What language has the most complex grammar in the world?

Hypothetically, if babies were born to only deaf parents on an island and were completely isolated from the rest of the world, would they develop any sort of language/pidgin/creole?

• I was reading Steven Pinker's A Language Instinct, and it speaks about how language is an in-born instinct and how some pidgins subsequently developed into creoles with corrected grammar rules when babies started learning the language. I'm really curious and wonder if babies could form a language of their own, given this hypothetical situation. Assuming that a lot of babies were on this isolated island (isolated from the rest of the world, of course) and they are free to interact with each other as they grow up. How long would it take them to develop some sort of voiced language? Would they develop some sort of sign language earlier? Either way, how long would a full-fledged language be developed?

• Answer:

  This has obviously never happened, but in looking for evidence on what might happen, it's important to recognize that https://en.wikipedia.org/wiki/Ontogeny does not recapitulate https://en.wikipedia.org/wiki/Phylogenyhttps://en.wikipedia.org/wiki/Recapitulation_theory was an intriguing, but disproven idea that the development of an adult from an embryo essentially mirrored the evolution of humans from single-cell organisms: In retrospect, a pretty ridiculous... but certainly creative... idea. The same principle applies to language such that we can't look for evidence on the evolution of language in humans ca. 200k years ago. Though we're still talking about biologically modern humans (homo sapiens), the context and environment that went into language evolution are not what would be relevant to your scenario. (By the same token, any evidence we would get from a scenario like this would not be definitive in terms of understanding how language evolved.) The evidence that we can use would depend, actually, on a detail in your scenario that was not clear, but really would be the crux of an answer, and that's the parents. You owe your parents a lot. And not just for putting food on the table. Because you not only need an child brain to learn a language as a native speaker, but you also need language input. In your scenario, either the parents are there, being parents, but incapable of communicating verbally. Or they're not really there at all, except to do the equivalent of sliding food under the door. Kind of, sort of similar stuff has happened that we can reference. Deaf Parents In the former case, it's certainly happened. Deaf parents have hearing kids. But, you might say, you mean that all the adults can't speak. That's happened too, if you flip the scenario on its head: Imagine a deaf child is born into a speaking world where there are no other signers. (Recall, we're all spoken-language biased, presuming that speech is somehow more language-y, but sign language is equivalent in all ways, except for modality.) The child and parents will develop a system called http://en.wikipedia.org/wiki/Home_sign, where some amount of grammar and regularity will emerge. There will also be some abstraction developed in terms of the lexicon. Instead of placing hands on head in a certain way to mean hat something a bit quicker and more abstract may evolve, especially if it's a frequent form. And what about a whole bunch of kids? What happens when they get together? We're had that scenario happen in real life, too. In the 70s and 80s, Nicaragua started a school for the deaf and so deaf  Nicaraguan children from rural areas would be able to interact with  other deaf children for the very first time. They all came together and the kids, communicating with their home sign, began to develop a fully fledged Natural Language,  with an even more intricate grammar and an even more abstract lexicon. This was particularly true for the latter generations and for the younger kids. Kids teaching each other NSL So, to put it in terms of pidgins, creoles, and language, the first contact amongst these home-signers would probably yield something pidgin-like. With the younger kids, though, you'd start to get the equivalent of a creole/language. No parents In this case, you're essentially asking about Tarzan crossed with Lord of the Flies. Documentary? For evidence, we can turn to the unfortunate cases of the unfortunately named http://en.wikipedia.org/wiki/Feral_children. The most notorious case was http://en.wikipedia.org/wiki/Genie_%28feral_child%29, but there were many others, e.g., http://en.wikipedia.org/wiki/Cambodian_jungle_girl or http://en.wikipedia.org/wiki/Ivan_Mishukov, most of questionable authenticity. It's difficult to figure out what to make of it, but if there is one things that we can glean from these cases it's that language development, cognitive development and nurturing all go hand in hand. You can't separate one apart from the other. In this case, it's probably best left to the writers and dreamers among us to imagine what might happen. Pidgins The last piece interesting evidence, which I would be remiss in not mentioning, is from the emergence of http://en.wikipedia.org/wiki/Pidgins. Pidgins are the languages that evolve when two different languages come into contact with each other and a third is created. The relevance here is the question of the role of children in a Pidgin becoming a Creole, or what some would consider a real language. One of my advisors, http://www.linguistics.ubc.ca/people/hudkam, has looked into the role of children in changing a Pidgin into a Creole. She did a language-learning experiment with adults and kids. They were both exposed to a language, but the language had some amount of irregularity in it. An analogy would be that "he runs" would be incorrectly said as "he run" some of the time. What she found was the adults tended to be veridical in their performance: if 3rd-sing. s was used 80% of the time, then they produce it 80% of the time. With the kids, though, they were found to over-regularized: If 3rd-sing. s was used 80% of the time (some) kids would use it 100% of the time. So, as with the NSL example above, children seem to over-generalize regularities that are part of the shift into a language system with what we would call grammar. That is what is supposedly part of the shift from a pidgin to a creole. So, whatever would happen would definitely reflect some sort of interesting properties of the human brain and its language faculty. Based on what we've observed via natural experiments and some constructed ones, there would likely be some sort of regularization of communication that would yield something that we would call language.