6.1 Edward Markham

In 1834, the English traveler Edward Markham made incidental sociolinguistic observations of northern New Zealand as part of his New Zealand or Recollections of It. They suggest that the son of one of the missionaries, a Mr. Butler, and a trader by the name of Jacky Marmon spoke the Māori language fluently (Markham Reference Markham and McCormick1963: 31, 38–39). In addition, Markham (Reference Markham and McCormick1963: 71) recognized a Mr. Hamlyn who had learned Māori so well as to be able to preach in it, to reflect a southern accent, and to pass as a native.

On another occasion, the record provides evidence for the Māori version of MPP, as suggested by an accompanying verbal or textual note of unknown provenance:

The supplementary interpretation of “Kiki” as Pidgin Māori matches the available historical attestations for kaikai ‘to eat’, and ultimately agrees with Markham's description of sociolinguistic circumstances in which Māori presumably used the Pidgin:

Markham thus suggests by inference that it is not only Māori sailors, but also their wives and Māori women of European sailors who would have used the Māori Pidgin on board of whalers, as Roger M. Keesing (Reference Keesing1988: 15–21) has similarly recognized native women in the role of linguistic and cultural brokers on whalers for other Pacific pidgins. The historical record, however, leaves no clues as to whether the children who grew up in these interethnic marriages creolized MPP.

6.2 Richard Henry Dana

As a primary literary source of the period, Richard H. Dana's Two Years before the Mast (Reference Dana1911 [1840]) describes his experiences as a sailor in the Pacific hide and tallow trade in 1835 and 1836. American or English officers and two or three experienced sailors “before the mast to do the work upon the rigging” worked side by side with crews of Hawaiians, known as Kanakas (< Hawaiian kanaka ‘human being, man, person, individual’) and appreciated as experienced boatmen in rough surf and excellent swimmers (Dana Reference Dana1911 [1840]: 68–75). Hawaiians also worked in the rigging, showed no fear in tackling sharks or rattlesnakes, helped cure hides, and served as messengers swimming after a bypassing vessel; they adapted easily to new conditions other than cold climates, but remained in a community of their own in San Diego (Dana Reference Dana1911 [1840]: 103, 177, 179, 195, 207, 217–218).

The literary sailor compared the sociolinguistic situation on board to Babel, and found several languages spoken on ships of Hawaiian, Italian, and Mexican origin with their mixed crews of Americans, English, Spaniards, French, and [American] Indians in addition to Hawaiians – with “all talking at once” and with the Hawaiians reportedly conversing continuously (Dana Reference Dana1911 [1840]: 103, 166, 186–187). On shore in southern California, one language prevailed in interlingual interactions:

On board, English did not figure as a major medium with the Hawaiians, only one of whom spoke a little of it (Dana Reference Dana1911 [1840]: 75, 80, 102), and was in use with Hawaiians of the San Diego community but in limited fashion (Dana Reference Dana1911 [1840]: 179–180). Whereas one Hawaiian by the name of Tom Davis reportedly was quite fluent in English, another older member of the community, Mr. Bingham, “spoke very little English, – almost none, and could neither read nor write” (Dana Reference Dana1911 [1840]: 181). Nonetheless, Dana had him respond largely in Pidgin English when teased about being a cannibal:

Unless we were to question Dana's observations about Bingham's linguistic skills, this example – frequently cited as prime evidence of Pidgin English in the early nineteenth-century Pacific (see, e.g., Clark Reference Clark1979: 29; Goodman Reference Goodman and Bickerton1985: 11; and Keesing Reference Keesing1988: 14) – suggests that it was not an accurate rendition of what the old man said. Instead, Dana evidently used Pidgin English as no more than a literary medium to represent the Hawaiian's speech in translation, the original of which may well have been in vernacular Hawaiian or Pidgin Hawaiian. Dana's short quote with the particle ua marking completed action and with the plural definite article nā point to vernacular Hawaiian rather than some reduced form, in agreement with historical data (see Dana Reference Dana1911 [1840]: 150, 175, 184, 207, 309, 334; T. Haunani Makuakāne-Drechsel, personal communication, 27 June 2008). What lends further support to this interpretation is the fact that Dana's samples of Pidgin English do not resemble later historical attestations of Hawai‘i Pidgin and Creole English in any fashion, as already recognized by Sarah J. Roberts (Reference Roberts1998: 14). Dana's attestations of Pidgin English do not match available historical evidence and would suggest a distinct Pidgin English for Hawaiians in California, for which there, however, was no historical or theoretical rationale.

Elsewhere, Dana (Reference Dana1911 [1840]: 180) actually referred to the Hawaiians as using “a sort of compromise, a mixed language…on the beach, which could be understood by all.” Ultimately, the only evidence identifying any form of linguistic compromise that Dana (Reference Dana1911 [1840]: 175, 334) offered consists of the following samples, the first one of which Ross Clark (Reference Clark1979: 60, fn. 25) and Roberts (Reference Roberts1995a: 19) already documented as Pidgin Hawaiian:

That the first utterance in (29) occurs in the immediate context of a sample in vernacular Hawaiian does not necessarily militate against interpreting it as Pidgin Hawaiian, since recordings of both may reveal no more than a situation in which speakers switched between the two depending on their audiences.¹ What instead marks Dana's examples above is the fact that they are interpretable with the underlying grammar of either a Polynesian language or that of a European language without creating any substantial differences in interpretation.

Dana's samples of “Hawaiian” therefore appear to represent a mixture of vernacular and Pidgin Hawaiian. Like some European and American colonists who had settled in the Hawaiian Islands, Dana had begun learning or had learned vernacular Hawaiian. This finding is not surprising in light of the fact that he showed much sympathy towards Hawaiians and associated with them on board and in San Diego over several months. During this period, Dana became a close friend or “aikane” (aikāne) of a Hawaiian by the name of Hope, whom he assisted in illness when others refused to help (Dana Reference Dana1911 [1840]: 102, 179–185, 308–311, 347). On another occasion, Dana (Reference Dana1911 [1840]: 311) “could not understand half of them [Hawaiians]”; but this instance was apparently due to several Hawaiians speaking at the same time, expressing their gratitude to him for his generous help, rather than for a lack of knowledge of Pidgin Hawaiian or Hawaiian.

In the end, Dana's recordings of both vernacular and Pidgin Hawaiian are consistent with comparative evidence, permit systematic reconstitutions, and thus are beyond suspicion as historical data. The same conclusions, however, do not hold true for his attestations of Pidgin English, which again raise questions of cultural blinders.

6.3 Dumont d’Urville on his third voyage to the Pacific

When in the years of 1837 to 1840 the French explorer Jules S. C. Dumont d’Urville returned to the South Pacific a third time for further explorations of the region (including Antarctica), he observed a mixed medium of Marquesan, English, and Spanish by an on-board visitor called “Moe” at Hiva ‘Oa, Marquesas Islands, in 1838:

Two of Dumont d’Urville's officers, Second Lieutenant Louis François Gaston Marie Auguste de Roquemaurel and Sub-Lieutenant Jacques Marie Eugène Marescot, reported the same incident in slightly different terms:

What Dumont d’Urville and his officers evidently perceived in slightly different terms need not reflect contradictions or raise doubts about the historical accuracy of their accounts. Rather, their different emphases on “a kind of mixed jargon of English, Spanish, and Nukuhivan” (Dumont d’Urville), “bad English” (Roquemaurel), and “some words of the Tahitian dialect,” English, and Spanish (Marescot) would seem to reflect different aspects of the same larger process of language contact in which Moe mixed MPP with Pidgin English and occasionally embellished it with Spanish loans. In addition, these descriptions apparently focus disproportionately on the European elements that Dumont d’Urville and his officers reportedly identified, and would neglect to recognize the non-European elements fully.

Regrettably, Dumont d’Urville and his officers did not offer any example of Moe's speech other than two short phrases of acknowledgment in English and a longer construction in Pidgin English: “me steer ship very well” (Dumont d’Urville Reference Dumont d’Urville1841–1846: III.231). The French explorer had only the following to say about Marquesan-English language contact:

Nor did Dumont d’Urville recognize any parallels between the Marquesan–English–Spanish jargon – i.e. a Marquesan variety of MPP with some English and perhaps Spanish loanwords – and the reduced Māori that he had recorded in New Zealand a decade earlier (see (24) in 5.7 above). On the other hand, if Marquesans indeed borrowed substantially from the English language in contact with Americans and perhaps other foreigners, the French explorer surprisingly did not draw any comparisons with Pidgin English of Patagonia, which he had documented in a part-Iberian sociolinguistic context, including Portuguese-derived pikinini ‘child’, in the first volume of his travel account (Dumont d’Urville Reference Dumont d’Urville1841–1846: I.150).

