Aύana has been referred to Skt. āpana, but there is nothing to suggest that it means “market” or “bazaar”. On the contrary, it is clear from the documents that it means “village”, “parish”, or “local township”, i.e. the small country towns or villages administratively dependent on Caota, the capital of the province. Further, -ύ- (= w) almost always occurs, not -v- which we would have expected out of an original -p-. It is possible, of course, that āpana might have been used in an extended sense “market-town”, but it can be explained perhaps easier out of Iranian. O.P. āvahana, Arm. lw. avan, both meaning “village”.

WHEN my translation of the Hudud Al-' Ālam was published in 1937, it was found impossible to swell the book1 by further remarks on the language of the Persian original, and I promised (p. xii) to deal with this problem at a later date. On the other hand, the quarter of a century which has elapsed since I began my work on the Hudud has brought a considerable number of new facts bearing on the interpretation of the data which I was trying to explain. In various ways my work has been continued, both in the articles which can be considered as a by-product of the Hudud2 and in contributions on fresh but cognate texts.3 The object of my present article is to bring the book up to date by integrating my own research and by completing it with references to the studies of other scholars.

In the Journal Asiatique for 1941–2 my friend the late Professor Sten Konow published an article entitled Une nouvelle forme aberrante de khotanais. This new type of Iranian is contained in the document P 410 brought back by the late Paul Pelliot from Tumšq, a ruined site near the modern Maralbashi.

For the study of the history of this region all the materials, unfortunately often fragmentary, which the various expeditions have recovered for us have proved and are still proving of great importance. The present Tumšuq fragment is a notable addition to this material.

Sten Konow gave with his study a facsimile of the MS. fragment, a transliteration (in which he had enjoyed the assistance of J. Filliozat) and a tentative translation, together with a glossary of the words according to his readings. Six years later he turned again to the document and in the Norsk Tidsskrift for Sprogvidenskap 14 (1947), pp. 156 ff., he published a second study of the document. It is a pleasure to recognize the merit of these pioneer studies, but neither could be considered as providing a clear interpretation. In one point, the reading of ai, the incorrect at is kept in the second study, although in the glossary to the first study J. Filliozat had pointed out that the sign was properly au.

The document is vitally important for Iranian dialectical studies. Hence a new treatment is well justified. The recognition that the document contains a type of Buddhist ordination service changed the whole problem of its interpretation.

The story of Sudhana is one of the most popular of the avadānas of Northern Hīnayāna Buddhism. There are two main versions of this story, one adopted by the Mahāsanghika school and the other by the mūla-Sarvāstivāda school. The former is preserved in the Mahāvastu under the title Kinnarījātaka, and a similar version of this story is found in a Chinese collection called Liu-tu-chi-ching ‘Collection [of tales to illustrate] the six pāramitās’, said to have been translated in approximately A.D. 270 (Taishō Tripiṭaka, in, no. 152, 44 f.) and accessible to us in Chavannes's Tripitạka (Cinq cents contes et apologues, no. 80). The mūla-Sarvāstivāda redaction is found in a Vinaya text of that school called the Bhaiṣajya-vastu. It was translated by I-ching and is referred to as Ken pen chouo … by Chavannes (iv, 133).

The exposition of different deity mantras forms an essential part of Tantric compendia. A mantra, a combination of letters thought to be of divine origin, is transmitted by a preceptor to a disciple in an initiation ritual whereby the disciple is authorized to practise its ritual repetition and strive for its perfection in order to attain supernatural powers. The different stages of the process from the selection of the mantra by the preceptor to its employment in desire–oriented rites by the practitioner shall be described in this paper mainly on the basis of chapters 15 and 16 of the Kulārṇavatantra (= KT), one of the most important texts of the Kaula School of Tantrism, written between the eleventh and fifteenth centuries A.D. These chapters were written at a stage when the technicalities of mantraśāstra had been fully developed and were widely accepted. Among other texts consulted were the Śāradātilakatantra (= ŚT) by Lakṣmaṇa Deśikendra with Rāghavabhaṭṭa's commentary entitled Padārthādarśa (= PĀ) (A.D. 1494), the Mantramahodadhi (= MM) by Mahīdhara (A.D. 1588/89) and the Tantrarāja(tantra) (= TR) with the commentary Manoramā (A.D. 1603/04).

The Xuanjie lu has long been recognized as an early recorded example of a printed text. The date of printing given in its preface suggests that its author, Hegan Ji, was about to be or had just been demoted from the governorship of Canton on the grounds of corruption, from where he was sent to Luoyang. This article argues for the latter place of publication, and suggests that the printing of ‘several thousand’ copies of this short work, while a notable achievement in itself, was perhaps primarily intended to recover the author's political fortunes—a goal in which the evidence suggests that he succeeded, to judge from a recently published inscription by one of his sons.

