One of my professors recently lent me a copy of Language Creation and Language Change, a compilation edited by Michel DeGraff. Lately, I've had something of an intellectual infatuation going on with Derek Bickerton, so I skipped right to his chapter (which happened to be chapter 2, so I didn't skip much). I'd like to summarize his main points (as I see them) here.
When adults of different linguistic backgrounds come together and have to figure out how to communicate with each other quickly and effectively (without the luxury of language classes), something called a pidgin might develop. This is basically just a mishmash of content words without much in the way of syntax. Pidgin utterances, much like the language of children under two, ". . . are devoid of hierarchical structure, . . . are extremely brief, . . . are incapable of expanding phrases, . . . regularly omit subcategorized constituents, . . . depend heavily on contextual and pragmatic clues for interpretation, and . . . are characterized by slow delivery punctuated by frequent pauses and hesitations," (DeGraff 2001: 64). Pidgins are highly variable both grammatically and lexically, and they tend to leave out pretty much all inflectional affixes and "grammar words".
These adult pidgin speakers also have access their native language and presumably speak it when they can find anybody around who'll understand it. But because the pidgin is what they have to use for day-to-day communication, and perhaps more importantly because this is what caretakers use when communicating with their children, this is what the children hear most often and, consequently, what should become their native language.
But the children aren't satisfied with this. They seem to have some pretty specific ideas of what they'd like to be to express already in their heads, semantic distinctions like [+/-anterior], [+/-realis], [+/-punctual], [+/-specific], and [+/-accomplished] (DeGraff 2001: 59). Bickerton calls these innate ideas "the Bioprogram". They use the inconsistent mess they hear around them mostly as a source of forms that they can adapt to express what they already want to. In the verb system, for example, it's ". . . as if they knew in advance what the parameter setting should be (nonpunctual) and then cast around for a word in the input that would appropriately express the setting. . . . These settings are not triggered lexically--they are triggered by the absence of TMA markers," (DeGraff 2001: 59-61).
Actually, the child is "under no particular pressure" to find words for these ideas if something similar enough is represented consistently in the target language (DeGraff 2001: 59), so these innate preferences don't necessarily show up as "unmarked parameter settings" in the world's languages (DeGraff 2001: 56). The language that these children of pidgin speakers develop is called a creole. Amazingly, the grammars of creoles around the world are essentially the same, though often quite different from that of their substrate and superstrate languages.
Now that I've given some explanation of what pidgins and creoles are and how they come to be, I'll cut to the chase. Somewhere "around the close of [their] twenty-fifth month," children progress very quickly from short, simple, pidgin-like utterances to longer, more complex utterances that make use of embedding and grammatical morphemes. Certain aspects of the grammar (such as irregular verb forms and proper distribution of pronouns and anaphors) take years to perfect, but the core of the grammar seems to be there almost overnight (DeGraff 2001: 61). At the same time, children experience unprecedented expansion of their vocabulary, but this does not seem to trigger their transition from a very limited communication system to a real language. If anything, the new grammar seems to trigger the growth of the vocabulary (DeGraff 2001: 63).
Actually, I'd love to see if this is indeed the case by looking at studies comparing normal to delayed acquisition. Do children who aren't exposed to any language until, say, their fourth year go through the same pidgin-like stage as normal children under two or do they progress much more rapidly having already developed their "language organ"?
So that's about the gist of it. Bickerton argues that ". . . the course of syntactic acquisition is consistent with the abrupt coming-on-line of a specific neurological module devoted to syntactic processing--a module that must be triggered within quite a narrow temporal window . . ." (DeGraff 2001: 65). Adults are unable to develop pidgins into creoles because they, like children under two, lack this Bioprogram for language. This also helps to explain why adults are such bad second language learners: they need "a rich and robust source of well-formed data" supplemented by "general cognitive capacities," (DeGraff 2001: 64).
This book was first published in 1999, and I'm not sure how long before that Bickerton's article was actually written, so it's just a bit dated. Bickerton's latest work, Bastard Tongues, was published earlier this year, and I highly recommend it as an introduction to and general overview of pidgins, creoles, and what they can teach us about how children learn language and what language actually is anyway.References:
- DeGraff, Michel. 1999. Language Creation and Language Change: Creolization, Diachrony, and Development. Cambridge: The MIT Press.