In weighing these two omissions, we would have to consider the second one as more serious. Dumont d’Urville drew on a greater familiarity with English than with Polynesian languages for the recognition of shared English features. In addition we would have to recognize a much shorter time span of only some seven months between his sociolinguistic observations in Patagonia and those in the Marquesas – in contrast to more than ten years that had lapsed since his observations on a reduced Māori. With the Marquesan–English–Spanish medium reportedly displaying a major English component, similarly to Pidgin English of Patagonia, we would expect the French explorer to have taken notice of some resemblance, given his earlier experience in philological research with Pacific languages (see Dumont d’Urville Reference Dumont d’Urville1833–1834) and his penchant for comparative anthropological observations within and beyond the Pacific. That the French explorer made no note of such parallels to Pidgin English in Patagonia therefore raises doubts about any extended English impact on the Marquesan–English–Spanish medium, although we may otherwise trust the observations by Dumont d’Urville and his officers about the presence of English influences by the fact of their own French cultural preferences.

In the end, Dumont d’Uville did not remain long enough in the Marquesas or the Society Islands to include any further observations on MPP in the massive ten-volume account of his third voyage to the South Pacific. Nor did he or his officers take special notice of the Pidgin during their revisit of New Zealand in 1840 (Dumont d’Urville Reference Dumont d’Urville1841–1846: IX.93–204), possibly because of its lack of novelty. Instead, Dumont d’Urville (Reference Dumont d’Urville1841–1846: IX.178) or his posthumous editor C. A. Vincendon-Dumoulin translated a single quote by Māori chiefs from what likely was Māori or Pidgin Māori, perhaps intermixed with some English, into standard French at a time when New Zealand came under increasing British control. According to Roquemaurel (in Dumont d’Urville Reference Dumont d’Urville1841–1846: IX.281), Māori also greeted the French by “le nom de Oui-Oui ou Yanapa, mots qui, pronouncés souvent par les marins, ont sans doute frappé les naturels,” that is ‘with the name Oui-Oui ‘Yes-Yes’ or Yanapa [a distortion of il n'y a pas ‘there isn't’ in French]’ – words that, often pronounced by the sailors, undoubtedly struck the natives. In the same vein, the lieutenant did not, however, hesitate to label his hosts in the same breath as children or to insult them as ‘very vile, defiled, very much below their reputation’. The latest travel accounts by Dumont d’Urville yield little of the great interest that he had in the languages of the Pacific, as his editor Vincendon-Dumoulin (in Dumont d’Urville Reference Dumont d’Urville1841–1846: IX.58–59) already recognized.

6.4 Edward Lucett

While voyaging across the Pacific from 1840 through 1849, the English pearl trader Edward Lucett left incidental impressions on how Māori, Tahitians, and other Pacific Islanders related to Europeans by language. While in New Zealand in 1840 or 1841, he explicitly recognized a “broken Maori” with elements of “broken English”:

Evidently, Lucett described here Pidgin Māori that gradually underwent language mixing from Pidgin English, and illustrated the process with incidental observations and several examples of crosscultural encounters:

Among the above examples, one clearly stands out as MPP:

The word “utn” in this instance is an obvious misprint of “utu” or Māori utu ‘return for anything, satisfaction, ransom, reward, price, reply’. The appearance of the accompanying article te here does not constitute much surprise in light of earlier instances with the definite article (see, e.g., (7), (15), (17), and (27) in Chapters 4 and 5 above). Although “te” or the variable form “ti” appears in mixed expressions with “kai-kai” such as “kaikai te dinner” ‘we eat some dinner’, “Kai-kai ti supper?”, and “kai-kai te (breakfast)” above, it did not function as the definite article te or even some Polynesianized rendition of the English article “the.” Instead, it derived from a shortened form of (i)ti ‘small, little, few’ (< Tahitian, Marquesan, Māori, Hawaiian) as a consequence of the deletion of the initial high vowel of iti due to its coalescence with the final disyllabic sequences of vowels in kaikai and as indicated by the English gloss ‘some’ in the second instance. These constructions then really read as

One could similarly reinterpret the previous example “Omai ti utn” [sic] ‘give me the payment’ as Homai (i)ti utu ‘Give a little/some payment’ by assuming that Lucett did not fully understand the request; but the original gloss, including a definite article, would speak against such an interpretation. In either case, the constructions in (30) and (31) suggest an order of the modifier preceding the modified noun, if they still leave us in the dark about the position of the subject within the sentence. These examples also reveal that Lucett did not have much understanding of the use of the definite article or the near-homophonous short form of (i)ti ‘little, some’. Lucett indeed admitted to having no more than limited knowledge of the Māori's language (Lucett Reference Lucett1851: I.87, 88, 108) and apparently relied on others who could speak it (Lucett Reference Lucett1851: I.109, 115). In the end, the English pearl trader mentioned several other incidents of Māori–English encounters not in Pidgin Māori, Pidgin English, or any mixture of both, but in a translation of English, which probably cannot stand as an accurate historical record (Lucett Reference Lucett1851: I.97, 101, 102, 111, 114–115, 134–135).

For Tahiti or other, neighboring South Pacific islands of the early 1840s, Lucett did not explicitly suggest the existence of Pidgin Tahitian or some related contact medium with elements of Pidgin English or French. In its place, he cited an incident by “[o]ne of our divers who could talk a little English” and who warned him in Pidgin English about an imminent attack by “Angatans” or Easter Islanders (Lucett Reference Lucett1851: I.249–251). He also recognized Marquesan Islanders who had sailed on whale ships to speak “in perfectly intelligible English” (Lucett Reference Lucett1851: II.195). Still, some attestations of Pidgin English again raise the question of translations. Lucett relied regularly on native interpreters in contact with Pacific Islanders (Lucett Reference Lucett1851: I.237, 250, 252, 255, 305; II.129) – one of these interpreters was a Hawaiian who apparently understood no more than a few words of English, although he had served on European or American whalers since his childhood (Lucett Reference Lucett1851: I.325, 327). Moreover, the English pearl trader reported that his own “unacquaintance with the language [on Hao or Haorangi, Tuamotu Archipelago, for example] prevent[ed] recording much that might have proved curious and interesting” (Lucett Reference Lucett1851: I.321–322). On the other hand, secondary clues suggest a reduced form of Polynesian in contact with Pacific Islanders in the Society Islands, the Tuamotus, and the Marquesas. Not only did Lucett (Reference Lucett1851: II.195) attest to the use of his own “bush rambles” in the Marquesas, but he finally provided a one-word example of his own in what he called “the Kanaka language” and what by its context was MPP, apparently on Tuamotu:

One of the two women possibly responded the way that she did because of her blindness; but what Lucett described above suggests a limited verbal exchange characteristic of pidginization by the other woman. Ultimately, Lucett again revealed but an incomplete understanding of MPP, as he evidently failed to recognize any parallels in the sociolinguistics of New Zealand and that of the Society Islands or neighboring islands.

6.5 Herman Melville

Prime sources for MPP of the period were Herman Melville's semi-fictional, autobiographical book Typee of 1846 (Melville Reference Melville, Hayford, Parker and Tanselle1968a) and, to a lesser extent, his sequel novel Omoo of 1847 (Melville Reference Melville, Hayford, Parker and Tanselle1968b). Subtitled A Peep at Polynesian Life, the first book drew on his personal experiences of about a month on Nuku Hiva in the Marquesas Islands in 1842, after he had jumped ship from a whaler and had gone into hiding among the Taipi people, turning into a gentleman-beachcomber. His second book, subtitled A Narrative of Adventures in the South Seas, built primarily on his visit to Tahiti and neighboring Mo‘orea (“Eimeo”), after he had escaped from his Marquesan hosts turned captors.

Like other exploratory accounts of the Pacific, Melville's novels have come garnished with several widely used terms of Polynesian origin: “arva” for ‘kava’, “Hoolah Hoolah” for ‘feast’, “Kannaka” for ‘South Sea Islander’, “mortarkee” for ‘good’, “taboo” for ‘forbidden’, “tappa” for ‘tapa’, “taro” for ‘taro’, and “whihenies” or “whihenee” for ‘young girls’, among others (Melville Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 14, 74, 75, 85, 88, 132, 139, 165, 237, 238; Melville Reference Melville, Hayford, Parker and Tanselle1968b [1847]: 20, 117, 173, 174, 181, 272). Except for a few items that without further comparative research remain unreconstitutable and thus unidentifiable as to their sources, most of Melville's vocabulary evidently derives from a common Eastern Polynesian origin with a few characteristically Marquesan and Tahitian words and several Hawaiian entries; it mainly relates to the semantic domains of tribal identity and subsistence, and apparently reflects the circumstances in which the author found himself in contact with the Taipi on Nuku Hiva. These words, however, appear in the accompanying vocabulary (Drechsel Reference Drechsel2014) only if Melville confirmed their use in a larger construction of MPP.

