A new mixed language in the news

« previous post | next post »

Lately we've seen a number of hair-tearing Language Log posts (including a couple of mine) about bad linguistic pseudo-hemi-demi-quasi-science getting into major science journals and the popular press.  But sometimes the news media get it right, and here's one example: thanks to effective publicizing by the Linguistic Society of America, a new article by Carmel O'Shannessy, who has been observing the emergence of a new mixed language in Australia for many years, is being widely reported nationally and internationally, for instance here and here.

Back in 2004 I gave a talk on `The birth of bilingual mixed languages' at the annual meeting of the American Association for the Advancement of Science.  A prominent linguist in the audience protested during the comment period that I had no actual evidence that such languages actually existed and were learnable, since my evidence came from historical situations.   (I still think my evidence was solid, but I'm pretty sure I didn't convince the doubter. )   Carmel's research (which wasn't yet published in 2004) would have been an effective response to that objection: she shows that young children have been participating in the creation of Light Warlpiri, and she shows conclusively that the language is being learned by younger children.

Share:



42 Comments »

  1. mike said,

    June 18, 2013 @ 11:40 am

    Question about this:

    "Another innovation of the newfound language is a word form that refers to both the present and past time, but not the future. Light Warlpiri speakers created a new form, such as "yu-m," which means "you" in the present and past time, but not the future. In other words, this verbal auxiliary refers to the "non-future" time, which is a word form that does not exist in English, Kriol or traditional Warlpiri, O'Shannessy said."

    How does that compare to a construct like "I have been seeing someone" –about dating–that (in my interpretation) includes both the past and the present?

  2. NW said,

    June 18, 2013 @ 12:06 pm

    English makes a single tense distinction on its finite verb forms: past (-ed or strong) and present – which can be used as future too, so call it non-past. In addition it has a syntactic combination, the perfect, that typically means past extending into present. So it can indicate non-future though it hasn't got a non-future basic marking.

    Australian languages typically have past/present/future verb forms, or past/non-past, or in some cases future/non-future. As with English, this is separate from whatever other means they have of making the aspect more explicit. Light Warlpiri has apparently evolved a different basic distinction from Warlpiri or English: a non-future basic marking.

  3. Michael said,

    June 18, 2013 @ 12:16 pm

    Very neat to see some research on bilingual mixed languages. In my undergrad thesis (by no means a particularly impressive piece of research, but an interesting project), I suggested that Yiddish might be classified as a bilingual mixed language (I started by looking at diminutive and pejorative suffixes which are largely of Slavic origin).

  4. James C. said,

    June 18, 2013 @ 12:58 pm

    “…I had no actual evidence that such languages actually existed and were learnable” — Seriously? To take a single example, there are still living Michif speakers, and I even know one. Maʼa is still around too, and Erromintxela is still said to be transmitted intergenerationally. There’s also Media Lengua, though I guess people argue about it.

    Another mixed language that hasn’t seen much research is Chiac. This is a bilingual French and English mix spoken in New Brunswick in Canada. It’s less ‘sexy’ because it involves two European languages, rather than a European language an a lesser-known language (cf. French+Cree Michif, Russian+Aleut Mednyj Aleut). Speakers are of course fluent in both French and English, but Chiac stands apart from both according to a linguist I know who grew up in the area (not a speaker). From what I’ve heard and read it seems to qualify as a mixed language, but I don’t know if anyone’s seriously studied it enough to establish some good examples from it.

  5. Sally Thomason said,

    June 18, 2013 @ 1:21 pm

    @ James C.: Yeah, well, for some theoreticians, if it isn't dreamt of in their philosophy, it doesn't exist.

  6. J.W. Brewer said,

    June 18, 2013 @ 1:30 pm

    Three or four years ago I found a cache of amusing videos on youtube of some musical group from NB (maybe Moncton?) rapping in Chiac. Alas I am unable to quickly retrace my steps to find them (perhaps "Chiac" is not the only searchable keyword that Kids Today use to describe that language variety); I think I may have originally been led there via a description by linguablogger Nick Nicholas (who is Australian but at least at one point traveled to Canada on business from time to time)?

