The first two volumes, entitled The American Soldier, of a four volume work on Studies in Social Psychology in World War II are announced by the Princeton University Press for publication early in May.

These volumes are produced under the auspices of a special committee of the Social Science Research Council, comprising Frederick Osborn (chairman), Leonard S. Cottrell, Jr., Leland C. DeVinney, Carl I. Hovland, John M. Russell, and Samuel A. Stouffer.

“It involved study of the attitudes of more than half a million soldiers, in the United States and overseas.”

The research on which the publications are based was carried on in the Research Branch, Information and Education Division of the War Department. It involved study of the attitudes of more than half a million soldiers, in the United States and overseas. After the War, the Social Science Research Council obtained the release of the basic data for further analysis and presentation to the social science professions. Funds for carrying out the work were granted to the Council by the Committee of Trustees on Experimental Programs associated with the Joint Army and Navy Committee on Welfare and Recreation, and by the Carnegie Corporation of New York. Security restrictions were removed on classified reports based on more than 200 separate research studies and the original IBM punched cards and code books were supplied to the editorial subcommittee: Samuel A. Stouffer (chairman), Leonard S. Cottrell, Jr., Leland C. DeVinney, and Carl I. Hovland.

Volume I, The American Soldier: Adjustment During Army Life, analyzes a wide variety of problems of the Army as a social institution and of the adaptation of men to the requirements of military life.

Volume II, The American Soldier: Combat and Its Aftermath, deals with the sociological and psychological factors involved in behavior under stress, and also analyzes special problems arising with the end of hostilities.

Volume III, Experiments on Mass Communication, scheduled for publication this summer, is an analysis of the logic of controlled experimentation in social psychology, as exemplified by a large number of experiments in influencing the minds of men through motion pictures and radio.

Volume IV, Measurement and Prediction, scheduled for early fall publication, funds the experience of the Research Branch in developing and using new theories of socio-psychological measurement and in predicting future behavior.

The present discussion is limited to Volumes I and II. These volumes, unlike Volumes III and IV, are primarily substantive rather than methodological. Nevertheless, they provide an exemplification of the application of new and old techniques to a very large variety of problems and are expected to be of considerable interest to social research technicians and to teachers seeking illustrative examples of technical operations.

The main objectives of Volumes I and II are to enhance our body of knowledge, both theoretical and empirical, about social behavior.

In Volume I, Chapter 1, the work of the Research Branch is described and the point of view of the volumes, in the context of social science, is elaborated. After reviewing some of the thinking of the past which has converged to make up the general orientation of the volumes, the authors write:

The chapters are not organized around any single conceptual scheme. To do that, in the present state of the knowledge of social psychology and sociology, would be as sterile as, say, it would have been to organize data in medicine a century ago around a theory of convulsive action as advanced by Dr. Benjamin Rush. We know now that not one overarching conceptual scheme but rather many limited conceptual schemes were to open the road to progress in medicine. A germ theory of disease was to be useful for one class of phenomena, a deficiency theory for another, still other theories for others. In the present state of the social sciences, it is imperative that we keep an open mind with respect to the potential utility of conceptual models which have not been subjected to the kind of rigorous verification which we can expect social scientists of the next generation to demand.

Conscious of their obligation to present Research Branch findings in a form which will maximize their utility in the future, the authors…have adopted a compromise position with respect to introduction of explicit conceptualization. On the one hand, these reports are not conventional chapters of history. On the other hand, while theory is used both explicitly and implicitly, the data have not been selected merely because of their relevance to some general proposition now current in the psychological or sociological literature. Where the problem area is one which can be expected to concern social scientists in the future, a considerable body of factual data has often been introduced, even if the data are in no sense definitive in resolving conflict between alternative hypotheses which might now be advanced or even if the data do not seem relevant to any current hypothesis.1Vol. I, pp. 32–35.

In the chapters which follow, illustrated by several hundred charts and tables, and textually as compact as the authors could make it, is a portrayal of the Army and its impact on men.

