Chomsky’s genetic argument for grammar as inherent and universal in human language can be refuted in a less ambitious way by looking at acquisition of language by human children. Chomsky argues that no matter the amount of training, a kitten will not acquire human language, whereas humans do acquire language. And that is generally true, children are noted for their capacity to mimic. Yet, that is not universally true.
Feral children are those very rare children who are not raised by human parents. Often they are said to be raised by animals. They may or may not in fact be so raised. All we really know about the rearing of feral human children is that they did not die of starvation or get eaten. Well, almost all.
A unique thing about feral human children is that they do not acquire language. Even when inserted into society, even when extensive resources are devoted to training such teenaged or adult children — they simply do not acquire human language,(1) being limited to few or even no words relying on grunts and body language, as most animals do.
On the one hand Chomsky could argue this shows the fact that language is biologically conditioned, since most feral children are reinserted into human society after their sixth years of age – mostly as pre-teens but sometimes as teenagers or even as adults. In all cases, they do not acquire language, not in the sense you or I use language. Chomsky would argue that this is evidence that language is biologically conditioned. That does not however correspond to the fact that adults can be as adept or even more adept at language acquisition than children, and that adults face entirely different social circumstances in their acquisition of new language.
However, if human language were an inherent inevitable biological quality of humans then feral children would develop language on their own in the wild. They do not. The reason is because language is a social construct, not a biological inevitability.