A response to commentators

I would like to thank all the respondents for their comments, which offer me the opportunity for a number of clarifications. Many relate to the topic of generalization. “To generalize is to be an idiot,” warned Blake: “to particularize is the alone distinction of merit.” The only trouble with that is that it is a generalization. One needs details, and nuances, which are what get shorn off in summary papers; and it is certainly more comfortable to remain with the details. But in science no progress is made that way. As the great evolutionary biologist and palaeontologist George Gaylord Simpson wrote:

Science, truly to be such, must centre not on descriptions and names but on principles—that is, generalizations, theories, relationships, interconnections, explanations about and among the facts. (Simpson, 1963, p. 82)

That requires imagination, making analogies, seeing connections—and the willingness to stick your neck out. Again I quote, this time from Sir Peter Medawar:

Scientists should not be ashamed to admit, as many of them apparently are ashamed to admit, that hypotheses appear in their minds along uncharted by-ways of thought; that they are imaginative and inspirational in character; that they are indeed adventures of the mind. (Medawar, 1964, p. 43)

No generalization can go unqualified—even this one. While some are unfounded, some are the result of years of experience, reflection and research. When the research is shorn off and the conclusions stated baldly, without space for qualifications, one offers hostages to fortune. In a short article, I had to be very general, and I had to be very brief. I had hoped that might be obvious.

But the charge of “generalization” brings up a point of central relevance: not seeing this accounts for almost all the misunderstandings of my hypothesis. Science begins from defined questions. Many, if not most, neuroscientists are engaged in localizing functions in the brain. They therefore are used to people starting from some commonly agreed “function,” and asking the question “where in the brain is function x?” To which the answer will usually be “a bit here and a bit there, involving both hemispheres.” The questions I am asking are quite different, namely “Why is the brain divided? Is this because they are engaged in delivering different aspects of human experience? And if so, what are those differences?” I am thus working in the opposite direction: not starting from a circumscribed function and trying to localize it, but starting from a system (a hemisphere) and trying to characterize it. That difference is all-important for evaluating my hypothesis. For example, Kundu & Smith state that I am “too biased to finding whole hemisphere differences.” But my purpose is, precisely, to discover what those differences are. Note that “whole” can mean two different things in this context: “different when taken as a whole,” or “wholly different.” The first sense, the one that interests me, does not imply the second. It is the differences—however small or large—that are of significance to my hypothesis, not the similarities. If I am to characterize the difference between Albert Einstein and Donald Trump, it is pointless objecting that the differences I notice are few compared with the countless points they have in common. Or in buying a car, it may be quite true that a new Ferrari and an old banger can each carry out much the same “functions.” They just do them in an importantly different way. And in order to characterize that way, it is precisely the differences, not however many similarities, that matter.

As it happens, there are emergent phenomena from the whole interconnected system at the hemisphere level, and it is increasingly recognized that we must no longer think in terms of isolated modules, but rather in terms of complex widespread networks. The hemispheres are massively more intraconnected within themselves than they are connected to one another and are the most important examples of such widespread networks in the brain. I do not deny the very obvious fact that regions within hemispheres are important, if you want to identify all the brain correlates of a specific function. (In The Master and his Emissary there is, in fact, much detail about localization). However that is to answer a different question from the one I am asking.

Another issue is at stake: how to make generalizations and how to rebut them. Without generalizations we can get nowhere: it is fair to say that in the life sciences, as in life, almost all claims are generalizations, admitting of counter-examples. Men, for example, are indisputably taller than women: but many women are taller than many men. It therefore matters that one's generalizations reflect a thorough knowledge of the evidence; and equally they are not to be disproved by a small minority of cases that do not fit an otherwise clear pattern. There is never 100% agreement among experimental evidence. The art of doing science properly means acquainting oneself thoroughly with the literature and making a just assessment of where the balance of evidence lies. Rebutting a generalization means doing the same and reaching a different conclusion which can then be judged in turn. And never to make any general judgments, even if it were possible, which it clearly is not, would buy immunity from reproach only at the price of emptiness.

I will begin with de Haan's piece, since his misunderstandings are the most instructive. De Haan opens by exhibiting our old friends, the homunculus fallacy and the mereological fallacy, and claims that I commit both. This follows from his belief that I purport to explain psychological phenomena, and in doing so reintroduce the same explanandum as if it were an explanans.

The homunculus fallacy

De Haan gives an example of the sort of thing he has in mind: empathy being “explained” by the empathic abilities of mirror neurones. It is true that, regrettably, many neuroscientists and some philosophers do make empty claims of this kind: e.g., that falling in love is explained by the “neural circuitry” which appears to “operate” when someone falls in love. I make no claims of this kind. I do not link phenomena to brain regions in order to reveal how it is that we come to have the experience, or what the experience “really is.” I do not therefore require a homunculus.

It would be worth distinguishing at least four kinds of explanation of any phenomenon. Each of these is a different kind of explanation, addressing different kinds of question.

1. What is it?

   There is nothing else in terms of which to approach experience. I do not claim to know what it is any more than anyone else, except that it is the condition on which I know anything at all. It is what we know (kennen) better than anything, and yet know (wissen) least of all about.

2. What gives rise to it?

   Despite there being an extraordinarily intimate connection between brain events and experience, in which if one is altered, so is the other, I make no claim as to how it works. I do not even assert that brain events originate experience at all (they may or may not), let alone how they do it. (Similarly Kundu & Smith are dismayed that I “conflate” brain localization of function which is measurable, with “how experience arises from neural processing, which is metaphysical.” I do not address how experience arises from neural processing at all, and so do not “conflate” it with anything.) Unlike Dennett, who notoriously claims to explain consciousness, I do not; likening our logical positions shows that de Haan has got hold of the wrong end of the stick. “How do the hemispheres yield their very own coherent experiential worlds? To explain this magical yielding would solve the mind–body problem,” writes de Haan. Quite so; but I do not claim to have solved it. And I do not need to. If the brain (however it may do it) yields our “take” on the world—which seems a reasonable formulation, since in its absence we don't have one, and by altering the brain we alter our “take”—then the hemispheres can also, in principle, yield “takes” on the world, since each is capable of sustaining consciousness on its own. Those two versions of the world have predictably different qualities: the focus of interest in my paper.

3. What constrains its possible manifestations?

   All experience in this life as we know it (and this applies whether we conceive the brain as originator, or as transducer, of consciousness) comes to us through the brain, and is therefore inevitably constrained, and shaped, by it. At the simplest level it is obvious that some animals can experience things we can't, and vice versa, since our nervous systems are different; and when the nervous system is damaged, through some stroke, injury or other brain insult, aspects of normal experience disappear from the subject's world. There are specific patterns to this which reveal the specific contributions of each hemisphere, and it is this that I aim to elucidate.

4. What determines the particular shape it takes at any one moment?

   This I make no claim to “explain.” Here there are a whole range of possible explanations, and what counts as a possible cause depends on the context of the question. I do not invoke the brain in order to trump other orders of explanation, be they psychological, physiological, anthropological, sociological, environmental, genetic, economic, or political, especially since they are inextricably intertwined one with another. My argument is not in opposition to such explanations, but complementary to them. Arnold supposes that I am “explaining” religions: “McGilchrist's many enthralling insights regarding brain-lateralization cannot, in principle, explain anything like the rise of Calvinism,” and so on. Of course not. But when Calvinism arises, for whatever constellation of reasons, we can recognize a pattern to the phenomena we observe: no longer isolated elements, the Gestalt coheres as the expression of a way of being in the world we can observe, mutatis mutandis, elsewhere and at other times, and which is associated with disregarding to some extent what the right hemisphere might help us see in the world. We begin to see patterns in the history of ideas driven by the human need for expressing both of two conflicting, but necessary, ways of being. We are, I assume, free to make choices, though there are a large number of influences on those choices: and yet some aspects of experience are not in our power to choose. Thus, I am free to choose to visit Iceland rather than Indonesia, and many different kinds of explanation could account for my decision. However I am not free to choose the climate, landscape or language I will encounter when I get there. Thus many paths led to the Reformation; but nonetheless the Reformation expresses—although in its own way—a recognizable and coherent pattern to those who understand hemispheric differences. Such an insight does not compete with, let alone replace, but, rather, complement, other ways of thinking about shifts in the history of ideas.

