Seventy-Sixth Year
EDITED AND MANAGED BY STUDENTS OF THE UNIVERSITY OF MICHIGAN
UNDER AUTHORITY OF BOARD IN CONTROL OF STUDENT PUBLICATIONS

Faculty Review: rPrisoner's Dilemma'

Where Opinions Are Free,
Truth Will Prevail 42MAYNAR ST., ANN APBOR, MICH.

NEws PHONE: 764-0552

Editorials printed in The Michigan Daily express the individual opinions of staff writers
or the editors. This must be noted in all reprints.
WEDNESDAY, FEBRUARY 2, 1966 NIGHT EDITOR: JOHN MEREDITH

Rhodesia: Economic
Sanctions Aren't Enough

FIREY OUTBURSTS from the Tory
benches greeted Prime Minister Har-
old Wilson's announcements of stricter
economic sanctions against Ian Smith's
rebel Rhodesian regime this week.
The additional sanctions themselves
were not that much more drastic than
those already in existence-stricter im-
port restrictions, announcement of com-
plete export restrictions and a warning
against extending credit to the rebels-
and the blasts against them thus indi-
cated that England, like most of the rest
of the world, has a long way to go before
black Africans will truly be free.
Despite all the outcry against Wilson's
plans, and despite considerable efforts by
Smith's English propaganda machine, La-
bor and Conservative seem to be in close
agreement on many issues relating to
Rhodesia. Both still agree that the rebel
government must be made to renounce
its declaration of independence, that eco-
nomic sanctions ought to be taken to
achieve this, and that after it is achieved
the white minority ought to rule restrict-
edly until blacks are ready to enter the
government. These in practice are the
essentials of the case.
BUT WHILE THE ENGLISH front has
few areal cracks in it, neither does the
Rhodesian. The primary contributor to
Rhodesian tenacity has been the general
failure of the English economic sac-
tions.
Despite the seeming logic of the case,
which says an economy cannot operate
without key inputs, and despite the sup-
posed implications of Smith's riding a
bicycle to work each morning, it is a
known fact that national economies are
very durable things. The surprising in-
effectiveness of Allied bombing of Ger-
many during World War II (the German
index of industrial production reached its
wartime peak in late 1944, the same
months that saw the most intense bomb-
ing raids) and of American raids on
North Viet Nam now testify to the fact
that no such thing as a "strategic mater-
ial" really exists. *
More specifically, 63 per cent of Rho-
desia's energy needs are supported by
the coal of which she has plenty. Ten
per cent more is met by hydroelectric
power from the Kariba Dam.
NEITHER HAVE the English financial
embargoes had the effects one might
think. Before independence was declared
the Rhodesians managed to move between4
12 and 14 million pounds sterling away
from British control. In addition they
have 3.5 million pounds within the coun-
try. Moreover South Africans, though pre-
vented by the Bank of England's influ-
ence from returning any of the 7 million
Rhodesian pounds they held, have been
extending credit to Rhodesia with the 7
million pounds as collateral.
Business Staff
CY WELLMAN, Businesseg Manager
ALAN;01UEC;KMAN........ Advertising Manager.
SUSAN CRAWFORD ..... Associate Business Manager
JOYCE FEINBERG.... ........Finance Manager
Subscription rate $4.50 semester ny carrier ($5 by
mal); $8 yearly by carrier ($9 by mal) ,
second class posage paid at Ann Arbor, Mich.

