TIME DEPENDENT CHANGES SN CONDITIONED
SUPPRESSION

Tnesis for the Degree of M. A.

MiCHiGAN STATE UNWERSITY

RODNEY CHARLES HOWARD
1977

’f L13 1-;- R Y
Michiga: State
gr? Umw 1y

 

ABSTRACT

TIME DEPENDENT CHANGES IN CONDITIONED SUPPRESSION
by

Rodney Charles Howard

The Kamin effect was investigated using a conditioned
suppression procedure in a within-subjects design. Four
groups of pigeons received Pavlovian conditioning "off the
baseline" immediately followed by operant conditioning.
During the Pavlovian phase, two groups received a forward
pairing of a tone with shock, one group received a backward
pairing and one group received a truly random pairing. One
of the forward pairing groups also received a delay between
the Pavlovian and operant phases. For all groups, keypeck-
ing was reinforced on a variable interval schedule during
the operant phase. Testing sessions were identical to
training sessions with the exception that the tone used
during Pavlovian conditioning was presented either 0 min,

15 min, 30 min, 45 min, or 60 min following the onset of the
operant phase. Testing sessions in which the Pavlovian

phase was omitted were also included. The results showed
U-shaped functions for 11 of the 12 pigeons in the forward
pairing groups and for 3 of the 5 pigeons in the truly random
group. The functions were flat rather than U-shaped for

pigeons in the backward pairing group.

TIME DEPENDENT CHANGES IN CONDITIONED SUPPRESSION

BY

Rodney Charles Howard

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS
Department of Psychology

1977

ACKNOWLEDGMENTS

I would like to thank Dr. Mark Rilling for his aid
and supervision of my thesis research and for his substantial
assistance in the preparation of my thesis manuscript.

I would also like to thank the other members of my
thesis committee, Drs. M. Ray Denny and Glenn I. Hatton for

their constructive comments and their valuable suggestions.

ii

LIST OF TABLES . .
LIST OF FIGURES .
INTRODUCTION . . .
MTHOD O O O O O 0
Subjects . .
Apparatus .
Procedure .

Pretrai

Trainin

Testing
RESULTS . . . . .
DISCUSSION . . . .

LIST OF REFERENCES

TABLE

ning
g 0 0

OF CONTENTS

iii

Page

iv

12
21

28

LIST OF TABLES

Table Page

1 Average Response Rate During the Tone
and Pre-Tone Periods . . . . . . . . . . . 16

iv

Figure

LIST OF FIGURES

Page

The level of response suppression as a
function of the testing interval for
individual pigeons in the forward pairing

(F) group . . . . . . . . . . . . . . . . 13

The level of response suppression as a
function of the testing interval for
individual pigeons in the forward

pairing with operant delay (FD) group . . 14

The level of response suppression as a
function of the testing interval for
individual pigeons in the backward

pairing (B) group . . . . . . . . . . . . 18

The level of response suppression as

a function of the testing interval

for individual pigeons in the truly

random pairing (R) group . . . . . . . . 19

Group averages showing the level of

response suppression as a function

of the testing interval for the forward
pairing (F), forward pairing with

operant delay (FD), and backward pairing

(B) groups . . . . . . . . . . . . . . . 20

INTRODUCT ION

In most studies of learning, retention is a monotoni-
cally declining function of the amount of time since original
learning. An exception to this rule is the U-shaped func-
tion, known as the "Kamin effect," obtained following
aversive conditioning with infra-human organisms. Kamin
(1957), in a between subjects design, incompletely trained
rats to avoid shock in a shuttlebox apparatus using a tone
as the CS. Following this training, the subjects were
divided into groups. One group was immediately given
additional training on the shuttlebox task while the other
groups received delays between initial and subsequent con-
ditioning that ranged from 30 min to 17 days. The mean
number of avoidance responses during retraining was used as
the measure of performance. The results showed a decline in
successful avoidance responses up to the one hour delay
period. Groups with delay intervals greater than one hour
showed progressively higher successful avoidance responses.
For the 24 hour delay group, the number of avoidance
responses did not differ significantly from the group
retrained immediately. When performance was plotted as a
function of the delay interval, a U-shaped function was
obtained with the greatest deficit in performance occurring

at the one hour delay interval.

The "Kamin effect" is always measured as a change in
performance obtained after various retention intervals follow-
ing the occurrence of aversive conditioning (Anisman, 1975;
Brush, 1971; Klien & Spear, 1970). The function reflects
changes in the ability of an organism to learn or perform an
avoidance response. It is reasonable to assume, however,
that if an organism's response to a CS used in Pavlovian
aversive conditioning was measured over time, a U-shaped
function would also be obtained.

