504 THE AMERICAN JOURNAL OF SOCIOLOGY
Classification will of itself render impossible the method of reasoning from isolated examples. It will force the most recalcitrant intellects to genuine induction, because it will show that there are different groups of social types, and that what is true of one may not be true of another. Nothing will aid us better to avoid premature generalizations, the actual scourge of our science. Under the dominance of classification the presumption will be against crediting a group with a trait unless it has first been observed in that group. The present tendency is precisely the contrary. We now pass by involuntary shocks from generalization to limitation. Then we shall advance more prudently from the latter to the former.
It cannot be denied that there is at present among the sociologists a cer- tain horror of admitting an exception. We shall outgrow this repugnance
after the habit of classification has become firmly fixed We shall at
last develop a passion for the experimentum cruets, a passion which is the severe and infallible touchstone of the genuine scientist. With this perfec- tion of method there will be no chaos of facts that we cannot reduce to order. What real experimentation is to the physical sciences, the experimentum cruets is to those sciences that cannot make use of genuine experimentation. With- out it there can be no complete induction. We may not accept a law as cer- tain without having discovered and satisfactorily explained the exceptions to the rule. What is the value, for example, of the assertion that woman is in a servile and degraded condition among hunters, if I am unable either to demonstrate by a complete induction that there are no exceptions to the law, or to explain the exceptions by the increasing influence of some other cir- cumstance ?
This careful search for exceptions will, moreover, have the further advantage of putting us on the track of new laws, for the exceptions must themselves be manifestations of some law. There is no rest for the genuine scientific investigator until the exceptions are reduced to new laws. Hence we shall advance to the demonstration of laws that will be more precise and more true, because we shall have determined the conditions under which they prevail, *. e., the sphere of their action. We shall discover their depend- encies upon each other and their relation to more general laws. The laws then formulated will be no longer vague and ambitious assertions. They will rather contain candid expression of all the certainty and all the generalization to which our evidence entitles us. Then at last chatter will be banished from sociology, as it already is from other, more fortunate sci- ences. The chatterers will be stigmatized as dilettantes, who have no lot with true science, which is devoted to discovery.
In the constructive portions of his monograph Steinmetz dis- cusses proposed classifications of society by use of selected schemes grouped as follows :
I. Artificial classifications ; e.g.:
