This paper presents a method for the analysis of connected speech (or writing). The method is formal, depending only on the occurrence of morphemes as distinguishable elements; it does not depend upon the analyst's knowledge of the particular meaning of each morpheme. By the same token, the method does not give us any new information about the individual morphemic meanings that are being communicated in the discourse under investigation. But the fact that such new information is not obtained does not mean that we can discover nothing about the discourse but how the grammar of the language is exemplified within it. For even though we use formal procedures akin to those of descriptive linguistics, we can obtain new information about the particular text we are studying, information that goes beyond descriptive linguistics.

When in 1925 I set about the task of writing an Indo-European comparative dictionary, I was hampered by lack of material on Albanian. I turned aside from the main theme to collect what I could both inside and outside Albania, and my IE materials were enriched thereby. Walde-Pokorny appeared in 1929–32, at a time when my own work was already well advanced. My results, while happily coinciding with theirs at many points, are nevertheless more fully documented, and are drawn up according to a different plan. In this article I offer some of these results as far as they concern Albanian.

Linguistic geography may simply present the facts about a linguistic detail on which there has been speculation; it may enable a speaker to judge the social acceptability of a linguistic form on the basis of usage rather than prejudice; or it may throw light on the history of the language, and suggest the forces that have led to the prestige of linguistic forms. Conclusions of all these types can be drawn from an examination of the occurrence or loss of initial /h-/ before semivowels in the states of the Atlantic seaboard, as recorded in the field records for the Linguistic Atlas of the United States and Canada.

The Semitic languages of Ethiopia occupy a geographical domain in which Cushitic was and still is employed. When the Semites of South Arabia conquered Ethiopia, they imposed their Semitic language on this Cushitic domain. A stage of bilingualism followed, which still endures. The Cushitic group lost ground, most probably because of the political prestige of the conquering Semites, but not without exerting a great influence on the structure of the Semitic languages. In fact, the Semitic languages of Ethiopia are in some respects quite far removed from the typical Semitic pattern. Some of these differences are no doubt due to the internal development of Semitic in Ethiopia; others may be inherited from the Semitic group that came from South Arabia, concerning which we are not sufficiently informed; but many of them are certainly a result of Cushitic influence.

Bariba is spoken by an estimated 150,000 people in the Cercles of Parakou, Kandi, and Natitingou in Dahomey, and by an undetermined number of people in adjacent sections of Nigeria. The position of Bariba among the West Sudanic languages must for the time being remain undecided. It certainly belongs in the general classification of Voltaic, but is not closely related to Gurma in the Eastern group of Voltaic languages. Bariba is remarkably similar in structure to languages in the Senufo group in the Ivory Coast and Sudan, but my knowledge of other Voltaic languages is insufficient to justify a more specific statement.