The So called Science of Sociology, by H. G. Wells

An Englishman Looks at the World

It has long been generally recognised that there are two quite divergent ways of attacking sociological and economic questions, one that is called scientific and one that is not, and I claim no particular virtue in the recognition of that; but I do claim a certain freshness in my analysis of this difference, and it is to that analysis that your attention is now called. When I claim freshness I do not make, you understand, any claim to original discovery. What I have to say, and have been saying for some time, is also more or less, and with certain differences to be found in the thought of Professor Bosanquet, for example, in Alfred Sidgwick's "Use of Words in Reasoning," in Sigwart's "Logic," in contemporary American metaphysical speculation. I am only one incidental voice speaking in a general movement of thought. My trend of thought leads me to deny that sociology is a science, or only a science in the same loose sense that modern history is a science, and to throw doubt upon the value of sociology that follows too closely what is called the scientific method.

The drift of my argument is to dispute not only that sociology is a science, but also to deny that Herbert Spencer and Comte are to be exalted as the founders of a new and fruitful system of human inquiry. I find myself forced to depreciate these modern idols, and to reinstate the Greek social philosophers in their vacant niches, to ask you rather to go to Plato for the proper method, the proper way of thinking sociologically.

We certainly owe the word Sociology to Comte, a man of exceptionally methodical quality. I hold he developed the word logically from an arbitrary assumption that the whole universe of being was reducible to measurable and commeasurable and exact and consistent expressions.

In a very obvious way, sociology seemed to Comte to crown the edifice of the sciences; it was to be to the statesman what pathology and physiology were to the doctor; and one gathers that, for the most part, he regarded it as an intellectual procedure in no way differing from physics. His classification of the sciences shows pretty clearly that he thought of them all as exact logical systematisations of fact arising out of each other in a synthetic order, each lower one containing the elements of a lucid explanation of those above it—physics explaining chemistry; chemistry, physiology; physiology, sociology; and so forth. His actual method was altogether unscientific; but through all his work runs the assumption that in contrast with his predecessors he is really being as exact and universally valid as mathematics. To Herbert Spencer—very appropriately since his mental characteristics make him the English parallel to Comte—we owe the naturalisation of the word in English. His mind being of greater calibre than Comte's, the subject acquired in his hands a far more progressive character. Herbert Spencer was less unfamiliar with natural history than with any other branch of practical scientific work; and it was natural he should turn to it for precedents in sociological research. His mind was invaded by the idea of classification, by memories of specimens and museums; and he initiated that accumulation of desiccated anthropological anecdotes that still figures importantly in current sociological work. On the lines he initiated sociological investigation, what there is of it, still tends to go.

From these two sources mainly the work of contemporary sociologists derives. But there persists about it a curious discursiveness that reflects upon the power and value of the initial impetus. Mr. V.V. Branford, the able secretary of the Sociological Society, recently attempted a useful work in a classification of the methods of what he calls "approach," a word that seems to me eminently judicious and expressive. A review of the first volume the Sociological Society has produced brings home the aptness of this image of exploratory operations, of experiments in "taking a line." The names of Dr. Beattie Crozier and Mr. Benjamin Kidd recall works that impress one as large-scale sketches of a proposed science rather than concrete beginnings and achievements. The search for an arrangement, a "method," continues as though they were not. The desperate resort to the analogical method of Commenius is confessed by Dr. Steinmetz, who talks of social morphology, physiology, pathology, and so forth. There is also a less initiative disposition in the Vicomte Combes de Lestrade and in the work of Professor Giddings. In other directions sociological work is apt to lose its general reference altogether, to lapse towards some department of activity not primarily sociological at all. Examples of this are the works of Mr. and Mrs. Sidney Webb, M. Ostrogorski and M. Gustave le Bon. From a contemplation of all this diversity Professor Durkheim emerges, demanding a "synthetic science," "certain synthetic conceptions"—and Professor Karl Pearson endorses the demand—to fuse all these various activities into something that will live and grow. What is it that tangles this question so curiously that there is not only a failure to arrive at a conclusion, but a failure to join issue?

