Journal oj Personality and Social Psychology

1973, Vol. 26, No. 3, 446-456

THE YIN AND YANG OF PROGRESS IN SOCIAL PSYCHOLOGY:

SEVEN KOAN 1

WILLIAM J. McGUlRE-
Yale University

We describe the current dissatisfactions with the paradigm that has recently
guided experimental social psychology—testing of theory-derived hypotheses
by means of laboratory manipulational experiments. The emerging variant
of doing field experiments does not meet the criticisms. It is argued that an
adequate new paradigm will be a more radical departure involving, on the
creative side, deriving hypotheses from a systems theory of social and cogni-
tive structures that takes into account multiple and bidirectional causality
among social variables. On the critical side, its hypotheses testing will be
done in multivariate correlational designs with naturally fluctuating variables.
Some steps toward this new paradigm are described in the form of seven koan.

THE PARADIGM RECENTLY GUIDING which I hasten to translate as follows

EXPERIMENTAL SOCIAL On a temple bell
PSYCHOLOGY Settled, asleep,
When the XlXth Congress met 3 years A butterfly.
ago in London, and certainly a half-dozen We social psychology researchers know all
years back at the Moscow Congress, social too well that the peaceful temple bell on
psychology appeared to be in a golden age. It which we were then displaying ourselves has
was a prestigious and productive area in now rudely rung. During the past half-dozen
which droves of bright young people, a suf-
years, the vibrations which could be vaguely
ficiency of middle-aged colonels, and a few
sensed at the time of the Moscow meeting
grand old men were pursuing their research
with a confidence and energy that is found in have gathered force. Now the temple bell has
those who know where they are going. Any tolled and tolled again, rudely disturbing
moments of doubt we experienced involved the stream of experimental social psychologi-
anxiety as to whether we were doing our thing cal research and shaking the confidence of
well, rather than uncertainty as to whether it many of us who work in the area.
needed to be done at all. The first half of this paper is devoted to
The image of these golden boys (and a few, describing the three successive waves of this
but all too few, golden girls) of social psy- current history. First, I shall describe the
chology, glowing with confidence and chutz- experimental social psychology paradigm that
pah only 6 years back at the Moscow Con- has recently guided our prolific research.
gress, blissfully unaware of the strident at- Second, I shall discuss why this recent para-
tacks which were soon to strike confusion into
digm is being attacked and what, super-
the field, brings to mind a beautiful haiku
of Buson that goes ficially at least, appears to be emerging in its
place. Third, I shall say why I feel the seem-
Tsurigane-ni ingly emerging new paradigm is as inadequate
Tomarite nemuru as the one we would replace. Then, in the
Kocho kana second half of this paper I shall offer, in the
1
This paper is based on an address given at the form of seven koan, my prescriptions for a
Nineteenth Congress of the International Union of new paradigm, more radically different from
Scientific Psychology at Tokyo in August 1972. the recent one, but more in tune with the
2
Requests for reprints should be sent to the
author, Department of Psychology, Yale University, times and the march of history than is the
333 Cedar Street, New Haven, Connecticut 06510. variant that is supposedly emerging.
446
 YIN AND YANG OF PROGRESS IN SOCIAL PSYCHOLOGY: SEVEN KOAN 447

