2 × 2 games played once
| DOI | 10.1177/002200277201600310 |
| Published date | 01 September 1972 |
| Date | 01 September 1972 |
| Author | Melvin J. Guyer,Anatol Rapoport |
| Subject Matter | Articles |
Gaming
2
x
2
games
played
once
MELVIN
J.
GUYER
Mental Health Researcli Itistihc te. University
of
Mich igati
and ANATOL KAPOPORT
Departtrients
of
Psychology and Mathernotics, University
of
Toronto;
Mental Healtli'Researcli bstihlte, University
of
hfichigati
Ititrodtiction
Experimental games have become a stand-
ard technique for the study
of
social inter-
action. Although reqearch in this area is
relatively new, the earliest studies having
been reported some
20
years ago,
a
large
body of literature now exists. Thus one
bibliography of experimental gaming re-
search contains some 1,000 citations (Guyer
and Perkel, 1972). The popularity of ex-
perimental ganies
as
a research tool lies in
the fact that they provide the social scientist
with a well-defined simple laboratory task
that can be explained to subjects in a few
minutes and that gives the experimenter
simple quantitative dependent variables with
which to work. And by selecting appropriate
payoff matrices, a number
of
different
models of social situations can be created in
the laboratory.
For
example situations that
pose the problem of strategic preemption
can
be
modeled by the game of Chicken
(Rapoport and Chammah, 1966) and situa-
tions involving the elements of threat and
appeasement can be modeled by other sim-
ple two-person games (Guyer and Rapoport.
1970).
In many branches of science the appeal of
a simple model of an admittedly more
complex phenomenon is that it allows re-
search
to
be carried out on a small number
of
variables at one time while holding
constant a great number
of
others. The
assumption is that a systematic investigation
of the factors affecting the model will aid in
understanding the real-world processes
which are
our
concern. It hardly needs
410
hlELVINJ.
GUYER
ANDANATOL
RAPOPORT
repeating that the fundamental laws
of
physics were developed
in
terms
of
the
behavior of ideal objects moving through
frictionless space.
It
is in these same terms that experi-
mental games can be viewed, for they may
be said
to
represent or model the simplest
and most abstracted instances of social
interaction. In particular the two-person
two-strategy games contain the minimal set
of elements necessary for social interaction
to occur. That is, social interaction requires
that there must be
at
least two social actors
and that each must be able
to
make
a
behavioral response, a simple choice between
at
least two alternatives. It is this apparent
simplicity which has contributed to the
extensive use of the
2
x
2
game
as
a research
tool in
a
variety of studies.
An overview
of
studies using
2
x
2
games
suggests an underlying basic dichotomy
in
orientation toward the theoretical function
that games
serve
for the experinienter.
Stated most simply the dichotomy can be
characterized in terms of whether the game
is used
as
the independent variable
of
central
interest
or
whether the game
is
used instead
as an assessment device for generating de-
pendent variables related to nongame-rele-
vant experimental manipulations. In the
former case the game itself is seen as being a
“cause”
of
the behavior of subjects in that
game. In the latter case the game servcs only
to scale behavior following some experi-
mental treatment; used in this way the game
functions in a fashion similar
to
an attitude
questionnaire
or
a
sociogram.
It can be argued that this latter use
of
games is associated with more traditional
approaches to the theoretical study
of
social
behavior.
Thus
the variables of interest are
typically taken
as
being the products and
consequences
of
a learning process, either
individual
or
social. These may be said
to
be
“l~urnan” variables since they involve inter-
nal
changes in the organism through learn-
ing, or they are seen
as
a
consequence of
biological traits affecting temperament and
potential for performance in a variety of
different situations.
This
traditional ap-
proach employs as basic concepts in explain-
ing behavior, such theoretical constructs
as
personality, attitudes, motivations, social
role, etc. Essential
to
each of these concepts
is some process of prior learning. Simply
put, an individual
or
individuals are viewed
as
behaving as they do because
of
modes of
responding which they have previously ac-
quired and wliicli predispose them
to
act
Li
characteristic ways in current situations.
And in the traditional view the
enviromient
in which the behavior occurs is regarded
as
being of secondary importance in determin-
ing behavior, at best serving a signaling
function by eliciting patterns of behavior
previously acquired in other settings.
Further it may be claimed that the several
competing traditional theories of social be-
havior, though they may differ in what they
take to be the crucial variables affecting
behavior (one perhaps emphasizing “role
structure” and another “social institu-
tions”), do not however differ in terms
of
their focus on acquired behavior
as
the basis
for variables to explain and predict behavior.
Perhaps most importantly the traditional
approaches
to
social behavior have been
occupied with the
products
and
conse-
quences
of
a learning process. This concern
lias
been in terms of the way in which
particular acquired characteristics of indi-
viduals predispose them to behave in
a
variety
of
defined situations.
For
example
there is a considerable literature dealing with
the ways in which social attitudes, which are
acquired characteristics, predispose
an
indi-
vidual to behave when in standard test
settings. Another example is found in the
literature on social class affiliation.
This
CorifIict
Resohrtioii
f
Volirriie
XVI
Niiniber
3
To continue reading
Request your trial