The Master and His Emissary: The Divided Brain and the Making of the Western World - Iain McGilchrist (2009)



IN WHAT HAS GONE BEFORE, I HAVE DELIBERATELY FOLLOWED NEUROPSYCHOLOGICAL practice in focussing on a set of discrete tasks or functions, that can be defined and measured, since that is the way we have gathered information about the brain, and it is the way we are used to thinking about it. I now want to look at this material in a different light. I want to draw it together, and suggest that the hemispheric differences are not just a curiosity, with no further significance, a bunch of neuropsychological facts, but actually represent two individually coherent, but incompatible, aspects of the world.

That will lead naturally to an exploration of why it is we have language – for communicating, for thinking, or for what? And what is the role of music? Do the answers to these questions cast any light on the asymmetrical expansions of the left and right hemispheres?


One of the findings mentioned in the last chapter (see pp. 40 and 75 above) was the important discovery by Goldberg and Costa, confirmed by subsequent research, that new experience of any kind – whether it be of music, or words, or real-life objects, or imaginary constructs – engages the right hemisphere. As soon as it starts to become familiar or routine, the right hemisphere is less engaged and eventually the ‘information’ becomes the concern of the left hemisphere only.1

Understandably this has tended to be viewed as a specialisation in information processing, whereby ‘novel stimuli’ are preferentially ‘processed’ by the right hemisphere and routine or familiar ones by the left hemisphere. But this already, like any model, presupposes the nature of what one is looking at (a machine for information processing). What would we find if we were to use a different model? Would perhaps something else emerge?

I want to suggest a different way of looking at the role played by the brain in forming our experience of the world. This involves concerning oneself with the nature of knowledge itself.

We use the word ‘know’ in at least two importantly different senses. In one sense knowledge is essentially an encounter with something or someone, therefore with something ‘other’ (a truth embodied in the phrase ‘carnal knowledge’). We say we know someone in the sense that we have experience of him or her, so that we have a ‘feel’ for who he or she is, as an individual distinct from others. This kind of knowledge permits a sense of the uniqueness of the other. It is also uniquely ‘my’ knowledge. If another person were to ask ‘what is she like?’, you might begin by trying to describe her in a few words (‘quick-tempered’, ‘lively’, modified by qualifying phrases such as ‘quite’, ‘a bit’, ‘very’ and so on), but you'd soon be frustrated by the feeling that these general terms didn't really help get it across. You might resort to retelling instances of things she'd said or done. You might get out a photograph – we learn a lot from faces. But if the questioning continued, you'd have to say: ‘Look, you'll just have to meet her – I'll introduce you.’ It's also ‘my’ knowledge, not just in the sense that I can't pass it on to you, but in the sense that it's got something of me in it. What I know about her comes from the fact that it was I who encountered her. Another person might allow other aspects of her to come forward and might know her as someone rather different. But it would also be odd if everyone who knew her found her to be someone altogether different. That would imply that there was no stable underlying entity to know. We would expect a consensus of those who knew her to emerge. This is the kind of knowledge we think of first when talking about the living.

Coming on the heels of the last chapter, I hope that some of these points may ring a bell. Certain aspects appear familiar. It's the way we naturally approach knowledge of a living being; it's to do with individuals, and permits a sense of uniqueness; it's ‘mine’, personal, not something I can just hand on to someone else unchanged; and it is not fixed or certain. It's not easily captured in words; the whole is not captured by trying to list the parts (‘quick-tempered’, ‘lively’, etc); it has at least something to do with the embodied person (the photograph); it resists general terms; it has to be experienced; and the knowledge depends on betweenness (an encounter). These are all, in fact, aspects of the world ‘according to’ the right hemisphere.

This kind of knowledge derives from a coming together of one being or thing as a whole with another. But there is another kind of knowledge, a knowledge that comes from putting things together from bits. It is the knowledge of what we call facts.2 This is not usually well applied to knowing people. We could have a go – for example, ‘born on 16 September 1964’, ‘lives in New York’, ‘5ft 4in tall’, ‘red hair’, ‘freckles’, and so on. Immediately you get the sense of somebody – who you don't actually know. Either it's a read-out from a police database, or it's one of those cheesy magazine profiles of celebrities (‘latest book read’, etc). What's more, it sounds as though you're describing an inanimate object – ‘chest of drawers, two single over three double, bun feet, circa 1870, 30 × 22 × 28in’ – or a corpse. It is the only kind of knowledge permitted by science (though some of the very best scientists have used subterfuge to get away with the other kind). It concerns knowledge in the public domain – the local train timetable, the date of the Battle of Trafalgar, and so on. Its virtue is its certainty – it's fixed. It doesn't change from person to person or from moment to moment. Context is therefore irrelevant. But it doesn't give a good idea of the whole, just of a partial reconstruction of aspects of the whole.

This knowledge has its uses. Its great strength is that its findings are repeatable. Its qualities are the inverse of those previously outlined, and they are associated with the left hemisphere: an affinity with the non-living; with ‘pieces’ of information; general, impersonal, fixed, certain and disengaged.

Both kinds of knowledge can be brought to bear on the same object, of course. My knowledge of you can be informed by knowing your age, height and place of birth, but that is not in itself at all what I mean by knowing you. These ways of knowing are so different that in many languages other than English they are referred to by different words: the first by, for example, Latin cognoscere, French connaître, German kennen; the second by Latin sapere, French savoir, German wissen – and so on. What I want to suggest is that, just as wissen could sometimes be applied to people and living things, kennen can be applied to a lot more than our acquaintances. This kind of knowing may help us to understand, rather than simply to amass information about, a host of things in the world, animate and inanimate. In fact there is clear evidence that we used to do this in the past, but have lost the habit or perhaps even the ability (see Part II).

To take one example of an apparently non-living entity that appears to require us to know it in the sense of kennen rather than wissen, think of a piece of music. The approach to music is like entering into relation with another living individual, and research suggests that understanding music is perceived as similar to knowing a person; we freely attribute human qualities to music, including age, sex, personality characteristics and feelings.3 The empathic nature of the experience means that it has more in common with encountering a person than a concept or an idea that could be expressed in words.4 It is important to recognise that music does not symbolise emotional meaning, which would require that it be interpreted; it metaphorises it – ‘carries it over’ direct to our unconscious minds. Equally it does not symbolise human qualities: it conveys them direct, so that it acts on us, and we respond to it, as in a human encounter. In other words, knowing a piece of music, like knowing other works of art, is a matter of kennenlernen. Coming to us through the right hemisphere, such living creations are seen as being essentially human in nature. In an earlier book I argued that works of art – music, poems, paintings, great buildings – can be understood only if we appreciate that they are more like people than texts, concepts or things.5 But the perception is ancient: Aristotle, for example, compared tragedy to an organic being.6

What Goldberg and Costa may be uncovering is not just something about novelty and familiarity but about two whole ways of knowing in the two hemispheres. To know (in the sense of kennen) something is never fully to know it (in the sense of wissen) at all, since it will remain for ever changing, evolving, revealing further aspects of itself – in this sense always new, though familiar, in the original sense of coming to belong among our chosen ones, those with whom we stand in close relation, our familia (in Latin literally our ‘household’). To know (in the sense of wissen) is to pin something down so that it is repeatable and repeated, so that it becomes familiar in the other sense: routine, inauthentic, lacking the spark of life. I think what one might deduce from their study is that the first apprehension of anything is by the right hemisphere while it remains new, and, I would suggest, while we are still getting to know it (in the sense of kennenlernen); but that it is soon taken over by the left hemisphere, where it becomes familiar, in the sense that it is now known (gewußt) and therefore certain (gewiß). Knowledge of the whole is all too soon followed by knowledge of the parts.

Jung said that ‘all cognition is akin to recognition’.7 By this he meant that we come to know in the sense of ‘cognise’ (wissen) something only by recognising (erkennen) something we already knew (kennen). In the process it becomes clear, familiar, where before it was latent, intuitive. This is, I believe, an expression of the same process that Goldberg and Costa describe at the neurological level; the new becomes old. In fact Nietzsche goes further and seizes the nub of the matter, when he expresses a similar idea: ‘it is through “knowing” [erkennen, re-cognition] that we come to have the feeling that we already know [wissen] something; thus it means combating a feeling of newness and transforming the apparently new into something old.’8

As Gregory Bateson says, all knowledge has to be knowledge of distinction, and it is of something other than the self.9 Equally one might say that all experience is experience of difference. Even at the sensory level we cannot experience anything unless there is a change or difference: our sensory nerves quickly ‘fatigue’, and we become accustomed, for example, to a smell, or to a sound. Our senses respond to the difference between values – to relative, not absolute, values.10 (It seems that knowledge and perception, and therefore experience, exist only in the relations between things. Perhaps indeed everything that exists does so only in relationships, like mathematics or music: there are aspects of quantum physics that would support such a view.)

This fact, that knowledge comes from distinctions, implies that we can come to an understanding of the nature of any one thing, whatever it might be, only by comparison with something else we already know, and by observing the similarities and differences. However, just as everything changes its nature, however slightly, when it changes its context, what we choose to compare a thing with determines which aspects of it will stand forward and which will recede. Thus comparing a football match with a trip to the betting shop brings out some aspects of the experience; comparing it with going to church brings out others. The model we choose to use to understand something determines what we find. If it is the case that our understanding is an effect of the metaphors we choose, it is also true that it is a cause: our understanding itself guides the choice of metaphor by which we understand it. The chosen metaphor is both cause and effect of the relationship. Thus how we think about our selves and our relationship to the world is already revealed in the metaphors we unconsciously choose to talk about it. That choice further entrenches our partial view of the subject. Paradoxically we seem to be obliged to understand something – including ourselves – well enough to choose the appropriate model before we can understand it. Our first leap determines where we land.11

If we assume a purely mechanical universe and take the machine as our model, we will uncover the view that – surprise, surprise – the body, and the brain with it, is a machine. To a man with a hammer everything begins to look like a nail. But because we can come to know things only in terms of other things we know, every ‘explanation’, however convincing, is merely a model; a comparison of something with something else. All one can say when trying, for example, to interpret scanning data to understand what the brain is doing when I imitate the action of someone else, is that there appears to be some sort of correlation between whatever it is we observe going on in the brain – say a ‘hot spot’ on a brain scan, suggesting increased metabolism in my right frontal lobe – with the experience I am having at the same time. The nature of that correlation, however, remains obscure, because it is instantiated in a unique substance, which is the human body – there isn't anything that that can be compared with, to get a handle on it, except other living bodies, which does not get us much further. You can compare it with a machine, if you wish; but the analogy is bound to be a poor one in every respect except, of course, whatever it is that the body and a machine have in common, and that is all the comparison will reveal (the catch is that to those who have bought into this model as the way to illumination, everything about the body will come to look more and more mechanical, and so the model comes to seem more and more apt: the original choice eventually seems confirmed as a perfect fit). Talk of ‘functions’ and ‘mechanisms’ leads us up this particular garden path. The model of the machine is the only one that the left hemisphere likes; remember that it is specialised in dealing with tools and machines. The machine is something that has been put together by the left hemisphere from the bits, so it is understandable purely in terms of its parts; the machine is lifeless and its parts are inert – the tappets don't change their nature with their context.

I want to try to stand back a bit from the question of which ‘functions’, therefore, the supposedly machine-like hemispheres are performing, and think of them instead more globally as having a disposition, or stance, towards the world – having a ‘take’ on it, if you like. This is to suggest that they may share some aspects of human mental life, I know; but is that really as strange as it sounds – or any stranger than supposing that all they did was compute like machines? If it is true that consciousness, whatever it is, arises from the sheer complexity of self-interconnected neuronal activity,12 why should the hemispheres, the two largest and most densely self-interconnected neuronal masses, each amounting to half the brain, and each capable of sustaining conscious life on its own, not have some of the features of normal consciousness? The separated hemispheres in split-brain patients each have a distinct personality, with characteristic tastes and preferences, according to one of those most closely involved with the study of such patients.13 The unconscious, while not identical with, is certainly more strongly associated with, the right hemisphere (for a fuller discussion of this issue, see pp. 187–8 below). To that extent, it might be expected that the separate hemispheres have distinct personalities and values: as Freud wrote, the unconscious is ‘a particular realm of the mind with its own wishful impulses, its own mode of expression and its peculiar mental mechanisms which are not in force elsewhere’.14

This is just another model, and like all models it should be taken for what it is, a comparison, not an identification. The most one can hope is that it may allow something to stand forward, something different at any rate from what stands forward in the conventional cognitivist account. Applying a different model, not that of a machine but of a person, to each hemisphere reveals different aspects, and enables us to get a sense of each hemisphere as a whole rather than an assemblage.


