Notes to Lindstrom

1. The symbol [barred i] represents a mid central vowel. I thank two anonymous reviewers and also participants at the University of Tulsa’s 16th Annual Comparative Literature Symposium, “Towards a Unified Framework in Developmental Linguistics 2” (2006), for helpful comments on this paper.

2. These transcribed recordings are twenty-some years old. My hope in the future is to record similar decision-making meetings to assess the amount and type of Bislama mixings in contemporary island talk and compare this with my 1980s baselines. It would be equally interesting to discover the current status of Taiap, Gapun village’s vernacular, vis-à-vis Tok Pisin, given Kulick’s 1992 prediction of its dimming future.

3. Alpher and Nash (1999), in counterpoint, have argued that word tabooing has had little effect on rates of linguistic change, at least in indigenous Australia.

4. Watt also produced several hymnals, one of which is still in use today, along with a series of primers and catechisms (see Watt 1919).

5. Kwamera, for example, possesses six vowels. Watt’s orthography used a, u, and e variably to transcribe mid central [barred i] (mostly the latter). In words where the [e]/[barred i] contrast is particularly important, Watt used ei and e respectively to  represent the two sounds. My statement that the New Testament is accessible is based on observations of younger Islanders’ use of the book, and my comparisons of the text with contemporary Kwamera lexicon and grammar. I have not, however, quantified this accessibility with formal analysis.

6. From this point, I italicize local and borrowed Kwamera words and underline Bislama terms. To assist the reader, I also use English rather than Bislama orthography to represent Pidgin forms.

7. Watt was able, for example, to improvise Kwamera translations of Holy Ghost, cross, prostitute, altar, hell, sin, Lord, revelation, incense, miracle, and so on.

8. Winford (2003:164–165) notes other difficult cases of classifying mixes as insertion, alternation, or congruent lexicalization.

9. Sankoff (1980) provides a more detailed example of debate Pidgin, as does Kulick (1992:149), who describes oratorical performances in Papua New Guinea where use of Pidgin connotes wisdom.

10. Certain Bislama terms are common in debate. These include against, agree, answer and question, background, clear, example, fine, history, judge, law, lose and win, meeting, point, punishreport, right, scale(m), side, story, trouble, witness, and wrong.

11. In some cases, an entire debate was recorded; in others, I missed the beginning and/or conclusion of debate.

12. This form may be relatively common in Oceanic languages. Similar nativization devices exist, for example, in Abelam and Buin (Laycock and Wurm 1977:198, 200–201).

13. Echo subjects are common in southern and some central Vanuatu languages. This is the plural Kwamera form (m + ha), see Lindstrom and Lynch 1994:33.

14. This is obvious, given the verbal nativization device -o. This readily incorporates Bislama verbs (e.g., -o letem (let), -o pass, -o shake, -o supportem, -o wait, etc.). In many cases, moreover, it is difficult to decide whether a Bislama element has been incorporated in its verbal or nominal form (e.g., -o thinkthink, -o agree, -o play).

15. Most terms were single words. In some cases, I counted several words as a single term (e.g., cases of reduplication such as playplay or instances where Bislama has combined two English words into a single lexeme, such as hurry-up, number-one, one-time). I also counted as a single instance cases in which the same term serves several functions (e.g., sorry as both an interjection and a noun (‘compassion’), or thinkthink as both a verb and a noun (‘idea’ )).

16. Again, I ignore grammatical function here. For example, I combine in my gross count uses of right employed as a noun and an adjective.

17. This must remain a suspicion on my part. Although I worked carefully over the transcripts with linguistic informants, I did not quiz them specifically about meanings they might infer from the presence of particular Bislama items in people’s speech.

18. Many Bislama speakers use uncle to refer to their sister’s children.

19. See Woolard 1987, which describes a code-mixing situation in which mixes serve to level a social boundary rather than to highlight this.

20. Watt also provided a brief Kwamera grammar for MacDonald’s compilation of South Pacific languages (MacDonald 1891:146–171). Discounting obvious errors on Watt’s part and the effects of squeezing an Austronesian language into European grammatical molds, the Kwamera of the 1890s is virtually identical to that of the 1980s. One difference involves a few verbal aspect markers (affixes) that do not today exist (numbers 4, 6, and 10 in Watt’s table (1891:159–160)). The first and third of these may actually be morphophonological variations (conditioned by environment) of affixes #4 and #9—mistakenly identified as independent markers by Watt. Marker #6 involves the reduplica-tion of #5—a reduplication that does not occur today.

21. In this case, however, insaed ‘inside’ takes an alternative possessive construction, associated with one set of directly possessed objects in Kwamera. Instead of nakwai nbarred iteta (‘canoe’s interior’), the possessive relationship is morphologically patterned along the lines of nakau ia pukah ‘pig’s rib’).

22. In that this is a future construction, nbarred ipbarred in ordinarily would be preceded by the marker tbarred i (tbarred inbarred ipbarred in, or tbarred inpbarred in). Speakers do not, however, say *tbarred i taem ‘time’. Here, there is simplification, although tbarred i maintains itself throughout the rest of the language.