The Concept of Complexity in Sociology: II
ARE SOCIAL PHENOMENA MORE INTANGIBLE?
THE idea that social phenomena are more complex than other natural phenomena because they are more intangible is based fundamentally upon two misconceptions: first, as to the nature of consciousness; second, as to the nature of knowing,-our old philosophical problems of mind and epistemology.
There have been, and unfortunately still are, many sociologists who regard social
( 370) phenomena as "intermental response," and "mental" means "mind" and "mind" means some kind of metaphysical stuff that exists outside of, independent of, and unconditioned by, living protoplasm. Mind is a kind of superphysical substance, stuff, essence or entity which is "revealed" to us through cerebral activity, "secreted as the liver secretes bile," or occurs parallel to neuro-muscular activity. This idea is perhaps best exemplified by the writings of Durkheim, Fouillée, and other members of the French School, but representatives can be found in every country in every age. It is a belated hangover of Platonic realism, very dear to the hearts and minds of philosophers, preachers, and poets. It is one aspect of the vicious dichotomy of the universe into mind and matter, soul and body, flesh and spirit.
In the attempt to unify a universe thus dualistically severed, one type of thinkers since Descartes has chosen the monistic-materialism horn of the dilemma and another has postulated some variety of spiritualism. Happily, both of these views are rapidly going into the discard and we seem to be developing a neoholozoism, or energism, which regards "reality" as. neither mind nor matter. Mind and matter are conceived as special types of energy-organization as responded to by the energy-systems we call organisms. The kind of responses made by the cortico-muscular systems of men to physical and social objects is what we usually mean by mental activity. Mindmonism is as dead as matter-monism. Consciousness is variously interpreted as implicit verbal responses (Watson), temporary cortico-muscular reflex arcs (Bode), symbolic verbalism (Markey), neuro-psychic behavior (Bernard), and in many other similar ways, but practically all modern writers agree that there is no mind stuff.  Hence, social phenomena involving "consciousness" must be dealt with as all other natural phenomena are, i.e., by observing, recording and generalizing the behavior of defined units.
When we think of intermental responses as existent always and only in some form of observable response, at least so far as the purview of science is concerned, the alleged intangibility of social phenomena largely disappears. Only sensible data are subject-matter for science. Can we observe social phenomena by our senses? Obviously so; and obviously we can do so without instrumental aids to our senses more easily and more accurately than can the physical scientists in any field when they transcend commonsense observation. To observe social data we need no eye-glasses, spy-glasses, micrometric scales, stress machines, amplifiers, testtubes, or other tools. Social data are usually crude, obvious, ultra-tangible to naked eyes, ears, noses, fingers. Our problem is not difficulty of observation because of the intangibility of our data, but difficulties of hypothesis, selection, accurate, comprehensive records, classification, mathematical and logical manipulation, elimination of subjective bias, and
( 371) anthropocentric frames of reference. So long as we try to base our science upon speculations regarding mental states, subjective motives, subconscious drives, non-sensible, hypothetical desires, wishes, and so on, we shall not escape our present magico-theologico-metaphysical myths. We shall continue to write many profound volumes that add little to the empirical generalizations of commonsense. 
The second fallacy regarding the tangibility of social phenomena hinges on epistemological confusion. From the point of view of science, we can know only what we can respond to by our senses. But something more than this is necessary. Other people must have sensory responses of a similar sort in the presence of similar objects. Another requisite is that this similarity of response must be communicable. This is done largely by means of words and gestures. We "know," and know that we know, when there is considerable agreement between ourselves and other competent observers regarding the similarity of our responses. All of this consciousness of similarity is dependent upon symbolical reference of some sort. These symbols are chiefly words, ranging in accuracy from the crude, empirical, vaguely understood words of commonsense to the highly developed, carefully defined technical terms, mathematical symbols, and formulas.