  7. Stephen Chrisomalis said,

    June 18, 2013 @ 1:32 pm

    The Chiac group in question is Radio Radio. I'm partial to their song 'Jacuzzi': https://www.youtube.com/watch?v=7cRPH4lb8UI

  8. Stephen Chrisomalis said,

    June 18, 2013 @ 1:37 pm

    I'm afraid I have to disagree on the quality of the news reporting on the new article in _Language_ and specifically with the LSA press release 'New language discovery reveals linguistic insights' which begins, " A new language has been discovered …". This makes it sound as if the Light Warlpiri is newly discovered and as if the article in Language is the first time it's ever been mentioned in the literature.

  9. Dan Lufkin said,

    June 18, 2013 @ 1:42 pm

    What's the boundary between a mixed language and creole?

  10. bulbul said,

    June 18, 2013 @ 2:10 pm

    Dan,

    it's essentially a question of how the language came about: creoles are (at least going by the textbook definition) thought to arise from pidgins which emerge when populations come together which do not share a common language, often under harsh conditions (e.g. slaves on plantations) and in an environment where there is a socially dominant language (say, that of the overseers). As the pidgin is learned as a native language, it takes on the characteristic of that dominat language and creoles can therefore usually be traced to a single language, referred to as lexifier. One therefore speaks of Spanish-based creoles (Papiamento), French-based creoles (Kriyol) or English-based creoles (Tok Pisin). A mixed, or as some prefer it, intertwined language, emerges when two populations have been in contact long enough and close enough, but without one achieving a particular dominance so that no language shift takes place. Whereas with pidgin, the members of the community do not speak each other's languages, with intertwined languages, the opposite is true.

    "The striking thing about Light Warlpiri is that most of the verbs come from English or Kriol, but most of the other grammatical elements in the sentence come from Warlpiri,"

    This is indeed striking, because something similar is true of a stereotypical intertwined language, Michif, only in Michif, it's the Cree component that provides the verbs.

  11. bulbul said,

    June 18, 2013 @ 2:17 pm

    Michael (and others),

    I assume you are familiar with Bakker and Mous "Mixed Languages:
    15 Case Studies in Language Intertwining" and the follow-up by Matras and Bakker "The Mixed Language Debate". If I recall correctly, the latter volume dismisses the idea of Yiddish as a mixed language.

  12. Etienne said,

    June 18, 2013 @ 2:36 pm

    Dan Lufkin: the boundary between mixed language and creole is stark. A creole draws practically all its vocabulary from one language (called the "lexifier" in creolistics) but lacks most, if not all, the bound morphemes of the lexifier (or indeed bound morphemes of any origin). As a result a creole language is always, typologically, a heavily isolating language. The fact that the same is true of pidgin languages is what leads many (myself included) to the conclusion that creoles are nativeized pidgins.

    (There do exist a few creoles –Saramaccan, Angolar, Berbice Dutch Creole– which are quite mixed lexically: but in the case of the first two comparative evidence makes it plain that this lexical mixedness postdates the birth of the earliest form of the creole language in question).

    A mixed language, on the other hand, combines the vocabulary and/or the bound morphemes of two languages. Thus, Michif not only has French nouns and Cree verbs: its verbs also have Cree bound morphemes, and its nouns have French bound morphemes. Another type of mixed language combines vocabulary from one source with bound morphemes from another: Media Lengua, with Spanish vocabulary and Quechua morphology, is a case in point.

    Incidentally, in 2004 the evidence of Light Warpiri would have been superfluous to a demonstration that mixed languages can be acquired by children: Michif is no longer being acquired by children, but it still has native speakers, despite the fact that it must (according to Peter Bakker) have arisen as a separate language in the first half of the nineteenth century, and thus was acquired by several generations of speakers (found in a number of geographically discontinuous communities), many of whom, crucially, knew neither French nor Cree. So it is difficult indeed to deny Michif the status of "natural language".

    (Indeed, as the only known mixed language with speakers who are fluent in neither of the component languages which combine in this mixed language, Michif is in a class of its own).

    James C.: conversely, "Chiac" doesn't qualify: quite apart from the fact that its speakers speak French and English, it is my understanding that Chiac is an idiolect acquired and used in high school by native French speakers fluent in English. That could change in the future, of course.

  13. Pompeius said,

    June 18, 2013 @ 2:57 pm

    Non-future is interesting. What's the percentage of languages with a futur/non-future distinction, but no difference between present and past?

