Volume 69 1960 > Volume 69, No. 3 > The New Caledonian vocabularies of Cook and the Forsters (Baland, 1774), by A. G. Haudricourt & K. J. Hollyman, p 215-227
THE NEW CALEDONIAN VOCABULARIES OF COOK AND THE FORSTERS (BALAD, 1774)
In this article Dr. K. J. Hollyman publishes further results of his research into New Caledonian languages, with the co-operation of Monsieur A. G. Haudricourt, Research Assistant at the National Council for Scientific Research in Paris, who spent some time in New Caledonia last year investigating the languages there.
NEW CALEDONIA was sighted by Cook on September 4, 1774, and he spent eight days at Balad in the far north, from the morning of the 5th to dawn on the 13th. Both he and J. R. Forster recorded in accounts of the voyage vocabularies collected at Balad and in the environs. These words appear in comparative tables. 1 A few words are also quoted in the text of their accounts by Cook and Georg Forster. The words from the tables only were listed (not without error) by S. H. Ray, along with the parallel terms published by Fabre, Inglis, and von Gabelentz. 2 Beyond noting that, by these standards, the vocabularies of Cook and J. R. Forster were “fairly accurate”, 3 Ray drew no conclusions. The purpose of this paper is to make a more detailed study of all the words recorded at Balad by Cook and both Forsters, by checking them initially against Leenhardt's Tableau 4 of NC 5 languages, but more particularly against the results of an investigation in the area, during the autumn of 1959 by A. G. Haudricourt. 6
Writers about New Caledonia usually classify the words recorded by the early European visitors as belonging to “the language of Balad”.- 216
The name of the language spoken there is Nyalazyu (noted by Leenhardt as Yalasu). Along with Belep, Nenema and Koumac, this language forms the northernmost group of NCN languages, characterised in particular by the same phonemic structure.
The main phonetic features of the language of Balad are as follows. There are ten vowels, five short and five long: a, e, i, o, u, ā, ē, ī, ō, ū. Leenhardt did not consistently note vowel length, but it is clearly phonemic: then “to close”, thēn “to plant”; khat “tuber”, khāt “small lobster”. The vowels are subject to nasalisation when in proximity with a nasal or nasalised consonant, but this nasalisation is non-phonemic. The consonants may be broadly tabulated as follows (phonetic transcriptions are given in brackets where necessary):
MAP II- 218
Northern New Caledonia, to illustrate Cook's visit, with an inset (after Leenhardt) showing the linguistic areas.
Of the aspirates in column 3, th is always occlusive, but the remainder are commonly realised as spirants. In intervocalic position, and within word-groups during rapid speech, the consonants in columns 2 and 3 are realised as the voiced fricatives in column 4. Of the nasals in column 5, ny and ng are rare: ny is mainly heard as y, which also represents an earlier zy [Dj]; ng is difficult to distinguish from the prenasalised g. The apparent aspirates in column 6 are voiceless, tense counterparts of the voiced, lax consonants in column 5.
It is also useful to note here the possessive suffixes used with names of parts of the body:
We begin our review of the information given by Cook and the Forsters with their comments on the phonetic characteristics and family relationships of the Balad language.
. . . the inhabitants of New Caledonia have many nasal sounds, or snivel much when speaking. 7
They are likewise niggards of speech, and we rarely met with individuals among them, who took a pleasure in holding converse with us. Their language therefore seems to be uncultivated, and their pronunciation so indistinct, that the vocabularies which several of our shipmates collected, disagreed remarkably. Though they have few harsh consonants, they have a frequent return to gutturals and sometimes a nasal sound, or rhinismus, which commonly puzzled those who were not acquainted with any other language than English. Perhaps, their plantations lying remote from each other, are the means of preventing that familiar intercourse which would gradually give the pleasures of society. 8
NC languages generally have a complex phonemic system as compared with English or German (the latter being the Forsters' native tongue). 9 The outstanding feature with the language under discussion is the high level of nasality: five prenasalised and eleven nasal consonants, with the contextual nasalisation of vowels. It is not surprising then that this characteristic should have struck the first European visitors. As for the “frequent return of gutturals”, there are only four of these, but no statement can at present be made about their frequency.- 219
. . . but they eat of some yams, which we happened to have yet left, calling them Oobee. This name is not unlike Oofee, as they are called at most of the islands, except Mallicollo; nevertheless, we found these people spoke a language new to us. 10
For oobee, see Part III, no. 302, below. The PN and NC words are related, being both reflexes of AN *qubi(h).
