natives are contained in the Eskimo publications printed in Greenland, which are devoted both to topics of daily interest and to ancient folk-lore. Similar conditions prevail in the Cherokee material collected by James Mooney, and in some of the daily papers printed in aboriginal languages. Even when good written records are available, control by means of the spoken language is necessary, because the expression of the written language may differ considerably from the spoken form.
Up to this time too little attention has been paid to the variety of expression and to the careful preservation of diction. We have rather been interested in the preservation of fundamental forms. Fortunately, many of the recorded texts contain, at least to some extent, stereotyped conversation and other formulas, as well as poetical parts, which give a certain insight into certain stylistic peculiarities, although they can seldom be taken as examples of the spoken language.
An added difficulty in the use of texts written by natives is that most are written by Indians who have had a modern school education. It may be observed in all parts of America that the native languages are being modified by the influence of European languages, not only in vocabulary, but also in phonetics and grammar. The far-reaching influence of these causes may be observed in a most striking manner in modern Mexican and other Central American languages, that have been under Spanish influence for centuries, and which not only have lost large parts of their vocabularies, that have disappeared with the ancient ideas, but which have also developed a new syntax, and, in part at least, new morphological forms. Modifications of this type are common in those regions where the intercourse between Indian and white is intimate, and particularly where the children are segregated from the parents. On the Pacific coast, for instance, the articulation of the glottalized fortis loses much of its strength, old words disappear, and new syntactical forms develop. Even the old facility of composition of stems tends to disappear. It is therefore necessary to obtain text material also from the older generation, because it is required for the study of the recent development of the languages.
On account of the difficulties and expense involved in the collection of texts, collectors have not only hesitated to obtain similar material from different individuals, but they have also confined themselves largely to the collections of native traditions. In some cases, native poetry has been included in the collections. Albert Gatschet recognized the need of varied material and collected texts on diverse topics in his studies of the Klamath, and J. Owen Dorsey published a collection of letters. The contents of the Eskimo publications and the native newspapers previously referred to also form a notable exception to this rule. Among later collectors, Drs. Goddard and Sapir have given particular attention to the collection of texts of varied contents. On the whole, however, the available material gives a one-sided presentation of linguistic data, because we have hardly any records of daily occurrences, every-day conversation, descriptions of industries, customs, and the like. For these reasons the vocabularies yielded by texts are one-sided and incomplete.
Notwithstanding the progress that during the last few decades has been made in the character of the material recorded, both as regards the accuracy of phonetic transcription and the character of the matter recorded, there is ample room for improvements of method.
With the extent of our knowledge of native languages, the problems of our inquiry have also assumed wider and greater interest. It is quite natural that the first task of the investigator was the registering and the rough classification of languages. It appeared very soon that languages are more or less closely related, and that comparison of brief