Well, there is a certain not too clearly recognised order in the sciences to which I wish to call your attention, and which forms the gist of my case against this scientific pretension. There is a gradation in the importance of the instance as one passes from mechanics and physics and chemistry through the biological sciences to economics and sociology, a gradation whose correlatives and implications have not yet received adequate recognition, and which do profoundly affect the method of study and research in each science.

Let me begin by pointing out that, in the more modern conceptions of logic, it is recognised that there are no identically similar objective experiences; the disposition is to conceive all real objective being as individual and unique. This is not a singular eccentric idea of mine; it is one for which ample support is to be found in the writings of absolutely respectable contemporaries, who are quite untainted by association with fiction. It is now understood that conceivably only in the subjective world, and in theory and the imagination, do we deal with identically similar units, and with absolutely commensurable quantities. In the real world it is reasonable to suppose we deal at most with practically similar units and practically commensurable quantities. But there is a strong bias, a sort of labour-saving bias in the normal human mind to ignore this, and not only to speak but to think of a thousand bricks or a thousand sheep or a thousand sociologists as though they were all absolutely true to sample. If it is brought before a thinker for a moment that in any special case this is not so, he slips back to the old attitude as soon as his attention is withdrawn. This source of error has, for instance, caught nearly the whole race of chemists, with one or two distinguished exceptions, and atoms and ions and so forth of the same species are tacitly assumed to be similar one to another. Be it noted that, so far as the practical results of chemistry and physics go, it scarcely matters which assumption we adopt. For purposes of inquiry and discussion the incorrect one is infinitely more convenient.

But this ceases to be true directly we emerge from the region of chemistry and physics. In the biological sciences of the eighteenth century, commonsense struggled hard to ignore individuality in shells and plants and animals. There was an attempt to eliminate the more conspicuous departures as abnormalities, as sports, nature's weak moments, and it was only with the establishment of Darwin's great generalisation that the hard and fast classificatory system broke down, and individuality came to its own. Yet there had always been a clearly felt difference between the conclusions of the biological sciences and those dealing with lifeless substance, in the relative vagueness, the insubordinate looseness and inaccuracy of the former. The naturalist accumulated facts and multiplied names, but he did not go triumphantly from generalisation to generalisation after the fashion of the chemist or physicist. It is easy to see, therefore, how it came about that the inorganic sciences were regarded as the true scientific bed-rock. It was scarcely suspected that the biological sciences might perhaps, after all, be truer than the experimental, in spite of the difference in practical value in favour of the latter. It was, and is by the great majority of people to this day, supposed to be the latter that are invincibly true; and the former are regarded as a more complex set of problems merely, with obliquities and refractions that presently will be explained away. Comte and Herbert Spencer certainly seem to me to have taken that much for granted. Herbert Spencer no doubt talked of the unknown and the unknowable, but not in this sense, as an element of inexactness running through all things. He thought of the unknown as the indefinable beyond to an immediate world that might be quite clearly and exactly known.

Well, there is a growing body of people who are beginning to hold the converse view—that counting, classification, measurement, the whole fabric of mathematics, is subjective and deceitful, and that the uniqueness of individuals is the objective truth. As the number of units taken diminishes, the amount of variety and inexactness of generalisation increases, because individuality tells more and more. Could you take men by the thousand billion, you could generalise about them as you do about atoms; could you take atoms singly, it may be you would find them as individual as your aunts and cousins. That concisely is the minority belief, and it is the belief on which this present paper is based.

