This work is a descriptive grammar of the Niuwozi variety of Prinmi, spoken by the Pumi people of northern Yunnan, China. It is an impressive work which draws upon both a wide-ranging, culturally rich corpus of data, and on the author's long-term experience with the language. It consists of 12 chapters, two appended texts and a glossary.

The book has already been the subject of reviews by both Daudey (Cahiers de Linguistique, 2015), who gives an overview and a summary of each chapter, and Jacques (Linguistics of the Tibeto-Burman Area, 2016), who considers the book from a typological perspective. The present review considers two aspects of the grammar that are deserving of much applaud, as well as a few minor areas which may have benefitted from more attention.

One of the book's major strengths is its meticulous descriptions of variation across multiple social dimensions. Social dimensions that are frequently discussed include region, age, village and clan. Intra-speaker variation as well as rapid speech phenomena are also discussed. The role of gender in variation, however, is only mentioned in passing. The segmental and suprasegmental phonology of the language offer many opportunities for the author to comment on variation. A sample of the phonetic and phonological variation captured includes: sibilant frication of voiceless rhotics (p. 20); complex-onset reduction and transfer of rhotic qualities from the onset to the nucleus (pp. 21–2); vowel centralization and lowering (pp. 33–40); the replacement /u/ with a bilabial trill following dental plosives (p. 44), debuccalization of velar initial plosives as well as voiceless nasals (p. 47), and also the realization of tonal contrast (pp. 50–2).

The emphasis on variation is not limited to the chapters on phonetics and phonology, but also extends to the description of the language's typologically interesting morphological features. For instance, several Prinmi varieties have undergone a morphological shift from an older polypersonal verbal agreement paradigm, in which verbs inflect for the person and number, to a more prototypically Tibetic system, which distinguishes egophoricity and volition. The data presented shows that the Niuwozi as a hybrid system with gradient, lexically specific levelling of the old paradigm.

Another example can be seen in verbal proclitics. Like many languages of Sichuan, Prinmi has a complex set of directional orientation markers which have grammaticalized into preverbal markers of aspect and mood. There is often a conventionalized collocation between the stem and the directional it takes, and speakers from the same region but different clans may use a different directional prefix as a perfective for certain verbs. To conclude this point, Ding's inclusion of variation at multiple levels allows for a picture of a dynamic system in a state of fluctuation as opposed to a static one.

The second strong quality of the work is the author's discourse-functional and typologically informed approach to grammatical analysis. An emphasis on usage-based descriptions is reflected in the naturally occurring discourse data in the corpus, which includes proverbs, procedural texts, traditional stories and conversations, as well as elicited data.

The types of data collected and a generally usage-based approach to explanation inform Ding's discussion of key areas of the grammar such as lexical categories, which show a wide range of multifunctionality (pp. 72–87); a system of pragmatically based agentive marking with no grammaticalized syntactic pivot (pp. 146–55); the relational, deictic nature of directional orientation markers (pp. 109–14); the detailed discussion of topic and comment in the broader context of information structure (pp. 297–317), and in details about how speakers can mark ad hoc control for non-volitional verbs for pragmatic purposes (p. 240). The author's functional approach to these issues allows balance, and in-depth description of the grammatical constructions as they are used by speakers.

This work does have a few minor shortcomings. First, the examples lack information on which subset of the corpus they represent. This sometimes renders the reader uncertain as to which examples come from naturally occurring discourse, and which are from elicited translations. Second, loan words are not marked as such in the glossaries. Aside from the forms mentioned in earlier reviews, I find the following Tibetan loans: ɹə^Lwɑ^H < re.ba ‘hope’; mĩ^Lto^H < men.tok ‘lotus’; tʃhi^F < chos ‘religion’; pu^F < spos ‘incense’; n̥jɛ̃^H < nyung ‘be few’; mo^H < mɑr ‘butter’; mo^H dʒe^L < mɑr.ja ‘butter tea’; ʒ̩ĩ^H < zhim ‘delicious’; ɹ̥ɑ̃^H < ring ‘be long’; m̥jɛ̃^H < sman ‘medicine’; tɕhjɑ^H < ? phyoks ‘direction’. Suspected Chinese loan words include: ɥa^R < wă ‘tile’; kwɑ^H < guā ‘melon’; lɜ^H < láng ‘wolf’; and ʒjɛ̃^H < yān ‘tobacco’. The words p^hɜ^Htɛj^L ‘umbrella’, (p. 165), do^L ‘damn’ (p. 140), ri^R ‘get’ (possibly < Tibetan rag) (p. 328) are not found in the glossaries. Lastly, given the highly productive nature of compounding in the language, it is unfortunate that the glossary does not provide the internal composition of lexemes.

These oversights do little to detract from an overall excellent grammar which will prove a useful resource for future historical-comparative work, and also as a model for linguists working on descriptive grammars.