Tioned: American Anthropologist

116 A M E R I C A N ANTHROPOLOGIST [N. s.
, 43, 1941
are, curiously, those which deal with the simplest and most basic phenomena. The
authors have given a brilliant and impartial analysis of the three leading theories of
learning, pointing out that each of these theories (of Substitution, Effect and Ex-
pectancy) is descriptive of a different basic learning context; and that each theory
can be made to cover all conceivable contexts if only we (and the experimental ani-
mal) be willing to see the less suitable contexts through Procrustean spectacles which
will convert these less suitable contexts into special cases of the basic context on
which the theory is based.
As the book works on towards more complex matters culminating in a chapter on
personality and experimental neurosis, the anthropological reader becomes restive
knd disappointed-not a t the authors but a t the thinness of the available data.
These above all are the experiments from which he might expect enlightenment, and
it is, indeed, exciting that anything approaching neurosis can be experimentally
induced. But the research in these complex fields is inchoate, and the induced neuros-
es are typically brought on not by repression but by failure to discriminate between
similar stimuli of opposed significance. We lack data on the inhibition of uncondi-
tioned responses which might be comparable with the repression of sexual and other
basic tendencies in socialized man.
This gap in the research is due almost certainly to the rigid frame in terms of
which the experimenters devise their experiments. The typical event sequence with
which they work is too short-terminated by the reward or punishment, and there-
fore problems relating to inhibition of the terminal unconditioned response are
necessarily ignored. T o tackle them will demand a lengthening of the typical se-
quence-a matter of very great difficulty. When this is accomplished, we shall be
able to introduce time into our concepts of life-space and, probably, to study learn-
ing not only in contexts where the subject plays a purely passive role but also in
contexts where he is given the illusion a t least of free will or dominance.
GREGORY BATESON
NEWYORKCITY
Mathematico-deductive Theory of Rote Learning: A S t u d y in Scientific Methodology.

CLARKL. HULL,CARLI. HOVLAND, ROBERTT. Ross, MARSHALL HALL,DONALD
T. PERKINS,FREDERIC B. FITCH.(Published for the Institute of Human Rela-
tions. xii, 329 pp., one plate, 40 figs.,$3.50. New Haven: Yale University Press,
1940.)
This great book is the first attempt a t exhaustive formal exposition of our knowl-
edge within one small area of the learning and conditioning field. I t s major function
for anthropologists is that it provides us with a lesson in how to think, and even
those who do not propose to work through the massive apparatus of definitions,
postulates and theoreums will do well to read the thirteen pages of introduction,
Concerning Scientific Methodology. The naive will discover that when knowledge is
formally exposed, the order of procedure is the reverse of that adopted in reporting
upon research. The ordinary scientist thinks inductively, deriving his general con-
clusions from his particular observations, but in this book the most abstract con-
BOOK REVIEWS 117
clusions become “Postulates” from which the author deduces the phenomena which
can be observed or ought to be observable in the laboratory.
With this deductive method there goes necessarily a new stringency in thinking.
I n anthropology we have only to look a t facts and then induce conclusions from
them and, to do this, we scarcely need to think. We just tinker with words until we
have a phrasing which will cover all the cases that we know of, and all too often the
resultingconclusion isanegative one. The case would be terribly altered if we solemn-
ly sat down to list the general conclusions of which we feel sure and then attempted
to deduce the phenomena from these. We should find in the first place that many of
our conclusions were so loosely stated as to be utterly useless as postulates; and then,
when we had introduced all sorts of new precision into our conclusions, we might well
find that a very vast range of phenomena could be deduced from a very few, very
simple postulates-in fact, that the phenomenal world is by no means as complex as
the addicts of “organismic” thinking would have us suppose. These are the two chief
morals to be drawn from Hull’s book-the need for absolute stringency in our for-
mulations and the inspiring hope that such stringency will be amply rewarded by
fruitful predictions.
As to the content of Hull’s system very little can here be said. I t is of course to
the Postulates that we must turn for ideas relevant to our subject, the Theorems
being for the most part concerned with details of phenomena in the rote laboratory.
The system contains eighteen postulates, of which the first states the existence of
“stimulus trace.” This concept is in no sense a causal entity, but is rather a conven-
ient linguistic peg upon which to hang values of other entities. The remaining seven-
teen Postulates build up into a picture somewhat like this: that the occurrence of a
correct response in a rote learning context depends in the main upon excess of Ex-
citatory Potential promoting that response over Inhibitory Potential opposing it,
(the inhibition being due to competition from possible incorrect responses). The
building upof thesepotentials is postulated as maximally simple, proceeding in equal
increments. The decrease in these potentials is also maximally simple, its rate being
proportional to the potential present a t the given instant.
Putting the matter loosely, we may see the Inhibitory Potential as a sort of
threshold which has to be exceeded by the Excitatory Potential. But this is not
enough, and Hull goes on to postulate a further threshold which he calls “Reaction
Threshold.” Specifically, the difference between the two potentials must be greater
than this new threshold for correct response to occur, i.e., the expression (Excitatory
Potential minus Inhibitory Potential), must be greater than the Reaction Threshold.
It is a far cry from the two Potentials to the subject matter of anthropology
but this extra entity, the Reaction Threshold, is of a higher order of complexity
which begins to be relevant to us. Discussing it, Hull lets drop a hint which is preg-
nant with meaning for anthropology. H e says “The fact that successive rote series
are, on the average, learned with progressively fewer repetitions suggests that (the
Reaction Threshold) may be reduced by preceding practice. This problem presents
serious theoretical difficulties which ultimately must be met by any satisfactory
theory of rote learning. I t is not treated in the present system.” I n these three sen-
118 A M E R I C A N ANTHROPOLOGIST [N. s., 43, 1941
tences, we may see indicated the future bridge between Hull’s system and cultural
anthropology. We, anthropologists, are continually dealing with problems of this
order; we want to know, not how fast people learn, but what methods of learning are
standardized among them. Hull’s statement means to us that the experimentalists
are beginning to think about these problems of learning to learn. The answers to these
problems will unite our discipline with theirs.
GREGORYBATESON
NEWYORKCITY
Explorations i n Persomilily. HENRY