Now, what is called the scientific method is the method of ignoring individualities; and, like many mathematical conventions, its great practical convenience is no proof whatever of its final truth. Let me admit the enormous value, the wonder of its results in mechanics, in all the physical sciences, in chemistry, even in physiology—but what is its value beyond that? Is the scientific method of value in biology? The great advances made by Darwin and his school in biology were not made, it must be remembered, by the scientific method, as it is generally conceived, at all. He conducted a research into pre-documentary history. He collected information along the lines indicated by certain interrogations; and the bulk of his work was the digesting and critical analysis of that. For documents and monuments he had fossils and anatomical structures and germinating eggs too innocent to lie, and so far he was nearer simplicity. But, on the other hand, he had to correspond with breeders and travellers of various sorts, classes entirely analogous, from the point of view of evidence, to the writers of history and memoirs. I question profoundly whether the word "science," in current usage anyhow, ever means such patient disentanglement as Darwin pursued. It means the attainment of something positive and emphatic in the way of a conclusion, based on amply repeated experiments capable of infinite repetition, "proved," as they say, "up to the hilt."

It would be, of course, possible to dispute whether the word "science" should convey this quality of certitude; but to most people it certainly does at the present time. So far as the movements of comets and electric trams go, there is, no doubt, practically cocksure science; and indisputably Comte and Herbert Spencer believed that cocksure could be extended to every conceivable finite thing. The fact that Herbert Spencer called a certain doctrine Individualism reflects nothing on the non-individualising quality of his primary assumptions and of his mental texture. He believed that individuality (heterogeneity) was and is an evolutionary product from an original homogeneity. It seems to me that the general usage is entirely for the limitation of the use of the word "science" to knowledge and the search after knowledge of a high degree of precision. And not simply the general usage: "Science is measurement," Science is "organised common sense," proud, in fact, of its essential error, scornful of any metaphysical analysis of its terms.

If we quite boldly face the fact that hard positive methods are less and less successful just in proportion as our "ologies" deal with larger and less numerous individuals; if we admit that we become less "scientific" as we ascend the scale of the sciences, and that we do and must change our method, then, it is humbly submitted we shall be in a much better position to consider the question of "approaching" sociology. We shall realise that all this talk of the organisation of sociology, as though presently the sociologist would be going about the world with the authority of a sanitary engineer, is and will remain nonsense.

In one respect we shall still be in accordance with the Positivist map of the field of human knowledge; with us as with that, sociology stands at the extreme end of the scale from the molecular sciences. In these latter there is an infinitude of units; in sociology, as Comte perceived, there is only one unit. It is true that Herbert Spencer, in order to get classification somehow, did, as Professor Durkheim has pointed out, separate human society into societies, and made believe they competed one with another and died and reproduced just like animals, and that economists, following List, have for the purposes of fiscal controversy discovered economic types; but this is a transparent device, and one is surprised to find thoughtful and reputable writers off their guard against such bad analogy. But, indeed, it is impossible to isolate complete communities of men, or to trace any but rude general resemblances between group and group. These alleged units have as much individuality as pieces of cloud; they come, they go, they fuse and separate. And we are forced to conclude that not only is the method of observation, experiment, and verification left far away down the scale, but that the method of classification under types, which has served so useful a purpose in the middle group of subjects, the subjects involving numerous but a finite number of units, has also to be abandoned here. We cannot put Humanity into a museum, or dry it for examination; our one single still living specimen is all history, all anthropology, and the fluctuating world of men. There is no satisfactory means of dividing it, and nothing else in the real world with which to compare it. We have only the remotest ideas of its "life-cycle" and a few relics of its origin and dreams of its destiny …