De Haan does not distinguish between the claim that mirror neurones possess “imitative and empathetic abilities,” which they don't (I doubt if many people believe that they do), and the perfectly valid claim that knowing about mirror neurones tells us something about how closely perception and action are linked in the brain: that at times they share the very same neuronal activity. This tells us something about how closely watching something done by or to another is linked to experiencing it oneself. Here is an example of brain structure and function telling us something about the psychological phenomena of perception and action, without itself being the psychological phenomena, or claiming anything at all about how the link between brain and mind works: it is enough that it does. No homunculus is involved.

Where I think de Haan goes wrong is that, having decreed linguistic distinctions that are too rigid, he then finds that these rigid distinctions are not a good fit with experience and that something is being left out. This is, I find, a common problem for what one might call “triple-A” [Anglo-American Analytic] philosophy. Maintaining important differences does not require rigid distinctions. Both de Haan and Kenny, to whom he refers over the homunculus trope, do make revealing qualifications, which suggest that they realize there is something wrong with hard and fast distinctions between persons and their brains. Why, then, does de Haan make them? Because he commits the mechanist's fallacy, and he is drawn to do so by the language he employs. Once you have effectively reduced the human to the mechanical, it is you who find the need for a homunculus in order to smuggle the human back in. I don't commit the mechanist's fallacy, so for me there is no need.

The mechanist's fallacy

This widespread fallacy lies in relying on analogy with a mechanism as the best, if not the only, way to understand a complex, organic phenomenon. One complication of the fallacy is that, when applied to people, it results in a Cartesian split of mind and body, because there is no place for a ghost in a machine. The fallacy is sometimes explicit, but more often implicit, and is by now so pervasive in neuroscience culture and triple-A philosophy that most are not even aware they are making it.¹ We gain an understanding of anything by making an analogy with something else of which we are prepared to say “I understand that.” Machines are often the chosen model because machines are peculiarly easy to understand—after all, we made them from parts, for our own purposes; but they are also, for that very reason, unlike almost everything else that exists—especially organisms.

The language we use and the model we espouse govern not only what we see, but what sort of thing we take it to be. We do not have a particular language for describing what a brain is or does, partly for the obvious reason that we have not had time to evolve one. We have therefore two options, neither of them ideal, but one much more damaging than the other. We can talk about it as though it had the characteristics of a machine (an almost invariable rule in science and philosophy publishing nowadays); or as though it had at least some characteristics of a person—my preferred choice, since every change in the brain is immediately reflected in a change in the person's experience. The “person” model looks like a much better fit than the “machine” model, especially since a hemisphere is at least a manifestly important part of a person, and isn't like any machine we have even got near inventing.

But of course the brain isn't actually either simply a machine or on its own a person. These are both just models that illuminate and obscure different aspects of what they model. We might need both. If you prefer not to abandon the person model, you will inevitably say that the left hemisphere “thinks,” since it is unreasonable to repeat endlessly “a person relying on the left hemisphere thinks.” Most people find that easy enough to understand.

One thing that Spezio, and possibly to a lesser extent de Haan, clearly find perplexing is my reluctance, wherever possible, despite decades of indoctrination, to use approved “neurospeak”: the usual “circuitry,” “modules,” “functions” and “mechanisms” of every conceivable (and many inconceivable) kinds. “Why not employ the language of neuroscience and NMP [New Mechanistic Philosophy] to talk about the brain?” asks de Haan. There are a number of answers to that question.

The first is that my whole purpose is to get away from the assumption that the brain is effectively a machine, because I think talking about it in this way skews our understanding in unfortunate ways and prematurely answers questions that still need to be independently addressed. More than this, much discussion in neuroscience has lost sight of the context of the whole human person, and their being-in-the-world, something vital to the real meaning of all neuroscience research. You can see this in the responses from Kundu & Smith and, especially, Spezio. When the phenomenal level enters into discussion, it is already reduced to terminology suggesting that, e.g., attention, memory, or affect are themselves quasi-mechanistic functions, viewed “from the outside.” We need to rehumanize discussion of the brain, and its relation to the whole person's experience “from the inside.” I therefore avoid the assumptions embedded in mechanistic language wherever possible. For example, attention is not just another brain “function,” but nothing less than the manner in which we dispose our consciousness towards the world, which has many consequences for what we take the world to be, and how we see ourselves and our relationship with it. Machines can't dispose their consciousness, not having any. Neurospeak obscures this very important perspective (as it does when applied to memory, emotion, and so on).

Furthermore, organisms, in general, are simply not mechanisms (Nicholson, 2013). If you restrict your focus to a tiny part of what is in reality a massively complex system, you can isolate, and, if desired, interfere effectively with, a simple chain of causation—a mechanism. So the model tells us something important at the local level, I agree; but it is a poor model of the whole. Most importantly for this discussion, machines involve a quite different part-whole relationship from that which obtains in an organism, though there are many other important differences.²

Not only do the two hemispheres have radically different phenomenological takes on the world, they also somehow possess their “own goals, concerns and values” and have “quite distinct ‘take[s]’ on every aspect of human experience and behaviour.” These are extraordinary attributions!

So de Haan observes, with more than a hint of mirth. Let us pass over, for now, the logical necessity for this to be the case, if (a) the hemispheres pay different kinds of attention (they indisputably do); and (b) the nature of attention changes the qualities of the world (it indisputably does). That aside, how extraordinary these attributions seem is inversely proportionate to how much you know about hemisphere lateralization.

De Haan refers to “purported scientific evidence that ‘each hemisphere on its own is capable of yielding a coherent experiential world.’” There is, however, nothing “purported” about it. Each hemisphere is indeed irrefutably capable of yielding a coherent experiential world, as is obvious from the case of hemispherectomy patients, split-brain patients, stroke, tumor and head injury patients. Not only that, but seeing how such a world is differently characterized according to laterality is manifest from a rich early twentieth century, largely German (e.g., Hoff, Pötzl, Zingerle, Lange, Pichler, Ehrenwald) and French (e.g., Hécaen, Ajuriaguerra, Alajouanine, Morlais, Tatossian) literature, as well as subsequent contemporary phenomenological or naturalistic accounts, such as those of Cutting, Ramachandran, Sacks, Bisiach, Feinberg, Brugger and others.

In a 2003 paper, four researchers who spent many years examining split-brain patients commented in a jointly authored paper in the classic textbook, Clinical Neuropsychology, that, with experience,

examiners spontaneously refer to the two hemispheres as if they were distinct people, e.g., “the LH was upset at the RH responses today.” While such references may be regarded as shorthand for patterns of behaviour with specific lateralized stimuli and responses, they nonetheless express a strong phenomenological sense of two coexisting streams of consciousness. Both hemispheres can probably be simultaneously and independently conscious; both can simultaneously possess conflicting wills so that the split-brain can exhibit two distinct, and possibly incompatible, loci of moral responsibility.³

Equally research on split-brain patients taught Roger Sperry, who won a Nobel prize for this work,

that the two disconnected hemispheres function independently and in effect have each a separate mind of its own. Each of the separated hemispheres appears to have its own private sensations, perceptions, thoughts, feelings, and memories. Each hemisphere has its own inner visual world, each cut off from the conscious awareness of the other.⁴

More recently, Onur Güntürkün, who, in 2013, won Germany's most prestigious science award, the Gottfried Wilhelm Leibniz-Preis, precisely for work on lateralization in the nervous system, has publicly supported—referring explicitly to my work—the idea that the hemispheres underwrite differing aspects of personality, as well as, of course, attention and cognition.⁵ That Bennett & Hacker, who may or may not have studied neuropsychiatric patients, have a problem with all this is evident, but I suggest that the problem is of their own making, because of the limited model of a person on which they rely.

