r^c rcC -it w~a t
nr c. a C. cC c p C. C r rm c im
o; cr. C O T e, C z c. C - = C C.
7r f C co c, O i 1.7 In L-1 Lr, C C -
M C. CN C uC Lu W c C C rCp
. . . . . . . .
go, %Z r- .- C r4 C rf C a J
0 .o ~c: C C. C' -3 C'
'L* N N C.. CO C .C S
Is J c a" NC C *' C" c *. I? Pa -
- N '. < N r 0 ". C -I N i' L cIc
.. . . . . . . .
U I' . . . . . .
3"' "I I
i- c u '
^ 0 0 tnrM'7*^ C~t^TCtr^ rn
^ r ry -. 'c \K TO !f ^T --g'
< * * .
PcsponsrTe 'nr.itudos [(Clq .c) lor C'roI; c l' I ;rst
Trinis rollowin, IReinforced Trials
Three UCS -ulor.
1 2 .
1.17 1.73 1.17
2 .3. .'2 .84
3 1.18 1.2l 1.2e
4 1.27 1.40 1.04
5 1.07 .Q5 .9n
G 1.48 1.34 1.45
7 .48 .19 .18
8 1.13 1.33 1.31
9 .82 .81 .7J
10 .93 1.00 .82
11 1.10 .23 .15
12 1.85 1.57 1.15
13 .94 1.23 1.23
14 .73 .57 .88
15 1.22 1.09 1.09
UCR (fnanitude (lo! Ac) Difference Scores Between 'ean of Final Two
Reinforced Trials and %ean of First Fhree UCS-alone Trials
Following Reinforcement for Croup Cucs
S Difference Score
I. -C C c m r W ?' L 0 n c
C ,LI -M rIM C c ,
. . I I I I I I
c C N c c r i . C.
CO .-1 o5 c c c c- a- r
C NL .. .C K .. ..- C -
- N V.^ ". N C C: -r NK-iO
S . . . . .. . .
I C a r I
me.-ac oor -oNo ov
u Cl.'i-'cloci c 0 CCc i t C; 0-c*
C OOM N CCMCCCC
v3CO l GOO
CO CC C'O tC-clC C
Mo c- C o C0 oo
O C C:OOC C. NCC
-4-4 -4t -N -4
~< t < < - C^ ~* -
C TO 0 MOCC C Cr
C OC CNimBr1fk
C C 0 C. c C 1 .11 Ce C0 0 a
C 0 C C OC W Cc CD nc C m'v
CO C C)c OCC CCC lNC C
0 CC. CC
CO C C O C'CT OC)~'C.
SC O -' C 0 O 0 0 n 0 C C.
C. C L- C C 0 C O C C
- 0 .' i. .z r. .O . 7; '-. *
C3w c a' D"oc rc ar
CCr-.=OfaC. O rOC"r-ooo;
- i rvn v Lf, z r- to c* c r i I ) t,
(-N Wi ifi M Z ci C r4i M l t
N -O O 0 00N e0'3
004 00 N. N 0 o
-N rl Ir- 0 i tS Ia0 'mC' n -
CMIDMOO tr ItMNN< O
'30 0 t.. N 0
I 1 I I II
CO TOfI N OOt 3O^ I
I I I I I SI
U^ Li c ('. '3 (N IN O 0 tI DO
BICC.PA'IIICAL SLI TCO!
Ilitcnel Crawford aorrnw wns torni Octoner 21, 1932, .it
Louisville, Alah.amn. lie artoe.doi public schools at Jacksinville,
lrorida, nnd w:as Rraduzted from Hobert E. Lee Conior Iligh Ccl.ool
in January, 1952.
Mr. uorrow entered the University of rloridn in eptetmber,
1954, after serving three years with tho II. S. Ar:n. lie received
the degree Rachelor of Arts in psychology in January, 19Ot, and
in frbrluary, 196n, enrolled in the graduate school of the Univer-
sitv of rlorida. After ljeinp jiwarde,]c the depree l:ister orf rts
in psychology in .January, 1962, "r. 'orrod continued his training
at the Ilniversity towards the decree Doctor of Philosphyv.
Mitchel Crawford 'borron is currently an Assistant Proftssor
of Psychology at the University or .'nuth FIr.rida. 'le is married
to the former Marifrances Tucker, and is a meabor of Psi Chi nt
the University o I'loriia
This dissertzition was prepared under the direction of the chairman
of the candidate's supervisory comnittct and hao !,een apnrovoue by all
members of that comittcn. It was submitted to the noan of t!e Colle, e
of Arts and Sciences and to the Craduamte Council, and was approved ss
partial fulfillment of the rcqu)irements Cor the de-yreo o.f Doctor of
December 19, 1964
I.' I, L. i v .: rtA r .in.
'c rn, r',r' ,ln te ,ch, ln