Since what we think about the world and what we know of it is, whether we like it or not, mediated largely by language, it's worth taking a closer look at the nature of language, and its relationship to the hemispheres. It is likely to tell us a lot about hemisphere differences, about how each hemisphere relates to the world at large, and even about how the hemispheres relate to one another, since it is itself nothing less than a version of the world, a type of experience, which therefore bridges the physical and the abstract, the unconscious and the conscious, the implicit and the explicit.

Language is the province of both hemispheres and, like everything else, has different meanings in either hemisphere. Each uses it differently, and different aspects of it stand out in the use that either hemisphere makes of it.

Let's return to the structure of the brain and take another look at that strange asymmetry in the left parietal region, where language is said to reside. Isn't that obviously what it's for? What's the problem with that explanation?

While it is true that the left hemisphere expansion is now associated with language functions,15 there are difficulties with the belief that it is language that necessitated the expansion.16 For one thing, fossil records of primitive humans from the earliest periods, long before anthropologists believe language developed,17 already show this typical pattern of brain asymmetry.18 Even more striking is the fact that some of the great apes, and possibly other large primates such as baboons, which clearly have no language,19 already show a similar asymmetry to that of the human brain, with enlargement in the same area of the left hemisphere that in humans is associated with language.20 The planum temporale, which in humans is certainly associated with language, and is generally larger on the left than on the right, is also larger asymmetrically, also on the left, in orang-utans, gorillas21 and chimpanzees.22 And Yakovlevian torque, too, is present not only in fossil humans, but in the great apes.23

What is more, now that we know more about the functioning of our own brains, we know that it is not actually true that language is subserved by one hemisphere: its functioning is distributed across the two. If it is true that most syntax and vocabulary, the nitty-gritty of language, are in most subjects housed in the left hemisphere, it is nonetheless the right hemisphere which subserves higher linguistic functions, such as understanding the meaning of a whole phrase or sentence in context, its tone, its emotional significance, along with use of humour, irony, metaphor, and so on. But if it is the right hemisphere that, in linguistic terms, paints the picture, it is still the left hemisphere that holds the ‘paint box’. Following a left-hemisphere stroke, the right-hemisphere painter has lost his materials. Hence the old view that the left hemisphere was ‘dominant’: in its absence no picture is painted – there is no coherent speech. But the argument that language had to be held together in one place, thus explaining the left-hemisphere expansion, just doesn't hold water.

So what is this expansion in the left hemisphere about? Perhaps, it has been suggested, it is a consequence of right handedness. But this begs a further question, namely why we should have developed right handedness. The usual assumption is that, man being the tool-making animal, extra skill was needed in the manufacture of such tools, requiring specialisation. But it is not obvious why skill is best acquired in one hand only. Skilful operators could be even more skilful if they could use both hands equally well, and the brain is not subject to some economic regulation that means that the development of one hand must be at the expense of the skill acquisition of the other. It is, however, an evolutionary fact that, for using and making, we have tended strongly to prefer the right hand, which is controlled by the left side of the brain – in fact by part of the brain that is, as it happens, very close to Broca's area, the part of the left hemisphere that has come to subserve the expressive power of syntax and vocabulary, the names of things and how we put them together.

Moreover if we look once again at the higher apes, it also turns out that some of them begin to prefer the right hand to grasp things – despite the fact that, though they may use sticks and stones, they are certainly not tool-makers in the human sense.24 Any asymmetry in their brains is unlikely to be due to the need for brain space to house the complex skills of tool making. Their grasping right hands must be a sign of something else.

Most bizarrely, it would seem that it is not an expansion at all. It's just that there is a deliberate inhibition of expansion in the corresponding area in the right hemisphere. And we even know the genes that do it. The researcher who found them comments: ‘It is safe to assume that the asymmetry that ended up leading to language is unlikely to have appeared because of language … It is likely to have appeared for some other reason and basically got co-opted by language.’25

Lateralisation of brain function, and asymmetry of its structure, occurred without language or tool design. I am not saying that, once hemisphere asymmetry came to accelerate in humans, it had nothing to do with language or handedness: it obviously did.26 My point is only that these could not have been the drivers, the origin of the phenomenon; the asymmetries associated with language and handedness must be epiphenomena of something else, something more fundamental, more primitive. What was that?


In order to understand this asymmetry better, we need to step back for a moment and ask ourselves why we have language at all. Because the world without it is so unimaginable, and because language is the medium through which we appraise all things, including language itself, it is harder than it may at first seem to bring it into focus. Just what sort of a thing is language? What sort of plan might it form part of, and has it got anything in common with handedness?

Fairly obviously, one might think, language must have developed for communication. But that is not as obvious as it seems. Some 300–400,000 years ago or longer, homo heidelbergensis, the common ancestor of homo sapiens and homo neanderthalensis, had a large brain and a vocal apparatus comparable to those of modern humans, and, although we cannot be sure of the earliest date such features arose, it may well have been as long ago as about half a million years.27 However, the evidence suggests that we did not develop the degree of sophisticated symbol manipulation that language requires until a much later point, possibly as little as 40,000 years ago, but at any rate not earlier than a mere 80,000 years ago, when the first cultural artefacts, along with evidence of visual representation, suddenly and profusely arise, and humans began to adopt ritualised burial of the dead.28 It would seem, then, that for most of human history, despite a large brain and presumably high intelligence,29 they managed to communicate satisfactorily without language as we understand it. Admittedly they were not civilised in the true meaning of the word. But they survived and thrived as social animals, living in groups. How did our ancestors communicate adequately, if not by language?

Addressing these questions necessitates looking at yet another curiosity in what the fossil record tells us.

We know that spoken language is dependent, not only on sufficient brain space to house the dictionary and grammar, but on quite specific features of the vocal apparatus (not just the vocal cords, but the articulatory bits and pieces of the tongue and mouth) enabling us to articulate a wide range of sounds, as well as on a remarkable degree of respiratory control, allowing us to sustain long, fluent, articulated phrases, and to modulate intonation subtly over the length of a single breath. All known languages require these features. Monkeys and apes do not have any such control, which is one of the reasons why attempts to train them to speak have been so unsuccessful. Birds alone can imitate human speech, while our nearest relatives cannot hope to do so: the birds' vocal apparatus, or syrinx (literally, ‘flute’), and their sophisticated control of respiration, explains why.30

When did humans develop these capacities? It might be thought a hopeless task to assess at what point in human history our ancestors developed the sophisticated control of vocal apparatus and respiration that we now possess. However, some ingenious observations allow a reliable inference to be made. In order to reach the tongue, the nerve which supplies it, the hypoglossal nerve, has to pass through an opening in the base of the skull, called the anterior condylar canal. The amount of work a nerve has to do is reflected in its size; in turn the size of the hole through which it passes indicates the size of the nerve. So by measuring the size of the canal in the base of the skull, we can get a very good idea of how much articulatory work the tongue of the skull's ‘owner’ had to do. Similar considerations apply to the thoracic vertebral canal in supplying the nerves that control respiration to the muscles of the chest wall. And what we find, as we might expect, is that apes and monkeys have much smaller canals, in relation to the nerves both ofarticulation and of respiration, than modern humans. But, and here is the thought-provoking oddity, examination of the earliest human skeletons, from long before the time we believe language arose, reveals canal sizes almost indistinguishable from those of modern humans. Why is that?

The most likely answer is a surprise, and requires a bit of a frame shift for most of us. For the explanation of this sophisticated control and modulation of the production of sound, in the absence of language as we know it, has to be that it was for a sort of non-verbal language, one in which there was intonation and phrasing, but no actual words: and what is that, if not music?

There are significant similarities between music and language, suggesting at least a common origin. For example, many subtle aspects of language are mediated by regions of the right hemisphere which also mediate the performance and experience of music. Furthermore these right hemisphere regions are the homologues of areas in the left hemisphere that are involved with language production and comprehension – they are in the ‘same’ position on the other side of the brain. Music and language have a shared architecture, built out of intonational phrases related by a kind of ‘syntax’, although the syntax of music has more to do with the overall shape of the whole piece over many minutes (or, in the case of Wagner,31 hours) than with the specific relationship of rapidly successive elements in a linear progression. In both music and speech, the phrase is the basic unit of structure and function, and both speech phrases and musical phrases have melody and rhythm, which play a crucial role in their expressiveness. There is even a close semantic relationship between music and language: musical phrases convey specific meanings that, if required, we will intuitively associate with specific words.32

When it comes to understanding the origins of language, however, there is less agreement, and speculation has followed one of three paths. There are those who believe that music is a useless spin-off, or epiphenomenon, of the development of language; there are those, on the contrary, who believe that language itself developed out of musical communication (a kind of singing); and finally there are those who hold that music and language developed independently but alongside one another, out of a common ancestor, which has been dubbed ‘musilanguage’.33 It seems to me that this last option is hard to distinguish from the ‘music came first’ position. That is because, while ‘musilanguage’ may not have been very sophisticated music, it must have been more like music than like referential language, and, for it to have been a ‘language’ in any sense, it must have relied on what we think of as the musical aspects – the non-verbal aspects – of language, such as pitch, intonation, volume, rhythm and phrasing. The very existence of the concept of ‘musilanguage’ merely points up how much the musical aspects of language do contribute to meaning, in that they could in themselves provide the basis for communication of meaning.

The evidence of the fossil record is, as I say, that the control of voice and respiration needed for singing apparently came into being long before they would ever have been required by language. But is there any reason, apart from this, why we should adopt the view that music came first?


There are, if nothing else, some indications on the matter. In the first place, the ‘syntax’ of music is simpler, less highly evolved, than that of language, suggesting an earlier origin. More importantly, observation of the development of language in children confirms that the musical aspects of language do indeed come first. Intonation, phrasing and rhythm develop first; syntax and vocabulary come only later. Newborns are already sensitive to the rhythms of language;34 they prefer ‘infant-directed speech’ – otherwise known as ‘baby talk’ – which emphasises what is called prosody, the music of speech. In response to this, mothers expand the pitch excursions, broaden the repertoire and raise the overall pitch of their speech, as well as slowing the tempo and emphasising its rhythm, as soon as their child is born.35 Newborn infants can distinguish the timbre and intonation of their mother's voice, and prefer it to any other;36 and can distinguish the unique intonation of their ‘mother’ tongue, which again they prefer to others.37 These capacities for distinguishing the characteristic inflections of a language, or even of an individual speaker, are not signs of an inborn talent for language as such: they rely on aspects of right-hemisphere holistic processing capable of making fine discriminations in global patterns and having little to do with the analytic processing of language by the left hemisphere.38 Indeed even primates can identify individual voices, using such features. These processes, then, in newborns have more to do with the activation of areas of the brain which subserve the non-verbal, the musical, aspects of speech. There is a (not wholly reliable) principle that ‘ontogeny recapitulates phylogeny’, that, in other words, the development of individuals of a species follows a similar course to that taken by the development of the species as a whole: a simple example is the early development by the human embryo of a tail, which it later loses. To the degree that this principle holds here, then music came before language. An observation to this effect was made even by Salomon Henschen: ‘The musical faculty is phylogenetically older than language; some animals have a musical faculty—birds in a high degree. It is also ontogenetically older, for the child begins to sing earlier than to speak.’39

Ultimately music is the communication of emotion, the most fundamental form of communication, which in phylogeny, as well as ontogeny, came and comes first. Neurological research strongly supports the assumption that ‘our love of music reflects the ancestral ability of our mammalian brain to transmit and receive basic emotional sounds’, the prosody and rhythmic motion that emerge intuitively from entrainment of the body in emotional expression: ‘music was built upon the prosodic mechanisms of the right hemisphere that allow us affective emotional communications through vocal intonations.’40 Presumably such ‘mechanisms’ were highly important for group survival. They were also likely to have deep roots: ‘the deeply emotional stirrings generated by music’, writes the influential anthropologist Robin Dunbar, ‘suggest to me that music has very ancient origins, long predating the evolution of language.’41

This conclusion has not been universally welcomed. There are a number of reasons, but one stands out, at least as far as concerns geneticists. Developments must demonstrate evolutionary advantage. Language, it is reasoned, gives a huge advantage in the power it confers to its possessor: but what has music to do with power – what advantage can it yield? It doesn't apparently put you in a position to deliver a knockout blow to the opposition, and doesn't look like a way of pushing your genes (despite unconvincing appeals to what one might call the ‘Tom Jones factor’). So music has been seen as a pointless ‘exaptation’ of language: that is to say, an adaptation of a skill, originally developed for its competitive advantage in one area, to a quite different purpose.42 Thus typing could be seen as an exaptation of the digital skill developed for making tools: it was not the pressure to out-publish one's colleagues that caused the skill to develop in the first place, any more than we have legs in order to give employment to tailors. Music has to be, on such an account, an irrelevant spin-off from something with more of a competitive cutting edge – namely, language: Steven Pinker certainly sees it as such, and even suggests that music is as meaningless and self-indulgent as pornography or a taste for fatty food.43 Nonetheless the evidence does not stack up in favour of music being an exaptation of language – rather the reverse. If language evolved later, it looks like it evolved from music. So the evolutionary problem remains (I hope to offer a possible resolution of the problem in due course).

