Title: Maintenance and suppression of responding under concurrent schedules of electric-shock presentation
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00097560/00001
 Material Information
Title: Maintenance and suppression of responding under concurrent schedules of electric-shock presentation
Physical Description: x, 60 leaves. : illus. ; 28 cm.
Language: English
Creator: Webbe, Frank Michael, 1947-
Publication Date: 1974
Copyright Date: 1974
 Subjects
Subject: Electric shock   ( lcsh )
Reinforcement (Psychology)   ( lcsh )
Psychology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis -- University of Florida.
Bibliography: Bibliography: leaves 57-59.
Additional Physical Form: Also available on World Wide Web
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00097560
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000582512
oclc - 14118997
notis - ADB0887

Downloads

This item has the following downloads:

PDF ( 2 MBs ) ( PDF )


Full Text











MAINTENANCE AND SUPPRESSION OF RESPONDING
UNDER CONCURRENT SCHEDULES OF ELECTRIC-
SHOCK PRESENTATION

















By

FRANK MICHAEL NEBBE


A DISSERTATION PRESENTED TO TIIE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IMPARTIAL FULFILLMENT OF THE REQUIRE'-CNTS FOR, TiF
DEGREE OF DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA


1974















ACKNOWLEDGMENTS

The doctoral dissertation is a milestone in the winding path of

inquiry, learning, and scholarly achievement. As such, it represents

not only the labor of the person whose name it beare, but also che

influence, assistance, and guidance of other persons who have formed

the environment in which it has reached fruition.

To my parents, I owe a debt which can be recognized but ever

repayed. TlIrough them I came to life, and through them 1 have, been

provided with the spiritual and material guidance mad support that hi.s

allowed me to live that life fully. Words can only incompletely

express my love and gratitude to them.

To Professor Edward F. Malagodi, I owe a debt of intellectual

stimulation and scholarly guidance. As a renaissance man, eclectic iin

the extreme, he has both broadened the scope of my inquiry, and enabled

me to channel that broadness in the development and exploration of new

areas of endeavor.

To Dr. Jo DeWeese, I owe a debt of friendship, both personal and

professional. Her encouragement helped me to maintain high levels of

academic accomplishment, and her presence allowed the shared exper-

iences of discovery and elation, disappointment and heartbreak.

To my professional colleagues and mentors, Professors Marc Branch,

Bob Isaacson, Merle Meyer, Hank Pennypacker, Carol YanHartesveldt,

Dick Willis, and Bill Wolking, I express my gratitude for innu-berable









insights into the way things are and may be, and for their assistance

in overcoming many of the obstacles which have hindered my progress

at one time or another.

To Elayne Insler, I am indebted for the many hours spent assisting

me in the collection of the data that are reported here. Her cheerful

aid made the process much more enjoyable than it might otherwise have

been.

And finally, to my wife, Ellen, I owe a debt of love for her

constant support through more than four years of living a hand to mouth

existence. At times when I was ready to surrender in the face of

seemingly insurmountable obstacles, her love, encouragement, and

unselfishness provided me with the strength to "kick against the goads."

They have now been kicked against thoroughly, and her repayment shall

begin directly.















PREFACE

The experiments reported here represent the combined endeavor of

persons involved in the operations of the Operant Behavior Laboratory

at the University of Florida. As my name on the title indicates, I

have borne the daily responsibility of experimentation and supervision,

as well as the onus of textual composition. However, the experiments

and many of the ideas reported here were germinated in an atmosphere

of scholarly and experimental inquiry of which I was a part.

The science of behavior has passed the point wherein any one per-

son bears full responsibility for the creation and implementation

of experimental ideas. Without the work of previous investigators,

and without the individual contributions of those who shared in this

study, the endeavor never would have been accomplished.















TABLE OF CONTENTS


Page

ACKNOWLEDGMENTS ii

PREFACE iv

LIST OF TABLES vi

LIST OF FIGURES vii

ABSTRACT viii

INTRODUCTION 1

EXPERIMENT I 3

Method 3
Results 6
Discussion 9

EXPERIMENT II 18

Method 20
Results 25
Discussion 34

OVERVIEW 39

EXPERIMENT III 43

Method 45
Results 46
Discussion 52

REFERENCES 57

BIOGRAPHICAL SKETCH 60

















LIST OF TABLES


Table 1. The Order of Experimental Conditions
for Monkey SM-38 23

Table 2. The Order of Experimental Conditions
for Monkey SM-432 26

Table 3. Mean Response and Shock Rates for
Monkey SM-38 32

Table 4. Mean Response and Shock Rates for
Monkey SM-432 33

Table 5. Relative Measures of Responses and
Shocks 37
















LIST OF FIGURES


Page

Figure 1. Representative Cumulative Records
for Monkey SM-38 under Avoidance
and VI 3 8

Figure 2. Response Rates and Shock Rates for
Monkey SA-38 during 70 Sessicns of
Avoidance and VI 3 11

Figure 3. Response Rates and Shock Rates for
Monkey SM-432 during 70 Sessions of
Avoidance and VI 3 13

Figure 4. Representative Cumulative Records
for Monkey SM-38 during Concurrent
VI, VI schedules 29

Figure 5. Representative Cumulative Records
for Monkey SM-432 during Concurrent
VI, VI schedules 31

Figure 6. Response Rates and Shock Rates for
Monkey SM-38 during Concurrent VI,
EXT schedules 48

Figure 7. Response Rates and Shock Rates for
Monkey SM-432 during Concurrent VI,
EXT schedules 50

Figure 8. Representative Cumulative Records
for Monkey SM-432 Showing Transition
from Response Maintenance to Response
Suppression under Concurrent VI, EXT
Schedules 54















Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy



MAINTENANCE AND SUPPRESSION OF RESPONDING
UNDER CONCURRENT SCHEDULES OF ELECT.IC-
SHOCK PRESENTATION

By

Frank Michael Webbe

August, 1974

Chairman: E. F. Malagodi
Major Department: Psychology

Following a variety of experimental histories, squiirel monkeys

continue to respond under conditions in which schedules of electric-

shock presentation remain as the only progranamed consequences of re--

sponding. Ihe rates and patterns of responding arc similar to those

that ordinarily characterize responding under similar schedules of food

or water presentation.

The experiments reported here were concerned wlth investigating

further the conditions under which this phenomenon might occur.

Specifically, squirrel monkeys were trained first to respond under

variable-interval (VI) schedules of shock presentation. Subsequently,

concurrent variable-interval schedil.es were introduced. That is, the

experimental chamber contained two levers, and schedules of shock pres-

entation were programmed for respoflding on eech lever.

Two basic procedures were employed for developing concurrent

responding. In one procedure, the response lever wr. either in the








right or the left hand position during each of the last 60 sessions

under a 3-min variable-interval (VI 3) schedule of shock presentation.

Subsequently, two levers were inserted into the chamber at the same

time, and equal and independent VI 3 schedules were programmed for

responding on each lever. A bi-directional changeover delay was pro-

grammed such that changeover responses initiated a 0.50-sec interval

during which programmed shocks could not be delivered.

In the second basic procedure, concurrent avoidance schedules

were introduced following 107 sessions under a VI 3 schedule of shock

presentation. Two levers were inserted into the chamber, and concurrent

and independent 5-sec shock shock intervals were programmed, as were

concurrent and independent 30-sec response shock intervals. Con-

current variable-interval schedules were introduced 30 sessions later.

After exposure to one of these training histories, the two monkeys

continued to respond under several parameter values of the concurrent

VI, VI schedules of electric-shock presentation. Both monkeys

responded at higher rates on one lever than on the other. Responding

on each lever was constant in rate, similar to the patterns usually

maintained by comparable schedules of food or water presentation.

Following a minimum of 70 sessions under concurrent VI, VI

schedules, the schedule for one of the levers was changed to extinction;

that is, programmed shocks were no longer delivered as a consequence of

responding on that lever. The VI 3 schedule remained in effect on tlhe

other lever. Under these conditions, responding on both levers

decreased to zero rates. Similar results were obtained with both of

the monkeys in two separate phases in which extinction was programmed









first on the left lever and then on the right lever.

In suunary, two separate operants were acquired and subsequently

maintained under VI, and concurrent VI schedules of electric-shock

presentation. The rates and patterns of responding were comparable

to those observed under similar schedules of food or water presentation.

lfen one of the concurrent operants no longer produced shocks,

responding decreased to zero rates, first on the lever for which

shocks were still programmed, and subsequently on the lever for which

no consequence was programmed. These results are exactly opposite to

the effects of a comparable manipulation during concurrent responding

under schedules of food or water presentation. The results indicate

the necessity of further study of the response-produced shock

phenomenon in complex situations. The generality of the phenomenon

may be limited to situations providing no alternative responses

with qualitatively different consequences.