And so in many ways the situation
looks like a standoff with neither side
having to concede to the other. Yet in a
wider sense Rhodesia's attempt at na-
tional segregation may already be said to
have failed; Rhodesia's recent independ-
ence has enabled the vast majority of
the world's nations to go on record as
soundly condemning it. Smith's govern-
ment's legitimacy has been revoked.
With the withdrawal went the aura of
legitimacy that surrounds other such
African nations: Portuguese Angola and
the Union of South Africa, to name the
two most outstanding.
BUT AS THE STALEMATE in the Rho-
desian case is currently illustrating,
calling world attention to such govern-
ments' lack of legitimacy is not enough.
For South Africa is not Arkansas and it
is not possible to force local segregation-
ist governments into submission there
as it is in the United States.
On an individual localized basis, such
as that in which it exists in the U.S.,
segregation is one of the few things that
can be assumed to be basically an evil.
When such an evil becomes a national
policy, as, for example, murder did in
Germany under the National Socialists,
the practicing nation can reasonably be
defined to be behaving in an actually
criminal manner.
In other words, there is no difference
between segregation in the U.S. or Eng-
land and segregation in Rhodesia. But
there is most certainly a difference be-
tween the governments of America and
England on the one hand and that of
Rhodesia on the other. That difference is
enough to define the government of Rho-
desia, and the whites and blacks who
support it, as criminal and to treat them
as such.
THIS IS IMPORTANT to remember not
sodmuch in the Rhodesian case, where
the die has evidently been cast, as it
will be in the cases which will follow.
For none should make the mistake of
thinking the Rhodesian trouble is the end
of African crises; like Viet Nam, it is most
profitably regarded as a sign of things to
come. Many south African countries share
her general population distribution and
social structure, and the combination of
the two is inherently unstable.
The cases which will follow the Rho-
desian can only realis'tically be approach-
ed from the standpoint that such gov-
ernments do not actually bear the com-
forting stamp of popular support. Rho-
desia has been approached exactly as if
the rebel government enjoyed that sup-
port ,as if its leaders did not consist al-
most wholly of men who have, for one
reason or another, ignored the needs and
desires of a large proportion of their na-
tion's inhabitants.
Approaching Rhodesia from that
standpoint, adn thus employing eco-
nomic sanctions, is evidently not the ap-
proach that is required. For it soon be-
comes apparent that the economic ship,
though perhaps not invincible, is certain-
ly more stable than earlier might have
been supposed; and once such men are
embued with that confidence, little will
bring them down short of force.
-LEONARD PRATT

By KENNETH E. BOULDING
IT MAY BE a little odd to ap-
ply the term "beautiful" to a
book which is addressed primarily
to experimental psychologists,
which describes and analyzes a
series of narrowly controlled ex-
periments on the behavior of hu-
man subjects, and which re-
quires an above average degree of
mathematical sophistication to
understand. If beauty, however, is
the apparently effortless achieve-
ment of a difficult but necessary
order, then this work shines with
it. It has all the air of a minor,
or perhaps not very minor, clas-
sic, and it is in addition a beau-
tifully printed and produced vol-
ume of which the University of
Michigan Press should be ex-
tremely proud.
The Prisoner's Dilemma is a
species of "game" in the theory
of games, a somewhat ill-named
body of mathematical theory
which deals with situations in
which two parties have two or
more choices, and in which the
outcomes of the choices, in terms
of the welfare of the parties,
can be ordered or measured, and
depend on the combination of the
:choices made. The choices and the
payoffs are expressed in what is
technically known as the "nor-
mal" form of the game in a pay-
off matrix such as Figure 1. Here,
the rows, C1, D1, represent the
choices of the first party, the col-
umns, C2, D2, represent, the
choices of the second party; the
numbers in the squares represent
the payoffs, the left-hand num-
ber of the first party, the right-
hand number of the second.
Let Rapoport himself explain:
"The nickname Prisoner's Dilem-
ma, attributed to A. W. Tucker,
derives from the original anec-
dote used to illustrate the game.
Two prisoners, held incommuni-
cado, are charged with the same
crime. They can be convicted only
if either confesses. Designate by
-1 the payoff associated with con-
viction on the basis of confessions
by both prisoners and by +1 the
payoff associated with acquittal.
Further, if only one confesses, he
is set free for having turned
state's evidence and is given a
reward to boot. Call his payoff
under these circumstances +2.
The prisoner who has held out is
convicted on the strength of the
other's testimony and is given a
more severe sentence than if he
had also confessed. Call his pay-
off -2," (pages 24-25). Both pris-
oners will clearly be better off
in the C1-C2 box at the top
left, where neither confesses (C
here stands for cooperate). If,

