The parrot that can talk - or not

Almost every sentence of this story is misleading: Parrot's oratory stuns scientists.

A parrot with a fairly above-average ability to mimic human language, no doubt after years of training by a dedicated owner, is entertaining, but not ground-shattering, and no reason for any scientists to be stunned or have to revise any theories.

The claim that the parrot will "invent his own words and phrases if he is confronted with novel ideas with which his existing repertoire cannot cope - just as a human child would do" is particularly misleading, as I can tell you, even without seeing any evidence, that there's absolutely no chance that the parrot behaves anything like a human child. Children 'invent' words and phrases in a manner which is consistent with the innate structure and variables that make up the grammar of human language, whilst the parrot will be simply stringing together patterns of sound that it's picked up from humans around it. Whilst the parrot is alleged to have coined the regularisation of 'flew' into 'flied', this was already pretty common in English, particularly in the Baseball phrase 'flied out', (as noted by Pinker et al in a paper [PDF] in 1991), so it's quite possibly that the parrot was just mimicking this too.

This kind of claim, that some animal somewhere is remarkably advanced in having learnt language, is all too familiar to academic linguists, so much so that they barely give it any of their time. The problem is always that the people who own or work with the animals, and who conduct the research, are so keen for the animals to understand language that they make all kinds of scientific errors, introducing biases into experiments, over-interpreting utterances, and ignoring any data which doesn't back-up the answer they're hoping to get. I'm sure that in the course of owning a talkative parrot over many years, the parrot will come out with all sorts of phrases which seem to have remarkable significance within the context, but this is really no reason to tear up our understandings of language and animal intelligence.

The worst culprit is possibly, whose documentary broadcast last year, Chimps are People Too, managed to introduce a falsity into the very title of the show. His programme exemplifies the argument that, if we somehow can prove that certain animals have language, they must be just like humans, and so we just change our perceptions of them and the way we treat them. This is simply a dead-end argument, like trying to prove that animals have a deep appreciation of Shakespeare. If you think we should be nicer to animals, you can argue that on plenty of other grounds without resorting to complete fallacies.

Non-human animals do not have language, and will never be proved to do so. So resist the (natural) temptation to empathise with animals, anthropomorphasising them, and instead just treat them as wonderful creatures.

More on the parrot, 'N'kisi', can be found on Wikipedia (N'kisi (parrot)), which currently does a bad job of giving a neutral account, and on The Skeptic's Dictionary, which debunks it rather better.