Their language, if we except the word areekee and one or two more, had no affinity with any one of the various languages which we had heard in the South Sea before. This was the more surprising to us, as we had found one language, or at least dialects of it, in all the easterly islands of the South Sea, as well as at New Zealand. 11
Aliki is, of course, a borrowing from the PN of Uvea. 12
They communicated a number of words of their language to us, which had no affinity with those we had learnt before in other islands; a circumstance sufficient to discourage the greatest and most indefatigable genealogist. 13
[In Malekula, Tanna and New Caledonia] some words are found, which seem to have a distant resemblance to those that go before [i.e., to the PN] . . . Yet, whether these may not have been accidentally introduced, is hard to determine; because they frequently use two words to express the same thing; as for instance, in New Caledonia, they call a star both Peejoo, and Fyfatoo: the first seems most consonant to the general composition of their language, whereas the second differs very little from Efaitoo or Whettoo, the name of a star at Otaheite. When they mention Puncturation, it is commonly called Gan or Gangalang, but sometimes they say Tatatou, which is almost the same as Tatou, used to express the same thing at Otaheite and Amsterdam [Island]. 14
Cook is here making very plain the existence of PN borrowings, in common use alongside the native MN word.
From their vicinity to New Holland, one might have been apt to suppose that they had the same origin with the people of that continent; but upon comparing all the accounts of former voyagers, who have visited New Holland, its inhabitants bear no resemblance to one another, and as a farther proof, the vocabulary of both nations is totally different. 15
There is, however, some similarity between their manufactures, and those of Tanna . . . The language, which on these occasions is the surest guide, is totally dissonant. 16
Were I to judge of the origin of this nation, I should take them to be a race between the people of Tanna and of the Friendly Isles; or between those of Tanna and the New Zealanders; or all three; their language, in some respects, being a mixture of them all. 18It would have been difficult for an amateur at the time to have come much closer to the view of modern linguists.
The full list of words recorded by Cook and the Forsters gives translations of fifty English words. It has appeared convenient to consider them in the order in which they occur in Leenhardt's Tableau, and the accompanying number is that of the relevant item there. In each case, the Balad form recorded by A. G. Haudricourt is quoted, without the language name. Where another language is involved, its name is given. It has not been considered necessary to reproduce the orthographical devices used by Cook and the Forsters to render the pronunciation clearer. Unless otherwise indicated, the source is always the Tables of Cook or J. R. Forster.
Cook: garmoing? Unidentified.
Forster: maninya. mhwadija “our noses”.
Forster: noo-anya. nūāja “our mouths”.
Cook: pennawein. penīān “his teeth” (or penīāng “my teeth”).
Forster: poon. If Forster is using German spelling, as seems to be confirmed by no. 26, this is pon “hair (of head or body)”; if his spelling is English, which seems less likely (but cf. 383, 955-58, 966 below), this is Mwenebeng pun (same meaning).- 221
Forster: poonwang. pōnūāng “my beard”. Since Forster's words are consistently recorded with plural possessive suffix, except for this one case, it seems probable that the singular is used here because he himself was clean-shaven. 21
Cook: whanbooeen. phwabuin “his navel” (or phwanbuing “my navel”). The sense of phwa is “hole, opening”, of bui- a convex area such as the calf of the leg, etc. (Belep).
