6l8 THE AMERICAN JOURNAL OF SOCIOLOGY
deliberate attempts to construct plans of research that would conform to the principles of exact science. The consequence is that, while sociology up to date can show comparatively little in the way of absolutely new knowledge about society, it has accu- mulated a wealth of perception about the' value of different por- tions of knowledge, and about ways in which knowledge of society must be tested and organized. Although these percep- tions are not yet coordinated in any system which is generally accepted by sociologists, there is an unformulated consensus about standards of objectivity and correlation which is steadily reducing sociological speculation to the soberness of observa- tional and experimental science."
Each of the chief types of sociological theory has contrib- uted something to this result. Perhaps the largest contributions have been not direct, but indirect. There may be close paral- lelism here between the merit of the sociologists and that of the philosophers of history. The share of the sociologists in the result may be quite different from the spirit of their own prem- ises. We may trace, however, in the progress of sociological theory, first, a reaction and a protest against speculative social philosophy ; second, a struggle by men still wearing the shackles of speculative tradition to perfect a positive method ; third, attrition among pseudo-positive methods. Reciprocal criticism of schools and programs of sociological inquiry is still the order of the day and unfortunately the chief employment of the sociolo- gists. Out of all this preliminary maneuvering a sociological method is emerging. It is an organization of ways of knowing society as it is. This is a substitute for the ways in which people
' " Unfortunately, the relation of facts is always less simple than we think ; the demand of our intellect for unity is often a little too strong, especially in the realm of social science. Hasty conclusions are still the order of the day. One assumes some- thing, not because it is so, because one has actually so observed it, but because it would agree so finely with something else. This is all very unscientific, but it suits our best thinkers not seldom. Really, we proceed still from the theory and seek facts merely for illustration. If one does otherwise, starts from the facts and goes no farther than they permit, then people are astonished that his result is not so beauti- fully rounded off, not so faultless, as their own fancies. That the latter, even if ever so consistent, harmonious, complete, are yet absolutely worthless — that does not appear to such people." (Steinmetz, Zeitschrift fur Socialwissenschaft, August, 1898.)
