I’m in a pout about this book; I’m conflicted. On the one hand, there are several respects in which it seems to me to be very good. Mithen knows a great deal and he writes well by the received standards of cognitive science (which are not daunting). So his book is both edifying and a pleasure to read. If you’re in the market for a summary of what’s known (a little) and what’s surmised (a lot) about the evolutionary history of our species, I’d be hard put to think of a better one to recommend. Also, and more to the point, the question to which the book wants to address itself is thoroughly fascinating to, as Mithen says, ‘anyone who has an interest in the human condition’. Namely: ‘Why should we be so compelled to make and listen to music?’ And if all that’s not enough, there’s a theory of the origin of language (that again!) thrown in for free.
But then there’s the other hand. It seems to me that Mithen’s book isn’t really about the question it purports to address, and I don’t believe he has thought through some of the large issues that lurk behind that question. In fact, a lot of his discussion depends on undefended theses which, though they are currently quite widely endorsed in the cog. sci. community, I doubt can bear the weight that Mithen wants them to.
Consider, for a start, the questions that motivate the book. What’s so good about music? Why do we care about it so? Why does it move us so? Why is it so pervasive in (it appears) most or all human societies? Music gets under one’s skin and behind one’s defences in a way that nothing else does. I try to be as hard a case as possible: I’m dry-eyed in museums, and I can get through novels and poems without excessive loss of cool. But I’ve been known to weep at the opera. What on earth is that all about?
It’s a promising topic, but Mithen gets off on the wrong foot. I’d have thought the natural way to read ‘what’s so good about music?’ is ‘what is it about music that’s so good?’ By contrast, Mithen reads it teleologically, as ‘what is music good for?’ What is the point of music? What, to put it vulgarly, is in it for us? The consequence is that, while his book contains all sorts of interesting stuff, it has, oddly enough, practically nothing to say about music. A considerable variety of topics that spring to mind as likely to be relevant, and that have traditionally perplexed musicologists and aestheticians, aren’t so much as mentioned by Mithen. How is music structured? What makes it different from mere sound? What makes it different from other art forms? What makes it different from mathematics? How are we to understand the composer’s relation to music, or the hearer’s, or the performer’s, or the critic’s? And so on through a great litany of standard issues. Everyone to his own hobby-horse; there’s no obligation to think of Mithen’s problem as proprietary to musicology or aesthetics. But if it’s the teleology of music that you care about – if your question is ‘what’s the social or biological function of music?’ – you have to face the possibility that it hasn’t got one. Or, that if it does, it’s not because it does that music moves us. I think that’s not implausible. As far as I can tell just on the basis of introspection, I don’t go to the opera for anything at all; I go because I like opera. I would have thought that is pretty generally the case for opera-goers; it could be that everybody but me is bribed to attend, but I doubt it.
So why would a person like Mithen, who is obviously sensible and clearly well informed, take it simply for granted that our interest in music is – must be – basically pragmatic? Thereby hangs a long and unhappy tale. He supposes – rightly, I imagine – that there’s a good case for our interest in music being somehow innate. (Nobody knows, exactly, what that means: something, presumably, about its being genetically carried rather than acquired by any sort of learning experience.) His main argument for the innateness of music parallels a standard argument for the innateness of language: viz its universality across what are, as far as we know, historically unconnected cultures. There are those who would resist this line of thought; who would, indeed, resist any form of nativism about our minds. But I’m not among them; it strikes me as a perfectly plausible working hypothesis that our interest in music belongs to our human nature.
But of course that doesn’t, in and of itself, answer the question of why music matters to us. We have, after all, lots of traits and behaviours that are pretty clearly genotypic but that we don’t much care about. The knee-jerk reflex, for example; or our preference for a cyclical sleep schedule; or our handedness; or, for that matter, our preference for languages that have sentences with segmental structures. In none of these examples – and, surely, there are many others – does the presumption of innateness carry any particular implications about value. I don’t care a bit that my leg jumps if you strike my knee; I’d be perfectly content for it not to. So, whence Mithen’s assumption that if the preference for music is innate, there must be some historical/ teleological account of why we like it? I strongly detect the deadening hand of the psychological Darwinist; that is, of somebody who thinks it’s true quite generally that, if a mental trait is genetically carried, then there must be some adaptationist story about its evolutionary history. If we like music a lot, that must be because, back in the prehistory of our species, there was something that hearing or making music or both did for our likelihood of survival (or of self-preservation, or of transmitting our genes to posterity – different versions are in fashion from one decade to the next). And if it couldn’t be that we like music just for itself, then it couldn’t be that Neanderthals (or whatever) did either. Perhaps they liked it because it was somehow good for their fitness as measured, say, by the number of thriving little Neanderthals that hearing music caused them to have. (If so, things have changed a lot since then: in all the opera houses I’ve been to, copulation during performances is frowned on; and what you splurge on mezzo-sopranos you can’t spend on your kids.)
A lot of people think that if a mental trait is innate then it must be (or have been) adaptive; indeed, that ‘the scientific world view’ requires one to believe something of the sort. It’s not at all clear to me why they think so; but I’d guess it’s because they take for granted that it’s the individual traits of creatures, severally as it were, that are the units of natural selection: traits that can’t pay their way in contributions to fitness are sooner or later selected out. Whether that view is defensible is itself a very large topic and the verdict is by no means in. Suffice it to say for present purposes that alternative assumptions seem at least equally plausible. Suppose, for example, that what selection pressures select for is the fitness of whole organisms. Then what matters for a creature’s survival (or its reproductive success, or whatever) is its relative fitness overall. That would leave open the possibility that, if a kind of creature is pretty fit overall, there may be room for it to acquire traits that don’t themselves contribute to its fitness; or, indeed, traits that militate against its fitness. A passion for opera, for example.
