Pidgins and creoles: vol. 1, Theory and structure; vol. 2, Reference survey by John Holm (review)

380LANGUAGE, VOLUME 67, NUMBER 2 (1991) Pidgins and creóles: vol. 1, Theory and structure; vol. 2, Reference survey, by John Holm. Cambridge: Cambridge University Press, 1988; 1989. Pp. xix, 257; xxvi, 259-704. Cloth $59.50; cloth $69.50, paper $24.95. Reviewed by Salikoko S. Mufwene, University of Georgia As critical as I am below of some aspects of these two books, they are useful resources and in several ways a welcome addition to the increasing reference literature on pidgins and creóles (PCs). Theory and Structure (T&S), which I discuss first, contains an Introduction (1-12), a Conclusion (216-18), an extensive historical survey of studies of PCs, especially regarding their genesis (Ch. 2, 13-70), and a detailed discussion of some formal features shared by several of those lexified by European languages, primarily those of the Atlantic and Indian Oceans (Chs. 3-5, 71-218). The common features provide part of the justification for discussing these PCs as a special group of languages. The other part derives from the historical socioeconomic circumstances of their genèses, viz. contacts motivated by trade and/or plantation economy following the European explorations of Africa, Asia, the New World, and the Pacific since the second half of the 15th century. Details of these are covered in the Reference Survey (henceforth RS). The similarities discussed by Holm are primarily lexicosemantic (Ch. 3, 71 — 104), phonological (Ch. 4, 105-43), and syntactic (Ch. 5, 144-215). The features apply to PCs as a group to the extent that one subscribes to the Wittgensteinian family resemblance model, as H admits (12). For instance, the distribution of the copula is not the same in Jamaican Creole, Gullah, and Papiamentu (182). PCs lexified by European languages vary according to whether they use their durative marker also for habits or use a null marker for the latter (158). Moreover , there are creóles like Gullah that have both patterns. As may also be expected from such a broad coverage of languages and subjects , only a few languages are selected to exemplify each formal feature; the selection varies from one feature to the next, although some creóles, such as Guyanese, Haitian, Hawaiian, Jamaican, and Saramaccan, are regularly cited. It is regrettable that the survey of features has been restricted to Atlantic and, to some extent, Indian Ocean creóles. Given the title of T&S, a discussion here of some features of, for instance, Melanesian PCs and/or those not lexified by European languages would show explicitly and more obviously that the features discussed 'are [not] necessarily traits of all creóles (the fundamental flaw of extreme universalists)—much less traits of all pidgins' (12). H's observation is a welcome antidote to the false impression given by part of the creolistic literature that there are universal PC features or that PCs, even those putatively formed by children (see below), form a typological group comparable to, for instance, the agglutinating or ergative or SVO languages. For instance, as is evident from RS, not all PCs do without the inflections of their lexifiers. Kituba and Lingala have preserved in a productive way, though to different extents, various inflections typical of Bantu languages. These PCs also use a copula in some contexts where none is used in PCs lexified by European languages. And they use special habituative markers/constructions, different from the durative REVIEWS381 ones; and Kituba (though not Lingala) uses an uninflected verb in the narrative tense, which is not always interpreted as the unmarked verb in PCs lexified by European languages. Although space constraints imposed by the publisher may be responsible for this shortcoming, T&S would definitely look better if H had also discussed features such as these and had attempted to account for the variation. Such a discussion could help articulate more explicitly the conditions under which substrate influence (a leitmotif of T&S; see below) or for that matter any influence, prevails in the genesis of PC systems. The subjects are covered fairly evenly, in contrast to the typical overemphasis on morphosyntax in creolistics: 32 pages are devoted to phonology (Ch. 4), 64 pages to lexicosemantics (Ch. 3), and 71 pages to [morpho]syntax (Ch. 5). H must be praised...