To determine if time dependent changes in an organism's
reaction to a CS occurs, it is necessary to use a procedure
that does not require a specific avoidance response. Con-
ditioned suppression, a decrease in the rate of an operant
response during the presentation of an aversive CS, fulfills
this requirement. Hunt and Brady (1951) developed a varia-
tion of the conditioned suppression procedure, known as
"off the baseline" conditioned suppression, in which aver-
sive Pavlovian conditioning occurs prior to testing the CS
on an operant baseline. This method permits the manipula-
tion of the retention interval between classical condition-
ing and testing of the CS on an operant baseline.

An "off the baseline" procedure has been used by
McMichael (1966) and Tarpy (1966) to measure time dependent
changes in the level of conditioned suppression. In both
experiments, rats received avoidance training in a shuttle-
box, followed after various retention intervals by testing

in an operant chamber. Testing consisted of the presentation

of the tone used in training while lever pressing was rein-
forced with food. Their results showed a monotonic increase
in conditioned suppression as a function of the retention
interval. The failure of these studies to obtain a U-shaped
function led Brush (1971) to tentatively conclude that a
U-shaped function is not obtained when a conditioned suppres-
sion procedure is employed.

In an unpublished study from this lab, rats were given
avoidance training in a shuttlebox. Following training, the
rats were immediately placed in an operant chamber. A
suppression measure for bar pressing was then taken at vari-
ous retention intervals. Results showed a U-shaped function
with rats at intermediate retention intervals showing maxi-
mum levels of suppression. The purpose of the present study
was to determine if a U-shaped function is obtained follow-
ing aversive Pavlovian training when the tone used in train-
ing is superimposed upon a baseline of operant behavior.
The present study employed Pavlovian conditioning to avoid
any problems resulting from the acquisition of an avoidance
response during the training phase. To avoid any problems
resulting from conducting training in a different chamber
from the one used in testing, pigeons received both training
and testing in the same chamber. Pigeons were selected as
the experimental organisms because of the paucity of data
on the "Kamin effect" with this species.

In contrast with most previous studies on the "Kamin

effect," the present study used a within-subjects design.

The use of a within-subjects design permitted an evaluation
of time dependent changes in conditioned suppression for
individual pigeons. Klein and Spear (1973) also used a
within-subjects design to investigate the "Kamin effect"
but reported only group averages.

The present design consisted of an experimental group
and three control groups. The experimental group received
a tone followed immediately by shock during Pavlovian con-
ditioning. To ensure that subsequent suppression of oper-
ant responding during the tone was the result of the pairing
of the tone with shock, control groups were necessary. The
traditional control group in Pavlovian conditioning has
been a backward pairing of the CS with the US. However,
Rescorla (1969) has argued that a CS which follows a US may
become inhibitory because it predicts the absence of the US
for a period of time. He suggested a "truly random" corre-
lation of the CS with the US as the appropriate control
condition. In a "truly random" control procedure, the prob-
ability of the US in the presence of the CS is equal to the
probability of the US in the absence of the CS. Initially,
Rescorla (1969) predicted that conditioning to the CS does
not occur using a "truly random" procedure. However,
Benedict and Ayres (1972), Kremer and Kamin (1971), and
Quinsey (1971) demonstrated conditioning to the CS using
"truly random" procedures. Rescorla (1972) later predicted
that excitatory conditioning could occur using the "truly

random" procedure but that with extended training,

conditioning would dissipate. Given this lack of consensus
about the appropriate control group, both the traditional
backward pairing control group and the "truly random" control
group were used in the present study. The appropriate con-
trol group should fail to demonstrate time dependent changes
in conditioned suppression thereby producing a flat function.
A third control group received forward pairings of
tone with shock during the Pavlovian phase of daily training
identical to the experimental group. However, while in the
experimental group an operant phase immediately followed the
end of Pavlovian conditioning, the control group received a
delay period of varying lengths between the end of Pavlovian
conditioning and the beginning of operant conditioning. This
group was included to determine if responding on an operant
baseline during the retention interval, the period of time
between the end of Pavlovian conditioning and the suppres-
sion test, had any effect on the U-shaped function. If only
the length of time between Pavlovian conditioning and test-
ing is the important determinant of the U-shaped function,
then the gradients in both the experimental group and the

third control group should be similar.

