It is quite evident, that, owing to the lack of knowledge of the historical development of American languages, convincing proof of genealogical relationship may be impossible to obtain, even where such relation exists; so that, from both a practical and a theoretical point of view, the solution of the problems of genetic relationship presents a large number of attractive problems.
Considering the complexity of this question, and the doubts that we entertain in regard to some of the principles to be followed in our inquiry, it seems probable that a safer basis will be reached by following out dialectic studies. Very little work of this kind has been done on our continent. James Owen Dorsey was able to point out a few phenomena pertaining to the inter-relation of Siouan dialects. Similar points have been made in regard to the Salish languages and in a few other cases, but no penetrating systematic attempt has been made to clear up the processes of differentiation by which modern American dialects have developed. It is fortunate for the prosecution of this study that quite a number of linguistic families in America are broken up into numerous strongly divergent dialects, the study of which will help us the more in the investigation of the relations between distinct languages, the more markedly they are differentiated. Siouan, Algonquin, Muskhogean, Salishan, Shoshonian, Wakashan, Caddoan, are languages of this type. They present examples of divergence of phonetic character, of differences in structure and vocabulary, that will bring us face to face with the problem of the origin of these divergent elements.
The more detailed study of American languages promises rich returns in the fields of the mechanical processes of linguistic development and of the psychological problems presented by languages of different types. In many American languages the etymological processes are so transparent, that the mechanism of phonetic adaptation stands out with great clearness. Contact-phenomena, and types of sound-harmony that affect more remote parts of words, occur with great frequency. Phonetic shifts between related dialects are easily observed, so that we can accumulate a large mass of material which will help to solve the question in how far certain phonetic processes may be of more or less universal occurrence.
Remotely related to this problem is the question that was touched upon by Gatschet, in how far the frequent occurrence of similar sounds for expressing related ideas (like the personal pronouns) may be due to obscure psychological causes rather than to genetic relationship. Undoubtedly, many hitherto unexpected types of processes will reveal themselves in the pursuit of these studies.
The variety of American languages is so great, that they will be of high value for the solution of many fundamental psychological problems.
The unconsciously formed categories found in human speech have not been sufficiently exploited for the investigation of the categories into which the whole range of human experience is forced. Here, again, the clearness of etymological processes in many American languages is a great help to our investigation.
The isolation of formal elements and of stems, or of co-ordinate stems, whichever the case may be, is easily performed, and the meaning of every part of an expression is determined much more readily than in the innumerable fossilized forms of Indo-European languages.
Lexicographic differentiation corresponds to the morphological differentiation of languages. Where ideas are expressed by means of separate stems or by subordinate elements, generalized stems will be found that express a certain action regardless of the instrument with which it has been performed; while, in languages that are not provided with these formal elements, a number of separate words