The corpus of Aramaic incantation bowls and their related texts opens a new approach to the lexicographical study of the Aramaic dialects in Babylonia of late antiquity. Some of these texts were copies of “Vorlagen” that the scribes no longer understood. Nevertheless, they are more reliable text sources than one supposes. Errors, garbled spellings, miscomprehensions and misreadings are always feasible, and are typical phenomena of copied texts. In the case of new text variants one can approach the puzzling words and text passages anew. This study deals first and foremost with words that are obvious corrupted spellings or scribal errors caused by text transmissions. There are also cases of the breakdown of standard spellings and orthographic conventions from the dialect of “Vorlage” that hide the lexical assignment of a word. Since one is dealing here with the earliest text material of the late Aramaic period, they can be taken as a significant contribution to the placing of many lexemes in existing dictionaries.

As a normative discipline Qur'anic exegesis shares both principles and terminology with the other Islamic sciences, and as such is not likely to have been articulated before the third/ninth century. Prominent in the fully elaborated system, as for example exhibited in the work of al-Zamakhshari (d. 538/1143), are the complementary principles of qiyās and taqdīr. While the former is commonly understood to represent the hermeneutical instrument called analogy, a typological description of qiyās will distinguish between applications of the principle which depend upon a textual similarity, and those which are derived from a rational or causal relation. Of the first type it may be said that there are as many kinds of analogy as there are means of establishing external (grammatical and lexical) affinity between different scriptural contexts. Underlying the second type of analogy is a unifying principle (ratio) independent of textual similarities, both explicit and implicit. In theory, if not always in practice, one may differentiate the two types of qiyās by reference to the incidence in each of what I have called the complementary principle of taqdīr. This term, of which the most common rendering ‘supplementation’ (Ergänzung), alluding to only one aspect of the procedure in question, is not quite satisfactory, signifies reconstruction or restoration (Wiederherstellung: restitutio in integrum), namely, of a scriptural context or passage. Now, while the elaboration of grammatical qiyās by the so-called ‘Basran school’ was characterized by an almost unlimited application of that principle, two reservations must be made.

The long-awaited first volume of the astronomical diaries, the life-work ofthe late Abraham Sachs, has been admirably brought to fruition by HermannHunger, in Sachs and Hunger, Diaries and related texts from Babylonia.2Although one laments the passing of Abraham Sachs before the completion ofthis momentous task, Hunger has earned the acclaim of his colleagues for thepolished presentation of the finished product, particularly as the Sachs Nachlasswas not yet in publishable form when he died. The importance of these diariesreaches far beyond the narrow confines of philology, since they includecalendrical and astronomical data, lists of commodity prices for the particularday or month, weather reports and often a tantalizing allusion to a currentevent, being the closest ancient counterpart to a modern news report.

This paper examines the evidence for the marginal feminine endings *-ay- and *-āy- in Proto-Semitic, and the feminine endings *-e and *-a in Proto-Berber. Their similar formation (*CV̆CC-ay/āy), semantics (verbal abstracts, underived concrete feminine nouns) and plural morphology (replacement of the feminine suffix by a plural suffix with -w-) suggest that this feminine formation should be reconstructed to a shared ancestor which may be called Proto-Berbero-Semitic.

Italy in the late fifth century a.d. and Persia in the mid-thirteenth were lands of ancient and deeply-rooted culture. One had been at the centre of the western Roman Empire; the other had formed a significant part of the society of medieval Islam, and its traditions stretched back further still. Both lands found themselves invaded, conquered, and ruled by wandering ‘barbarians’, Ostrogoths and Mongols respectively, from beyond the then recognized frontiers of ‘civilization’.

The paper addresses the frequent collocation /ir(hu)waliyan parittarwaliyan/, which occurs in Luwian incantations embedded in Hittite cuneiform ritual texts. Despite the relatively clear context, the meanings of this and similar collocations have remained obscure. Using combinatory and etymological methods, we intend to demonstrate that it hides the merism “internal (or) external”, which modifies various sorts of supernatural negative phenomena. Furthermore, we intend to argue that such an interpretation is compatible with the Late Bronze Age Anatolian beliefs about potential sources of evil. A collateral result of our demonstration is the elucidation of a number of additional contexts in the Hittite, Luwian, Lydian, and official Aramaic languages.