The Army is viewed as a social institution, contrasting in some ways with civilian institutions, although the differences are often differences of degree rather than kind. Thus it has an authoritarian organization, demanding rigid obedience; a highly stratified social system, in which hierarchies of deference are formally and minutely established by official regulation, subject to penalties for infraction, on and off duty; an emphasis on traditional ways of doing things which discourages initiative. Like other social institutions it has its informal as well as formal codes of behavior. The mobilization of informal social controls, in combat and in garrison duty and training, is an important object of study.

In the general analysis of how personal adjustment varied in the Army, an immense amount of data from scores of surveys all over the world is consolidated in a relatively small number of tables. One table alone, at the end of Chapter 5, is based on an analysis of the differences between 8,554 pairs of percentages, each pair based on samples of men matched on 6 characteristics!

“The correspondence between verbal behavior and nonverbal behavior is high enough to indicate that the attitudes supposedly reflecting adjustment are important in the sense that they are related to other behavior reflecting adjustment.”

Verbal behavior of the men, as manifested in responses to questions on attitudes, is compared with nonverbal behavior reflecting adjustment in the Army—such as success in the Army as represented by promotions, or failure as represented by isolation in the guardhouse or in the hospital psychoneurotic ward. The correspondence between verbal behavior and nonverbal behavior is high enough to indicate that the attitudes supposedly reflecting adjustment are important in the sense that they are related to other behavior reflecting adjustment. But this relationship is only half the story. The other half is that attitudes reflecting adjustment, though all positively correlated with nonverbal behavior, represent profiles which vary (a) with the personal background characteristics of the individual soldier, such as education, age, and marital condition; and (b) with various factors in Army experience, such as his branch of service, whether he was overseas, how long he was in the Army, and at what stage of the War his attitudes were studied. To illustrate: personal esprit, personal commitment, status-satisfaction, and general approval of the Army, as revealed in attitudes, are each positively correlated with objective indices of success or failure in the Army. Education is also positively correlated with these indices. But education is positively correlated with attitudes reflecting personal esprit and personal commitment, and negatively correlated with those reflecting status-satisfaction and general approval of the Army. The concept of varying profiles may, the authors think, be important for social psychology.

Other conceptual tools, notably a theory of relative deprivation, also are introduced to help in more generally ordering a somewhat remarkable body of otherwise disparate findings.

One of the most interesting illustrations of the relationship between verbal and nonverbal behavior is a study reported in Volume II, Chapter I of the attitudes of 12,295 enlisted men in 108 rifle companies and 34 heavy weapons companies prior to the invasion of Normandy. It was possible to compare their attitudes before combat with subsequent performance in Normandy as measured by nonbattle casualties, many of which were psychogenic. Careful statistical analysis shows that within a given regiment, the three companies with the worst attitudes before combat had, on the average, more than 50 percent higher nonbattle casualties than did the three companies with the better attitudes. A companion study, based on individual ratings of men in combat by platoon leaders, showed that the men with the better attitudes before combat also were judged to have made the best record in combat.

The Army status system and reactions of officers and men to it are studied in special detail. An ambivalence, involving profound dislike of a system of privilege which is alien to the democratic mores of our civilian society (except for some minority groups) and at the same time involving a desire to use the system to acquire status, is manifested throughout the data.

In the description of problems of leadership, it turns out that the antagonism to officers, though endemic, was least among combat troops and greatest among troops in inactive overseas theatres or the rear areas of active theatres. The theory of relative deprivation helps show how such a finding is related to the extent to which officers and enlisted men shared the same discomforts and dangers.

The analysis of problems of job assignment and job satisfaction reveals many parallels with problems in civilian industry. World War I taught the importance of taking account of aptitudes. If the implications of these volumes from World War II are fully realized, an additional impetus will be given to the importance of taking account of attitudes as well.

“There was only one answer to Pearl Harbor, and there was no defeatism, but the conflict was not viewed as a war to end war or as a crusade for democracy.”