The Old Paradigm from people inside and outside social psychol-
ogy, reflecting both the increasing social
What was the experimental social psychol- concern of researchers themselves and the
ogy paradigm which until recently had been demands of an articulate public for greater
unquestioningly accepted by the great ma- payoff from expensive scientific research.
jority of us but which now is being so vigor-
While many of us still insist with Lewin that
ously attacked? Like any adequate paradigm "There is nothing so practical as a good
it had two aspects, a creative and a critical theory," the extent to which the pendulum
component (McGuire, 1969, pp. 22-25). By
has swung from the theoretically relevant
the creative aspect, I mean the part of our
toward the socially relevant pole is shown in
scientific thinking that involves hypothesis
the recent upsurge of publications on socially
generation, and by the critical aspect, I mean
important topics of ad hoc interest, such as
the hypothesis-testing part of our work. bystander intervention, the use of local space,
The creative aspect of the recent paradigm the mass media and violence, the determi-
inclined us to derive our hypotheses from nants of love, responses to victimization, non-
current theoretical formulations. Typically, verbal communication, etc.
these theoretical formulations were borrowed At least as strong and successful an assault
from other areas of psychology (such as the
has been launched on the critical aspect of
study of psychopathology or of learning and the recent paradigm, namely, the notion that
memory), though without the level of refine-
hypotheses should be tested by manipula-
ment and quantification which those theories
tional laboratory experiments. It has been
had reached in their fields of origin.
urged that laboratory experiments are full of
The critical, hypothesis-testing aspect of
artifacts (such as experimenter bias, demand
the recent paradigm called for manipulational
character, evaluation apprehension, etc.)
experiments carried out in the laboratory. The
which make their results very hard to inter-
experimental social psychologist attempted to
pret. Ethical questions also have been raised
simulate in the laboratory the gist of the
against the laboratory social experiments on
situation to which he hoped to generalize, and
the grounds that they expose the participants
he measured the dependent variable after
to an unacceptable amount of deception, co-
deliberately manipulating the independent
ercion, and stress.
variable while trying to hold constant all other
In place of the laboratory manipulational
factors likely to affect the social behavior
experiment, there has been a definite trend
under study. In brief, the recent paradigm
toward experiments conducted in field set-
called for selecting our hypotheses for their
tings and toward correlational analysis of
relevance to broad theoretical formulations
data from naturalistic situations. A variety of
and testing them by laboratory manipulational
recent methodological advances (which we
experiments. McGuire (1965) presented an
shall list under Koan 5) has made alterna-
emphatic assertion of this recent paradigm in
tive hypothesis-testing procedures more at-
its heyday. tractive.
Assaults on the Old Paradigm The attacks on the old paradigm of theory-
derived hypotheses tested in laboratory ma-
During the past several years both the cre- nipulational experiments have certainly shaken
ative and the critical aspects of this experi- confidence in that approach. At the same time,
mental social psychology paradigm have come there is some suggestion of an emerging new
under increasing attack. The creative aspect paradigm which has as its creative aspect the
of formulating hypotheses for their relevance derivation of new hypotheses for their ad hoc
to theory has been denounced as a mandarin interest and social relevance. And in its criti-
activity out of phase with the needs of our cal aspect, this new paradigm involves testing
time. It has been argued that hypotheses these hypotheses by field experiments and,
should be formulated for their relevance to where necessary, by the correlational analy-
social problems rather than for their rele- sis of naturalistic data. McGuire (1967, 1969)
vance to theoretical issues. Such urgings come described in more detail the worries about
448 WILLIAM J. McGuiRE

the recent paradigm and the nature of the zations. The simple sequential model had its
purportedly emerging one. Higbee and Wells uses, but these have been largely exploited in
(1972) and Fried, Gumpper, and Allen past progress, and we must now deal with the
(1973) suggested that reports by McGuire, complexities of systems in order to continue
by Sears and Abeles (1969), etc., of the de- the progress on a new level.
mise of the recent paradigm may be exag- The real inadequacy of the theory-derived
gerated, but perhaps they have underesti- hypotheses of the recent paradigm is not, as
mated the time that must intervene before a those now advocating socially relevant hy-
change of vogue by the leaders shows up in potheses insist, that it focused on the wrong
mass analysis of the methods used in pub- variables (those that were theory rather than
lished research. problem relevant). Rather, the basic short-
coming of the theory-relevant and the socially
MORE BASIC QUESTIONS REGARDING BOTH
relevant hypotheses alike is that they fail to
THE RECENT AND EMERGING PARADIGMS
come to grips with the complexities with
My own position on the relative merits of which the variables are organized in the indi-
the recent paradigm and this supposedly vidual and social systems.
emerging new paradigm is a complex and
developing one which I have detailed in print Inadequacies of the Critical Aspect of the
(McGuire, 196S, 1967, 1969) so the reader Recent Paradigm
will be spared here a recital of my Byzantine The critical, hypothesis-testing aspect of
opinions on this issue. Instead, I am raising the purportedly emerging paradigm also has
the more fundamental issue of whether or not the defect of being but a minor variant of
both the recent and the seemingly emerging the recent experimental social psychology
paradigms which I have just described fail to paradigm rather than the fundamental de-
come to grips with the deeper questions which parture which is called for. Let me first de-
lie behind our present unease. It seems to me scribe some of the deep epistemological un-
that any truly new paradigm that ultimately easiness some of us have been expressing
arises from the present unrest is going to be about the manipulational laboratory experi-
more radically different from the recent one ment that was the hypothesis-testing proce-
than is the supposedly emerging paradigm I dure of the recent paradigm. The crux of this
have just depicted. It will represent a more objection is that we social psychologists have
fundamental departure on both the creative tended to use the manipulational laboratory
and the critical sides. experiment not to test our hypotheses but to
demonstrate their obvious truth. We tend to
Inadequacies on the Creative Side start off with an hypothesis that is so clearly
The switch from theory relevance to social true (given the implicit and explicit assump-
relevance as the criterion in the creative, hy- tions) and which we have no intention of re-
pothesis-generating aspect of our work seems jecting however the experiment comes out.
to me to constitute only a superficial cosmetic Such a stance is quite appropriate, since the
change that masks rather than corrects the hypothesis by its meaningfulness and plausi-
basic problem. Socially relevant hypotheses, bility to reasonable people is tautologically
no less than theoretically relevant hypothe- true in the assumed context. As Blake said,
ses, tend to be based on a simple linear proc- "Everything possible to be believ'd is an
ess model, a sequential chain of cause and image of truth."
effect which is inadequate to simulate the true The area of interpersonal attraction will
complexities of the individual's cognitive sys- serve to illustrate my point. The researcher
tem or of the social system which we are might start off with a really obvious proposi-
typically trying to describe. Such simple a- tion from bubba-psychology, such as "The
affects-b hypotheses fail to catch the com- more someone perceives another person as
plexities of parallel processing, bidirectional having attitudes similar to his own, the more
causality, and reverberating feedback that he tends to like that other person." Or a
characterize both cognitive and social organi- somewhat more flashy researcher, a little hun-
 YIN AND YANG OF PROGRESS IN SOCIAL PSYCHOLOGY: STCVTCN KOAN 449

