124 THE AMERICAN JOURNAL OF SOCIOLOGY
society, nation, sovereignty, government, property, etc., are constructions of many past minds. They are not necessarily coherent constructions. We might throw into the air pieces of material which might for a moment assume shapes conformed to geometric definition. It would by no means follow that these fantastic shapes are permanent and useful correlations of that material. The like is true of social concepts. They may be merely the product of a series of psychological accidents. They may thus be very monstrosities of thought. It is impossible to tell how far back in the process of building up concepts out of concepts the vitiating false synthesis may have occurred. Hence the fully equipped sociologist must be able to criticise the synthetic process from the beginning. It is to be noticed, further, that synthesis may be of threefold stages, corresponding to those of analysis.
1. Elementary synthesis. This stage, to be sure, is usually taken for granted in elementary analytical synthesis.
To go back to our tree swaying in the gale, there has been a synthetic process which results in the judgment : " Tree is not man, nor mountain, nor horse." This elementary synthesis is taken for granted when analysis begins to separate the attributes in tree, so that there can be discrimination of oak, elm, maple tree, etc.
2. Causal synthesis, on the contrary, is a highly important stage of critical knowledge. It is a serious component of experimental processes. It has its place not merely in confirming the results of analysis by reversing the method of approach ; it brings new things to light.
3. Logical synthesis operates in all mathematical or other thought con- struction. These constructions are sometimes facilitated by concrete obser- vation (Anschauung), as in synthetic geometry, or in mechanical invention, when deliberately studied out, or in improvement of a social combination, like a railroad association, or a mutual benefit order. In other cases these con- structions follow out purely conceptual associations of ideas, as in the case of the demonstrations of Euclid or Hegel's dialectics.
While the method of Euclid is the least objectionable of these dififerent processes (of logical synthesis), it still shows plainly that it is really a process of reproductive synthesis. This does not always come to light in other appli- cations. In fact, the method of Euclid simply recasts analytical results in the synthetic form. When this is not the case, as in the synthetic (construc- tive) attempts of philosophical dialectics, it is all too easy for arbitrary com- binations of idejis to insinuate themselves into the place of valid logical constructions.'
In general, the synthetic method is of less wide application than the analytic. Facts defy synthetic construction after they have attained a certain degree of complexity. For example, physics and chemistry analyze the
•An illustration is the vagary of Novicow in deriving institutionalized aristocracy from the organic concept. In the same logical category is Bellamy's nationalism, George's single tax, Tolstoi's non-resistance, etc.