It is not only geneticists who may have difficulty with the idea. We now tend to think of music as peripheral, if not useless. We are all inclined to raise an eyebrow when our ancestors are referred to as the ‘singing Neanderthals’.44 But in fact many theorists of language, including Rousseau in the eighteenth century, von Humboldt in the nineteenth century and Jespersen in the twentieth, have thought it likely that language developed from music, so that the theories of Mithen and others in the twenty-first century do not come out of the blue.45

That we could use non-verbal means, such as music, to communicate is, in any case, hardly surprising. The shock comes partly from the way we in the West now view music: we have lost the sense of the central position that music once occupied in communal life, and still does in most parts of the world today. Despite the fact that there is no culture anywhere in the world that does not have music, and in which people do not join together to sing or dance, we have relegated music to the sidelines of life. We might think of music as an individualistic, even solitary experience, but that is rare in the history of the world. In more traditionally structured societies, performance of music plays both an integral, and an integrative, role not only in celebration, religious festivals, and other rituals, but also in daily work and recreation; and it is above all a shared performance, not just something we listen to passively.46 It has a vital way of binding people together, helping them to be aware of shared humanity, shared feelings and experiences, and actively drawing them together. In our world, competition and specialisation have made music something compartmentalised, somewhere away from life's core. So Oliver Sacks writes:

This primal role of music is to some extent lost today, when we have a special class of composers and performers, and the rest of us are often reduced to passive listening. One has to go to a concert, or a church or a musical festival, to recapture the collective excitement and bonding of music. In such a situation, there seems to be an actual binding of nervous systems …47

But if it should turn out that music leads to language, rather than language to music, it helps us understand for the first time the otherwise baffling historical fact that poetry evolved before prose.48 Prose was at first known as pezos logos, literally ‘pedestrian, or walking, logos’, as opposed to the usual dancing logos of poetry. In fact early poetry was sung: so the evolution of literary skill progresses, if that is the correct word, from right-hemisphere music (words that are sung), to right-hemisphere language (the metaphorical language of poetry), to left-hemisphere language (the referential language of prose).

Music is likely to be the ancestor of language and it arose largely in the right hemisphere, where one would expect a means of communication with others, promoting social cohesion, to arise.


The predominance of language, and, above all, of the effects of the written word, may itself have contributed to the decline of music in our culture. (I hope in later chapters to show that the culture of the written word tends inevitably towards the predominantly left-hemisphere phenomenon of a competitive, specialised and compartmentalised world.) We may find it initially hard to accept the primacy of music, since we are trapped inside a culture that is so language-determined and language-dependent that we cannot imagine it being any other way. Because the part of communicating that we are aware of lies in the choice of words, we imagine wrongly that that must be where most, or perhaps even all, communication lies. What we are not conscious of, and need for most purposes to remain unconscious of, is that the majority of the messages we communicate are not in words at all. Animals communicate with one another, and even co-operate, without language, so why shouldn't we?

Some animals adopt a form of ‘musilanguage’, using intonation, not just body language, to communicate with humans: look at the domestic dog. Amongst one another they communicate preferentially by scent, and body language. But they have achieved awareness of the fact that intonation is an important part of human communication. Good vocal communicators as they are, the possibilities for them are limited: they have neither the range of concepts to convey, nor the vocal or respirational apparatus that early humans would have possessed with which to convey them.

In case one is tempted to think that music could never provide a flexible or extensive enough means of communication for hominids, one should remember that the extensive social lives of some of the most intelligent non-human animals, not only bonobos, but aquatic mammals, such as whales and dolphins, down to and including the complex attack manoeuvres of killer whales, are co-ordinated entirely by what one might call ‘music’ – a ‘language’ of pitch, intonation and temporal relation.49 And our pre-lingual children communicate with us – not always, it is true, with desired clarity, but effectively enough. Even left-hemisphere stroke sufferers, who have suddenly and devastatingly lost the use of language, can learn to communicate an array of intentions and meanings, as well as, obviously, emotions, despite the fact that they are at the immense double disadvantage that they did not grow up learning instinctively and intuitively the skill of communicating without words, and that they live in a culture which has organised itself wholly around communication through language.

Perhaps the most striking evidence, though, is that there are extant tribes in the Amazon basin, such as the Pirahã, a hunter-gatherer tribe in Brazil, whose language is effectively a kind of song, possessing such a complex array of tones, stresses, and syllable lengths that its speakers can dispense with their vowels and consonants altogether and sing, hum or whistle conversations.50

For our primate ancestors, who clearly had no speech, body language played a vital role in social cohesion, especially in prolonged sessions of mutual grooming. One theory is that singing, a sort of instinctive musical language of intonation, came into being precisely because, with the advent of humans, social groups became too large for grooming to be practical as a means of bonding. Music, on this account, is a sort of grooming at a distance; no longer necessitating physical touch, but a body language all the same. And, the theory goes, referential language was a late evolution from this.51 It is estimated that even now over 90 per cent of communication between humans is by non-verbal means, through body language and perhaps especially through intonation. Communication, after all, does not only mean the kind of language we use to talk about things. Music is communication – but it speaks to us, not about things. It does not refer (to a third party): it has an ‘I–thou’ existence, not an ‘I–it’ existence.

In fact, even without the anthropological evidence, we might well be doubtful that language was needed for communication, if for no other reason than that language, unlike more intuitive, musical, forms of communication, is the perfect medium for concealing, rather than revealing, meaning. The fact that we now communicate mainly through language shouldn't blind us to the equally important fact that we do not need it for most types of communication. Most of our communication goes on without it anyway, whether we like it or not, and language may even make some kinds of communication more problematic.


Surely, though, it may be said, even if language isn't strictly necessary for communication, its advent was necessary for humans to become the thinking beings they are, capable of forming concepts, making judgments, taking decisions, solving problems, all that is characteristic of our highest functions? Well, not really – in fact, not at all. The belief that one cannot think without language is yet another fallacy of the introspective process, whereby thinking in words about language only serves to confirm the importance of the verbal process. When we consciously introspect, or retrospect, on our own thought processes, and try to construct what happens, how the mind works, we can do so only as we would under those circumstances try to achieve the task, consciously, putting it in words. But the mind is not like this. We carry out most mental processes that would normally constitute what we mean by thinking without doing anything consciously, or in language, at all. We make sense of the world, form categories and concepts, weigh and evaluate evidence, make decisions and solve problems, all without language, and without even being consciously aware of the process.

Indeed, many of these things can be achieved satisfactorily only if we do not become too explicitly aware of the process, which would otherwise have a limiting and inhibiting effect. Many examples exist of famous scientific problems that were solved without language. After much cogitation, Kekulé seized the shape of the benzene ring, the foundation of organic chemistry, when the image of a snake biting its tail arose from the embers of his fire; Poincaré, having spent 15 days trying to disprove Fuchsian functions, suddenly saw their reality, as, after a cup of black coffee, ‘ideas rose in crowds – I felt them collide until pairs interlocked’; later their relation to non-Euclidean geometry occurred to him at the moment he put his foot on a bus, though he was in the middle of a completely unrelated conversation (‘on my return to Caen, for conscience’ sake I verified the result at my leisure’). The structure of the periodic table of the elements came to Mendeleyev in a dream. Einstein wrote that ‘the words or the language, as they are written or spoken, do not seem to play any role in the mechanism of my thought …’52 Similar points were made by Gauss and Helmholtz. Mathematical thinking, which is principally right-hemisphere-mediated, takes place in three dimensions. Rudolf Arnheim wrote in his classic work, Visual Thinking, as powerful today as when it was written in 1969: ‘What we need to acknowledge is that perceptual and pictorial shapes are not only translations of thought products but the very flesh and blood of thinking itself.’53 Note that expression: the ‘very flesh and blood of thinking’. I will have more to say about this later in a discussion of language and the body. But the point is this: the fact that we are more aware of those times when we do think explicitly to ourselves in words – and now conceive of all thought as taking place in words – should not deceive us into believing that language is necessary for thought. It could even be an impediment to it. Most forms of imagination, for example, or of innovation, intuitive problem solving, spiritual thinking or artistic creativity require us to transcend language, at least language in the accepted sense of a referential code. Most thinking, like most communication, goes on without language.

What is more, in evolutionary terms, thought, including concept formation, clearly predates language. Quite apart from the already mentioned existence of sophisticated pre-lingual hominids, we know that animals can think and form concepts. Any sentient being living in an environment where it needs to defend itself from predators and find enough to eat must be capable of forming concepts and placing things in categories. Without it, it would be forced to start from scratch with every encounter with friend or foe, with plant or poison, and wouldn't last long. These assumptions are borne out by the evidence of studiesdemonstrating that indeed categorical perception is not unique to humans, and cannot therefore depend on language.54 The ability to categorise is in fact almost universal.55 Pigeons, for example, can categorise different types of leaves, or fish, or people. They can even distinguish a human face in a crowd, and artificial from natural objects.56

In fact they have also been shown to be able to categorise cartoon pictures, and to discriminate examples of modern art, for example a Monet from a Picasso.57 Trained to peck a key that will give access to hempseed in response to one or other type of picture, they were able to choose appropriately, and could even begin to generalise from Monet to Renoir, and from Picasso to Braque. And their discrimination was not based on one dimension, such as colour, or contour, since they were also tested in black and white and out of focus. Similarly pigeons can tell Bach from Hindemith, or Stravinsky.58 Mind you, carp can tell blues from classical music; even – and I suspect the investigator must have indulged a sense of humour here – Muddy Waters from the Trout Quintet (in case you're wondering, they had to press a puck with their snout: a correct choice yielded a food pellet).59 And animals are more broadly capable of mental representation, of the ability to generalise, and to form categories and reason, though they do not have language.60Dogs even understand the apparently arbitrary connections between words and actions or things.

Concept formation, together with the ability to see the relations between things and events, and the ability to link concepts with signs of some kind, presumably physical movements, arose through natural selection, and formed the substrate of language long before the emergence of modern humanity. Forward planning, hitherto thought to be a hallmark of human cognition, is clearly present in birds that have no language (a point worth making since, for example, Irene Pepperberg's African grey parrot, Alex, was able to communicate, plan and reason – but he had quite a vocabulary).61 Even ‘theory of mind’, the ability to attribute mental states to others, which has become the shibboleth of complex, multilayered thought – since children are commonly said not to acquire it till about the age of four, and some subjects, particularly those with autism, may never acquire it at all – is intact in human subjects who have lost language,62 and may be present to a degree in chimpanzees and primates.63 Clearly, therefore, ‘theory of mind’ cannot depend on language, either.