grier for novelty, might hypothesize the oppo- the obvious correctness of our hypothesis can
site. That is, he could look for certain circum- be demonstrated.
stances in which the generally true, obvious It is this kind of epistemological worry
hypothesis would obviously be reversed. He about manipulational laboratory experiments
might hypothesize exceptional circumstances that a half-dozen years back caused a number
where attitudinal similarity would be anxiety of observers (e.g., McGuire, 1967) to urge
arousing and a source of hostility; for exam- social psychology to search for interrelations
ple, if one loves one's wife, then one might among naturally varying factors in the world
actually dislike some other man to the extent outside the laboratory. Out of these urgings
that one perceives that other as also loving has come the critical aspect of the apparently
one's wife. Or another exceptional reversal emerging paradigm which I have described
might be that some people may think so above, calling for research in the field rather
poorly of themselves that they think less well than in the laboratory.
of another person to the extent that the other
person is like themselves. If the negative rela- Inadequacies of the Critical Aspects of the
tionship is not found, we are likely to con- Purportedly Emerging New Field-Experiment
clude that the person did not have a suffi- Paradigm
ciently low self-image, not that the hypothesis Recently, I have come to recognize that
is wrong. Both the original obvious hypothesis this flight from the laboratory manipulational
and the obvious reversed hypothesis are rea- experiment to the field study, which I myself
sonable and valid in the sense that if all our helped to instigate, is a tactical evasion which
premises obtained, then our conclusion would fails to meet the basic problem. We would
pretty much have to follow. grant that in the field we put the question to
Experiments on such hypotheses naturally nature in a world we never made, where the
turn out to be more like demonstrations than context factors cannot be so confounded by
tests. If the experiment does not come out our stage management proclivities as they
"right," then the researcher does not say that were in the laboratory. But in this natural
the hypothesis is wrong but rather that some- world research, the basic problem remains that
thing was wrong with the experiment, and he we are not really testing our hypotheses.
corrects and revises it, perhaps by using more Rather, just as in the laboratory experiment
appropriate subjects, by strengthening the in- we were testing our stage-managing abilities,
dependent variable manipulation, by blocking in the field study we are testing our ability
off extraneous response possibilities, or by as "finders," if I may use a term from real
setting up a more appropriate context, etc. estate and merchandising. When our field
Sometimes he may have such continuous bad test of the hypothesis does not come out
luck that he finally gives up the demonstra- correctly, we are probably going to assume
tion because the phenomenon proves to be so not that the hypothesis is wrong but that
elusive as to be beyond his ability to demon- we unwisely chose an inappropriate natural
strate. The more persistent of us typically setting in which to test it, and so we shall try
manage at last to get control of the experi- again to test it in some other setting in which
mental situation so that we can reliably dem- the conditions are more relevant to the hy-
onstrate the hypothesized relationship. But pothesis. Increasing our own and our graduate
note that what the experiment tests is not students' critical skill will involve making us
whether the hypothesis is true but rather not better hypothesis testers or better stage
whether the experimenter is a sufficiently in- managers but rather better finders of situa-
genious stage manager to produce in the lab- tions in which our hypotheses can be demon-
oratory conditions which demonstrate that an strated as tautologically true. Though I shall
obviously true hypothesis is correct. In our not pursue the point here, other objections to
graduate programs in social psychology, we the laboratory experiment, including ethical
try to train people who are good enough and methodological considerations, that have
stage managers so that they can create in the been used (McGuire, 1969) to argue for more
laboratory simulations of realities in which field research could similarly be turned
450 WILLIAM J. McGuiRE

