From the Mouthes of Babes

Derek Bickerton is an interesting fellow, an odd man out who goes his own way even when he writes his memoirs. Most memoirs are, by definition, accounts of a person’s life. “I was born on a sunny day in the year of our lord 19__.” No such sentence is to be found in Bickerton’s memoirs. He does have a scene in which a driver almost crashes head on into an oncoming car, but this bit of personal recounting is so unusual that it left me wondering how he had happened to put that moment into his book. Bastard Tongues: A trailblazing linguist finds clues to our common humanity in the world’s lowliest languages is an account of “the world lowliest languages” and how they came to be as they are. The memoir side of it appears only in the fact that it describes how the author came to understand pidgin and Creole languages in his fashion. I should think the book will irritate scholars who have a different understanding of the subject. Fortunately, however, I’m not a scholar and I’m in complete agreement with Bickerton on his main theme.

In a letter to the editor of the New York Times Book Review almost 30 years ago I made the same argument (using Gullah as my example) that Bickerton makes at the close of his book. So there is nothing new in this book, moral-wise. Its strength is in its gathering of data from around the world to make the case that the urge and ability to use language is built into us: … the evidence from bastard tongues shows beyond doubt that a major part of language learning comes from the brain rather than experience (p. 246).

Pidgins are the simple languages people cobble together when they have to speak to one another, even though they do not share a common language. A Greek-speaking tourist in the mountains of Peru trying to haggle with an Inca-speaking roadside vendor will fall into a pidgin. Neither has any idea of the other’s language, but perhaps each can toss together a mix of Spanish and English words (while paying no attention to any syntactical rules) and negotiate a sale. Surely soldiers in Baghdad have some kind of ritualized, bastard Arabic-English stew that they use to bark orders at check points. Factory workers in India may labor side-by-side without sharing a common language (India is said to have over 29 languages with at least a million speakers) but they have to co-ordinate some things. Pidgins emerge spontaneously in such situations.

Bickerton makes a striking point about pidgins. They are not just without rules, they actively destroy rules:

…in any language words come embedded in a grammar of some kind. Start plucking words at random from different languages, and any consistent grammatical structure disappears. That’s not happenstance; it’s a logical and inevitable consequence of macaronic speech. [p. 218]

No wonder then that pidgins are exhausting and alienating. Even as their speakers come together they are reminded that they are members of separate communities. If people are forced to speak only pidgin, they are perpetually reminded of how they are outside of a world they where once they belonged. It would seem inevitable then that people forced to live in a community of pidgin speakers would transform their speech into something more expressive. Those more expressive languages are called Creoles and there has long been a question of just how a language can be transformed from an ad hoc, clumsy, syntaxless mess into something with rules and a vocabulary. Bickerton forever won himself a place in linguistics history by solving that problem. The children transform pidgin into a true language capable of supporting a community.

Much of Bickerton’s memoirs describe how he traveled around the world examining the Creole languages of the Indian Ocean, the Caribbean, and the Pacific Ocean to discover and then prove his thesis. The single best piece of evidence in favor of Bickerton’s it’s-the-children hypothesis comes from the way so many Creole languages share a common syntax. But to prove that the syntax comes from a common way of thinking Bickerton had to disprove the counter-thesis that Creoles all descend from a common first language. Bickerton does not refer respectfully to this older idea, and indeed is so scornful that his memoirs are a bit puzzling. Why did he have to keep working so hard at proving what he had already seemed to have proved, and at disproving what seemed disproved?

Nevertheless, he does keep working and establishes that Hawaiian pidgin was made in Hawaii and not imported from some rival first language. He showed that the records from around 1900 report that children were speaking Creole forms while adults used pidgin ones. He showed that when pidgin was the only available language, children will turn it into a Creole in one generation, and that this new language contains grammatical structures that are common to Creoles around the world, but were not part of any of the languages spoken by the children’s parents.

What’s in that universal grammar?

• It uses verbs to link subjects and objects of sentences.
• These linkages can be serial. In English, we use conjunctions to describe serial actions (e.g., I kicked the ball and scored a goal), but the universal grammar does not seem to have conjunctions (I kicked ball scored goal). English verbs may also imply serial actions. For example, the verb  to bring compounds two actions—carrying and traveling—and the verb to bring you compounds three actions, adding giving to the other two. In languages with serial linkages, however, all the verbs must be spelled out, so I brought you a book  becomes Me carry book come give you. 
• The linkages can tell us something about the relative time of the event (e.g., I see Tom versus I saw Tom). They can also tell us if the linkage really happened (e.g., I saw Tom versus I could have seen Tom). They can tell us if a linkage has been completed or not (e.g., Bill Clinton ran for president versus Hillary Clinton is running for president).
• Sentences can include a variety of markers indicating spatial location (e.g., We went to the house versus We went into the house), habitual action (e.g., He lied versus He always lied), or state of certitude (e.g., the dog barked versus a dog barked versus dogs bark). 

This list does not imply that all grammars have rules for expressing these features. They are available if need be. Serial verbs are particularly important because they appear to be universal in Creoles and not so common in other languages, perhaps because we have other ways of expressing the same thing. The important point is that Bickerton has identified a series of relationships that are universally perceived although pidgins have no rules for expressing them, but which have universally been brought under syntactical control by Creole speakers.