Once again, not just animals, but the young of our own species, indicate that it is wrong to assume that meaning depends on language, though our conscious left hemisphere may be unable to conceive of meaning that is not conveyed in words. Meaning, and the ability to communicate meaning, antedates language in human development. What is more such meaning is mediated, not by the left, but by the right hemisphere. The attunement of emotionally expressive facial expressions between mother and baby in the child's early maturing right hemisphere means that, long before the infant either comprehends or speaks a single word, it possesses an extensive repertoire of signals to communicate its internal state.64

That thought does not depend on language is also demonstrated by those who have developed aphasia – lost the power of speech. Those that recover are able to describe their experience and we are fortunate to have the description left by Jacques Lordat, a professor of physiology at the University of Montpellier, a man who, somewhat ironically, had made a study of aphasia. In 1843 he published a paper in which he gave a detailed description of an aphasic episode, lasting several weeks, that he himself had experienced following a stroke. Lordat noticed that

when I wanted to speak I could not find the expressions that I needed … the thought was all ready, but the sounds that had to express it as intermediary were no longer at my disposition … I was unable to accept … the theory that verbal signs are necessary, even indispensable for thought.65

In fact subjects who have suffered a stroke demonstrate that even complex reasoning and mathematical calculation do not depend on language.66 Syntactic structure is distinct from logical structure: subjects that have lost their grasp of syntax following a left-hemisphere stroke remain able to use sophisticated thought processes, as complex as the structure of complex syntax, and can calculate and reason perfectly well.67 Patients with semantic dementia, too, can perform calculations, sometimes exceptionally well.68

That we do not need words in order to hold concepts is also demonstrated by some beautiful research carried out amongst tribal peoples with quite differently structured vocabularies from our own. It turns out that, for example, numerical concepts do not depend on the pre-existence of linguistic terms for them. Tribes with limited words for numbers (such as the Amazonian tribe, the Mundurukú, who have no word for a value greater than three) can succeed in arithmetical tasks that involve values as great as 80. Some members of the Mundurukú speak both their own language, with its extremely limited vocabulary for numbers, and Portuguese, in which there are an unlimited range of number words, while others speak only Mundurukú. The two groups of speakers nonetheless perform comparably on calculation tasks (whether or not, in other words, their number vocabulary goes further than ‘3’), and both groups perform as well as French-speaking controls; and this is the same for adults and children.69 The idea of ‘recursive infinity’ – that is, that you can keep adding one indefinitely to get ever larger numbers – comes naturally to us, even when it does not figure in our established symbolic systems, though Chomsky and his colleagues have claimed that it is derived from the recursive property of natural languages.70 The Oksapmin of Papua New Guinea, who use body-parts as a counting system, quickly adapted the system to a generative counting rule (i.e. being able to count higher and higher, up ‘levels’ of magnitude clearly too great for there to be an adequate number of body parts) when times changed and money was introduced to their system.71 It would seem that such key concepts are innate, rather than being culturally imposed as a product of a learnt language, and this is further borne out by research on the way children develop concepts of number.72

But, it may be said, surely we need language in order to discriminate, or at any rate to make fine discriminations, among the things we experience: how can we organise experience if we do not have ‘labels’ for what it is we perceive? This also turns out to be untrue. For example, not having a word for a colour does not mean we can't recognise it. Quechi Indians have only five colour terms, but can differentiate hues as well as any Westerner; and, nearer to home, Germans, who do not have a native word for the colour ‘pink’, can of course recognise it all the same.73 However, words can influence our perceptions. They can interfere with the way in which we perceive colours – and facial expressions, for that matter – suggesting that colour words can create new boundaries in colour perception, and language can impose a structure on the way we interpret faces.74 In other words, language is necessary neither for categorisation, nor for reasoning, nor for concept formation, nor perception: it does not itself bring the landscape of the world in which we live into being. What it does, rather, is shape that landscape by fixing the ‘counties’ into which we divide it, defining which categories or types of entities we see there – how we carve it up.

In the process, language helps some things stand forward, but by the same token makes others recede. Observation of child development confirms this:

It has been suggested that our concepts are determined by the language that we speak (the Sapir–Whorf hypothesis). However, this is no more than a half or quarter truth. Children certainly often get the concept first and then quickly learn the word to describe it, which is the wrong way round from the Sapir–Whorf point of view. Moreover there is evidence that five-month-old babies have a concept, to do with tightness of fit, which they subsequently lose if their native language does not embody the same concept.75

The Sapir–Whorf hypothesis has partial truth – if you don't have the word, you are likely to lose the concept; but this research demonstrates that the concept can arise without the word, and is therefore not dependent on it. So thinking is prior to language. What language contributes is to firm up certain particular ways of seeing the world and give fixity to them. This has its good side, and its bad. It aids consistency of reference over time and space. But it can also exert a restrictive force on what and how we think. It represents a more fixed version of the world: it shapes, rather than grounds, our thinking.76

Language is not essential, then, to communication or to thinking – and may interrupt or interfere with both. So we return to the question: why did language actually come about, and what is it for?

There may be a clue in the fact that the other conventionally defining human characteristic, apart from language, is tool making, and that this is associated with the development of right-hand skill: interestingly enough with the same area of the left hemisphere as semantics and syntax. But even if the development of right-hand skill had caused the outgrowth of the left petalia (which it can't have done, because the outgrowth long antedates it) that would still leave us with the question of why the process had to be asymmetrical. Unless, that is, the agenda of the two hemispheres are quite different.


The location of grasp in the left hemisphere, close to speech, is not accidental and tells us something. We know from experience that there are many connections between the hand and language. For example, there clearly is a close relationship between spoken language and the wealth of gesture language that often accompanies it. In normal subjects, restricting hand movement produces an adverse effect on the content and fluency of speech.77 Ramachandran even reports the case of a young woman, who was born without upper limbs. She experiences phantom arms; and the fact that she has phantom arms at all, replicating a number of such findings in the congenitally limbless, is interesting enough (phantoms are usually thought of as being the residual of a limb that is lost, that in other words must have been there originally).78 But, even though these phantom arms do not, for example, swing by her side as she walks, she cannot stop them gesticulating when she speaks. Even though she has never been able to use an arm or hand, speech activates these areas of her brain.79

At the neurophysiological level, too, it turns out that there are similarities between the skills required for speech production and those required for hand movement, specifically movement of the right hand.80 In fact, according to Marcel Kinsbourne, language develops specifically in relation to ‘right-sided action and, particularly, rightward orienting’.81 It is, according to him, an ‘elaboration, extension and abstraction of sensorimotor function’, originating in a proto-language formed by the ‘utterances that were coincident with and driven by the same rhythm as the movement in question’.82 As if to confirm the close connection between language and the body, especially pointing and grasping movements of the right hand, babies and young children can be seen to point while they babble, and the child ‘always points while naming and does not name without pointing – stretching out the right hand … Babbling can also be heard in conjunction with the motor sequences that are sequelae of the orienting response – locomotion, grasp, manipulation.’83 And the association holds not just for the child: even in the adult, language, gesture, and bodily movement are ‘different actualizations of the same process.’84

Manipulospatial abilities may have provided the basis of primitive language, and such abilities and referential language require similar neural mechanisms.85 The syntactic elements of language may well derive from gesture.86 And not just from gesture, but from the more functional, more manipulative, hand movements: tool making and speech are both ‘serial, syntactic and manipulatory behaviours based on complex articulations of biomechanical patterns.’87 In fact, so strong is the connection that one theory is that referential language may have evolved, not from sounds at all, but actually direct from hand movements – not only that, but specifically from motions to do with grasping.88 The closeness of function is imaged in anatomy by the proximity of the area for speech and the area of the brain designed to promote grasping. As mentioned, Broca's area, the motor speech area of the frontal lobe, involves certain specialised nerves called mirror neurones which are involved in finger movements, and are also activated in watching others carry out hand movements.89

This complicity of language and grasping movements of the hand is not just an interesting neurophysiological and neuroanatomical finding. It is intuitively correct, as evidenced by the terms we use to describe linguistic comprehension and expression. It is not an accident that we talk about ‘grasping’ what someone is saying. The metaphor of grasp has its roots deep in the way we talk about thinking in most languages (e.g. the various Romance derivatives of Latin com-prehendere, and cognates of be-greifen in Germanic languages). In his fascinating study of the human hand, the German-speaking Hungarian psychologist Géza Révész writes:

In German the notion of ‘handeln’ embraces all meaningful and goal-directed human activities. It characterises unequivocally the total personality of man. This idea is not limited to external manipulation, that is actions which effect changes in the outer world. It also includes inner action, the purposeful activities of the mind. In his mode of manipulation man experiences his real ‘I’. Through it he acquires power over physical nature, gathers a rich fund of material from experience, enlarges his range of effectiveness, and develops his capacities … Impulses, aspirations, wishes, decisions press for realisation, and this takes place chiefly through the mediation of the hand … If we wish to convey that we have acquired something mental, we say that we have grasped it.90

He points to words such as Erfassen, Begriff, begreiflich, Eindruck, Ausdruck, behalten, auslegen, überlegen. We have them too in English – not just grasp and comprehend, but words such as impression, expression, intend, contend, pretend (from Latin tendere, to reach with the hand).

Among other things which Révész draws attention to is the fact that, though touch is the first, most basic and most convincing of the senses (the simplest organisms have only tactile hairs or cilia), it can nonetheless provide only a piecemeal image of something. Handling something gives one bits of information at a time, and one has to put together the parts: it does not deliver a sense of the whole. He also points out that in distinguishing things with the hand it is a question of what type of a thing, and not of which particular one.

All of this – this grasping, this taking control, this piecemeal apprehension of the world, this distinguishing of types, rather than of individual things – takes place for most of us with the right hand. And so it is not surprising that hidden in these reflections are clues to the nature of left-hemisphere processes. In all these respects – not just in the taking control, but in the approach to understanding by building it up bit by bit, rather than being able to sense the whole, in the interest in categories of things, rather than in individuals – grasp follows a path congenial to the operations of the left hemisphere. It is also through grasping things that we grant things certainty and fixity: when they are either uncertain or unfixed, we say we ‘cannot put our finger on it’, we ‘haven't got a hold of it’. This too is an important aspect of the world according to the left hemisphere. The idea of ‘grasping’ implies seizing a thing for ourselves, for use, wresting it away from its context, holding it fast, focussing on it. The grasp we have, our understanding in this sense, is the expression of our will, and it is the means to power. It is what enables us to ‘manipulate’ – literally to take a handful of whatever we need – and thereby to dominate the world around us.

And, as if to confirm the deep relationship between the left hemisphere and instrumentality, an attitude of grasping and use, it will be remembered that tool use is preferentially represented in the left hemisphere even in a left-handed individual. This is surely a remarkable finding. Even though the individual's brain is so organised that the right hemisphere governs day-to-day actions using the left hand, concepts of tool use preferentially activate not the right hemisphere but the left.91 And still further evidence, just as remarkable, shows that, again even in left-handers, actions specifically of grasping are associated with left-hemisphere control – the concept is separate, therefore, from control of the hand as such.92 Meanwhile exploratory, rather than grasping, motions of the hand activate the right superior parietal cortex, even when the hand that is doing the exploring is the right hand.93 These findings from imaging are in keeping with clinical experience. Subjects with right-hemisphere damage tend to grasp anything within reach, or even brandish their right hand about in empty space, as if searching for something to grasp. And this is not just the release of a primitive reflex: unlike those subjects who exhibit a grasp reflex, they are able to loosen their grip immediately, when asked to do so. It is volitional.94 And the contrast with those with left-hemisphere damage, therefore relying on their right hemisphere, could not be more stark:

The patient, when asked to copy the examiner's gestures, tries to put his own hands on the examiner's … When his hands are brought into action, it seems as if they are seeking not to remain isolated, as if trying ‘to find companionship in something that fills up the empty space.’95


I am not the first to have surmised that referential language originated in something other than the need to communicate. The philosopher Johann Gottfried Herder, who in 1772 published one of the most important and influential essays on the origins of language, noticing that what I would call the ‘I–thou’ element in communication at the most intuitive level, the empathic force that is present in music, is hardly characteristic of human language, concluded that ‘language appears as a natural organ of reason’.96 That might require qualification, since, as I have emphasised, reasoning goes on without it; but what he points to here is the importance of language primarily as an aid to a certain particular type of cognition. Nearer our own time, the distinguished American neuroscientist Norman Geschwind ventured that language may not, after all, have originated in a drive to communicate – that came later – but as a means of mapping the world.97 I would agree with that and go further. It is a means of manipulating the world.