against experiments conducted in the natural will be hammered out by theoretically and
environment. empirically skilled researchers in a hundred
What I am arguing here is that changing eyeball-to-eyeball confrontations of thought
from a theory-relevant to a socially relevant with data, all the while obscured by a thou-
criterion for variable selection does not con- sand mediocre and irrelevant studies which
stitute a real answer to the basic problem will constitute the background noise in which
with the creative aspect of our recent social the true signal will be detected only gradu-
psychology paradigm. And again, the switch ally. Trying to predict precisely what new
from laboratory to field manipulation does pardigm will emerge is almost as foolish as
not meet the basic objection to the critical trying to control it.
aspect of the old paradigm. Neither the recent But there is a subsidiary task with which I
paradigm nor the supposedly emerging one feel more comfortable and to which I shall
really supplies the answer to our present devote the rest of this paper. I have come to
needs. The discontent is a quite healthy one, feel that some specific tactical changes should
and we should indeed be dissatisfied with the be made in our creative and critical work in
recent paradigm of testing theory-derived social psychology so as to enhance the mo-
hypotheses by means of laboratory manipula- mentum and the ultimate sweep of this wave
tional experiments. But our healthy discon- of the future, whatever form it may take. I
tent should carry us to a more fundamentally shall here recommend a few of these needed
new outlook than is provided by this sup- innovations and correctives, presenting them
posedly emerging variant paradigm of testing as koans and commentaries thereon, to mask
socially relevant hypotheses by experiments my own uncertainties.
in natural settings.
Koan 1: The Sound of One Hand Clapping
SOURCES OF THE NEW SOCIAL PSYCHOLOGY . . , and the Wrong Hand
The Ultimate Shape of the New Paradigm One drastic change that is called for in our
What I have written in the previous section teaching of research methodology is that we
suggests my general vision of what the more should emphasize the creative, hypothesis-
radically different new paradigm for social formation stage relative to the critical, hy-
psychology will look like. On the creative pothesis-testing stage of research. It is my
side, it will involve theoretical models of the guess that at least 90% of the time in our
cognitive and social systems in their true current courses on methodology is devoted
multivariate complexity, involving a great to presenting ways of testing hypotheses and
deal of parallel processing, bidirectional rela- that little time is spent on the prior and more
tionships, and feedback circuits. Since such important process of how one creates these
complex theoretical formulations will be far hypotheses in the first place. Both the crea-
more in accord with actual individual and tion and testing of hypotheses are important
social reality than our present a-affects-b parts of the scientific method, but the cre-
linear models, it follows that theory-derived ative phase is the more important of the two.
hypotheses will be similar to hypotheses se- If our hypotheses are trivial, it is hardly
lected for their relevance to social issues. worth amassing a great methodological arse-
Correspondingly, the critical aspect of this nal to test them; to paraphrase Maslow, what
new paradigm involves hypothesis testing by is not worth doing, is not worth doing well.
multivariate time series designs that recog- Surely, we all recognize that the creation of
nize the obsolescence of our current simplistic hypotheses is an essential part of the scien-
a-affects-b sequential designs with their dis- tific process. The neglect of the creative phase
tinctions between dependent and independent in our methodology courses probably comes
variables. neither from a failure to recognize its impor-
But I feel somewhat uncomfortable here in tance nor a belief that it is trivially simple.
trying to describe in detail what the next, Rather, the neglect is probably due to the
radically different paradigm will look like. It suspicion that so complex a creative process
 YIN AND YANG or PROGRESS IN SOCIAL PSYCHOLOGY: SEVEN KOAN 451

