December 18, 2006

The Portuguese pidgin that never died?

For some reason, the idea that the world's creole languages all, or mostly, trace back to a 15th century Portuguese trade pidgin language seems to stick in people's minds. Today, a New York Times review of David Crystal's new HOW LANGUAGES WORK reminds us of this tasty factoid, whose memory-friendly nature likely has something to do with its piggy-backing on our school-days familiarity with the Great Explorers, Vasco Da Gama, Magellan, and so on.

Too bad it isn't true.

There indeed was a Portuguese pidgin used on the west coast of Africa starting in the 1400s. It was used in contacts with Africans, and was later adopted by subsequent explorers of the coast such as the English, French, Dutch, Spanish and even Swedes. Varieties of the pidgin were used as far afield as India and today's Indonesia.

In the fifties and sixties, a few linguists proposed that the reason that creoles like Jamaican "patois," Haitian Creole French, Papiamentu Creole Spanish, etc. have interestingly similar grammars is because slaves brought to those colonies had learned the Portuguese pidgin in Africa, and plugged words from the European language of their colony into the Portuguese pidgin's sentence structure.

So, in Haiti, the result would be a French creole, while over in Jamaica, the creole would be English -- but both would share a sentence structure inherited from the Portuguese pidgin. Creole language specialists call this the monogenesis hypothesis.

This idea was a nimble surmise based on the evidence available back in the day. But a great deal of linguistic and historical research has been done since then, and no working creolist has subscribed to the monogenesis hypothesis for over twenty years.

There is no evidence that an appreciable number of slaves spoke Portuguese pidgin -- it was used by African traders but would have been of little use to a typical villager inland. There is no creole that looks anything like a descendant of the pidgin in terms of sentence structure -- rather, the pidgin left behind a few words like SAVVY "to know" and PICANINNY "small" in creoles around the world, just like O.K. is sprinkled around the world from English.

Probably the most summary and conclusive statement that put the monogenesis hypothesis to rest was an article by Morris Goodman in the anthology PIDGIN AND CREOLE LANGUAGES edited by Glenn Gilbert in 1987.

There's a whole passel of ideas as to why creoles' grammar tend to be almost oddly similar. Derek Bickerton of the University of Hawaii has argued that children, deprived of coherent linguistic input by pidgin-speaking parents, spontaneously generated creoles as direct products of our innate mental configuration for Universal Grammar. Others have argued that the native languages spoken by slaves in most colonies were all quite similar themselves, such that the creoles the slaves created are equally similar. As always, the truth is likely a combination of all of the hypotheses bandied about.

That is, of the LIVING hypotheses. The monogenesis hypothesis does not, in the scientific sense, explain or predict anything about how creoles are currently known to pattern grammatically or geographically. It is, at this point, an archival matter.

That as awesome a scholar and writer as Crystal has included the Portuguese pidgin idea in his entry on creoles in HOW LANGUAGE WORKS is, however, not really his fault. There is a kind of ingrained custom in textbooks on pidgins and creoles to genuflectively list the monogenesis hypothesis along with the living ones, such that one pretty much has to be one of the small gang of working creole specialists to know that the monogenesis hypothesis today dwells with the likes of phlogiston and the four humours.

Posted by John McWhorter at December 18, 2006 03:58 PM