however, we are in the Cl-C2 po-
sition, it pays either of the pri -
oners to defect, as shown by the
arrows marked d2 and dl. If
prisoner (1) defects, we go from
C1C2 to D1C2, following the ar-
row dl, and prisoner l's payoff
increases from 1 to 2. On the oth-
er hand, if prisoner 1 defects, it
also pays prisoner 2 to defect.
following the arrow e2, for he
will be better off at -1 than at
-2. Similarly, if prisoner 2 de-
fects, following the arrow d2, it
will pay prisoner 1 to defect, fol-
lowing the arrow el; and in both
cases we end up in the bottom
right-hand box, with both pris-
oners worse off than they would
have been if they had cooperated.
This may seem very abstract,
but it represents a model of a
constantly recurring and impor-
tant human situation. Suppose,
for instance, C represents con-
cord, as represented by disarma-
ment, and D discord, represented
by armament. If both parties are
in concord (disarmed), they are
clearly better off than if both
parties are in discord (armed).
Nevertheless, if both parties are
in concord (disarmed), in the
short run it pays one party to
arm. Then, however, it pays the
other party to arm, so that both
of them end in the discord box.
The international system is a
standard example of the process
whereby the dynamic processes of
the system make everybody worse
off.
One sees the same pattern in
family quarrels, in crime, in what
might be called non-economic
conflicts of all kinds. In general
terms, indeed, it is the theory of
the social contract. If both par-
ties maintain the contract and.
are "good" (C2C2), they are both
better off. If one party is good,
however, it pays the other party
to be "bad," and we move, say,
to C1D2 or to D1C2. Then it pays
the other party to be bad, and
we move into the D1D2 box, the
social contract is broken, every-
body is worse off. The importance
of this model, therefore, can
hardly be overestimated, for it is
the key to an immense amount
of human misery.,We can almost
say that all human society is
achieved by precarious coopera-
tion. It is of vital importance,
therefore, to know as much as we
can about the conditions under
which people cooperate and the
conditions under which they de-
feet.
RAPOPORT and Chammah
have set up an elaborate experi-
mental procedure for studying

Prisoner's Dilemma, A Study
in Conflict and Cooperation
By Anatol Rapoport and
Albert M. Chammah
TherUniversity of Michigan
Press, Ann Arbor, 1965
258 Pages, $7.50
this problem. The experimental
apparatus is simple in the ex-
treme. Each subject is asked to
choose simply "right" or "left";
sometimes the payoffs are known
in advance, sometimes they are
not; and the payoffs themselves'
can be varied experimentally.
The game is simplified by making
the payoffs symmetrical, in which
case there are four major para-
meters of the system: P, the pay-
off to each player when both
choose the defecting strategy,
D1D2, which in Figure 1 is -1;
R, the payoff to each player
when both choose the cooperative
strategy, C1C2, in the case of
Figure 1, +1; S, the payoff to
(Cooperate)
(Good)
C2

erative or defecting plays. One of
the most interesting conclusions
of the experiments is that players
tend to get "locked in" either on
the mutually cooperative position,
C1C2, or on the mutually defect-
ing position, D1D2, unless one of
them happens to be a "martyr"
who takes constant punishment
in the hope of reforming the
other.
These dynamic parameters have
the interesting property that they
reflect in a highly abstract form
certain virtues and vices which
are the commonplaces of folk
knowledge, such as trustworthi-
ness, trustfulness, deceitfulness,
and so on. This may represent an
important step towards the clari-
fication of moral concepts which
are of great importance in the
dynamics of social systems, but
which are very hard to reduce to
an abstract and exact form. The
exact pattern of responses de-
pends, of courses on the payoffs
(Defect)
(Bad)
D2^

(Cooperate)
(Good)
Cl

d2
3.1-2 2
() (R) (,S) (T) j
d el
1 4 1 l
(T) (s) P p
e2

might perhaps express this by
saying that we start out with
naive trust. naive trust is be-
trayed and leads to naive distruct
and mutual defection, and that
this eventually is overcome and
leads to mature trust. The lessons
of this pattern for social systems
in real life are to be taken very
seriously.
Although the study did not set
out to examine the different re-
sponses of different personality
types or kinds of people, one
fascinating conclusion did emerge,
which is that women tend to be
less cooperative than men. Male
pairs are the most cooperative,
female pairs the least, and mixed
pairs about half way between
the two (page 196). Mixed pairs,
furthermore,start out with a
very high level of naive tr~ust.
are rapidly disillusioned, and
then recover to about the origi-
nal level. All three types of pairs
follow the pattern of naive trust,
disillusion, and recovery, with the
females recovering much less than
the males.' What this means I
would not venture to speculate,
It is clear that what is opened
up here is a whole new branch
of experimental science, of which
this book is merely the first pio-
neering effort. The Prisoner's Dil-
emma game, indeed, may turn
out to play something of the role
in the social sciencesbithat the
fruit fly has done in biology, of
providing a relatively simple pat-
tern which is capable of wide ex-
perimental variation and which
has nevertheless a profound im-
plication forthe larger and more
complex systems which arl not
accessible to experimental con-
trol. It will not be at all surpris-
ing if this work produces a real
revolution in the way we th:nk
about social, political, and even
economic systems.
THIS BOOK alone justifies all
the expenditure which has been
put into. the Mental Health Re-
search Institute, and the Univer-
sity of Michigan can well be
proud of having nurtured a situa-
tion where research of this kind
was possible. +
Prof. Kenneth E. Boulding of
the economic department is di-
rector of the Center ;for, Re-
search on Conflict Resolution.
Hopefully he is also the first
in a series of faculty reviewers
of books of major interest writ-
ten by both members of the Uni-
versity community and academe
at large. Suggestions for reviews
and expressions of interest in
reviewing are welcome.

