Back in grad school, I used to TA for Dr. Mischel of Nim Chimpsky fame. A lot of researchers badly wanted to prove language in chimps back in the 70's, but it

Back in grad school, I used to TA for Dr. Mischel of Nim Chimpsky fame. A lot of researchers badly wanted to prove language in chimps back in the 70's, but it all amounted to naught, and other than a few holdouts, the field moved on. People got excited about animal communication, but confused that with language. Language has a syntax, such that one can utter a sentence you've never spoken before and you can understand a sentence you've never heard before. Language is impossible under conditioning-based learning, but most of the chimp "utterances" could still be explained by conditioning. E.g., chimp Washoe infamously signed "water" and "bird" when a swan happened to be nearby, but that could be explained with operant conditioning: she associated signing "water" and "bird" in the presence of waters and birds with getting treats. Eventually, researchers started taking a sober look at the "corpuses" generated by animals. When cherry-picking individual phrases, things looked impressive; but when they really started to analyze these statistically as a whole, the results were unimpressive, closer to random utterings designed to elicit treats, rather than communication.