The tendency in recent years has been to downplay clear hemisphere differences—possibly part of the general denial that surrounds, or has surrounded until very recently, the topic of hemisphere difference. When they have been referred to, the point made has almost invariably been, not that Sperry was wrong to see what he saw—Sperry arguably spent more time observing such patients than anyone who ever lived, and his conclusions were in accord with those of other observers, so that would be unlikely—but that it is a mistake to extrapolate from the observed differences in the hemispheres of a split-brain subject to the hemispheres of a normal subject, where such differences are not commonly manifest.

But this is not right, either. Real differences may be apparent only when circumstances permit. Each hemisphere can obviously sustain consciousness on its own—and appears to sustain not just different “cognitive strategies,” but different goals, values, opinions and emotional timbres, all of which are inevitably interconnected with one another, and with cognition. The human psyche is not neatly compartmentalized into distinct “functions” in the way that, for example, Kundu & Smith and Spezio seem to think. The evidence is overwhelming that personality, emotion, cognition and action are not isolable one from another, but form continua that are only artificially divided by the conventions of the laboratory. As I have pointed out, on introspection we are unable to detect that we have two kinds of experiential world, for the very good reason that the process of fusion goes on below the threshold of consciousness. The fact is manifest, however, as soon as one hemisphere, for whatever reason, stops functioning. This can be seen experimentally by temporarily suppressing activity in one hemisphere at a time in normal subjects—it doesn't require a commissurotomy. These different experiential worlds are then clearly, immediately, demonstrable. Are we supposed to believe they are suddenly created ex nihilo?

The precision fallacy

I mentioned that Kenny and de Haan reveal an unease about the possibly too rigid boundaries they draw. Thus de Haan states that referring to mirror neurones as an “explanation” for empathy is, “without any additional qualifications,” a pseudo-explanation, prompting the question, what additional qualifications would be needed to make it a genuine explanation? Kenny, quoted by de Haan, refers to “the reckless application of human-being predicates to insufficiently human-like objects.” This raises the by no means negligible question, what would be a sufficiently human-like object? Hacker points out that for us to walk we need our motor cortex to function, but we don't walk with our brains (Smit & Hacker, 2014, p. 1082). Indeed—that's not how we express it. But we do say we walk with our legs: does Hacker not walk with his? And are our legs wholly separate from our brains? Clearly not. This is ground where it is wise not to be too clear-cut in one's judgments. As a general principle, I believe one should be every bit as precise as the situation permits, but not a whit more. Precise means (L, prae-cisus) cut off too soon, which sometimes turns out to be remarkably apt. Sometimes to be more precise is to be less accurate (L, ad+ -curare, to take care). The precision fallacy, in other words.

The so-called mereological fallacy casts further light on the problems of insufficiently flexible categorization. Let us turn to it next.

The mereological fallacy

This is most often imputed in situations where the fallacy lies not in the disputed assertion but in the imputation of fallacy itself: the “mereological fallacy” fallacy.

In de Haan's words, the “mereological fallacy concerns the mistake of ascribing to parts, attributes and predicates that can only be attributed to a whole.” Everything here hangs on the kind of whole, and the kind of parts, one is dealing with. Although I understand triple-A philosophers’ desire to be cut and dried, it isn't helpful here. Distinctions need to be made and graded judgments reached.

It is one thing to claim, as I do, that the whole is not the same as the sum of the parts, and quite another to claim that attributes of the whole cannot be attributed to a part. Note, however, that de Haan does not in fact make the latter claim. Rather he describes it as a mistake to ascribe to parts “attributes and predicates that can only be attributed to the whole”—which begs the question which can, and which can't.

There are indeed many instances where such attribution may be perfectly in order, and these include the cases of most wholes that are not mechanisms. This is, as I have mentioned, because mechanisms have quite particular properties that mark them off from everything else that exists. If we could not attribute any of the attributes or predicates of wholes to parts, we could not ascribe any of the attributes of, say, Elgar's Nimrod (or Jupiter's red spot, or a waterfall, or a landscape) to any part of it—and the same might indeed be said of most aspects of the world that are not mechanical.⁶

The concept of the “mereological fallacy” is the brain child, if they will pardon the expression, of Hacker & Bennett. Hacker's model is, as models in the Cartesian tradition are, mechanical: to quote him (and co-author Smit) on the subject, “it is the bracket clock as a whole that keeps time, not its fusée” (Smit & Hacker, 2014, p. 1085). However, the important question is, surely, not what keeps time—both the parts and the whole are seamlessly involved in that (a metronome keeps time, and similarly a pendulum that won't keep time produces a clock that won't keep time), but what tells the time: a human consciousness external to the clock-system has to be imported. We, however, are not mechanisms, nor are we homunculi distinct from, and reading out from, a mechanism (the brain), much as a person reads a clock. We experience and tell time seamlessly, brains and all; and as it happens we can't tell time properly if we have only our left hemisphere to rely on. Temporality, a very important part of personhood, then becomes grossly abnormal.

Machines are, indisputably, made of parts. Organisms are not: they grow and differentiate in the process, leaving us to create “parts” artificially post factum. Much depends also on the degree to which one decides to take apart (“analyze”) a whole. Though 10 bars of Nimrod has many of the same qualities as the whole piece, the single notes of which it is composed clearly do not. Closer to home, brains have attributes that neurones don't, and persons have attributes that brains don't. But that doesn't mean the distinctions are absolute.

Some degree of hesitancy over the relations between person and body (or brain) is not only natural, but a benign sign, a sign that one has not fallen into any of the obvious traps. Turning from clocks to brains, Smit & Hacker seem, encouragingly, to hedge their bets about the relationship between brains and persons. “Perhaps,” they begin promisingly,

we would hesitate to say that the brain is part of a person. But even if we grant that for the sake of argument, it does not follow that the brain is not part of the human organism—the human being—that we are. (Smit & Hacker, 2014, p. 1084)

Since parts of the person alter or disappear at a stroke (literally) when parts of the person's brain alter or disappear, it seems that we could do worse than consider the brain to be part of a person. Still, Smit & Hacker's agreement with “human being” is good enough for me.

So, now that's agreed, what is the relationship between the part and the whole? As far as I’m concerned, a part is never the whole; and no, we are emphatically not just our brains, despite the suggestions of some neuroscientists to that effect. That is because we are complex, irrefrangible wholes, which means we can't just dissect out the brain and discard the person. But, note, this “cuts” both ways: Hacker, de Haan and their ilk can't dissect out the person and discard the brain. People and their bodies are far more deeply interfused than they seem to realize.

The topic of parts and wholes is vitiated by the belief that the linguistic terms refer to delimitable entities in the real world. If we think of ourselves as limited to the bit of our consciousness which is aware of itself and capable of verbalizing experience, we have drawn the limits too narrowly. It's not as if our conscious selves have a very close relationship with our brains and with our body as a whole, such as a married couple might have: we are, at least during life, completely inseparable, a relationship that is strictly sui generis, and so cannot be properly compared with anything else, especially not with a clock. The rest of the body is inseparably interconnected with the brain in a myriad of complex ways: and the body is part of what it means to be us, since we are embodied beings. I suggest it is wise neither to collapse persons to their brains, nor deny their sometimes common features.

I like the reverence for the whole expressed in the idea of the person: it speaks of an insight into experience.⁷ Looking inwards we see persons and brains as mutually self-constituting; looking outwards, we see persons and societies as mutually self-constituting. In life, and in living beings, the whole and the parts cannot be cleanly dissected in the way triple-A philosophers seem so desperately to crave. Some degrees of attribution of some attributes to some parts of some wholes will be fallacious to an important degree, others will be either not at all or at any rate less fallacious than the alternative, to deny common attributes altogether. Triple-A philosophers are often blinded by their supposedly clarifying terminology to such an extent that they fail to see what is in front of their noses. They engage in clarifying their terms to such a degree that they obfuscate life. Words must form themselves around reality, not reality around words.

If we stand back from this person-chopping—not to say logic-chopping—fest, and think about what really matters here, we can see that we must urgently stop using machine language about parts of people, and start talking about them as if they were just that—integral parts of real people.