  14. Dan Lufkin said,

    June 18, 2013 @ 3:26 pm

    @bulbul — Thanks, that clears up a lot. I'm sitting here in the midst of translating a lengthy Afrikaans medical document and constantly admiring how a language with so little grammar can be so solid and unambiguous. Perhaps going through creolization removes the fluff from a language.

    Afrikaans has been described as Dutch with Malay grammar.

  15. Sally Thomason said,

    June 18, 2013 @ 3:37 pm

    @ Dan Lufkin: I'd disagree with parts of the responses to your query about mixed lgs. vs. creoles. First, I've argued that pidgins & creoles, like bilingual mixed languages, *are* mixed languages — historically, none of these languages can be viewed as primarily the descendant of a single parent language, in contrast to the vast majority of the world's languages. Second, to my mind the crucial distinction between a bilingual mixed language and a creole (or a pidgin) is that bilingual mixed languages arise in bilingual contexts, whereas creoles & pidgins, more or less by definition, don't. The creators of Michif, for instance, *must* have been bilingual in both French and Cree, because there is very little structural distortion in either component of the grammar (French noun phrases, Cree verb phrases); the creators of Mednyj Aleut must have been bilingual in both Russian and Aleut; the creators of Media Lengua must have been bilingual in both Spanish and Quechua; and so forth. Not so in pidgins & creoles, where the amount of grammatical structure from the lexifier language is extremely limited. There are borderline cases, of course — we're talking about history here, fuzzy boundaries are to be expected. Reunionese, a French-lexicon creole, has quite a bit of French grammar, and Barbadian English Creole (or "Creole") has a lot of English grammar. The distinction doesn't rest on the presence of absence of lexifier-language bound morphemes; syntax, lexical semantics, and phonology have to be taken into account, and it's quite possible to have a creole whose lexifier language had few bound morphemes to begin with. Moreover, creoles that developed from lexifier and other languages that were typologically similar *do* sometimes have quite a bit of bound morphology; an example is Kituba, with a fair amount of inflectional morphology because all the input languages, lexifier and others, were Bantu, with typical elaborate Bantu morphology. The creole doesn't have the full morphological arrays of any of the input languages, lexifier included. But it's not an isolating language.

    Finally, creoles that evolve out of pidgins don't necessarily move closer to the lexifier language in grammatical structure! Depends on whether the lexifier language is in close contact with the creole (especially if it's dominant wrt the creole). And it isn't at all clear that all creoles evolve out of pidgins. There are still a few candidates, possibly viable ones, for abrupt creolization. Pitcairnese, for instance.

  16. Coby Lubliner said,

    June 18, 2013 @ 5:23 pm

    Isn't the Brussels dialect known as Marols or Marollien a mixed language?

  17. Etienne said,

    June 18, 2013 @ 5:58 pm

    Dan Lufkin: Professor Thomason is entitled to her own views, but most scholars (myself included) would not regard pidgins or creoles as mixed languages. The reason, in my case, being that the isolating structure of these languages appears to be universal: not matter how morpheme-rich the lexifier and/or substrates are, the resulting pidgin/creolees are always heavily isolating languages.

    Indeed, pidgin/creole grammars are so alike across the globe that they cannot owe more than secondary details to the influence of other languages: and to repeat my point above, lexically pidgins and creoles are by no means mixed. Hence my disagreement with Profesor Thomason's claim.

    Professor Thomason: while the creators of Michif must indeed have been bilingual in French and Cree, this is NOT true of the creators of Media Lengua. Inasmuch as the language shows Spanish lexicon but Quechua (and not Spanish) morphosyntax and grammar it is quite possible that the creators of Media Lengua never mastered Spanish morphosyntax and grammar and only knew Spanish vocabulary (perhaps they only spoke a Spanish pidgin for example). This scenario may or may not be true, but to my knowledge nothing about the structure of Media Lengua would preclude it.

  18. marie-lucie said,

    June 18, 2013 @ 6:42 pm

    ST: A prominent linguist in the audience protested during the comment period that I had no actual evidence that such languages actually existed and were learnable

    I am puzzled by that linguist's comment: "and were learnable". Surely, if a phenomenon observed among humans deserves the name of "language", it must be learnable! What use would a non-learnable language be? The proof of anything being "learnable" is that it has indeed been learned by some humans, including even linguists.