Melville already took exception to the widespread opinion of the period that the Pacific Islanders’ language was devoid of words by which to describe any moral issues, and remarked sarcastically

He thus anticipated any claims that Pacific Islanders were limited in their linguistic expressions. Unlike many contemporary witnesses of the early colonial Pacific, Melville enriched his novels with tidbits of his conversations, real or imagined, with indigenous people and related linguistic observations, and so often revealed glimpses of MPP amounting to a grammatical sketch. Even more than John Savage or John Liddiard Nicholas (see (14), (17), and (18) in 4.9 and 5.3 above), the beachcomber-writer offered highly regular linguistic data to permit the deduction of clearly identifiable grammatical paradigms (see (32) through (40) below).

A major case in point is the author's illustration of the syllable structure of native speech, when he recounted a principal Taipi chief's attempts at pronouncing the narrator's assumed name of Tom:

In describing Mehevi's pronunciations of “Tom” as something like [tomo], [toma], and [tomi], Melville thus displayed an awareness of the pattern of syllables ending in a vowel for Marquesan, whether standard or reduced.

The American beachcomber also documented a quite extensive use of reduplications to indicate mass or plurality, repetition or continuation, or intensity (Melville Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 72, 81, 84, 91, 94, 103, 116, 176, 237, 238, 270; Melville Reference Melville, Hayford, Parker and Tanselle1968b [1847]: 125, 127, 157):

Some of these reduplications prove to be retentions from their Polynesian source languages (as is the case with fe‘efe‘e, hulahula, kaikai, and paepae); others appear to be innovations unique to MPP.⁶

Melville, moreover, appeared to be aware of a distinct postnominal and postverbal pattern for modifiers in noun and verb phrases respectively (Melville Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 17, 266; Melville Reference Melville, Hayford, Parker and Tanselle1968b [1847]: 129, 258):

Melville's two samples of a noun immediately followed by a modifier in (33) would seem to exhibit a zero copula before a non-verbal predicate. Speakers of English and other European languages might interpret such constructions as corresponding to a nominal phrase in which the subsequent modifier functions as an adjective; but this pattern also reflects the Polynesian syntactic pattern of a noun and an apparent adjective, actually a stative verb, in postnominal position, which Roberts (Reference Roberts1995a: 6) has already described as “a clausal juxtaposition” or “a zero-marked clause embedding” for Pidgin Hawaiian and which is equivalent to a relative clause in English. What Europeans interpreted as a sequence of a noun and a predicate with a zero copula, then, was a noun and a stative verb to Polynesians, although in the end their grammatical differences apparently led to few fundamental misunderstandings.

The same underlying dynamics of clausal juxtaposition and noun-predicate constructions in (33) extended to what Melville (Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 69, 71) clearly understood as full sentences, as is evident from the following examples:

Predicate constructions consisted of not only a subject plus an adjective or – rather – a stative verb, but also a subject plus a noun or adverbial phrase, as determined from passages in Omoo for Tahiti (Melville Reference Melville, Hayford, Parker and Tanselle1968b [1847]: 177, 178, 196, 254–255, 282):

The first and third examples in (35) again illustrate the principle of zero subject, that is the omission of an already expressed or implicit subject. On the other hand, the compound noun phrase such as tūtae ‘āuri, literally meaning ‘excrement’ and ‘iron’,⁷ in the first example above raises questions about its internal structure: Does the modifier follow the head noun like verbal modifiers (‘adjectives’) in noun phrases, as Melville suggested with another compound, “arva tee” or ‘ava tī ‘ti-root liquor’ (see the last example in (40) below)? Alternatively, should we stipulate the reverse order of modifying noun and head as indicated by a single other compound offered by Melville, pihi kanaka ‘fisherman’ (Reference Melville, Hayford, Parker and Tanselle1968b [1847]: 214), an apparent loan translation from English? These two examples, excluding the dubious entry of mikanale ta‘ata in the last utterance of (35) above, unfortunately do not provide sufficient data to resolve this question satisfactorily.

Other constructions reveal what speakers of Polynesian languages would interpret as action verbs in contrast to stative verbs (Melville Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 100, 129, 135, 207, 240, 245, 247):

Four examples in (36) again illustrate the use of pi mai mai/pi‘i mai ‘come here’ (< Hawaiian pi‘i ‘to go inland or overland, to go or walk up, to climb’ + mai directional particle indicating direction towards speaker), which Derek Bickerton and William H. Wilson (Reference Bickerton, Wilson and Gilbert1987: 64) as well as Roberts (Reference Roberts1995a: 8; Reference Roberts1995b 112) have considered as characteristically Pidgin Hawaiian in contrast to Hawaiian hele mai ‘come here’. Melville, however, offered the alternate form haere mai as well, as in (39) below.

Yes–no questions, of which Melville (Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 75, 266) provided two samples, apparently exhibited the same word-order pattern as statements; they presumably differed from statements only by intonation, and offer no further clues about their pattern. For so-called “Wh”-questions, Melville (Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 242) provided a single instance, which constitutes insufficient evidence to determine their structure beyond perhaps a claim that MPP speakers fronted the “Wh”-constituent to the beginning of the question:

The single instance of a “Wh”-question initially created a major puzzle for reconstitution, for which Wilson (personal communication, 14 June 2002) thought Melville's “Arware” to be no more than a rendition of Pidgin English ‘a [,] where he’ in place of a Hawaiian source. In that case, Wilson apparently picked up, probably quite unintentionally, an argument by Richard M. Fletcher (Reference Fletcher1964: 137) suggesting that “Here Melville obviously sought to duplicate the sounds someone half asleep might make in English when mumbling: ‘Where are you going, Tommo?’” Aside from the fact that Fletcher's claim to an English duplication half asleep is not obvious on either philological grounds or a textual basis, the interpretation by Fletcher and Wilson still raises the question of why Melville employed such an awkward spelling instead of Pidgin English, included elsewhere in his semi-autobiographical novels of the Pacific (see Drechsel Reference Drechsel, Barnum, Kelley and Sten2007b). Did he use his own spelling solely for dramatic purposes, or did he not even recognize this phrase as Pidgin English? In its place, I suggest an elegant and consistent alternative solution, which recognizes Melville's tendency to hear low central vowels with hypercorrected r-coloring (see the discussion leading up to footnote 9). When we disregard this secondary feature for the reconstitution of “Arware,” we are left with “Awae,” a close representation of ‘Auhea ‘where’ by the standards of English spelling conventions, which also matches the translation given by Melville. This example, however, still leaves unresolved the question of whether “Tommo” functioned as a nominative together with a zero subject for ‘you’ (as suggested by Melville's punctuation) or perhaps as a grammatical subject.

If MPP followed a predominant pronoun-predicate word order and applied this pattern even to questions, Melville (Reference Melville, Hayford, Parker and Tanselle1968b [1847]: 178) in Omoo furnished an example of questions with just the reverse order:

At first sight, this pattern would simply reflect substrate influences from Polynesian languages by which mikanele functioned as a verb ‘to be a missionary’, by means of the widespread categorial multifunctionality of MPP's lexicon. Then again, the source of the pattern of (38) was Anglophone Melville rather than a speaker of Tahitian or some other Polynesian language. The answer to this question, produced by a young Tahitian woman in Pidgin English, had a pronoun-predicate order: “Yes, me mikonaree” (Melville Reference Melville, Hayford, Parker and Tanselle1968b [1847]: 178) ‘Yes, I am a missionary’, as if this conversation displayed a reversal of sociolinguistic roles.

Broader considerations suggest an alternative explanation. First, Melville provided no systematic documentation of verb-initial word order in his samples with the exception of imperatives as in (39) below. In contrast to the answer by Miss Ideea, with its pronoun-predicate order, the inverted copular order in Melville's question could also have drawn on English grammar, which has made use of copula inversion, although infrequently. In that case, the predicate-pronoun order functioned as a syntactic marker of questions in place of merely a change in intonation patterns; conceivably, the reversed order even indicated focusing of the predicate by giving it syntactic prominence in what was an obvious expression of surprise by Melville, equivalent to ‘Are you a missionary?’ in English. The question in (38) therefore raises the problem of possible superstrate influence from English in Melville's MPP, whether directly or indirectly by way of Pidgin English; but it also illustrates an inadvertent convergence of Polynesian grammatical patterns with those of Pidgin English and English (see Drechsel Reference Drechsel, Barnum, Kelley and Sten2007b).

