Creoles, Pidgins and Conlangs

In Derek Bickerton’s book Bastard Tongues, he describes a thwarted plan to essentially do collaborative, naturalistic* conlanging for the good of science. A couple of subsistence farmers from radically different linguistic backgrounds would be put on a uninhabited island for a year or three *with their wives children* and the linguists would observe the creation of a pidgin and a creole. The project was killed for fear of doing something unethical to the study participants.

* naturalistic, but not Tolkienesque.

And I think that is in part because most the the creoles that exist came out of horrible situations: slavery, brutal indentured servitude, war, genocide, mass forced migrations. I hope people have the logical clarity to keep these separate from the languages that happened. And you will have to continue to hold that clarity because there are really interesting similarities between creoles and conlangs, mostly on the technical language-features level, as less so on the social level.

Some interesting findings from studies of creoles is that they:
- are usually easy to learn, compared to well established languages
- are more regular, maybe more logical, because they didn’t have the time to accumulate inexplicable idioms and conventions
- in creole form have remarkably little in common with their superstrate* or substrate, but a lot in common with other creoles. They do have a superficial similarity that comes from borrowing the vocabulary of the super and substrate languages. But typically the grammar gets reinvented from scratch.
- creoles often embody a sort of “default” grammar, the sort that a child would instinctively use if something was missing from the language of the home as it clearly is when the parents and neighbors are all pidgin speakers.

* The superstrate is the prestigious language of the contact language community, the sugar mill owners language, for example. The substrate is the many languages of the slaves, servants and workers.

Pidgins rely on context, have few words, little or no predictable grammar. They fall somewhere between an animal communication system and a full language. I suppose one could create a con-pidgin, it wouldn’t take long, but it wouldn’t be very rewarding.

toki pona, in part inspired by things like tok pisin, fits poorly into the pidgin/creole taxonomy. toki pona has as few words as a pidgin, relies on context and some things are just to hard to express with the tools available. But toki pona has a very strict grammar, the grammar has sophisticated things like particles, a well defined lexicon (where as in a pidgin, you’re allowed to use any word you suspect the other person might know!).

Conlangs-with-a-community (to distinguish them from auxlangs which, except for Esperanto, usually don’t have a community), are a kind of contact language. But we don’t get too much creole type text because the rules of grammar are (usually) clearly posted and most successful communities have some talented polyglots acting as tutors to move people from the pidgin stage to the somewhat competent 2nd language user stage. Real life pidgin speakers don’t have that. Also real life conlang fans don’t normally expose their children to their conlang hobbies, so they kids don’t get a chance or the incentive to inject conlang like the children of pidgin speakers.

Conlangs, to some extent have some intersection with pidgins. For example, conlang fans and really most second language learners go through a “pidgin” phase where they either use their mother tongues grammar, or simply unpredictable grammar when forming sentences. Conlangs on the other hand tend to have grammars, unlike pidgins, and sometimes these grammars are much more sophisticated than a typical creole. On the other hand, people are lousy random rule generators, so they tend to make conlangs that are especially regular, like a creole.

An Experiment if there are any 10 year old Conlangers out there
I conjecture, but can’t prove it, that conlangers, especially younger ones, are more likely than not to pick creole type rules, although they will inevitably be influenced by what they just read. So if they were to conlang in a vacuum, we’d get a creole type conlang. If they just read about Esperanto, a 10 year old conlanger will create yet another Esperanto clone.

An Experiment if there is a bunch of Linguistics PhD’s out there with kids
The “create a creole” experiment could still be done, without funding if they were ex-patriots and old spoke their L1 at home and took their kids to daycare with other children of ex-patriots of different L1s. As long as the weren’t planning to teach the kids the language of the community (say because they aren’t planning to live abroad forever) then the kids at daycare could conceivably create a creole. The ideal place for something like this would be a University in somewhere like Latvia, where it is unlikely that visiting professors would invest much time in learning Latvian. The best way to deal with the caretaker question would be to have a sign language only caretaker. The kids learn the sign language, but would probably rather create a creole for talk amongst themselves.

Tok Pisin
Tok Pisin is an interesting creole, not covered in this book very well, because it was created, *in the creole form* largely by adults. Pidgin speakers use communication strategies, like “try all the words in all languages, in all orders until they seem to understand” But creole speakers use rules and generally it’s only the children that will spontaneous create rules and follow the rules that other people in the community are using. Pidgin speakers are oblivious to any hint of rules in other pidgin speakers language and pidgin speakers aren’t known to learn creole from their children either!

So in a way, tok pisin is one of the largest adult created collaborate conlangs in the world.

This entry was posted in creoles. Bookmark the permalink.

2 Responses to Creoles, Pidgins and Conlangs

  1. in creole form have remarkably little in common with their superstrate* or substrate, but a lot in common with other creoles.

    Of course, 99% of creoles that have received any attention at all have a superstrate language of IE descent.

    For a creole experiment, cf. this (and then, for a conlang, this).

    • matthewdeanmartin says:

      I re-read you article– it reminds me of my Icelandic study groups. The effect of motivation really swamps the effects of any other dynamic going on. Yours indeed was a case where maybe it would have helped to have had some slavery and whips going on to get people to focus on the task of communicating. Anyhow, it gives me an idea for my next blog post about identity & motivation.

      And don’t forget to look for contact language features in the writings of fans with a non-English, non-professional-polyglot background. Those are the ones most likely to spontaneously create creole-type constructions if bioprograms exist and if the bioprogram applies to adults.