  19. bulbul said,

    June 18, 2013 @ 7:17 pm

    Professor Thomason,

    creoles that evolve out of pidgins don't necessarily move closer to the lexifier language in grammatical structure
    I certainly didn't mean to imply that by referring to the characteristics of the dominant language, that comment was meant to refer to the lexical component of the creole.

    pidgins & creoles, like bilingual mixed languages, *are* mixed languages — historically, none of these languages can be viewed as primarily the descendant of a single parent language
    Perhaps so, but that only speaks to the (lack of) genetic affinity of a particular creole with its lexifier. It does not make creoles mixed (or intertwined) in any meaningful sense of the word.

    the crucial distinction between a bilingual mixed language and a creole (or a pidgin) is that bilingual mixed languages arise in bilingual contexts, whereas creoles & pidgins … don't
    Then the question is, whence the mixing in creoles? Wouldn't this imply that a creole is essentially a restructured form of the lexifier and anything not coming from the lexifier is due to substrate influence or some universal of language contact? If so, this is very different from a mixed language like Michif.

    The creole doesn't have the full morphological arrays of any of the input languages, lexifier included. But it's not an isolating language.
    Here I agree with Etienne: creoles as a type do tend to exhibit very strong analytical and even isolating tendences and the replacement of inflection with free lexemes (in comparison to the lexifier) is generally considered one of the hallmarks of a creole. That is not to say that no creole has any inflectional morphology (e.g. the oft-mentioned Berbice Creole Dutch perfective suffix -te), but as a rule, creoles really don't have much inflection and describing them as isolating is appropriate.

  20. Doug said,

    June 19, 2013 @ 2:55 am

    For those interested in hearing Carmel present on Light Warlpiri, here's a webcast of a seminar she gave at AIATSIS (Canberra) recently (scroll down to the Special Seminars during Semester 1).

  21. minus273 said,

    June 19, 2013 @ 8:03 am

    From the Wikipedia page, chiac just looks like good old French with a little bit of loans from English.

  22. marie-lucie said,

    June 19, 2013 @ 1:18 pm

    Minus, I don't know what you mean by "good old French". Leaving aside the English words (and some English syntax), some of the morphology and syntax are Acadian, not Québécois or Standard. Acadian French has many things in common with dialects of the French mid-Atlantic coast, from which the Acadians originated.

  23. Asya Pereltsvaig said,

    June 19, 2013 @ 3:39 pm

    Etienne:
    (There do exist a few creoles –Saramaccan, Angolar, Berbice Dutch Creole– which are quite mixed lexically: but in the case of the first two comparative evidence makes it plain that this lexical mixedness postdates the birth of the earliest form of the creole language in question).

    What about Russenorsk, which was quite mixed lexically though it was a pidgin, not a creole? It didn't have a clear lexifier—which didn't prevent it from being a pidgin…

  24. Dan Lufkin said,

    June 19, 2013 @ 4:06 pm

    I doubt that there were enough children who ever spoke Russenorsk for it to become a real pidgin.

  25. bulbul said,

    June 19, 2013 @ 4:30 pm

    Asya,

    as a pidgin, Russenorsk was unstable, not to mention functionally limited, so its hardly appropriate to make any judgments on the composition of its lexicon. Which is also why I don't think it makes sense to speak of a lexifier in case of a pidgin – creolists certainly don't.

    Dan,

    by definition, if children speak it, it's usually already a creole. Pidgins arise in a contact between two groups of people who do not speak the other's language and need to communicate in a limited and scenario, like when they work together or trade.

  26. minus273 said,

    June 21, 2013 @ 4:41 am

    marie-lucie, I just mean that there seems to be very little that is structurally English in the language. It looks French and feels French. Which dialect of it is another matter.

  27. Asya Pereltsvaig said,

    June 21, 2013 @ 5:30 pm

    bulbul:

    I don't think it makes sense to speak of a lexifier in case of a pidgin – creolists certainly don't

    … but those who believe that creoles arise from pidgins do, no? After all, isn't it the case that the usual explanation for the fact that creoles have one lexifier is that pidgins typically have one lexifier?