Additionally, MPP permitted imperatives, for which Melville (Reference Melville, Hayford, Parker and Tanselle1968b [1847]: 277, 310) offered two instances:

Imperatives could occur alone or with a vocative in a variable initial or final position. These examples also illustrate haere mai ‘come here/hither/towards the speaker’ as an alternative form to pi mai/pi‘i mai ‘come here’, considered by Bickerton and Wilson (Reference Bickerton, Wilson and Gilbert1987: 64) as well as by Roberts (Reference Roberts1995a: 8; Reference Roberts1995b: 112) as characteristically Pidgin Hawaiian. Non-Polynesian speakers of MPP likely lost any semantic difference between haere mai and pi mai/pi‘i mai in the form of ‘going inland, going or walking up, climbing’.

For both Nuku Hiva and Tahiti, Melville (Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 88, 102, 103; 1968b [1847]: 256, 272, 273) moreover offered a few “complex” utterances, including some with double arguments and an SVO word order. Like the first example in (35) above, the following examples reveal sequential linking (parataxis), as the American novelist already appeared to recognize when he used ellipsis or a dash between the sentences, suggesting a closer semantic relationship between parts:

The fourth example of (40), nuinui hanahana potato, presents somewhat of a structural enigma in that it is not clear how best to interpret the parts semantactically and how to relate them to each other. Does nuinui modify the following verb hanahana ‘to make’ or possibly the entire phrase hanahana potato ‘to make potato’ rather than the object, in which case we should understand it as an adverb? Melville did not provide sufficient comparative data or other clues to resolve this issue. Nor do his notes give any indications for tense or aspect, even when he recognized the past or the future in his translations. Also noteworthy in all of these examples is the consistent absence of articles. The only exception remains haole ‘stranger, foreigner, white man’, apparently used with the definite singular article ka as in Hawaiian when one addressed a person with a common noun (Pukui and Elbert Reference Pukui and Elbert1986: 106), although any such grammatical distinction was undoubtedly lost in the Pidgin. Ka haole here evidently recalls the use of the semantic opposite, te taio ‘the friend’, with a definite article in (7) of 4.4 above.

In addition, Melville provided descriptions in English of interactions in what apparently was said in MPP as a prime medium. By his own account, initial contacts came about rather arduously and with apprehension among Marquesans, although or perhaps because Melville (Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 68–69) and his American companion Toby enhanced their communications with gestures and pantomimic expressions, raising suspicion among the Hapa‘a people. A little tobacco, however, proved an effective gift to make them render their services to Melville (Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 71), and the Islanders could not hide their curiosity about their visitors (Melville Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 74–75). Regrettably, the beachcomber-novelist did not document any extended verbal exchanges with Marquesans or other Pacific Islanders.

In contrast to the linguistic data, the extralinguistic information by Melville remains disappointingly meager in that he gave but few samples in a broader sociocultural context of who spoke what under what other extralinguistic circumstances. In his first semi-autobiographical novel Typee, Melville presented most Pacific Islanders in the Marquesas as speaking in MPP and had native characters respond in Pidgin English only on a few occasions. A character illustrating this sociolinguistic behavior was Manu, a neutral intermediary and interpreter protected in his interactions with foreigners and known as a “tabooed kanaka” (Melville Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 74), who conversed in Pidgin English possibly because of his role as intermediary (Melville Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 139, 140, 241). Melville (Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 249) recognized another “tabooed kanaka” by the name of “Karakoee” or Kealaka‘i (?; William H. Wilson, personal communication, 24 July 2012), who too spoke broken English. He had originally come from O‘ahu (Hawaiian Islands), and together with Manu would eventually help the novelist escape from his Taipi captors. In contrast to Typee, Melville's sequel novel Omoo presented native characters in Tahiti of a year later as speaking predominantly in Pidgin English – without giving any obvious linguistic or sociohistorical reason.

Melville's attestations do not permit many conclusions about the use of MPP by any particular characters such as chiefs, neutral intermediaries or ‘tabooed kanaka’ (Melville Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 74, 139–140), or particular communities. The American beachcomber like many of those preceding and following him apparently did not recognize functional differences between the Pidgin and any vernacular language until late, if at all (see Melville Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 224). The lack of accompanying sociolinguistic information, especially in light of their proportional linguistic richness, then puts a serious constraint on Melville's attestations of MPP as historical data. Foremost, his literary use of the Pidgin in two semi-autobiographical novels of the Pacific raises questions of how the author applied information not only in his own interactions, but also in writing Typee and Omoo. These questions become all the more pressing in light of the fact that the novelist, like other writers of the colonial Pacific, took literary license for dramatic effects that have questioned the legitimacy of his semi-autobiographical novels as ethnographic-historical documents (Anderson Reference Anderson1966 [1939]). Much less did Melville reveal where or how he had learned the Islanders’ language or MPP in the first place.

Melville's semi-fictional, autobiographical novels of the Pacific would fall short of the required minimum of five years for the author's length of stay in the community, as determined by Philip Baker and Lise Winer (Reference Baker, Winer, Baker and Bruyn1999: 104–108) in their criteria of what constitutes a pidgin's or creole's trustworthy historical text. This issue thus raises as a prime sociolinguistic mystery the question of where Melville obtained his linguistic data on MPP and how he learned the Pidgin. By his own account, the novelist did not keep a journal of his travels to the Pacific, if there remain doubts that Melville wrote his novels solely from memory (Anderson Reference Anderson1966 [1939]: 70, 189). He left no clues of his own about how he had obtained information on MPP when he claimed familiarity with it (see, e.g., Melville Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 68) except to suggest that he had already learned it before arriving in the Marquesas by stating that “we had not dropped from the clouds upon them” (Melville Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 68–69). He did not even identify Manu or “Karakoee” as a source, although in his role as neutral intermediary and interpreter either might have served in this role. In other passages of Typee, the novelist shared only that he did not understand much of what a Taipi chief called Mehevi (Melville Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 80–81) or Manu (Melville Reference Melville, Hayford, Parker and Tanselle1968a [1846]: 137) said, in which case they may have spoken vernacular Marquesan rather than MPP.

Already during his times, the American novelist became the target of a devastating critique by the previously cited English pearl trader Edward Lucett (Reference Lucett1851: I.293–296), a contemporary who had apparently met Melville in Tahiti in 1842:

Supposedly, the beachcomber perverted the facts and possessed insufficient skills to write both Typee and Omoo; much less could he speak the Islanders’ language according to Lucett (Reference Lucett1851: I.294), who ironically was in no position to draw on much linguistic expertise of his own (see 6.4 above).

Melville's accounts of his experiences on Nuku Hiva and Tahiti indeed have not remained beyond doubt for their integrity as sources of historical-ethnographic information. When Charles Roberts Anderson (Reference Anderson1966 [1939]) reviewed Melville's novels of the Pacific for their historical-ethnographic accuracy, he found him to have spent a shorter time in the Marquesas than he had claimed – merely four weeks rather than four months. Within a period of less than a month, Melville could not, then, have made more than a limited acquaintance with island life. To compensate for missing first-hand experiences, Melville supposedly drew on earlier accounts of travels in the Pacific, without acknowledging them fully or at all. As primary sources for the beachcomber's novels reportedly served Georg H. von Langsdorff's Bemerkungen auf einer Reise um die Welt in den Jahren 1803 bis 1807 (Reference Langsdorff1812) or the English translation, Voyages and Travels in Various Parts of the World (Reference Langsdorff1813), David Porter's Journal of a Cruise Made to the Pacific Ocean (Reference Porter1822), and Charles S. Stewart's Visit to the South Seas (Reference Stewart1831); to a lesser extent, Melville perhaps also drew on William Ellis’ Polynesian Researches (Reference Ellis1831) and Edmund Fanning's Voyages round the World (Reference Fanning1833; see Anderson Reference Anderson1966 [1939]: 70–74, 96, 118–119, 190–192). Moreover, Melville supposedly used the same kind of reduced Polynesian as Nicholas’ Narrative of a Voyage to New Zealand (Reference Nicholas1817) and a few other, unidentified writers of the period, as already suggested by Clark (Reference Clark1977: 29–30).

A closer examination of these early sources, however, reveals no more than single, at best short, examples of MPP for these accounts except Nicholas’, and provides few, if any, clues for any sources of Melville's linguistic data for MPP above. Langsdorff's Bemerkungen contain sporadic observations about the Marquesans’ language and single words throughout his account of the South Pacific, plus a vocabulary of ten pages with a few phrases or full sentences at the end. While offering no relevant linguistic observations in the first volume, Porter enriched the second volume of his Journal with widely used native terms and intermittent comments on language use in the South Pacific, and included only one instance (see (16) in 5.2 above). Stewart's Visit displays even greater generosity with syntactic data by regularly offering compound phrases and short sentences in Polynesian languages, as do Ellis’ Polynesian Researches throughout multiple volumes; but as missionaries who aimed at learning vernaculars, Stewart and Ellis were primarily concerned with presenting Polynesian languages in their vernacular forms, and did not include many, if any, data on MPP. Fanning's Voyages, too, offered only a few single words and two phrases along with an observation on language use in the South Pacific. Counter to Anderson's claim, these sources could not have provided the fairly rich data for MPP in (32) through (40) above.

