Friday 28 April 2023

REVIEW ARTICLE The Experience Machine

 From The Philosopher CXI No. 1 Spring 2023

In this optical illusion, the two orange circles are exactly the same size; however, the one on the right appears larger


The Experience Machine: How Our Minds Predict and Shape Reality

Andy Clark is a professor of ‘cognitive science’ at Sussex University and talks briefly about how he started there when the department was entirely novel. Coincidentally, I also remember this excitement, as I studied at the same university around this time, and was offered a choice of modules. The choice was social science, Marxism, or this new thing, ‘cognitive science’. I took the first option and my career has never recovered. Cognitive science, on the other hand, has become highly fashionable. But what is it exactly? I was suspicious then that the subject was really an uncomfortable blend of computing and biology - the study of the mechanisms of the brain.

The thing is, I don’t think the human mind is a computer – far less that you can work out how it operates by studying electrical signals in the brain circuits. 

Clark says that his approach “challenges a once traditional picture of perception, the idea that light, sound, touch and chemical odors activate receptors in eyes, ears, nose and skin, progressively being refined into a richer picture of the wider world”. The new idea, the “new science” as he puts it, “flips that traditional story on its head”. Perceptions are “heavily shaped from the other direction, as predictions formed deep in the brain… alter responses”.

Yet, having offered this radical reversal, Clark brings back the ‘outside-in’ approach by allowing that sense perception “helps correct errors in prediction”. What does ‘error’ mean here? That there is a real world out there that sense perceive accurately? It seems to be an uncomfortable attempt to ride two horses at once.

“Predictions and prediction errors are increasingly recognised as the core currency of the human brain, and is is in their shifting balances that all human experience takes shape” adds Clark undeterred. The brain is however in the driving seat, “painting the picture”, with sensory perception “mostly to nudge the brushstrokes”. Switching to computer language, he explains:

“Instead of constantly expending large amounts of energy on processing incoming sensory signals, the bulk of what the brain does is learn and maintain a kind of model of body and world– a model that can then be used , moment by moment, to try to predict the sensory signal”

As Clark mentions, we come across this usually hidden effect when we look at optical illusions, like the one where two figures are the same height yet one is made to seem much more larger by virtue of tricks with the background. Clark goes so far as to say that what we really see are “controlled hallucinations”.

Talking of which, the placebo effect is discussed and Clark notes how studies have found not only that people suffering from back pain benefit both from pills which contain active ingredients and those that don’t, but that this effect survives even when the patients are told the pill has no active ingredients. (Not that Clark goes there, but this certainly points to a possible justification for the infinitesimal treatments of the homeopaths.)

What he does say, however, is that anything which boosts confidence in an intervention will enhance its prospects of success – but here I think he misses the difference between conscious cues and subconscious ones. Odd lacuna? But then he actually argues that “predictive brains” involve the “active construction” of human experience.

A lot of the claims here are offered flat, as “science says”, yet surely deserve some scepticism. Research for example, that, when shown religious images, “religious  subjects rated a sharp pain as less intense than atheists shown the same images”. Or that isting the risks of side effects on medical treatments can “actually bring about the side effects they describe”. Let alone that dentists telling patients that the injection will only be a tiny poin prick reduces the pain experiences. On the contrary! Those words of reassurance signify to many of us that a very nasty pain is about to follow! Okay, maybe my point looks a bit like a joke, , but actually, one big concern I have with this account is how it removes the complexity of human thought processes. But that’s cognitive science!

Common-sense notions of causality are also reversed in the phenomenon noted by the German philosopher, Hermann Lotze and, later, William James too, that actions come about because we mentally represent the completed effects of the action. Clark gives the example of pulling the strings on a puppet. We are interested in (say) the puppet waving its hand – not in the details of how the string moves which bit of the puppet in what may be a complicated sequence. 

Likewise, it seems that when we have a drink of water to assuage thirst, we get immediate satisfaction of the thirst, even though the water has not had enough time to have had any physical effect.

More ominously, things like a police officer’s elevated heart rate when investigating a possible threat, can be taken by the “predictive brain” as themselves evidence that there really is a threat. (We’ve seen too many cases of such things in American in recent years, with police shooting householders or motorists out of misplaced conviction of a threat.)

Having put the brain in control of our environment, Clark then backtracks and offers examples of how out environment can be adapted to help our brain. Alzheimers sufferers, for example, he says, may arrange their homes with lots of visual clues, from written nots to arranged objects, to “take over the functions that were previously played by their biological brains”. A biological part of the brain is replaced by external, physical substitutes. Clark suggests we are all increasingly doing this – relying on calculators to do our maths, on search engines to remember things.“Most of our daily devices especially our smartphones and other wearables, are already starting to act as woven resources. They are devices whose constant functionality has become factored deep into our brains’ ongoing assumptions of what kinds of operations can be performed and when.” Actually, this is an idea Clark set out earlier in another book, co-authored with David Chalmers, called The Extended Mind. Clark mocks the “chauvinism” of those who say such devices cannot be considered part of our ‘minds’, as they are outside our heads. Yet he does not seem to have considered that all our thinking might be better understood as social, particularly given that so much of it is framed in words and concepts that are produced socially and made concrete in human languages.

Towards the end of the book, which is a reasonable place to do it, Clark sums up his theory: “To perceive is to find the predictions that best fit the sensory evidence”. This rather underlines how little philosophy there is in the book. “The sensory evidence” seems to be still there, just as John Locke and the other philosophers supposed centuries ago, steadily being processed by humans. The only new thing is that at a certain level of the conscious mind, the perceptions are being reorganised, largely in line with expectations based on previous experience. Clark declares this is big progress, writing that “understanding the way predictions of many kinds alter and adjust the shape of experience is our biggest clue yet to what experience really is”. But if that’s the takeaway from the book, it’s rather meagre. Plato wrote about perceptual illusions and what they told us about perception, two thousand odd years ago. Cognitive science, it seems, is a new name for a very old study.

Reviewed by Martin Cohen

Book art The Experience Machine

The Experience Machine: How Our Minds Predict and Shape Reality

By Andy Clark

Pantheon  (Penguin–Random House) 2023 ISBN 9781524748456