Journal of Experimental Psychology

1973, Vol. 101, No. 1, 155-158

BLOCKING OF THE RABBIT'S CONDITIONED NICTITATING

MEMBRANE RESPONSE IN KAMIN'S
TWO-STAGE PARADIGM '
HORACE G. MARCHANT, III, AND JOHN W. MOORE 2
University of Massachusetts

Blocking of the rabbit's conditioned nictitating membrane response was

demonstrated in Kamin's 2-stage paradigm. Blocking of a conditioned
stimulus (CS) of strong salience (76-db. tone) with a CS of weaker salience
(light) required extended conditioning to the light in Stage 1 of Kamin's
paradigm. This result is consistent with predictions of the Rescorla—Wagner
model.

This report describes a successful exten- Method

sion of Kamin's (1968, 1969) 2-stage Subjects and apparatus. The 5s were 20 naive
blocking paradigm to a classical condition- albino rabbits weighing approximately 2 kg. each.
ing preparation other than the conditioned A detailed description of the apparatus may be
emotional response (CER). Briefly, the found elsewhere (e.g., Marchant, Mis, & Moore,
2-stage blocking paradigm consists of ex- 1972). Briefly, 4 animals were run concurrently
in individually ventilated and soundproofed file
citatory conditioning to Stimulus A (Stage drawers while restrained in Plexiglas boxes like
1) followed by conditioning to a simultan- those described by Gormezano (1966). A "mini-
eous compound of A and B (Stage 2), torque" potentiometer mounted on the rabbit's
After what would ordinarily be a sufficient head was connected by a small metal hook and
number of trials for B to become condi- thread to a nylon loop sutured into the nictitating
membrane (NM) of the right eye. Movement of
tioned, tests in extinction reveal little or the right NM produced a dc signal which was ampli-
no conditioning to B. It is then said that fied and recorded on a Grass 5D oscillograph. A
the prior conditioning to A blocked con- conditioned response (CR) was defined as a pen de-
ditioning to B. Wagner and Saavedra flection of 1 mm. (corresponding to an extension of
the membrane of less than 1 mm.) or greater within
(Wagner, 1969) have previously reported the 500-msec. CS-UCS interval.
blocking in the rabbit eyelid preparation The CSs were a pure tone (T) of 1,200 Hz. and
using a paradigm which differed from 76-db. sound-pressure level delivered via a speaker
Kamin's in that reinforced A and AB located directly in front of the animal's head
trials were alternated from the outset, i.e., (Stimulus A) and the onset of 2 4.5-v. incandescent
lights (L) mounted on either side of the speaker
Stages 1 and 2 were in essence run at the and behind translucent white plastic screens (Stimu-
same time. lus B). The unconditioned stimulus (UCS) was an
ac shock of 2 ma. and of 50-msec. duration delivered
through 2 stainless-steel wound clips (Clay-Adams
EXPERIMENT J 9 mm.) applied to the infraorbital region of the
Experiment I sought to demonstrate right eye. A low level of masking noise was pro-
vided by blowers ventilating each file drawer.
blocking of a light conditioned stimulus Procedure and design. All animals were sutured
(CS) with a tone in Kamin's (1968, 1969) and habituated to restraint and their file drawer
2-stage paradigm against a variety of con- for a period of 20 min. on the day prior to the first
trol procedures. conditioning session. Each session consisted of 100
trials at a constant intertrial interval (ITI) of 30
1
This research was supported by National Science sec. Except for 2 yoked control groups described
Foundation Grant GB-24557 to the second author. below, each rabbit received 4 conditioning sessions
The authors wish to thank F. W. Mis, P. Solomon, in Stage 1 unless a level of 80% CRs was attained
L. Remal, and D. Messenger for their assistance in over Sessions 2 and 3, in which case 5 sat out
various aspects of this study. Session 4 and began Stage 2 training on the next day
2
Requests for reprints should be sent to John W. (Day 5). The 20 rabbits were randomly assigned
Moore, Middlesex House, Department of Psychol- to 5 groups of 4: (a) a blocking group (T-TL) re-
ogy, University of Massachusetts, Amherst, Massa- ceived 3-4 daily conditioning sessions to A in Stage
chusetts 01002. 1 followed by 5 sessions with the compound AB
155
156 HORACE G. MARCHANT, III, AND JOHN W. MOORE