A comparison of Nicholas’ samples with Melville's, especially their longer constructions in (18) in 5.3 and (40) in 6.5 as best candidates to reveal any traces for copying, provides no better evidence. Most of Nicholas’ examples consist of short, reduplicative phrases from his two-volume Narrative (1817), and include only a few longer constructions, none of which shows any immediate resemblance to Melville's examples other than overall grammatical patterns (see (17) and (18) in 5.3 above). Archibald Campbell's Voyage round the World from 1806 to 1812 (Reference Campbell1816) or John Slade's Old Slade; or, Fifteen Years Adventures of a Sailor (Reference Slade and Denison1844) with their many grammatical inconsistencies (see (15) and (27) in 5.1 and 5.11 above) do not fare any better as sources for the American novelist, except perhaps for the vocabulary (Campbell Reference Campbell1816: 229–251). A review of other early accounts of the South Pacific, examined for this purpose so far, has produced no obvious resemblances to Melville's linguistic data beyond general linguistic patterns to suggest any copying. Significantly, none of the earlier sources of MPP except that by Savage (Reference Savage1807) even offered as much of a substantial body of homogeneous data at the time when Melville wrote his two semi-autobiographical novels, and Savage's data too are distinct from Melville's. By all indications, early accounts thus could not have served as primary models for the descriptions of MPP by the American beachcomber-novelist, even if Melville frequently used the same idiosyncratic orthography for native terms and names as Porter (Anderson Reference Anderson1966 [1939]: 98).

Ignoring any borrowings from earlier accounts of the Pacific as well as hyperbolic embellishments and imaginary exaggerations in truly Romantic literary tradition, Anderson (Reference Anderson1966 [1939]: espec. 117–178, 237–323) already found Melville's other descriptions to be unexpectedly accurate and consistent with independent information, including modern ethnographic data. In checking Melville's accounts against historical-ethnographic descriptions by other authors, Anderson (Reference Anderson1966 [1939]: 156–162) also identified some of the given native terms with modern equivalencies, but erred in a few instances. The most obvious one is “hoolah-hoolah” for ‘festival grounds’ as supposedly equivalent to Marquesan tohua ‘sol, place, terre, terrain, place publique’ (Dordillon Reference Dordillon1931–1932: 394) or ‘ground, floor, square, soil, earth’ rather than as a reduplication of hula or related forms in various Polynesian languages or even hulahula ‘feast’ (see Drechsel Reference Drechsel2014). Anderson apparently did not consider the possibility that what Melville had learned on Nuku Hiva was a contact medium rather than vernacular Marquesan.

Nonetheless, Melville's use of Polynesian linguistic materials has remained under a continuous shroud of doubt among students of literature, who for all their many misgivings have not been above suspicion for their own methodological or theoretical foundations. When examining Melville's use of the Marquesan language in his early novels, Fletcher (Reference Fletcher1964: 135) argued that Melville “consciously and consistently rephras[ed] or even purposely misus[ed] the Marquesan vocabulary” for poetic and especially satirical license. Fletcher (Reference Fletcher1964: 136–138), however, provided no supportive evidence for his negative evaluation and intermittently found the novelist's linguistic information on Marquesan actually to be quite accurate in contradiction to his very own claim. Others recognizing Polynesian words in Melville's novels such as Shigeru Maeno and Kaneaki Inazumi (Reference Maeno and Inazumi1984: 514–515) classified these terms into geographic groupings only to dismiss larger constructions unceremoniously in substratophile terms: “Polynesian passages in Melville's works are not Polynesian, because they have the same word orders as English.” A recent doctoral dissertation entitled “A Critical Dictionary of Herman Melville's Polynesian Terms” (Christodoulou Reference Christodoulou2006) has placed greater confidence in Melville's knowledge of “Polynesian” when reviewing his use of primarily personal, place, and other names of Polynesian origin. Sadly, this study lacks a solid foundation in comparative-contrastive linguistics, fails to distinguish carefully a phonetic or phonemic transcription from Melville's Anglicized renditions, and does not attend to problems of interpretation for names or tackle questions of morphosyntactic interpretation for Polynesian languages. Much less has the author taken the next methodological-theoretical steps by addressing problems of philological reconstitution or raising the possibility of pidginization for Melville's data, whence this dictionary offers little insight for the present review. All these studies have characteristically lacked an expanded comparative basis for contrasting Melville's Polynesian data with historical linguistic evidence or a solid theoretical-methodological foundation (including a model of pidginization) for the interpretation of such comparisons.

A solitary exception in this critical discussion remains the already mentioned essay “In Search of Beach-la-Mar” by Clark (Reference Clark1977: 29–30), who has thought Melville's linguistic record to correspond so nicely with other historical documentation of the period and the area as to make an accurate testimony: “[T]he general nature of Melville's jargon agrees so well with Nicholas's [Narrative of 1817] and with occasional examples from other writers of the period, that I believe it is probably a faithful record of the type of communication system prevailing in the situation.” Clark (Reference Clark1977: 31) cited as his only reservation Melville's “promiscuous” mixing of Marquesan, Tahitian, and Hawaiian in contrast to Nicholas’ primary reliance on Māori as the source of lexicon. A mixed lexicon per se would, however, not cast any doubts on the validity of Melville's linguistic attestations, as there are indications for such in earlier historical sources and lexical mixture occurred in other jargons and pidgins (Mühlhäusler Reference Mühlhäusler1997: 135–136, 160–162, 175–185). In a follow-up essay by the same title, Clark (Reference Clark1979: 29, 33–34) eventually drew on selected samples from Melville's writings for illustration, apparently without further reservation about his mixed lexicon. These circumstances therefore warrant greater confidence in Melville's linguistic attestations and observations, including sociolinguistic considerations not even taken into account by Clark. They also confirm the current approach of comparing Melville's attestations with not only a broad basis of Polynesian and possibly other Pacific bodies of linguistic data, but also historical records of MPP beyond those by Nicholas (Reference Nicholas1817) for a maximum database in interpreting and reconstituting the novelist's recordings.

Once reconstituted, Melville's linguistic data prove surprisingly consistent not only internally, i.e. among each other, but also in comparison with earlier and recent recordings of MPP. His attestations differ from other transcriptions such as those on Pidgin Hawaiian (Roberts Reference Roberts1995a; Reference Roberts1995b) only in that these do not provide any evidence for the sentence-initial negative except in the antonym of ‘a‘ole motaki ‘bad’ (see (40) above). Apparently, the sentence-initial position of the negative was too odd for Melville to incorporate in his writings, a fact that leaves open some questions about the representativeness of his data, despite their overall high quality. That Melville indeed did not record a sentence-initial negative in accordance with Pidgin Hawaiian grammar suggests that he did not perceive negative constructions correctly or that he possibly transcribed his samples from recollection, which in either case resulted likely from interference from his first language, English. The high uniformity of Melville's with other data, however, lends strong support to the suggestion that he did not obtain such a consistent body of data by borrowing from multiple historical sources. If we were still to insist on some other source for Melville's data, these would by all indications have originated from a single person who obviously had a good command of the Pidgin, notwithstanding any remaining idiosyncrasies in spelling.

Without contrary evidence for matching samples in earlier sources, we must first assume Melville's sentence constructions to be original ones, that is the novelist's own. This conclusion does not mean that he invented them. Quite on the contrary, the overall structural consistency of Melville's sentences with independent attestations makes it unlikely that he simply fabricated these. Had he simply conceived them, his rendering of MPP would have revealed little or no structure with all utterances made up according to his imagination; alternatively, they would have exposed obvious English grammatical patterns in place of, for example the postnominal positions of modifiers, on account of him not knowing any other language for a model (see Siegel Reference Siegel2008: 11–42). These circumstances are indeed what we would expect in any initial situation of second-language learning. Melville's attestations, then, appear quite authentic samples that originated from his own travel experiences in the Pacific. Notably, these samples except perhaps for a few constructions in (40) do not even reveal any hyperbolic language or other extraordinary features of expression, although the novelist obviously used these instances for literary embellishment and dramatic purposes as part of his English narrative.