METHOD

Subjects

Twenty-four white Carneaux pigeons were maintained at
80 percent of their free-feeding weight. Each pigeon was
implanted with a stainless steel wire through each side of
its pubic arch. The wires were attached to a phono plug

mounted on the pigeon's back.

Apparatus

 

Two standard 3 key, Lehigh Valley Electronics experi-
mental chambers were used. The center key of each chamber
was illuminated with a black vertical line on a white back-
ground or a green (555 nm) wavelength. The stimuli were
projected onto the key using an Industrial Electronics
Engineers inline projector (model #10-0W78-1820-L). A
houselight, consisting of a GE #1820 lamp, remained on
during sessions. A GE #1820 lamp illuminated the food
hopper during reinforcement. The onset of reinforcement was
controlled by a Lehigh Valley Electronics Photosensor
(model #221-10). A speaker, mounted on the front panel of
the operant chamber was used to present the auditory
stimulus. A B & K sound pressure meter was used to set the
auditory stimulus at 65 db. above the ambient noise level.
The chamber was modified for shock delivery as described by
Klein and Rilling (1974). A high internal resistence‘

6

AC milliampere power supply, insensitive to external resis-
tence changes up to 5000 ohms, was used as the shock source.
A fan, providing masking noise and ventilation, remained on
during sessions. The chambers were maintained in separate
rooms separated by a central room in which programming and

recording equipment was maintained.

Procedure

 

Pretraining

The pigeons were initially magazine trained and auto-
shaped to peck the center key. An autoshaping trial con-
sisted of the illumination of the key with the 555 nm
stimulus for 8 sec followed by access to a grain filled
hopper for 3.5 sec. The autoshaping trials were controlled
by a variable time (VT) 60 sec schedule in which the 555
nm stimulus followed by the access to grain was presented
at variable times with a mean of 2 min independent of any
response by the pigeon. Each session of autoshaping was
60 min in duration and was terminated when a pigeon emitted
100 or more responses to a 555 nm stimulus during the session
for two consecutive sessions.

After autoshaping, the pigeons received one 60 min
session on a variable interval (VI) 30 sec schedule followed
by two 60 min sessions on a VI 60 sec schedule. Next, the
pigeons received two, 60 min sessions of exposure to a 1000
HZ tone. A session consisted of 2 min presentations of the

tone while a VI 60 sec schedule was in effect. The tones

were presented randomly throughout the session according to
a VT 2 min schedule. During all of the sessions in which a
pigeon pecked for reinforcement, the key was illuminated with

the 555 nm stimulus.

Training

Following pretraining, the pigeons were divided into
four groups: (a) a forward pairing (F) group, (b) a forward
pairing with operant delay (FD) group, (c) a backward pair-
ing (B) group, and (d) a truly random (R) group. Each group
consisted of six pigeons.

A training session consisted of two phases, a Pavlovian
conditioning phase and an operant conditioning phase. In
the Pavlovian phase, the pigeons received presentations of
a 1000 HZ tone and a 2 ma shock. In the operant phase, the
pecking response was reinforced on a VI 60 sec schedule.
Grain was never available during the Pavlovian phase and
shocks never occurred during the operant phase.

The onset of the Pavlovian phase was signalled by the
illumination of the houselight. The duration of the 1000
HZ tone was 2 min and the duration of the 2 MA Shock was 1.5
sec. For the forward pairing (F) group and the forward
pairing with operant delay (FD) group, the presentation of
the tone was immediately followed by shock. The intertrial
interval (ITI) was varied following a VT 2 min schedule.

For the backward pairing (B) group, a trial consisted of

the presentation of shock followed, after a 10 sec delay, by

the tone. The tone was followed by a 30 sec safe period
before the next ITI began. A VT 2 min schedule determined
the ITI. For the truly random (R) group, tone and shock
presentations were programmed on independent VT 2 min
schedules. For all groups, the Pavlovian phase ended when
a pigeon had received 15 presentations of the tone and 15
presentations of the shock.

At the end of the Pavlovian phase, the center key was
illuminated with the 555 nm stimulus for all groups except
the forward pairing with operant delay (FD) group. During
the operant phase, responding was reinforced on a VI 60
sec schedule. No tones or shocks occurred during this
phase prior to testing. For the forward pairing with
operant delay (FD) group, a delay between the Pavlovian
and operant phases was imposed. This delay was indicated
by the illumination of the center key with a vertical line
stimulus. The delay period was either 0 min, 15 min, 30
min, 45 min, or 60 min in duration. The delay intervals
were randomized across sessions with each subject receiving
two sessions at each delay interval. During the delay
period, no shocks or tones were presented, nor was reinforce-
ment for responding available. At the end of the delay
interval, the key was illuminated with the 555 nm stimulus
and responding was reinforced on the VI 60 sec schedule.