De Haan raises three further objections, which, I will have to deal with quickly, so as to move on to other contributions. One concerns levels of experience; another concerns normative judgments; and the third concerns the “serious empirical problem” of the validity of hemisphere differences.

Levels of experience

De Haan is concerned that I have committed an “absurdity” by imagining that there can be an aspect of “experience” (that made possible by one hemisphere) that “no one experiences.” But this is surely trivial. Is what is experienced by the body and mind as a whole, though not entering into the awareness of the conscious ego, experienced by no-one? I am conscious of only a tiny part of my activity. What is present to me now as I write these words is the tiniest fragment of all that I encompass; we discriminate, reason, make judgments, solve problems, take decisions, weigh possible outcomes, imagine possibilities and struggle to balance competing desires and moral values all without being aware of it. Or should I repudiate “we” and say that all this is done by “no-one”? Regardless of what you choose to call it, all this activity still goes on, and in so doing helps constitute my world and who I am. Sooner than place the Cartesian ego first, I would place first and foremost thoughts, feelings, imaginings and responses to the world, embodied and embedded in the world—experience—and allow the ego to take a back seat. Even Descartes derived the ego from experience, not experience from an ego. Both hemisphere “takes” can be and often are taken into account together, sometimes more consciously, sometimes less.

Normative judgments

De Haan rightly observes that I often prefer the “take” of the right hemisphere, but can see no grounds on which to base a normative judgment. Arnold makes the same, perfectly valid, point in his paper. I deal with the issue of which hemisphere may be considered to deliver a more reliable account of the world at length in a book I am currently writing. Assume, though, for now, that I could show you evidence that the left hemisphere takes in less of the world than the right; is more likely than the right to be, literally, deluded; is less perceptive; is more likely to jump to false conclusions, and to make poorer judgments; is more likely to misunderstand social situations; and makes less of a contribution to both emotional and cognitive intelligence. To me that would represent grounds for making a judgment in favor of the right hemisphere as more reliable. And I do have such evidence.

Arnold additionally comments “Ought we, though, to prefer deontology to utilitarianism in ethics because the former reflects RH predominance?” (I remind the reader that the word “deontology” is derived from the Greek word for duty, δέον, not from the Latin word for God, deus). I have just given a partial answer, but there is more on this specific matter. Suppressing the right temporoparietal junction leads to concluding that an accidental poisoning is a morally worse act than a seriously attempted poisoning that fails (Young, Camprodon, Hauser, Pascual-Leone, & Saxe, 2010; see also Miller et al., 2010). Those with frontal damage, especially right frontal damage, make more utilitarian judgments.⁸ Deontological moral judgments are associated with normal social emotion processing (Martins, Faísca, Esteves, Muresan, & Reis, 2012); those with emotional blunting make more utilitarian judgments (Koenigs et al., 2007). The tendency to make utilitarian judgments is more marked in those with reduced aversion to harming others (Cushman, Gray, Gaffey, & Mendes, 2012), lower trait empathy (Choe & Min, 2011), and higher psychoticism (Wiech et al., 2013). Taken together, that seems to me significant.

The “serious empirical problem”

At this point I can move from de Haan to a more general discussion, since the validity of the hemisphere differences I describe is a matter of more general debate. De Haan says that I rely on “a certain construal of evidence from neuropsychology and cognitive neuroscience that has been confuted,” and that the hypothesis is therefore “untenable.” He insinuates that my understanding lacks complexity and nuance (I will come to that next).

But first, what makes him so sure that my construal of a large body of evidence has been confuted? My construal and hypothesis are crucially not a repetition of any that went before. It is not the same one that according to him has been confuted. There would have to have been a far more complex and nuanced debate than any of which I am aware, of my specific hypothesis, not just of the earlier (I agree, mistaken) formulations of hemisphere difference from which I repeatedly distance myself. The “confutation” of the old hemisphere story is where I begin, not end. My claims are new, supported by a mass of research that would need careful evaluation, and are many stranded: the whole is indeed complex and nuanced.

Brain science is a bit like particle physics: there are multiple hypotheses, and each has its adherents. Science, being done by fallible human beings, has its fashions and its factions. Even the world's greatest expert, necessarily on a limited field, must take everything outside his field in science on authority. Many authorities in my area, however, including such names as VS Ramachandran, Jaak Panksepp and Howard Gardner (there are many others), wouldn't agree with de Haan that there is nothing in the divided brain hypothesis I put forward. Real scientists should be curious about something different or new, not pretend to know in advance of a detailed examination that it must be wrong.

About nuance and complexity I have said something at the outset. There is an insurmountable problem in condensing a long, detailed exposition over about 500 pages, decomposable into many sub-theses, involving disquisitions on many topics, with qualifying excursuses, many, at my editor's correct insistence, in footnotes, into a few paragraphs. This is what I had to do in a couple of pages at the outset of the target paper. Furthermore in a short piece you are forced to make general statements that would otherwise be qualified—the price of not doing so is that you end up covering very little ground. I wanted to cover a lot of ground, because the Gestalt I am offering requires seeing the whole: it is not like building a wall, one brick at a time.

Which leads to a further problem: that of the status of the synoptic view in science today. Again I have alluded to that at the outset. The modern academic world is less congenial to synoptic thinkers than it used to be; academics are now trained to be more skilled at seeing the trees than the wood. (That doesn't mean that the science establishment doesn't make important generalizations, and start from largely unexamined assumptions.) Amassing details is important, but there is no point in finding out more and more details if one has no idea of the big picture they go to compose or enrich. Indeed, in doing so, one mistakes the nature of the details, too.

In summary, taking a synoptic view (though one that is argued for at book length, in detail, and backed by extensive reference to the literature) is a risk; drastically compressing it for the purposes of a paper increases the risk by an order of magnitude; and unpacking its implications for religious experience in what cannot but be general terms compounds it still further. When you take such risks, you know you will not get through to some readers. But, no, it has not been confuted and is by no means untenable. It is only just beginning to be debated.

Turning to the paper by Arnold, he makes a number of points that I have already addressed, one being that I “trade” in generalizations. More specifically Arnold makes one point repeatedly: that “Oriental religions” are not monolithic. Oddly enough, although I mention Buddhist meditative practices in passing, I do not refer to “Oriental religions” (or “Eastern religions”) in my paper at all, nor do I anywhere in The Master and his Emissary. There I did speak briefly about experimentally verified differences in perception and conceptualization between specifically East Asian (Korean, Chinese and Japanese) and Western cultures, though they were not presented as absolute. I had nothing to say about the Indian subcontinent (Arnold's principle topic of concern), or its religions. But Arnold is surely right that distinctions between Oriental cultures are important. Like everything, though, it depends on the context. For example, it is usual to refer to Western culture, when one is characterizing its difference from other cultures, though naturally there are many cultural strands within it, that in another context would need to be distinguished.

The art of understanding involves both seeing differences and commonalities. Academia is better at the first than the second. However I think there are patterns that will simply be invisible unless one is prepared to make some generalizations. Joe Henrich, the Harvard anthropologist who is credited with coining the acronym WEIRD—Western, Educated, Industrialized, Rich and Democratic—to describe the peculiarity of the modern West, told me that when his team were investigating the psychology of pre-industrial peoples across many continents, they constantly commented that their subjects were “more Chinese than the Chinese.” The generalization might scandalize Arnold, but has real meaning. They observed what I would describe, in shorthand, as a different balance of hemispheric influence in their thought when compared with modern Westerners.

Incidentally, contra Arnold's implication, I do not claim that “right-hemispheric modes of attention are predominant among people in Asia.” In The Master and his Emissary I say: “What the evidence suggests, if reviewed in greater detail than I have here, is that the East Asian cultures use strategies of both hemispheres more evenly, while Western strategies are steeply skewed towards the left hemisphere.” (McGilchrist, 2009, p. 458).

