That prediction was proven correct, since all those "experiments" ended in embarrassing failure.Įdit - To be clear, as a disclaimer, I'm talking about UG in the "broad" sense and not as any specific generative grammar framework, which is obviously going to be way more uncertain and contested.īut if children spontaneously correct creolized languages or otherwise generate "correct" grammar despite access only to "incorrect" grammar (whether that be because the adults all have bad grammar due to speaking creol or due to learning a spoken language only late in life as with e.g. The predictive value is in the fact that a newborn from a remote Amazonian tribe can naturally learn to speak German as easily as any native speaker, the fact that animals have shown no signs of any such cognitive faculties and the fact that there's a critical period in development that can expire without stimulus and leave you without any language ability, just as being blindfolded through the critical period for vision development can leave you functionally blind.Ī prediction, for example, is that Skinnerean behaviorists would have exactly as much luck getting Nim and Koko to talk as they would getting penguins to fly. It's the basis of modern linguistics, and the disagreements are about the specifics of things like language acquisition.