Sourcebook on Tomini-Tolitoli languages: General information and word lists (review)

Readers of linguistics publications are familiar with various categories of language descriptions, including grammars, dictionaries, collections of conference papers, and survey volumes, but what is a "sourcebook"? Himmelmann's book comes closest to the last of the familiar categories in that it surveys a collection of some eleven languages (Balaesang, Pendau, Dampelas, Dondo, Totoli, Taje, Ampibabo-Lauje, Tajio, Lauje, Tialo, Boano) spoken by about 145,000 people on the island of Sulawesi, Indonesia. What makes it different from many survey volumes is its thoroughness and attention to detail.

The Tomini-Tolitoli (TT) languages take their name from two geographical features, the town of Tolitoli near the northern end of their territory, and the Tomini Gulf, which stretches eastward from the narrow base of Sulawesi's northernmost peninsula. As H points out, the term 'Tomini-Group" was first used in the classical work of Adriani and Kruyt (1914), and although speakers of these languages have a sense of common identity distinct from that of the Gorontalic languages to their north or the Kaili-Pamona languages to their south, it is possible that Totoli and Boano do not subgroup with the other TT languages.

H's primary concern is to provide basic descriptive materials for an area that previously was among the linguistically most neglected parts of Sulawesi. Following a 13-page introduction in Indonesian the book continues with a 44-page "General Introduction" that includes an areal map indicating language boundaries, an exhaustive review of the limited literature on these languages, a discussion of language names and subgroups, language distribution by both territory and village (with 1991 preelection census figures for more than 170 villages), the sociolinguistics of language usage, general information on geography and local topography, socioeconomic stratification, recorded history, settlement patterns, vegetation and economy, modern infrastructure, education, and cultural practices and religion. Chapter 3 is a 57-page 'Introduction to word lists" that covers sources of material, including information on the Indonesian terms used in elicitation; geographical location and reliability of the word lists; format of the entries; orthography and notes on segmental phonology; the representation of affixes in the wordlists (what H calls 'morphonology'); a note on NP markers; examples that illustrate how to read an entry; and a 28-page section with comments on semantic fields, including problems encountered in elicitation, negative responses, etc. Chapter 4, "Tomini-Tolitoli word lists" (115-333) is organized by semantic fields, using a decimal system adopted from Tryon (1995) to mark proposed degrees of inclusion (08.550 = branch, 08.555 = twig- 8.556 = bunch, stalk, 08.560 = leaf, 08.570 = blossom, bud, 08.572 = flower, etc.). Entries are set off by an English headword and its Indonesian equivalent in boldface. This is followed by three pages of references, and 99 pages of indices that serve as finder lists for the material in chapter 4. These include reverse indexes in English and Indonesian, and an alphabetical listing of all TT lexical material, together with the numerical code but no gloss (Balaesang aasi 16.350 = pity/kasihan, etc.). H makes is clear that the comparative vocabulary he presents is a wordlist and not a dictionary, drawing on the basic Saussurean distinction between signification (finding the best test-list equivalent) and meaning (providing the full set of referential connections) to distinguish the two types of lexical compilations (58).

The core of this book is the set of comparative wordlists that make up slightly over half of its volume, and these have been "normalized" in terms of dialectal and idiolectal variation. Unlike the authors of many survey volumes, however, H liberally shares the difficulties of his field experiences with the reader: the TT languages as a whole lack prestige, and some speakers are ashamed to use them (43); the social setting essentially precluded his working with women, with the result that the informant base was almost exclusively male, potentially concealing important information in such semantic domains as kinship...