Understanding the nature of language depends once again on thinking about the ‘howness’, not the ‘whatness’. The development of denotative language enables, not communication in itself, but a special kindof communicating, not thinking itself, but a special kind of thinking.

It is certainly not so important for personal communication within a relationship, and may even be a hindrance here. Telephone conversations, in which all non-verbal signals apart from some partially degraded tonal information are lost, are unsatisfactory not only to lovers and friends, but to all for whom personal exchange is important; one would not expect the medium to work well as a means of, for example, conducting therapy sessions, or for any type of interviewing. It is unattuned to the ‘I–thou’ relationship. Where words come into their own is for transmitting information, specifically about something that is not present to us, something that is removed in space or time, when you and I need to co-operate in doing something about something else. It almost unimaginably expands the realm of the ‘I–it’ (or the ‘we–it’) relationship.

And what about the role that language plays, now that it exists, in thinking? Once again, language's role is in giving command over the world, particularly those parts that are not present spatially or temporally, a world that in the process is transformed from the ‘I–thou’ world of music (and the right hemisphere) to the ‘I–it’ world of words (and the left hemisphere). Words alone make concepts more stable and available to memory.98 Naming things gives us power over them, so that we can use them; when Adam was given the beasts for his use and to ‘have dominion’ over them, he was also the one who was given the power to name them. And category formation provides clearer boundaries to the landscape of the world, giving a certain view of it greater solidity and permanence. That may not have begun with humans, but it was obviously given a vast push forward by referential language. Language refines the expression of causal relationships. It hugely expands the range of reference of thought, and expands the capacity for planning and manipulation. It enables the indefinite memorialisation of more than could otherwise be retained by any human memory. These advantages, of memorialisation and fixity, that language brings are, of course, further vastly enhanced when language becomes written, enabling the contents of the mind to be fixed somewhere in external space. And in turn this further expands the possibilities for manipulation and instrumentalisation. The most ancient surviving written texts are bureaucratic records.

Language in summary brings precision and fixity, two very important features if we are to succeed in manipulating the world. And, specifically, though we may not like to recognise this, it is good for manipulating other human beings. We can't easily hide the truth in non-verbal communication, but we can in words. We can't easily direct others to carry out our plans without language. We can't act at a distance without language. Language, it would seem, starts out with what look like imperial aspirations.

Of course there is nothing wrong with manipulation in itself, with having designs on things that we can control, change, or make new. These are certainly basic, human characteristics, and they are the absolute foundations of civilisation. In this sense language is, as it is conventionally but simplistically conceived, a vastly precious and important gift.

Reverting to the needs of the frontal lobes, it provides the framework for a virtual representation of reality. Language enables the left hemisphere to represent the world ‘off-line’, a conceptual version, distinct from the world of experience, and shielded from the immediate environment, with its insistent impressions, feelings and demands, abstracted from the body, no longer dealing with what is concrete, specific, individual, unrepeatable, and constantly changing, but with a disembodied representation of the world, abstracted, central, not particularised in time and place, generally applicable, clear and fixed. Isolating things artificially from their context brings the advantage of enabling us to focus intently on a particular aspect of reality and how it can be modelled, so that it can be grasped and controlled.

But its losses are in the picture as a whole. Whatever lies in the realm of the implicit, or depends on flexibility, whatever can't be brought into focus and fixed, ceases to exist as far as the speaking hemisphere is concerned.

It also shifts the balance towards the concerns of the left hemisphere, which are not always consonant with those of the right. There are many links between language and grasp, and they have a similar agenda. Both sharpen focus on the world: mental grasp, like physical grasp, requires precision and fixity, which language provides, making the world available for manipulation and possession. Was it the drive for power, embodied in the will to control the environment, which accelerated symbol manipulation and the extension of conceptual thought – already present in some apes,99 and present in our early ancestors – resulting in the expansion of the left hemisphere before language and grasp evolved? In this light, language and grasp can be seen as expressions at the phenomenal level of a deeper lying drive in the left hemisphere: effective manipulation of the world, and beyond that competition with other species, and with one another. Once the capacity for manipulation was established in the left hemisphere, and no doubt especially once the power of abstraction was embedded there with the beginnings of a referential language, the preferential use of the right hand to carry out the literal manipulation of the environment would naturally have followed.


Language functions like money. It is only an intermediary. But like money it takes on some of the life of the things it represents. It begins in the world of experience and returns to the world of experience – and it does so via metaphor, which is a function of the right hemisphere, and is rooted in the body. To use a metaphor, language is the money of thought.

Only the right hemisphere has the capacity to understand metaphor.100 That might not sound too important – like it could be a nice thing if one were going to do a bit of lit crit. But that is just a sign of the degree to which our world of discourse is dominated by left-hemisphere habits of mind. Metaphoric thinking is fundamental to our understanding of the world, because it is the only way in which understanding can reach outside the system of signs to life itself. It is what links language to life.

The word metaphor implies something that carries you across an implied gap (Greek meta- across, pherein carry). When I call language metaphorical, I am not thinking only of Keats addressing the Grecian urn – ‘Thou still unravish'd bride of quietness, / Thou foster-child of silence and slow time’. Here there are clearly many complex, interacting metaphors, and that this creates something new and different from a factual description of the Sosibios Vase is obvious. This is metaphorical language in a dramatic sense. But there are two other, broader, but related, senses in which language is metaphorical. Speaking metaphorically, one might say that language is open to carry us across to the experiential world at the ‘top’ and at the ‘bottom’.

At the ‘top’ end, I am talking about any context – and these are not by any means to be found in poetry alone – in which words are used so as to activate a broad net of connotations, which though present to us, remains implicit, so that the meanings are appreciated as a whole, at once, to the whole of our being, conscious and unconscious, rather than being subject to the isolating effects of sequential, narrow-beam attention. As long as they remain implicit, they cannot be hijacked by the conscious mind and turned into just another series of words, a paraphrase. If this should happen, the power is lost, much like a joke that has to be explained (humour is a right-hemisphere faculty).

At the ‘bottom’ end, I am talking about the fact that every word, in and of itself, eventually has to lead us out of the web of language, to the lived world, ultimately to something that can only be pointed to, something that relates to our embodied existence. Even words such as ‘virtual’ or ‘immaterial’ take us back in their Latin derivation – sometimes by a very circuitous path – to the earthy realities of a man's strength (vir-tus), or the feel of a piece of wood (materia). Everything has to be expressed in terms of something else, and those something elses eventually have to come back to the body. To change the metaphor (and invoke the spirit of Wittgenstein) that is where one's spade reaches bedrock and is turned. There is nothing more fundamental in relation to which we can understand that.

That is why it is like the relation of money to goods in the real world. Money takes its value (at the ‘bottom’ end) from some real, possibly living, things – somebody's cows or chickens, somewhere – and it only really has value as and when it is translated back into real goods or services – food, clothes, belongings, car repairs – in the realm of daily life (at the ‘top’ end). In the meantime it can take part in numerous ‘virtual’ transactions with itself, the sort of things that go on within the enclosed monetary system.

Let me emphasise that the gap across which the metaphor carries us is one that language itself creates. Metaphor is language's cure for the ills entailed on us by language (much as, I believe, the true process of philosophy is to cure the ills entailed on us by philosophising). If the separation exists at the level of language, it does not at the level of experience. At that level the two parts of a metaphor are not similar; they are the same. The German thinker Jean Paul (Johann Paul Friedrich Richter) wrote in 1804, in his Vorschule der Ästhetik:

Ingenious figures of speech can either give soul to the body or body to the spirit. Originally, when man was still at one with the world, this two-dimensional trope did not yet exist; one did not compare that which showed no resemblance, but one proclaimed identities: metaphors were, as with children, necessary synonyms for body and mind. Just as, in the case of writing, pictures preceded the alphabet, metaphor (insofar as it designated relations and not objects) was the first word in spoken language, and only after losing its original colour could it become a literal sign.101

A metaphor asserts a common life that is experienced in the body of the one who makes it, and the separation is only present at the linguistic level. Our sense of the commonality of the two ideas, perceptions or entities does not lie in a post hoc derivation of something abstracted from each of them, which is found on subsequent comparison to be similar, or even one and the same thing; but rather on a single concrete, kinaesthetic experience more fundamental than either, and from which they in turn are derived. Thus a clash of arguments and a clash of cymbals are not seen to have something in common only after the disembodied idea of a ‘clash’ is abstracted from the one and from the other, and found – aha! – to be similar; it is rather that these experiences – a clash of arguments and a clash of cymbals, or, for that matter, a clash of swords, or a clash of colours – are felt in our embodied selves as sharing a common nature.

When the metaphor is paraphrased or replaced, whatever had been extralingual, unconscious, and therefore potentially new and alive in the collision of these two entities gets reconstructed, this time in terms only of what is familiar. The point of metaphor is to bring together the whole of one thing with the whole of another, so that each is looked at in a different light. And it works both ways, as the coming together of one thing with another always must. You can't pin one down so that it doesn't move, while the other is drawn towards it: they must draw towards each other. As Max Black says: ‘If to call a man a wolf is to put him in a special light, we must not forget that the metaphor makes the wolf seem more human than he otherwise would.’102 And Bruno Snell, discussing the way in which Homer likens the brave warrior to a rock amidst the crashing waves, the rock then, in turn, being described as ‘steadfast’ by analogy with human behaviour, writes perceptively:

This peculiar situation, namely that human behaviour is made clear only through reference to something else which is in turn explained by analogy with human behaviour, pertains to all Homeric similes. More than that, it pertains to all genuine metaphors, and in fact to every single case of human understanding. Thus it is not quite correct to say that the rock is viewed anthropomorphically, unless we add that our understanding of rock is anthropomorphic for the same reason that we are able to look at ourselves petromorphically … man must listen to an echo of himself before he may hear or know himself.103

Metaphor (subserved by the right hemisphere) comes before denotation (subserved by the left). This is a historical truth, in the sense that denotative language, even philosophical and scientific language, are derived from metaphors founded on immediate experience of the tangible world.

Metaphor is centrally a matter of thought, not just words. Metaphorical language is a reflection of metaphorical thought … Eliminating metaphor would eliminate philosophy. Without a very large range of conceptual metaphors, philosophy could not get off the ground. The metaphoric character of philosophy is not unique to philosophic thought. It is true of all abstract human thought, especially science. Conceptual metaphor is what makes most abstract thought possible.104

It is also a truth about epistemology, how we understand things. Any one thing can be understood only in terms of another thing, and ultimately that must come down to a something that is experienced, outside the system of signs (i.e. by the body). The very words which form the building blocks of explicit thought are themselves all originally metaphors, grounded in the human body and its experience.

Metaphor embodies thought and places it in a living context. These three areas of difference between the hemispheres – metaphor, context and the body – are all interpenetrated one with another. Once again it is the right hemisphere, in its concern for the immediacy of experience, that is more densely interconnected with and involved in the body, the ground of that experience. Where the right hemisphere can see that metaphor is the only way to preserve the link between language and the world it refers to, the left hemisphere sees it either as a lie (Locke, expressing Enlightenment disdain, called metaphors ‘perfect cheats’)105or as a distracting ornament; and connotation as a limitation, since in the interests of certainty the left hemisphere prefers single meanings.

