In lieu of an abstract, here is a brief excerpt of the content:

Reviewed by:
  • The Austronesian languages of Asia and Madagascar
  • Robert Blust
Alexander Adelaar and Nikolaus P. Himmelmann, eds. 2005. The Austronesian languages of Asia and Madagascar. London and New York: Routledge Language Family Series. xxi + 841 pp. ISBN 0-7007-1286-0. £180.00, hardcover.

This book, part of the Routledge (formerly Curzon) Language Family Series, forms the natural complement to Lynch, Ross, and Crowley (2002), the two volumes partitioning the geographically enormous Austronesian (An) language family into roughly equal portions in terms of scope (924 pages for the first volume, 841 pages for the second). From a practical standpoint, this division is understandable and perhaps unavoidable, but in other respects it is artificial, since the An languages of Asia and Madagascar constitute nearly two-thirds of the family, and, unlike the Oceanic languages, do not form a single subgroup. Like its earlier companion, the present volume—hereafter ALAM—begins with five survey chapters, and then proceeds to sketches of individual languages.

In chapter 1, "The Austronesian languages of Asia and Madagascar: a historical perspective," Alexander Adelaar introduces the reader to (i) important contributions to An comparative historical linguistics, (ii) traditional scripts, (iii) Proto-Austronesian and Proto–Malayo-Polynesian (PMP) linguistic history, (iv) subgrouping, (v) theories about the An homeland and An migrations, and (vi) language contact. This section of the book, written by a scholar who is well informed about the major issues in An comparative linguistics, is a valuable summary of much of the relevant scholarship up to about 2002. Due to limitations of space, I will comment in detail only on the discussion of subgrouping, which at 18 pages is by far the largest section in a chapter of 34 pages (excluding references).

Adelaar recognizes the force of the arguments for an East Formosan subgroup (Basay, Trobiawan, Kavalan, Amis, Siraya) presented in Blust (1999), but continues to believe in "a possible close link between Malayo-Polynesian and East Formosan languages." The idea of such a unit was first proposed by Harvey (1982) and Reid (1982), under the name "Amis–Extra-Formosan" Although the arguments offered in support of it were never strong, this claim has been repeated periodically for nearly a quarter of a century. Adelaar presents a new twist on the old argument, suggesting (14) that "Siraya shares a number of lexical features with Malayo-Polynesian that are not found in Amis or other Formosan languages." As support for this claim, he cites reflexes of *Ratus 'hundred' (Siraya ka-xatux-aη), *laba 'spider' (rawa), *k<əm>uliC 'to peel' (k<əm>urit), *tuRut 'follow' (t-muxot),1 and *qaRus 'stream' (axu-aŋ 'river'), plus the syncope in *paŋudaN > pandal 'pandanus', and cluster reduction in *biRbiR > bibix 'lip'. However, a likely reflex of *Ratus is also found in Hoanya matala-gasut '100' (matala = 'one'; Tsuchida 1982); syncope of the medial vowel in the word for 'pandanus' is shared by Kavalan paŋzan 'Pandanus tectorius', Mayrinax Atayal paŋran, Saisiyat parŋraŋ 'pineapple' (Zeitoun n.d.);2 reflexes of *k<əm>uliC 'to peel' also appear in the Tsouic languages (Tsuchida 1976: 225–26); it is by no means certain that Siraya tmuxot and axuaŋ [End Page 302] reflect the protoforms to which they have been assigned; and, as Adelaar himself notes, the word for 'lip' is recorded as vixbix in the so-called "Utrecht Manuscript dialect" of van der Vlis (1842). Moreover, given the extreme diversity of the Formosan languages, it is relatively easy to find lexical items in individual Formosan languages that are shared exclusively with Malayo-Polynesian (MP)—this is a natural consequence of the structure of the An family tree. Examples from Thao include fakun (PAn *bakuŋ) 'Crinum asiaticum', fanuz (PAn *baŋuN) 'awake, wake up', fciq (PAn *beCiq) 'germinate, explode', fuar (PAn *bual) 'bubbling spring', ian (PAn *ian) 'dwell, reside, live in a place', lhakup (PAn *Rakup) 'gather in the cupped hands', pataqaz (PAn *pasaqaN) 'shoulder pole; carry with a shoulder pole', qa-qutilh 'chase, pursue' (PAn *qusiR), and qaruf 'knee' (PAn *qaleb). Similar collections of MP cognates that are shared exclusively with individual Formosan languages...