It was pointed out above that all words are abstract. They center attention upon only certain aspects and relations of the relatively stable events for which they stand. We usually consider a symbol "concrete" if it refers to objects immediately present to our senses, "abstract" if it refers to generalizations regarding objects not present. But it is very clear that when we respond to this present object by the symbol "table," we are merely symbolizing one of the countless aspects of the energy-system, "table." We have selected only one of the possible human responses (relations), and have centered our attention upon that. In other words, our "table" is an abstraction. We can never "observe" the table as a "whole," because the possible responses to it and relations of it to other objects are innumerable. We could spend our lifetime "studying the table," i.e., making all of these possible human responses to it, and yet never "know it as a whole." Such an intensive study would require a mastery of all knowledge. There is truth in the poetry about the flower in the crannied wall."
From this point of view, then, social phenomena are no more intangible, and no more "complex," than physical phenomena. We observe them both by
( 372) symbolic reference. We observe all things by selecting certain aspects of them and symbolizing these aspects. We see things partially, "through a glass darkly." It is as easy to observe a "whole" city or nation, as to observe a "whole" table; i.e., both are impossible except as we abstract certain features of them and neglect the remainder. We observe them partially, make generalizations by symbolic reference, and let our abstractions stand for the whole. It is an elaborate synecdoche. The city is as tangible as the table to those who have proper symbolical means of knowing cities. The only question the scientist raises is whether the generalizations are based upon sense experience which other competent observers can have. 
So we may dismiss the doctrine of the complexity of social phenomena as due to their intangibility because we "know" them in the same way that we know physical phenomena, and because the "intermental responses" are always "revealed" through sensible behavior. In one sense, social phenomena are more tangible because we can observe most, if not all of them, without the assistance of sense-extending instruments, though we must use sense-objectifying aids in observing and recording our data if they are to be made useful for science. Only thus can we get sufficient numbers of comparable data to reveal the relatively repetitive uniformity of behavior which is the sole object of science.
ARE SOCIAL PHENOMENA MORE DIFFICULT TO UNDERSTAND?
The foregoing discussion implicitly answers this question. We must, however, grant that much social interaction is not understood at present or else is misunderstood. In the same breath we must assert that the same is true of much physical phenomena. It is probably true that empirical (non-scientific) generalizations regarding social phenomena are more valid than many of those regarding physical phenomena. This would naturally follow from the greater tangibility, greater orderliness and simplicity of the commonly experienced social phenomena. The partial invalidity of the commonsense generalizations of folklore, adages, proverbs, moral judgments, religious and political theories, processes of leadership, effects of weather, and topography, etc., on societal phenomena, is certainly no greater than the nonsense of empirical generalizations regarding the growth of stones, weather-signs, folklore of gestation, evolution, creation, effects of drugs, "essence" of herbs, etc., in chemistry, physics and biology
At the same time it must be admitted that the physical sciences have far outrun the social sciences in demonstrating the nonsense of commonsense generalizations. But it remains notable that many of these empirical "principles" of physical science are at least partial truths. Ordinarily, the scientific "explanations" are radically different from those of commonsense, but the observations and generalizations (implicit quantifications) of commonsense have often proved ''true."
( 373) But sociology and other social sciences are rapidly building up a vast body of sound scientific knowledge which by reason of its quantic nature and criticism of commonsense, or both, sounds as strange to the man on the street as endocrinology, radioactivity, or theories of immune sera. I am not sure that such ideas as comparative ethics, the newer ideas of mind and consciousness, the nature of self, acculturation, miscegenation, birth control, criminology, folkways, and mores, relation of machines to culture patterns, relations of parents to children, of men to women, whether married or unmarried, the irrational nature of customs and conventions, etc., are not just as dramatic, incomprehensible, and astounding to the "average" man as the latest physical, chemical, and biological inventions and discoveries.
If by "understanding," we mean understanding the technique by which the generalizations were deduced, it is obvious that the ordinary man is much more nearly able to understand and use the technique of social science than he is that of most physical science. If we mean by "understanding," acceptance, I think a good case could be made for at least as great, if not greater, speed in accepting the findings of social science. It is true there are fewer hunted and hounded physical than social scientists today, but the hunting of social scientists is not so violent as the hunting of physical scientists in the days of Galileo and Copernicus. However, evolutionary biologists and geologists do not yet rest on beds of roses in all parts of the land. The physicists (some of them) are being accepted because radioactivity, cosmic rays, energism, etc., "prove the Bible" and the existence of the "spirit."