INTRODUCTION

Since 1966, more than 10 studies have reported on the development

and chronic maintenance of responding under schedules of electric-shock

presentation. After initial histories of responding under schedules of

food presentation, shock postponement, or shock termination, the

occasional presentation of intense electric shocks remained as the only

programmed consequences of responding. These studies have progressed

in complexity from "simple" demonstrations of the phenomenon under

fixed-interval (FI) schedules with squirrel monkeys (Kelleher and

Morse, 1968; McKearney, 1968; Stretch, Orloff, and Dalrymple, 1968)

and with cats (Byrd, 1969), to maintenance conditions under fixed-ratio

(FR) schedules (McKearney, 1970) and variable-interval (VI) schedules

(McKearney, 1972). More complex behavioral processes have also been

investigated using this phenomenon, including extinction and recovery

of responding (Kelleher and Morse, 1968; McKearney, 1969), the effects

of interrupting behavior maintained under FI schedules by alternating

SD and S-delta periods (Scretch, Orloff, and Gerber, 1970), stimulus

and schedule control during multiple schedules (McKearney, 1970, 1972),

the establishment cf second-order schedules of brief stimulus presenta-

tions (Byrd, 1972), examination of the effects of the Estes-Skinner

procedure (McKearney, 1972), and the generation of schedule-induced

hose-biting during FI schedules (Malagodi, De'eese, Webbe, iad Palenno,

1973).

In tach of these reports, the patterns of behavior nsintained









under schedules of shock presentation (other than FR 1) have been

comparable to the patterns maintained under similar schedules of food

or water presentation. By increasing the complexity of the experimental

situation, each succeeding study has allowed for the further general-

ization of the conditions under which the presentation of response-

dependent electric shocks will maintain behavior.

The series of experiments reported here were similarly concerned

with extending the generality of conditions of behavior maintenance

by response-dependent shock presentations. The experiments proceeded

in three stages. First was the development and chronic maintenance

of behavior under variable-interval schedules of electric-shock

presentation. The second stage was the development of concurrent

operants, each maintained by independent variable-interval schedules

cf shock presentation. The third stage was an examination of the

effects upon rates and patterns of concurrent responding of changing

one of the concurrent schedules specifically, changing from a VI

schedule to extinction.















EXPERIMENT I. DEVELOPMENT AND MAINTENANCE OF RESPONDING UNDER
VARIABLE-INTERVAL SCHEDULES OF ELECTRIC-SHOCK PRESENTATION


McKearney (1972) reported the development and chronic maintenance

of responding in two monkeys under variable-interval (VI) schedules of

electric-shock presentation. The two monkeys had histories of exposure

to shock-postponement and FI schedules of shock presentation. Seven

sessions of shock-postponement training immediately preceded intro-

duction of a 3-min variable-interval (VI 3) schedule of shock presen-

tation. After a total of 33 sessions under VI 3, response rates had

stabilized at higher values than were occurring for the sessions of

shock postponement. Similar poeterns of responding occurred during

shock postponement and VI 3 conditions. That is, the steady rates

characteristic of the two types of schedules were maintained throughout

each session.

Experiment I was essentially a direct replication of McKearney's

procedures. The only significant methodological deiaarture in the

present study was the alternation of the lever position (either on the

left or right side of the front wall) from session to session. This

manipulation was performed in anticipation of developing concurrent

responding in Experiment II,

IMethod

Subjects

Two adult, male, jungle-born squirrel monkeys (Saimiri sclureus)

were used. The monkeys werv housed in individual cges with food and






4

water continuously available. Both monkeys were maintained at their

free-feeding weights, about 1,000 g. A 12-hr light-dark cycle was

maintained in the colony room, and experimental sessions occurred during

the first quarter of the light cycle. The two monkeys had histories of

responding under shock postponement and Fl schedules of shock presenta-

tion.

Apparatus

A Plexiglas chair similar to that described by lake and Azrin (1963)

was used. The monkeys were restrained only at the waist, with no

restrictions placed upon movements of the upper torso, arms, neck, and

head. The tail was confined in a non-conductive brass clamp and Plexi-

glas stock. Electric shock (6 mA for 100 msec) was delivered by a BRS-

Foringer ac shock generator (BRS SG-901) to two brass electrodes resting

on a, shaved portion of the tail. A non-corrosive electrode cream

(Grass, Inc., EC-2) was applied to the tail prior to each session to

maintain a low electrical resistance between the electrodes and the

tail.

The front wall of the chamber could be fitted with one or two Le-

high Valley rat levers (LVE #1352). The lever receptacles were located

on the left and right side of the front wall, 6 cm above the waistplate

and 8 cm apart. Lever depressions with a force in e-cess of 20 g

(0.20 N) produced a loud click from a feedback relay mounted behind the

front wall. Two yellow 7-w stimulus laps and a 4-ohm speaker were

mounted at the top of the front wall, 42.5 cm above the waistplate.

The entire experimental chamber was housed within a ventilated,

sound-attenuating exterior chamber. A Plexiglas vie:.jng port in the

exterior ch.ibcr allowed continuous observations of experimental





5

sessions via closed-circuit television equipment. Standard electro-

mechanical scheduling and recording devices were located in an adjoining

room.

Procedure

The two monkeys, SM-38 and SM-432, had previously responded under

FI schedules of electric-shock presentation for more than 150 sessions.

During these sessions the response lever was always located on the left

side of the front wall.

The first experimental condition of the present study was exposure

to shock-postponement schedules (Sidman, 1953). In the presence of

white noise and illumination of the yellow stimulus lamps, shocks were

scheduled to occur every 5 sec (S S = 5 sec). Each effective lever

press postponed the next scheduled shock for 20 sec (R S = 20 sec).

With Monkey SM-38, the lever was located in the left position, while with

Monkey SM-432, the lever position (right or left) was alternated random-

ly from session to session.

After 15 days of training, the shock-avoidance procedure was

discontinued, and a 3-min variable-interval (VI 3) schedule of shock

presentation was introduced. Under this condition, shocks no longer

occurred in the absence of responding, but rather, the first response af-

ter an average of 3 min produced shocks. The film tapes scheduling the

times when shocks could be produced were derived from Fleshler and Hoff-

man's (1962) equal-probability distribution of intervals. Individual

intervals ranged from 5 sec to 12 min.

With both monkeys, the lever position was changed randomly to

either the left or right side of the front wall during the 60 sessions

of exposure to VT 3. Experimental sessions were conducted 6 days per









week, lasted for 60 min, and were preceded and terminated by 4-min

periods of timeout (TO). The yellow lights and white noise were off

during TO, and a low frequency clicking noise was present. Lever press-

es during TO had no programmed consequences, and shocks were never pre--

sented.

Results

Figure 1 shows cumulative records from the last session of shock

avoidance, and from the 5th and 55th sessions of VI 3 for Monkey SM-38.

Avoidance responding was characterized by steady rates at about 10 per

min as is shown in Record A. Typically, 1 to 6 shocks per hour were

delivered. Introduction of the VI 3 schedule of shock presentation

resulted in an immediate increase in response rates to more than 20 per

nin, and in increase in shock rates to an average of 20 per hour

(Record B). Characteristic steady rates of responding were maintained

throughout the sessions. By the end of the 60-session phase, response

rates had increased to about 30 per min, as is depicted in Record C.

Similar results were obtained with Monkey SM-432. Avoidance

responding was maintained at steady rates of 10 responses per min, with

1 or 2 shocks per hour delivered. Introduction of the VI 3 schedule

resulted in marked increases in response rates, with an average of 20

shocks per hour being produced. Constant rates of more than 30 re-

sponses per min characterized performance after 60 sessions.

Effects of Changing_ Lever Position

Figures 2 and 3 show response rates and shock rates for Monkeys

SM--38 and SM--432, respectively, during the last 10 sessions of avoidance

and during the 60 sessions of VI 3. With SM-38, che leve- was in the










































Figure 1. Representative cumulative records for Monkey SM-38

showing final performance under the avoidance schedule, and

early and late stages in the development of responding under

the VI 3 schedule of shock presentation.