Arnold raises a new point to do with reductionism, making the claim that Buddhism is reductionist. Once again, I do not mention reductionism in the paper, apart from stating that the purpose of Michael Trimble's work is not reductive (as neither, of course, is mine). I understand reductionism to mean the assumption that the best way to understand anything is by reducing it to the parts of which it appears to be made. Nothing in Arnold's account suggests that this is a Buddhist view. Reductionism is usually materialist, though that is another issue; he quotes approvingly, there is “no other philosophical view that is more radically opposed to the tenets of Buddhism than materialism.” Nothing here makes me want to characterize the “Buddhist tradition as being, in McGilchrist's terms, generally left-hemispheric.”

In any case, I do not argue that all aspects of all religions are right-hemispheric in nature—clearly they are not. Indeed I make the point that some problematic forms of religion, such as fundamentalism, seem dominated by left hemisphere thinking.

Non-standard categories, wholes and parts

Kundu & Smith complain that I do not adhere to “the domains of function that have [been?] standardized by consensus in neuropsychology.” This is an example of mistaking the forms of science for science itself, a point which is even more obvious when we come to the case of Spezio. That, for example, dimensions such as reasonableness v rationality, certainty v flexibility, fragmentation v wholeness, optimism v pessimism are not ways of thinking about human psychology that get included in modern lab work does not necessarily mean that there is something wrong with these dimensions; it might suggest that the categories in which cognitivist psychology now thinks are too restrictive. The conventional categories fail to capture the fact that the brain is not a computer, but part of a person—to whom all such elements of experience are of the highest importance, as they are in philosophy. Real science, as I feel sure Kundu & Smith would recognize, is tentative, adventurous, involved in questioning assumptions, using the imagination and recasting whatever it is we think we know; the institutional form of what now passes for “science” in many university departments is nothing of the kind.

Moreover this goes back to the matter of which questions one asks. If I were engaged in localizing, say, a “reward circuit,” it would be important for me to use the same categories and definitions as other researchers, in order to make my results comparable. But I am starting without preconceptions about any such thing as a “reward circuit,” and instead asking “what is it like for human experience when this brain region is active or ceases to function?” That may show up in all sorts of human ways that we recognize in life, rather than in a list of functions drawn up by neuroscientists themselves. Let me be clear: my aim is not to find “centers for,” e.g., certainty or optimism. My aim is the humanistic one of relating the brain to phenomenology, and understanding the overall difference between the hemispheres.

Scans

Like Spezio, Kundu & Smith assume that “low resolution” methods such as the study of brain lesions are inferior to “high resolution” techniques such as functional brain imaging. As it happens I cite evidence from many different methodologies, including a great deal from functional imaging, but also including many types of behavioral psychological studies in both normal and split-brain subjects, EEG, TMS, tDCS, and lesion studies. Functional imaging is not superior to other methods, but has some advantages, and some disadvantages, in comparison with other forms of localization.⁹ In particular, it is likely to fail to find important lateralization effects.¹⁰ (I will refer to others below.) As Rorden & Karnath point out, “techniques such as fMRI (which measures brain activation) offer a weaker level of inference than the lesion method (which measures brain disruption) … a task that is processed entirely in the left hemisphere might produce bilateral activation because of the strong homotopic neural connections between the two hemispheres.” Rorden & Karnath's paper comparing the strengths and weaknesses of different methods conclude that “[t]he strength of cognitive neuroscience comes from using convergent tools to investigate the same theoretical question.” (Rorden & Karnath, 2004). The long and short of it is that, as Price & Friston suggest, lesion studies are a vital ingredient in a mix which includes imaging studies; and a combination of approaches is best (Price & Friston, 2002). Or as neuroimager Stanislas Dehaene puts it: “Some people believe that psychology is just being replaced by brain imaging, but I don't think that's the case at all … It's the confrontation of all these different methods that creates knowledge.”¹¹

Errors of fact

Kundu & Smith also claim that I am factually mistaken in places. About the divergent forms of attention in the two hemispheres, they comment that “it is apparent that such diametric opposition is not supported.” But it is (see, e.g., Benson & Barton, 1970; Benton & Joynt, 1959; Bisiach, Mini, Sterzi, & Vallar, 1982; Brederoo, Nieuwenstein, Lorist, & Cornelissen, 2017; Chechlacz, Mantini, Gillebert, & Humphreys, 2015; Corbetta, Miezin, Dobmeyer, Shulman, & Petersen, 1991; Çiçek, Gitelman, Hurley, & Nobre, 2007; de Renzi & Faglioni, 1965; Dee & van Allen, 1973; Delis, Kiefner, & Fridlund, 1988; Delis, Robertson, & Efron, 1986; Deouell, Ivry, & Knight, 2003; Dimond, 1979a, 1979b; Flevaris & Robertson, 2016; Godefroy & Rousseaux, 1996, 1996; Godefroy, Lhullier, & Rousseaux, 1996; Halligan & Marshall, 1994; Heilman & van den Abell, 1979; Howes & Boller, 1975; Jansen, Sturm, & Willmes, 1992; Jerison, 1977; Korda & Douglas, 1997; Leclercq, 2002; Lewin et al., 1996; Mesulam, 2000; Nakamura & Taniguchi, 1977; Parasuraman, Greenwood, Haxby, & Grady, 1992; Pardo, Fox, & Raichle, 1991; Rueckert & Grafman, 1996; Salmaso & Denes, 1982; Siéroff, 1990, 1994; Sturm et al., 1999; Sturm & Büssing, 1986; Sturm, Reul, & Willmes, 1989; Tartaglione, Bino, Manzino, Spadavecchia, & Favale, 1986; Whitehead, 1991; Wilkins, Shallice, & McCarthy, 1987) (Diametric opposition is a statement about the difference in characteristics of the two types of attention, not an assertion of the absolute nature of hemisphere differences). They tax me with proposing a “‘religion circuit’ localized [to?] the RH”, but I propose no such thing. They wrongly reject the “claims that the LH is ‘unreasonably optimistic’ and RH ‘tends towards the pessimistic.’” The evidence, however, is plentiful.¹²

Moving on to Alcorta's interesting paper, it brought to my notice additional evidence supporting the divided brain hypothesis, for which I am grateful. She draws attention to the well-attested fact that hemispheric asymmetries have “deep phylogenetic roots.” In fact, as she will be aware, lateralization goes back into prehistory. The lateralized distinction between narrow-beam, highly focussed attention and broad sustained vigilance has been repeatedly described not only in mammals, but in amphibians, reptiles and birds,¹³ as well as in fish, which are more likely, for example, to bite targets viewed on the right, the left hemisphere focusing on prey, than on the left (Barth et al., 2005). The same predilections have recently been demonstrated in cuttlefish, who attack prey in the right visual field and scan for predators using the left visual field (Jozet-Alves et al., 2012; Schnell, Hanlon, Benkada, & Jozet-Alves, 2016, p. 620). Indeed lateralized behaviors have also been demonstrated in many invertebrates (Frasnelli, Vallortigara, & Rogers, 2012), including molluscs (Byrne, Kuba, & Griebel, 2002), bees (Frasnelli, Anfora, Trona, Tessarolo, & Vallortigara, 2010; Rigosi et al., 2015), ants (Frasnelli, Iakovlev, & Reznikova, 2012), spiders (Ades & Ramires, 2002; Heuts & Lambrechts, 1999), slugs (Matsuo, Kawaguchi, Yamagishi, Amano, & Ito, 2010), and nematode worms (Hobert, Johnston, & Chang, 2002; Taylor, Hsieh, Gamse, & Chuang, 2010). Asymmetries of behavior can be traced back more than 500 million years, and have been identified from fossils of trilobites (Babcock & Robison, 1989). In fact the earliest known instance of a neural network, that of Nematostella vectensis, a sea anemone over 700 million years old, already exhibits lateral asymmetry.¹⁴ As Güntürkün puts it, “if animals from nematodes to humans show asymmetries of brain and behavior, we should expect that left-right differences provide some important fitness benefits” (Güntürkün & Ocklenburg, 2017). Denying the importance of lateralization is simply not an option: it is the scientist's job, not to deny hemisphere differences, but to find out what lies behind them.