Cook: bandonheen. There is no individual name for “hand”. bwaodaen “his fingers” (or bwaodaeng “my fingers”). The initial bwa- is a collective prefix.
Forster: hea. Unidentified. The absence of any sign of a possessive suffix makes the word particularly suspect.
Forster: ghung. Unidentified.
Forster: ooma. mwa “hut, dwelling”. Forster may have met a Iaispeaking visitor from Uvea, where uma is the word, but [Nmwa] could easily be heard as ooma by a European.
Forster: halap. This does not resemble Leenhardt's word for “earth”. For “the heart of the tribal land”, Belep has halafoemwa: 22 halap could be a partial rendering of this. It could also represent holap “oven covering”. 23 However, the recent enquiry recorded bwe alnap “sandy beach” (bwe = “top, surface”), with a corresponding kalnap for Koumac. As this Koumac k—Balad zero correspondence is a regular one, more information is needed before we can be certain of identifying alnap with Forster's halap, because of the initial h-.
Forster: dallai. dalac “sea”.
Forster: at. at “sun”.
Cook: ooe. we “water”, see 189 above. ora = “rain”, we ora = “rainwater”. 24
Forster: ooba. uva [uBa] “taro leaves”.
Forster: munghee. Mwenebeng, munji “native banana”.
359 (cf. 553). Cloth
Cook: hamban. habwān “bark-cloth covering”.
383. Coconut 26
Forster: ta eeka. Uvea, ika “fish”.
503. Hatchet 28
Forster: babbanew. Leenhardt records bāwadu, for which we may postulate an earlier bāvwaju [mba:Bwandju], much closer to Forster's word.
Forster: aït. ac [atj] “man, person, male”.
Cook and Forster: tama. thāmwa, term used to call the women.
Forster: halleek. phalic “ill”.
Cook: eeva, eeba. Unidentified, but cf. Mwenebeng bwa.
Throughout the numeral series, Cook gives the set used of large objects such as ships, and Forster the set used of inanimate objects such as yams.
(a) Cook: wannoonaiuk. wādulnik.
(b) Forster: parooneek. pwadulnik.
Forster: alee; mailee. Unidentified.
Forster: hoot. huc “eat (of starchy foods)”.
Cook: oodoo, oondoo; Forster: hyndoo. udu.
Cook: wyoo. wayu “whistle with the lips”, distinct from yēdit “whistle with the tongue” (neither word listed by Leenhardt).
There would be little profit in analysing closely the accuracy of representation shown by Cook and the Forsters. Naturally enough, they use their own native spelling systems, which were not adequate to the need. Their native phonemic systems were also dominant, and as a result they misrepresent most of the new sounds heard. The unfamiliar j, for example, is recorded by Cook as g (14, 767, 774), by J. R. Forster as g (14, 322, 769, 774), ny (10, 17, 18), and n (503). The unusual nasal l (ln) is registered as n by Cook (14, 451, 776), as l (14, 236), r (767) and n (776) by J. R. Forster. The labiovelars are noted as simple occlusives (57, 359, 503, 767-776) or nasals (10, 119, 236, 621), and there is confusion over the presentation of the prenasalised consonants. None of the men, of course, was prepared for the task of recording the complex phonology of MN languages. Their knowledge of PN, with its simpler phonemic organisation, was of little help, and it is doubtful whether Georg Forster's acquaintance with other European languages, which he used to test the phonetic abilities of South Sea Islanders, 35 gave him any edge on Cook. Before reaching Balad, the men had a brief apprenticeship in the New Hebrides, at Tanna and Malekula, but that was all. Their achievement is all the more admirable, particularly since so few of the words are unidentifiable.- 225
The overwhelming majority of the words do belong to the expected language, but there are some which do not, and their presence must be explained. There are several PN words, at least one Mwenebeng, and a possible one from Iai.