0 M -53


SESSION 15


AVOIDANCE


B EARLY VI 3 SESSION 20


. /l









left position during all sessions of avoidance and during the first 10

sessions of VI 3. The initial effects of changing to the right position

were overall decreases in response rates. Additionally, response rates

were lower when the lever was on the right side than when it was on

the left side. These differential rates persisted for about 9

sessions. Gradually, over a period of about 40 sessions, response

rates increased from less than 10 per min to more than 25 per min.

During the last several sessions response rates were higher when the

lever was on the right side than when it was on the left side.

With SM-432, the lever position was alternated randomly from

side to side beginning with the 4th session of avoidance. Approximate-

ly equivalent rates were obtained regardless of lever position during

all sessions of avoidance and the first 50 sessions of VI 3. Higher

response rates occurred when the lever was on the right side during

the last 10 sessions of VI 3.

Discussion

Characteristic steady rates of responding were maintained by the

VI schedules of electric-shock presentation. In all respects, the be-

havior was similar to that reported by McKearney (1972). The chronic

maintenance of characteristic response patterns with VI schedules of

shock presentation also complements results obtained with FI (e.g.,

Kelleher and Morse, 1968) and FR schedules (McKearcvy, 1970).

Eubanks, Killeen, and Hamilton (1973) reported a failure to main-

tain lever pressing in squirrel monkeys under VI schedules of electric-

shock presentation when each shock was not follower by a period of

timeout from the experjnienta. coe.tingoucic;. 'The betiavior was weil

maintained when 10- to 150-sec TOs followed uach saock under VI 2.































M )0
C I 4J 1
o a a

o ar u -
S 4 C0 4O

'H 4H
co Co Co 0 Cf
I > H 0) 0
Z H C rHI


Co f a
a)' S "
C i 0 C

u Lo o
Ea 0 o 0 4


Q Co j 0
13 Co Co 'H H
a0 cO 0 0 C




Co In -c
wo ri 44
C o Co w H u
o m 1 Co





'U 4 13 'H Co
Co Co H
Cu Co

M 0 0 U S 4
o 0U u

0 > 0)

-H 0
Co m


o C) 'C 'H

0 *'H 4 0) 0C



M Co 41

13 0 13 'H U
3 U Co 0o 'Hl
to 44 Co '0
'H 0 C C Co
f N *H m o H





























o o










000
0 0 000

LL- a_-








o ( CD 0
'NI' d3d "SdS3A


UI H3d SO30!US
So










CC
o
O




0




3 oI_


0
---4






























*H I l

I H (U 0 0
U ) a G) >
l *H -H u)
S Cu *
--4
-t Cu Cc -a c



> 0 4H H


0 0 ) 0 JC

C I 4 0

o n & c


ca O 0 60
H0 a C
> Cu Cu





W 4J .C
S0 C -1
C 3 0 O *-H



'I -H H G)
o Co 4




(f Cu 4 C
C H Q ) -H Cu





O Lo
Cu 0 C 0 *)
Cu 0 4J
u) cu in u ca
Cu O > u0


I o Cu H 4 u








r U C C 4-A
Cu C Cu C

to 4- c Cu
Cti N- -H Cu H





















































I- 0o LO LO
Co (N
T'II' i3d *SdS34


UH H3d
C
(N


S 0 DHS
,


0




0,









Responding rapidly decreased to low rates and was not maintained

when the TOs were discontinued. In comparing their results to chose

obtained by McKearney (1972), Eubanks et al. suggested that differ-

ences in the method of derivation of the VI schedules might have

accounted for the disparity in results. McKearney used an arithmetic

distribution of intervals, while Eubanks et al. used a constant-prob-

ability distribution. Although each interval occurs with a fixed rel-

ative frequency in both types of distributions, the increasing proba-

bility of shocks as a function of the passage of time since the last

shock is more readily discriminable in an arithmetic distribution

(cf., Fleshler and Hoffman, 1962). Moreover, the probability of

shocks following shocks closely in time is lower with an arithmetic

distribution than with an equal- or constant-probatility distribution.

Accordingly, Eubanks et al. raised the possibility that an implicit

TO following shocks might have been present in McK-arney's experiment.

This possibility, in combination with their own failure to maintain

behavior without TOs (using a constant-probability distribution), led

Eubanks et al. to suggest that timeout following shock may be a

necessary condition for the maintenance of behavior under VI schedules

of electric-shock presentation. However, in the present study, a

constant-probability distribution of intervals was always used, TOs

were never scheduled following shocks, and the behavior was well main-

tained.

The present results with respect to changing lever positions may

bear upon the failure to maintain -behavior that was reported by

Eubanks et al. With Monkey SM-432, the new lever position was intro-

duced before the introduction of the VI schedule of shock presentation.









Even in the early sessions of VI 3, characteristic rates and patterns

of responding occurred regardless of the lever position. With Monkey

SM-38, however, introducing the new lever position during the llth

session of VI 3 resulted in marked decreases in response rates.

Additionally, the steady-rate patterns of responding were disrupted,

such that oscillations between high and low response rates occurred.

This disruption in rates and patterns of responding was not limited

to those sessions when the lever was in the new (right) position, but

also occurred during sessions when the lever was it the original

(left) position. The behavior gradually recovered from low to high

maintained rates of responding over a number of seEsions.

The effects of disrupting behavior maintained by schedules of shock

presentation has been discussed previously by Morse and Kelleher

(1970). The behavior is characterized by the interaction of stereo-

typed, ongoing behavior patterns with the schedule of shock presen-

tation. Any intervention which disrupts the patterning results in a

change in the temporal relationships and dynamic interactions between

responding and shock presentation. Depending upon the severity of the

disruption, similar, or new and different stable patterns may develop

and be maintained. For example, in the present experiment, the

steady, moderate rates of avoidance responding interacted with the VI

schedule of shock presentation such that the patterns of responding

remained the same, although response rates increased. The monkeys

responded at the same steady rates both before and after shock presen-

tations. The introduction of the novel lever position resulted in

a decreased race of responding for the one monkey, bur the patterns

of responding cere not altered drastically. Thus, the interactions









between the behavior and shocks remained approximately the same before

and after the change in lever position. Consider, however, the effects

of an intervening treatment which results in an actual cessation of

responding during a VI schedule of shock presentation. For example,

a loose electrical connection in the stimulus lamp panel causes the

lights to flicker. The monkey may orient towards the lamps and stop

responding for a time. If a shock has been scheduled during the time

that the monkey has stopped responding, the next response will produce

the shock. The interaction between responding and shock presentation

has been altered. Previously, shocks occurred during steady, moderate

rates of responding. After the stimulus lamp malfunction, a response

produced a shock following a pause. This new interaction is similar to

an FR 1 schedule of shock presentation, and may result in further

pausing and the eventual suppression of behavior (cf., Kelleher ar.d

Morse, 1968; McKearney, 1969).

The term metastability has been used by Staddon (1965) and Morse

and Kelleher (1966, 1970) to describe "two different stable patterns

of responding maintained under the same schedule parameters, one

before and one after an intervening treatment" (Morse and Keileher,

1970, p. 161). Metastability is a phenomenon associated with any

behavior maintained by its consequences in the environment. However.

it is most clearly seen in situations in which the maintenance of

behavior also depends critically upon prior history and ongoing

response patterns such as behavior maintained under schedules of

shock presentation.

These considerations of the conditions maintacir.in behavior under




17




schedules of shock presentation. mid the modification of these

conditions by disrupting stimuli, suggest that the failure to maintain

VI behavior reported by Eubanks et al. (1973) may have been a

result of the behavioral disruption accompanying the abrupt removal

of TOs. Additionally, since VI behavior was maintained in the

present study using constant-probability distributions of intervals

and with no TOs following shocks, it appears that the presence of

TOs following shocks is not a necessary condition for the maintenance

of responding under VI schedules of shock presentation.















EXPERIMENT II. DEVELOPMENT AND CHRONIC MAINTENANCE OF RESPONDING
UNDER CONCURRENT VARIABLE-INTERVAL, VARIABLE-INTERVAL
SCHEDULES OF ELECTRIC-SHOCK PRESENTATION


In progressing towards the formulation of laws, science typically

proceeds from the observation and manipulation of simple and isolated

phenomena to increasingly complex experimentation. In the field of the

Experimental Analysis of Behavior, for example, the initial focus of

study is usually the thorough scientific understanding of variables

controlling the emission of isolated operant responses. However, as

more variables are identified, and experimental control becomes more

precise, further experimentation is directed towards more complex

occurrences of the response. The initial observations of the response

in isolation are dictated by experimental exigencies: it is easier and

more economical to identify major controlling variables when the

response occurs in isolation from other responses. Temporarily, the

response is studied much as if it were a discrete event. However, the

heuristic value of this approach dwindles as the initial goals of

identification of major variables are realized. There comes a point

when the response must again be studied in the setting from which it

was abstracted, that is, as one vicissitude in the continuous process

of the behavior of an organism.