But if the social scientists of today and tomorrow are having an easier time than the physical scientists of today and yesterday, it is largely because more people are willing to extend scientific technique to the study of social phenomena and to accept the findings as "true." Physical science has pioneered the way. We have the "scientific habit." It has required less than thirty years to get sufficient understanding of the scientific study and treatment of juvenile offenders so that there is scarcely a city in the land that does not have a juvenile court. It is true that most of them leave much to be desired in the way of organization and technique, but most people understand and approve the idea. Millions of farmers neither understand, approve, nor practice such simple techniques of physical science as crop rotation and scientific breeding and feeding. It took fifty or sixty years to convince English farmers that iron plow-shares did not "poison the soil" and that "breaking grass" was not sure disaster. One might go on forever. Physical phenomena are not easier, not so easy, to understand, either scientifically nor popularly, as social phenomena are; nor are the ideas and techniques of the former always popularly accepted more readily than those of the social sciences. Both have to overcome the terrible irrational inertia of the "animal that laughs but seldom thinks."
The foregoing discussion has merely pointed out that some of the scientific formulations of sociology are more easily understood and more readily accepted than many of those in physical science. This does not meet the contention that social phenomena are more difficult to understand, i.e., to explain, .scientifically, than are physical phenomena; therefore the former are more "complex" than the latter. This is one of the outstanding defense mechanisms of sociologists. The foregoing discussion has largely exploded it, but one common error, which is largely responsible for the condition must be
( 373) discussed briefly. This is what may be called the Fallacy of Explanation by Reduction. This has been, and still is the bête noire in the sociologist's zoo of methodological monstrosities.
It is a commonplace axiom or dogma of science that there are certain empirical autonomous orders or levels of natural phenomena. A common categorization of them is physical, biological, psychological, sociological. The number of levels is immaterial, but from the time of Comte, these have had general acceptance. Some writers make a distinction between social and cultural which is probably valid. The materialistic and spiritualistic monists have tried in vain to bridge the gaps between the various levels. While it is obvious that each merges into the other almost imperceptibly, still, in general, we have no difficulty in classifying most sense-objects as belonging to one of these levels. There is also fairly complete agreement that they represent, in a rough way, the chronological order of appearance, at least on this earth. Still, we are beginning to suspect that all four orders of phenomena may be appearing (as they are obviously disappearing,-or at least changing their unit energy-organization) on earth at the present time. 
The difficulty that arises in connection with this view is closely connected with the Comte-Spencer hierarchy which has been implicitly if not explicitly accepted by subsequent writers. The fundamental fallacy arises when we begin to consider each successive order as dependent upon the preceding, more complex and unstable, less tangible, and explicable only in terms of the preceding. This is the fallacy of Explanation by Reduction. Thus Hall writes, "True types of character can be determined only by studying the animal world." Allport pushes the reduction back to chemistry and physics. The battle rages at present between some
( 375) sociologists and most psychologists as to whether psychology is basic to sociology. Trotter, Wallas, McDougall, Ellwood, and many others seem to take this view. The assumption is that while social phenomena are sui generic, yet to be "explained" they must be "reduced" to psychological phenomena; which, in turn are reduced to biological; which, in turn, are explained in physico-chemical terms,and so by this declension we plunge into the all-engulfing waters of metaphysics where nothing can be "explained."
If the nominalistic view of this paper be accepted, it follows inevitably that any Explanation by Reduction does not "explain" because it immediately changes the universe of discourse and we are no longer talking about the aspects of the world with which we started. "Table" is explained on the sociological level when we classify it by its uses, cost, design, etc., and describe it in terms of human values and relations. We can similarly explain it on the organic and inorganic levels. Taking the psychologic level to refer to individual mental behavior, we might "explain" it on this level by describing its relations to a particular individual in his non-social behavior, if any exists. All human psychological behavior is sociological. The residue belongs on the biologic level. Human psychology appears to be in a process of fission, one half going back to biology from which it came, and the other becoming social psychology, or a phase of sociology. Comte's "cerebral (or neuromuscular) physiology" was a good term.