TABLE 1 tions), Group T-TL yielded only 1 CR in

SUMMARY OF EXPERIMENTAL DESIGN AND RESULTS extinction, indicating a powerful blocking
OF EXPERIMENT I effect when contrasted with Group sit-TL
Stage 3 Total no.
and each of the other groups in turn, Mann-
Group* Singe I Stage 2 (2 clays
of ex-
of CRs in
Stage 3
Whitney U = 0, p < .03, for each compari-
(3-4 days) (5 days) tinc- out of son. Except for Group sit-L, which was
tion) 240
never conditioned to the tone, all groups
T-TL T TL L 1 gave approximately the same number of
sit-TL — TL L 77
TL-T TL T L 66 CRs to the tone during testing (44%-49%
T-L T L I. 104 CRs) and had similar generalization gradi-
sit-L — L L 161 ents (not shown) along the audiofrequency
dimension. Consequently, the blocking of
Note. Terms: CR = conditioned response; T = tone; L =
light;
a
sit = yoked control group. the light in Group T-TL cannot be
In all groups, n = 4.
attributed to gross differences in resistance
reinforced. Stage 2 was followed by 2 successive to extinction produced by generalization
daily extinction sessions with 30 presentations of L decrement.
randomly intermingled with 70 presentations of T A comparison between Groups T-L and
in each session. (A larger number of tone test trials sit-L in Table 1 suggests that Stage 1 con-
than light test trials were presented in order to ob- ditioning to the tone may reduce condi-
tain generalization gradients to other tonal fre-
quencies.) The other 4 groups were tested in the tioned responding to the light in Stage 3,
same way. (V) The 5s in Group sit-TL were yoked even though the 2 CSs are not compounded
to the blocking group during Stage 1. These animals in Stage 2. Unlike the comparison between
experienced handling and restraint in their file Groups T-TL and sit-TL, however, this
drawer but experienced neither CS nor UCS until
Stage 2. Rabbits in Group sit-TL received com- difference was not statistically significant,
pound conditioning trials to the tone plus the light U = 5, p < .30. Although acceptance of
in Stage 2 followed by testing in extinction, as in the null hypothesis in this case is tenuous
Group T-TL. (c) A Third group (TL-T) was in- at best, it is doubtful that any proactive
cluded in order to evaluate the effect of order of inhibitory influence of Stage 1 conditioning
treatment on blocking. A similar control was run
by Kamin (1969, p. 46). Animals in this group on Stage 3 responding to the light was as
received compound conditioning trials in Stage 1 for great for Group T-L as for Group T-TL
3-4 sessions followed by 5 additional sessions to the when each group is contrasted with its re-
tone alone, (d) Group T—L was included in order spective control. Another observation
to determine the extent to which prior conditioning
to the tone inhibits conditioned responding to the which seems intuitively counter to the ap-
light during testing even though the light was not parent proactive inhibition in Group T-L
compounded with the tone in Stage 2. (e) A fifth is the fact that this group made 70% CRs
group (sit-L), yoked to Group T-L, was included on the first day of Stage 2 compared with
as the logical control for any proactive interference only 8% for Group sit-L, U = 0, p < .03.
in Group T-L. Animals in this group experienced
handling and restraint in their file drawers during
Stage 1 but no conditioning. Stage 2 for these EXPERIMENT II
animals consisted of 5 conditioning sessions to the
light alone. All groups were responding in excess Experiment II attempted to replicate
of 90% CRs for at least 2 sessions prior to Stage 2 the blocking effect observed in Experiment
except for Group T-L, which had a mean CR rate I but with the light CS as Stimulus A and
of 87% over Sessions 3 and 4 of Stage 1. The experi-
mental design and procedures are summarized in the tone as Stimulus B.
Table 1.
Method
Resulls
Eight naive rabbits were divided into 2 groups of
The total number of CRs to Stimulus 4. Group L-LT received 3-4 days of conditioning
B (light) over the 2 extinction test ses- to the light in Stage 1 using apparatus and pro-
cedures identical in all other respects to those em-
sions are given in the right-hand column ployed in Experiment I. After attaining a mean
of Table 1. Out of a trial of 240 test level of responding of 91% CRs over Sessions 3 and
trials (4 Rabbits X 60 Light Presenta- 4, rabbits in Group L-LT received 5 additional con-
 BLOCKING THE RABBIT'S NICTITATING MEMBRANE CR 157

