i'm not sure it even makes sense to try and ask questions about grammar as they relate to sign language outside of the context of a spoken and written language, and i think it opens up a lot of questions regarding the circularity of it. i know there's different ideas about it, but i have a hard time separating grammar from written language. it seems to me that it's the writing that enforces the grammar, rather than the other way around. when you look at tribes that don't have written languages, the grammar may exist but is often rather basic - and they have thousands of years of linguistic evolution to get to the point, whereas koko only has her lifetime and a set of limited tools to express herself. i don't really have to hypothesize about taking europeans and putting them on a different planet without writing - you can look at how the language has broken down in areas of australia and north america, where the written component is not great. that is, you take the writing out and the grammar demonstrably starts to fracture. so, i just don't see how this experiment is able to produce any kind of meaningful conclusions on the question. to answer that question, you'd have to carry the experiment out over generations, teach them how to use written language and then construct something that gives the gorilla more ability to use grammar than signing.
but, i'm not falling for this idea that the gorilla is being conditioned. i've seen very little, of course - youtube videos. but the bit i've seen is just overwhelmingly in favour of an independent agent producing independent thought.
if the gorilla can understand over a thousand signs, it could conceivably understand just as many key combinations on a keyboard. that would eliminate a lot of ambiguity. perhaps using chinese style writing (or even something roughly comparable to hieroglyphs) may be a better way to start.
after all, humans didn't start with a complicated alphabet, either. we built it up over time. we started with pictorial representations that expressed ideas.
so, it's not really fair to grab a gorilla and expect it to grasp a modern roman alphabet with the complexity of a modern language right off the bat. i wouldn't even expect that a pre-neolithic member of our own species would be able to do that.
everything we know about plasticity and evolution nowadays suggests that whatever is inherent must have developed over the time we've been using language and grammar. so, if you want to do this and draw any meaningful results, you need to control for that by emulating the same kind of systems that early humans used, not the fully developed ones we use now.
i mean, we have no idea what ancient egyptians sounded like when they talked to each other. we can take some guesses. but there's not really any serious way to really understand how complex their grammar was, at the time.
chinese would be better for that reason, but it might be too complicated.
if tolkien can construct a new language, it can't be that hard to make one for some apes using a simple but "correct" grammar and then transliterate it with pictures constructed with combinations on an oversized keyboard.
and i'm suggesting this because i think the results will be shocking to certain people and put some questions to rest rather permanently.