For all groups, the operant phase was 60 min. Each

session consisted of a Pavlovian phase followed by an operant

10

phase. There were 10 training sessions.

Testing

Testing sessions were identical to training sessions
with one exception. During the operant phase the 1000 HZ
tone was presented without shock for 2 min either 4 min, 15
min, 30 min, 45 min, or 60 min from the beginning of the
phase. The Pavlovian phase preceded the operant phase as
in training. The tone was presented only once per session
for testing during the operant phase. The schedule of
reinforcement remained in effect during the presentation
of the tone. At least 24 hr intervened between testing
sessions. For the forward pairing with operant delay (FD)
group, the delay separating the Pavlovian and operant phases
was terminated 4 min prior to the scheduled tone presenta-
tion. For the 4 min testing interval, no delay occurred.

All groups also received sessions in which the Pav-
lovian phase was omitted. These sessions were always 24
hours following a normal testing session. During these
testing sessions, the tone was presented at one of the test-
ing intervals and all testing intervals were sampled across
pigeons.

The order of testing was randomized within and between
pigeons with all testing intervals sampled across pigeons
per day. The randomized schedule of testing was replicated
three times giving each pigeon four tests at each interval.

Each pigeon received a total of 24 testing sessions.

ll

Suppression ratios were calculated using Kamin's (1961)

formula. The formula is:

T
pre T + T

 

(1)

Where T is the number of responses emitted during the tone
presentation and pre T is the number of responses emitted

in the two minute period preceding the tone presentation.

RESULTS

The results for the forward pairing (F) group are
shown in Figure l. U-shaped functions, indicating maximum
suppression at intermediate intervals, were obtained for 5
of the 6 pigeons in this group. The point of maximum sup-
pression varied for individual pigeons with the maximum for
P6517 and P10535 occurring at 15 min, for P6449 and P5701
at 30 min, and for P5767 at 45 min. For P4435, which did
not produce a U-shaped function, maximum suppression occur-
red at the 4 min interval and suppression monotonically
decreased as a function of the retention interval. The 1440
min interval on the graph represents the suppression obtained
during the session in which the Pavlovian phase was omitted.
For this interval, no systematic changes were observed in
suppression as a function of the placement of the tone
within the session.

The results for the forward pairing with operant
delay (FD) group are shown in Figure 2. U-shaped functions
were obtained for each of the six pigeons with maximum
suppression occurring for P500 and P373 at 15 min, for
P438 and P431 at 30 min, and for P922 and P4879 at 45 min.

A two-factor, mixed design, analysis of variance was
performed on the data from the two forward pairing groups.
The test was conducted to determine if the level of

12

Figure l. The level of response suppression as a function
of the testing interval for individual pigeons
in the forward pairing (F) group.

SUPPRESSION RATIO

 

l3

FORWARD (F) GROUP

P5767

 

 

 

‘ P6449

 

. r A 4 #—
4 IS 30 45 60 I440

P6517

 

4“" """°"“"T"—-" 7"- T-‘rh-

 

 

.1
P4435
O
A //'/N
1"" f I I U 1 4 '
Pl0535
O
J
1 fi+V—‘_

 

 

4 GI 50‘4360 I440

RETENTION INTERVAL (MIN)

Figure 2. The level of response suppression as a function
of the testing interval for individual pigeons
in the forward pairing with operant delay (FD)
group.

SUPPRESSION RATIO

 

l4

FORWARD/DELAY (FD)

 

 

 

 

 

51 ‘\‘\‘\‘/’ o

4 .

.34

.2 4

.l' P922

0 .#4

.5 I .

.4. \//~.

.3.*

.2 4

J P500

0 . 11qu

5I o

44 v.”

3 1

.2 ‘

.I P373
. . . . r4144—
4 I5 30 45 60 I440

RETENTION

'I

 

 

GROUP

\//. .

P438

P4879

 

A

 

P43|

 

INTERVAL

v v v r H—F—
4 I5 30 45 50 I440
(MIN)

15

suppression varied significantly as a function of the reten-
tion interval. The results showed a significant effect of
the retention interval on the level of suppression (F5,50 =
2.81, p < .025). The two groups were also found to differ
significantly in the total amount of suppression over all
testing intervals (Fl,10 = 23.8, p < .001). A comparison

of the two groups showed suppression levels in the forward
pairing (F) group to be significantly higher than suppres-
sion levels in the forward pairing with operant delay (FD)
group (see Figures 1 and 2). The analysis of variance showed
no significant interaction between the group tested and time

dependent changes in suppression (F .55, p > .10).