In the expanded study of the interaction of the response with other

concurrently available responses, experimental attention focuses upon

identification of controlling variables arising from the complex








situation itself. In many instances, studies of complex occurrences

of the response result in the discovery of major independent variables

which are not readily apparent in unelaborated situations (cf.

Findley, 1962).

Previous studies (including Experiment I of this paper) showed

that occasional presentations of response-dependent electric shocks

maintained schedule-appropriate patterns of lever pressing. Experi-

ment II was designed essentially to ask four questions concerning

the development and maintenance of this behavior in a more complex

situation. First, would monkeys press each of two levers when the

only consequences of lever pressing were the occasional presentations

of electric shocks. Second, if the monkeys did initiate responding

on each lever, would they continue to respond, or would the responding

be-suppressed. Third, if the monkeys continued to respond, what types

of patterns and what rates of responding would develop and be main-

cained. Fourth, could the interactions between the two operants be

predicted on the basis of previous studies of concurrent operants

maintained under schedules of food or water presentation.









Method

Subjects

Monkeys SM-38 and SM-432 were used.

Apparatus

The basic apparatus described in Experiment I was used. Plexi-

glas plates, designed to restrict lateral and vertical movements

of the monkeys, were constructed. Their use is described in the

following section delineating experimental procedures.

Procedure

Following the manipulations described in Experji.nt I, the two

monkeys were responding under a VI 3 schedule of electric-shock presen-

tation. The response lever was randomly alternated between the left

and right positions from session to session. After 60 sessions, charac-

teristic VI patterns of responding were being maintained at rates be-

tween 20 and 40 responses per min.

Development of concurrent responding with SM-38

With SM-38, two levers were inserted into the chamber at the same

time following the 60 sessions under VI 3. The levers occupied the left

and right positions in which they had previously been alternated.

Concurrent 3-min variable-interval schedules of electric-shock pres-

entation were programmed as consequences of responding on each lever.

That is, responses on the left lever resulted in the presentation of

shocks once every three minutes, on the average. Simultaneously,

responses on the right lever also resulted in shock presentation accord-

ing to an equal and independent VI 3 schedule. Folljoing established

procedures, a bi-directional changeover delay (COD) was prograner.d.

This protective contingency ensured that changeover responses (e.g.,









a response on the left lever following a response on the right lever)

were never followed immediately by shocks. This procedure specified

that a changeover or switching response initiated a 0.50-sec interval

during which scheduled shocks could not be delivered. The first

response on the same lever to which the monkey had switched produced

the scheduled shock, providing that the interval had elapsed.

Responses on the alternate lever initiated another 0.50-sec interval.

Within three sessions after the introduction of concurrent

schedules, it was apparent that the monkey's topography of responding

was not conducive toward the development of concurrent responding.

The monkey oriented toward a corner of the chamber, bending down and

resting his head on the waistplate. This same orientation had occurred

previously during avoidance and single-lever VI 3. In this position,

the monkey could reach only one lever at a time, responded exclusively

on that one lever (sometimes tne left and sometimes the right), and

did not look at the other lever. In an attempt to develop a topography

in which the monkey oriented towards both levers at the same time, and

which facilitated more frequent changeovers from one lever to the other,

the chamber was modified in the following way. Two large Plexiglas

plates were mounted on each side of the monkey to restrict lateral

movements. Another Plexiglas plate was mounted between the two

side plates to prevent the monkey from lowering his head below the

levers. This plate extended up from thp middle of the waistplate in

front of the monkey, and then extended horizontally until it reached

the front wall, just below the two levers.

The introduction of the restraint plates severely disrupted the

monkey's behavior, suc hint few responses occurred on either lever.










In order to facilitate the development of responding with the restraints

in position, the concurrent VI schedule was discontinued, one Lever

was removed from the chamber, and the shock-avoidance schedule was

reintroduced. During 24 sessions of avoidance, with the lever position

alternating from session to session, steady response rates of about 10

per min were obtained. Introduction of the VI 3 schedule of shock

presentation resulted in significant response rate increases. After

20 sessions of VI 3, the monkey was responding with rates and patterns

similar to those observed in Experiment I, but with the new restraints

in place.

Two levers were again introduced into the chamber at the same

time, and concurrent VI 3, VI 3 schedules of shock presentation

were programmed. The 0.50-sec COD was in effect. The parameters of

the concurrent schedule was subsequently changed to cone VI 3 (left),

VI 1.5 (right) and then to cone VI 1.5 (left), VI 1.5 (right). The

order of experimental conditions and the number of sessions of

exposure to each procedure are presented in Table 1.

Development of concurrent responding with Monkey SM-432

A different procedure was used for developing concurrent responding

with Monkey SM-432. The two Plexiglas side restraints were first

introduced while the VI 3 schedule was still programmed. The presence

of the restraints had little effect upon rates or patterns of respond-

ing when the lever w-s in the left position. However, responding on

the right lever was disrupted, and response rates decreased to less

than 5 per min after several sessions. In order to facilitate the

redavelopneint of responding when the lever was in .th right position,















Table 1

The Order of Experimental Conditions and the Number of Sessions

at Each Condition for Monkey SM-38, during the Development and

Maintenance of Responding under Concurrent VI, VI Schedules

of Electric-Shock Presentation.



Condition Number of Sessions

Cone VI 3, VI 3 3
(restraints introduced)

Avoidance 24
(lever position alternating)

VI 3: shock 20
(lever position alternating)

Cone VI 3, VI 3 25

Cone VI 3 (left), VI 1.5 (right) 15

Cone VI 1.5, VI 1.5 30









the shock-avoidance schedule was programmed for sessions when the

lever was in that position. The VI 3 schedule of shock presentation

continued in effect when the lever was on the left side. After

9 sessions of avoidance, the VI 3 schedule was again programmed for

right-lever responding. Characteristic steady rates of responding

were maintained regardless of lever position, but the rates were much

higher when the lever was on the left side than when it was on

the right side. In an attempt to facilitate the development of

equivalent rates of responding regardless of lever position, and

with the restraints in place, concurrent avoidance schedules were

programmed. Two levers were inserted into the chamber at the same

time, and concurrent and independent 5-sec shock shock intervals

were programmed, as were concurrent and independent 20-sec response -

shock intervals. A bi-directional changeover delay was programmed,

sucn that each changeover response initiated a 0.50-sec interval

during which responses were not effective in avoiding shocks.

Acquisition of concurrent avoidance responding was rapid. Within

the first session, the monkey was responding on both levers, and

within 5 sessions, shocks were being delivered at a rate of less than

10 per hour. Occasionally, shocks were delivered following a change-

over response before the COD had elapsed. Response rates stabilized

after about 10 sessions, although left-lever rates were higher than

right-lever rates. Generally, concurrent avoidance responding was

similar to that reported by Sidnar (1962) and Verhave (1961).

Following 27 sessions of concurrent avoidance, concurrent VI

schedJles of shock presentation were introduced foi the first time with









this monkey. Initially, the schedule was cone VI 3, VI 3; however,

the parameters were changed to cone VI 3 (left) VI 1.5 (right) after

40 sessions. The order of experimental procedures and the number of

sessions of exposure to each procedure is presented in Table 2.

With both monkeys, experimental sessions were conducted 6 days

per week, lasted for 60 min, and were preceded and terminated by

4-min TO periods.

Results

Topography and Patterns of Concurrent Responding

With the Plexiglas restraint plates in position, the monkeys

assumed an upright posture, facing the front wall, and resting the

left hand en the left lever, and the right hand on the right lever.

Left- and right-lever responses occurred with the left and right hands,

respectively. Upon the introduction of the concurrent schedules,

both monkeys responded at higher rates on the left lever than on the

right lever. (It should be noted that at the end of Experiment I,

responding was occurring at higher rates when the lever was on

the right side.) After about 7 sessions, Monkey SM-38 responded at

higher rates on the right lever than on the left Irver.

