Remarks on Method in the Study of Society

Arthur F. Bentley


The background.—Things material and things psychic are approximations in statement that point to problems of process to be solved. Things social must not be sought to add to them, rather the social aspect of bifurcated facts should be the clue to their examination in common process. Visibility and language.—We see by the aid of describing, and describe by the aid of vision; and these interdependent processes make up science. Social science is notoriously in that early stage in which set eyes and dead words still hamper each other. The framework of personality.—We still see our facts as located in "personalities," which are themselves but linguistic tools for the practical uses of daily life. We must force our way onward until, through and across personalities, we can "see" directly the facts with which we must deal. The framework of language.—Any single language limits the world of its addict to its own content and structure. Not merely many languages, but all types of language, must enter solution before we can secure the clarified vision which will then be our fact.

Current psychological and philosophical terms, from souls to larynges, are attempts at fixating the verbal values of practical talk, the form of fixation being the "thing," a form from which science flees. In science, as today developed, such terms as "gold" or "the moon" are merely indications of problems; and the trend, so far from being toward their heightened specification as things, is toward the blending of all their indicated processes into the wider processes of scientific experience.

Sociology deals with material things and with analogous psychic things, but until it passes beyond these in their approximate values as things, and goes through them into their full processes, it can hardly qualify itself for the term "science" in the recent sense in which the word is used above. Not fixation of things, but processing, is what it must seek.

The other sciences, except psychology, need not bother much with the process of experience; their field is in a content of experience. Psychology works inside a process of experience, assuming an environment. It is the peculiarity of sociology that it must deal with great procedures that work through the experiencing process.


Sociology has had its struggles to locate a social thing in addition to various material and psychic things; but in that way does not lie science for it. It must cut through the bifurcations, subject-object among them, if it is to do its work. In the social aspect it must find, not an added type of thing, but the clue to unification and to analysis of process in the unified material. It may approach this problem in disregard of the techniques of all theologies, philosophies, and psychologies for the very reason that the tests and goals of those systems of approach are to be found within its own material.

This problem must be solved, to the extent at least of a working scheme, before any advance whatever may be considered as having been made toward a statement of the process of social living with values comparable to those of other contemporary sciences.

The following remarks can hardly be of interest to any investigators who regard the problem as one merely curious or incidental in character; or to any who satisfy themselves in retaining "things" as their material with some unifying hypothesis as a sop to their minds; or to any who by reason of their dominating practical goals do not feel the need of cutting under those goals themselves. They are intended only for investigators who refuse to regard themselves as making progress at all until they can deal with a knowing-known direct, instead of with the hypothesized interactions of some knowing thing and some thing known.

Of method much is written. With method little is done. Roundabout ways of satisfying self, or others, that something is—or may, or should, be—as one wants it, are legion.

But for knowledge upon which all may agree good eyesight and accurate words are alone required.

Visibility and language are the conditions of science—they are its substance.

Nor are they in the end separate. As instrumentalities they coalesce.

Visibility has been low in the social sciences. Facts have presented themselves borne in the personalities, the individualities, of men. Through these personalities we have tried to look—as through a glass—darkly.


Terminology has been poor in the social sciences, drawn as it has been from the language of everyday life—from the vocabularies of the manipulation of one man by another.

But not the point of view of one toward another is what we seek, rather the very processing itself of the ones-with-others. The everyday language has been a hindrance, not capable of purification.

A generation ago the atom was a hypothesis. Chemists still assumed that bare-eye reports, or magnified eye-reports, were the known facts. Today the electron is not a hypothesis, in the older sense, but—thanks to Millikan—rather in the class of things seen, and vision has become instrumental in new ways.

Then, spectroscopic vision was in its infancy. Today, the spectroscope reports as accurately on hundred-thousand light-year intervals as the keen-sighted woodsman on what moves in the treetops. This is not to make a creed of the electron; it is only to refuse to make a creed of the squirrel in the tree-top.

Then, mathematics was a thin-air dweller: abstract, unreal, but helpful when not mere curiosity. Today, mathematics is the language of physics, an aid to the vision of physics, and the recorder of that vision. Mathematically stated vision is physics.

The method of social science is likewise vision and statement, both processes instrumental.

Quantity and quality in social science consist in how much and how well we can see, and in how much and how well we can tell what we see. For "how well" understand "how coherently." Quantity and quality both are to be found in the extent to which our observations and reports of observations fit together fluidly, flexibly, coherently—every observation, every statement of every worker, ready promptly to destroy itself in accordance with the ever-changing requirements of all the other observations, all the other statements, in the system, the whole system of social science.

Social science, as I have said, has long been looking at the facts of men's social living through personality. It has been reporting these facts largely in individual point-of-view language, that is, in psychological terms. It has been assuming that psychological terms indicated psychological existences—a nave assumption.

(459) True, indeed, but true like a nymph in a tree trunk.1 It has been assuming that personality was a resting point for social fact, a starting-point for social interpretation.

Sweep it away. That is the first of the two points of method which I wish to emphasize.

Sweep what away? Not the personality-in-the-body, nymph-in-the-tree idea, feeling, belief. That idea-feeling-belief is very lovely, very true, very useful—in its own time and place.

Sweep away only the naive assumption that that is the method to see and record facts in social science.

Reduce, then, the personality system of seeing and recording to the 'position of one among many possible systems2 of seeing and recording.

Try the other systems, and see which gives the most complete, the most coherent delivery.

The personality form of statement is a good form for domestic quarrels, but a bad form for theories of crime and punishment. It is nice for hero-worshipers, illusion for Buddha, bedrock for Western theology, multiplicity-in-unity for psychoanalysis, hypothesis for the philosopher, and the most uncertain thing in the world to its closest students.

Incidentally it is accepted datum to most sociologies. Almost certainly it ought not to be.'

The president signs the bill with the treasured pen. His act is physical and personal and social. It is prospectively capable of complete statement as physical, of complete statement as personal, of complete statement as social.

Under present technique, we know the physical statement of presidential pen-pushing cannot be perfected to our satisfaction.


The personal statement we know to be confused. The literature of the confusion is in evidence.

The social[4] statement, not as complement or supplement to physical or personal, but as alternative, in its own right, for complete observation and description, is steadily advancing in power.

The yield for this social form of observation and statement in the immediate future will depend apparently not so much on whether we can command the words to report its observations as upon how far we can make the effort successful to see full facts under its form. One grits one's teeth over it; sometimes one gnashes them. But words are both the material and the tools of social study. We must see through them before we use them.

Therefore we come to the second of the two points in regard to method which I desire here to make.

It is entirely hopeless to expect a satisfactory technical sociological statement to develop in any one existing language.

It is equally hopeless to expect any worker to make real progress unless he controls several languages, at the very least, in their symbolic values.

One-language terms drag in the absolute, make the worker himself an absolute, and therefore wreck him and his work. Multiple-language terms shimmer. They compel knowledge. 


  1. Spirit occupancy of a tree is of course a true statement of tree life if that is the best system of expression the observer commands in which to make his reports.
  2. For such systems the phrase "verbal frames of reference" is probably more satisfactory. See my book, Relativity in Man and Society.
  3. "There is also the feeling that any adding in of `conscious' factors which cannot be measured and do not obey the same laws as the rest of nature must play havoc with all hopes of satisfactory explanations: and this feeling is justified" (C. К. Ogden, The Meaning of Psychology, p. 166). This passage refers to psychological, not to sociological, investigations, but is nevertheless pertinent.
  4. The word "social" is arbitrary here. "X" would do as well. Every adult student in this field knows there's the rub.

Valid HTML 4.01 Strict Valid CSS2