I would not however agree with Alcorta's suggestion that I am “dualistic.” I did not invent hemisphere differences; nature got there first. And while Alcorta is quite right to point out that “broad functional categories … subsume a multitude of finer-grained tasks involving structures from both brain hemispheres,” she is wrong to suppose I suppose differently. I must repeat that I know that each hemisphere is involved in every single thing we do—just in a quite different way. It's the differences that matter, and for that it's back to “the how, not the what”; a point which many scientists, used to thinking in terms of mechanisms, sometimes seem (I do not by any means necessarily include her) to have difficulty taking on board. She, Spezio and Kundu & Smith all appear to assume that I am starting from a function and trying to localize it, whereas I am working in the opposite direction: starting from a system (a hemisphere) and trying to characterize it. That difference is all-important for evaluating my hypothesis.

Now to Spezio. It is worth quoting here some important words of Oliver Sacks's that refer to the fascinating case of Dr P, the very “man who mistook his wife for a hat.” Dr P exhibited a panoply of symptoms suggestive of right temporoparietal failure. Sacks reflects:

The entire history of neurology and neuropsychology can be seen as a history of the investigation of the left hemisphere … it is the right hemisphere which controls the crucial powers of recognising reality which every living creature must have in order to survive. The left hemisphere, like a computer tacked onto the basic creatural brain, is designed for programmes and schematics; and classical neurology was more concerned with schematics than with reality … our mental processes, which constitute our being and life, are not just abstract and mechanical, but personal, as well—and, as such, involve not just classifying and categorising, but continual judging and feeling also. If this is missing, we become computer-like, as Dr P was. And, by the same token, if we delete feeling and judging, the personal, from the cognitive sciences, we reduce them to something as defective as Dr P—and we reduce our apprehension of the concrete and real. By a sort of comic and awful analogy, our current cognitive neurology and psychology resemble nothing so much as poor Dr P! … Our cognitive sciences are themselves suffering from an agnosia essentially similar to Dr P's. Dr P may therefore serve as a warning and parable—of what happens to a science which eschews the judgmental, the particular, the personal, and becomes entirely abstract and computational. (Sacks, 1986, p. 2, 3 & 19)

There is much to be concerned about in the contemporary cognitive science establishment. Not overprovisioned with self-doubt, rigid, narrowly focussed, opposed to any new ideas that would require thinking differently, or using imagination, mechanistic in outlook and neglectful of the big picture, while not being aware of all that it is missing—to a clinician, such as Sacks or myself, the likeness to certain of our patients is striking. Spezio is defensive about my suggestion that the culture has become left-hemisphere dominant. But by his own efforts he does nothing to persuade me I am wrong.

Speaking as a physician, and as someone with a training in the humanities, I cannot see the brain as merely a technical device. I am fascinated with what it is like for someone when their brain radically changes, and with it not just some “function,” but often their whole experienced world changes—as, from experience, I can tell Spezio for free it very clearly does. Understandably, then, my way of talking is not to his liking. Spezio complains I speak of what can't be measured. I don't “operationalize” or express myself “in terms of neural measures.” He points out that neuroscientific work I cite “actually never moves into his language or concepts of “mode” or “version” but remains grounded in cognitive systems and information processing theory.” The point being? Spezio insists I should speak of “neural coding,” as in “the activation in ventromedial prefrontal cortex encodes the expected reward probability of choice A”; or of “specific cognitive subprocesses, often modeled as computations.” I have explained why I go out of my way to do otherwise. And, once again, we have the problem that I am thought to be answering questions different from the ones I address.

On what grounds would a good scientist merely assume that there are no significant global differences between hemispheres? If Spezio makes that assumption, would he not at least be duty bound to take a careful look—and with an open mind? If he doesn't make that assumption, what would he think the differences might be? We are keen to learn. Meanwhile he creates the impression that there is no evidence for my thesis, yet any reader who knows a little about the area will not be tempted to believe him, as I will shortly demonstrate.

“Neuroscientific reductionism about cultural histories, at best, and a set of easily falsified claims at worst.” That is Spezio's summary. The first is a complete misunderstanding, as I have already explained; and the second is simply false. As I have already pointed out, the art of doing science properly means acquainting oneself thoroughly with the literature and making a just assessment of where the balance of evidence lies. This is what Spezio calls generalization. He makes it sound like a short cut to being just in one's appraisal; rather it is the goal of a long journey and requires a thoroughness I do not find in his contribution. To light on a handful of counter-examples and put them forward as a rebuttal is rhetorically cheap, but it is not science. It is to misuse evidence and show a cavalier disregard for the value of evidence as a whole. Spezio uses the phrase “sweeping claims” no less than 11 times, so that one begins to get a sort of feeling that Spezio doesn't like generalizations. But Spezio himself makes sweeping statements—this paragraph begins with one.

In his piece, Spezio certainly scatters his fire; indeed he makes so many idly dismissive claims, that it would be tedious to pick him up on them all, and I would need much more space and time than I have available to walk the reader patiently through the evidence as a whole. I will therefore limit myself to taking as representative examples his attempts to rebut the local/global distinction, and the metaphoric/literal distinction, since they are both undoubtedly relevant to my argument.

On the local/global issue, Spezio tries to discredit what is probably the single best attested distinction between the hemispheres (see e.g., Çiçek et al., 2007; Delis et al., 1988, 1986; Deouell et al., 2003; Dimond, 1979a; Halligan & Marshall, 1994; Leclercq, 2002, p. 16; Ivry & Robertson, 1998; Jerison, 1977; Kitterle, Christman, & Hellige, 1990; Kitterle & Selig, 1991; Martinez et al., 1997; Mesulam, 2000; Posner, 1995, pp. 615–624; Posner & Petersen, 1990; Robertson & Lamb, 1991; Robertson, Lamb, & Knight, 1988; Rousseaux, Fimm, & Cantagallo, 2002, pp. 289–290; Sergent, 1982; Siéroff, 1990, 1994; Siéroff, Decaix, Chokron, & Bartolomeo, 2007; Sturm et al., 2002, p. 370; van Kleeck, 1989) by quoting one paper that relates to the auditory modality (Sanders & Poeppel, 2007). First, discriminating local from global in the auditory domain is far less reliable than in the visual domain: it is even a matter of dispute what it means, let alone how one tests for it, while it is obvious in the visual realm what is meant. That aside, Spezio's one study is of 24 subjects, in an event-related potential (ERP) paradigm, all female. What the reader needs to know is that lateralization of any “function” is rarely found in ERP studies where otherwise well-attested studies report it (Han & He, 2003; Han, Fan, Chen, & Zhuo, 1997, 1999; Han, He, & Woods, 2000; Johannes, Wieringa, Matzke, & Münte, 1996); and that women show weaker lateralization effects than men (e.g., Ingalhalikar et al., 2014). In the very paper he cites—but it would seem may not have read—the authors affirm:

Perhaps the most compelling distinction between local and global visual processing is the differential lateralization in the brain. Behavioral, neuropsychological, and neuroimaging studies with neurally intact and lesioned adults indicate that local information is preferentially processed in the left hemisphere and global information in the right hemisphere [italics mine].

In the light of this they explain their own null results by commenting that their finding “is not particularly surprising in light of the lack of lateralized effects in most ERP studies of local and global visual processing with similar design.” Which does make one wonder why they wasted their time.

Meanwhile the important point is that the large majority of other studies, even of the auditory domain, despite its problems of reliability, confirm the local(LH)/global(RH) distinction in both humans and animals (e.g. Alexandrou, Saarinen, Kujala, & Salmelin, 2018; Alpherts et al., 2002; Ben-Dov & Carmon, 1984; Peretz, 1990; Wetzel, Ohl, & Scheich, 2008). Two recent (2016 & 2017) reviews of the literature confirm the general validity of that distinction, borne out by decades of research, across all domains (Brederoo et al., 2017; Flevaris & Robertson, 2016). And deficits in global feature processing are predicted by disruption of two long association white matter tracts, both of them in the right hemisphere (the superior longitudinal fasciculus and the long segment of the perisylvian network) (Chechlacz et al., 2015). Spezio mentions none of all this.

