Tangut is not only one of the earliest attested Sino-Tibetan languages but also has a rich native tradition of lexicography and phonological analysis. Yet a century of tremendous progress in Tangut linguistics has not made much impact in Sino-Tibetan studies. Jacques’ book is a much-needed solid bridge between the insular world of Tangutology and the rest of the Sino-Tibetan realm.
Jacques opens by placing Tangut in a sub-group with Pumi in a Macro-rGyalrongic group which in turn forms a Burmo-Qiangic sub-group with Lolo-Burmese (p. 2). Readers hoping to see an immediate justification for this bold classification must be patient, as Jacques does not present evidence for his tree until much later. The rest of the first chapter covers sources of Tangut phonological data (e.g. fanqie), Gong Hwang-cherng's reconstruction of Tangut, and the phonological system of Japhug, a rGyalrong language. I wish I had read such an introduction when I first started working on Tangut nearly twenty years ago.
Readers expecting to see more on Pumi may wonder if Jacques chose Japhug simply because he has studied it in depth. That is not the case; Japhug is much more phonetically conservative than Tangut or Pumi and preserves many features that have either been altered beyond recognition or lost in those languages. Hence Japhug and another conservative (albeit less closely related) language, Tibetan, are essential for Jacques’ efforts to reconstruct pre-Tangut, an earlier, less eroded stage of Tangut.
In the second chapter, Jacques presents sound correspondences between Tangut, Japhug and Tibetan, with occasional references to other languages such as Pumi. He proposes sound laws to explain how Tangut came to look so unlike its sisters. Some of his laws are similar to ones in my article in Tanguty v central'noj azii (Moscow, 2012; not in his bibliography) and on my blog (amritas.com): e.g. consonant lenition. Others are not: e.g. his account of the raising of low vowels. In general, I prefer his solutions to my own since they are supported by comparative data that I did not have.
However, his use of Gong's reconstruction raises two major issues that are unresolved in his book.
First, Gong reconstructed Tangut Grade III syllables with a medial -j- that almost always corresponds to zero in Tibetan transcriptions of Tangut and in other Sino-Tibetan languages. Jacques projects this dubious -j- back into his pre-Tangut and does not posit a source for it. I think that Grade III was characterized by high vowels like Grade III of Pulleyblank's reconstruction of Early Middle Chinese in Middle Chinese (Vancouver, 1984), and I discuss its origins in my 2012 article.
Second, Gong reconstructed long vowels that do not correspond to long vowels in Tangut transcriptions of Sanskrit or to anything in other Sino-Tibetan languages. This questionable length also appears in Jacques’ pre-Tangut. I am agnostic about the precise phonetic nature of this feature; figuring out what it was may help in the search for its ultimate source.
Jacques’ Tangut forms differ from Gong's in one respect. Although native Tangut dictionaries distinguished between the initials that Gong reconstructed as w- and the glottal stop-glide sequence .w-, Jacques writes both initials as .w-. There are no w-initial words in Jacques’ Tangut index. I do not know whether this is intentional or accidental. In either case, reconstructing .w- instead of w- complicates lenition slightly; it is simpler to have labials weaken to w- without the addition of a glottal stop. Jacques’ formulation of lenition does not mention the glottal stop: “*C-p- > w-” (p. 35).
A final phonological issue is the origin of the two Tangut tones. On p. 260, Jacques suggests that Tangut tones may be traces of prefixes, but he does not explore this hypothesis further. His pre-Tangut lacks tones or sources of tones. Conversely, in my 2012 article, I proposed that the Tangut second tone may be a trace of lost final glottals like the “rising” and “departing” tones of Middle Chinese, but this hypothesis predicts incorrect tones for Tibetan and Chinese borrowings. As Jacques noted, tonogenesis in Macro-rGyalrongic as a whole is unclear; progress in understanding the development of tones in other branches of Macro-rGyalrongic (and perhaps Pumi in particular) may shed light on the tones of Tangut.
Jacques’ third chapter demonstrates that Tangut is relatively conservative in terms of morphology despite the massive phonological changes described in the previous chapter. Jacques provides many textual examples for various morphemes. The superficially Chinese-like Tangut script may lead one to believe that Tangut is an “isolating” language, but Jacques analyses the Tangut verb as having up to ten slots (p. 266). The first is for directional prefixes which are crucial to his sub-grouping of Tangut with Pumi. Three out of the six directional prefixes of Pumi have Tangut cognates. Only one of those three has rGyalrong cognates (p. 289), so the other two may be innovations common to Pumi and Tangut.
The final chapter, on classification, lists three types of evidence supporting Jacques’ tree diagram from the first chapter: phonological, lexical and morphological. The lexical evidence is the strongest of the three. On pp. 302–3, Jacques paints pictures of the worlds of Proto-Burmo-Qiangic and Proto-Macro-rGyalrongic speakers using etyma that might be shared innovations. The phonological evidence is the weakest, as it is not clear whether the velarized vowels of rGyalrong and pre-Tangut are innovations or retentions, and the schwa of pre-Tangut is certainly a retention.
Despite my criticisms, I regard this book as a landmark in both Tangut and Sino-Tibetan linguistics. It is the foundation for all future work on Tangut language history. Every page is a mine of ideas to be tested and explored.