For the left hemisphere, consequently, language can come to seem cut off from the world, to be itself the reality; and reality, for its part, comes to seem made up of bits strung together, as the words are strung together by syntax. The left hemisphere is bound to see language like this because it understands things by starting from the observation of ‘pieces’ and builds them up to make something, and this is the only route it has to understanding both the world and language itself, the medium with which it does its understanding, including its understanding of language.


Metaphors, even the simple ones hidden in expressions like feeling ‘down’, derive from our experience of living as embodied creatures in the everyday world.106 The body is, in other words, also the necessary context for all human experience.

In fact even language, historically and within the story of each living individual as he or she acquires language, demonstrates that it is not a theoretical system or set of procedures, made up of bits strung together by rules or algorithms, but an embodied skill, and its origins lie in the empathic communication medium of music and the right hemisphere, where it is deeply connected with the body.

I mentioned earlier that there were those who believed that language arose, not from music, but from gesture. There is, however, no necessary conflict between such beliefs. Music is deeply gestural in nature: dance and the body are everywhere implied in it. Even when we do not move, music activates the brain's motor cortex.107 Music is a holistic medium, ‘multimodal’ as Mithen puts it, not limited to a distinct modality of experience. To the extent that the origins of language lie in music, they lie in a certain sort of gesture, that of dance: social, non-purposive (‘useless’). When language began to shift hemispheres, and separate itself from music, to become the referential, verbal medium that we now recognise by the term, it aligned itself with a different sort of gesture, that of grasp, which is, by contrast, individualistic and purposive, and became limited to one modality.

But language, if we attend to it rightly, still trails the clouds of glory from its origins in the right hemisphere. The eighteenth-century German philosopher Herder, in his Essay on the Origin of Language, points out that language can help to blind us to the intrinsically synaesthetic nature of experience, but suggests that some of this must, in spite of language, be caught in the word-sounds arising from it:108

We are full of such interconnections of the most different senses … in nature all the threads are one single tissue… . The sensations unite and all converge in the area where distinguishing traits turn into sounds. Thus, what man sees with his eye and feels by touch can also become soundable.109

Yet with the rise of Saussurian linguistics in the twentieth century, it has become fashionable to insist on the arbitrary nature of the sign – a fascinating and counterintuitive move, designed to emphasise the ‘freedom’ of language as far as possible from the trammels of the body and of the physical world it describes. There is, however, plenty of evidence that the sounds of words are not arbitrary, but evocative, in a synaesthetic way, of the experience of the things they refer to. As has been repeatedly demonstrated, those with absolutely no knowledge of a language can nonetheless correctly guess which word – which of these supposedly arbitrary signs – goes with which object, in what has become known as the ‘kiki/bouba’ effect (‘kiki’ suggesting a spiky-shaped object, where ‘bouba’ suggests a softly rounded object).110However much language may protest to the contrary, its origins lie in the body as a whole. And the existence of a close relationship between bodily gesture and verbal syntax implies that it is not just concrete nouns, the ‘thing-words’, but even the most apparently formal and logical elements of language, that originate in the body and emotion. The deep structure of syntax is founded on the fixed sequences of limb movement in running creatures.111 This supports evidence that I will examine in Chapter 5 that the very structures and content of thought itself exist in the body prior to their utterance in language.

Why do I emphasise this bodily origin of thought and language? Partly because it has been denied in our own age, not by any means only, or even mainly, by de Saussure and his followers. More than that, the fact of its denial seems to me to form part of a general trend, throughout the last hundred years or so, towards the ever greater repudiation of our embodied being, in favour of an abstracted, cerebralised, machine-like version of ourselves that has taken hold on popular thinking – even though there may be more recent trends in philosophy that attempt, with widely varying degrees of success, to point away from such conclusions. As Lakoff and Johnson make clear,

the very structure of reason itself comes from the details of our embodiment. The same neural and cognitive mechanisms that allow us to perceive and move around also create our conceptual systems and modes of reason … Reason is evolutionary, in that abstract reason builds on and makes use of forms of perceptual and motor inference present in ‘lower’ animals … Reason is thus not an essence that separates us from other animals; rather, it places us on a continuum with them.112

The flight of language from the enchantment of the body during the last hundred years represents, I believe, part of a much broader revolt of the left hemisphere's way of conceiving the world against that of the right hemisphere, the theme of Part II of this book.

The vehemence of the comparison of music to such useless exaptations as pornography and a taste for fatty food intentionally or unintentionally makes it hard to put forward the case that language, the precious tool of scientific cognition, comes ultimately from the mucky world of emotion and the body. Is that what is being denied? Or is it perhaps that scientific discourse, so heavily dependent on referential language, doesn't like acknowledging those skeletons in the family closet, its embarrassing bodily ancestors, grasp and manipulation? Whatever the motive, language certainly does its best to cover its tracks and deny its parentage.

Take the rise of the Chomskyan theory of universal grammar.113 The belief that the structures of analytic language are hard-wired into our brains helps to perpetuate the idea that the brain is a cognitive machine, a computer that is fitted with a rule-based programme for structuring the world, rather than its being an inextricable part of an embodied, living organism that develops implicit, performative, skills through an empathic process of intelligent imitation. While I am not in a position to do full justice to an issue that is still a matter of lively debate among the experts, it is uncontroversial that the existence of a universal grammar such as Chomsky conceived is highly debatable. It remains remarkably speculative 50 years after he posited it, and is disputed by many important names in the field of linguistics.114 And some of the facts are hard to square with it. Languages across the world, it turns out, use a very wide variety of syntax to structure sentences.115 But more importantly the theory of universal grammar is not convincingly compatible with the process revealed by developmental psychology, whereby children actually acquire language in the real world. Children certainly evince a remarkable ability to grasp spontaneously the conceptual and psycholinguistic shapes of speech, but they do so in a far more holistic, than analytic, way. They are astonishingly good imitators – note, not copying machines, but imitators.

Imitation can certainly be reduced to a matter of copying by rote: breaking an action down into a series of steps, and reproducing them mechanically. Deliberate, explicit copying of single gestures, out of context, would be like this. But it can also be driven by a feeling of attraction which results, by a process that remains mysterious, in our apprehending the whole and trying to feel what that must be like from the inside – by so to speak ‘inhabiting’ the other person. This is how we imitate someone else's voice, speech patterns or physical mannerisms, their way of talking or walking. I use the term ‘attraction’ in a sense that makes no necessary judgment about the worth of its object: if imitation can be the sincerest form of flattery, it can also be the sincerest form of mockery. But it does often carry a charge of positive attraction towards its object: we become who we are by imitating the models of people we admire or respect. It is also how we acquire most of our skills, even though at times we may resort to copying by rote. Such imitation is empathic, and involves identification. It plays an important role in human development, not only in skill acquisition – such, precisely, as a child's developing mastery of language – but in the development of values which form part of our individuality. I will return to the topic of imitation in Chapter 7, when I consider the possible ways in which shifts in the history of ideas could come about.

Skills are embodied, and therefore largely intuitive: they resist the process of explicit rule following. The Chuang Tzu, a classic of Taoist literature, contains several stories, such as that of Cook Ting cutting up an ox for Lord Wen-hui, designed to illustrate the fact that a skill cannot be formulated in words or rules, but can be learnt only by watching and following with one's eyes, one's hands and ultimately one's whole being: the expert himself is unaware of how he achieves what it is he does.116 As Dreyfus and Dreyfus put it in their book Mind Over Machine: ‘an expert's skill has become so much a part of him that he need be no more aware of it than he is of his own body.’117 Despite powerful suggestions to the contrary, language is not an abstraction from life, a game – with its suggestions of autonomy, triviality and definition by rules. No, it is an extension of life. In Wittgenstein's famous phrase: ‘to imagine a language means to imagine a form of life’ – not a virtual representation of life, but a form of life.118

A child does not acquire the skill of language, any more than the skill of life, by learning rules, but by imitation, a form of empathic identification, usually with his or her parents, or at any rate with those members of the group who are perceived as more proficient. I have suggested that such identification involves an (obviously unconscious) attempt to inhabit another person's body, and this may sound somewhat mystical. But imitation is an attempt to be ‘like’ (in the sense of experiencing what it is ‘like’ to be) another person, and what it is ‘like’ to be that person is something that can be experienced only ‘from the inside’.119 Not just the acquisition of language, but the everyday business of language in itself involves just such an inhabiting. Communication occurs because, in a necessarily limited, but nonetheless crucially important, sense, we come to feel what it is like to be the person who is communicating with us. This explains why we pick up another person's speech habits or tics, even against our will (a stammer is a sometimes embarrassing case in point); it explains many of the problems of emotional entrainment in conversation, the countertransference that occurs, not just in therapy, but in ordinary, everyday life, when we experience in our own frames the very feelings that our interlocutor experiences. And empathy is associated with a greater intuitive desire to imitate.120

By inhabiting the body of the other: is this how language (‘musilanguage’) began? Rudolf Laban, who perhaps more closely observed the meaning of bodily movement in performance than anyone that ever lived, has some fascinating observations to make in this regard. In sub-Saharan Africa there is a form of communication using drumbeats which has been dubbed, perhaps somewhat infelicitously, ‘rhythmic drum telegraphy’. The technique is widespread, and by it apparently detailed messages can be communicated over long distances. According to Laban, there is no attempt, as the Westerner might imagine, to mimic the sound pattern of words or phrases; that would be rendered pointless by the many different languages spoken by different tribes occupying adjacent territories. Instead ‘the reception of these drum or tom-tom rhythms is accompanied by a vision of the drummer's movement, and it is this movement, a kind of dance, which is visualised and understood.’121 Communication occurs because the listener inhabits the body of the person who drums and experiences what it is that the drummer is experiencing. Even if language no longer seems to us in the West to ‘body forth’ meaning in this way, it may be that at least our understanding of music still shares this inhabiting of the movements of the other – the performer, the singer, perhaps even the composer. Laban again: ‘It is … interesting to note that orchestral music is produced by the most precise bodily movements of the musicians,’ and he suggests that perhaps one of the reasons we like to see, as well as hear, music performed is exactly that we can better inhabit the performer's body, a perception that appears to me intuitively correct.122

To recapitulate, then: language originates as an embodied expression of emotion, that is communicated by one individual ‘inhabiting’ the body, and therefore the emotional world, of another; a bodily skill, further, that is acquired by each of us through imitation, by the emotional identification and intuitive harmonisation of the bodily states of the one who learns with the one from whom it is learnt; a skill moreover that originates in the brain as an analogue of bodily movement, and involves the same processes, and even the same brain areas, as certain highly expressive gestures, as well as involving neurones (mirror neurones) that are activated equally when we carry out an action and when we see another carry it out (so that in the process we can almost literally be said to share one another's bodily experience and inhabit one another's bodies); a process, finally, that anthropologists see as derived from music, in turn an extension of grooming, which binds us together as physically embodied beings through a form of extended body language that is emotionally compelling across a large number of individuals within the group. At the least one can say that it forms bridges, as any mode of communication has to, between individuals, at every stage of its development and practice, historically and individually, and it does so by relying on our common corporeality, within a group – the image of which, furthermore, is the body. We call a group of people ‘a body’, and its constituents are seen as limbs, or ‘members’. Their relationship within the group is not additive merely, as it would be in a mechanical assembly of items, but combinatory, producing a new entity that is more than the sum of its parts. If it were a chemical, one would say that it were a compound, rather than a mixture.