And on metaphor, it is a similar story. He is right to imply that the large body of research is not unanimous: a large body of research rarely is. It needs therefore more careful attention, being sure not to neglect essential distinctions. One must clarify, first, what one is examining, and second, how one had best go about it.

All language is metaphorical in nature (see e.g., Lakoff & Johnson, 1980; Lakoff & Johnson, 1999; McGilchrist, 2009, passim). So one needs to be clear about what one is separating from what. There is all the difference in the world between a fresh metaphor that still works as a metaphor, and a cliché which no longer does. The left hemisphere is predominant in the processing of familiar and literal language. Success in determining the localization of metaphor comprehension therefore depends on using highly familiar versus unfamiliar, or conventional versus unconventional, metaphors as discriminatory (Schmidt & Seger, 2009). Not all researchers have been careful to do this, and those using only conventional phrases have, as might be expected, found little evidence of right hemisphere involvement in “metaphor” processing (Eviatar & Just, 2006; Giora, Zaidel, Soroker, Batori, & Kasher, 2000; Lee & Dapretto, 2006; Rapp, Leube, Erb, Grodd, & Kircher, 2004; Rinaldi, Marangolo, & Baldassarri, 2004; Stringaris, Medford, Giampietro, Brammer, & David, 2007). However when the metaphor is new or in any degree imaginatively demanding, the kind encountered in poetry rather than cliché, the right hemisphere involvement is clear (Bottini et al., 1994; Eviatar & Just, 2006; Faust & Mashal, 2007; Foldi, Cicone, & Gardner, 1983, pp. 51–86; Kaplan, Brownell, Jacobs, & Gardner, 1990; Mashal & Faust, 2008; Mashal, Faust, & Hendler, 2005; Mashal, Faust, Hendler, & Jung-Beeman, 2007; Schmidt, DeBuse, & Seger, 2007). A recent meta-analysis of imaging studies actually referred to by Spezio makes the point that if tests avoid purely conventional expressions, involve a phrase in a context, not single words out of context, or generally make other than minimal demands on understanding, the right hemisphere is preferentially involved in metaphorical understanding (Yang, 2014).

Methodology is also important. Above all there is little point in aggregating foci of activity across different scanning studies in what is inevitably a linguistic task and not expecting very considerable left hemisphere involvement, as in one of the three studies referred to by Spezio (Rapp, Mutschler, & Erb, 2012). Not all scanning studies are that poor, but this is why lesion studies are so crucial. Let us look at the evidence.

1. Lesion studies clearly demonstrate that individuals with right, but not left, hemisphere damage exhibit inappropriate literal thinking, and failure to understand metaphorical meaning, for which reason, they often find it hard to understand ordinary conversation (Bell & Karnosh, 1949; Foldi et al., 1983, pp. 51–86; Kaplan et al., 1990; Kempler, VanLancker, Marchman, & Bates, 1999; Mackenzie, Begg, Brady, & Lees, 1997; Mackenzie, Begg, Lees, & Brady, 1999; Rinaldi et al., 2004; Winner & Gardner, 1977). This evidence is extensive, consistent and compelling.

2. Transcranial magnetic stimulation (TMS) of the right, but not the left, dorsolateral prefrontal cortex, caused improvement in understanding of metaphors in patients with Parkinson's disease (Tremblay, Monetta, Langlois, & Schneider, 2016). In other studies TMS demonstrates the critical involvement of right, but not left, posterior temporal cortex, in the understanding of novel metaphors (Pobric, Mashal, Faust, & Lavidor, 2008); and confirms the left-hemisphere advantage in understanding stock phrases (Papagno, Oliveri, & Romero, 2002).

3. Visual field studies confirm a right hemisphere advantage for metaphor comprehension (Anaki, Faust, & Kravetz, 1998a, 1998b; Faust & Mashal, 2007; Faust & Weisper, 2000; Mashal, Faust, Hendler, & Jung-Beeman, 2008; Schmidt et al., 2007).

4. Event-related potential studies (ERP), despite their frequent failure to detect lateralization, nonetheless confirm the key role of the right hemisphere (Proverbio, Crotti, Zani, & Adorni, 2009; Pynte, Besson, Robichon, & Poli, 1996; Sotillo et al., 2005).

5. Scanning. The evidence from scanning is more mixed. In accord with the evidence from other sources, many, but not all, scanning studies do also confirm right hemisphere involvement in metaphor comprehension (Mashal, Faust, Hendler, & Jung-Beeman, 2009; Mashal et al., 2007, 2008; Schmidt et al., 2007). In particular they confirm the role of the right temporal lobe (Bambini, Gentili, Ricciardi, Bertinetto, & Pietrini, 2011; Bottini et al., 1994; Diaz, Barrett, & Hogstrom, 2011; Mashal et al., 2005, 2007, 2008; Pobric et al., 2008), the right temporo-parietal junction (Prat, Mason, & Just, 2012), the right parietal region (Cardillo, Watson, Schmidt, Kranjec, & Chatterjee, 2012; and the right insula [Schmidt & Seger, 2009]), as well as the right medial frontal and prefrontal cortex (Ahrens et al., 2007; Bottini et al., 1994; Lee & Dapretto, 2006; Mashal et al., 2007, 2008; Schmidt & Seger, 2009; Stringaris et al., 2006). In other words, metaphor comprehension activates networks widely distributed throughout the right hemisphere. The phenomenon is hemisphere-wide—only the occipital cortex is not fingered.

But this leads to a methodological issue about scanning. Scanning should always be interpreted in the light of evidence from other sources. This is for a number of reasons, which I have detailed elsewhere,¹⁵ but an important one is that it fails to identify broader patterns. This point is well made by one of the best meta-analyses of imaging studies on metaphor to date (Bohrn, Altmann, & Jacobs, 2012). Having noted the frequent reports of right hemisphere involvement in the majority of imaging studies, the authors tried to account for the lack of any one right hemisphere area's standing out in their meta-analysis. They concluded that this could be explained by the activation patterns in the right hemisphere being more widespread, and exhibiting less overlap, than those in the left. Making this adjustment to their interpretation brought their findings “in line,” as they put it, with results from other types of study. In other words, the very fact that the activation is so broadly hemispheric acts to obscure its importance in an analysis of restricted “modules.”

Every study that I am aware of (with one exception which employed only decontextualized word pairs as examples of novel “metaphors”) (Forgacs, Lukacs, & Pleh, 2014) in which there has been a failure to find specific right hemisphere recruitment during metaphor comprehension was both a scanning study and employed only familiar or conventional metaphors, which cannot be expected to recruit the right hemisphere (Lee & Dapretto, 2006; Rapp et al., 2004; Rapp, Leube, Erb, Grodd, & Kircher, 2007). Thus, according to the authors of one such paper, “the metaphors in our study were very simple statements, so-called predicate metaphors (e.g. ‘the alarm clock is a torturer’)” (Rapp et al., 2004). Reference to “torture” is an overused way of expressing displeasure—“it was torture having to drag myself out of bed this morning.” Lee and Dapretto, who found no right hemisphere pre-eminence in metaphor, used single word sequences to exhibit the so-called “figurative” meaning of, for example, the word “cold” meaning unfriendly, usages which have little or no metaphoric status, lack context and are banal (Lee & Dapretto, 2006). The study of Stringaris and colleagues was similarly unlikely to test true metaphorical understanding: the paradigm “metaphor” they report using—“some surgeons are butchers”—is hardly a metaphor at all (Stringaris et al., 2007). Elegantly, however, another study was able to differentiate the two kinds of phrases within the same subjects: e.g., “rainclouds are pregnant ghosts” (right hemisphere) and “babies are angels” (left hemisphere) (Schmidt et al., 2007). It remains true that when the metaphor is of the kind encountered in poetry, rather than cliché, it is clearly the right hemisphere that is involved (Faust & Mashal, 2007; Mashal et al., 2005; Schmidt et al., 2007).

(6)	Meta-analysis confirms this impression: an overview of more than 60 behavioral and scanning studies to date found a coherent pattern: the right hemisphere is better wherever there is non-obvious, non-literal meaning—including, of course, metaphors (Mashal et al., 2008).

