In rare cases, however, children do develop an entire language of their own, and amazingly, all full-blown twin languages spontaneously develop the same structure, regardless of the language spoken at home. Aarhus University linguist Peter Bakker told me that twin-language structure is unlike that of any established language, and its syntax doesn't simply reflect the usual mistakes made by children. (Deaf children not taught sign language who invent their own also use this structure, by the way.) This "gives us a potential insight into the nature of language," Bakker said, into mankind's "first language," now lost to history.
Twin languages are simple, just as simple as necessary, one might say. For one, they freely mix subjects, verbs, and objects, putting the most important item first in any context. In an Estonian study a child said, in his private language, "Again I foyer toward write come." (Estonian grammar would have dictated, "I come again to the foyer to write.") Negation appears at the sentence's beginning or end, regardless of where it appears in the native language. Thus one Swiss child said, "Bobby, here drive no!" instead of, "Bobby, don't drive here!" Verbs aren't conjugated. There's no way to locate things or events in time and space. And finally, twin languages almost never use pronouns, just proper names. Language can get simpler than this, but not much. ...
If language originated between just two people, it might well have looked like this: The seemingly universal twin-language structure is blissfully easy to use in one-on-one conversation. However, that first language would have had to evolve quickly to be useful to a larger community. Societies need "unambiguous ways to distinguish between subject and object," Bakker says. "In the twin situation these can be dispensed with, but not in languages in which it is necessary to refer to events outside the direct situation." ...
Linguist Bernard Comrie at UC-Santa Barbara cautions that research into the birth of language is still in its infancy. "First we were told that creole languages would provide us with insight into 'first language,' then when that didn't pan out interest shifted to deaf sign language (also with mixed results)—I guess twin language will be the next thing," he wrote me. Twin language is particularly difficult to test because children give it up quickly, except when they are very isolated. And you can't just isolate kids on purpose—not anymore, anyway. Gone are the days when the pharaoh Psammetichus I could send two infants off to be raised by goats, or the Holy Roman Emperor Frederick II could forbid children's caregivers from speaking to them.
Project Nim, the 1970s attempt to teach the chimpanzee Nim Chimpsky to use sign language to disprove Noam Chomsky's ideas about what Steven Pinker calls the "language instinct," was in the tradition of Psammetichus and Frederick. It's an interesting question.
Lackman recalls a 1981 movie in which a primitive tribe has only one phrase: "It will be mine." That's not too far off from many of the things Nim had to say for himself, such as:
“Me banana you banana me you give.”
“You me banana me banana you give.”
“Banana me me me eat."