In my earlier examination of the phenomenon of taḍmīn in Arabic poetry, I attempted to show how it evolved from the status of a defect into a poetic device. But I could not at that time offer any explanation that would reasonably account for this development. Moreover, I have come to realize that my treatment of the term as referring to those instances where the full meaning of an idea started in one line emerges only in the succeeding line(s), viz. enjamb-ment, did not adequately account for its subtleties; and, more importantly, it ignored other phenomena which are also subsumed under the term, though they have no obvious connection with enjambment. It is precisely these lacunae that I intend to fill in the present study. For purposes of analysis I will designate the over-running of lines as ‘grammatical taḍmīn’ since the relationship in such cases is either syntactic or semantic, although still other, subtler relationships within this broad classification are demonstrable. The term is also used for cases where a poet deliberately quotes, with or without indication, from poems or statements by others: this will be discussed under ‘rhetorical taḍmīn’. Yet another use of the term arose from the introduction of philosophical thought into theoretical speculation, specifically in the area of scriptural interpretation: this I will designate as ‘hermeneutical taḍmīn’.

The Year 9 A.H. is known as 'ām al-wufūd, ‘the year of the delegations', on account of the many tribal delegations which came at that time to the Prophet at Medina. One of the most impressive was that of Tamīm. The accounts of its meeting with the Prophet are rich in detail, and involve both Qur'ānic verses and poems. As is often the case with our sources, there is a considerable amount of confusion between the various accounts. In this instance, however, the confusion is the result not merely of falsifications of failures of recall but also of the conception which guided the Muslim scholars in their redaction of the material. The compilers, rather than their primary sources, are responsible for the confusion which is the main subject of thispaper.

Legal manuscripts excavated from tombs serve as important materials for research on Qin and Han laws. These manuscripts differ from received legal texts or law documents found at archaeological sites in nature and function, as they were stored as funeral texts in tombs. This article studies the Ernian lüling manuscript in terms of its nature and function. It argues that the manuscript compiled in the second year of Empress Lü (186 bce) nearing the death of the owner was not produced for official use but specifically for burial in the tomb. This article further proposes that the burial of the Ernian lüling manuscript may have taken place to illustrate the social status and official capabilities of the owner to the underworld. The investigation of the Ernian lüling manuscript in its archaeological context helps us achieve a stronger understanding of the dating, origins, completeness, and compilation of its text.

The notion of text, along with the issues relating to how variable or fixed a text may be, has been a central issue in the study of oral literature for some time. In this article it will be argued that the notion of the poetic text as being unchangeable is central to Somali maanso poetry, and that such poems are instances of what become, through the composition process, ‘definitive texts’. Arguments in favour of this will be presented from a number of different perspectives. Maanso poetry will be contrasted with traditional hees and the characteristics of both outlined, with a view to showing how the ‘texts’ of each type of poetry differ from an extra-textual viewpoint, that is from the point of view of people's expectations regarding the different types of poetry. Intratextual characteristics will also be mentioned with a view to showing the objectivity and autonomy of maanso definitive texts. The development of modern hees from the heello and also the development of the use of writing in Somali poetry will also be considered and shown to support the idea of the maanso definitive text.

When the Aceh War finally broke out in 1873 it did not come as a surprise to most contemporaries. In colonial historiography Aceh, a ‘nest of pirates’, according to the Dutch, had obviously precipitated the war by its ‘treason’ in turning to Turkey, America and other foreign powers for support against the Dutch. The discussions with the American consul in Singapore early in 1873 on a possible Acehnese-American treaty constituted a casus belli in the eyes of the Dutch.

It has long been argued that the peculiar syntactic traits which typify the modern Semitic Ethiopian languages can be accounted for by the Cushitic substratum, notably the word order of the sentence and the system of qualification (the qualifier-qualified order of elements). It has been suggested that among these languages Tigre is ‘more Semitic’ since (a) it is much less rigid as regards word order, and (b) ‘it may optionally have either the Semitic pattern or the Cushitic one’. A parallel diachronic statement would be that as regards its syntax Tigre was more successful in resisting Cushitic influence or, for some reason, less exposed to it. Here, to avoid the problem of measuring the syntactic features of Tigre in general Semitic terms, Gə'əaz is usually taken as the mode for a Semitic language. In this respect it follows that descriptively Tigre is more Gə'əz-like than the other modern Semitic Ethiopian languages. In spite of the plausibility of such an explanation, it seems to be a matter of scholarly tradition rather than the result of thorough investigation, i.e. the belief that Tigre is the odd man out among its modern relations by virtue of its syntax is an impression rather than a proven observation.