III.
Already, in biology the nature of the object had compelled scientific men to start from the consideration of the whole to reach that of the parts, to proceed from the complex to the simple. With still greater reason, the same inversion of method imposes itself in sociology. For, although the individual elements of society appear to be more separable than those of the living being, the social consensus is still closer than the vital consensus.[240]
The spirit of the sociological method will then be always to consider simultaneously the various social aspects, whether in statics, or in dynamics. Undoubtedly each of them can be the object of a special study, by the way of “preliminary elaboration.” But, as soon as the science is sufficiently advanced, the correlation of phenomena will serve as a guide for their analysis. Political economy has proved by facts that the isolated study of a series of social phenomena is condemned to remain irrational and barren. Those then who, in the system of social studies, wish to imitate “the methodical parcelling out, which belongs to the inorganic sciences,” misunderstand what the essential conditions of their subject require. Here the most general laws must be known first. It is from them that science must then descend to the more particular laws.
The more complex the phenomena, the more numerous are the processes of method at our disposal for studying them. This law of compensation is verified again in the present case. Sociology, over and above the processes made use of by the preceding sciences possesses some which are peculiarly its own. To put it more plainly, in its capacity of final science, the whole positive method belongs to it. As method is only learnt by practice, the sociologist will therefore have to be formed by a complete scientific education from mathematics, which will give him the feeling of positivity, to biology which will teach him the comparative method. The Cours de philosophie positive precisely retraces this methodical ascent, which leads the human mind, by successive degrees, up to social science. And, since the intellectual evolution of the individual reproduces that of the species, the sociologist will cover the same ground to reach the same end.
At any rate, if a mathematical education is indispensable so as to accustom him to the positive mode of thought, he will, however, acknowledge that social phenomena do not allow of the use of numbers or of mathematical analysis, nor more especially of the calculation of probabilities. Comte treats Laplace’s attempt upon this point as absurd, an attempt which has been taken up again by other mathematicians. He likes to quote it as a proof of the lack of the philosophical spirit among geometers. Indeed, according to him, to apply the calculation of probabilities to historical events, implies a failure to understand that these phenomena are subject to invariable laws like all other phenomena.
In default of the powerful instrument furnished by mathematics, sociology makes use of the methods employed in the physical and natural sciences. Of these observation is the first. Social phenomena seem easy to observe, because they are very common, and the observer takes part in them more or less. But, on the contrary, these two circumstances render sociological observation very difficult. We only observe well on condition that we place ourselves outside what we observe.[241] Sociological facts ought then to appear objective to us, detached from us, independent of the state of our individual consciousness. Nothing is more difficult to realise. In order to obtain, and more especially to maintain, “such an inversion of the spontaneous point of view,” the mind must already have partly constructed what it wishes to see. Were it not already provided with a preliminary theory, for the most part the observer would not know what he must look for in the fact which is taking place under his eyes. It is therefore by the preceding facts that we learn to see the following ones. There lies “the immense difficulty” of sociology, in which we are thus obliged, in a certain measure, to determine simultaneously the facts and the laws. If we are not already possessed of the necessary speculative indications to grasp them, the facts remain barren and even unseen, although we are, so to speak, immersed in them.
Consequently, a social fact can have no scientific significance if it is not brought into relation with another fact. In an isolated condition, it remains in the state of a simple anecdote, capable at most of satisfying “idle curiosity,” but unfit for any rational use. An infinite number of facts may be useful to sociology, apparently very insignificant customs, all kinds of monuments, the analysis and the comparison of languages; but the mind must be provided for their observation with general points of view. Only on this condition will a mind, well prepared by rational education, be able to transform the actions which take place beneath its eyes into sociological indications, “according to the more or less direct points of contact, which he will be able to discern in these actions with the highest notions of science, in virtue of the connexion of the various social aspects.”
There can be no question of experimenting in sociology.[242] Not that we cannot act upon the social phenomena: they are, on the contrary, the most modifiable of all. But an experiment properly so-called consists in comparing two cases which differ from each other by a certain definite circumstance, and by that one alone. We have no means of determining two cases of this kind in sociology. It is true that in the absence of direct experiments nature presents indirect ones. They are the pathological cases, unfortunately too frequent in the life of societies, the more or less serious perturbations which they undergo through accidental or passing causes. Such are the revolutionary periods which correspond to diseases in living bodies. If we properly extend Broussais’ principle to sociology, that is to say, if we admit that morbid phenomena are produced by the effect of the same laws as normal phenomena, then social pathology will in some measure replace experiments. It will be said that this study has been fruitless up to the present time. But the reason of this is, according to Comte, that direct or indirect experimenting ought, like simple observation, to be subject to rational conceptions. Both are only productive in a sociology already possessed of its essential laws.
The comparative method, so useful to the biologist, is also precious for the sociologist. It draws together the various states of human society which coexist on the different parts of the earth’s surface, and among peoples independent of one another. Undoubtedly, if the total development only is considered, the evolution of Humanity is one. It nevertheless remains true that very considerable and very varied populations have as yet only reached the more or less inferior degrees of this evolution. We can thus observe them simultaneously and compare their successive phases. From the Fuegians to the most civilised nations in Europe, we can imagine no “social shade” which is not at present realised on some portion of the globe. Frequently, within the same nation, the social condition of the various classes represents states of civilisation which are very far removed from one another. Paris to-day contains more or less faithful “survivors” of nearly all the anterior degrees of social evolution, especially from the intellectual point of view.[243] This comparative process holds good for social statics as for social dynamics. Even in statics a comparison can be established between animal societies and human society.
However, this type of method is not devoid of inconvenience in sociology. It does not consider the necessary succession of the various phases in the social evolution: it seems on the contrary to consider them all as simultaneous. Consequently, it prevents us from seeing the filiation of social forms. It also runs the risk of falsifying the analysis of the cases which are observed, and of causing simple secondary factors to be taken for main causes. This is what happened to Montesquieu who compared indifferently the cities of antiquity, the France of the Middle Ages, the England of the XVIII. century, the republic of Venice, the government of Byzantium, the Empire of the Sultan, and that of the Shah of Persia.
So the comparative method is only an auxiliary process in sociology. Like observation and experiment, it has to be made subordinate to a rational conception of the evolution of humanity. The latter in turn depends upon the use of an original method of observation, belonging to social phenomena, and free from the dangers presented by the preceding ones. This specific sociological method, this “transcendent” process, by which the positive method is completed, is, says Comte, the historical method.[244]