ethology meets linguistics

This summer I read two interesting things about linguistics (both relating linguistics to cognitive science, as I work in a computational cognitive science lab). The first was part of Daniel Dennett's The Intentional Stance. For the most part I found that book rather difficult reading, but I thought his essay on language and monkeys was interesting (even though I can't really remember how it fit into his argument). I think he concluded that you couldn't really tell if monkeys were using a language or not.
The other interesting thing I read was a paper written by the professor whose lab I work in describing how to use Bayesian statistics in conjunction with probabilistic grammar.

Today I was reading these papers about how both 8-month-old infants and cotton-top monkeys can determine segmentation of nonsense words after relatively little training (very little for the infants, about two minutes in fact). It got me thinking again about the papers I read this summer.

First off, if monkeys really do derive grammar through the same mechanisms that humans do, that's just... weird. Even if monkeys have a language, they don't use human speech, and the fact that they can somehow figure out human speech is very interesting. If so, I think it would mean that they have figured out the phonology of our language far faster than we figured out the phonology of theirs. Somehow human language is hardwired into monkeys, while monkey is not hardwired into humans.

The other thing that I found interesting was the application of probabilstic learning to animals. I find it hard enough to believe that humans learn based on Bayesian conditional probabilities (or at least it can be modeled well with them), so going further to apply that to monkeys seems crazy to me.

Assuming, however, that monkeys do have language, and that the rules for the derivation of their own languages are conditionally probabilistic the way that they are for us. What would that mean?

I'm not sure exactly. I know it puts the Catholics uncomfortably closer to animals on the soul food-chain. I know it takes away the special character of humanity as opposed to animals. I'm not sure about its cognitive implications.

So really this post has no point except that I am thinking about this, and it's interesting. I think people should perform more cognitive experiments on monkeys. If monkeys have the probabilistic architecture for language acquisition, it doesn't seem an unreasonable leap to assume that they may have other similar Bayesian architectures.