'I

4
4

(Defect)
(Bad)

Prisoner's Dilemma

1s

the cooperating player when the
other player choosesE the defect-
ive strategy, -in Figure 1 this is
-2; and T, the payoff to the de-
fecting player when the other
player chooses the cooperative
strategy. In Figure 1 this is +2.
In some respects the differences
(T-R), representing the arrows
d1 and d2 in Figure 1, and (P-S),
representing the arrows el and
e2, ale more significant. An im-
portant part of the experimental
strategy is of course to vary the
parameters and see what happens
to the results. A further set of
parameters which emerges out
of the study of repeated plays
relates to the succession of coop-

which are used. Not surprisingly,
there is a greater tendency to-
wards cooperation when the pay-
offs for this are greater.
THE EXPERIMENTAL data,
however, suggest some interesting
properties of the learning process
which might not otherwise have
been suspected and which need
to be explained. A very common.
pattern seems to be to start off
with trust and mutual coopera-
tion, and this is followed by in-
creasing defection and sometimes
a locking in in the D1D2 box..
Very frequently, however, there is
another learning process by which
cooperation increases again. We

Letters: Student Involvement in the 'U'

To the Editor:
A TTHE RISK of exposing my-
self to the merciless knives of
your writers, let me make some
suggestions about student involve-
ment in University affairs.
(1) Power and influence are
frequently not a matter of legal
voting rights, formal meeting ar-
rangements, or pressure tactics.
Persuasion can go a long way,
and it goes' farther if the facts,
the logic, and the evidence of
widespread informed support are
on your side. Has this been the
case with the proposed bookstore?
(2) Policies are more often
changed by a series of specific
decisions,. each decided on its
merits, than by formal votes on
policy. Hence, the students need
to focus on one thing at a time.
Perhaps the bookstore was an at-
tempt in this direction, if an un-
fortunate one. How about another
try?
(3) What is the point in asking
for a universal hunting license?
Any organization is bound to re-
sist requests for access to all its
files, perpetual discussions without
specified topics or agenda, or de-
tailed plans andrrules of opera-
tion. They have seen what a hos-
tile legislator or "investigating
committee' can do with such
freedom (license). Furthermore,

there is expense involved in long
discussions that can only be jus-
tified if there is a clearly identi-
fied problem, and some clear in-
terest or expertise represented by
the participants.
(4) Money mingles, and it shows
a lack of perception of the finan-
cial processes of any organization
to ask where funds from some
particular source go. This is like
writing on your church contribu-
tion that it must all go to mis-
sions. Such action only forces the
church to devote more of some-
one else's money to paying the
janitor. Anyhow, segregation of
funds is more likely to lead to in-
efficiency and waste than to
anything else.
THE OVERALL allocation of
all the funds is an important
policy matter, and one in which
students and faculty as well as
administration and Regents and
the Legislature have an interest.
In this area students might well
want to ' express opinions. But
opinions must, to be useful, be
about the choices which must be
made. Would we rather have a
$3 million theatre at a cost in
unrestricted University funds (po-
tentially available for something
else) of say $1.5 million, or $1.5
million more put into classrooms?
Or would we rather have a $2

million faculty club for which the
faculty put up $.5 million (to
dream up an example the facts
of which are probably wrong).
But if students want to get into
such discussions, they must un-
derstand both economics and ac-
counting, and invest the time to
become acquainted with the
choices that are actually in ex-
istence. And you may have ei-
ther to accept or check out
someone's assertion that the Leg-
islature will or will not appro-,
priate funds for some given pur-
pose, etc. Sniping or intermittent
heavy involvement followed by
complete absence from the scene
have too often characterized past
student involvement. For instance,
intermittent requests for the right
to attend meetings of the Senate
Advisory Committee and its sub-
committees seem to be followed
by failure of the delegates to at-
tend regularly.
I could suggest some campaigns
I'd like to see the students get
going on, but that would be out
of place here.
-James N. Morgan
Professor of Economics
Deferments
To tlhe Editor:
THE FOLLOWING statement
was approved Monday night
by the membership of Voice Po-
litical Party:
We are in fundamental opposi-
tion to what we see as an immor-
al, illegal and dehumanizing war
in Viet Nam. It is deplorable
that in such a conflict, fought
supposedly for "freedom," it is
domestic freedom and domestic
democracy which become sub-
ordinated to the mobilization ef-
forts.
Typical of this trend away from
democracy is the selection and
deferment policy with which the
military staffs its ranks. The
present system of defering college
students (2-S) and defense em-
ployes (Z-A) results in an im-
posed policy of blatant economic,
social and racial discrimination.
To a very large extent, a stu-
dent's educational level is a func-
tion of his economic, social and
racial background:
0 The expense of higher edu-
cation tends to discriminate
against the poor and the chil-
dren of non-college-educated par-
PDnts_ At +hg UTniversity (in, the