as hypothesis formation is something that own work on deriving hypotheses about tech-
cannot be taught. niques for inducing resistance to persuasion,
I admit that creative hypothesis formation where I formulated hypotheses by analogy
cannot be reduced to teachable rules, and with the biological process of inoculating the
that there are individual differences among us person in advance with a weakened form of
in ultimate capacity for creative hypothesis the threatening material, an idea suggested in
generation. Still, it seems to me that we have earlier work by Janis and Lumsdaine. A
to give increased time in our own thinking fourth creative procedure is the hypothetico-
and teaching about methodology to the hy- deductive method, where one puts together a
pothesis-generating phase of research, even at number of commonsensical principles and
the expense of reducing the time spent dis- derives from their conjunction some interest-
cussing hypothesis testing. In my own meth- ing predictions, as in the Hull and Hovland
odology courses, I make a point of stressing mathematico-deductive theory of rote learn-
the importance of the hypothesis-generating ing, or the work by Simon and his colleagues
phase of our work by describing and illustrat- on logical reasoning. The possibility of com-
ing at least a dozen or so different approaches puter simulation has made this hypothesis-
to hypothesis formation which have been used generating procedure increasingly possible and
in psychological research, some of which I popular.
can briefly describe here, including case study, A fifth way of deriving hypotheses might
paradoxical incident, analogy, hypothetico- be called the functional or adaptive approach,
deductive method, functional analysis, rules of as when Hull generated the principles on
thumb, conflicting results, accounting for ex- which we would have to operate if we were to
ceptions, and straightening out complex be able to learn from experience to repeat
relationships. successful actions, and yet eventually be able
For example, there is the intensive case to learn an alternative shorter path to a goal
study, such as Piaget's of his children's cog- even though we have already mastered a
nitive development or Freud's mulling over longer path which does successfully lead us to
and over of the Dora or the Wolf Man case that goal. A sixth approach involves analyz-
or his own dreams or memory difficulties. ing the practitioner's rule of thumb. Here
Often the case is hardly an exceptional one— when one observes that practitioners or
for example, Dora strikes me as a rather mild craftsmen generally follow some procedural
and uninteresting case of hysteria—so that it rule of thumb, we assume that it probably
almost seems as if any case studied intensively works, and one tries to think of theoretical
might serve as a Rorschach card to provoke implications of its effectiveness. One does not
interesting hypotheses. Perhaps an even surer have to be a Maoist to admit that the basic
method of arriving at an interesting hypothe- researcher can learn something by talking to
sis is to try to account for a paradoxical inci- a practitioner. For example, one's pro-
dent. For example, in a study of rumors circu- grammed simulation of chess playing is im-
lating in Bihar, India, after a devastating proved by accepting the good player's heur-
earthquake, Prasad found that the rumors istic of keeping control of the center of the
tended to predict further catastrophes. It board. Or one's attitude change theorization
seemed paradoxical that the victims of the can be helped by noting the politician's and
disaster did not seek some gratification in advertiser's rule that when dealing with pub-
fantasy, when reality was so harsh, by gen- lic opinion, it is better to ignore your opposi-
erating rumors that would be gratifying rather tion than to refute it. These examples also
than further disturbing. I believe that at- serve to remind us that the practitioner's rule
tempting to explain this paradox played a of thumb is as suggestive by its failures as by
more than trivial role in Festinger's formula- its successes.
tion of dissonance theory and Schachter's A seventh technique for provoking new hy-
development of a cognitive theory of emotion. potheses is trying to account for conflicting
A third creative method for generating results. For example, in learning and attitude
hypothesis is the use of analogy, as in my change situations, there are opposite laws of
452 WILLIAM J. McGuiRis

primacy and of recency, each of which some- than the exception that the effects act back
times seems valid; or in information integra- on the causal variables. Hence, students of
tion, sometimes an additive or sometimes an cognitive and social processes must be en-
averaging model seems more appropriate. The couraged to think big, or rather to think com-
work by Anderson trying to reconcile these plexly, with conceptual models that involve
seeming conflicts shows how provocative a parallel processing, nets of causally inter-
technique this can be in generating new the- related factors, feedback loops, bidirectional
ories. An eighth creative method is accounting causation, etc.
for exceptions to general findings, as when If we and our students are to begin think-
Hovland tried to account for delayed action ing in terms of these more complex models,
effect in opinion change. That is, while usually then explicit encouragement is necessary since
the persuasive effect of communications dissi- the published literature on social and cogni-
pates with time, Hovland found that occa- tive processes is dominated by the simple
sionally the impact actually intensifies over linear models, and our students must be
time, which provoked him to formulate a warned against imprinting on them. But our
variety of interesting hypotheses about de- encouragement, while necessary, will not be
layed action effects. A ninth creative tech- sufficient to provoke our students into the
nique for hypothesis formation involves re- more complex theorizing. We shall all shy
ducing observed complex relationships to away from the mental strain of keeping in
simpler component relationships. For exam- mind so many variables, so completely inter-
ple, the somewhat untidy line that illustrates related. Moreover, such complex theories al-
the functional relationship between visual low so many degrees of freedom as to threaten
acuity and light intensity can be reduced to the dictum that in order to be scientifically
a prettier set of rectilinear functions by hy- interesting, a theory must be testable, that is,
pothesizing separate rod and cone processes, disprovable. These complex theories, with
a logarithmic transformation, a Blondel-Rey- their free-floating parameters, seem to be
type threshold phenomenon to account for adjustable to any outcome.
deviations at very low intensities, etc. Hence, we have to give our students skill
But our purpose here is not to design a and confidence and be role models to encour-
methodology course, so it would be inappro- age them to use complex formulations. To this
priate to prolong this list. Let me say once end we have to give greater play to techniques
again, to summarize our first koan, that we like computer simulation, parameter estima-
have listened too long to the sound of one tion, multivariate time series designs, path
hand clapping, and the less interesting hand analysis, etc. (as discussed further in Koan 5
at that, in confining our methodology discus- below), in our graduate training programs.
sion almost exclusively to hypothesis testing.
It is now time to clap more loudly using the Koan 3: Observe. But Observe People Not
other hand as well by stressing the impor- Data
tance of hypothesis generation as part of In our father's house there are many rooms.
psychological methodology. In the total structure of the intelligentsia,
there is a place for the philosopher of mind
Koan 2: In This Nettle Chaos, We Discern and the social philosopher, as well as for the
This Pattern, Truth scientific psychologist. But the scientific psy-
I stress here the basic point that our cog- chologist can offer something beside and be-
nitive systems and social systems are com- yond these armchair thinkers in that we not
plex and that the currently conventional sim- only generate delusional systems, but we go
ple linear process models have outlived their further and test our delusional systems against
heuristic usefulness as descriptions of these objective data as well as for their subjective
complex systems. In our actual cognitive and plausibility. Between the philosopher of mind
social systems, effects are the outcome of and the scientific psychologist, there is the
multiple causes which are often in complex difference of putting the question to nature.
interactions; moreover, it is the rule rather Even when our theory seems plausible and so
 YIN AND YANG OF PROGRESS IN SOCIAL PSYCHOLOGY: SEVEN KOAN 453