If language began in music, it began in (right-hemisphere) functions which are related to empathy and common life, not competition and division; promoting togetherness, or, as I would prefer, ‘betweenness’. By its nature as a means of communication, language is inevitably a shared activity, like music, which begins in the transmission of emotion and promotes cohesion. Human singing is unique: no other creature begins to synchronise the rhythm, or blend the pitch, of its utterances with that of its fellows, in the way that human singing instinctively does. It is not, like birdsong, individualistic in intention and competitive in nature (remember that birdsong, like other instrumental utterances, is grounded in the left hemisphere, not, like human music, in the right). Everything about human music suggests that its nature is sharing, non-competitive. And so it has been argued by a number of anthropologists that the development of musical skill must have been a product, not of individual selection, but of group selection, a process whereby ‘reproduction of all genes present in a group is influenced in a similar manner by newly developed behaviours’.123 Natural selection exploits the difference in individual rates of successful reproduction within the group, but here the whole group would have benefited – in terms of its cohesion as a group – from something the whole group would have evolved.124

And indeed referential language too would have to have been a product of group selection if it really had much to do with communication. Either everyone in the group develops it, and the group benefits, so that the members of the group flourish, or it doesn't develop – since it's not much good being a solitary expert communicator if your fellows can't pick up the message. And this makes perfect sense: the advantages of spoken language, such as more efficient hunting, would have benefited all individuals in the group, even if there was a range of development in linguistic skills, since the products of hunting would have been shared. Classical natural selection, by contrast, would have to demand that the skill be positively hidden or kept secret from others – a skill, however, which, by its very nature, is concerned with sharing information. So it looks as if language may have started out, not as the product of ruthless competition, but as another area in which humanity has done better by co-operation and collaboration.125 Returning to the puzzle of music's (apparent lack of) competitive advantage, we have conventionally been unable to see any advantage because we are used to thinking in terms only of individual pitted against individual, not in terms of the group, in which individuals work together.

This is the argument from utility. Like so many other things that we are so often told are useful to the group – music and dance, a sense of beauty and a sense of awe – language helped make us more effective competitors at the group level, if not at the level of the individual. Ultimately, though, I believe that the great achievement of humankind is not to have perfected utility through banding together to form groups, but to have learnt through our faculty for intersubjective experience, and our related ability to imitate, to transcend utility altogether. We can, through our ability to imitate, make our own choices about the direction we take, mould our thinking and behaviour, and therefore our human future, according to our own values, rather than waiting to be driven by the blind process of genetic competition, which knows only one value, that of utility. We can choose to imitate forms of thinking or behaving; and by so doing both speed up our evolution by many orders of magnitude, and shift it away from the blind forces of chance and necessity, in a direction or directions of our own choosing.

It is rather odd to find Dunbar referring to dance as useless: ‘dancing, a phenomenon that probably ranks, along with smiling and laughter’, he writes, ‘as one of the most futile of all human universals’.126 I say it is odd because he of all people ought to be able to see past its apparent uselessness to the individual, to its supposed usefulness to the group. Perhaps he does, and calls it ‘futile’ tongue in cheek. But I'd rather agree with him, nonetheless, that smiling, laughter and dance are – gloriously – useless: how many of us really believe that when we dance, laugh, or smile we do so ultimately because of some dreary utility to the group to which we belong? Perhaps there is no end in view. Perhaps these spontaneous behaviours are pointless, with no purpose beyond themselves, other than that they express something beyond our selves. Perhaps, indeed, the fact that so many of our distinguishing features are so ‘useless’ might make one think. Instead of looking, according to the manner of the left hemisphere, for utility, we should consider, according to the manner of the right hemisphere, that finally, through intersubjective imitation and experience, humankind has escaped from something worse even than Kant's ‘cheerless gloom of chance’: the cheerless gloom of necessity.127

Deprived of an explanation in terms of a final cause – the reason that makes sense of a behaviour in terms of its outcome – scientists sometimes think they have accounted for a phenomenon by redescribing it on another level. Thus Dunbar explains our indulging in ‘futile’ activities by reference to endorphins. Grooming, music, togetherness, love, religion – all turn out mysteriously to release endorphins. ‘Sound familiar?’, he queries, on one such occasion, pleased with the simplicity of his solution, but aware that it has clocked up a few air miles by now: ‘Well, of course it is: it is the endorphin story all over again.’128And that is supposed to explain at last why we need, enjoy and take comfort in such things. But is this really any different from proudly announcing that, after prolonged research, we have discovered that the reason we dislike being belittled, ignored or hit over the head with a shillelagh is that it causes depletion of endorphins, reduced bioavailability of serotonin, secretion of cortisol or overdrive of the sympathetic nervous system? In the real world, however, we do not choose to engage in activities because they release endorphins, and endorphin release is a blast; it's that when we engage in what, for a myriad complex and subtle reasons, has meaning and importance for us, we are happier, endorphins merely being part of the final common pathway for happiness at the neurochemical level – just as we avoid a mugger not because we'd like to maintain our levels of serotonin as long as possible, but because he's likely to attack us and make life miserable for us, depleted serotonin just being the final common pathway of misery.

So language is a hybrid. It evolved from music and in this part of its history represented the urge to communicate; and to the extent that it retains right-hemisphere empathic elements, it still does. Its foundations lie in the body and the world of experience. But referential language, with its huge vocabulary and sophisticated syntax, did not originate in a drive to communicate, and from this point of view, represents something of a hijack. It has done everything it can to repudiate both its bodily origins and its dependency on experience – to become a world unto itself. Despite all that, however, the urge to speak still does not come from Broca's area, where the motor speech act originates. That's evident from the fact that subjects with lesions in Broca's area usually seem desperate to communicate. No, it comes from the anterior cingulate, a deeper lying region profoundly implicated in social motivation. Subjects with damage to this area exhibit akinetic mutism, the lack of desire to communicate despite having perfectly normal speech function. ‘This reinforces the conclusion that speech is fundamentally a social act, and it has only been tortuously bent for scientific ends. Parenthetically, dolphins and whales have rich neural expansions in this area of the brain [anterior cingulate], and they do seem to be highly communicative.’129 It might also be pointed out that these animals, famous for their intelligence and sociability, communicate by music.

Language has done its best to obscure its parentage. It has increasingly abstracted itself from its origins in the body and in the experiential world. It developed its current form to enable us to refer to whatever is not present in experience: language helped its re-presentation. This had the effect of expanding its usefulness to communication and thinking for some purposes, but reducing it for others. In the process important aspects of language, the denotative elements that enable precision of reference and planning, have taken up residence in the left hemisphere, while other aspects of language, broadly its connotative and emotive functions, have remained in the right hemisphere. And the understanding of language at the highest level, once the bits have been put together, the making sense of an utterance in its context, taking into account whatever else is going on, including the tone, irony, sense of humour, use of metaphor, and so on, belongs once again with the right hemisphere.

The way these aspects of language have sorted is, as we have seen, not random, but in keeping with the overall nature of each hemisphere. Metaphor is the crucial aspect of language whereby it retains its connectedness to the world, and by which the ‘parts’ of the world which language appears to identify retain their connectedness one to another. Literal language, by contrast, is the means whereby the mind loosens its contact with reality and becomes a self-consistent system of tokens. But, more than this, there is an important shape here which we will keep encountering: something that arises out of the world of the right hemisphere, is processed at the middle level by the left hemisphere and returns finally to the right hemisphere at the highest level.


We have talked a lot about the left hemisphere and its world. What of the right hemisphere? At the same time as developing this specialised narrowly focussed view of the world, we cannot afford to lose track of the totality of experience in all its richness. It's all very well having a virtual world, but first and foremost one has to carry on inhabiting the real world of experience, where one's ability to manipulate can be put to effect. Man's success has been not just in manipulating the environment, as the ‘tool-making animal’, but in creating close-knit societies, the basis of civilisation.

It is the right hemisphere that enables us to do just that, by maintaining its broader remit, and, in light of what the frontal lobe development opens up to us, take it much further. Already specialised in social bonding, it would be the natural place for the relational, empathic skills of man, the ‘social animal’, to be further developed: and this is exactly what one finds. As I have mentioned, if one looks at the brain's structure, one notices that it is not just the left hemisphere that has an asymmetrical enlargement, but the right, too.

As we have seen, asymmetries similar to those found in the human brain are also found in monkeys and apes, and I have mentioned the existence in them of a left-hemisphere expansion. But it is not just in this that our evolutionary ancestors anticipate us. They have this right frontal petalia, too.

The evidence as to which petalia came first, right or left, is divided.130 Being frontal, unlike the more posterior left-hemisphere expansion, the right petalia may well have arrived later on the scene. In the human foetus, however, the frontal regions of the right hemisphere develop before the occipital regions of the left hemisphere – does ontogeny here recapitulate phylogeny?131 In general the right hemisphere matures first, though in the second year of life the left hemisphere overtakes it, with the laying down of the speech and language areas;132 but there is also evidence that the right hemisphere then continues developing after the left hemisphere has matured, with the more sophisticated emotional and prosodic elements of language developing in the fifth and sixth years of life.133 If true, there is an interesting parallel here between the developmental history of the hemispheres (right The Master and his Emissary left The Master and his Emissary right) and their functional relationship. In any case, a right frontal petalia begins to be found in some of the more social monkeys, such as macaques, and in apes,134 but reaches its most pronounced in humans – in whose brains it is in fact the most asymmetrical region of all.135

If it is even more pronounced than the left petalia, and even more particular to humans, why have we paid so little attention to it? Could that be because we have focussed on the left hemisphere, and what it does, at the expense of the right, and what it does? Until recently everything about the right hemisphere has been shrouded in darkness. It was, after all, considered to be silent; and to the verbal left-hemisphere way of thinking, that means dumb. Is the right frontal lobe responsible for anything that might compare with the achievements, in terms of grasp and denotative language, of the left hemisphere?

We know that it is the right frontal lobe which enables us to achieve all the rest of which language is capable; which makes empathy, humour, irony possible, and helps us to communicate and express not just facts, but our selves. Here language becomes not a tool of manipulation but a means of reaching out to the ‘Other’. But it is, of course, not just in the realm of language that its significance lies – far from it.

Indeed, most of the remarkable things about human beings, the things that differentiate us from the animals, depend to a large extent on the right hemisphere, and in particular on the contributions of the region of this right-hemisphere expansion, the right frontal lobe. If asked to name the characteristics that ultimately differentiate humans from animals, the classic answers, reason and language, seem like a poor stab. Plenty of animals show, in their degree, capacity to deduce (deductive reasoning is importantly associated with right-hemisphere function, in any case): crows can reason, even bees have language of a kind. Of course, even the most highly evolved animals are incomparably inferior to ourselves in both respects, but the point is that they do show at least glimmerings of such, utilitarian, functions. But there are many things of which they show no evidence whatsoever: for instance, imagination, creativity, the capacity for religious awe, music, dance, poetry, art, love of nature, a moral sense, a sense of humour and the ability to change their minds. In all of these (though as always both hemispheres undoubtedly play a part), a large part, and in most cases the principal part, is played by the right hemisphere, usually involving the right frontal lobe. Where the left hemisphere's relationship with the world is one of reaching out to grasp, and therefore to use, it, the right hemisphere's appears to be one of reaching out – just that. Without purpose. In fact one of the main differences between the ways of being of the two hemispheres is that the left hemisphere always has ‘an end in view’, a purpose or use, and is more the instrument of our conscious will than the right hemisphere.


I suggest that there are two opposing ways of dealing with the world that are both vital but are fundamentally incompatible, and that therefore, even before humans came on the scene, required separate treatment, even neurological sequestration from one another. One tendency, important for being able to get things from the world for one's own purposes, involves isolation of one thing from the next, and isolation of the living being, perceived as subjective, from the world, perceived as objective. The drive here is towards manipulation, and its ruling value is utility. It began in my view by colonising the left hemisphere, and with the increasing capacity for distance from the world mediated by the expansion of the frontal lobes as one ascends the evolutionary tree, resulted in a physical expansion of the area designed to facilitate manipulation of the environment, symbolically and physically, in the higher monkeys and apes. Eventually that expansion became the natural seat of referential language in humans.

The other tendency was centripetal, rather than centrifugal: towards the sense of the connectedness of things, before reflection isolates them, and therefore towards engagement with the world, towards a relationship of ‘betweenness’ with whatever lies outside the self. With the growth of the frontal lobes, this tendency was enhanced by the possibility of empathy, the seat of which is the right frontal expansion in social primates, including humans.

