BRĀHMĪ, Indian script used for a variety of languages in Chinese Turkestan, including Iranian languages.
From the Tarim Basin (Xinjiang, China) we have first-millennium documents in Brāhmī script in several Iranian languages. There are substantial quantities of Khotanese (q.v.), at least fifteen Tumshuqese (q.v.) manuscripts, and there exists one small piece of Sogdian (q.v.). The scripts of the documents in question differ markedly from each other with reference to the shapes of individual signs. It also must be noted that significant divergences exist among the writing systems with respect to the ways in which the sounds of the languages are accorded graphic representation.
Paleography. The standard description of the evolution of Brāhmī character shapes in the Tarim Basin is given by Sander (esp. pp. 2-6, 43-47, 181-88 and tables 29-41). She distinguishes four stages: 1. Indian Gupta, 2. Turkestan Gupta, 3. Early Turkestan Brāhmī, and 4. North and South Turkestan Brāhmī. The North Turkestan Brāhmī (NTB) was once called the “Slanting Gupta” and is most familiar from Tocharian. The South Turkestan Brāhmī was formerly labelled the “Upright” or “Formal Gupta” and is best known from Old Khotanese.
It is further necessary to distinguish “literary” from “business” scripts; these are sometimes inaccurately described as “formal” and “cursive” respectively. Although the two types are closely related, the more efficient and quickly written business style is not simply a cursive derivative of the carefully written literary script. They evolved in tandem within the same culture, and the business script may in fact preserve archaic features which have been lost in the literary style.
A small number of Tumshuqese documents are in the North Turkestan literary shape while the majority are written in the North Turkestan business script. With Khotanese the situation is complicated. Apparently, in the early period Old Khotanese was written in the literary South Turkestan as well as in a very different business script which is usually, but misleadingly, called “cursive” and so here will be referred to as Cursive. At some point, probably during the Late Khotanese stage, a new literary style was derived from the Cursive and subsequently the South Turkestan Brāhmī fell into disuse. Thus, when Late Khotanese monks recopied Old Khotanese texts in South Turkestan Brāhmī, they transliterated it into the literary Cursive type of character. This explains why we find just Old Khotanese written in South Turkestan Brāhmī but both Old and Late Khotanese in the Cursive.
Linguistic principles. Because the Brāhmī was originally used to write Indian languages, certain additions and modifications were necessary in order to adapt the system to Iranian, Turkish, and Tocharian. Significantly disparate innovations were used in Tumshuq and Khotan which probably means that the two new systems are not directly connected. However, they do share at least two non-Indic features, the vowel sign ä and the ligature rr, which suggests some kind of relationship through an unattested third system.
Old Khotanese. The adaptations include digraphs (combinations of consonant signs), doubling of consonant signs, and new signs. The digraph ys stood for /z/, tc for /ts/ and js for /dz/. Doubled śś and ṣṣ denoted voiceless sibilants while single ś and ṣ marked their voiced counterparts. Doubled tt was voiceless and single t voiced, which permitted d to represent the voiced dental fricative /δ/. Doubled gg denoted the voiced velar stop while g alone stood for the fricative /γ/. With respect to new signs, Khotanese used what is commonly referred to as a subscript “hook,” now usually transliterated by an apostrophe. This sign mostly indicates “the recent loss of an internal sound,” usually /ẓ/ (Emmerick, 1979, p. 9). Two new vowel graphs existed, the two dot diacritic ä and an e stroke with a crossbar, transcribed as ei.
Late Khotanese. The devices are the same as in the older language with the important difference that singly written ś and ṣ were voiceless while these graphs in conjunction with the subscript “hook” denoted the voiced counterparts, i.e., śʾ/ź/ and ṣʾ/ẓ/.
Tumshuqese. Our understanding of the script is handicapped by our still limited knowledge of the language. Only one digraph, ts, was used and there was no character doubling, but many new signs were added. Some are modifications of existing graphs, others are apparently borrowed. xš may be seen as derived from ṣ, the voiced śˊ from ś, and kh from kh, in all cases by the addition of a tail. ẓ appears to be from rr with an added loop at the base. w is actually older o, assigned a new function, and ṟ (ra = rä) is identical with subscript ṛ in Sanskrit. The new signs z, ḏ, dz, g may be borrowed from Manichean script z, t, ǰ, g respectively (Hitch, pp. 298-300). z′ may actually be two distinct signs, one a gh with a tail and the other of unclear origin. The Tumshuqese, Tocharian, and Old Turkic Brāhmī systems are related, but of these three only Tumshuqese and Turkic share z, ḏ, dz, g, and ź. It is unlikely that the Tumshuqese system is derived from the Tocharian.
Sogdian. The single known Sogdian Brāhmī fragment (Mainz 639) was discovered by D. Maue (Giessen), who will publish it in association with N. Sims-Williams (London). It has the new graphs z, ź, g,w, ḏ, as well as -ṯ, -ḵ, -ṟ as specialized final signs. The writing system is related to the Tumshuqese, Tocharian, and Turkic forms of Brāhmī, but the nature of these relationships is unclear.
R. E. Emmerick, A Guide to the Literature of Khotan, Studia Philologica Buddhica, Occasional Paper Series 3, Tokyo, 1979.
D. Hitch, “A Brahmi-Manichaean Hybrid Script,” Orientalia Lovaniensia Periodica 14, 1983, pp. 293-312.
L. Sander, Paläographisches zu den Sanskrithandschriften der Berliner Turfansammlung, Verzeichnis der orientalischen Handschriften in Deutschland, Supplementband 8, Wiesbaden, 1968.
For samples of the script see for Old Khotanese and South Turkestan Brāhmī: R. E. Emmerick, 1970, The Khotanese Śūraṅgamasamādhisūtra, London, 1970 (20 plates).
Late Khotanese and literary Cursive: J. P. Asmussen, The Khotanese Bhadracaryādeśanā, Hist. Filos. Medd. Dan. Vid. Selsk. 39, no. 2, Copenhagen, 1961 (15 plates).
Tumshuqese and business and literary North Turkestan Brāhmī: S. Konow, “Ein neuer Saka-Dialekt,” SPAW, phil.-hist. Kl., pp. 772-823 (pls. 1-7 business script, pl. 8 literary script).
(Douglas A. Hitch)
Originally Published: December 15, 1989
Last Updated: December 15, 1989
This article is available in print.
Vol. IV, Fasc. 4, pp. 432-433