ingenious that it deserves to be true, we are Koan 4: To See the Future in the Present,
conditioned to follow the Cromwellian dictum Find the Present in the Past
(better than did the Lord Protector himself) One idea whose time has come in social
to consider in the bowels of Christ that we psychology is the accumulation of social data
may be wrong. archives. Leaders of both the social science
But I feel that in our determination to and the political establishments have recog-
maintain this difference we have gone too far. nized that we need a quality-of-life index
In our holy determination to confront reality (based perhaps on trace data, social records,
and put our theory to the test of nature, we self-reports obtained through survey research,
have plunged through reality, like Alice etc.). Such social archives will also include
through the mirror, into a never-never land
data on factors which might affect subjective
in which we contemplate not life but data. All
too often the scientific psychologist is observ- happiness, and analyses will be done to tease
ing not mind or behavior but summed data out the complex interrelations among these
and computer printout. He is thus a self- important variables. The need for such ar-
incarcerated prisoner in a platonic cave, where chives is adequately recognized; the interest
he has placed himself with his back to the and advocacy may even have outrun the
outside world, watching its shadows on the talent, energy, and funds needed to assemble
walls. There may be a time to watch shadows them.
but not to the exclusion of the real thing. In this growing interest in social data ar-
Perhaps Piaget should be held up as a role chives, one essential feature has been ne-
model here, as an inspiring example of how a glected, namely, the importance of obtaining
creative mind can be guided in theorizing by
time series data on the variables. While it
direct confrontation with empirical reality.
Piaget's close observation of how the develop- will be useful to have contemporaneous data
ing human mind grapples with carefully de- on a wide variety of social, economic, and
vised problems was much more conducive to psychological variables, the full exploitation
his interesting theorizing than would have of these data becomes possible only when we
been either the armchair philosopher's test of have recorded them at several successive
subjective plausibility or the scientific en- points in time. Likewise, while a nationwide
trepreneur's massive project in which as- survey of subjective feelings and attitudes is
sistants bring him computer printout, inches quite useful for its demographic breakdowns
thick. at one point in time, the value of such a
The young student typically enters gradu- social survey becomes magnified many times
ate study wanting to do just what we are pro- when we have it repeated at successive points
posing, that is, to engage in a direct confron- in history. It is only when we have the time
tation with reality. All too often, it is our series provided by a reconstructed or pre-
graduate programs which distract him with planned longitudinal study that we can apply
shadows. Either by falling into the hands of the powerful methodology of time series
the humanists, he is diverted into subjectiv- analyses which allow us to reduce the com-
ism and twice-removed scholarly studies of plexity of the data and identify causality.
what other subjectivists have said; or, if he Hence, my fourth koan emphasizes the use-
falls under the influence of scientific psycholo- fulness of collecting and using social data
gists, he becomes preoccupied with twice- archives but adds that we should collect data
removed sanitized data in the form of com- on these variables not only at a single con-
puter printout. I am urging that we restruc- temporaneous point in time, but also that we
ture our graduate programs somewhat to keep should set up a time series by reconstructing
the novice's eye on the real rather than dis- measures of the variables from the recent and
tracting and obscuring his view behind a wall distant past and prospectively by repeated
of data. surveys into the future.
454 WILLIAM J. McGuiRE