A. MURRAY and Associates. (xiv, 761 pp. $8.50.
New York: Oxford University Press, 1938.)
Some psychologists are content with a slow circuitous approach to the study of
man and glad to maintain life and activity by unobjectionable if picayunish pieces of
research which will “sometime by someone” be integrated into coherent knowledge.
H. A. Murray is not among these. He has poured his vivid and rebellious spirit in-
to the clinic which he leads. He wants a psychology right now, or the beginnings of
one, and has taken the first steps (as he sees them) to get it.
First he assembled a “diagnostic council” consisting of five persons who had had
some experience in dealing with real people with all their complexities. Next he got
some people to study (28 in the best observed group), and then he and his associates
worked out a conceptual scheme and a series of experiments and interviews with
which to evaluate the personalities of the subjects.
The foremost analytical concept is that of need, a n inferred process that precedes
behavior. Personality is conceived of primarily as driven by needs, organic and
psychogenic. The attempt is made to define all relevant needs and to recognize them
in behavior. Examples would include Cognizance, Defendance, Deference, Exhibi-
tion, Nurturance, Dominance and Blamavoidance. The diagnostic council attempted
to rate the subjects on these and other needs, to test agreement on such ratings, to
predict behavior on the basis of the need analysis of a given person, to discriminate
latent or unconscious from conscious needs, and finally to reconstruct the whole
personality out of these elements. A brief review can scarcely convey the ingenuity
and subtlety of the conceptual attack on the problem; but one sees plainly the evi-
dences of intellectual courage, intense application, and high talent for the task.
Often, indeed, Murray’s paraphrases of the pet ideas of others improve upon the
originals.
For a n existing psychology which makes a good attempt a t grappling with im-
portant personality problems one must turn, of course, to Freudian psychoanalysis,
and this Murray and his co-workers have done. Many of their major orientations are
drawn from this source, and they give an honest (if somewhat ambivalent) recogni-
tion to the bone labor already done in the field. They do not, to be sure, make use of
Freud’s best invention, to wit, his interviewing method, but this seems to have been
due to the technical difficulty of getting so many subjects to accept it in a university
community, rather than to a lack of appreciation of its utility. The use of Freud’s
hypotheses without employing the method necessary for validating them does,

Tioned: American Anthropologist

Uploaded by

Copyright:

Available Formats

Tioned: American Anthropologist

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tioned: American Anthropologist

Uploaded by

Copyright:

Available Formats

116 A M E R I C A N ANTHROPOLOGIST [N. s.

Mathematico-deductive Theory of Rote Learning: A S t u d y in Scientific Methodology.

Explorations i n Persomilily. HENRY

You might also like