Anyonc interested in demotic Tibetan will encounter serious problems in this new monograph, not the least of which will be the identity of the language that the author is describing. Denwood gets off to a bad start by flatly equating "'Central Tibetan' and/or 'Lhasa Tibetan'" (p. 34); he attempts to protect himself with generalized hedging ("the town could hardly have a uniform dialect"; "informants...may not always hold identical views on what constitutes Lhasa Tibetan or Central Tibetan" (p. 35); but he never mentions the specific isoglosses separating Central Tibetan from Lhasa that have been clear in the literature since the mid-1950s. As early as 1985 A. Rona-Tas lamented, "[l]eider vermischt man noch heute die zwei verschiedenen Begriffe bzw. Namen [i.e., Central Tibetan and Lhasa]" (WSTB 13 : 160); plus ca change.... Denwood labels all his transcriptions of the modern language "Lhasa Tibetan" without question or quibble. Unfortunately, this is hardly the case.
What we actually find in D's hundreds of transcribed examples is rather the well-known koine of the post-1956 Tibetan diaspora, a language that in most of its phonological particularities is essentially Central Tibetan, sporadically modified from time to time with smatterings of Lhasa-like forms depending upon the taste, background, and social circumstances of the individual speaker. And in a number of points D's language appears to agree with no Lhasa (or even Lhasa-like) locutions ever recorded elsewhere by anyone else.
For the genitive and instrumental forms of the 1st per. sg. the received version of Lhasa familiar from the current Western literature has n[epsilon][epsilon] and n[epsilon][epsilon], resp. (typically in M. C. Goldstein and N. Narkyid, English-Tibetan Dictionary of Modern Tibetan [Berkeley, 1984], 286, 208; A. Agha, Structural Form and Utterance Context in Lhasa Tibetan [New York, 1993], 99). D has [_n[epsilon]:] throughout for both forms, overlooking (or denying?) the morphological difference between them, but at the same time apparently hearing a fronting of the initial [n] of [na] 'I' to [n] that hardly anyone else has ever noticed (or when they have, they have treated it not, as does D, as a significant distinction but only as an allophone: thus, A. M. Han, An Investigation of the Tones of Lhasa Tibetan [Huntington Beach, Cal., 1980]. 35, 38). In too many other cases neither the segmentals nor the suprasegmentals for routine lexical items as heard by D agree in detail with those for the same words as rep orted by others. For 'a native of Lhasa' D has [""[epsilon] sa:] (p. 91); for the same word Hu Tan, "Researches sur les tons du tibetain (dialecte de Lhasa)," Cahiers de linguistique, Asie orientale 11.1 (1982):28, has ["" sa? ]. For '8' D has [_gi[epsilon]:] (p. 80); Hu has [[epsilon]?] (p. 35). For 'language, voice' D has variously [kj[epsilon]:] and [kje:] (p. 79), but Goldstein and Narkyid have/q[epsilon][epsilon]/ (p. 241a), and Hari has [k[epsilon]?] (p. 191a).
Particularly in view of this great uncertainty about the overt shapes of many of the very ordinary words in his version of Lhasa Tibetan, it is all the more difficult to understand why D retranscribes, and thus alters into his own idiosyncratic version of the language, citations he draws from the literature where the original he is thus tampering with is surely closer to the real thing than are his data (e.g., p. 236, where he silently rewrites as [trAci: _drare _du:] an example drawn from Agha [p. 124], where it appeared as [tAsii tore tuu]).
Perhaps the most striking phonological isogloss separating the Central Tibetan koine from genuine Lhasa Tibetan is the correspondence of retroflex stops [t, t', d] in the former with retroflex affricates in the latter. Reported for Lhasa in Chinese transcriptions from early on (Laufer, TP 17 : 409, 436), then in modern times canonically described in Y. R. Chao's well-known monograph Love Songs of the Sixth Dalai Lama, Academia Sinica Monographs, series A, no. 5 (Peking, 1950), and later verified by numerous in situ investigators, including Chin P'eng, Schubert and Richter, Hu Tan, and Huang Pufan, these characteristic affricate phonemes [ts, ts'] were well documented for speakers of a specific sociolinguistic level, notably lay officials, in the city of Lhasa itself. But they were absent from the speech of most of the diaspora-speakers who became the first-generation of informants for Western linguists. These speakers mostly replaced the affricates with their own less elegant Central Tibetan retroflex stop equivalents: that much is clear.
Rather less clear is why D treats these sounds on the gross phonetic level (the only level of description with which he is concerned) as clusters of stops plus [r] or [r] (p. 70), and especially why he at the same time insists that his sequences [tr, tr, dr] that result are (sometimes) not really clusters at all but only unitary "simple consonants" (p. 75).
