Patricia J. Hamel. 1994. A grammar and lexicon of Loniu, Papua New Guinea. Pacific Linguistics C-103. Canberra: The Australian National University. 275 pp.

There are so many languages in Papua New Guinea and so few descriptive grammars available that each new one is a major contribution. Hamel’s work is of particular value in being the most complete description so far of an Admiralty Islands language. It includes the basics of Loniu grammar (1–172), a Loniu–English lexicon (173–234), an English–Loniu finderlist with appendices for specialized vocabulary (235–264), and two texts with interlinear glosses and free translations (266–272).

The languages of the Admiralties are of interest for several reasons. They appear to constitute a first-order subgroup of Oceanic (Ross 1988). They do not appear to have been influenced (at least directly) by Non-Austronesian [end page 248] grammatical patterns, as languages of the New Guinea mainland have been. And they are among the closest Melanesian languages to Micronesia, lying roughly 1000 km due south of Trukic-speaking Lamotrek and Satawal in the Caroline Islands, not much farther away from those islands than Guam is. (Satawal is the home of the best-known traditional navigator in the contemporary Pacific, Mau Pialug.)

Several scholars have noted tantalizing morphosyntactic similarities between the languages of Micronesia and the Admiralties, which Ross (1988:326–329) has summarized. The pertinent phenomena are already well-described in many representative Micronesian languages. H now offers a fuller picture of the matching phenomena in a language typical of the Admiralties. Among the similarities are: (a) the direct suffixation of possessive pronouns on a larger class of nouns than the typical inalienables such as body parts, kin terms, and spatial relationships (43–50); and (b) a fairly elaborate system of numeral classifiers in which the numeral precedes the classifier (54–66). (In Loniu at least, the order is reversed for higher numbers: ma+numeral+classifier > ma+classifer+numeral+‘tens/hundreds’.) Loniu also typifies its neighbors in rendering the numbers 7–9 as subtractive compounds: ma+aru+‘3’ = ‘7’; ma+aru+‘2’ = ‘8’; ma+aru+‘1’ = ‘9’ (56). This unusual pattern, which Ross (1988:344) reconstructs for Proto–Eastern Admiralty, is found in only one other Austronesian language that I am aware of—Yapese, an Austronesian language in Micronesia that has so far resisted genetic classification. (Marshallese, at the other end of Micronesia, exhibits fossilized compounds for the numbers 6–9, but they are not indisputably subtractive.)

H’s description also hints at the possibility of another morphosyntactic similarity between the languages of Micronesia (including Yapese) and those of the Admiralties (if Loniu is typical). Loniu has many verbs that have what H calls "long" and "short" alternates (71). The latter tend to be intransitive, while the former are lengthened by one of the transitive suffixes -ani, -eni, or -i. Compare celu ‘stand’ v. celuweni ‘stand (NP) up’ or tang ‘cry’ v. tangesi ‘mourn for’. In some cases, however, either form can precede an object noun phrase, as in yó ma kutiye/kutiyani mwalih ‘I want to tell a story’ (72, 74). H tested for differences between specific and nonspecific objects, noting that yó kutiyani cah? ‘What (story) shall I tell?’ has a nonspecific object with a morphologically transitive verb, while yó to yetehe ’um a yó ‘I am thatching my house’ has a specific object with a morphologically intransitive verb (transitive yetahani occurs in the context ‘Has he finished thatching his house yet?’). It may be that Loniu is simply well on its way to losing any morphological distinction between transitive and intransitive verbs. Many of the Oceanic languages of Morobe Province have lost the same distinction. But it is also possible that further careful elicitation may reveal a more subtle distinction between transitives and intransitives, perhaps along the lines of Hopper & Thompson (1980) or—of greater historical interest—Sugita (1973), who describes variants of a common Micronesian [end page 249] pattern of "semitransitive" verbs and object incorporation. Some of Sugita’s examples from Trukese follow. In (1), intransitive ppek ‘shoot’ occurs with an incorporated object, which in some Micronesian languages cannot be specific. There are no such restrictions on the objects of fully transitive verbs, as in (2). However, Trukese and Marshallese—but not Ponapean or Kosraean (or Yapese)—also allow morphologically intransitive verbs to occur with specific objects, as in (3), which might also be translated ‘[I will] engage in the activity of shooting in relation to the bird’ (Sugita 1973:405).

 (1) Wúpwe ppek macchang

     I-will shoot bird

     ‘I will shoot birds.’


 (2) Wúpwe pekkiiy ewe macchang

     I-will shoot-it the bird

     ‘I will shoot the bird.’


 (3) Wúpwe ppek ewe macchang

     I-will shoot the bird

     ‘I will participate in shooting the bird.’ 

Unfortunately, H did not have the luxury of many years of research on Loniu, as many of the researchers on Micronesian languages have had. She began work on the language at the University of Kansas with a Loniu-speaking student and followed up with eight months of fieldwork in Loniu village. However, she apparently made the most of what time she had. Her description includes many observations about individual variation, and she recorded several texts that furnish a useful corpus against which to test the inevitable hypotheses that never seem to arise until after further fieldwork is no longer possible.

H appears to have done a particularly careful job of sorting out the sometimes tricky morphophonemics, neutralizations, and harmony requirements. However, her choice of orthography is typographically inefficient (although, to be fair, devising an orthography was not one of her goals). The ubiquitous lower mid vowels are rendered as /E/ [open e] and /O/ [open o], while the much rarer high mid vowels are rendered as /e/ and /o/. I prefer the orthography in Hamel (1993), where the lower mid vowels are written e and o, the higher mid vowels are written é and ó, and where ñ replaces /n/ [palatal n], c replaces /c/ [hacek c], and ’ replaces /?/ [glottal stop]. In fact, I have adopted the revised orthography for the Loniu forms cited in this review.

Although based on H’s dissertation, this monograph has undergone considerable revision, incorporating referee comments as well as important later works such as Ross (1988). (In one case [69], however, H was too quick to adopt what I believe to be an erroneous referee suggestion that Loniu’s nominalizing—or adjectivizing—suffix -n derives from the Proto-Oceanic nominalizing suffix *(a)na [sic]. The latter should be *anga [Ross 1988:459], whose consonant should [end page 250] show up as ng in Loniu.) However, the analysis of serial verbs (called "co-verbs" in the monograph) is further revised in Hamel (1993), which was written after the monograph was largely complete.

In general, Loniu verb serialization resembles that found in the Oceanic languages of Morobe Province (such as Jabêm or Numbami) more than it does anything found in Micronesia. The obligatory subject prefixes on Loniu verbs, however, are severely eroded. Only the 2d person singular prefix is unique; 1st and 3d person singular share another shape; and a third prefix serves for all nonsingular subjects. As a result, all main verbs (except 2d person singular) require nouns or independent pronouns in the subject NP slot (Hamel 1993:113). To make matters worse for those trying to distinguish same-subject from switch-subject serial constructions (both of which can be found in Loniu), some verbs appear to take frozen subject prefixes when serialized. Nevertheless, Loniu provides much of interest to those working on verb serialization, especially in the unusually wide range of functions performed by serialized ‘come’ or ‘go’.

H deserves much credit for taking the considerable time and trouble to make her valuable language data available to other linguists, especially during these dark ages for descriptive linguistics in the United States.

Joel Bradshaw
University of Hawai‘i at Mânoa