These findings need not contradict observations that Melville reportedly had “a bad ear” and had little interest in literal ethnohistory, as suggested by Paul Lyons (personal communication, 29 May 2002). If MPP was the primary means used by newcomers to communicate with Polynesians, but there existed few earlier, easily accessible descriptions of it as sources of information, there simply remained no choice for Melville but to transcribe it on his own, short of making up the indigenous peoples’ speech. Although different from anything that Melville had encountered before, MPP proved phonologically transparent enough such that even a stranger with “a bad ear” could record it without many difficulties or misrepresentations beyond standard Anglophone spellings.

Significantly, Melville's recordings give us valuable clues about his own speech. Comparative evidence suggests, for instance that the literary sailor regularly heard [r], probably retroflexed, after the low central vowel in Polynesian words where there was none in the speech by Pacific Islanders, as is evident in “arva” for ‘ava ‘kava’, “Happar” for Hapa‘a ‘Hapa‘a people’, “mortarkee” for motaki ‘good’, and “puarkee” for puaka ‘pig’ among other instances above. Perhaps the most telling example is the English/Spanish loan “portarto” (Melville Reference Melville, Hayford, Parker and Tanselle1968b [1847]: 256), reconstituted in (40) above as potato rather than as poteito ‘potato’. Melville apparently perceived these words as equivalent to instances in need of hypercorrection, as has been characteristic of speakers from eastern New England and New York with a non-rhotic accent and a broad “a” or the so-called Boston accent (see Labov Reference Labov1972b). Melville had indeed been born and raised in New York; he had also lived many years in the city and identified with it, and his family had roots in Boston (Evelev Reference Evelev2006 and Kelley Reference Kelley1996). This biographical information has in turn helped in the reconstitution of still other, problematic examples such as “Arware” as ‘Auhea ‘where’ in the third instance of (37) above, which indicates that Melville extended the tendency to hear low central vowels as r-colored to diphthongs with the low central vowel. In this case, r-coloring included lip rounding (Albert Schütz, personal communication, 18 October Reference Schwaller2012), consistent with the reconstitution of “Arware” as ‘Auhea, but remained a secondary feature of articulation in other instances without a preceding or subsequent rounded vowel. All in all, the phonological feature of r-coloring lends a mark of authenticity to Melville's recordings, inasmuch as it is fully consistent with the available sociohistorical information and thus weakens suggestions for simple copying or arbitrariness in his transcription.⁹

The question still remains as to whether Melville could have learned enough MPP within a month to offer several sentences of structurally consistent data and a small vocabulary of Polynesian words. Such an inference might indeed seem unreasonable, if one concludes that the novelist learned the Pidgin while in the Marquesas and subsequently in the Societies. Alternatively, this option gains weight when we take into consideration the fact that MPP functioned as an interlingual medium not only on Pacific islands, but also on European and American ships with Polynesian crews serving as replacements for sailors who had jumped ship. This very situation had already applied to Adelbert von Chamisso (Reference Chamisso and Hitzig1856a [1835]: 157–158), who in 1816 had learned fundamentals of Pidgin Hawaiian on board the Russian vessel Rurik and at that not even from a native Hawaiian, but from John Elliot de Castro, a passenger of English-Portuguese ancestry and a former pearl trader on O‘ahu. Likewise, the missionary wife Lucia Ruggles Holman had learned her first sentences of Pidgin Hawaiian on board the brig Thaddeus in 1820 (Holman Reference Holman1931 [1820–1821]: 13). Melville may well have gone through the same experience on a whaling ship before arriving in the Marquesas, in which case he could have had a considerably longer exposure to MPP than merely four weeks.

Support for the claim that Melville learned the Pidgin on board comes from internal linguistic evidence, specifically several entries that etymologically appear more Hawaiian than either Marquesan, Tahitian, or even Māori: ka haole ‘stranger, foreigner, white’, kanaka ‘human, man, native person/people, Pacific Islander’, koko ‘sweet potato’, make ‘to die, dead, sick, ill’, mikanele ‘missionary’, pihi ‘fish’ (< English), and pū‘ā‘ā ‘to flee, to go’. Most distinctive is the Hawaiian sound correspondence of the velar stop k where all other Polynesian languages have retained t from Proto-Polynesian *t (see Biggs Reference Biggs and Sebeok1971: 480–481), as is also evident in reconstitutions with the definite singular article ka. Hawaiian could also have served as the source of much of Melville's other vocabulary except ‘aita ‘no, not’, fe‘efe‘e ‘elephantiasis’, Ferani ‘French people’, maiore ‘breadfruit’, motaki ‘good’, puaka ‘pig, hog’, and tūtae ‘āuri ‘nonbeliever’ because of conspicuous phonological or lexical-etymological differences (William H. Wilson, personal communication, 14 June 2002; see also Drechsel Reference Drechsel2014). By the vocabulary's overall composition, Melville apparently learned MPP from either a Hawaiian crew member or somebody who had spoken a Hawaiian-based variety.

Alternatively, these linguistic findings may yet indicate that the beachcomber-novelist picked up much of the Pidgin during his stay on the Hawaiian Islands following his visit to the Marquesas and Tahiti, as suggested by the Melville scholar Paul Lyons (personal communication, 29 May 2002).¹⁰ In that case, we would have to conclude that Melville embellished his novels Typee and Omoo with Pidgin Hawaiian next to some vocabulary from South Pacific languages – presumably the source of his “promiscuous” mixing of Marquesan, Tahitian, and Hawaiian (Clark Reference Clark1977: 31) – rather than actually having documented most or all of the MPP attestations in situ in the Marquesas or the Society Islands. Still, in comparison to any translation into Pidgin English, any such Hawaiianized version of MPP would have constituted a comparatively minor distortion of historical-sociolinguistic reality in light of contemporaneous independent evidence for local varieties on the Marquesas and in the Society Islands. The reason simply is that pidginized varieties of Marquesan and Tahitian closely resembled Pidgin Hawaiian in grammar and vocabulary. Melville's recordings would therefore require some reinterpretation in Marquesan and Tahitian terms without losing much of their overall historical interest.

6.6 Max Radiguet

As secretary to staff headquarters of the French colonial expansion in the Pacific on the frigate La Reine Blanche from 1842 until 1845, Max Radiguet made ethnographic and other observations on Marquesan life in Les derniers sauvages (Reference Radiguet1929 [1860]). His view on the linguistics of the Marquesan Islands – apparently inspired by a friend called Monsieur Rohr, who had spent several years on Nuku Hiva as an artillery officer and had studied the local language and customs (Radiguet Reference Radiguet1929 [1860]: 161, fn. 1) – by itself suggests a reduced form of Marquesan:

Much of Radiguet's book includes personal conversations in French, which, interspersed with occasional Polynesian words, represent evident translations of native speech. Occasionally, the author provided larger-than-single-word samples of Marquesan in both vernacular and reduced forms, the latter of which usually appeared in exchanges with foreigners as evident in the following instances (Radiguet Reference Radiguet1929 [1860]: 32, 34, 177, 182, 220):

We can reconstitute Radiguet's short examples of reduced Marquesan as follows:

Most interestingly, Radiguet's samples confirm not only the otherwise well-attested pi mai/pi‘i mai ‘to come’ for the Marquesas (as already recognized by Roberts [Reference Roberts1995a: 20]), but the use of synonyms of Marquesan-derived motaki and Tahitian-derived maita‘i ‘good’ side by side for the enhancement of mutual understanding. Radiguet also documented apparent English loans for ‘money’ and ‘tobacco’, probably introduced by English-speaking sailors and traders such as the Americans Joseph Ingraham and David Porter. Because Radiguet and his source were speakers of French, these English loanwords are less suspect than if they had appeared in a document by an English or American observer. For their syntactic pattern, Radiguet's samples also exhibit consistency in word order and predicate constructions with other examples of MPP.

Radiguet (Reference Radiguet1929 [1860]: 148, 192) further included text materials of two songs among which the second contains two grammatically reduced constructions interpretable as part of the Pidgin (Gabriele H. Cablitz, personal communication, 27 August 2009):

These two examples reveal a basic pattern of a question word equivalent to English “Wh”-question words and a noun, which would have been intelligible to speakers of Polynesian languages as well as European languages. Radiguet (Reference Radiguet1929 [1860]: 192), however, provided no explanation for the great discrepancies between his interpretation and the word-for-word translations; nor did he offer any indication or other clues as to why the singer switched from vernacular Marquesan to MPP in this circumstance.

Though syntactically rather meager, Radiguet's examples provide perhaps the most explicit association of MPP with a colonial navy, that of France, and its representational institutions such as a Catholic mission and a satellite court. The French official also confirmed the use of the Pidgin by women on Nuku Hiva, but thought they spoke it only sparingly and with blank expressions.