It may well be that we, and the great apes before us, are not the originators of the asymmetry in hemisphere function – not even the originators of the nature of that asymmetry – but inheritors of something much older than ourselves, which we have utilised and developed in peculiarly human ways to peculiarly human ends. It is not just human beings who have found that there are needs, drives or tendencies, which, while equally fundamental, are also fundamentally incompatible: an essentially divisive drive to acquisition, power and manipulation, based on competition, which sets individual against individual, in the service of unitary survival; and an essentially cohesive drive towards co-operation, synergy and mutual benefit, based on collaboration, in the service of the survival of the group. Before the arrival of language or tool making in the left hemisphere, with their need for Lebensraum, could have ‘driven it out’, the higher apes show signs already of having segregated the expression of social emotion, as we have, to the right hemisphere of the brain – kept away from the areas of useful abstraction; and abstraction, just as importantly, kept from exerting its corrosive effect on experience.

Both of these drives or tendencies can serve us well, and each expresses an aspect of the human condition that goes right to the core. It is not inevitable, ultimately, that they should be in conflict; and in fact it is best that they should not be. In some human brains, it appears that they can more closely co-exist, and I will return to that in the conclusion of this book. But the relationship between the hemispheres is not straightforward. Difference can be creative: harmony (and counterpoint) is an example. Here differences cohere to make something greater than either or any of the constituents alone; which is why it would be a mistake to see the divisive tendency as purely negative. Before there can be harmony, there must be difference. The most fundamental observation that one can make about the observable universe, apart from the mysterious fact that it exists at all – prompting the ultimate question of philosophy, why there is something rather than nothing – is that there are at all levels forces that tend to coherence and unification, and forces that tend to incoherence and separation. The tension between them seems to be an inalienable condition of existence, regardless of the level at which one contemplates it. The hemispheres of the human brain, I believe, are an expression of this necessary tension. And the two hemispheres also adopt different stances about their differences: the right hemisphere towards cohesion of their two dispositions, the left hemisphere towards competition between them.

Since the right hemisphere is more distinctive of the human condition than the left, it remains a puzzle why it has been neglected. It seems part of what one might call the ‘minor hemisphere’ syndrome. Yet we know it is the hemisphere on which experience is grounded and which has the broader view, the one that is open to whatever else exists outside the brain. How, then, has this neglect occurred? Is it just that the left hemisphere has control of language and analytic argument, and that therefore whenever scientists (who depend on such methodology to build up a view of anything from the ‘bits’ of information) look at the brain, they do so only with the left hemisphere, and see only what it sees? Is it just that such means are not capable of understanding the world as it is as a whole, and that therefore the left hemisphere prefers its own version, which at least makes sense to it? Or is there something else going on here?

The sheer vehemence with which the right hemisphere has been dismissed by the representatives of the articulate left hemisphere, despite its overwhelming significance, suggests a possible rivalry. I believe there has been until very recently a blindness among neuroscientists to the contributions made by the right hemisphere. In 1966 R. C. Oldfield wrote that ‘a certain conspiracy of silence prevails among neurologists about the lack of anything much for the right hemisphere to do’.136 Until John Cutting's The Right Cerebral Hemisphere and Psychiatric Disorders and his Principles of Psychopathology, and Michael Trimble's recent The Soul in the Brain: The Cerebral Basis of Language, Art, and Belief,137 it has not only received little credit, but has been the object of some, at least superficially, inexplicable animus. There would seem to be a partisanship amongst scientists in the left hemisphere's favour, a sort of ‘left-hemisphere chauvinism’ at work. One sees it even in the language used by the most objective writers to describe the hemisphere differences: for example, the smart left hemisphere's need for precision leads to ‘fine’ processing, the lumpen right hemisphere's to ‘coarse’ processing. No mention here of the dangers of over-determination, or the virtues of a broader range, of subtlety, ambiguity, flexibility or tolerance.

In this respect, it is perhaps worth reporting the unbiased impression of a ‘naïve’ reader, the composer Kenneth Gaburo, approaching the neuroscientific literature from outside, who picked up that ‘there is something extraordinarily pejorative’ in the language used to describe the right hemisphere.138 Amongst other references to the left hemisphere as ‘dominant’, and the right hemisphere as ‘minor’, ‘silent’, and so on, he refers to the influential paper which Salomon Henschen, one of the giants of the history of neuropathology, and a former Professor of Medicine at Uppsala, contributed to Brain in 1926. The situation is indeed worse even than Gaburo implies, since Henschen's actual words are:

In every case the right hemisphere shows a manifest inferiority when compared with the left, and plays an automatic role only … This fact shows the inferiority of the right hemisphere, especially of the right temporal lobe … A person who is not able to understand words after destruction of the left temporal lobe sinks to the level of primitive man … The right temporal lobe is, of course, sufficient for the more primitive psychical life; only by using the left temporal lobe can man reach a higher level of psychical development … it is evident that the right hemisphere does not reach the same high level of psychical development as the left … The question therefore arises if the right hemisphere is a regressing organ … it is possible that the right hemisphere is a reserve organ.139

Michael Gazzaniga, one of the most distinguished living neuroscientists and hemisphere researchers, carries on the tradition, and his language is in the tradition of Henschen. There are ‘shocking differences between the two hemispheres’: ‘the left hemisphere has many more mental capacities than the right one … [The right hemisphere] is a distant second with problem-solving skills … It knows precious little about a lot of things.’140 He once wrote that ‘it could well be argued that the cognitive skills of a normal disconnected right hemisphere without language are vastly inferior to the cognitive skills of a chimpanzee’.141 In a more recent article, he has written: ‘A brain system (the right hemisphere) with roughly the same number of neurons as one that easily cogitates (the left hemisphere) is incapable of higher order cognition—convincing evidence that cortical cell number by itself cannot fully explain human intelligence.’142 Yet when the right hemisphere can be shown to outperform the left at some fairly basic task of prediction, he interprets this as a sign of the intelligence of the left hemisphere, on the grounds that animals are also capable of outperforming the human left-hemisphere strategy. In fact the problem is that the left hemisphere just loves a theory, and often this is not helpful in practice – which is why it gets it wrong. Of attention he writes that

the left-dominant hemisphere uses a ‘guided’ or ‘smart’ strategy whereas the right hemisphere does not. This means that the left hemisphere adopts a helpful cognitive strategy in solving the problem whereas the right hemisphere does not possess those extra cognitive skills. But it does not mean that the left hemisphere is always superior to the right hemisphere in attentional orienting.143

True enough. In fact, as he knows, since he refers to the fact, the right hemisphere is predominant for attentional orienting, a topic that will be familiar to the reader. What he refers to as a ‘smart’ and ‘helpful’ strategy – ‘those extra cognitive skills’ – are in fact neither smart nor helpful, when compared with the open, undogmatic stance of the right hemisphere. They lead to less accuracy, not more. But you might not know that from the language used.

Just as tedious, of course, is the tendency to see what commonly passes for the ‘left hemisphere’ in pop parlance as wholly without redeeming features. Often, it seems to me, such positions conceal an undercurrent of opposition to reason and the careful use of language, and once words slip their anchors, and reason is discounted – as some quite influential post-modern and feminist critics have advocated – Babel ensues.144 Doubts about the extent of rationalism, the belief that reason alone can yield all truth, do not make one anti-rational: to decry reason itself is utter folly. Poetry and metaphor, like science, hold no brief for sloppiness, quite the opposite, just as it is reason, not its unfettered disregard, that leads to scepticism about misplaced and excessive rationalism. But language and reason are the children of both hemispheres, not one alone. The work of the left hemisphere needs to be integrated with that of the right hemisphere, that is all. The left hemisphere is the Master's most prized counsellor, his valued emissary.

I have already suggested that there is a need for the hemispheres to keep their distance from one another and to be able to inhibit one another. More than that, the concerns of the left hemisphere with getting and using make it by nature competitive – it may be remembered that it is confident, unreasonably optimistic, unwitting of what goes on in the right hemisphere, and yet in denial about its own limitations. What if it should turn out – and it does – that the left-hemisphere advantage gained by right handedness has been the result, not of an increase of skill in the right hand, but of a deficit in the left?145 There may be another interesting asymmetry here.

Marian Annett, perhaps the greatest living authority on handedness, believes that we may have developed ‘over-dependence on the left hemisphere at the expense of right hemisphere skills’. She points to the unexpectedly large number of left-handers amongst artists, athletes, and ‘skilled performers of many kinds’.146 A marked difference between the performance of the two hands in right-handers is associated with a slight improvement in the right hand, but the price for this, according to Annett, is that ‘the left hand declines dramatically’, a finding that has been corroborated by many other researchers.147

This pattern of a specific relative right-hemisphere handicap is borne out at the anatomical level.148 The planum temporale, as mentioned in the first chapter, is asymmetrical in most human brains, with the left being up to a third bigger than the right. But in cases where, unusually, the two hemispheres develop symmetrically, it's not that the two plana are the same size as the usual right (smaller) planum, but the size of the usual left planum: in other words, they are both large. In normal brains of right-handers, therefore, it's not that the left planum is increased, but that the right planum is decreased, in size. Recent research to find the gene or genes responsible for brain asymmetry in the language region expected to find a gene which operated on the left hemisphere to cause it to expand. Instead they found genes that acted on the right hemisphere to prevent its expansion: of the 27 genes implicated, most were more highly expressed on the right, and the most important gene was dramatically more so. Christopher Walsh, a professor of neurology at Harvard who led the research, comments: ‘We tend to assume teleologically, because of our focus on language being that most beautiful thing, that it must be endowed by some special mechanism in the left hemisphere … in fact, it may just be normally repressed in the right hemisphere and allowed to take place in the left.’149

The ‘normal’ situation, then, is associated with right-hemisphere losses, both anatomical and functional.150 The mechanisms inducing human cerebral asymmetry operate by reducing the role of the right hemisphere.151

Why? Not to be lateralised at all is a disadvantage, as we have seen.152 This has to be because there are trade-offs associated with the specialisation of the ‘dominant’ hemisphere, the one with control of language and grasping. Isolation of left-hemisphere-type function makes it that bit easier for it to do what it has to. It functions more efficiently if it is not having to deal with the conflicting ‘version’ of the world put forward by the other, so-called ‘minor’, hemisphere. So the non-dominant hemisphere has to be put at a disadvantage. But take the process too far, and the obvious losses occasioned by hobbling the right hemisphere outweigh the advantage to the left. It is an inverted U-shaped curve. Speed in moving pegs on a board with either hand is a measure of the skill of the contralateral hemisphere: strong right-handers are slower than non-right-handers, especially with their left hand.153 Equally the relatively few strong left-handers, whose brains may mirror those of strong right-handers, are at a disadvantage, too. In fact Annett surmises that the high numbers of left-handers among mathematicians154 and sports professionals is not so much due to an intrinsic advantage for left-handers as to the absence of strong right-handers (who are at a disadvantage).155 Those more likely to have anomalous patterns of lateralisation, such as left-handers, and those with dyslexia, schizophrenia, bipolar disorder, and autism, for example (together with their relatives, who may, advantageously, carry some, but not all, of the genes for the condition), are the least likely to show what might be called ‘left-hemisphere encapsulation’. In other words, in the normal brain the serial processing that forms the basis of left-hemisphere function is carefully segregated from functions that it might impair, but the corollary of this is that the holistic approach of the right hemisphere is not available to the same extent for language and conceptual thought. In anomalous lateralisation patterns, this segregation no longer occurs, with reciprocal advantages and disadvantages. This would result in some gaining access to particular talents from which the rest of us are debarred (‘flying mathematicians’), and some faring worse, and losing the evolutionary advantages of specialisation (‘inhibited trapeze artists’). This view is compatible with the available large body of evidence that there are both special talents and handicaps associated with anomalous cerebral organisation in these conditions, and in the relatives of those with such conditions.

This is clearly a huge topic, which deserves more analysis. The point I wish to emphasise here is that the left hemisphere has to ‘blot out’ the right hemisphere in order to do its job at all. That is surely the import of the functional and anatomical evidence that left hemisphere superiority is based, not on a leap forward by the left hemisphere, but on a ‘deliberate’ handicapping of the right.

If we are to understand the relation between the hemispheres, and their possible rivalry, we need to compare the two experiential worlds that the hemispheres produce.