  28. Etienne said,

    June 21, 2013 @ 10:11 pm

    Coby Lubliner: Marollien is indeed a Dutch-influenced variety of French, but it isn't a mixed language the way Michif or Media Lengua are: obviously it is difficult to draw a clear line, but if Marollien is to be called a mixed language then I suspect a majority of the languages of the world would deseve the title.

    Dan: what Bulbul said. Some scholars have claimed that a pidgin becomes a creole when it becomes a generation of children's L1, whereas others claim that the use of a pidgin as the dominant language of a community (leading to its becoming the L1 of the next generation of the community in question) is what transforms it into a creole. Either way, a language with L1 speakers cannot be called a pidgin.

    Asya: Russenorsk was mixed, but tellingly, it did not nativize and turn into a creole. I think this is no coincidence. The best-known creoles –those of the West Indies– arose in a radically inegalitarian context, and as a result the language of the dominant (European) element supplied the vocabulary of the pidgin which subsequently became a creole. In the egalitarian context of Norwegian-Russian trade neither group dominated the other, and thus a lexically mixed pidgin arose, which neither group could impose upon the other.

    All: I left some comments clarifying (some of) my views on pidgins, creoles and genetic relationship on the May 2 thread, "Ayiti Pare", which I encourage interested readers to have a glance at.

  29. Sally Thomason said,

    June 22, 2013 @ 2:15 pm

    I'm with Asya on the pidgin vs. creole and the pidgin-lexifier issue(s). Chinook Jargon is another pidgin with a significantly mixed lexicon; it is, or was, demonstrably stable, though with dialect variation; and it's not at all clear that it arose in a mainly egalitarian context — the Chinooks kept slaves, and at least one other Northwest pidgin/prepidgin/jargon (no data, so impossible to tell which) is reported to have arisen in talk by and to slaves. As for languages with native speakers being by definition creoles, one must surely allow for fuzzy boundaries here. There were, and for all I know are, at least a few native speakers of Chinook Jargon, but there was never a whole community of native speakers. So, a borderline case, but best treated, in my opinion, as a pidgin that was nativized to a limited extent. (As far as I know there's no data on L1 speakers of Chinook Jargon vs. L2 speakers, so no way to compare them. Speakers on the Grand Ronde [sic] reservation historicallly spoke a more "Chinooky" variety, since some or most of them were descended from Chinook speakers and their Chinook-Jargon-speaking ancestors presumably still knew some Chinook as well as the pidgin.

    It's easy to rule out cases like Russenorsk and Chinook Jargon and for that matter Kituba (a creole with quite a bit of morphology) by decreeing them atypical and therefore, implicitly, irrelevant for theorizing about pidgins and creoles. But that impoverishes our theories of pidgin/creole genesis and later development, so I don't think it's a good idea.

  30. marie-lucie said,

    June 23, 2013 @ 9:08 am

    Chinook Jargon and slaves

    The typical creoles of the Caribbean arose in a situation not only of slavery but of plantation slavery: the majority of slaves, drawn from many linguistic origins, worked together on large plantations, lived in separate quarters, and apart from their supervisors, had little contact with Europeans. Only a small proportion, especially women, were employed as domestic servants. Under those conditions, early pidgins rapidly developed into creoles, as the children of slaves grew up with the common language of their parents. Slavery on the Pacific coast was not of the same type: slaves, usually war captives, did a lot of the hard or menial work but were not physically separated from their owners. Nor were they a compact group or a large percentage of the population as in the plantation areas. Slaves were sometimes freed, sometimes redeemed by their own relatives.

    The Chinook, living on the Columbia River, engaged in trade along the river, encountering several other languages, and the Jargon is supposed to have arisen from the traders' deliberate simplification of Chinook (a language with complex morphology), with addition of vocabulary of various origins (especially Wakashan and Salishan, later French and finally English). No doubt many slaves (most of them speaking one of the indigenous "donor" languages) would have found it difficult to learn Chinook itself, and the masters would also have tended to speak to them in a simplified version of their own speech, but the slaves would also have heard normal Chinook spoken around them. I doubt that CJ originated as a means of communication with just slaves, rather than with non-Chinook speakers in general. In any case, unlike the creoles, CJ remained an intertribal and even international auxiliary language and did not start to acquire native speakers until several tribes, speaking a total of five languages, were grouped together on two large reservations and had no other common language. Other historical circumstances prevented it from acquiring more native speakers outside the limits of those reservations.