Sociology, it is evident, is, upon any hypothesis, no less than the attempt to bring that vast, complex, unique Being, its subject, into clear, true relations with the individual intelligence. Now, since individual intelligences are individual, and each is a little differently placed in regard to the subject under consideration, since the personal angle of vision is much wider towards humanity than towards the circumambient horizon of matter, it should be manifest that no sociology of universal compulsion, of anything approaching the general validity of the physical sciences, is ever to be hoped for—at least upon the metaphysical assumptions of this paper. With that conceded, we may go on to consider the more hopeful ways in which that great Being may be presented in a comprehensible manner. Essentially this presentation must involve an element of self-expression must partake quite as much of the nature of art as of science. One finds in the first conference of the Sociological Society, Professor Stein, speaking, indeed a very different philosophical dialect from mine, but coming to the same practical conclusion in the matter, and Mr. Osman Newland counting "evolving ideals for the future" as part of the sociologist's work. Mr. Alfred Fouillée also moves very interestingly in the region of this same idea; he concedes an essential difference between sociology and all other sciences in the fact of a "certain kind of liberty belonging to society in the exercise of its higher functions." He says further: "If this view be correct, it will not do for us to follow in the steps of Comte and Spencer, and transfer, bodily and ready-made, the conceptions and the methods of the natural sciences into the science of society. For here the fact of consciousness entails a reaction of the whole assemblage of social phenomena upon themselves, such as the natural sciences have no example of." And he concludes: "Sociology ought, therefore, to guard carefully against the tendency to crystallise that which is essentially fluid and moving, the tendency to consider as given fact or dead data that which creates itself and gives itself into the world of phenomena continually by force of its own ideal conception." These opinions do, in their various keys, sound a similar motif to mine. If, indeed, the tendency of these remarks is justifiable, then unavoidably the subjective element, which is beauty, must coalesce with the objective, which is truth; and sociology mast be neither art simply, nor science in the narrow meaning of the word at all, but knowledge rendered imaginatively, and with an element of personality that is to say, in the highest sense of the term, literature.

If this contention is sound, if therefore we boldly set aside Comte and Spencer altogether, as pseudo-scientific interlopers rather than the authoritative parents of sociology, we shall have to substitute for the classifications of the social sciences an inquiry into the chief literary forms that subserve sociological purposes. Of these there are two, one invariably recognised as valuable and one which, I think, under the matter-of-fact scientific obsession, is altogether underrated and neglected The first, which is the social side of history, makes up the bulk of valid sociological work at the present time. Of history there is the purely descriptive part, the detailed account of past or contemporary social conditions, or of the sequence of such conditions; and, in addition, there is the sort of historical literature that seeks to elucidate and impose general interpretations upon the complex of occurrences and institutions, to establish broad historical generalisations, to eliminate the mass of irrelevant incident, to present some great period of history, or all history, in the light of one dramatic sequence, or as one process. This Dr. Beattie Crozier, for example, attempts in his "History of Intellectual Development." Equally comprehensive is Buckle's "History of Civilisation." Lecky's "History of European Morals," during the onset of Christianity again, is essentially sociology. Numerous works—Atkinson's "Primal Law," and Andrew Lang's "Social Origins," for example—may be considered, as it were, to be fragments to the same purport. In the great design of Gibbon's "Decline and Fall of the Roman Empire," or Carlyle's "French Revolution," you have a greater insistence upon the dramatic and picturesque elements in history, but in other respects an altogether kindred endeavour to impose upon the vast confusions of the past a scheme of interpretation, valuable just to the extent of its literary value, of the success with which the discrepant masses have been fused and cast into the shape the insight of the writer has determined. The writing of great history is entirely analogous to fine portraiture, in which fact is indeed material, but material entirely subordinate to vision.

One main branch of the work of a Sociological Society therefore should surely be to accept and render acceptable, to provide understanding, criticism, and stimulus for such literary activities as restore the dead bones of the past to a living participation in our lives.

