> those children were already exposed to civilization and in contact with language-using humans.

Exposure to civilization is obviously not important, as all uncivilized people also speak. (Exposure to other people is important.)

Contact with language-using humans doesn't help in developing sign language either; deaf children exposed to speaking humans do not develop spoken language or sign language.

It is generally felt that what prompted the development of NSL was the fact that the children were put into a group. Children in a group want to communicate with each other. A hearing child in the same group would also have learned/participated in developing the language.

We tried to teach various animals language and mostly failed.

Thus there is indeed a class difference between humans and the rest of animals.

The rest of animals seem incapable of this ability, even when we try our best to teach them.

For all we know, animals could have been trying to teach us their languages for centuries

AFAIK it's only very recently that we began to try and learn the languages of animals rather than teach them

Parent is arguing that they need to witness language use between members of their own species, and then not learn, in order to show they are incapable of language use.

The first human language user managed without this. (Though, if used for communication, it takes two... so there couldn't have been a first. OTOH there's an argument that language is primarily for thought, not communication, so it only takes one person. Secondly, evolution doesn't usually build a complete perfect system, but iterates, gradually improving. A puzzle here is that abstraction seems to be all or nothing, you can't have part abstraction. One resolution is that the necessary neuro-machinary was built up for other purposes, and abstraction emerged when some crucial capstone happened to be put into place. Or... maybe you can have some kind of "partial" abstraction...?)

I think you could have partial abstraction. People move around physical objects which represent other things, all the time.

Maybe representation is more important than grammar? e.g. representational art

The bee-dance "represents" direction and distance, but only that.

I wonder if it's a matter of skill or desire

Or, evolutionary advantage (primarily in the meme sense, but also genetically, for improving physiological and neurological support for language).

In what contexts does abstract language confer a survival or reproductive advantage, beyond simple communication, which is a commonplace, e.g. predator warning, food location, mate-seeking.

That’s our failure. Not theirs.

One huge failure is thinking that animals should care to learn anything we teach them.

If animals are unable to direct their attention towards learning to do something, that's functionally identitical to them not being able to do it. Even in an alternate universe where the only reason dogs can't speak English is that they can't work up the focus to sit down with a copy of Strunk and White, it would still be correct to say that animals lack the capacity for language.

You're assuming that not doing something implies an inability.

Does the fact that I have not adopted your position imply that I am unable to do so?

Why can't I be unwilling?