128 THE AMERICAN JOURNAL OF SOCIOLOGY
we undertake to organize and direct economic action, to say what the policy of a community should be about a given subject — currency, taxation, hours of labor, etc. — we at once presuppose this economic abstraction, and a series of other abstractions synthesized into a social philosophy. The "great bad" of recent social theory has been the assumption that a working social philos- ophy is furnished by the economic abstraction alone.
(The process, just illustrated, is logically the same which the physicist, performs when he studies the breaking up of a pencil of light by passage through a prism. He watches the course of the different resulting rays, the angles of refraction, the distribution of the colors. He shuts out of view, e. g., all phenomena of heat, produced in the prism. Or he might reverse the abstraction, ignoring the optical and abstracting only the thermic phe- nomena.)
Generalizing abstraction consists in ignoring all the qualities of the phe- nomena under consideration which are liable to variation, from case to case ; and in emphasizing certain properties remaining common to the whole group of phenomena under discussion, and treating them as marks of a general concept. Thus the abstraction " solid " has no reference to special forms of solids, as cube, sphere, cone, etc., nor to particular substances that may be found in these forms. It abstracts the one trait of having the three dimen- sions, length, breadth, and height, in relatively permanent form. Or again, the abstraction "charity" does not take into account the diversified types of charity — hospitals, lodging houses, soup kitchens, bureaus of justice, etc.; it abstracts the single trait of voluntary effort for the welfare of others.
Generalizing abstraction falls into two subdivisions: (i) In case the sub- ject-matter of analysis is actual objects of observation or thought. In this case concepts of species result from the analysis, as in zoSlogy, or in political science when governments are classified, and traits common to some or all are, in turn, abstracted. (2) In case the subject-matter of analysis is propo- sitions vi\\\ch formulate certain relations oi objects. Here a process of isolat- ing abstraction is presupposed, and the product of this form of generalizing abstraction is abstract rules or laws. An instance is the physical law that "action and reaction are equal;" the psychical law that "as the twig is bent the tree's inclined;" or the social law that " individual interest yields to common weal."
Ordinary mental processes which precede scientific processes prepare in some measure both forms of abstraction. These uncritical processes have comparatively little effect upon isolating abstraction. On the other hand, they lend themselves far too readily to generalizing abstraction. Concepts which have been formed by uncritical generalization are caught up into learned tradition without being made to furnish scientific credentials, and henceforth they introduce confusion, fallacy, and contradiction. Yet the source of the trouble remains long undetected. A familiar form of the involved fallacy is the tacit assumption that concepts formed by elementary