5,50 =
A trend analysis was performed on the data to deter-
mine what type of function best fit the changes obtained in
suppression as a function of the retention interval. The
results showed that a quadradic, U-shaped, function provided

the best fit (F = 6.35, p < .025). This test indicates

1,50
that the functions obtained in the forward pairing groups
were significant U-shaped functions.

Table 1 shows the average response rates during the
pre-tone period and during the tone for all groups. The
table was derived from averaging within and across pigeons
and suppression ratios cannot be obtained from the data.

A two factor, mixed design, analysis of variance was per-
formed on the response rates during the pre-tone period for

the two forward pairing groups. The test was performed to

determine if any systematic changes in response rate occurred

16

as a function of time. No significant effect of the reten-
tion interval on pre tone responding was obtained (F5,50 =
.18, p > .10). A significant effect, however, was obtained
between groups (Fl,10 = 6.57, p < .05). The group effect
resulted from the higher level of responding in the forward
pairing with operant delay (FD) group as compared to the

forward pairing (F) group.

Table 1

Average Response Rate During the
Tone and Pre-Tone Periods

 

 

 

 

Group Retention Intervals (Min)
4 15 30 45 60 1440

F a

Pre-Tone 72 76 74 72 72 73

Tone 36 28 24 31 28 30
FD

Pre-Tone 111 115 113 110 113 109

Tone 103 95 88 92 108 99
B

Pre-Tone 54 51 51 46 53 48

Tone 53 47 50 47 49 48
R

Pre-Tone 44 57 51 50 50 49

Tone 34 35 30 29 32 45

 

a 0 0 0
Response rate 13 in responses per minute.

The results for the backward pairing (B) group are
shown in Figure 3. Suppression was minimal in this group

and the rate of responding during the tone was roughly

17

constant across retention intervals. All pigeons within
this group show relatively flat functions. The results from
the truly random (R) group are shown in Figure 4. One pigeon
in this group was dropped for failure to respond during the
operant phase. Two of the remaining five pigeons, P4706
and P1110, show no suppression and flat functions. Three
pigeons, P217, P2478, and P1062, show both suppression to
the tone and U-shaped functions. The interval of maximum
suppression for these pigeons varied from 15 min to 45 min.
Thus, when suppression was obtained in the truly random (R)
group, the resulting functions were similar tothose obtained
for the forward pairing groups while when no suppression was
obtained, the functions were similar to those obtained with
the backward pairing (B) group.

A comparison of group averages is presented in Figure
5. Due to the varied results obtained with the truly random
(R) group, its group average is not included in the figure.
For both the forward pairing (F) and the forward pairing
with operant delay (FD) groups, the average point of maximum
suppression occurred at the 30 min retention interval. The
suppression level for the 1440 min retention interval approxi-
mately equaled the amount of suppression at the 60 min reten-
tion interval. A t test between the 60 min interval and the
1440 min interval was found to be nonsignificant for both
the forward pairing (F) group (T5 = .69, p > .10) and the
forward pairing with operant delay (FD) group (T5 = 1.67,

p > .10).

Figure 3. The level of response suppression as a function
of the testing interval for individual pigeons
in the backward pairing (B) group.

SUPPRESSION RATIO

l8

BACKWARD (B) GROUP

 

 

 

 

 

 

 

 

 

 

 

 

41 .
.34 .
.2. I
J‘ P600 ‘ PIGBI
0 . 4 . -41.: f . r . n!
5« M O . w .
.4~ «
.3~
.ZI
‘4 PI936 I PI203
0 . . . T .w r . - . #r
ST W O 1 M O
.41 -
.34
.2- «
.II P4078 4 PI3
.0 . . . . fiv—r . . - f . .fi.
4 I5 30 45 60 I440 4 I5 30 45 60 I440

RETENTION INTERVAL (MIN)

Figure 4. The level of response suppression as a function
of the testing interval for individual pigeons
in the truly random pairing (R) group.

SUPPRESSION RATIO

.4<

.31

.2~

.5~

.4.