In discussing Allport's paper referred to above, Goldenweiser said something which should be pasted on the blotter of every sociologist. "Modern science conceives of explanation as conceptualized description. . . . . If a fact in one level is explained in unit terms of the same level, the advantage of the procedure lies n the fact that the autonomy of the level is preserved and the mystery (or at least the puzzle) of the transformation of its terms into those of another level avoided. . . . If, on the other hand, a fact in one level is explained by unit factors from another level, this leads to an ultimate conceptual unification of the universe, to a monistic world-view." Furthermore, it means that the fact is not "explained" at all, because it no longer exists when it is viewed from the aspect of the other levels. It is a fine case of throwing out the baby with the bath. The result of such explanation is that we either merely make a series of true statements about two or more energy-structures, or else we erect an elaborate analogy between them. In either case, the result is confusion and that illusion of knowledge which is a dangerous thing. We can make a number of true statements that will apply equally well to tables, chairs, men, poems, inorganic compounds, organic compounds, religious institutions, etc., but no one imagines we have thus "explained" any of them. It is true that iron and hydrogen both expand in the presence of heat, both have mass, both have atoms, etc., but such statements do not "explain" either. The only procedure that does so is to describe their behavior in quantitative units. The fact that some of the same units can be used to describe both is fortunate from the standpoint of economy in recording behavior, but that is not the essence of the explanation. We can use some of the same units, or multiples of them, to describe phenomena that are related to so-
( 375) -cial behavior, e.g., the correlation of size (mass) of men with occupations, leadership, etc. But some of the units necessary to explain iron and hydrogen are different, as their atoms and molecules, tensile strength, coefficient of elasticity, indefinite expansibility, and so on.
The chief difficulty with Explanation by Reduction is that it often transforms terms, deals in crude analogy, blurs the differences, assumes a unilinear continuity in energy-emergents, and uses general terms that apply equally well to many things. Referring to the inability of psychology to explain culture, Lowie says, "We get simply general formulae about feelings and will that are equally applicable to the case of the man's beating his wife or to the boy's resisting the temptations of the lolly-pop. " American ethnologists have long been accustomed to explain cultural patterns on the basis of the situation level; i.e., to be content with conceptualized description of their data; the behaviorists are beginning to do this in psychology; the sociologists must follow. It does us no good to know all about the instincts (if any), reflexes, hormones, enzymes, synapses, musculature, bio-chemistry, and inorganic chemistry of men, so far as a sociological explanation is concerned, for just as soon as we change our attention to these things the social phenomenon no longer exists as an object of attention.
Therefore, it is contended that the claim that sociology has to "wait for psychology (or any other science) to develop," is unsound and absurd. If psychology knew all things on its own level (if it has one), it would not help sociology to understand sociological data. What we have been doing is to wait for psychology to develop into social psychology or sociology. This fallacy of Explanation by Reduction, based upon the fallacy of a dependent hierarchy of natural phenomena levels of increasing complexity, intangibility, and instability is largely responsible for the arrested development of sociology. Happily, we are now beginning to see the error of our ways.
CAUSES OF CONFUSION CONCERNING THE CONCEPT OF COMPLEXITY
It is the thesis of this paper that social phenomena are no more "complex" than other types of natural phenomena, and in some respects, less complex; i.e., they are less numerous, at least as stable and orderly, are more tangible, and therefore easier to "understand," both popularly and scientifically. Some of the reasons advanced to account for the "illusion of complexity" are the following.
I. The Comte-Spencer Hierarchy of Sciences. While it is agreed that the conception of a chronological order of cosmic appearance of certain types or levels of phenomena may be sound, it was pointed out that there is reason for doubting even this. Certainly, there are no convincing historical or logical reasons for the assumption of a unilateral, dependent, chronological order in the appearance of the various sciences, nor is there any reason to suppose that the later emergent levels are more complex, heterogeneous, and unstable than the earlier.
2. The Fallacy of Explanation by Reduction. This is a logical result of the Spencerian evolutionary concept. It inevitably leads to a world monism of some sort. Most physical science has been based explicitly or implicitly upon a materialistic, mechanistic monism. Most all particularistic, simplistic social theories commit this fallacy of Explanation by Reduction. Crude analogies, non sequiturs, transforma-
( 377) -tion of terms, and speculative inferences abound in these reduction explanations. If we start with the assumption that social phenomena must be explained in terms of sciences of the lower levels,of course no science of sociology is necessary or possible. Many sociologists fall into this pit of confusion. While asserting the autonomous nature of social data, they still rely upon Explanation by Reduction.
