That would have been, indeed, the one fundamental mistake, as the whole Congress work was planned in the service of the great synthetic movement which pervades the intellectual life of to-day. The undertaking would be useless and even hindering if it were not just the newer and deeper tendencies that came to most complete expression in it. Everything depended, therefore, upon the fullest possible representation of scientific endeavors in the plan. But no one can become aware of this manifoldness and of the logical relations who does not go back to the ultimate principles of the human search for truth. We have, therefore, to enter now into a full discussion of the principles which have controlled the classification and subdivision of the whole work. The discussion is necessarily in its essence a philosophical one, as it was earlier made plain that philosophy must lay out the plan, while in the realization of the plan through concrete work the scientist alone, and not the logician, has to speak. Yet here again it may be said that while our discussion of principles in its essence is logical, in another respect it is a merely historical account. The question is not what principles of classification are to be acknowledged as valuable now that the work of the Congress lies behind us, but what principles were accepted and really led to the organization of the work in that form in which it presents itself in the records of the following volumes.
The problem of dividing and subdividing the whole of human knowledge and of thus bringing order into the manifoldness of scientific efforts has fascinated the leading thinkers of all ages. It may often be difficult to say how far the new principles of classification themselves open the way for new scientific progress and how far the great forward movements of thought in the special sciences have in turn influenced the principles of classification. In any case every productive age has demanded the expression of its deepest energy in a new ordering of human science. The history of these efforts leads from Plato and Aristotle to Bacon and Locke, to Bentham and Ampère, to Kant and Hegel, to Comte and Spencer, to Wundt and Windelband. And yet we can hardly speak of a real historical continuity. In a certain way every period took up the problem anew, and the new aspects resulted not only from the development of the sciences themselves which were to be classified, but still more from the differences of logical interest. Sometimes the classification referred to the material, sometimes to the method of treatment, sometimes to the mental energies involved, and sometimes to the ends to be reached. The reference to the mental faculties was certainly the earliest method of bringing order into human knowledge, for the distinction of the Platonic philosophy between dialectics, physics, and ethics pointed to the threefold character of the mind, to reason, perception, and desire; and it was on the threshold of the modern time, again, when Bacon divided the intellectual globe into three large parts according to three fundamental psychical faculties: memory, imagination, and reason. The memory gives us history; the imagination, poetry; the reason, philosophy, or the sciences. History was further divided into natural and civil history; natural history into normal, abnormal, and artificial phenomena; civil history into political, literary, and ecclesiastical history. The field of reason was subdivided into man, nature, and God; the domain of man gives, first, civil philosophy, parted off into intercourse, business, and government, and secondly, the philosophy of humanity, divided into that of body and of soul, wherein medicine and athletics belong to the body, logic and ethics to the soul. Nature, on the other hand, was divided into speculative and applied science,--the speculative containing both physics and metaphysics; the applied, mechanics and magic. All this was full of artificial constructions, and yet still more marked by deep insight into the needs of Bacon's time, and not every modification of later classifiers was logically a step forward.