 

 

l9

RANDOM
\/ o

PIOGZ

P2478

 

 

P2I7

__lL
if r

4 I3 30 is 63 I440

(R)

 

GROUP

P4706

J1

 

PIIIO

 

4—..—

 

A

I15 50 43 60* I440

RETENTION INTER VAL (MIN)

Figure 5. Group averages showing the level of response
suppression as a function of the testing interval
for the forward pairing (F), forward pairing with
operant delay (FD), and backward pairing (B)
groups.

SUPPRESSION .RATI O

 

20

GROUP AVERAGES

W O
\\ ‘// A
\ /

A
\ \M/ H B GROUP
A—-A FOGROUF
I—-I FGROUP
' I—- 4 '
\ ./
I
\,I/

4 IVS 3‘0 4‘5 6'0" [440
RETENTION INTERVAL (MIN)

DISCUSSION

The U-shaped functions obtained in this study demon-
strate that time dependent changes occur in conditioned
suppression. The phenomenon was reliably produced in eleven
of the twelve pigeons which received a forward pairing of
tone with shock. Although both forward pairing groups pro-
duced U-shaped functions, the forward pairing with operant
delay (FD) group showed significantly less overall suppres-
sion. The reason for this difference is not clear. This
group also showed a significantly higher rate of responding
during the pre-tone period. The higher rate of responding
in this group may represent a "contrast effect" resulting
from the imposition of a delay period between the Pavlovian
and operant phases. The high rate of responding may in
turn, have led to an attenuation of response suppression
during the tone.

The use of a Within-subjects design permitted an
evaluation of time dependent changes in conditioned suppres-
sion within individual pigeons. The results showed that the
point of maximum suppression varied from 15 min to 45 min
across pigeons. Typically, "Kamin effect" studies have not
used within-subjects designs and have not, therefore,

detected this variance in the U-shaped function.

21

22

The minimum of the U-shaped function in the group
averages of the two forward pairing groups occurred at the 30
min retention interval. In "Kamin effect" studies using rats,
the minimum of the U-shaped function has typically occurred
at the one hour retention interval. The difference in the
minimum of the function between the present study and pre-
vious studies may have resulted from a species variable.

The minimum_of the U-shaped function in "Kamin effect"
studies has been found to vary extensively as a function of
the species used (reviewed by Squire, 1975). The experi-
mental design may also have effected the minimum of the
U-shaped function. The present study differed from previous
"Kamin effect" studies by utilizing a within-subjects

design, by using conditioned suppression as the dependent
measure, and by not removing the pigeon from the experimental
chamber during the retention interval. The minimum of the
U-shaped function may be effected by such design variations.

A comparison of the control groups for Pavlovian con-
ditioning shows an absence of conditioning in the backward
pairing (B) group but the presence of conditioning and
U-shaped functions in the truly random pairing (R) group.
Although Rescorla (1972) has predicted that conditioning
may initially occur with the "truly random" control, con-
ditioning should dissipate with continued training. The
pigeons, in the truly random pairing (R) group which showed
conditioning, suppreSsed responding to the tone after over

30 days of Pavlovian conditioning. Therefore, it seems

23

unlikely that the conditioning demonstrated in these pigeons
was a transitory effect resulting from the "truly random"
procedure as suggested by Rescorla (1972). The presence of
suppression in some pigeons and the absence of suppression
in other pigeons in this group may have resulted from dif-
ferences in the total number of forward pairings received.
Some pigeons may have received a greater number of chance
pairings. It seems unlikely however, that there would have
been more then a negligible difference in the total number
of forward pairings between pigeons after over 30 sessions
on the "truly random" schedule. Benedict and Ayres (1972)
have shown that random schedules with forward pairings
occurring at the beginning of training are more likely to
result in conditioning. In their study (Experiment 2) rats
received either initial pairings or initial nonpairings of

a tone with shock under computer generated "truly random"
schedules. The rats were then placed on an operant schedule
for food and the tone presented during barpressing. The
results showed that the rats which received pairings of the
tone with shock during the first 10 percent of training
showed subsequent suppression of barpressing during the tone
while groups which received an equal number of tone-shock
pairings at other times during training showed no subsequent
suppression of barpressing to the tone. In another study,
Ayres, Benedict and Witcher (1975) selectively eliminated

pairings of a tone with shock in "truly random" schedules.

24

They found that the total number of chance pairings did con-
tribute to the presence or absence of conditioning to the
tone. They also found that the absence of conditioning to
the tone was significantly related to the number of shock
presentations prior to the initial chance pairings of the
tone with shock. It seems likely that the difference in
conditioning between pigeons in the truly random pairing
(R) group was the result of both the number of chance pair-
ings which occurred during the early sessions of training
and the number of shock presentations which occurred prior
to the first chance pairing.