The Difficulty of Escaping Anthropomorphism and Anthropocentrism. Every science has this trouble, but it is especially vicious in the social sciences, perhaps because social phenomena are so tangible, because social adjustments are so often unconscious and irrational, because human values are so much more closely connected with social data than with most non-social data. But it is also largely due to the fact that sociology is so often defined in terms of social betterment, welfare, progress, etc., and is so often taught and practiced by men who came into the field without sound scientific orientation. The note of evangelism sounds clear in much so-called sociological literature. Sociology, probably more than any other natural science, has recruited its professors from the ranks of disillusioned, discouraged preachers, missionaries, theological students, social workers, reformers and other religio-ethico-minded persons. It is difficult indeed for these leopards to change their spots. Some have done so, but many are distinctly torn between conflicting impulses. Many of us hold theoretically to the idea that we should study human beings as mycologists study fungi, but in practice, we usually come tardy off. We try to preserve the "dignity of man," the hegemony of the human, and are led into the fallacy that sociology is the "most important science," the most difficult, the most complex. Behold what great men are we! The vatic frenzy of much sociological writing is the proof of anthropocentrism. We think we must prove and promote "progress."
4. The Failure to Use Scientific Method. This grows out of the above misconceptions. Most of us are not yet clear as to the nature of mind and the nature of knowing. We lean too heavily upon the more "advanced" sciences. We are not thoroughly committed to the doctrine that social phenomena are natural phenomena and that social science must be natural science if it is to be science at all. From the security of our arm-chairs we still try to deal with data that are not sensible; hence, much of our "science" is nonsense. Most sociologists should read the "Grammar of Science" once a year.
It should be noted, however, that this criticism does not apply to many recently trained sociologists. The contempt of other natural scientists is forcing us to acquire a sounder scientific point of view. We are rapidly putting our house in order -so rapidly that from a research as well as a theoretical point of view, many sociologists know more about the logic and methods of science than do many of their most vitriolic critics.
5. Finally, it is contended that the assertion of the greater complexity of social phenomena is largely a defense mecha-
( 378) -nism of sociologists, a rationalization of their failure to deliver the "scientific goods."
From the point of view of this paper, complexity means nothing except in relation to organic adjustment. A situation is complex when adjustment is inadequate. It is "simple" when satisfactory adjustment had been attained. Consider a man driving in congested traffic for the first time. The situation is terribly "complex." He does not have adjustment mechanisms adequate to the situation. Soon, however, he is thinking about his business, talking animatedly to seatmates, and doing all manner of things other than "paying attention" to his driving. The same situation, formerly very "complex" is now ridiculously "simple." He has acquired adequate, automatic, habitual responses.
The same reasoning applies to all natural phenomena, including the social. Complexity merely means the non-existence of scientific categories describing relatively stable uniformities. Whenever natural phenomena fall into such energy-systems and we are able to make accurate (statistical) generalizations regarding their repetitive (predictable) behavior, complexity becomes simplicity. All scientific formulations are simple (though the techniques of making them may be very difficult), because they enable us to make more satisfactory adjustments within our universe of discourse and previously defined frames of reference.
It has been my purpose to show that there is no intrinsic reason for believing that social phenomena are more complex than physical phenomena. For the most part, social phenomena are at present conceived in commonsense terms; the empirical adjustments made on this basis are simple, tangible, and understandable to everyone. It is only when we transcend the narrow irrational limits of commonsense space-and-time spans that social phenomena seem complex. But as soon as we develop adequate adjustment patterns on the basis of scientific generalization, they suddenly become as simple as driving a car in heavy traffic, or dressing for dinner. We thus "learn our way about" in the scientific social world and it becomes very "simple."
The moral of the tale is that sociologists should spend less time pleading complexity and more time doing the research that changes "complexity" into simplicity. In many respects our task is much simpler, more tangible, less tedious, and more dramatic than the tasks of astronomers and bio-chemists.