Leonard MlodinowThe subtitle of Leonard Mlodinow’s book Subliminal makes bold claims: The Revolution of the New Unconscious and what it Teaches us about Ourselves.

Mlodinow is a talented fellow: I first became aware of him as Stephen Hawking’s co-author on The Grand Design (I blamed him then for the terrible jokes in that book, but the evidence of Subliminal, which is amiable but wince-free throughout, I think Hawking was probably to blame for them after all). Being Hawking’s colleague is probably the nearest the modern world can offer to being God’s assistant, but in addition Mlodinow has done impressive original work in physics and written successful screenplays.

The book is a wide-ranging compilation of a lot of interesting stuff. In the early stages of the book, it seems Mlodinow is basing his claims on contemporary technology and fMRI in particular: he tells us it is transforming our knowledge. But in fact not much of the research he reports is dependent on scanning. It feels as if the book might have changed direction in the writing, as Mlodinow found that most of the stuff he wanted to include actually didn’t involve advanced technology after all, but retained in the text the laudatory stuff about fMRI which it no longer really justifies.

How come the scanners don’t feature more strongly? One possible reason is sort of indicated when Mlodinow talks about how experimental subjects were shown to rate wine more highly when told it was expensive. Mlodinow wants to say that the tasters did not merely give the ‘expensive’ wine better ratings, but actually enjoyed it more: so he tells us that fMRI scans showed activity in the orbitofrontal cortex, ‘a region that has been associated with the experience of pleasure.’  Has been associated (not necessarily by me) associated with (not necessarily controlling or unambiguously diagnostic of). If the trumpet sounds as uncertainly as that, we must ask whether whether we’re really being told anything of value. Of course we know why Mlodinow is so hesitant. First, nobody has a really clear idea of what the orbitofrontal cortex does; it seems to be involved in addiction and motivation – but for Mlodinow’s purposes we need to be talking about the qualia-laden appreciation of fine bouquets and the like, which may well be an unrelated matter.  Second, fMRI is a fuzzy and ambivalent tool  and the wider implications of the data it produces are always debatable. Third, this business of real pleasure is a philosophical swamp: put aside all the Hard Problem issues: were the subjects experiencing real pleasure, or did they just think they were experiencing pleasure, or were they just thinking about pleasure? Was the pleasure straightforwardly gustatory, or did it come from thinking what smart guys they were and wishing their friends could see them now? These are not mere quibbles; the latter case for example, would be much less interesting than the radical and somewhat implausible claim that beliefs about price really change the experience.

That points to a general difficulty: books of this kind often give us stuff that is interesting, new, and well-founded; but the stuff that is well-founded isn’t new and the stuff that is interesting is debatable and looks over-interpreted. I wouldn’t say Mlodinow escapes this pitfall entirely. He tips his hat generously to Freud, which is nice, but that’s surely the Old Unconscious. The wine experiment – how many eyebrows would that have raised around the table at Plato’s symposium? Perhaps not many. Mlodinow tells us yet again the story of how Nixon lost out to Kennedy in 1960; people who could see him on TV were less inclined to think he had won the debate than those who merely heard him on the radio. Well, we’ve known that people are influenced by candidates’ appearance at least since Pericles took to appearing in a helmet which both reminded the electorate of his generalship and concealed the weird shape of his head. Do we even know that people were unconscious of being influenced by Nixon’s appearance? It seems quite possible that some of them drew the entirely conscious conclusion that he looked too rough and too shifty to be credible (a verdict which some would argue was borne out by later history, incidentally). On the other hand, Mlodinow reports research showing that people named ‘Brown’ are significantly more likely to marry other people called ‘Brown’ than statistical chance would warrant. Is that true?  Or is there some quirk here  – perhaps there are ‘Brownsvilles’ where by chance or history a concentration of Brown families mean you’re more likely to meet people of that name than random population matching would suggest? I don’t know, but I’m left in doubt, and as a human being myself I need something pretty strong to convince me to give up my strong intuitive understanding that surnames are not generally relevant to my species’ mating decisions.

The point that electors may have assessed Nixon’s appearance emotionally but consciously leads us to another difficulty: quite a bit of the research Mlodinow recounts doesn’t really bear on his thesis about the unconscious. He recounts the experiment, by now fairly well-known, in which an experimenter asked a stranger for directions: accomplices interrupted the conversation by carrying a door between the two, behind which the experimenter was switched for someone else: subjects often resumed the conversation without noticing the change in their conversational partner (the book here sort of undercuts the experiment by including pictures in which it is clear that the two experimenters were not that dissimilar in looks, and, if I may be rude,  also of a rather unstriking generic appearance, too).

The experiment is interesting, but how does it show that the unconscious is more important than we thought? Is there any suggestion that the difference was recognised unconsciously while being ignored consciously? Well, no: in fact we might think that this is the sort of thing the conscious wouldn’t deal with, leaving itself to be warned by unconscious processes, so if anything the hit is against the effectiveness and influence of the unconscious. Simply showing errors in conscious beliefs does not establish a revolution in favour of a new unconscious.

But then Mlodinow never formulates what he means by either the old or the new unconscious. We don’t even know whether he thinks the unconscious really amounts to one thing, several different unconsciousnesses, or simply a lot of default non-conscious mechanisms. The word ‘consciousness’ notoriously covers a number of different entities or processes, but we never get told explicitly which of them Mlodinow believes in or which of them he wants to dethrone. If you want to carry out a revolution against one form of mental activity and in favour of another, you really need to offer a pretty clear of view about what those different forms actually are and what roles they play, don’t you? Mlodinow would never try to get away with such vagueness if he were trying to sell us a revolution in physics, so the fact that he seems to think it will do for consciousness suggests an unattractive casualness, to say nothing worse. Perhaps in a way it’s evidence in his favour that Mlodinow never seems to have noticed consciously that so much of his material doesn’t really bear on his thesis; perhaps his unconscious is subtly offering us a different verdict.

That may be just a little hard: there’s a lot of very readable stuff about genuinely interesting research here, but the Revolution of the New Unconscious seems to me to have gone missing.

One Comment

  1. 1. Vicente says:

    Very interesting, pure neuro-psychology. I just got the book.

    I was thinking that may the Integrated Information Theory (phi-function), could be a good tool to understand better why some neural processes are conscious…

    If we could plot phi-function for the different processes running in the brain (the pandemonium), maybe we could see, that some of the surpass a certain value, above which they become conscious, i.e. the threshold. So it is matter of how the information is structured, rather than the nature or amount of information itself.

    Some processes could be inherently integrated, like those supporting perception, sensing qualia, still the would need some time raise the threshold value… others, rational processes, would need more stringent processing or synchrony…

    Maybe that’s way meditators that achieve higher and more stable periods of EE waves synchrony, expand their “conscious volume”, as a result of facilitating the information integration (plus previous filtering probably).

Leave a Reply