The results of the present study indicate that excita-
tory conditioning to the tone is essential for the occurrence
of a U-shaped function. Furthermore, for the present study,
the backward pairing (B) group, which received explicitly
unpaired presentations of the tone with shock, provided the
most appropriate control procedure for Pavlovian conditioning.
In this group, all pigeons showed an absence of suppression
and flat functions. The ”truly random" control procedure
as suggested by Rescorla (1967) was found to result in sup-
pression and U-shaped functions in three of the five pigeons
tested.

In the present study, U-shaped functions were obtained
while in the McMichael (1966) and Tarpy (l966)studies a
monotonic increase in suppression was observed. The numerous
differences in the present design as compared with these

previous studies makes a specific analysis of the differences

25

in results difficult. One factor which may influence
whether a U-shaped function is obtained or not is the use
of different chambers for training and testing, character-
istic of previous studies, as opposed to the use of a single
chamber for both training and testing, as in the present
study. Brush (1970) has suggested that the use of two
chambers in the McMichael (1966) and Tarpy (1966) studies
may have led to problems with stimulus generalization during
testing.

Anisman (1975) has attributed the "Kamin effect" to
.a decrease in the organism's ability to initiate voluntary
responses at intermediate retention intervals. This "re-
sponse inhibition" is said to result from changes in the
levels of various neuro-transmitters produced by the
organism's direct exposure to an aversive US. Under such
a model, it is primarily the response which is inhibited
at intermediate retention intervals with the CS-US associ-
ation playing only a minor role. In the present study,
the operant response should have varied systematically with
the retention interval in the absence of the tone. The
fact that it did not, and the lack of a U-shaped function
in the backward pairing (B) group, indicates that condi-
tioning to the tone was necessary for the production of
a U-shaped function.

Klein and Spear (1970) have attributed the "Kamin

effect" to a memory retrieval failure at intermediate

26

retention intervals. They have suggested that a type of
"state dependent learning-dissociation" effect may be the
mechanism responsible for retrieval failure. The organism
learns the avoidance response under particular internal cues
which are produced by the stress of shock. At intermediate
retention intervals, the internal stimulus conditions which
were present during original learning are no longer present
or they cannot be re-established quickly. Organism's can
acquire a conflicting task at intermediate retention inter-
vals at a rate roughly equivalent to niave animals because
the memory of the original learning task cannot be retrieved
and therefore cannot interfere with the acquisition of a

new task. Under this model, the level of performance is a
function of the amount of transfer between training and
testing. A memory retrieval interpretation of the "Kamin
effect" would predict that, at intermediate retention
intervals, an organism would have difficulty retrieving CS-US
associations (Bryan & Spear, 1976). To be consistent,

the model would also predict that if time dependent changes
occurred in conditioned suppression, a decrease in the level
of suppression would be observed at intermediate retention
intervals. If the pigeons in the present study could not
retrieve the association of the tone with shock, they should
have shown less suppression at the 30 min interval. The
results, however, show the greatest level of suppression at

the intermediate retention intervals.

27

Denny (1958) proposed that the "Kamin effect" resulted
from an incubation of fear over time.) Under this model,
fear reactions to the CS increase as a function of the
retention interval to an asymptotic level. Fear then
begins to dissipate to a base level approximately equal
to the level of fear during initial training. If retraining
of the avoidance response occurs when fear is at a maximum
level, the animal reacts to the CS and the situational cues
with fear reactions, such as freezing (Stein, Hoffman, &
Stitt, 1971), which are incompatible with the appropriate
avoidance response. The idea that freezing may be one of
the behaviors which interferes with the active avoidance
response has also been suggested by Anisman (1975). In
the present study, maximum suppression occurred at inter-
mediate retention intervals. Numerous theorist (Estes,
1969; Konorski, 1967; Mowrer, 1960; Rescorla & Solomon,
1967) have proposed that conditioned suppression reflects
the conditioning of central motivational states. Under
this position, and by labeling the central motivational
state resulting from aversive conditioning as "fear," the
present results can be considered to represent an "incubation
of fear" effect. Regardless of whether or not one accepts
conditioned suppression as a measure of "fear," the present
results show changes in a pigeon's reaction to a tone paired
with shock which are consistent with the Denny (1958)

interpretation of the "Kamin effect."