Polynesian. Of the PN words listed, the following are Uvean: 36 masina (236), fetū or fetu'u (241), 'ufi (302), niu (383), vaka (411), ika (432), 37 aliki (612). Where the Uvean word is not known, the languages of the original PN settlers on Uvea have been quoted: Wallis, Samoan, Futunan. 38 These explain vai (189), tātatau (558), 39 io and oe (722-23). It is therefore very clear that, although none of the visitors comments, as Labillardière was to do later, on the presence of Polynesians in the Balad area, the Wallis settlers had already arrived in Uvea and had already visited Balad before 1774. 40 As a result, Balad inhabitants themselves—and Cook, as we have noted, points this out—had alternative words available for a number of concepts, and it was natural that they should use them in speaking to strangers who, no doubt, tried to communicate with them in PN. In two particular cases (411 and 612), a Balad pronunciation of a PN word is registered.
The possible Iai word, uma (119), if correct, would not be surprising in view of the known relations between the Loyalties and the north coast of the mainland.
Mwenebeng. The Mwenebeng word is munji “banana” (322), which has no apparent cognates in Leenhardt's list. While it could be explained by the presence of a group of visitors from the Pwebo area, the true answer is more likely to lie in a different direction. In Northern New Caledonia the clans fell into two groups which were hereditary enemies, the Ohot and the Hwahwap. 41 The enmity had as counterpart the interchange of wives, a process which the New Caledonians themselves give as the major reason for linguistic change:
The women teach their children their mother tongue; whatever area they come from, they prepare their son for the day he will go to the home of his maternal uncle, and for this purpose the knowledge of the kanya's language will always seem to them indispensable. In the same way, the daughter will have to understand the language of the “brother”, boru nya, who will one day be her husband. And if the women coming from a clan speaking a different language are numerous, their tongue will little by little be understood by the whole generation. Thus, in the region of Cap Bayes, towards Poindimié, we saw Camuki withdraw in twenty years before Pati simply because the women, who previously came from Tuo, were taken in the preceding generation from Bayes and Poneriwen. 42- 226
Camuki areas lost to Pati as a result of a change in the clans with which intermarriage took place.
The Balad and Arama clans were Ohot, while the Belep Islanders and Mwenebeng speakers were Hwahwap. Hence the appearance of a Mwenebeng word at Balad. Nearly every New Caledonian was at least bilingual, speaking his own and, usually, the neighbouring language. The custom of using synonyms to which Cook refers in connection with PN words, was in fact just as current with MN words from different languages or dialects, and was a characteristic feature of oratory and songs. Leenhardt quotes from a Wailu pilu chant: go kuru, go sala ro, which New Zealanders could represent by “I sleep, ka moe ahau”—kuru being the Wailu word, sala the Bwewe, for “sleep”. These semantic pairs are also found in everyday speech. 43
In conclusion we may sum up by saying that the linguistic record left by Cook and the Forsters as a result of their visit to Balad is surprisingly accurate in view of their lack of training and limited experience. The PN words they registered make it quite plain, whatever the indications of Wallis and Uvean genealogies, 44 that PN speakers had been in the Balad area well before 1774, with the result that MN speakers had adopted some Uvean words. The recording of at least one Mwenebeng word suggests that the effect of intermarriage on linguistic change was operative nearly 200 years ago, and that it may legitimately be considered in the explanation of anomalies when reconstructing earlier stages of NCN languages.- 227
1 Cook 1777:365 (hereafter referred to as Table); Forster 1778: facing p. 284 (hereafter: Table).
2 Ray 1926:12-13.
3 Ray 1926:16.
4 Leenhardt 1946:245-647.
5 For the abbreviations used, see Hollyman 1959:357 n. An additional one is NCN, for the Northern group of New Caledonian languages (see Map I). Phonetic transcriptions appear in square brackets; the symbols used are those of the International Phonetic Association, except that typographical difficulties have made it necessary to use small capitals for four consonants: [B] represents the voiced bilabial fricative; [D] the voiced dental fricative; [G] the voiced velar fricative; and [N] the velar nasal corresponding to [n].