To his credit, D does not belabor this last delicate point, instead simply citing as "a justification" for his analysis a long article by R. K. Sprigg, BSOAS 35 (1972): 546-87. The argument found there may easily be epitomized: these sequences of consonants are not initial clusters because the Lhasa language does not have initial clusters; conversely, the Lhasa language does not have initial clusters because these sequences are not initial clusters. Not much can be made of this, unless it is intended to illustrate D's rather bewildering allusion elsewhere in the book to "the intellectually demanding terrain of Buddhist philosophy, in whose farthest reaches language itself is tested to its destruction" (p. [xi]); perhaps we are intended to conceptualize these clusters that at one and the same time both are and are not clusters along the lines of the celebrated Mahayana conundrum that explained why "a white horse is not a horse"? At any rate, if not "language itself" then at least conventional linguistic termi nological praxis is indeed "tested to destruction" by thus playing fast and loose with entirely obvious phonological-structural categories.
Much of this is additionally complicated by D's topsy-turvy treatment of such loanwords from modern subcontinental vernaculars as Central Tibetan [l0nd0n] 'London' and /mota/ 'motor(car), automobile'. These D has heard and writes as Lhasa [l0ndr0n] (p. 163) and [motra] (p. 184), forms that he then equates with putative written lon.dron and mo.tra. But these written Tibetan spellings with -dr- and -tr- are demonstrably quite as incorrect as the alleged pronunciation of the words going with them is unlikely to involve dr- and tr- "non-cluster" clusters. Tashi Tshering, English-Tibetan-Chinese Dictionary (Peking: Mi.rigs dpe.skpan.khang, 1991), 567b, 625a, has, regularly, the spellings lon.don and mo.tar, employing the proper retroflex consonant graphs of the Tibetan script for I and t. All this of course merely reflects the normal Anglo-Indian approximation of the English dental-alveolar stops by means of the subcontinent's retroflex stops.
The traditional Tibetan script and its orthography are discussed in chapter 5, "Graphology" (pp. 55-68), employing the term here and throughout the book not in its usual sense of "the study of handwriting for determining character or diagnosing nervous diseases," but rather a la M. A. K. Halliday to indicate "a level of language [sic], whatever the nature of the script, [that is] characterized by its own distinctive pattern, just as is the level of phonology" (OED, s.v., 6:766a ).
This definition's basic, and fatal, conflation of writing and language does more than short-circuit much of chapter 5; it continues to surface throughout the book, which, as a consequence, too often becomes less a linguistic study and more a handbook on how to read aloud a text written in Tibetan. Given the nature of the script and its history vis-b-vis the language, this is rather like providing instructions about how to read a Late Latin text in modern Tuscan--something not entirely impossible but probably too replete with exceptions to be of general utility. And the reader will search in vain for the promised "transliteration of Tibetan [script] in underlined form" (p. 55): all written Tibetan transcriptions in this chapter and for that matter throughout the entire book are not underlined but printed in italics.
Be that as it may, the account of "graphology" stands head-and-shoulders above the following chapter 6, "Phonology" (pp. 69-85), which despite its title deals neither with phonology nor phonemics. Its goal, simple but probably impossible, is to nail down the sounds of Lhasa Tibetan in IPA notation. When unexpected, rare, or otherwise strange sounds turn up, as they persist in doing, they are symbolized as well as the author can manage to manipulate the IPA; but they are never structurally related to any larger system within the language.
An unexplained [[FOREIGN CHARACTER NOT REPRODUCIBLE IN ASCII]] pops up once in a single word (p. 92); uvular voiced [[gamma]] (p. 74) and an undescribed [x] (p. 76) put in brief appearances but are then forgotten; and an (apparently) voiceless syllable-final [r] is reported "in a few [actually, only two] loanwords," but loans from what or loans from where is not specified (p. 71). The first of the two, [plr] 'paintbrush', is, of course, a well-known early borrowing from Chinese, the word having come along with the thing (more precisely, 'writing brush'); but the second, [par] 'print', though sometimes alleged on next to no evidence to be another borrowing from Chinese, is without question etymologically Tibetan. This was convincingly demonstrated, and at considerable length, by W. Simon, BSOAS 25 (1962): 72-80. D mentions Simon as one of his teachers at SOAS "without the previous efforts of [whom] [t]he writing of this book would have been impossible." Unfortunately, the student's proper high regard for this fine scholar seems not to have led him to s tudy his publications.
Most briefly put, what we find in this chapter are neither phonemes, allophones, morphophonemes, nor archiphonemes: they are noises, and as a consequence the chapter does not really deal with phonology at all. A valiant attempt to arrange at least the principal articulatory elements that D has managed to detect from among all these noises on a two-way grid (table 6.1, pp. 69-70) fails to inform: column I has been left entirely blank, and it will be the task of the perceptive reader to renumber the rest, and/or to rewrite all the corresponding acoustic descriptions, if anything at all is to be salvaged here-not a good omen, especially so early on in the book.