6.7 Louis-Théodore Bérard

For 1850, a French settler and former naval officer by the name of Louis-Théodore Bérard, often confused with Admiral Auguste Bérard of a few years earlier, offered a description of his campaign to New Caledonia with incidental attestations of what would at first sight appear to be MPP:

Bérard (Reference Bérard1854: 136) also included relevant observations by an oarsman named Hervé:

These attestations are reconstitutable as follows:

These three examples are consistent with earlier recordings of MPP for their grammatical pattern, including the SVO word order. Exceptions apply only to the proper names of “Bonéone” and “Bonérate” in French transcription with the final mute “e” (variably articulated as a weak schwa), the ethnonym “Belep,” and the short form for aliki ‘chief’, which did not conform to the standard Polynesian syllable pattern with a final vowel.

Throughout the second half of his account, Bérard (Reference Bérard1854: 103, 108, 109, 110, 114, 120, 127, 149, 160, 180) cited not only the usual alik(i) ‘chief’, kaikai ‘to eat, food’, and kakino ‘bad, immoral’, but also other words and short phrases of Polynesian provenance, most of which have corresponding forms in West Uvean on nearby ‘Uvea Atoll (“Ouvéa”) (see Hollyman Reference Hollyman1987):

These words and short phrases would be of little interest as evidence for MPP if Bérard's listings had drawn from a single rather than two or more Polynesian languages or if they applied to a Polynesian linguistic environment. In actuality, many of these entries reflect origins in other languages: epikopo (< Latin), kakino (< Māori, Marquesan?), taio (< Tahitian), and kanaka (< Hawaiian; see Crowley Reference Crowley1990: 59). For West Uvean, K. J. Hollyman (Reference Hollyman1976: 37–40; Reference Hollyman1987) identified an even larger list of not only six Polynesianized European loans, but also twenty-three terms of diverse Polynesian origins. Terry Crowley (Reference Crowley1990: 58) has since added two loans to the list, which he tellingly considers the result of contact with Polynesian languages to the east. Significantly, Bérard's samples with repeated references to locations in northern New Caledonia – the Arama valley, the Belep people, and Kumak (“Coumac”) – also applied to a non-Polynesian sociolinguistic environment. However, he did not disclose any information about the identity of his sources, presumably Bonéone's slave, a fair-haired child identified as “Blondin,” and Bérard's guards. These lexical cum sociolinguistic indications would therefore favor an interpretation of the citations as MPP, extendable also to Bérard's extra lexical data. They discourage interpreting them as the West Uvean vernacular on nearby ‘Uvea Atoll, although these words undoubtedly remained intelligible to speakers of Uvean (see Hollyman Reference Hollyman1987).

Even so, Bérard's linguistic data raise some fundamental questions of analysis. Whereas the French settler expectedly failed to recognize Chief Bonérate's native language (Bérard Reference Bérard1854: 115), presumably some local language of northern New Caledonia, surprisingly he and his cohort did not, by his own account, do any better with the Māori upon their arrival in New Zealand in early 1851:

With some knowledge of MPP based substantially on the vocabulary of related Polynesian languages (including Māori cognates), Bérard should have had little difficulty recognizing at least portions of his Māori hosts’ speech, whether reduced or not. This inference would apply, even if in the most unlikely case we were to argue that Bérard's Polynesian medium had developed locally on New Caledonia and without any input from MPP, for both shared obvious structural and lexical parallels. That the French naval officer and settler still could not establish basic communication with the Māori raises troublesome questions about whether the above linguistic data were really his own or whether he possibly projected them from eastern Polynesia onto New Caledonia without any complementary understanding.

Manifestly, Bérard's attestations for an exclusively Polynesian-based medium on New Caledonia have not so far received any independent documentary confirmation by other contemporary observers in the western Pacific except New Zealand to the south. Instead, for 1842, the Australian trader Andrew Cheyne (Reference Cheyne and Shineberg1971: 98) had already recorded what evidently was another contact medium as part of the sandalwood trade on nearby Lifu north of New Caledonia:

Aside from obvious Polynesian influences such as aliki ‘chief’, Jean Guiart (in Cheyne Reference Cheyne and Shineberg1971: 98, fn. 16) has interpreted this quote as follows:

Cheyne's quote has also attracted the attention of Terry Crowley (Reference Crowley1990: 56–57), who found it of interest not only for its lexical mixture of Polynesian and two Southern Oceanic languages, but for the Pidgin English translations of the chief's speech:

The contact medium described by Cheyne, then, was what Crowley (Reference Crowley1990: 52–60) interpreted as a partly Melanesian-based variety of South Seas Jargon.

At no point did the Australian trader, however, make reference to any form of English spoken on the islands of the southwestern Pacific that he visited. Wherever he dropped anchor, he depended on local interpreters or his new linguistic skills, as was the case on ‘Uvea Atoll (“Ouvéa”) and Pohnpei (Cheyne Reference Cheyne and Shineberg1971: 108, 110, 120, 157, 159). Cheyne (Reference Cheyne and Shineberg1971: 175–179) also provided a list of Pohnpeian words, which J. L. Fischer has interpreted as part of the local foreigner talk or a kind of Pohnpeian pidgin and which shows no immediately obvious Polynesian influences. The apparent Pidgin English equivalencies for the future and plural markers would therefore have been little more than English translations, and Crowley's suggestions about Cheyne's use of English remain speculations that are short of supporting historical documentation. However, they illustrate a recurrent association of Pidgin English with an indigenous contact medium and a premature interpretation of early language contact in Europhile terms. Even if we take into consideration the earlier experiences by Slade (Reference Slade and Denison1844: 37–39, 42–44) farther east on ‘Uvea Island in 1830, with its native population speaking a language closely related to West Uvean of ‘Uvea Atoll (“Ouvéa”) off New Caledonia (see the discussion following (27) in 5.11 above), the local contact medium was so different from MPP as to have been mutually unintelligible with it.

Cheyne's description matches the historical picture of a growing number of Melanesian laborers who not only were engaged in the trade of sandalwood and trepang (bêche de mer or sea slug/cucumber) as well as in whaling, but also came to serve as crew members on foreign sailing ships northeast of Australia around the mid-nineteenth century (Crowley Reference Crowley1990: 62–64). In contrast, Bérnard's account does not quite match the historical scenery of the period and requires an explanation. It is as if the French naval officer overlooked non-Polynesian influences in the medium of New Caledonia, but then failed to recognize Polynesian resemblances in New Zealand. Without independent confirmation from contemporary observers, his linguistic attestations remain in doubt as reliable evidence for MPP, at least in New Caledonia, in an exclusively Polynesian format; they indeed leave unresolved nagging questions about a substantial non-Polynesian portion missing in what evidently was a jargon or pidgin quite different from MPP. As long as we cannot resolve this apparent discrepancy between structure and use, Bérard's data remain defective evidence of the Pidgin, to be excluded from the overall analysis below (Chapter 7) and the accompanying vocabulary (Drechsel Reference Drechsel2014).

6.8 James F. Munger

For October 1851, an American sailor by the name of James F. Munger (Reference Munger1852: 22, 38, 67) reported that “Sandwich Islanders” taught him “Hawaiian,” while serving on a whaler in the Pacific and the Arctic in place of an American crew who had jumped ship in the Hawaiian Islands:

Regrettably, Munger did not include any actual samples except the words of “Poa” or poi for ‘a sort of batter’ and “Mamalu” or māmalu for ‘umbrella’. However, his reference to “common conversation” and his focus on “names of nearly everything in common use” among the Hawaiian crew (emphasis added) leave little doubt that he was speaking of Pidgin Hawaiian rather than vernacular Hawaiian, as Roberts (Reference Roberts1995a: 19) already recognized. This interpretation receives support from the very fact that Munger had learned much of the language within six months. On the other hand, he did not have a high opinion of Hawaiians – “Our Kanakas also left us here [in Hilo], and this, we were not sorry for; they are not the most agreeable ship mates in the world” (Munger Reference Munger1852: 39) – even if he recognized their lot to be similar to that of Native Americans and he, too, considered jumping ship in the Hawaiian Islands (Munger Reference Munger1852: 57, 60, 61). The American sailor would thus have had to demonstrate a more open mind toward Hawaiians in order to learn within six months basics of what must still have been an “exotic” language to him; but during that period, he could certainly have picked up essentials of Pidgin Hawaiian on board like Chamisso, Holman, and Melville.