6 This article was originally prepared by Dr. Hollyman on the basis of Leenhardt's transcriptions. It has been largely re-written as a result of the accurate notation of M. Haudricourt, whose willing collaboration is an excellent example of international co-operation. (Editor's note).
7 Cook 1777: Table.
8 Forster 1777:430.
9 cf. Lenormand 1952 or 1954; Hollyman 1960.
10 Cook 1777:106.
11 Forster 1777:381.
12 cf. Hollyman 1959:361.
13 Forster 1777:393.
14 Cook 1777: Table.
15 Forster 1777:324.
16 Forster 1777:426. With both these last extracts, compare Forster 1778:278.
17 The Tanna list is reproduced by Ray 1926:15-16.
18 Cook 1777:120.
19 The NC words appear to require *d- for *talinga (cf. Haudricourt 1951:148). This could be due to the influence of *deneR “hear”.
20 If we omit the unidentified garmoing? (1) and gangalang (558b), Cook actually records -ng “my” once (14) and -n “his” four times (17, 19, 42, 57). One might have expected a similar consistency in the possessive pronoun to that observable with Forster. But Cook records the sound ng [N] in another word (411) and does so regularly with Maori words (Hawkesworth 1773:70-71); it was also familiar to him in English. It has therefore seemed advisable to interpret the possessive suffixes literally, as he gives them.
21 See the 1784 portrait by Anton Graff reproduced in Kersten 1957: facing p. 32, from the collection of Herr Joseph Jans, Lucerne (reference kindly provided by Dr. J. A. Asher). The wearing of beards by New Caledonian men is evidenced not only by the portraits in Cook's account, but also in a specific reference by Forster 1778:239.
22 Leenhardt 1946: no. 188.
23 Leenhardt 1946: no. 512. The covering of the oven consisted of earth lying on a layer of leaves and topped by strips of niaouli bark (Leenhardt 1937:86; cf. Lambert 1900:133).
24 Leenhardt 1946: no. 253.
25 Cook 1777:106.
26 cf. Hollyman 1959:356 (Map II), 358-359.
27 Forster himself (1778:458 n.) remarks, in speaking of Tahitian canoe names: “The name E-waha is certainly pronounced by the natives of the Friendly-isles and in New-Zeeland with a stronger sounding of the aspirate h, by saying Te-wagga.”.
28 Described by Forster 1777:409, and illustrated by Chapman in Cook 1777: facing p. 122.
29 Cook 1777:116 (“alekee (chief)”).
30 Forster 1778: Table. In his text (1778:380), he writes: “In the more Western isles of Mallicollo, Tanna and New Caledonia, we observed likewise chiefs, under the denominations of aleeghee or areekee.” In the Table, areekee is attributed to Tanna.
31 Cook 1777:124 appears to be the only one of the three men who saw tea as a word: “Tea seems to be a title prefixed to the names of all, or most, of their chiefs or great men. My friend honoured me by calling me Tea Cook.” The Forsters appear to regard the word as part of the name: Tea-booma (1777:380), Te-abooma (1778:471).
32 Lambert 1900:61.
33 Forster 1777:380, 381 (“areekee or king”).
34 Colomb 1891:123.
35 Forster 1777:214.
36 As recorded by Leverd 1922.
37 cf. Hollyman 1959:371.
38 Bataillon 1932; Violette 1880; Grézel 1878.
39 cf. Hollyman 1959:380.
40 cf. Hollyman 1959:361.
41 Leenhardt 1930:105; Guiart 1953:97; Person 1954:30-31.
42 Leenhardt 1930:262. See Map III. We cordially thank Mr. W. R. Ambrose, who drew Maps II and III.
43 Leenhardt 1946:xvi-xviii.
44 cf. Hollyman 1959:361-362.