But it is in the second and at present neglected direction that I believe the predominant attack upon the problem implied by the word "sociology" must lie; the attack that must be finally driven home. There is no such thing in sociology as dispassionately considering what is, without considering what is intended to be. In sociology, beyond any possibility of evasion, ideas are facts. The history of civilisation is really the history of the appearance and reappearance, the tentatives and hesitations and alterations, the manifestations and reflections in this mind and that, of a very complex, imperfect elusive idea, the Social Idea. It is that idea struggling to exist and realise itself in a world of egotisms, animalisms, and brute matter. Now, I submit it is not only a legitimate form of approach, but altogether the most promising and hopeful form of approach, to endeavour to disentangle and express one's personal version of that idea, and to measure realities from the stand-point of that idealisation. I think, in fact, that the creation of Utopias—and their exhaustive criticism—is the proper and distinctive method of sociology.

Suppose now the Sociological Society, or some considerable proportion of it, were to adopt this view, that sociology is the description of the Ideal Society and its relation to existing societies, would not this give the synthetic framework Professor Durkheim, for example, has said to be needed?

Almost all the sociological literature beyond the province of history that has stood the test of time and established itself in the esteem of men is frankly Utopian. Plato, when his mind turned to schemes of social reconstruction thrust his habitual form of dialogue into a corner; both the "Republic" and the "Laws" are practically Utopias in monologue; and Aristotle found the criticism of the Utopian suggestions of his predecessors richly profitable. Directly the mind of the world emerged again at the Renascence from intellectual barbarism in the brief breathing time before Sturm and the schoolmasters caught it and birched it into scholarship and a new period of sterility, it went on from Plato to the making of fresh Utopias. Not without profit did More discuss pauperism in this form and Bacon the organisation of research; and the yeast of the French Revolution was Utopias. Even Comte, all the while that he is professing science, fact, precision, is adding detail after detail to the intensely personal Utopia of a Western Republic that constitutes his one meritorious gift to the world. Sociologists cannot help making Utopias; though they avoid the word, though they deny the idea with passion, their very silences shape a Utopia. Why should they not follow the precedent of Aristotle, and accept Utopias as material?

There used to be in my student days, and probably still flourishes, a most valuable summary of fact and theory in comparative anatomy, called Rolleston's "Forms of Animal Life." I figure to myself a similar book, a sort of dream book of huge dimensions, in reality perhaps dispersed in many volumes by many hands, upon the Ideal Society. This book, this picture of the perfect state, would be the backbone of sociology. It would have great sections devoted to such questions as the extent of the Ideal Society, its relation to racial differences, the relations of the sexes in it, its economic organisations, its organisation for thought and education, its "Bible"—as Dr. Beattie Crozier would say—its housing and social atmosphere, and so forth. Almost all the divaricating work at present roughly classed together as sociological could be brought into relation in the simplest manner, either as new suggestions, as new discussion or criticism, as newly ascertained facts bearing upon such discussions and sustaining or eliminating suggestions. The institutions of existing states would come into comparison with the institutions of the Ideal State, their failures and defects would be criticised most effectually in that relation, and the whole science of collective psychology, the psychology of human association, would be brought to bear upon the question of the practicability of this proposed ideal.

This method would give not only a boundary shape to all sociological activities, but a scheme of arrangement for text books and lectures, and points of direction and reference for the graduation and post graduate work of sociological students.

Only one group of inquiries commonly classed as sociological would have to be left out of direct relationship with this Ideal State; and that is inquiries concerning the rough expedients to meet the failure of imperfect institutions. Social emergency work of all sorts comes under this head. What to do with the pariah dogs of Constantinople, what to do with the tramps who sleep in the London parks, how to organise a soup kitchen or a Bible coffee van, how to prevent ignorant people, who have nothing else to do, getting drunk in beer-houses, are no doubt serious questions for the practical administrator, questions of primary importance to the politician; but they have no more to do with sociology than the erection of a temporary hospital after the collision of two trains has to do with railway engineering.

So much for my second and most central and essential portion of sociological work. It should be evident that the former part, the historical part, which conceivably will be much the bulkier and more abundant of the two, will in effect amount to a history of the suggestions in circumstance and experience of that Idea of Society of which the second will consist, and of the instructive failures in attempting its incomplete realisation.