ditioning sessions to the light-plus-tone compound. at the end of Stage 1, blocking of B will
This was followed by 2 successive daily extinction not be obtained. Since the values of 0 and
sessions, each consisting of 50 presentations of each
CS component presented separately in an unsyste- X depend upon UCS intensity, which was
matic order. Animals in Group sit-LT, yoked to held constant in the present investigation,
Group L—LT, experienced handling and restraint differences in the rate of growth of FA de-
in their file drawers for the 3-4 sessions of Stage pend on the parameter a, which is identi-
1, followed by S compound conditioning sessions to fied in the model with CS salience. Thus,
the light plus tone and 2 extinction test sessions as
in Group L-LT. the 3-4 days of Stage 1 conditioning to the
light may have been insufficient to increase
Results FA to the point where increments to B
would be negligible during Stage 2. This
The 4 animals in Group L-LT gave a argument has validity in terms of the
total of 138 CRs (out of a possible 400) Rescorla-Wagner model if and only if the
to the tone during extinction, while the initial rate of conditioning to the light is
control group (sit-LT) gave a total of 202. demonstratively lower than the rate of con-
Although this difference was not statisti- ditioning to the tone. Such a demonstra-
cally significant, the direction of the differ- tion is necessary because the 3-4 days of
ence is consistent with blocking. Neverthe- Stage 1 conditioning to the tone in Experi-
less, these results fail to replicate the ment I was sufficient to block conditioning
powerful blocking demonstrated in Ex- to the light in Stage 2.
periment 1 when roles of the 2 CSs were Although the mean percentage of CRs
reversed. over Days 3 and 4 in Stage 1 was 96% for
Group L-LT, gave a total of 143 CRs to tone (Experiment I) and 91% for light
the light, while Group sit-LT gave 71. (Experiment II), this difference was not
This difference (not significant) was prob- statistically significant. Therefore all exist-
ably due to the greater number of condition- ing data protocols from our laboratory con-
ing trials to the light in the former group, forming with the procedures employed in
although "overshadowing" by the tone in Stage 1 of the present investigation were
Stage 2 might have been partially responsi- pooled for the purpose of determining
ble for the comparatively low level of re- whether the light and tone CSs differed
sponding to the light in Group sit-LT (cf. with respect to rate of conditioning. Suit-
Rescorla & Wagner, 1972). In support of able data were available for 23 animals
the idea that overshadowing occurred is conditioned to the 76-db, tone and 12
the fact that there was a greater percentage animals conditioned to the light CS. Com-
of CRs given to tone than to light by all parison of CR frequency over 3 sessions of
sit-TL groups in Stage 3. 100 trials indicated a mean CR rate of 58%
for tone 5s and 41% for light 5s, t (33) =
EXPERIMENT III 2.44, p < .05. Since the tone is demon-
Rescorla and Wagner's (1972) theory of stratively more salient than the light in the
conditioning can account for the failure to sense of the Rescorla-Wagner (1972) model,
obtain significant blocking in Experiment increasing the number of conditioning trials
II if one assumes that conditioning to the to the light in Stage 1 should result in a
light failed to reach asymptote by the end sufficient excitatory strength (FL) to block
of Stage 1. Their equations describing the conditioning to the tone in Stage 2. Ex-
increments in CR strength (V for value) periment III provides evidence on this
to the A and B components of a reinforced prediction.
compound trial are as follows :
Method
FA = «A/3(X - FAB)
Eight naive rabbits were divided into 2 groups
FB = an/3(X - F AB ), of 4. Group L-LT received 10 successive daily
sessions of 100 trials, each consisting of simple con-
where FAB = FA + FB, 0 < a, ft < 1, and ditioning to the light CS. This was followed by 5
X > 0. If FA is substantially less than X additional conditioning sessions with the tone and
158 HORACE G. MAKCHANT, III, AND JOHN W. MOORE

