Tangut is the dead Tibeto-Burman language of the Buddhist empire of Xīxià, which was destroyed in 1227 by the Golden Horde of the Mongol warlord Temuüjin, more commonly known as Genghis Khan (c. 1162–1227). The Tangut empire was established in 1032 and comprised the modern Chinese provinces of Gānsù, Shānxī and Níngxià, extending from the Yellow River in the east to Kökö Nōr (Chinese: Qīnghăi Hù) in the west. The northern frontier of the Xīxià empire skirted the city of Qumul (Chinese: Hāmì), the river Edzin Gol (Chinese: Ruò Shuĭ), the Hèlán hills and the Yellow River. In the south, the empire extended down into the present-day province of Sichuān. The Xīxià capital was situated in what is now the city of Yinchuān (formerly Níngxiàfŭ) on the left bank of the Yellow River.

Examination of the verb paradigms of Omotic can easily engender a sense of bewilderment; there is such a variation from language to language, even when, in other respects, the languages concerned appear to be quite closely related. Coming to Omotic from a language group such as Eaśt Cushitic, the contrast is especially striking. For example, when investigating any little known East Cushitic language (and there are a few left!), one starts out with a strong expectation of finding in the verbal system certain morphological elements and patterns (e.g., t ‘2nd person, 3rd feminine singular’; n ‘1st person plural’ a palatal vowel signalling ‘perfect’; a guttural verb signalling ‘imperfect’; a rounded vowel signalling a ‘subordinate function’ or ‘non-indicative mood’; etc.), and one is never entirely disappointed; the well known formatives, along with the anticipated patterns of distribution, crop up with an almost monotonous regularity. Similar predictabilities are not obviously the case when we look at Omotic. With virtually every new language investigated, new and unfamiliar features emerge. I have remarked on this before (Hayward, 1984: 324), but I am not alone in having done so; Andrzej Zaborski has also emphasized the high degree of innovation seen in the Omotic verb (Zaborski, 1986: 528). But I think the essence of the problem is not that there are no identifiable cognate formatives so much as that we have been unable to understand their historical significance, andso supply them with meaningful labels.

The present paper describes a palatographic investigation of Javanese coronal stops. Javanese exhibits stop consonants at both dental and alveolar places of articulation. In the current roman orthography, the dentals are represented by the single letters t, d, while the alveolars are represented by the digraphs th, dh. Thus, for example, titik [titI?] ‘full stop’ contrasts with thithik [titI?] ‘a little’.

Generation- and age-set systems are found in many parts of the world. They are of particular importance in Africa, and especially in East Africa where some ethnic groups operate socio-political and cultural systems in which generation-sets play a dominant role. Ethnographic descriptions of generation-set systems abound, but their theoretical understanding seems still to be inadequate. With examples from the Toposa and Turkana, this paper aims to contribute to the theory of generation-set systems. Moreover, the processes described and the ideas expressed may also contribute to the general theory of socio-political and cultural systems.

I should like to discuss briefly three Arabic treatises that are quite different from one another with regard to content, authorship, and provenance. They do, however, share one interesting feature. Each one survives in just two manuscript copies, both in a Yemeni hand, the one in Arabic script, the other also in the Arabic language but written in Hebrew characters. The circumstances of their survival are instructive concerning the intellectual histories of both Jews and Muslims in the Yemen, and especially with regard to the relationships that obtained between these two communities. We may also add that none of these treatises, each of which is of some interest to its respective field, has been the subject of even a cursory study.

Bārgīrī is a New Persian term seemingly of transparent meaning: ‘the taking up of loads’. As such, it is listed by Steingass, with various special meanings ‘Conviction, criminal charge; taking-in of a ship's cargo’, none of which covers its full, or perhaps most important, semantic range. Moreover, the word is of note both for its relation to the nomadic life-style, and its interest to Orientalist scholarship. It seems worth collecting the references to provide a wider setting.

Leyla Azami (sister of the Iranian folklorist, Cheragh Ali Azami), recently discussed the word in connexion with the transhumant life of the Sangsarī tribe in the south-eastern Elburz. Here it is defined as tahiya-i muqaddamāt-i safar ‘the preparation of the preliminaries for a journey’, events which take place ‘in the third decade of the month of Urdibihisht’—the second month of the Iranian calendar, April–May, when the pastoralists assemble, preparing for their move to high ground with the advent of spring.

On a recent (December 1990) visit to India, I had the opportunity of visiting the new excavations at Mahabalipuram (Māmallapuram) 50 km. south of Madras, carried out in 1990 by the Archaeological Department of Tamilnadu, and still in progress.

The work is in the area immediately to the north of the ‘Shore’ temple. So far uncovered is an apsidal-shaped tank, its curved end aligned south towards the middle portion of the shore temple. This middle portion consists of a mandapa enclosing a rock-cut image of reclining Visnu, Anantaśayana. From the eroded nature of this piece, and the fact that the Pallava structure over it was clearly an addition, I have long felt this Anantaśayana to antedate the shore temple. The erection of the latter would thus have ‘converted’ this area to Śaiva use, since, as extant, the shore temple consists of two cellae with liiigas, one to the East of the Visnu-figure, the other to the West.