Forms of Inhibition in Animal and Human Learning

Abstract

Forms of inhibition were identified in human predictive learning that are qualitatively similar to those identified by P. C. Holland (1984) in rats. When P (positive) signaled the outcome and PN (N = negative) signaled the absence of the outcome, participants learned the discrimination, but the negative cue did not suppress responding to a transfer cue. Post-learning reversal training, in which N was followed by the outcome, did not abolish the original discrimination. These 2 results imply a configural form of inhibition. Negative transfer, which indicated a 2nd, elemental form of inhibition, was observed when neither PN nor N were reinforced during the discrimination stage. Under these conditions, negative transfer and the original discrimination were both abolished by individually pairing N with the outcome. Empirical parallels and differences with the animal conditioning literature are discussed.

Full text

Journal
of
Experimental
Psychology:
Animal
Behavior
Processes
1995,
Vol.
21, No. 2,
129-142
CopyriRht
1995
by the
American
Psychological
Association,
Inc.
0097-7403/95/S3.00
Forms
of
Inhibition
in
Animal
and
Human Learning
Douglas
A.
Williams
University
of
Winnipeg
Forms
of
inhibition
were
identified
in
human
predictive
learning
that
are
qualitatively
similar
to
those
identified
by P. C.
Holland
(1984)
in
rats.
When
P
(positive)
signaled
the
outcome
and
PN (N =
negative)
signaled
the
absence
of the
outcome,
participants
learned
the
discrimination,
but the
negative
cue did not
suppress
responding
to a
transfer
cue.
Post-
learning
reversal
training,
in
which
N was
followed
by the
outcome,
did not
abolish
the
original
discrimination.
These
2
results
imply
a
configural
form
of
inhibition.
Negative
transfer,
which
indicated
a
2nd,
elemental
form
of
inhibition,
was
observed
when neither
PN
nor
N
were
reinforced
during
the
discrimination
stage.
Under
these
conditions,
negative
transfer
and the
original
discrimination
were
both
abolished
by
individually
pairing
N
with
the
outcome.
Empirical
parallels
and
differences with
the
animal
conditioning
literature
are
discussed.
Animal
learning
and
cognitive psychology
are
often
de-
picted
as
separate
empirical
and
theoretical
domains
that
have little
in
common with each other except
for an
increas-
ingly
distant historical connection. Experiments
in
animal
conditioning
are
often
characterized
as
providing insight
on
how
reflexive associations
are
formed between stimuli
and
responses
(see
Rescorla,
1988).
However, humans have
been
characterized
as
using
abstract rules
to
learn
in
richer
and
more complex ways. Given
the
predominance
of the
levels-of-learning
analysis,
it is
intriguing
that
modern
re-
search
in
human predictive learning
has
resulted
in the
discovery
of
empirical principles that correspond with those
identified
in the
earlier animal conditioning literature (for
a
review,
see
Allan,
1993).
For
example, when several antecedent cues
are
presented
together
and
followed
by an
outcome,
the
cues compete
with
each other
to
signal
the
outcome. Thus,
if one
arranges
for
Cue A to
reliably signal
the
outcome
in
Stage
1, and Cue
B is
then added
in
Stage
2, Cue A
blocks
Cue B's
associ-
ation
with
the
outcome. This result holds regardless
of
whether
the
experimental subjects
are
rats
or
people
(e.g.,
Chapman
&
Robbins, 1990; Kamin, 1969)
and
even though,
on
an
absolute basis,
B is
always followed
by the
outcome.
As
empirical correspondences have accumulated,
the
dis-
tinction among theories
in the
animal
and
human domains
has
blurred. Rule-based theories
of
human induction
(e.g.,
J. H.
Holland,
Holyoak,
Nisbett,
&
Thagard,
1986)
have
been
offered
as
alternative accounts
of how
animals
act in
simple conditioning experiments
(e.g.,
Holyoak,
Koh,
&
Nisbett,
1989).
Although this possibility
has not
been well
received
in
animal learning,
the
empirical convergence
of
animal
and
human
experiments certainly allows
for the
This
research
was
supported
by a
grant
from
the
Natural
Sci-
ences
and
Engineering
Research
Council
of
Canada.
Correspondence
concerning
this
article
should
be
addressed
to
Douglas
A.
Williams,
Department
of
Psychology,
University
of
Winnipeg,
Winnipeg,
Manitoba,
Canada
R3B
2E9.
possibility that symbolic cognitive processes
are
responsible
for
the
generation
of
simple
conditioned
responses
in
lower
organisms. Perhaps more appealing
is the
claim
that
reflec-
tive judgments
of cue
predictiveness
in
humans
are
complex
manifestations
of
simple associative networks
(e.g.,
Shanks,
1987,
1993a,
1993b;
Van
Hamme,
Kao,
&
Wasserman,
1993).
That
is,
interevent
associations
such
as
those
ob-
served
in
animal
conditioning
may
explain
contingency
judgments. Another possibility
is
that
cognitive mecha-
nisms
of
different
levels
of
complexity
may
function
in
similar ways when constrained
by
similarities
in the
struc-
ture
of the
task. That
is,
humans
may use
abstract rules
to
solve discrimination problems
that
rats solve
by
associative
learning.
In
the
current experiments,
I
focused
on the
extent
to
which several critical findings about negative contingency
learning
in
animals
can be
generalized
to
humans.
An ap-
preciation that events
may be
correlated includes
a
sensi-
tivity
to
negative relations
in
which
an
antecedent
cue
signals
a
reduction
in the
probability
of the
outcome (Chat-
losh,
Neunaber,
&
Wasserman,
1985;
Rescorla,
1968).
Chapman
and
Robbins
(1990)
provided
one of the
best
demonstrations
of
negative contingency learning
in
humans
(see also Allan
&
Jenkins,
1983;
Neunaber
&
Wasserman,
1986).
They asked college students
to
predict whether
the
overall value
of a
fictional stock market would either rise
(outcome)
or not rise (no
outcome)
on the
basis
of
whether
the
values
of
individual stocks
had
risen (cue present)
or had
not
risen
(cue
absent).
In one
experiment, participants
received
a
P+/PN-
conditioned inhibition problem
(P
=
positive
cue;
N =
negative
cue;
+ =
followed
by
outcome;
— = not
followed
by
outcome)
in
which
a
rise
in
the
value
of
Stock
N
signaled that
the
stock
market would
subsequently
not rise. At the
termination
of the
experiment,
participants rated
the
contingency between
the
cues
and the
outcome
on a
scale ranging
from
—100
to
+100.
As ex-
pected, participants' ratings
of
Stock
N
approached
the
maximum
value
for a
negative
cue
(i.e.,
-100).
Impor-
129
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.