[tsimionescu]

There are numerous discoveries of the limits of animal intelligence, especially in the area of language acquisition. Especially when Skinner's behaviorism concept was in vogue, it had actually been assumed that at least some animals could learn human-like language with enough training. This turned out to be completely false even for animals like chimpanzees and gorillas. In fact, some of the exact limits were identified - especially around syntax and embedding (one phrase inside another).

You can read up on a chimpanzee called Nim Chimpsky to see one of the more infamous failures in this area (and quite a sad story of animal mistreatment). It also illustrates the dangers of wanting to believe, or of excessive anthropomorphizing - as initially the experiments were deemed successful, until more dispassionate analysis showed that the chimp's signing was far more random and determined than presented.