I agree with the philosopher Willard Quine, who is reported to have said that Bickerton’s empirical evidence for a universal grammar is far more persuasive than Chomsky’s theoretical arguments. It is also less dogmatic, for it carries no implications on just how that universal grammar works. It could be, as Chomsky proposes, a separate, syntactical module in the brain that generates its output to other areas (in which case some distinct, evolutionary account will be required), or it could be some part of our perceptual tracking system that has its roots deep in primate biology. Bickerton takes no position, contenting himself with the empirical task of establishing that there is a universal grammar which is sufficient to produce a language when children are thrown into a community of pidgin speakers.

Comments

[watercat]

The alienation you mention, I would think comes from not being able to share conventions, or rules, between peers, thus frustrating attempts to form emotional bonds and forge a community. Pidgins would be a remedy to this, rather than a cause.
If pidgins were truly an ad hoc, clumsy, syntaxless mess they would be indistinguishable from word salad, which they're not. Even pre-linguistic gestural communication has some structure, and in fact all the features you list as elements of UG are found in what Susan Goldin-Meadow calls the resilient properties of language, those that appear in attempts to communicate even when there is no linguistic input.

In both home signs and pre-linguistic gesture we find units that serve nominal, adjectival, or verbal functions (may be linked serially) arranged into predicate frames with markings for thematic roles such as subjects and objects; that communicate about the past, future, and even hypotheticals, relate narratives about the self and others and make metalinguistic references.

Starting with mutual eye gaze, people point to direct attention and establish shared topics, and use mimetic actions to add comments, resulting in a semantic frame of topic+comment that focuses attention on a complete predicate, and to which other elements can be added. Children express the elements of these predicate frames in habitual patterns, based first on semantics and later on distribution. Interacting with peers provides a cultural incentive to negotiate shared conventions that foster understanding, such as converging on certain word orders. If there is any pre-existing language in the environment, it provides a rich source of expressions to insert into these structures, which already exist: not in some nebulous language organ but in the human appetites for abstract thought and social bonding.

Throughout DB's book he keeps asking what the children create these languages out of. Obviously my answer is that he should look at language as the audiovisual object it really is. Whatever. This may be the best linguistics book I've ever read. It's not boring!

[BH]

I haven't kept up on Bickerton's work lately nor do I have references at hand, but I felt like chiming in that the "its-the-children" argument tends to fall apart when you consider creoles which we can see congealing (e.g., Tok Pisin) and historical evidence on some of Bickerton's (former?) central cases, like Hawaiian Pidgin and other plantation systems where children were rarer than newly imported laborers.

He may be right that pidgins and creoles can tell us something about language cognition, but the case for a creole prototype has suffered a lot in the last decade, at least when it's stated more specifically than above (as McWhorter and others have attempted to do). The idea of monogenesis has fell from favor among creolists some time ago (at least so far as I've seen), but this was in part due to the variety of pidgins/creoles in the world.

If I can find those references this weekend, I may post them. Searching for Bickerton in a linguistics database will certainly find you no shortage of criticism.

[watercat]

Cases like ABSL and Nicaragua conclusively disprove Monogenesis—the idea that all human languages derive from a single ancestral language.
Nor did the children draw gestures from their speaking caretakers since the co-occurring speech of the caretakers imposes an entirely different structure than the linguistic organization the children invent.
Similarities in the co-occurring gestures that accompany all the world's spoken languages require explanation no less than similarities that exist amongst all new languages created when users---adults or child---are denied a common language. The perception/attention pathway pursued in this blog provides such.

[BH]

Ack, I think I reached into my Scrabble bag and pulled out the wrong word entirely. I meant the single-origin hypothesis for the pidgins that emerged in the 16th through 19th centuries. Not monogenesis of all human languages. My mind is still blanking on what the correct term is.
-------------
BLOGGER: The only term I know for the single source of all human languages is the Mother Tongue Theory. But the critical issue for Bickerton is not the origin of pidgins, but the origins of Creoles. Were they created by separate groups of children (Bickerton's thesis) or are they all descended from a common Creole?

[watercat]

What common Creole could be ancestral to, for example, both Hawaiian Creole and Nicaraguan Sign Language?
----------------
BLOGGER: Obviously, none, but Bickerton had to devote much scholarly energy to the task of showing that, say, Hawaiian and Jamaican Creoles did not have a common ancestor.

[Jesus Sanchis]

From the point of view of mainstream Chomskyan linguistics, it seems that pidgins and creoles are seen as some kind of breaches in the normal image of languages as a perfectly organized set of rules. But that depends on how you look at it. Let's try to imagine pidgins and creoles not as the exception, but as the normal thing in language evolution. We live in the year 2008 and we know a lot about languages, grammatical rules, spelling rules, we know about grammar mistakes, verb forms, etc., but let's go back to a time when human societies had no sophisticated hierarchical structure or when there was no elite or domineering group telling the others how or what 'language' they had to speak. We see pidgins as something strange because they are really strange for us in the framework of our own modern sophisticated, highly-complex societies. But for many thousands of years, the mixture of languages was definitely a common phenomenon, and the most important factor in language evolution. Mario Alinei, the linguist who devised the Continuity Theory, talks about hybridization as the main reason for language change, and I think he's quite right. In my opinion, American linguists are far too obsessed with language rules. There's something more to this debate. Pidgins and creoles can definitely be seen from a radically different perspective. They may have a central role in linguistics, rather than a marginal one.

[JoseAngel]

Also, from the standpoint of integrational linguistics, one might consider pidgin-like babble (nongrammatical, nonregularized, nonChomskian, and with lots of signic interaction) as the primordial soup from which language arose. Grammar, regularity, a fixed vocabulary, etc., are invented later. Which is also a way of saying perhaps that the notion that Indo-European was a "language" in the sense that Greek or Latin or English are languages is a retrospective fallacy, one more version of the hindsight bias.