i9o6] ROYCE— PRINCIPLES OF THEORETICAL SCIENCE. 87 



initiating processes of logical deduction, ideas of which some general 

 and thoroughgoing account is possible, so that, although we cannot 

 predict the facts of the natural world, we can predict the forms in 

 terms of which we shall always be obliged to think the rational con- 

 nections of these facts in case we form any theory at all ? Are, then, 

 the internal conditions of theoretical science, the logical possibilities 

 upon which such a science depends, of a determinate range, and of 

 a knowable character? Such are the problems which are suggested 

 when we begin to inquire as to the logical position which quantita- 

 tive theories hold amongst the various types of theories which are 

 logically possible. 



The questions thus suggested are obviously of the most funda- 

 mental importance for any one who is interested in understanding 

 the workings of science. Science depends upon finding facts ; it 

 certainly also aims at the controlling of facts. The control which 

 is here in question may either mean the technical mastery of facts, 

 the power to produce them at will, or it may mean the prediction of 

 facts. But either kind of control is possible only in so far as we 

 possess something of the nature of a theory. And a theory involves 

 the construction and control, and logical linking of concepts which 

 have to be of our own making. Therefore, the study of the types 

 of concepts which we can construct and control and link, the study 

 of the forms and linkages which the nature of our thought makes 

 possible, is surely as serious a study, as the direct study of the facts 

 which we can hope to control through the use of our intelligence. 

 The pursuit of useful knowledge surely includes in the end a knowl- 

 edge of those logical processes of thought whereby we come to make 



an intelligent use of facts. 



II. 



The result of such considerations is that a science is needed 

 which I may provisionally call the morphology of theories. This 

 science is a branch of logic. And it is to this science that I now call 

 your attention. 



So far it is easy to define our problem, and to see that if solvable, 

 it must be an important problem. What the student unacquainted 

 with modern logic will find doubtful may be the assertion that such 

 a problem can at present be fruitfully studied. If you define this 