The pattern of responding was one of alternation between levers,

rather than one of simultaneous emission of the two responses. The

manner of alternation controlled the rate disparity between the two

responses. Bursts of responding on one lever (the left for SA-432,

and first the lfit, and subsequently the right for SM-38) were

followed by one or two responses on the other lever. This pattern of

responding interacted with the COD such that shocks progra,,iled for

responding on Lne "lower-rate lever" were held in abeyance until the














Table 2

The Order of Experimental Conditions and the Number of Sessions

at Each Condition for Monkey SM-432, during the Development and

Maintenance of Responding under Concurrent VI, VI schedules

of Electric-Shock Presentation.


Condition

VI 3: shock
(barriers introduced)
(lever position alternating)

VI 3 (left); Avoidance (right)

Cone AV, AV

Cone VI 3, VI 3

Cone VI 3 (left), VI 1.5 (right)


Number of Sessions

16



19

27

40

30









duration of a response burst exceeded the COD. Because of this

interaction, some shocks programmed for the lower-rate lever were

not actually produced until several minutes after they had been pro-

grammed. Conversely, responding on the other lever resulted in

the presentation of shocks soon after they were programmed. The

presentation of shocks resulted in rapid changeover responses be-

tween levers, but not in perseveration of responding on one or the

other lever. Occasionally, shocks programmed for responding on one

lever were produced during these rapid bursts of changeover responses

following shocks produced for responding on the other lever.

Figures 4 and 5 show representative cumulative records of

concurrent responding for Monkeys SM-38 and SM-432, respectively.

Despite the differences in overall rates of responding on the two

levers, steady rates of responding characteristic of VI schedules

were maintained throughout the sessions on each lever.

Effects of Manipulating Parameter Values

Tables 3 and 4 summarize the effects on response and shock rates

of changing the parameter value of one of the concurrent schedules.

No significant changes in response rates accompanied any of the

manipulations with either monkey. Increases in the scheduled density

of shock presentations had differential effects upon shock rates,

depending upon which lever the increase was scheduled for. For

example, increasing the scheduled frequency of shocks for right-

]ever responding with SM-38 [from cone VI 3, VI 3 to cone VI 3 (left),

VI 1.5 (right)] resulted in obtained shock rates approximating










































U



44(
U
ci
H,
ci






H





U





C

ci


U
C:
0
U

0 0


i ci

U uj
C ci


2 0.


4-4 U

ci z








2C
















0
c-J
CN
Nz






































L.JO
-J
Lai




































I U
O n



C ) 0




00




cM
00




)
41 U






C 00
en
E D >

a 0


CO 44



o 0

Co 0
> U

Co C 0




Q U
U O

1 n) nC




14r C' C




31









7o
CN)
Cr)C

C/)~
C-' cr~
























C)~




C-)








1*-*















Table

Successive Five-session Means of

on the Left and Right Levers for


Schedule

Cone VI 3
VI 3











Cone VI 3 (L)
VI .5 (R)


Cone VI 1.5
VI 1.5


Responses per Min

Left Right

24.5 7.4

16.1 18.6

13.9 16.2

8.5 20.8

10.4 22.1


10.2

13.0

12.5



11.9

11.6

9.0

8.4

9.5

9.5


23.0

21.1

19.3



20.1

18.4

18.3

18.5

17.5

17.8


3

Response and Shock Rates

Monkey SM-38.


Shocks per Hour

Left Right

20.0 10.0

20.0 19.0

18.0 20.0

14.3 20.2

14.0 20.0


15.0

14.5

17.5



21.0

17.5

16.0

17.8

16.0

16.0


36.0

37.0

37.2



36.0

37.0

33.0

38.0

37.0

35.0















Table

Successive Five-session Means of

on the Left and Right Levers for


Schedule

Cone VI 3
VI 3


















Cone VI 3 (L)
VI 1.5 (R)


Responses per Min

Left Right

31.1 9.5

34.0 9.5

28.5 9.2

24.2 7.0

26.4 6.9

23.2 6.8

25.0 5.2

27.9 3.7


31.3

28.4

25.5

24.8

25.8

25.3


4

Response and Shock Rates

Monkey SM-432.


Shocks per Hour

Left ight

20.0 16.0

19.5 12.0

21.0 13.8

20.0 14.5

19.0 12.0

18.5 14.0

20.5 10.2

20.5 8.0


20.0

18.5

20.0

20.0

20.0

20.0


12.5

9.0

13.5

8.5

6.5

5.5









the scheduled frequency, that is about 40 shocks per hour. However,

increasing the scheduled frequency of shocks programmed for left lever

responding had little effect upon obtained shock rates for that

response. Presumably, the patterns of response alternation in

interaction with the COD maintained the obtained shock rates at

invariant levels. Thus, shock rates and response rates on the left

lever were unresponsive to parameter changes of the schedule for that

lever.

A similar invariance of patterns of responding influencing

obtained shock rates was evident with SM-432. The initial increases

in shock rates for right lever responding when the higher density,

VI 1.5 schedule was first introduced were transient, and represented

only about 30 percent of shocks actually scheduled.

Discussion

In Experiment II, responding was developed and chronically

maintained on two levers under concurrent and simultaneous schedules

of electric-shock presentation. By responding on each of the

two available levers, the monkeys effected an increase in the frequency

with which shocks were presented. The patterns of responding on each

lever resembled the steady, moderate rates of responding characteristical-

ly maintained by VI schedules. These results complement previous re-

ports of the maintenance of characteristic response patterns by

FI (Kelleher and Morse, 1968; McKearney, 1968), FR (McKearney, 1970).

and VI (McKearney, 1972; Experiment I of the present study) schedules

in a single-response situation.

The patterns of responding on the two levers wsre also similar









to patterns occurring during concurrent variable-interval schedules of

food or water presentation. The lack of any systematic effect on rates

of responding of changing the parameter value of one of the concurrent

schedules, could also be predicted from the results of previous

studies of concurrent schedule interactions. As Baum (1974) has

pointed out, and Iglauer and Woods (1974) have confirmed, changes

in the scheduled frequency of reinforcement affect behavior only

when the behavior comes into contact with the change. (Or, put another

way, experimental changes may affect the behavior of the experimenter

but not of the subject.) In the present experiment, the stereotyped

patterns of alternation between the two responses negated the

scheduled increase in shock density, and the obtained frequency of

shocks remained the same from condition to condition on the low-rate

lever. It should be noted, however, that an increase in the obtained

shock frequency with SM-38, when the schedule was changed from

cone VI 3, VI 3 to cone VI 3 (left), VI 1.5 (right), resulted in a

relative increase in rates of the response with the lower shock density.

If this effect occurs in succeeding studies of concurrent schedules of

shock presentation, it would imply a preference for that response

which produces fewer shocks. No definitive statement of that nature

could be made on the basis of the present results, since in every

condition, the response producing the higher shock frequency was also

emitted aS a higher rate.

In predicting the interaction of responses under concurrent schedules

with qualitatively similar reinforcers, the relative frequency of one

response often approximates thc relative frequency with which it is









reinforced (Herrstein, 1961). The proportion of responses on the left

lever as a function of the proportion of shocks produced for those

responses during the last five sessions of each condition were

calculated, and are presented in Table 5 Since only five determinations

were possible for the two monkeys, the data are only suggestive, and

are presented merely to facilitate comparisons with other studies.

With SM-38, the values indicate "preference" (greater response output)

for the right lever response over and above that which is predicted

by the matching relationship. With SM-432, preference for the left

lever response shows a similar "overmatching." Bauc (1974) suggests

that the proportional expression of the matching relationship, while

useful, does not make allowances for the presence of a systematic

and invariant response bias such as position preference, hand

prepotency, etc. Since the topography of responding in the present

study may be considered such a bias, the relationsh-.p between responses

and shocks for the two levers were also calculated as the ratio of

the two responses and the ratio of the shocks obtained for the two

responses. Baum argues that logarithmic plots of tlhse ratios (for

several determinations) graphically portray bias as a systematic

deviation from the line of perfect matching. The obtained left-to-right

ratios are also presented in Table 3. Obviously, no systematic deviation

occurred between monkeys, nor with each individual nmnkey for the

limited number of determinations.

Although the matching relationship often correctly predicts

response preferences, it by no means accounts for response interactions

in all studies of concurrent responding (e.g., Holld and Davison,

1971; Trevett, Davison, and Williams, 1972). Additionally, in the















Table 5

Relative Measures of Responses and Shocks


I. Proportion of responses on the left
shocks obtained for those responses
of each condition

SM-38


Schedule

Cone VI 3
VI 3


Responses Shocks


lever, and proportion of
during the last five sessions


SM-432

Responses Shocks


Cone VI 3 (L)
VI 1.5 (R) .40 .33 .88 .84


Cone VI 1.5
VI 1.5 .37 .30



II. Ratio of left to right responses, and ratio of left to right shocks


Cone VI 3
VI 3 .38 .39 6.91 2.81


Cone VI 3 (L)
VI 1.5 (R) .67 .48 7.11 3.90


Cone VI 1.5
VI 1.5


.60 .45




38



present experiment, the ongoing topographical preference for one of

the levers prevented the observations of the effects upon rates of

responding of changes in shock rates for each of the operants, in a

manner comparable to many other studies of concurrent operants.















OVERVIEW

The results of Experiment II fully answered three of the four

questions asked at its outset. The monkeys did press each of the

levers, the responding was chronically maintained, and the patterns

of responding were comparable to those seen in Experiment I in

the single-response situation. Thus, the generality of the phenomenon

of behavior maintained by response-contingent presentations of shock

has been extended to a much more complex situation.

Unfortunately, two interrelated aspects of the experiment may

forestall comparisons with other studies of concurrent behaviors: 1)

the topography of the responses and 2) the lack of effects upon

response and shock rates of manipulations of the schedule parameter

values. The fact that the two responses could be emitted simultaneously,

and the degree to which this aspect of topography attenuated the

effects of experimental manipulations, may raise questions concerning

the propriety of including this study within the archival literature

of concurrent operants. The conditions responsible for this potential-

ity are rooted within the methodological history of the study of

concurrent operant behaviors, and are best explained within that

context.

Classification of Concurrent Operants

With respect to the classification of concurrent operants, a dis-

tinction is made between compatible and incompatible responses with re-

gard to topography. For example, in the present situation, since the









levers were a short distance apart, the monkeys could respond on the

two levers simultaneously or in rapid succession. The two responses

are topographically compatible, although the locus of topography is

different. If the environment were changed, such that the levers

were moved 10 feet apart, simultaneous responding on the two levers

would be impossible. The responses would be called incompatible.

Insofar as compatibility or incompatibility may affect the frequency

with which an organism engages in two or more behaviors, these

aspects of topography are of considerable importance in specifying the

determinants of the behaviors. Historically, topographical compati-

bility or incompatibility were not proposed as defining characteristics

of concurrent operants, but formed a basis for the description of the

interactions between responses, and were applied to both reflex and

operant responses (Skinner, 1938). However, conteTporary redefinition

of concurrent operancs have included these aspects of topography as

properties integral to formal laboratory investigations. For example,

Ferster and Skinner (1957) characterized concurrent operants as

two or more responses, of different topography at least with
respect to locus, capable of being executed with little mutual
interference at the same time or in rapid succession, under the
control of separate programming devices (e.g., responses to two
keys present at the same time under separate schedules). (p. 724)

Ferster and Skinner provided no rationale for the demand that concurrent

operants be amenable to easy mutual execution. Casual observations

of behavior reveal many instances of operants which are concurrent

with respect to locus of availability, but are not susceptible to

mutual execution once one of the Operants is emicted chosenr). Indeed,

the choice of one of a number of concurrently available operants may

render impossible the opportunity for engaging in other behaviors.








For example, if travel comprises one component of a behavior which is

chosen, the spatial and temporal separation from the environment in which

the other behaviors are available negates the possibility for simul-

taneous emission or rapid alternation.

Independence of Concurrent Operants

Although Ferster and Skinner (1957) appear to have favored studies

of concurrent operants employing responses capable of simultaneous or

nearly simultaneous emission, later investigators have focused

experimental attention almost exclusively upon incompatible concurrent

operants. For example,

It would appear that a necessary condition for the maximal
independence of two or more operants is that they be topographical-
ly compatible so that they can occur at the same time as well as
in succession....In practice, however, topographical compatibility
often leads to an incompatibility, with respect to separation
of the control of concurrents by their respective schedules of
reinforcement. (Catania, 1966, p. 21L)

and

If it is correct that the simultaneity of concurrent oparants
is incompatible with the independence of concurrent operant,,
concurrent procedures must provide that responses occur only in
succession. (Catania, 1966, p. 219-220)

Catania's insistence upon some degree of incompatibility among

concurrent operants reflects special problems involved in the analysis

of variables controlling one or more of the responses. Two behaviors

which occur simultaneously or in rapid succession may both be maintained

in an adventitious relationship by the consequences resulting from only

one of the behaviors. Concurrent superstitions of this type pose a

problem for the experimental analyst interested in studying the inter-

actions between concurrent behaviors and their respective consequences.

Just as the initial tactics of research dictated the importance of









studying a single response in isolation, so also have tactics led to

the study of concurrent operants in situations designed to ensure

incompatibility, and presumably, independence. By providing experimen-

tal situations in which each response may be controlled separately,

each by its own consequences, the identification of major variables

controlling concurrent interactions has proceeded steadily, if not

rapidly.

A danger inherent in adopting the tactic of primarily studying

incompatible and/or independent concurrent operants is the exclusion

from study of compatible and/or dependent concurrent operants. The

degree to which concurrent operants, compatible or incompatible,

may occur under independent stimulus and schedule control is itself

a variable of importance. Similarly, the extent of compatibility or

incompatibility is a variable, not a defining property of concurrent

operants. Viewed in these terms, compatibility and incompatibility

should be recognized as end points along a topographical continuum,

the parameters of which may be defined either spatially or temporally.














EXPERIMENT III. SUPPRESSION OF RESPONDING UNDER CONCURRENT
VARIABLE-INTERVAL, EXTINCTION SCHEDULES OF
ELECTRIC-SHOCK PRESENTATION


The preceding discussion exposed some of the analytic drawbacks

and conceptual predispositions which might hinder the comparison

of the results of Experiment II with other studies of concurrent oper-

ants. Indeed, one of these problems bears upon the rationale for

actually conducting the experiment. The primary objective of the

experiment was discovering whether or not a response whose only

consequence was the presentation of electric shocks would be maintained

when the opportunity to engage in other operant responses was also

available. Since both responses were maintained, regardless of

the locus of control, explicit or adventitious, the experimental

question was answered in the affirmative. However, as Catania (1966)

pointed out, if one response is adventitiously maintained by the

consequences of a second response, the first response must be analyzed

as a member of a chain, rather than as one of two alternatives in a

choice situation.

Experiment III was designed to assess whether or not the two

responses occurred in a chaining relationship, or whether they were

actually laiintained by the consequences programmed for each. The

most widely accepted procedure for demonstrating independence of

this type calls for the establishment of schedule control for each

of the responses. Changes in the consequences of one response should









then be reflected in changes in the rates or patterns of responding,

or in the relative distribution of responses. Those types of changes

did not occur in Experiment II when the schedule of shock presentation

for the low-rate lever was changed from VI 3 to VI 1.5. It was

contended that the interactions among topography of responding,

engagement of the COD, and stereotyped ongoing patterns of responding

allowed no opportunity for the behavior to come into contact with the

new contingencies.

Since increasing the scheduled shock density was a futile

manipulation without discontinuing the COD, that option was discarded.

Similarly, since the use of a COD is one of the main guarantors of

response independence, discontinuing that contingency might have been

a step backwards. One manipulation which was certain to come into

contact with the behavior was changing one of the VI schedules to

extinction. This procedure had several advantages. First, it would

result in a change in obtained shock rates. Second, the nature of

the control of responding would be clarified. If response rates did

not change, presumably the response was being maintained adventitiously.

If the rate did change, responding was presumably under the control

of the explicitly programmed consequences. Third, a number of studies

have examined the effects of extinction upon shock-maintained

responding (e.g., Kelleher and Morse, 1968). When responses no longer

produced shocks, response rates decreased to very low levels. The

responding returned to baseline levels when shocks were again presented.

Fourth, the effects of extinction on one of two responses maintained

under concurrent variable-interval schedules of food presentation









(Brownstein and Pliskoff, 1968; Catania, 1963; Herrnstein, 1961; Stubbs

and Pliskoff, 1969) and point production (Catania and Cutts, 1963) are

well documented. In most of these experiments, one of the VI schedules

was held constant, and the parameter value of the other schedule was

manipulated. In all cases, when the second schedule was extinction,

very few responses occurred on the manipulandum for which extinction

was programmed. Rates of emitting the other response which still

resulted in reinforcement either remained the same or increased.