Because the Research Branch was part of the Information and Education Division, which was responsible for trying to stiffen the ideological supports of the men, the role of ideology in World War II was studied in special detail. It perhaps comes as no surprise to learn, although the picture is a complicated one, how relatively impervious were our soldiers to attempts to describe the War in terms of positive goals. There was only one answer to Pearl Harbor, and there was no defeatism, but the conflict was not viewed as a war to end war or as a crusade for democracy. The failure to recover from the disillusionment of World War I, combined with distrust of some of our Allies—notably Russia—comes in for careful study. Moreover, the chapters on combat show that the closer the contact with the enemy, in either hemisphere, the less rather than the greater was hatred or vindictiveness.

Over a hundred pages are devoted to an analysis of race problems in the Army. A number of myths are dissolved by these studies, for example, the idea, quite commonly held by white officers, that the Negro soldiers preferred white officers. Only 4 percent of the Northern Negroes and 6 percent of the Southern Negroes said they did. Or the idea, often advanced by Southerners, that Negroes preferred Southern white officers to Northern white officers. Only 1 percent of the Northern Negroes and 4 percent of the Southern Negroes did. Or take the idea, firmly held, that white soldiers would never tolerate Negroes in the same company. Mixing Negro platoons into white companies was actually tried out in combat in Europe. Only 7 percent of the white soldiers in these mixed outfits disliked the idea after they had had experience with their Negro comrades in combat, but two thirds said they had disliked the idea at the beginning. It is significant that almost exactly this same proportion (two thirds) of white combat troops in divisions without such mixed companies said they would dislike trying the experiment in their outfits.

The Negro, it is shown, had mixed feelings about the War. He was less likely than the white to consider the War his affair; but many Negroes, and especially the most racially militant, saw in the War a chance to stake a claim for better treatment afterward. Hence, the racially militant were the most zealous to make a good showing in the Army.

There are many paradoxes in this study of Negro attitudes. For example, Northern Negroes stationed in the South, while disturbed by Jim Crow practices, were as well adjusted as or better adjusted than Negroes stationed in the North. This surprising finding makes sense, however, when it is noted that relative to Negro civilians in the South the Negro in uniform was very well situated; the opposite was true in many instances in the North, where he could compare his lot with that of friends earning “big money” in war industry.

More new data bearing on America’s race problem are assembled in these volumes than in perhaps any book since Myrdal’s An American Dilemma.

A total of eight chapters in Volume II is devoted to an analysis of social and psychological factors relating to combat—on the ground and in the air. No attempt will be made here to review systematically the manifold problems treated: combat motivations, effect of length of combat on attitudes, what kind of men stood up best, the control of fear and anxiety, etc.

“This chapter shows, too, that not only did attitudes deteriorate after a few months of combat but, contrary to the view widely held by the top command, combat efficiency also deteriorated.”

The kind of material on combat can be illustrated from Volume II, Chapter 6, on “The Combat Replacement.” Here it is possible to compare three types of men, namely, seasoned battle veterans, new replacements in veteran companies, and men in combat divisions which had not seen combat. How the new replacements quickly took on the attitudes of the veterans, to whom they looked up—both the cynicism and the pride, often bitter pride, of the veteran—is vividly illustrated. Also, in important respects, the replacements are shown to have been responding to the different situation that they faced. To them, for example, combat-tested leaders carried greater prestige than their fellow veterans would grant them, and the replacements surpassed veterans and men in new divisions alike in their acceptance of Army requirements. This chapter shows, too, that not only did attitudes deteriorate after a few months of combat but, contrary to the view widely held by the top command, combat efficiency also deteriorated. Such findings as these were to have a bearing on controversies accompanying the adoption of the point system establishing the order of demobilization in the Army.

The point system was actually invented by the Research Branch and “sold” to the Army on the basis of attitude studies made in all parts of the world. The inside story of the point system and the struggle for its adoption is told for the first time:

In planning for demobilization, the Army faced a problem unprecedented in American history.