light compounded and 2 extinction test sessions in ment I are consistent with blocking data ob-
which the light and tone were presented separately tained in the CER preparation (e.g., Kamin,
50 times per session in an unsystematic order. 1969). Second, the Rescorla-Wagner (1972)
Animals in Group sit-LT experienced handling and model, which had previously been success-
restraint in their file drawers for 10 sessions followed
fully applied to Kamin's results and the
by 5 compound conditioning sessions to the tone
and light and 2 extinction test sessions. Other Wagner and Saavedra experiment (Wagner,
aspects of the procedure were the same as in Experi- 1969) cited above, can account for the essential
ments I and II. features of the present results. The Rescorla-
Wagner model seems a particularly attractive
Results representation of the data because of its ex-
plicit and successful predictions regarding the
Mean percentage of CRs to L on Day 10 role of CS salience and amount of Stage 1
of Stage 1 was 99.5%. The 4 animals in training in blocking without recourse to selec-
Group L-LT gave a total of 32 CRs (out tive attention as an explanatory mechanism.
of a possible 400) to the tone during extinc-
tion, while the control group (sit-LT) gave REFERENCES
a total of 293. The difference between
GORMEZANO, I. Classical conditioning. In J. B.
groups is statistically significant, Mann- Sidowski (Ed.), Experimental methods and instru-
Whitney U = 0, p < .03, and conse- mentation in psychology. New York: McGraw-
quently blocking of the tone by the less Hill, 1966.
salient light was demonstrated. KAMIN, L. J. "Attention-like" processes in classical
The corresponding totals for the light conditioning. In M. R. Jones (Ed.), Miami Sym-
posium on the Prediction of Behavior, 1967: Aver-
were 257 for Group L-LT and 86 for sive stimulation. Coral Gables, Fla.: University
Group sit-LT, respectively. This differ- of Miami Press, 1968.
ence, U = 1, p < .06, reflects overtraining KAMIN, L. J. Selective association and conditioning.
to the light in Stage 1 of Group L-LT and In N. J. Mackintosh & W. K. Honig (Eds.),
quite possibly overshadowing of the light Fundamental issues in associative learning. Hali-
fax, Nova Scotia: Dalhousie University Press,
by the tone in Stage 2 of Group sit-LT. 1969.
MARCHANT, H. G., Ill, Mis, F. W., & MOORE, J. W.
DISCUSSION Conditioned inhibition of the rabbit's nictitating
The principal findings of this investigation membrane response. Journal of Experimental
Psychology, 1972, 95, 408-411.
are (a) preliminary conditioning of the rabbit RESCORLA, R. A., & WAGNER, A. R. A theory of
nictitating membrane response to one CS Pavlovian conditioning: Variations in the effec-
blocked conditioning to a second CS when the tiveness of reinforcement and nonreinforcement.
2 CSs were subsequently compounded; and In A. H. Black & W. F. Prokasy (Eds.), Classical
(6) the apparent difficulty in blocking a more conditioning II. New York; Appleton-Century-
salient CS (tone) with a less salient CS (light) Crofts, 1972.
in the 2-stage paradigm was overcome by in- WAGNER, A. R. Stimulus selection and a "Modified
creasing the amount of preliminary condition- Continuity Theory," In G. Bower & J. T. Spence
ing to the less salient CS in Stage 1. (Eds.), Psychology of learning and motivation.
Vol. 3. New York: Academic Press, 1969.
Two aspects of the present investigation
warrant emphasis. First, the data of Experi- (Received February 5, 1973)