In the study by Catania and Cutts, human subjects received points

on a counter according to a 30-sec variable-interval schedule. Two

buttons were located on the panel facing the subjects, but reinforce-

ment was programmed for pushing only one of the buttons. The presence

or absence of a 4.5-sec COD differentially affected rates and patterns

of pushing the button on which extinction was programmed. When this

protective contingency was in effect, very few pushes occurred cn the

extinction button, while high rates were maintained on the other

button. When no COD was programmed, equivalent rates of responding on

the two buttons often occurred. Generally, much higher rates of pushing

the extinction button occurred when the COD was not programmed, indicat-

ing that the response was adventitiously maintained by the consequences

programmed for the other button. In a similar experiment with pigeons,

the presence or absence of the COD had similar effects upon the

maintenance of the response under extinction.

Method

Subjects and Apparatus

Monkeys SM-38 and SM-432 were used with the sare apparatus.









Procedure and Results

At the beginning of this experiment, the two monkeys were respond-

ing concurrent VI schedules of electric-shock presentation.

With Monkey SM-38, the schedule was cone VI 1.5, VI 1.5; with Monkey

SM-432, the schedule was cone VI 3 (left), VI 1.5 (right). A 0.50-sec

COD was programmed.

After 30 sessions the schedule programmed for responding on the

right lever was changed to extinction; that is, responding on the

right lever no longer resulted in shock presentations. The results

of this manipulation are shown in the first two panels of Figures 6

and 7.

With Monkey SM-38, responding on the left lever, on which shocks

were still programmed, decreased from 15 responses per min to less than

1 response per min within 3 sessions. Responding on the right,

"extinction," lever, decreased from 25 to 7 responses per min. The

decreases in response rates were characterized by gradual decelerations

rather than by abrupt transitions from high to low rates. Although

left-lever responding was suppressed by the end of the 3rd session

of this manipulation (session 33 on Figure 6), right-lever respond-

ing continued into the next session before being suppressed entirely.

With Monkey SM-432, no immediate effects of the schedule change

were apparent. However, over a period of 50 sessions, while left-lever

responding remained constant at about 20 responses per min, right-lever

response rates increased from less than 3 per min to more than 11 per

min. By session 87, right-lever rates had reached 16 responses per min,

and were higher than left-lever rates for the first time under any





































> >
to

> 0
h a) i *







-u a a
0) 0 4
tm o 0 a)


-H o 0 0O


M O 1

CL 0 1
0 4' ) 0)





0 0 4J
o4 4 r
^ a a *e


U I- LU
g 0 4 4'






H ri
) ) 0 0
C.) tnl 0 0
o C LU (u
t G0 '0 0) 4






0 i C -4 >
' t o 0

01 o to 0)









lU H, CO 4 C
P Cd 4U C










x--
r I
UJ


u *'*

o










0
o

O LO;1~ -------- -*.*











o *^
= .
0


0 0 .0
tof -


P'P nP


0 0 0
to N -


N I!,N i nOH-i
S3SNOdSA3 S 3 0 H S


----l~-L~II _L --~i-_l


c0




C)






7--

o
. . . . .
"s? ^ -
7 ^ '
'< '

































04 4



ct 0 a 4
CO (U
>HC
>4 c: 4- -H






to

4. 0
CO Do Ci '4-1

0 4 c




0 -0 44 U
,C C 4C 4 I

C CD o -d
o 0 c
c4 M C 1 -4 C)


C o 0 co

o co t4 QH

a) w ED -D


m U 4J b 0
Q F *r *r
C 0 O

0 0 *Q)
cD r-i qC
m o fLa = c

O B0

0D co C
ro Ci -n 1 D
CD 4J -

a1) 0. *j 0)
c 0 Q- ) 0
:3 4 aD
-,4 Do 0-r
4 H 0. 6D










X--


uJ
0i

.. .. .


0 0 0 0



s N 0 s S3 P


0 0

iJn0OH
S >1 0 H S









condition of concurrent schedules with this monkey. From session 87

to session 100, left-lever rates decreased from 15 to less than 1

response per min. Typical VI patterns of responding occurred until

session 98, when short bursts of responding alternated with long

pauses. During the 13 sessions of suppression of left-lever behavior,

responding on the right lever occurred at steady rates between 6 and

12 responses per min. All responding ceased during the 71st session

of cone VI 3 (left), EXT (right) (session 101 on Figure 7).

Following the suppression of responding under the concurrent VI,

EXT schedule, concurrent VI, VI schedules of shock presentation

were reintroduced. Although occasional responses occurred on both

levers, usually followed by shoc-s, the behavior of both monkeys

was still suppressed after 10 sessions of concurrent VI, VI.

The sequence of procedures described in Experiment II were then

employed to redevelop concurrent responding. With Ionkey SM-38,

9 sessions of avoidance was followed by 10 sessions of VI 3; in both

conditions, the position of the lever was alternated from session to

session. On the 20th day of retraining, concurrent 1.5-min VI

schedules were programmed.

With SM-432, 18 sessions under the concurrent avoidance schedule

preceded introduction of cone VI 3, VI 3.

Reacquisition of the concurrent VI, VI perforrmnce was rapid.

As is shown in Figures 6 and 7, the two monkeys responded at slightly

higher rates on both levers than during the previous condition under

concurrent VI, VI. After 30 sessions, extinction was programmed for

responding on the left lever. With both monkeys, responding on the









right, "shock," lever decreased to rates of less than 1 response per

min within 10 sessions. Responding on the left, "extinction," lever

remained at rates of about 20 per min, until responding was suppressed

on the right lever. Thereafter, left-lever rates decreased to zero

in a negatively accelerated pattern within the next two or three

sessions. Cumulative records showing the transition from response

maintenance to response suppression during this phase for SM-432

are presented in Figure 8.

Discussion

The results of Experiment III are similar neither to the effects

of extinction of a response maintained by schedules of shock presen-

tation in a single-response situation, nor to the effects of extinction

of one of two concurrent responses maintained under schedules of food

or point presentation. In some respects, the results are similar

to the effects on a punished response of the availability of an

unpunished alternative response (Azrin and Holz, 1966). In that exper-

iment, pigeons' responses on one key resulted in both food and shock.

Responding was not significantly suppressed at low and moderate in-

tensities of electric shock. When the intensity was increased further,

response rates decreased, but a surprising number of responses were

still occurring even at high shock intensities. When a second key was

uncovered, and responses directed against this key resulted in food

presentation at equivalent frequencies to the first response, but

without the presentation of shocks, the pigeons responded on the

second key exclusively. Responding on the second key remained prepotent

even at intensities of shock on the first key which had little effect

































FiguLi 8 representativee cumulative records for I' ..ey S' 4-

sho. ':g The transition' from response maintenance! to respIij'

supprejsion during the second ec:poisure to the extin;'i* conditiie,

In each record, the schedule programmed for respond i. i'. he i '-r

leav.r is presented first. The sessions are nitr vred :io oo''rrespond

Figure 7.




54

S M 4 2


cone VI 3, VI 3 SESSION 146


LEFT


RIGHT, 3 SESO



cone EXT. VI 3 SESSION 15!


LEFT






RIGHT







LEFT


conc EXT, VI 3 SESSION 158
conc EX -, V/I 3 SESSION 158i


RIGHT









upon responding in the absence of the alternative response. Since

each response produced a characteristic consequence food presentation

- these results are not entirely applicable to the present experiment.

The main point of interest is the difference in rates of responding on

the punished key as a function of the presence of another response.

When considered within this context, the results of the present

experiment suggest that the degree to which behavior will be main-

tained by schedules of shock presentation may be greatly affected

by the availability of other responses. The results do not suggest

that this effect will necessarily result in the suppression of

behavior since the two operants were maintained whan each produced

shocks. Further research should examine conditions in which responses

on one manipulandum result in shock presentation, and responses on

a second manipulandum result in a qualitatively different stimulus -

for example, food presentation.

Experiment III did indicate that the two responses were under

the control of the consequences programmed for each. If one of

the two responses had been maintained in an adventitious relationship

by the consequences of the other response, then discontinuing shocks

programmed for that response should have had little effect upon

either response. The fact that similar effects on responding occurred

during extinction of each response in two separate conditions and

with both monkeys is evidence that the two operant responses wece

independent, and were under the control of their respective conse-

quences.









The failure to recover the previous rates and patterns of respond-

following the suppression of behavior under the concurrent VI, EXT

schedule by reprogracning the concurrent VI, VI schedule was not

surprising. The steady, ongoing response rates were no longer

present. Reintroduction of the concurrent VI, VI schedule resulted

in a low ratio of responses to shocks. After several sessions

of a few responses alternating with long pauses, no further

responding occurred. The behavior probably could have been restrength-

ened if the the concurrent VI, VI schedules were reintroduced

while responding was occurring at steady, although low, rates.