By contrast the left hemisphere positively disengages when the meaning is metaphorical or unfamiliar.¹⁶ Here, as elsewhere, the hemispheres have a relationship of reciprocal inhibition, which it is part of the function of the corpus callosum to maintain.

In any case, to answer the question about hemisphere difference it would be enough for there to be a strong tendency, not a 100 to nil distinction here. Trump has his intelligence: Einstein had his character flaws. They remain distinguishable to most people.

Finally I am impressed that Spezio can run a computer program that can tell him whether or not I am onto something. That will simplify academic life no end.

Of all respondents, Cottingham most fully understands my hypothesis, and I am grateful to him for his very thoughtful response. He raises the important question of how much it adds to tie the different “takes” to the brain.

I’d say three things about that. First, if my hypothesis about the brain is true, it is ipso facto worth knowing. Second, no-one to my knowledge has even seen, let alone forged a philosophical position on the basis of, the commonalities between, for example, the appreciation of uniqueness, music and time, a sense of humor, a capacity for reading body language, sustaining attention and the fight-or-flight mode; or between unreasonable optimism, manipulation, disembodiment, literalism, and preoccupations with detail, theory and body parts. They sound, put like that, improbably random assortments of characteristics, not corresponding to any pre-existing philosophical world views. Yet, after reading The Master and his Emissary, and once the bigger picture is painted, the reader will, I hope, see exactly how these elements go to form a consistent picture, one that falls into two coherent parts: a picture that also, I believe, illuminates our situation here in the West today. The neurological data lead us to see connections where a preconceived theory would have overlooked them. And finally it gives us an outside perspective on our minds and helps us see what it is we are missing.

Cottingham concludes that the brain science does indeed add something. His reasons could well act as an epigraph to this whole discussion:

Pointing to the bilateral structure of the brain, and the different modes of awareness and engagement associated with this, reduces the temptation to see our consciousness as arising from an indivisible unitary centre of awareness and decision-making (as Descartes, for example, supposed), to which we have transparent access at each moment; for what we naively take to be unproblematic awareness of the world in fact depends on a finely tuned and constantly interacting coalition of relatively autonomous subsystems. This in turn should generate humility about the danger of mistaking the thin surface of our logical abstractions and analytic dissections of a given phenomenon for a complete and self-sufficient grasp of its nature and significance.

Moral and religious discourse, he cautions, “resonates with us on many different levels of understanding, not all of them fully grasped by the reflective, analytic mind.”

I certainly do not aim, as I know Cottingham is aware, to validate religious experience by giving it a foundation in the brain, which might even be taken by some as explaining it away. My purpose was rather to show that a culture which is, I believe, over-reliant on left hemisphere-mediated patterns of cognition would be likely to be closing off the proper means by which alone awareness of the sacred and divine can be achieved. It might give us a new insight into the crassness of our culture. His emphasis on our embodiedness, and yet, at the same time, on the irreducibility of personal meaning to brain events, is profoundly wise.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 CS Peirce: “these observations escape the untrained eye precisely because they permeate our whole lives, just as a man who never takes off his blue spectacles soon ceases to see the blue tinge.” In Hartshorne and Weiss, 1931, p. 241.

2 For example, machines can be switched on and off; are near-equilibrium systems that are static unless set in motion by us; are not constantly rebuilding themselves millisecond by millisecond; are made of parts that pre-exist the wholes they go to form, parts that are made by us and are not themselves constantly changing; are not engaged in co-constructing themselves with their environment; have precise boundaries; work in a linear fashion; and have an instrumental purpose.

3 Zaidel, Iacoboni, Zaidel, and Bogen (2003, p. 388). See also Iacoboni, Rayman, and Zaidel (1996, pp. 197–202; but also MacKay & MacKay, 1982).

4 Sperry (1973, pp. 211–212): referring to Gazzaniga (1965; Gazzaniga, Bogen, & Sperry, 1967; Gordon & Sperry, 1969; Levy, 1969; Levy, Trevarthen, & Sperry, 1972; Milner & Taylor, 1971; Nebes, 1971; Nebes & Sperry, 1971; Sperry, 1968; Sperry, Gazzaniga, & Bogen, 1969).

5 “Ich glaube, dass McGilchrist auf einer tiefen Ebene recht hat. In den Hemisphären mit ihren unterschiedlichen Komponenten und Fähigkeiten stecken unterschiedliche Persönlichkeitsschwerpunkte” [“I believe that, at a deep level, McGilchrist is right. In the hemispheres, with their differing components and capabilities, lie different facets of the personality”]: Die Zeit, 13 June 2013.

6 Fractal structures (and holograms) are just obvious examples, but since many living forms are fractal, their parts will qualify for at least some attributes of the whole. Such is also true of simple multicellular creatures. As cells become more differentiated, and the structure of an organism more complex, attributes multiply, so the likenesses between parts and wholes, in terms of their attributes, will need to be correspondingly selective. But it will certainly still be there, since nature knows no completely hard and fast boundaries, most obviously not within organisms. All is a matter of degree.

7 It even leads Hacker to invent a person-like term for a brain in a vat—a “cerebroid”—in order to avoid saying the brain is thinking.

8 Eslinger, 1998; Gazzaniga, Ivry, and Mangun, 2009; Greene and Haidt, 2002; Mendez and Shapira, 2009; Koenigs et al., 2007 (in this study the orbitofrontal lesions were, as one might expect, more right-sided, though that is apparent only by inspecting the scan data); Anderson, Barrash, Bechara, and Tranel, 2006; Martins et al., 2012; Tranel, Bechara, and Denburg, 2002; Young, Cushman, Adolphs, Hauser, and Cushman, 2006; Mendez and Shapira, 2009; Demaree, Everhart, Youngstorm, and Harrison, 2005; Martins et al., 2012.

9 Some of these are briefly explored in The Master and his Emissary, 34–6 & 467 n20–21.

10 E.g.,: “[n]euroimaging studies may especially fail to shed light on hemispheric lateralization: contrasts designed to find the neural correlates of one process may be contaminated by the neural fingerprints of another, which may in turn make a truly lateralized process appear bilateral. We also necessarily make the neuroimager's fallacy in almost every case, as the vast majority of studies—even those that report lateralization results—only report clusters of activation, not direct contrasts across hemispheres.” Marinsek, Turner, Gazzaniga, and Miller (2014, p. 839). The term “imager's fallacy” appears to originate in Henson (2005, p. 216, 217).

11 Stanislas Dehaene, leading neuroimaging researcher and Professor of Experimental Cognitive Psychology at the Collège de France, quoted in Holt (2008, pp. 44–45).

12 This is most dramatically seen in right hemisphere-lesioned patients, who deny or minimize their handicaps, and make fatuously optimistic predictions. This is in stark contrast with left hemisphere-lesioned patients, who are more realistic, but tend towards pessimism. For discussion and many references, see The Master and his Emissary, 63ff, 84ff & 479–480 (notes 453–468). See also, Silberman and Weingartner (1986; Babinski, 1922; Denny-Brown, Meyer, & Horenstein, 1952; Hécaen, de Ajuriaguerra, & Massonet, 1951; Killgore, 2002). And depressed subjects appear to have increased reliance on the right hemisphere (Ahern et al., 1994; Bear & Fedio, 1977; Brosch, Sander, & Scherer, 2007; Emerson, Harrison, Everhart, & Williamson, 2001; Lazure & Persinger, 1992, p. 1058; Persinger & Makarec, 1991).

13 See The Master and his Emissary, 25–7, for references.

14 Watanabe et al., 2014: I am grateful to Professor Martin Rossor for drawing this to my attention.

15 The Master and his Emissary, 35–6 and 467 n19–28.

16 Cardillo et al., 2012; Lai, van Dam, Conant, Binder, & Desai, 2015, p. 44. It can be demonstrated, incidentally, that it is not the well-attested advantage of the right hemisphere for difficulty or novelty per se that leads to the right hemisphere advantage for metaphor, but metaphor itself (see McGilchrist, 2018).