- The syntax of Chichewa
One of the great strengths of Sam Mchombo's The syntax of Chichewa is that it is more than just the syntax of Chichewa; it is also an excellent introduction to Bantu syntax from a theoretical [End Page 651] linguist's point of view. Although the book is organized around the presentation of the major morphological and syntactic structures of Chichewa, the syntactic analyses are theoretically informed and the consequences for the general picture of Bantu syntax are always highlighted. M's theoretical perspective is certainly shaped by his commitment to lexical functional grammar (LFG), a theory to which his work has contributed mightily, but this commitment serves to give some of the argumentation a welcome edge, especially since the book is not overly technical, though it is, at times, elliptical.
Chichewa has many features typical of Bantu languages, including SVO order, a rich set of noun classes marked by prefixes that figure in extensive patterns of agreement, an agglutinative verb with prefixes occurring in a familiar fixed order, and suffixes that appear to influence verb argument structure. There are also cleft-like focus structures that appear to be built from the same mechanisms that build relative clauses, and direct questions, when they are not in situ, are built from focus structures.
After an introductory chapter on nouns and noun classes, the second chapter develops some key phonological domains as they relate to constituent distinctions, in particular the appearance of high or low tone on subjunctive suffixes as indicators of whether or not there is anything else in the VP.
Ch. 3 explores the verb prefixes as a reflection of clause structure. These include the subject and object markers, and the tense, modality, directionality, and negation markers, all of which are treated as clitics, that is to say, outside of the verb stem. There is complementary distribution between certain affixes in the slots where they can appear. Although I found this chapter informative, the actual structural analysis of these preverbal affixes is downplayed even when it is finally discussed (in Ch. 5), and it is unclear whether a complex verb with modals and directional affixes (interpreted like light verb 'come' or 'go') is a larger syntactic structure than one that lacks these affixes, or if these positions do not correspond to any syntax external to the verb. The latter position is apparently adopted (see the diagram on p. 70), but the boundary between morpho-syntactic and morpholexical processes is taken to be the verb stem, while the structure outside the verb stem, though still in the verb, looks much like the extended functional projections in principles-and-parameters proposals.
Ch. 4 on clefts, questions, and relatives outlines the major structures, and then discussion turns to the status of subject and object markers (SMs and OMs) as revealed by their interaction with these larger structures. Questioned constituents in Chichewa can be either left in situ or else displaced in a cleft-like structure: [Q copula WH-pron Agr-méné [.. .x. . .]]. In this structure, the complementizer méné agrees in noun class with the WH-pronoun, and in the clause that follows ([. . .x. . .]), an object will be missing if the object = x or a subject will be missing if the subject = x. The SM is always obligatory in Chichewa and clefted questions are no different, but the OM is normally obligatory in the absence of the object of a transitive verb, yet the transitive verb object can be missing and the presence of the OM is optional. The presence of the OM is treated as a resumptive structure, but it is not obvious what is assumed about antecedent-gap constructions when there is no resumptive pronoun, and all that is said is that 'the FOC element is linked to the TOP marker -méné which is functionally identified by the missing argument in the relative clause'. Here M appears to be tacitly relying on Bresnan & Mchombo 1987, and it is a lacuna for those not familiar with that paper. The basic conclusion...