Kelleher and Morse (1968) and McKearney (1969) have shown

previously that the same intensity of shocks either maintains or

suppresses behavior depending upon the schedule under which it is

presented. Similarly, depending upon the previous history of

responding, the same paramters of Fl schedules of shock presentation

will either suppress or enhance responding (Azrin and Holz, 1966;

Morse and Kelleher, 1970). In the present experiment, behavior was

both maintained and suppressed by the same schedules of shock presen-

tation, after the same history, depending upon the consequences pro-

grammed for a second, concurrent response. These results provide

further emphasis for Morse and Kelleher's (1970, Pp. 172-175)

conclusions that the effects of stimuli upon behavior change when

the conditions under which the behavior occurs are clanged.














REFERENCES


Azrin, N. H. and Holz, W. C. Punishment. In W. K. Honig (Ed.),
Operant behavior: areas of research and application. New York:
Appleton-Century-Crofts, 1966. Pp. 380-447.

Baum, W. M. On two types of deviation from the matching law: bias
and undermatching. Journal of the Experimental Analysis of
Behavior, 1974, 22, 231-242.

Brownstein, A. J. and Pliskoff, S. S. Some effects of relative
reinforcement rate and changeover delay in response-independent
concurrent schedules of reinforcement. Journal of the Experimen-
tal Analysis of Behavior, 1968, 11, 683-688.

Byrd, L. D. Responding in the cat maintained under response-independ-
ent and response-produced electric shock. Journal of the
Experimental Analysis of Behavior, 1969, 12, 1-10.

Byrd, L. D. Responding in the squirrel monkey under second-order
Schedules of shock delivery. Journal of the Experimental
Analysis of Behavior, 1972, 18, 155-167.

Catania, A. C. Concurrent performances: reinforcement interaction and
response independence. Journal of the Experimental Analysis of
Behavior, 1963, 6, 253-263.

Catania, A. C. Concurrent operants. In W. K. Honig (Ed.), Oparant
behavior: areas of research and application. New York:
Appleton-Century-Crofts, 19u6. Pp. 213-270.

Catania, A. C. and Cutts, D. Experimental control of superstitious
responding in humans. Journal of the Experimental Analysis of
Behavior, 1963, 6, 203-208.

Eubanks, J. L., Killeen, P., and Hamilton, B. Post-shock pause
reinforces responding in schedules of shock presentation. Pre-
sented at the meeting of the American Psychological Association,
Montreal, September, 1973.

Ferster, C. B. and Skinner, B. F. Schedules of reinforcement.
New York: Appleton-Century-Crofts, 1957.

Findley, J. D. An experimental outline for building and exploring
multi-operant repertoires. Journal of the Experimental Analysis
of Behavior, 1962, 5, 113-166.










Fleshier, M. and Hoffman, H. S. A progression for generating variable-
interval schedules. Journal of the Experimental Analysis of
Behavior, 1962, 5, 529-530.

Hake, D. F. and Azrin, N. H. An apparatus for delivering pain shock
to monkeys. Journal of the Experimental Analysis of Behavior,
1963, 6, 297-298.

Herrnstein, R. J. Relative and absolute strength of response as a
function of frequency of reinforcement. Journal of the Experimen-
tal Analysis of Behavior, 1961, 4, 267-272.

Hollard, V. and Davison, M. C. Preference for qualitatively different
reinforcers. Journal of the Experimental Analisis of Behavior,
1971, 16, 375-380.

Iglauer, C. and Woods, J. H. Concurrent performances: reinforcement
by different doses of intravenous cocaine in rhesus monkeys.
Journal of the Experimental Analysis of Behavior, 1974, 22, 179-196.

Kelieher, R. T. and Morse, W. H. Schedules using noxious stimuli. III.
Responding maintained with response-produced electric shock.
Journal of the Experimental Analysis of Behavior, 1968, 11, 810-838.

Malagodi, E. F., DeWeese, J. Webbe, F. M., and Palermo, G. Responding
maintained by schedules of electric-shock presentation. Presented
at the meeting of the Psychonomic Society, St. Louis, November,
1973.

McKearney, J. W. Maintenance of responding under a fixed-interval
schedule of electric-shock presentation. Science, 1968, 160,
1249-1251.

McKearney, J. W. Fixed-interval schedules of electric-shock presen-
tation: extinction and recovery of performance under different
shock intensities and fixed-interval durations. Journal of the
Experimental Analysis of Behavior, 1969, 12, 3C1-313.

McKearney, J. W. Responding under fixed-ratio and multiple fixed-inter-
val fixed-ratio schedules of electric shock presentation. Journal
of the Experimental Analysis of Behavior, 1970, 14, 1-6.

McKearney, J. W. Maintenance and suppression of responding under
schedules of electric shock presentation. Journal of the Experi-
mental Analysis of Behavior, 1972, 17, 425-432.

Morse, W. H. and Kelleher, R. T. Schedules using noxious stimuli. I.
Multiple fixed-ratio and fixed-interval termination of schedule
complexes. Journal of the Experimental Analysis of Behavior, 1966,
9, 267-290.









Morse, W. H. and Kelleher, R. T. Schedules as fundamental determinants
of behavior. In W. N. Schoenfeld (Ed.), The theory of reinforce-
ment schedules. New York: Appleton-Century-Crofts, 1970.
Pp. 139-185.

Sidman, M. Avoidance conditioning with brief shock and no exteroceptive
warning signal. Science, 1953, 118, 157-158.

Sidman, M. Reduction of shock frequency as a reinforcement for
avoidance behavior. Journal of the Experimental Analysis of
Behavior, 1962, 5, 247-257.

Skinner, B. F. The behavior of organisms: an experimental analysis.
New York: Appleton-Century-Crofts, 1938.

Staddon, J. E. R. Some properties of spaced responding in pigeons.
Journal of the Experimental Analysis of Behavior, 1965, 8, 19-27.

Stretch, R., Orloff, E. R., and Dalrymple, S. D. Maintenance of respond-
ing by a fixed-interval schedule of electric-shock presentation
in squirrel monkeys. Science, 1968, 162, 583-585.

Stretch, R., Orloff, E. R., and Gerber, G. J. Multiple interruptions
of responding maintained by a fixed-interval schedule of
electric-shock presentation in squirrel monkeys. Canadian Journal
of Psychology, 1970, 24, 117-126.

Stubbs, D. A. and Pliskoff, S. S. Concurrent responding with fixed
relative rate of reinforcement. Journal of the Experimental
Analysis of Behavior, 1969, 12, 887-895.

Trevett, A. J., Davison, M. C., and Williams, R. J. Performance in
concurrent interval schedules. Journal of the Experimental
Analysis of Behavior, 1972, 17, 369-374.

Verhave, T. Some observations concerning prepotency and probability
of postponing shock with a two-lever avoidance procedure.
Journal of the Experimental Analysis of Behavior, 1961, 4, 187-192.
















BIOGRAPHICAL SKETCH

The author of this dissertation, Frank M. Webbe, was born in Vero

Beach, Florida on the 13th of October, 1947, to Richard and Peggy

Webbe. He attended hig' school in Miami, and received the Bachelor of

Arts degree from the University of Florida in 1969. he is married to the

former Ellen Marie Kane, and he and Mrs. Webbe are the parents of

two children, Elizabeth St. Clare and Tristan Kane.

Since receiving the Bachelor of Arts degree, he has matriculated

through the Graduate School of the University of Florida, receiving the

Master of Science degree in 1971. lie has been active both in research

and teaching during his tenure as a graduate student, and plans

to continue an academic career.










I certify that I have read this study and that in my opinion it
conforms to acceptable standard; of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.




E. F. Malagodi, Chai .ian
Associate Professor -f Psychology



I certify that I have read this study and that in my opinion it
conforms t' acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.




Marc N. Branch
Assistant Professor of Psychology



I certify that I have read this study and that in my opinion it
conforms to acceptable st..dards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.




Henry S. Pennyp tkce-
Professor of sychoiogy



I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.




SCarol J. VanHartesveldt
Associate Profess r of Psychology











I certify that I have read this study and that in ry opinion it
conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.




William Wolkin.
Professor of Special Education



This dissertation was subrtted to the Graduate Faculty of the
Department of Psychology in the College of Arts and --lences and to
the Graduate Council, and was accepted as partial fui-Iillment of the
requirement for the degree of Doctor of Philosophy.

August, 1974


Demi, Gradue'e Srhjol




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs