The fruit of Roland Bielmeier's pioneering CDTD project is finally here, having defied all the odds against the publication, now regrettably posthumous, of his formidable work on the linguistic history of Tibetan. Certainly many will look on this volume (and another forthcoming) as a great tribute to Bielmeier's scholarship, the depth of which, as related by George van Driem in his Foreword, goes far beyond what a dictionary could contain. Nonetheless, those who will carry forth his legacy now have an etymologized cognate list, compiled from numerous major dialects documented in the last few decades, covering all subgroups, with by-dialect and by-gloss indices. This makes this volume the first true comparative dictionary of Tibetic languages, and deserves recognition.
The current volume – Volume 2, although it is published first – comprises verbs, which for historical linguists constitute the more interesting and possibly revealing word class. Morphological regularities of Written Tibetan verbs have been a wellspring of diachronic insights for more than a century, and with a comparative dialects dictionary one instantly looks forward to mining deeper for patterns of morpho-phonological change that will hopefully ground our script-mediated knowledge of Tibetic in vernacular data. Agreeably, the organizational features of this volume appear to serve that end well, though a consideration of some of them brings to light some imperfections.
Dialectal entries are etymologized under Written Tibetan forms. These are collated from Jäschke's dictionary (both German and English editions) and the Bod-rGya Tshig-mdzod Chen-mo (Zàng-Hàn Dàcídiǎn) – two classic lexicographic sources – although Goldstein's Modern Tibetan dictionary is also included (I think Paul Hackett's verb lexicon would have been more suitable). In determining paradigmatic forms, the authors choose to follow BTC unless it misses forms that the other sources record, though all variant forms are listed.
These decisions seem adequate for establishing a thorough Written lexical base with reputable spellings and definitions, not bothering too much with the whims of Tibetan lexicography (for which consult Hill's Lexicon of Tibetan Verb Stems). But readers of this volume should note historically meaningful variants that may be masked by the BTC forms. In 851 blug-blugs-blug-blugs “to pour”, the BTC paradigm is analogically levelled from the past form b-lug-s, compared with the Jäschke present ldug (< ḥ-lug, Conrady's law + Coblin's Law) and past blugs. For a comparative dictionary, the better choice in cases like these would seem to be the older form.
A bigger caveat is that modern dialectal forms may not correspond neatly with entire Written paradigms. I have argued elsewhere that “past” (really perfective) stems of transitive paradigms can become new intransitive verb roots through lexical–semantic shift, and thereupon be extended paradigmatically. The existence of such diachronic processes ought to make us healthily sceptical about treating four-part paradigms as natural units to be inherited as a whole.
Consider Tabo (Western Innovative) nɖil “to fall” and ʈil “to roll (vt. = make round)”. The authors have listed them respectively under 653 ḥdril “to fall, roll down” and 654 ḥdril-dril-ḥdril-dril “to make round, wrap round, etc.”, apparently relying on matching transitivity/control. Taken prima facie, it would seem that Written ḥdr :: Tabo nɖ and Written dr :: Tabo ʈ, but there is e.g. Tabo ɖu(p) “to sew together” under 605 drub-drubs-drub-drubs “to sew, embroider”, where it must be that Written Tibetan dr :: Tabo ɖ. Two Written Tibetan onset clusters ḥdr and dr cannot explain three Tabo initials nɖ, ɖ, ʈ; the background story will need to involve a third form/feature that is not represented in the dictionary paradigms. (Incidentally Tabo is a Western Innovative variety that typically devoices erstwhile voiced unprefixed onsets, so the problem would probably be ɖ.) The point is that users of this volume should be actively aware of sub-paradigmatic complexities that paradigm-based etymologization may hide.
Morphologically related entries are listed separately but cross-referenced. Cross-references include well-substantiated connections such as transitivity alternation (initial voicing, s-prefixation, m-prefixation), Conrady's Law (e.g. 1191 ral “to fall/be torn to pieces” – 651 ḥdral “to come open” < ḥ-ral), extension (e.g. 915 √bral “to be scattered, separated” < b-√ral,), etc., as well as many suspected relations (e.g. 925 √ḥbrel “to hang together, be connected”, superficially similar in shape to √bral with incidentally antonymic semantics). Except for a handful of cross-references that point to non-existent entries, I find this to be rather consistent.
Of interest are the entries without dictionary-sourced Written Tibetan forms, which are essentially reconstructions. The majority of them cross-reference related verbal, quasi-verbal (e.g. m-prefixed statives, deverbal adjectives) or nominal forms that are well-attested, and virtually all of them rely on Western Archaic and/or conservative Amdo for segmental information. To me these are as sound as any evidence could get for reconstructing verbs unattested in Written Tibetan, though the lack of an index for reconstructed entries makes studying them in conjunction difficult. I note with optimism that some of these reconstructed forms may benefit comparative work with non-Tibetic Bodish groups, e.g. the reconstructed form 504 btil “to press, hit (pound), spread” (cf. mthil “palm, bottom, centre”) etymologically clarifies Tamangic *A~Btil/tit “id.” (with final -t either due to merger or, more likely, a reflex of Old Tibetan past stem suffix -d).
Dialectal data are subgrouped under each entry. The subgrouping schema follows Bielmeier's approach of separating “archaic/conservative” from “innovative” for Western and Amdo varieties based on degree of cluster simplification. Central, Southern and Kham subgroups remain well established, with the notable inclusion of Hor as a distinct group. Two kinds of detailed data strike me as refreshing: modern paradigms in conservative Amdo varieties, and collocations with nouns/adverbials, which provide better semantic resolution especially for the connection from concrete to abstract senses (though Written forms for collocative elements are not given). Controlledness and case frames (ergative, absolutive, dative and instrumental) for a significant number of dialectal forms are also given.
Against the backdrop of Tibeto-Burman (or Trans-Himalayan) as a whole not readily lending itself to the classical comparative method, this volume places Tibetic one refreshing stride closer to the Neogrammarian ideal. Those of us who have felt the gap between solid reconstruction and an ever-expanding database of modern varieties will certainly find more direction here.
