Although there are dance types in Japan where movement is unaccompanied by song, dance to song comprises a major part of the various traditions of dance in its classical theaters and is widely found in the countryside as well. In some cases, of course, song includes meaningless words, or the voice is used like an instrument, simply to provide a musical element or beat to which to dance. But it has seemed to me that a unique aesthetic underlies a full appreciation of dance to song. What I have in mind is an associational process in operation when the movement of a dance is seen while one simultaneously hears and understands the verbal meaning of its song.

It may be true that since Japanese is not my native language I directed more attention, while watching dance of this sort, to the meaning of the song than a native speaker might.

This paper represents an attempt to raise three points with regard to the study of dance and social anthropology. First, I have stressed the importance of an anthropological perspective in contrast to other perspectives in connection with ethnographies of dance. Second, I have briefly outlined a few ontological and epistemological implications of treating social anthropology itself as a language-based, rather than as a Behavioural science. Third, I have only barely indicated the epistemological consequences involved in accepting the idea of a personal anthropology.

In fact, it is the latter point which suffers most from the following brief treatment, for while the subject matter for a deeper analysis is present in this essay (i.e. the parts of texts of articles written before I read anthropology), it has mainly been subjected to a fairly standard anthropological critique. There are those who might say that this could have been done without the benefit of the idea of a personal anthropology.

As Russian ballet has become more important as a subject of historical inquiry, western scholars have become increasingly aware of the need to consult the rich bibliography of appropriate source material in the Russian language. Of all the eras addressed in this extensive literature, the late imperial period has quite understandably cast the most powerful spell over us. It was then, of course, that Tchaikovsky's music, Petipa's choreography, and the artistry of many still-remembered ballerinas first graced the stage. Like ourselves, the contemporaries of that period were conscious of the wealth of talent around them and sought to record their impressions of it. As witness to this we have the histories of Konstantin Skalkovsky, Alexander Pleshchejev and Sergei Khudekov, to name but three. More impressive than any of these works and a reference that their authors used, is the monumental Yearbook of the Imperial Theaters [Ezhegodnik Imperatorskikh Teatrov] —unquestionably the richest single source of information concerning the late imperial stage. In the following pages I shall describe this periodical and point out its value to ballet historians.