6.9 Benjamin Boyd

During the second half of the same year, the Scottish-Australian entrepreneur Benjamin Boyd returned from a futile excursion to California's gold rush via the Hawaiian Islands to the South Pacific, sailing with a crew of mostly Pacific Islanders on his fast topsail schooner Wanderer – eventually to run aground in a gale off New South Wales, Australia. The captain of the accompanying supply ship, John Webster, then expressed surprise when he encountered a few Pacific Islanders, among them a Hawaiian guide and some Gilbert Islanders, speaking fairly fluent or at least good broken English, which they had acquired either on a visit to Great Britain or from the trade of trepang with local Europeans (Webster Reference Webster1863: 10, 51–52). In other circumstances, Webster (Reference Webster1863: 3) “experienced considerable difficulty in getting our native crew to understand the points of the compass,” presumably in part because of difficulties in communication, and depended on one of his own Gilbertese crew members, Timarare (“Timmararare”) or in short Tim, a native of Banaba, for interpretations with native peoples of Micronesia (Webster Reference Webster1863: 26–27, 35).

The captain did not specify what medium the crew of Pacific Islanders used among themselves or with the ship's officers except in one dramatic instance. When the ship was already out of the Hawaiian Islands, the crew unexpectedly found Timarare overboard, but shortly recovered him in a tub thrown after him:

Timarare apparently spoke MPP with several English words or Pidgin English enhanced by some Polynesian vocabulary – Te Manu clearly derives from the noun phrases for ‘the bird’ in several Polynesian languages, and mate from equivalent words for ‘to die’ in these same languages. What proves socio-linguistically intriguing is not only the traumatic circumstances from which Timarare's response arose, but also the fact that the victim was Micronesian instead of Polynesian and spoke Gilbertese (Kiribati) as his first language with a corresponding word for ‘to die’, even if with different terms for ‘bird’: te man ni kiba, literally ‘a/the animal for/in flight’ plus some seventy distinct terms for different kinds and species of birds (Sabatier Reference Sabatier and Oliva1971: 188, 219, 241, 272, 367; see Trussel Reference Trussel2003).

At first sight, Webster would seem to have been witness to a process by which a Polynesian medium gradually ceded to English, although in reduced form. Still, the incidental Polynesian vocabulary, combined with the extraordinary circumstances of Timarare's response and the fact that the majority of the Wanderer crew consisted of Pacific Islanders, raises the question of whether the captain gave an accurate rendition of the saved sailor's answer. Did Webster present an Anglicized rendition or stereotypical Pidgin English in place of MPP, like other writers before him, such as Dana in his Two Years before the Mast (Reference Dana1911 [1840]) and Melville in Omoo (Reference Melville, Hayford, Parker and Tanselle1968b [1847])? In that case, most of Timarare's utterances now remain lost and unreconstitutable.

6.10 Artemas Bishop

When in the mid-nineteenth century the Reverend Artemas Bishop published the first edition of what subsequently came to be known as the Hawaiian Phrase Book (Reference Bishop1854) in English, he recognized the existence of a “corrupted” form of Hawaiian. He denied any regularity of grammar or aesthetics, but offered no descriptive details in support of his argument:

What Bishop took as an obvious linguistic aberration served as a justification for his publication with exemplary English and corresponding phrases in vernacular Hawaiian. English-speaking visitors could thus learn “proper” or vernacular Hawaiian and also promote its usage in formal education, before English would become the new linguistic standard in the Hawaiian Islands.

6.11 Julien Viaud alias Pierre Loti

During his visit to Tahiti in 1872, the French naval officer Julien Viaud came to understand some of the sociolinguistic realities of Tahiti, an account of which we find in his autobiographical novel Le mariage de Loti, written under the pen name of Pierre Loti (Reference Loti1880: 415–416, 879). Loti's short-term residency in the Society Islands of less than two months (see Wright and Eleanor Frierson in Loti Reference Loti, Frierson and Frierson1976: xiv), however, raises major doubts that he learned to appreciate the grammatical complexities of Tahitian. Significantly, Loti observed that “Depuis longtemps je pouvais couramment parler le tahitien de la plage qui est au tahitien pur ce que le petit-nègre est au français” (Reference Loti1880: 415). That is, “For a long while I had been able to speak fluently the Tahitian of the beach, which is to pure Tahitian what petit-nègre [specifically a form of French pidgin associated with Blacks] is to French” (Loti Reference Loti, Frierson and Frierson1976: 89). The French officer did not specify how he had learned “Beach Tahitian” or Pidgin Tahitian; however, his account of amorous encounters with several Tahitian women, integrated in the single character of Rarahu in a historically otherwise accurate account (Frierson in Loti Reference Loti, Frierson and Frierson1976: xv, xvii), assigns a major role to young women. Pidgin Tahitian evidently was in use on the shores of Pape‘ete, among others, by young Tahitian prostitutes in contact with foreign visitors:

On his subsequent visit to the Hawaiian Islands and specifically Honolulu, use of Pidgin Tahitian also made Loti intelligible to Hawaiians with little difficulty, a fact that helped him feel less homesick for Tahiti:

It is noteworthy that Loti's reference to Pidgin Tahitian caught the attention of Hugo Schuchardt in his discussion of the South Seas whaler jargon as part of Beach-la-Mar:

The eminent creolist of the period noticed Polynesian-derived words such as pau, taio, and kaikai as well as foreign loans with a Polynesian pronunciation such as kaukau, pikenene, and Wiwi, most of which he had probably obtained from Chamisso's account (see (19) in 5.4 above). To my knowledge, Schuchardt did not pursue his inquiry about Pidgin Tahitian; nor did he consider the reverse possibility that the English-based whaler jargon emerged out of MPP, say, by relexification. Instead, T. L. Markey, translator and editor of a recent collection of Schuchardt papers, has mistaken Loti's reference to “Beach Tahitian” as “a sort of Beach-la-Mar, presumably Tahitian, French, and English. In any event, there is no trace of such a pidgin or creole in Tahiti today” (in Schuchardt Reference Schuchardt1979: 111, fn. 17). QED!

6.12 Herbert H. Gowen

In the late 1880s, people in the Hawaiian Islands were still speaking Hawaiian on a broad scale and required interpreters for speakers of English, according to the Anglican missionary Herbert H. Gowen (Reference Gowen1892: 8, 17–18, 34, 36). For early September 1887, Gowen eventually cited an incident in which a major storm forced the captain of an interisland vessel sailing from Kō-‘ele on Lāna‘i to Lahaina on Maui to return to Lāna‘i. The captain offered an explanation in Pidgin Hawaiian:

As already indicated by Roberts’ modified translation, the above example again illustrates sequential linking and does so even twice with a comparative and conditional. Gowen's punctuation between sentences in the form of commas rather than periods or semicolons suggests that he, too, was already aware of the semantic connections between the constituent phrases, even if he did not spell out their relationship in these or comparable terms.

Back on Lāna‘i, the vessel was “obliged to land at the very extremity of the island, where I found myself among some of the most primitive islanders of the group, ignorant of a word of English. ‘Makemake lio’ (want horse) was, however, sufficient to procure myself an animal to ride” (Gowen Reference Gowen1892: 132). This example again exemplifies MPP:

Gowen left the subjects of both instances with action verbs unspecified (zero subjects), which makes the confirmation of word order impossible.

Although Lāna‘i Islanders reportedly were “ignorant of a word of English,” Gowen had one woman respond in English or Pidgin English, the latter of which, however, received no mention in any other form throughout his book:

Gowen's final quote by the Hawaiian woman would reflect a genuine rendition of her response, although perhaps still as an exception to his observation on English on Lāna‘i because of its nature as a neighboring rural island.

Based on currently available historical information, MPP did not carry on in New Zealand beyond the first half of the nineteenth century, and faced an uncertain fate in the Marquesas, even though declarations of demise remain premature without better evidence. The Pidgin still endured in the Hawaiian Islands on multiethnic plantations until the end of the nineteenth century and even into the early twentieth (Roberts Reference Roberts1995a: 27–45; Reference Roberts2005: 107–123). MPP also survived in its original homeland of Tahiti in the form of Parau Tinitō or Chinese Pidgin Tahitian, a contact medium between Chinese storekeepers and their Polynesian customers in the late nineteenth century (Mühlhäusler et al. Reference Mühlhäusler, Dutton, Tryon, Wurm, Wurm, Mühlhäusler and Tryon1996: 449b–450a) and as recently as the 1970s (Wilson, personal communication, 14 June 2002). Parau Tinitō apparently consisted of a largely Tahitian lexicon and “a simplified and modified morpho-syntax, from which the standard Tahitian models are largely absent” (Mühlhäusler et al. Reference Mühlhäusler, Dutton, Tryon, Wurm, Wurm, Mühlhäusler and Tryon1996: 450a), and was much the same as Pidgin Hawaiian according to William H. Wilson (personal communication, 29 April 2012), who still witnessed both. How Parau Tinitō precisely related to MPP linguistically and sociohistorically remains a question for further investigation, as better sociolinguistic documentation of the first becomes available.