February 02, 2004

Talking birds vs. singing birds

Among the echoes of the BBC's telepathic talking parrot story, there has been a certain amount of renewed debate about animal communication in general, for instance in the comments to this post on Language Hat's site. I'll have a few things to say on this topic here later on, starting with a description of some wonderful work on cephalopod morphophonemics. However, there's a general tension in the air whenever this kind of question comes up, so the first thing I want to do is to get some of the feelings out in the open, by considering an analogous case that people discuss very differently.

Forget about language for a moment. Homo sapiens may be the only genuinely musical species.

At least, no other species seems to have songs in which musical intervals (small whole-number ratios between pitches) play a role. No other species seems to create rhythmic patterns by subdivisions of a regular beat, or by repetitions of sequences of rhythmic cells composed of units with small-integer time ratios.

There might even be some related evolutionary specialization of the human auditory system -- at least, I've read that the just noticeable difference in pitch perception is about an order of magnitude smaller for humans than for other mammals (though a recent attempt to find the citation came up empty). There may also be evolutionary specializations of the motor system, for example to allow stronger voluntary control of the vocal apparatus. (Darwin thought that sexual selection for performance of courtship duets was a crucial step in the evolution of human language).

Then again, maybe this is going too far. What about the songs of whales or songbirds? Don't they have real melodies? Don't apes sometimes beat out rhythms?

Frankly, I don't think so, but the point is, no one seems to get worked up about this question. Books and articles are not written about whether the songs of whales or sparrows are Truly Song, or whether chimps' tree-root-pounding displays are Truly Rhythm. There is plenty of study, scientific and otherwise, of the vocal displays of whales and songbirds, and some work on the tree-pounding of chimps. However, the interest is in what these animal acoustic displays are like -- what they sound like, how they are made, how they are learned, what they are used for -- and not whether they are basically the same as, or basically different from, human music.

By comparison, there is enormous and continuing controversy about whether various sorts of animal behavior, both natural and taught by human trainers, are Truly Language.


For some reason, people find it more interesting -- either more attractive or more threatening -- to think of animals as perhaps having command of language, than to think of them as having command of music. The debate often seems to take on a quasi-religious tone, as if the issue were whether animals have (or can acquire) souls rather than whether they have (or can acquire) languages.

As with many such controversies, it's hard to stay out of the fight. However, I'd like to suggest that strong quasi-religious convictions on this subject are a bad idea, in both directions, because they make it hard investigate the phenomena as they are. Many (though not all) researchers and commentators come into the arena to look for evidence to support their beliefs, rather than to investigate the facts.

That doesn't mean, by the way, that I disagree with anything that Geoff Pullum wrote here. I do think, though, that it's sometimes helpful to view animal behavior from the perspective of what one might call "the communicative stance", by analogy to Daniel Dennett's "intentional stance". This amounts to treating an animal as if it has (or recognizes and responds to) communicative intentions, irrespective of whether it actually does. I recognize that this is dangerous at best, and expect to get a certain amount of flack for it. But that's for later.

[Note: the above is a slightly expanded version of the start of a lecture that I wrote a few years ago for an undergraduate course].

[Update: Bill Poser emailed this:

My impression is that one major aspect of the division between linguists and others on animal language is that there were different ideas in different fields as to what it means to have language and also different expectations. As of, say, 1950, biologists and psychologists generally thought of "language" as the use of symbols and thought that only humans could use symbols. When they discovered that other primates could use symbols, this was a big deal, and for them, it constituted having language. Linguists, on the other hand, seem not to have had any strong beliefs about only humans being able to use symbols, so the discovery that other primates did wasn't really news to them. Furthermore, for linguists language consists of much more than symbol use, so in the absence of non-trivial syntax, they weren't interested. Although there are other issues, such as the problems of data collection and interpretation with Penny Patterson's work with Koko, it seems to me that this has had long term effects. Linguists tend to dismiss anything short of full human-like language as uninteresting, while non-linguists often don't quite understand what the difference is, and even if they do, feel miffed by the linguists' reaction.

I agree, basically, though I think there is still a lot of disagreement (and confusion) about what it means to "use" a symbol, specifically in the "theory of mind" aspects. However, basically the same thing could be said about the many differences between musicologists and others in understanding what "music" is, and it doesn't lead to the same sort of dynamic between experts and outsiders about (say) whether birds "sing" or not.]

Posted by Mark Liberman at February 2, 2004 02:36 PM