Koan 5: The New Methodology Where use of time series causal analyses like the
Correlation Can Indicate Causation cross-lag panel design. More training is also
needed in computer simulation and techniques
If we agree that the simple linear sequence
of parameter estimation.
model has outlived its usefulness for guiding
Mastery of these techniques will not be
our theorizing about cognitive and social sys-
easy. Because we older researchers have al-
tems, then we must also grant that the labora-
ready mastered difficult techniques which
tory manipulational experiment should not
have served us well, we naturally look upon
be the standard method for testing psycho-
this retooling task with something less than
logical hypotheses. But most graduate pro-
enthusiasm. We have worked hard and en-
grams and most of the published studies (Hig-
dured much; how much more can be asked of
bee & Wells, 1972) focus disproportionately
us? But however we answer that question
on descriptive and inferential statistics ap-
regarding our obligation to master these tech-
propriate mainly to the linear models from the
niques ourselves, we owe it to our students to
recent paradigm. The methods taught and
make the newer techniques available to those
used are characterized by obsolescent pro-
who wish it, rather than requiring all students
cedures, such as rigorous distinction between
to preoccupy themselves with the old tech-
dependent and independent variables, two-
niques which have served us so well in reach-
variable or few-variable designs, an assump-
ing the point from which our students must
tion of continuous variables, the setting of
now proceed.
equal numbers and equal intervals, etc.
It seems to me that we should revise the Koan 6: The Riches of Poverty
methodology curriculum of our graduate pro-
grams and our research practice so as to make The industrial countries, where the great
us better able to cope with the dirty data of bulk of psychological research is conducted,
the real world, where the intervals cannot be have in the past couple of years suffered eco-
preset equally, where the subjects cannot be nomic growing pains which, if they have not
assigned randomly and in the same number, quite reduced the amount of funds available
and where continuous measures and normal for scientific research, at least have reduced
distributions typically cannot be obtained. In the rate at which these funds have been grow-
previous writings in recent years, I have called ing. In the United States, at least, the last
attention to advances in these directions which couple of years have been ones of worry about
T mention here (McGuire, 1967, 1969), and leveling scientific budgets. It is my feeling
Campbell (1969) has been in the forefront that the worry exceeds the actuality. In the
in devising, assembling, and using such pro- United States' situation, psychology has in
cedures. fact suffered very little as compared with our
Our graduate programs should call the stu- sister sciences. As an irrepressible optimist I
dent's attention to new sources of social data, am of the opinion that not only will this
such as archives conveniently storing informa- privileged position of psychology continue but
tion from public opinion surveys, and to non- also that the budgetary retrenchment in the
reactive measures of the unobtrusive trace other fields of science is only a temporary one
type discussed by Webb and his colleagues. and that, in the long run, the social invest-
Our students should also be acquainted ment in scientific research will resume a
with the newer analytic methods that make healthy, if not exuberant, rate of growth. I
more possible the reduction of the complex recognize that this optimism on my part will
natural field to a manageable number of un- do little to cheer scientists whose own re-
derlying variables whose interrelations can search programs have been hard hit by the
be determined. To this end, we and our stu- financial cuts. To my prediction that in the
dents must have the opportunity to master long run social investment in science will grow
new techniques for scaling qualitative data, again after this temporary recession, they
new methods of multivariate analysis, such as might point out (like Keynes) that in the
those devised by Shepard and others, and the long run we shall all be dead.
 YIN AND YANG OF PROGRESS IN SOCIAL PSYCHOLOGY: SEVEN KOAN 455

I persist in my Dr. Pangloss optimism that senior researcher from wasting his efforts on
things are going to turn out well and even meeting the payroll rather than observing the
engage in gallows humor by saying that what phenomena.
psychological research has needed is a good I am urging here, then, that if the bud-
depression. I do feel that during the recent getary cutbacks continue instead of running
period of affluence when we in the United ever faster on the Big-Science treadmill, we
States could obtain government funds for make the best of the bad bargain by changing
psychological research simply by asking, we our research organization, our mode of work-
did develop some fat, some bad habits, and ing, and our priorities. I would suggest that
some distorted priorities which should now be rather than fighting for a bigger slice of the
corrected. While we could have made these diminishing financial pie, we redirect our ef-
corrections without enforced poverty, at forts somewhat. We should rediscover the
'least we can make a virtue of necessity by gratification of personally observing the phe-
using this time of budgetary retrenchment to nomena ourselves and experiencing the relief
cut out some of the waste and distraction so of not having to administer our research em-
that we shall emerge from this period of re- pire. Also, I think we should spend a greater
trenchment stronger than we entered it. portion of our time trying to interpret and
The days of easy research money sometimes integrate the empirical relationships that have
induced frenzies of expensive and exhausting been turned up by the recent deluge of stud-
activity. We hired many people to help us, ies, rather than simply adding new, undi-
often having to dip into less creative popula- gested relationships to the existing pile.
tions, and to keep them employed the easiest
thing to do was to have them continue doing Koan 7: The Opposite of a Great Truth is
pretty much what we had already done, re- Also True
sulting in a stereotyping of research and a What I have been prescribing above is not
repetitious output. It tended to result in the a simple, coherent list. A number of my urg-
collection of more data of the same type and ings would pull the field in opposite direc-
subjecting it to the same kinds of analyses as tions. For example, Koan 1 urges that our
in the past. It also motivated us to churn out methodology courses place more emphasis on
one little study after another, to the neglect the creative hypothesis-forming aspect of re-
of the more solitary and reflective intellectual search even at the cost of less attention to
activity of integrating all the isolated findings the critical, hypothesis-testing aspect, but
into more meaningful big pictures. then in Koan 5 I urged that we, or at least
Affluence has also produced the complex our students, master a whole new pattern of
research project which has removed us from hypothesis-testing procedures. Again, Koan 3
reality into the realm of data as I discussed urges that we observe concrete phenomena
in Koan 3. The affluent senior researcher rather than abstract data, but Koan 4 favors
often carried out his work through graduate assembling social data archives that would
assistants and research associates, who, in reduce concrete historical events to abstract
turn, often have the actual observations done numbers. My prescriptions admittedly ride off
by parapsychological technicians or hourly in opposite directions, but let us remember
help, and the data they collect go to card- that "consistency is the hobgoblin of little
punchers who feed them into computers, minds."
whose output goes back to the research asso- That my attempt to discuss ways in which
ciate, who might call the more meaningful our current psychological research enterprise
outcome to the attention of the senior re- could be improved has led me in opposite
searcher, who is too busy meeting the payrolls directions does not terribly disconcert me. I
to control the form of the printout or look remember that Bohr has written, "There are
diligently through it when it arrives. A cut- trivial truths and great truths. The opposite
back in research funds might in some cases of a trivial truth is plainly false. The oppo-
divert these assistants into more productive site of a great truth is also true." The same
and satisfying work while freeing the creative paradox has appealed to thinkers of East and
456 WILLIAM J. McGuiRE

