On the history of the Rejang vowels and diphthongs

Abstract

The development of the Proto-Austronesian (PAN) vowels in is of interest for two reasons. First, it exemplifies a pattern of phonemic splits, shifts and mergers of unusual complexity for an Austronesian (AN) language. Second, it shows that this pattern and exceptions to it cannot be stated in full without reference to semantic conditions. In connection with the first point Dyen (1949) described the development of the PAN vowels in Trukese. His central concern was to state the conditions under which a reconstructed four vowel system could be mapped systematically onto an attested nine vowel system. Superficially the history of the vowels appears simpler, as has only six vowel contrasts. However, the PAN vowels have in fact undergone more splits in than in Trukese. This greater diachronic complexity is partially reflected in a wealth of diphthongs possibly unequalled in any other AN language. In connection with the second point the often-heard assertion that meaning-based exceptions to sound change do not exist appears to conflict with the exceptional character of the pronouns and perhaps the kinship terms in relation to otherwise well established phonological changes. Background (in the Dutch and some English literature: Redjang) is the language of some 200,000 persons concentrated in the fertile highland valleys and plateaus of the Barisan Range, and extending for about * I am indebted to Richard McGinn, D. J. Prentice, H. Steinhauer and P. Voorhoeve for valuable criticisms of an earlier version of this paper. They are, however, in no way responsible for the accuracy of my data or the soundness of my conclusions. ROBERT BLUST, who majored in English at the University of California and took his B.A. in anthropology and Ph.D. in linguistics at the University of Hawaii, is at present Professor of Linguistics at the University of Hawaii. His main field of interest is compara tive Austronesian Linguistics, and major publications include 'Proto-Austronesian ad denda', Oceanic Linguistics 9 (1970): 104-62, The Proto-Oceanic palatals, Memoir 43, 1978, The Polynesian Society, Auckland, New Zealand, 'Early Austronesian social orga nization: the evidence of language', Current Anthropology21 (1980):205-47 and 'Austro nesian etymologies', Oceanic Linguistics 19 (1980): 1-181. Prof. Blust's current address is: Dept. of Linguistics, University of Hawaii, Honolulu, Hawaii 96822, U.S.A. This content downloaded from 157.55.39.100 on Wed, 06 Jul 2016 05:25:39 UTC All use subject to http://about.jstor.org/terms On the History of the Vowels and Diphthongs 423 100 km. along the Indian Ocean coast to the north of Bengkulu in southwest Sumatra. According to Voorhoeve (1955) it is closely related to Malay, but given the longstanding political and economic importance of the latter language in south Sumatra this impression is doubtless due in part to a prolonged history of unilateral lexical borrowing.1 Although the recent publication of a theoretically-oriented syntax by McGinn (1982) and the planned publication of a dictionary by Jaspan (ms.) promise to alter the existing situation substantially, remains at present among the least known languages of Sumatra. A few older works of limited scope and reliability provide some textual material and vocabulary. These are reviewed succinctly by Voorhoeve (1955). More recently Coady and McGinn (1982) and McGinn (1982) have devoted some attention to problems in the synchronie phonology, most notably to the phonetic realization and phonological interpretation of the barred nasals. In November, 1980, while at Tugu, west Java, I was able to collect a vocabulary of somewhat over 300 words in phonetic transcription, to gether with a limited number of inflectional paradigms from two speakers of Rejang. The first of these (Atika S. M.) was a native speaker of Bengkulu Malay who had lived among speakers and acquired fluency in the Lebong dialect. The second (Endarwati J.) was a native speaker of what she described as Rejang Rejang.2 The following analysis, which draws on this material, is intended to address a problem in the historical phonology of an area that has until now been all but totally neglected. 1. Synchronie phonology Both dialects of that I recorded have the following inventory of consonants and vowels. Various details not relevant to the aims of the present paper are omitted: consonants vowels t c k ? i u b d j g e o o m ? r) a m ? ?y , s 1 (r) w y In my material c is rare and r is restricted to suspected loanwords.3 Voiced obstruents do not occur word-finally, the glottal stop does not occur initially, and k is of very low frequency in final position. In accordance with inherited phonotactic constraints palatals may not occur word-finally, sequences of like vowels or of shwa followed by another vowel (as distinct from the diphthong /sa/) do not occur, and the This content downloaded from 157.55.39.100 on Wed, 06 Jul 2016 05:25:39 UTC All use subject to http://about.jstor.org/terms