Closer scrutiny of D's IPA renderings further inclines one to the view that, all in all, we are dealing here with frequently shifting perceptions of often trivial surface acoustic phenomena, and hence not really with elements of a linguistic system. In one and the same morpheme a semivowel [j] is sometimes reported before [e, [euro]], sometimes not (examples of both on p. 79 for the same word); in other morphemes in identical environments it seems always to be present (pp. 80, 81 in '8'); in still others it seems to be present but only by implication, apparently having exhausted itself in the course of fronting [[FOREIGN CHARACTER NOT REPRODUCIBLE IN ASCII]] to [[FOREIGN CHARACTER NOT REPRODUCIBLE IN ASCII]] (pp. 101, 135 and passim) in the oblique case-forms of [[FOREIGN CHARACTER NOT REPRODUCIBLE IN ASCII]a] 'I'.
Somewhere behind all this there probably lurks some lingering trace o f the classic distrust (and equally classic misunderstanding) of the phoneme as somehow representing a potentially subversive American artifact, a view that first surfaced in SOAS linguistic literature a generation ago (e.g., BSOAS 41 : 185). But whatever its cause, D's Consistent rejection of the phoneme in favor of random noises pressionistically recorded hardly represents an advance in the description of Tibetan. And given his descriptive confusion among the different levels of segmental phonology, not to mention his conflation of language and orthography, it comes as no surprise to learn that D's treatment of the language's admittedly involved and much-mooted suprasegmentals is even less lucid. No other features of modern Tibetan pronunciation have led to more disagreement in the literature over the past half-century than have its suprasegmentals; unfortunately nothing in this book contributes to resolving the problems with whic h these continue to confront the descriptivist.
D's transcriptions are said (p. 12) to mark "tonal classification" rather than "actual pitch" an apparent prima facie contradiction of his own IPA principles. But how either or both of these differ from his "tonemes," which suddenly put in an appearance on p. 71, is not explained, nor is the even more pressing question of what tonemes are doing in D's adamantine anti-phonemic system ever answered. (In view of all this inherent confusion about the suprasegmentals, they have been deleted from citations in the present review unless immediately relevant to the discussion.)
A surprisingly long portion of the remainder of the book (pp. 115-242) is concerned with "polarity, judgemental modality and part [sic] of the mood system," as these are said to operate "on the level of the verb phrase" (p. 115). Numerous interlocking schemes are described which, it is claimed, reveal sets of one-to-one connections between Tibetan verb-structures and its speakers' thought- and perception processes (e.g., "self-centered" vs. "other-centered," "witnessed" vs. "unwitnessed" vs. "unwitnessed" vs. "not necessarily witnessed," etc., etc.); each set is illustrated by short, fragmentary citations. These fine perceptual demarcations would be more convincing if they were supported by citations from longer, connected texts, where one could verify the situational content of the forms involved. As they stand here, supported only by isolated fragments taken totally out of context, they appear mostly to reflect the introspective musing of hard-pressed bilingual informants, and probably tell us more about t heir command of "unidiomatic" English (p. 156) than about what happens deep within the Tibetan psyche.
And as it happens, significant portions of this ostensibly recondite analysis are directly contradicted by the few short texts that do conclude the volume (pp. 275-77). Despite their limited scope these provide enough situational context to make possible verification of D's categories to a large extent. In these texts a speaker, describing life in Lhasa for the benefit of another Tibetan "who had never lived in Tibet," employs at least four verbals (translated "they live," "it's very cold," "it lies (without melting)," and "they keep things to eat") that embody structures analyzed elsewhere in the book as signifying "3rd person unwitnessed" and "no first-hand knowledge" (pp. 121, 122, 158-59). But the whole point of this text is precisely the opposite: the speaker is relating information for which he has not only firsthand knowledge but indeed a life-time of direct experience. Something is obviously very wrong here.
Since the Tibetan of D's description has neither phonemes nor phonological structure, it follows that it also has neither allomorphs nor morphophonemics; nor may its many forms that from time to time appear in different but obviously related shapes rigorously be associated with one another. Mostly no basic forms are postulated; but when they are, it is always obscure why a particular shape has been selected as the point of descriptive departure. Why, for example, is go selected as the allomorph under which to list the terminal or statement-closing particle (p. 358)? The form appears in one of ten possible different shapes, of which go is merely a single unrepresentative example. Such features require the postulation of basic forms and the formulation of rules predicting the surface realizations; there are none here.
Similarly, morphologically determined allomorphs of a small list of nouns are an important feature of the language (/chu/ 'water' in isolation and certain compounds, but /chub-/ in certain others, e.g., /chub-ja/'water-bird' with/-ja/ [less than]/ca/'bird';similarly,/cu/'10', but /cug-/ in '11' and/cur-/ in '19', rewriting forms from D's p. 76 in a probable phonemic version of his IPA renderings). In this book the morphological operations involved in such forms as these are neither described nor explained; their final /-b-, -g-, -r-/ archiphonemes of neutralization are recognized only in their trivial surface realizations and then assigned to the initials of the morphemes that follow in the composita in question. This in turn results in apparent (but surely not real) consonant-cluster initials for certain morphemes in what is otherwise, by D's fiat, a "non-cluster" variety of Tibetan.
In this particularly egregious bit of descriptive leger-demain it is of course D's knowledge of the Tibetan script that is operative ('water-bird' is written chu.bya; '11' is bcu.gcig: '19' is bcu.dgu), as in a hundred other places in the book. Even a perfectly routine articulatory sequence like [an] is split between two morphemes, with -a final in the former and n. Initial in the "cluster-initial" of the latter, for no other reason than because a word such as D's [tanda] 'now' is written da.lta (p. 76). Small wonder then that we read here of something called "phonetic vowels" (p. 79), leaving open the question of what other kind of vowels a language might possess; or that to be "invariable in its Tibetan spelling" (p. 87) is one of D's four formal criteria for defining his form-class "noun" (the fourth in his list, "[it] may occur in noun words only," is scarcely more informative).
D identifies three colleagues who helped him compile the "Bibliography" (pp. 306-56) that concludes the book; he ought to have checked their work before printing it. Spectacularly absent bath from these pages and from the text as well is Y. R. Chao's essential 1930 monograph, already cited supra; perhaps this silence reflects the fact that the Chao analysis carefully documents phonetic and phonological details that, in turn, refute D's allegation that the language he is writing about is Lhasa Tibetan.
But there are many other less easily explained omissions as well, notably of works that D does cite in his text. Of the eight sources he cites by author and year of publication in note 3, p. 19, five (Clauson and Yoshitake 1929; Uray 1955a; Hoernle 1916; Laufer 1918; Thomas 1951) are missing from the bibliography; missing also is Nishida 1970, frequently cited in the text (e.g., p. 23) but never further identified. Japanese scholarship on Tibetan is treated especially shabbily. Inaba 1955 (p. 323) and Kitamura 1955b (p. 325) are both simply bogus; no such works have ever been published. Similarly bogus but in a different fashion is Sato 1963 (P. 340), a study of the history of a Ming-period ecclesiastical lineage not remotely concerned with the Tibetan language; and far too many of the romanizations of Japanese book- and article titles are garbled to admit of correction here, even in the cases where the works cited do actually exist.
European-language citations fare almost as badly: for slokas and Sambhota (p. 308) read flokas and Sambhota; for de Koros (pp. 313, 331) read de Koros; for a l'histoire (p. 308) read a l'histoire; for misionnaries (p. 315) read missionaries; for Soren (p. 316) read Soren; for Andras (pp. 338-39) read Andras; for Tibetskago (p. 340) read Tibetskogo; for Graf (p. 349) read Grof; etc., etc. A curious scrambling of the dialect-names in the title (p. 324) of Chin P'eng 1958 ("Lhasa, Chamdo, Shigatse," where the monograph has "Lhasa, Shigatse, Chamdo") makes one doubt that the study in question was actually placed under contribution. So also for the abruptly truncated listing (p. 345) of Simonsson 1957: D might have avoided his embarrassingly naive morpheme-by-morpheme glossing of sangs.rgyas 'Buddha' as 'purify-vast' (p. 90) by consulting pp. 265-66 of Simonsson, who there demonstrated how, ever since its ninth-century coinage, this compound has been understood as meaning 'der Erwachte-Aufgebltlhte'. And even wh en earlier literature is correctly identified, the way it is cited in the text generally makes it clear that it is mostly only being cited but has not been read, since the citations are often alleged to support just the opposite of what the earlier author says (e.g., pp. 35, 44 with note 16).
The SOAS series editors boast that this volume is "reliable and up-to-date"; the publisher claims that it has "a comprehensive bibliography." Neither allegation is true. Had General Motors put out a product like this, the company would now be busy issuing a general recall for repairs. SOAS and Benjamins ought to do the same.
|Printer friendly Cite/link Email Feedback|
|Author:||MILLER, ROY ANDREW|
|Publication:||The Journal of the American Oriental Society|
|Article Type:||Book Review|
|Date:||Jan 1, 2001|
|Previous Article:||Elaborations on Emptiness: Uses of the Heart Sutra.|
|Next Article:||Mishnah and Tosefta: A Synoptic Comparison of the Tractates Berakhaot and Shebiit.|