West alike since Sikh sacred writings advise nothing seems certain but everything seems
that if any two passages in that scripture possible.
contradict one another, then both are true. I began this talk by describing the proud
The urging at the same time of seemingly and placid social psychology of a half-dozen
opposed courses is not necessarily false. It years back, just before the bell tolled, as sug-
should be recognized that I have been giving gesting Buson's beautiful sleeping butterfly.
mini-directives which are only a few parts of I close by drawing upon his disciple, the
the total system which our psychological re- angry young man Shiki, for a related but dy-
search and research training should involve. namically different image of the new social
Indeed, I have specified only a few compo- psychology which is struggling to be born.
nents of such a total research program. Any Shiki wrote a variant on Buson's haiku as
adequate synthesis of a total program must be follows:
expected to contain theses and antitheses. Tsurigane-ni
I have asserted that social psychology is Tomarite hikaru
currently passing through a period of more Hotaru kana.
than usual uneasiness, an uneasiness which is Or,
felt even more by researchers inside the field On a temple bell
than by outside observers. I have tried to ana- Waiting, glittering,
lyze and describe the sources of this uneasi- A firefly.
ness as it is felt at various levels of depth. I
REFERENCES
have also described a few of the undercur-
rents which I believe will, or at any rate CAMPBELL, D. T. Reforms as experiments. American
Psychologist, 1969, 24, 409-429.
should, be part of the wave of the future FRIED, S. B., GUMPPER, D. C., & ALLEN, J. C. Ten
which will eventuate in a new paradigm which years of social psychology: Is there a growing
will lead us to further successes, after it re- commitment to field research? American Psycholo-
places the recent paradigm which has served gist, 1973, 28, 155-156.
HIGBEE, K. L., & WELLS, M. G. Some research
us well but shows signs of obsolescence. trends in social psychology during the 1960s.
A time of troubles like the present one is a American Psychologist, 1972, 27, 963-966.
worrisome period in which to work, but it is McGuiRE, W. J. Learning theory and social psychol-
also an exciting period. It is a time of con- ogy. In O. Klineberg & R. Christie (Eds.), Per-
spectives in social psychology. New York: Holt,
tention when everything is questioned, when it Rinehart & Winston, 196S.
sometimes seems that "the best lack all con- McGuiRE, W. J. Some impending reorientations in
viction, while the worst are full of passionate social psychology. Journal of Experimental Social
intensity." It may seem that this is the day Psychology, 1967, 3, 124-139.
of the assassin, but remember that "it is he McGuiRE, W. J. Theory-oriented research in natural
settings: The best of both worlds for social psy-
devours death, mocks mutability, has heart to chology. In M. Sherif & C. Sherif (Eds.), Inter-
make an end, keeps nature new." These are disciplinary relationships in the social sciences.
the times when the "rough beast, its hour Chicago: Aldine, 1969.
come round at last, slouches toward Beth- SEARS, D. O., & ABELES, R. P. Attitudes and opinions.
Annual Review oj Psychology, 1969, 20, 253-288.
lehem to be born." Ours is a dangerous pe-
riod, when the stakes have been raised, when (Received December 1, 1972)