trade unions. Only 25 per cent of
the student body had parents
with less than a college education.
" Racially-discriminatory back-
grounds, cultural and residen-
tial isolation, and low academic
motivation resulting from these
make it difficult for those who do
manage to find their way into the
University to remain there. The
high dropout rate of the students
recruited through the Opportuni-
ty Awards Program points out
the much deeper problems en-
countered by the victims of dis-
crimination and deprivation. The
mentality of academic competi-
tion, so essential to successful
performance with the modern
complex university is largely the
result of the sort of background
to which the racially and socially
discriminated have not been ex-
posed.
* The effect of occupational de-
ferments is to exempt those who
derive their existence from the:
profits of an ever-expanding mili-
tary-industrial complex from the
obligation to fight to preserve it.
THEREFORE, as a result of
these descriminatory effects of
the 2-S and 2-A deferments, we
restate our opposition to them
and to the mentality which pre-
serves them.
In line with this general atti-
tude, we feel that there is a ba-
sic tension between the two major
roles of the university in society:
1) The functional--the univer-
sity must mass produce
trained personnel and tech-
nicians to staff the offices
of industry and government
and the military.
2) The intellectual-the univer-
sity as a community of
scholars committed to a rig-
orous search and to an
equally rigorous analysis
and criticism of their own
society.
Clearly, both of these functions
must be carried out, but there
are numerous times when those
who run the University in rela-
tive isolation from the needs of
the student body must choose to
lean in one of these two direc-,
tions. We feel that the present
issue of ranking students for the
purpose of drafting the lower 25r
or so per cent is one which de-
mands that the administration
lean quite unequivocally to the

WE FEEL that students in-
volved in extracurricular activi-
ties; students who must work to
support themselves; students who
take time from studies for crea-
tive pursuits; and students who
simply prefer to concentrate
greater efforts on their own pri-
vate learning and learning meth-
ods have at least as much right
to consider themselves fully legit-
imate members of the University
community as those who do noth-
ing to endanger their grade-
point averages.
The issue at stake in the ques-
tion of ranking is not one of the
inviolability of the 2-S deferment,
but rather, it is the far more im-
portant one of imposed social
values. To reward the competi-
tive and the ,privileges as well as
the academically proficient while
at the same time rejecting and
positively penalizing the poor, the
underprivileged, the uncompeti-
tive, and the individualistically
creative is to reinforce the very
values and class divisions which
we find most objectionable in our
society.
While it does seem necessary
for the University to train the
functionaries of society, it is
not at all desirable for it to
serve as an administrative arm of
the Selective Service System. The
types of pressures, quotas, and
criteria which form an integral
part of the SSS are largely in-
imical to the aims and atmos-
phere which should prevail in a
university community.
THE SITUATION is clear. The
2-S deferment is discriminatory;
ranking students is a further ex-
tension of this discriminatory
policy. Therefore, we reject the
2-S deferment and propose in-
stead that all those liable for
service be given the option to cre-
ate instead of destroy. We are
willing to risk our lives in the
South of the United States in
the cause of building freedom.
But we are not willing to fight
against people trying to achieve
freedom from foreign domination
in Viet Nam.
TO THIS END, WE PROPOSE
THAT THE UNIVERSITY:
-Reject the 2-S deferment as
discriminatory, and use its re-
search and educative facilities to
enlighten the people of the state
to this discrimination;

i


FEIFFER

L0kAT TH
WOAY 'TO56s
60' fIVE!
PR M.FUU OF
I COU0Q
"HAT-
.,WAY F!/3V
60Y9LA1K!
45T. ME~-

h .,
H - 11

Aor
1E-

AQP LUOOK AT T N UTt
KU iTCOFA ,
'5110P%3
7NEH
AIUZ.
TR 1R 2'

ANA~
I'D
K1UM.

4

SOS' A
AT ",

lcv

AWP 112


I'5 A

I

60T

\ /