  31. Etienne said,

    June 23, 2013 @ 4:59 pm

    Professor Thomason-

    I certainly would not exclude Russenorsk or Chinook Jargon in a discussion of pidgins and creoles. Yet it cannot be denied that their lexical mixedness is atypical. On the other hand, they are both typical pidgins in another way: to wit, a (near?)-total absence of productively used bound morphemes, of any source.

    This is why I simply cannot accept your claim that mixedness defines pidgins and creoles. Especially since mixed languages needn't be morphologically reduced. Consider: a mixed language such as Michif, with its Cree verbs inflected with Cree morphemes, and its French noun phrases, complete with French morphology, can indeed be argued to be more complex than either of its components.

    Thus, whereas Cree has two grammatical nominal genders (animate/inanimate, with intransitive verbs agreeing in gender with their subject and transitive verbs with their object) and French a masculine/feminine system of nominal gender (mostly involving agreement within the Noun Phrase), Michif combines both systems into a system of four grammatical genders (I remind the reader that Michif speakers often spoke neither French nor Cree).

    The contrast with creoles or pidgins, including Chinook Jargon or Russenorsk, is stunning. Indeed, both languages stick out like sore thumbs from their source languages and neighboring languages (Northern Russian and American Pacific Northwest languages do not lack bound morphemes, to put it mildly).

  32. Asya Pereltsvaig said,

    June 23, 2013 @ 5:06 pm

    I agree with Marie-lucie that plantation environment, rather than trade (symmetrical or symmetrical socially), is what's conducive to creolization. Two other Russian pidgins, both related to trade, were symmetrical in their social nature and had largely one lexifier, Russian. Yet neither became a creole. See more on them here: http://geocurrents.info/place/russia-ukraine-and-caucasus/siberia/russian-based-pidgins-in-siberia

  33. Asya Pereltsvaig said,

    June 24, 2013 @ 12:59 pm

    Here's my recent post on this "new mixed language" in Australia (going back to the original theme of Prof. Thomason's post), in which I mostly take a different angle on Light Warlpiri: http://geocurrents.info/cultural-geography/linguistic-geography/new-mixed-language-discovered-in-northern-australia

  34. Bril said,

    June 24, 2013 @ 2:10 pm

    Etienne said,
    June 21, 2013 @ 10:11 pm

    Coby Lubliner: Marollien is indeed a Dutch-influenced variety of French

    That's what it may look like now, because the degree of frenchification has gone up so far it looks mainly as a form of French. But since the language spoken by 'the people' ('les petites gens') in Brussels – before the French took over the country and the city after the French Revolution and imposed French as the administrative language and a tool for social pyramid-climbing – used to be a form of Dutch (or 'thiois') one might just as well rephrase your phrase as: 'Marollien is indeed a French-influenced variety of Dutch'.

  35. Bril said,

    June 24, 2013 @ 2:12 pm

    looks mainly like

  36. Etienne said,

    June 25, 2013 @ 4:20 pm

    Bril: Marollien cannot have begun its life as a variety of Dutch. Obviously the ancestors of many Marollien speakers today were native Dutch speakers, but while they probably spoke French-influenced Dutch, this French-influenced Dutch is not the ancestral form of present-day Marollien: Marollien is a later, changed form of French, not of Dutch. In language contact studies it is well known that speakers do not mix languages haphazardly: rather, they speak a given language (known as the "Matrix language") within which they insert elements from another language. Marollien is typical: it is a form of French with Dutch elements. Not a form of Dutch with French elements.

    Indeed, what makes Michif so remarkable is that PRIMA FACIE it is difficult to tell what the Matrix language originally was.

    Marie-Lucie: actually, what is conducive to a pidgin nativizing (=creolization) isn't so much a plantation environment as it is an environment where the pidgin becomes the language of a community. The only reason why Chinook Jargon at Grand Ronde failed to unambiguously nativize was because the spread of English took place practically at the same time that the pidgin began to nativize, so that there never was a generation within the community which was monolingual in Chinook Jargon: had the spread of English taken place a generation or two later I have no doubt Grand Ronde Chinook Jargon would have been a full-fledged creole.