BIBLIOGRAPHY

BIBLIOGRAPHY

Anisman, H. Time-dependent variations in aversively moti-
vated behaviors: Nonassociative effects of Cholin-
ergic and Catecholaminergic activity. Psychological
Review, 1975, 88, 359-385.

 

Ayres, J.J.B., Benedict, J.O. & Witcher, E.S. Systematic
manipulation of individual events in a truly random
control in rats. Journal of Comparative and Physio-
logical Psychology: 1975, 88, 97-103.

 

Benedict, J.O., & Ayres, J.B. Factors affecting conditioning
in the truly random control procedure in the rat.
Journal of Comparative and Physiological Psychology,
1972:.l§r 323-330.

Brush, R.F. Retention of aversively motivated behavior.
In Brush, R.F. (Ed.), Aversive Conditioning and
Learning. New York: Academic Press, 1971, pp. 402-
465.

 

Bryan, R. G., & Spear, N.E. Forgetting of a discrimination
after intervals of intermediate length: The Kamin
effect with choice behavior. Journal of Experimental
Psychology: Animal Behavior Processes, 1976, 8,
221-234.

 

Denny, M.R. The "Kamin effect" in avoidance conditioning.
American Psychologist, 1958, 18, 419.

 

Estes, W. K. Outline of a theory of punishment. In Campbell,
B.A., & Church, R.M. (Eds.), Punishment and Aversive
Behavior. New York: Appleton, Century, Crofts, 1969,
pp. 57-82.

 

Hunt, H.F., & Brady, J.V. Some effects of electroconvulsive
shock on a conditioned emotional response (anxiety).
Journal of Comparative and Physiological Psychology,
1951, 88, 88-89.

 

Kamin, L. J. Retention of an incompletely learned avoidance
response. Journal of Comparative and Physiological
Psychology: 1957, 88, 457-460.

 

 

28

29

Kamin, L. J. Trace conditioning of the conditioned emotional
response. Journal of Comparative and Physiological
Psychology, 1961, 88, 149-153.

 

 

Klein, M., & Rilling, M. Generalization of free-operant
avoidance behavior in pigeons. Journal of the Experi-
mental Analysis of Behavior, 1974, 81, 75-88.

 

 

Klein, S. B., & Spear, N.E. Forgetting by the rat after
intermediate intervals (Kamin effect) as retention
failure. Journal of Comparative and Physiological
Psychology, 1970, 11, 165-170.

 

 

Klein, S.B., & Spear, N.E. Sequential variables and the
Kamin effect. Learning and Motivation, 1973, 8,
357—365.

 

Konorski, J. Integrative Activity of the Brain. Chicago:
University of Chicago Press, 1967.

 

Kremer, E. F., & Kamin, L. J. The truly random control
procedure: Associative or nonassociative effects in
the rat. Journal of Comparative and Physiological
Psychology, 1974, 18, 203-210.

 

 

McMichael, J.S. Incubation of anxiety and instrumental
behavior. Journal of Comparative and Physiological
Psychology, 1966, 8;, 208-211.

 

 

Mowrer, O. H. Learning Theory and Behavior. New York:
Wiley, 1960.

 

Quinsey, V. L. Conditioned suppression with no CS-US
contingency in the rat. Canadian Journal of
Psychology, 1971, 88, 69-82.

 

 

Rescorla, R.A. Conditioned inhibition of fear. In Mackintosh,
N.J., & Honig, W.K. (Eds.), Fundamental Issues in
Associative Learning: Proceedings of a Symposium
held at Dalhousie University, Halifax, June, 1968.
Halifax: Dalhousie University Press, 1969.

 

 

 

Rescorla, R.A. Informational variables in Pavlovian condi-
tioning. In Bower, G. H. (Ed.), The Psychology of
Learning and Motivation: Advances in Research and
Theory (V01. 6). New York: Academic Press, 1972.

 

 

Rescorla, R.A., & Solomon, R. L. Two-process learning
theory: Relationship between Pavlovian conditioning
and instrumental learning. Psychological Review,
1967, 18, 151-182.

 

30

Squire, L.R. Short term memory as a biological entity.
In Deutsch, D., & Deutsch, J.A. (Eds.), Short-Term
Memory. New York: Academic Press, 1975.

 

Stein, N., Hoffman, H.S., & Stitt, C. Collateral behavior
of the pigeon during conditioned suppression of key-
pecking. Journal of the Experimental Analysis of
Behavior, 1971, 88, 83-93.

Tarpy, R. Incumbation of anxiety as measured by response
suppression. Psychonomic Science, 1966, 8, 189-190.