Anyhow, a lot of Mithen’s book is dedicated to imagining a variety of scenarios in which music would have been a useful thing to have Back Then. Often enough his adaptationist story seems, if not convincing, at least as good as any other one might cook up. Shared music might promote social solidarity; lullabies might be good for keeping babies satisfied; music might just help to make a glum Neanderthal feel better. (Mithen doesn’t question what’s so good about feeling better, though consistency might suggest that he should. There are, indeed, evolutionary psychologists who argue that it’s feeling bad that conduces to fitness.) ‘One is immediately struck by the thought of how good it would be if one were always surrounded by happy people – and if they were not happy on their own account, how nice it would be to induce a little happiness into their lives, perhaps by singing them a happy song.’ There, in a nutshell, is my idea of hell.
In any case, I strongly suspect that Mithen’s teleological account of the evolution of music is motivated by his taking psychological Darwinism for granted. Certainly, he doesn’t argue for his key assumption that innate traits must be adaptations. I find that very odd, given, on the one hand, its centrality to his project and, on the other, its face implausibility when music is the topic. I would have thought that it’s among the charms of music that it seems to be so utterly useless. One does weary of wanting things because one wants other things.
I should say a word about Mithen’s treatment of the evolutionary origin of language, which strikes me as also depending on background views which, though currently fashionable, quite possibly can’t be sustained. Mithen thinks that language and music co-evolved from some sort of ‘protolanguage’: that is, from some quite different system of communication. (I suppose that, if you’re pursing an adaptationist agenda, you have to hold something of that sort. An adaptation is, practically by definition, a trait that goes from earlier to later forms in consequence of selection pressures.) In fact, his book is largely about what this protolanguage might have been like and how selection might have moulded it, on the one hand, into the sorts of thing that Homo sapiens talks and, on the other, into the sorts of thing that Homo sapiens whistles. The account Mithen gives is, as you would expect, very speculative. It is also a bit naive in crucial places. For example, he seems pervasively to confuse the question of how language came to be segmented (specifically, into words) with the question of how it came to be compositional (whereby the syntax and semantics of complex expressions are determined by the syntax and semantics of their constituent parts). These are, to put it mildly, quite different issues, and solving the first would not be tantamount to solving the second.
Mithen thinks protolanguage was ‘holistic’ (meanings were typically conveyed by verbalisations that had no meaningful parts) and music-like, both because pitch contour and rhythm were much more important than they are in our languages, and because communicating the emotional states of the speaker was what protolanguage was mainly used for. I am, as I say, mostly unmoved by Mithen’s theory about what protolanguage must have been like or the evidence that he offers in its favour. But I’ll put that aside since there’s a deeper issue that needs to be raised.
Granting (for the sake of argument) that language is an adaptation, why should one suppose that it evolved from a prior system of communication? That is the natural assumption if you take for granted that communication is what language is for. Most contributors to current discussions do, in fact, assume that; though not, to my knowledge, on the grounds of any argument they’ve thus far made public. There is, in any case, at least one other option: that the primordial function of language is not communication but the externalisation of thought. What language is for is to say what we think. Sometimes we do that in order to communicate our thoughts, sometimes to edit them, sometimes just to hear how they sound when they’re said out loud, sometimes for no reason at all. (Cherubino talks about women all the time; he simply can’t stop. If nobody’s there to listen, he talks about them to the trees or to the fountains or to himself.) The point is, if the function of language is to externalise thought, then the problem that the origin of language solved was how to pair thoughts we have with sounds we make. How hard that would be (and how or whether our capacity for doing it evolved) depends on what kind of structure thought has. If, as one might well suppose, the structure of thought is quite like the structure of language, then language might have come about practically overnight. Correspondingly, the pressing question would not be how language evolved but how thought did (about which, it goes without saying, nothing is known).
In fact, I think the thought-first view simply has to be right. Mithen speculates that typical protolanguage utterances meant things like ‘give me that juicy bit over there’ or ‘watch out for the goddamn tiger’ or ‘would you like to drop in and see my cave paintings?’ But consider: how could a noise one makes mean any of these things unless one were already able to represent, in thought, tigers, food and caves (to say nothing of being able to mentally represent both oneself and one’s interlocutor)? And, once that’s granted, it doesn’t take much argument to see that such thoughts must have a segmental structure, a compositional syntax and semantics, and so forth. One can, in short, make sense of thought without language (or rather, of thought without communication), but not, I think, of language without thought. You can’t so much as consider saying what you think unless you can think what you are considering saying.
I’m pretty sure that much of what’s wrong with our current cognitive science is its commitment to not considering that there might be innateness without adaptation and that there might be thought without language. Everything changes if you take seriously either or both of these. Evolutionary psychologists, who generally assume a pre-emptive stance in their relation to ‘the scientific world-view’, almost always take both of these theses to conflict with it. If so, then so much the worse for the scientific world-view, since it’s quite likely that both of them are true.
Send Letters To:
London Review of Books,
28 Little Russell Street
London, WC1A 2HN
Please include name, address, and a telephone number.