    Indeed on plantations in the West Indies it seems clear that creoles were born well before the demographic disproportion between blacks and whites existed. The reasons why these creoles arose in the first place appears to be because the pidgin was the "in-group" language of African slaves, not because there were so few whites that blacks could not learn a full version of their language.

  37. Asya Pereltsvaig said,

    June 25, 2013 @ 6:03 pm

    Etienne,
    I think we are trying to say the same thing in different ways. I think plantations (as opposed to trade) are conducive to creolization exactly because in plantations a community of native speakers can arise. People are "stuck" on plantations all the time, whereas traders go home to wives and kids who do NOT speak the pidgin and have no need for it. Same applies to "master-servant pidgins" such as another version of Russian-Chinese pidgin that arose in Shanghai and Machuria: Chinese servants mostly went home when the workday was over, so there was no room for a COMMUNITY.

  38. marie-lucie said,

    June 25, 2013 @ 10:23 pm

    Etienne, Asya, I think we all basically agree. Slave plantations were communities, which did not include Europeans, so even if some slaves learned the dominant language (eg house slaves), it was not the language of their own community. For CJ, I mentioned "historical circumstances" which prevented full creolization: I did not want to go into details at that point, but those circumstances were precisely the ones Etienne mentioned. There were apparently a few children raised as L1 CJ speakers (who were probably bilingual with an indigenous language), but they seem to have been both the first and the last generation.

  39. Bril said,

    June 29, 2013 @ 3:12 pm

    @Etienne

    "Obviously the ancestors of many Marollien speakers today were native Dutch speakers, but while they probably spoke French-influenced Dutch, this French-influenced Dutch is not the ancestral form of present-day Marollien.

    OK. So in stead of a gradual take-over there would have been a clear switch from one "matrix language" to another one. When exactly?

  40. Etienne said,

    June 30, 2013 @ 12:01 am

    Bril: I am no expert on the linguistic history of Belgium, but I believe the linguistic gallicization of Brussels which led to the birth of Marollien was comparatively recent (ninetenth century): can anyone out there confirm or refute this? Perhaps with some references?

  41. Bril said,

    June 30, 2013 @ 11:57 am

    @Etienne

    Bril: I am no expert on the linguistic history of Belgium, but I believe the linguistic gallicization of Brussels which led to the birth of Marollien was comparatively recent (ninetenth century): can anyone out there confirm or refute this? Perhaps with some references?

    OK. Formulations like 'the birth of Marollien' and 'a Dutch-influenced type of French' and your rejection of its definition as 'a French-influenced form of Dutch' seem to reject a gradual metamorphosis and to presuppose a clear switch from matrix A to B. So it should be possible to point at the time it precisely happened and to describe the process, but you don't answer me . And you 'believe'. Mmm, doesn't sound very scientific. And you're no expert on the linguistic history of Belgium. And now you need help to clarify the facts your assumptions about le Marollien seemed based upon. Is there with all due respect any chance your statement about Marollien being a 'Dutch-influenced form of French' is more than an admirably bold shot into the dark?

    :=)

  42. Etienne said,

    July 1, 2013 @ 2:45 pm

    Bril: A fair question. Not a shot in the dark. Rather, dropping a stone in the dark and knowing that the law of gravity will apply and that it will fall until it hits something (the ground, my foot, whatever…)

    That is to say, in all observed instances of language contact (between languages that are clearly separate systems) there always is a Matrix language, and there always is a sharp divide between Matrix language A influenced by language B and Matrix language B influenced by language A. So despite my not being an expert in Belgian linguistic history I can state that Marollien, being a Dutch-influenced variety of French, cannot have emerged through ever-growing French influence upon what was originally a variety of Dutch. There indeed must have been an abrupt discontinuity: this I can claim on the basis of my knowledge of other instances of language contact. The details I leave to specialists in Belgian linguistic history.

    This is, incidentally, quite scientific: it is called uniformitariamism. Basically this principle states that, all other things being equal, we must assume that what is observed in the present holds true for the past, and that we do not have the right to postulate in the past events unobserved in the present.

    Does this answer your question?

RSS feed for comments on this post

Leave a Comment