With the defeat of Germany, it would be possible to release several million soldiers…Prudence required the assumption that another year or more might elapse before the capitulation of Japan…

To keep all men in the service was both unnecessary and politically unthinkable. Some would have to be discharged. Who should these men be?2Vol. II, p. 520.

The idea of a point system for demobilization…[in terms of what the soldiers themselves wanted, was] accepted by the War Department…Representative samples of men throughout the world were queried and from their responses the variables of length of service, overseas duty, combat duty, and parenthood emerged as most significant…Studies of reactions to the point system showed that the response to it was remarkably favorable, except among minorities who felt they were personally most disadvantaged by it—and the response to the idea of the point system remained predominantly favorable even after many men became angered by the alleged slowness of demobilization. The point system established the order, not the rate, of demobilization. While some men eventually confused the two ideas, the majority, though hostile to many if not most Army policies, continued to approve the point system.3Vol. I, p. 7.

In view of the explosive tensions in the early demobilization period, it has been suggested, history may eventually record that the establishment of an objective system for order of demobilization whose justice was accepted by most men may have saved the country from what could have been a crisis seriously damaging to American prestige.

The Research Branch was, of course, engaged in an engineering operation and its contributions to social science as funded in these volumes are a by-product rather than the main objective of its wartime activity. The point system was only one of many achievements of the Branch, though it perhaps was the best known. Most of its studies were directed at local specific problems. For example, what led men in the Southern Pacific not to use atabrine as regularly as the Army thought they should; what attitudes and practices which might be correctible enhanced the likelihood of getting trench foot; which of two kinds of huts did men prefer in Alaska; what were the preferences for winter clothing among frontline troops in Belgium, Luxemburg, and Germany; what kind of radio programs did men prefer; what did they like to read in Yank magazine; what was the laundry situation in Panama; what were the attitudes toward Chinese among troops in India-Burma? These were a few among many topics about which studies were requested.

Sometimes the studies required not only factual data but also predictions as to the future. One chapter in Volume II reproduces in full a memorandum written months before VE day by the Research Branch on morale problems among troops in Germany after it was occupied. The predictions are compared, point for point, with factual data obtained afterward from the occupying soldiers. Another chapter in Volume II tells of the soldiers’ plans for their future upon discharge. It was a Research Branch survey, made as early as 1943, which formed the basis for calculating the educational costs of the GI bill being drawn by a special presidential committee. Comparisons are shown of predictions of soldiers’ plans with what the same men actually were doing several months after discharge from the Army.

The Research Branch worked closely with many other agencies of the Army—with none more closely than with the Psychiatric Division of the Office of the Surgeon General. Out of this liaison came many studies, including, toward the end of the War, a psychiatric inventory which was employed routinely in all the induction stations of the United States.

These volumes are frank, not only in revealing the Army’s shortcomings, but also in pointing out explicit shortcomings of the research. In Volume I, Chapter 1 appear these words:

The social psychologists and sociologists who studied problems of motivation and social adjustment in World War II have an obligation…to report on their studies and thus to speed up the process of development of the science of man. Science, unlike art or literature, is cumulative, in the sense that a scientific achievement is most successful when it stimulates others to make the concepts and techniques it has used look crude and become obsolete as rapidly as possible. In this spirit the present volumes have been prepared.4Vol. I, p. 5.

The authors of Volume I are Samuel A. Stouffer, Edward A. Suchman, Leland C. DeVinney, Shirley A. Star, and Robin M. Williams, Jr. Stouffer, Star, and Williams are also authors of Volume II, with Arthur A. Lumsdaine, Marion Harper Lumsdaine, M. Brewster Smith, Irving L. Janis, and Leonard S. Cottrell, Jr.

This essay originally appeared in Items Vol. 3, No. 1 in March of 1949. Visit our archives to view the original as it first appeared in the print editions of Items.

Banner photo credit: National Archives.


Vol. I, pp. 32–35.
Vol. II, p. 520.
Vol. I, p. 7.
Vol. I, p. 5.