What Am I? A Strange Sense of Self

Review of I Am a Strange Loop by Douglas R. Hofstadter, New York, Basic Books, 384 pp, hardback £14.99, ISBN 978-0465030781

Reviewed in Nature 3 May 2007, 447, p 29-30

(This submitted version is reproduced here with permission. It may have been edited before publication.)


Who is this “I”; is it the author; is it me the reader; could it be the sentence itself (would that work?)? Readers of Hofstadter’s 1979 megabestseller Gödel, Escher, Bach will be familiar with such twisty questions, and probably delight in the pain, mental squirming, and occasional wondrous resolution such questions provoke. But whereas GEB conjured myriad mysteries, I am a strange loop tends rather to explain them, and the result is less magical and even slightly awkward.

At one level (and this is, of course, a deeply multi-layered work) the book pulls off some remarkable achievements. For example, in a matter of 40 readable, and even enjoyable, pages, Hofstadter manages to explain Gödel’s theorem in a way I have never seen attempted before, and much appreciated. We are taken from “the gloomy, austere, supposedly paradox-proof castle” of Whitehead and Russell’s great work Principia Mathematica (containing the formal system Hofstadter calls “PM”), through ways of mapping numbers onto theorems and hunting for patterns among squares, primes and “prims” (Hofstadter’s name for theorems that are provable in PM), to sentences that talk about themselves and so, finally, to make sense of a concise English translation of Gödel’s formula (the paradigmatic strange loop) “I am not provable in PM.” Wow! I really felt I learnt a lot.

Hofstadter speculates that Russell never saw the second level of meaning in his great work, like a dog that sees a TV screen as a mass of changing pixels, or a child who sees the people on the screen but fails to grasp the romantic plot. And then we are off again to tangle with ever more layers of paradox, and wonderfully mind-wrenching questions. What is the nature of mathematical truth? What is the nature of meaning? Could a machine be confused? Could it know it was confused? Could it believe that its unquestioned belief in the reality of its own “I” was a necessary illusion?

Along the way Hofstadter talks about himself, but his pacy mix of stories, metaphors, questions and explanations is sometimes spoiled by what seems like a lack of confidence. Instead of just blasting us with his rush of original ideas, Hofstadter apologises for a “corny pun”, a “hopefully amusing example” or just for telling personal stories at all. Yet these stories are delightful. My favourite is his first encounter with something that, he says, “runs in our human grain”; the irrational fear of loops. When little Douggie went with his parents to buy a video camera, one of the cameras in the store was plugged into a TV screen. So he pointed the camera at his father, then at himself and then … about to point it at the screen itself, he stopped. He remembers with shame that he was hesitant to close the loop! So he timidly asked the salesperson whether he might and was told “No, no – you’ll break the camera!”.

He then convinced himself that this couldn’t possibly be true, and went home to experiment with video feedback, finite and infinite regresses, corridors with curves or corners, and a completely unexpected pattern with precisely thirteen (prime, of course) spokes.

A sadder event, the discovery of his baby sister’s brain damage, began his fascination with the physical basis of consciousness. Aged twelve or so, it dawned on him that consciousness is a peculiar kind of mirage that perceives itself, and yet doesn’t believe it’s perceiving a mirage. This insight leads directly to the stated aim of this book: to try to pinpoint that “special kind of subtle pattern” that underlies, or gives rise to the “soul”, the “I”, “having a light on inside” or “being conscious”.

This is a grand aim, and he joins countless contemporary writers in struggling to explain consciousness. He derides zombies and qualia, has some harsh words for Dave Chalmers and John Searle, and skilfully sweeps away all sorts of nonsense, from old-fashioned kinds of dualism, to the more prevalent belief that consciousness is still something “extra” – an élan mental? Instead he argues that a self is a strange loop that automatically arises in a machine with a sufficiently sophisticated repertoire of categories. It’s a myth, a mirage, something like a satellite to your brain whose resident strange loop calls wherever that brain is “here” – and, he claims, once you have explained the self you’ve explained consciousness.

Herein lies the source of my dissatisfaction. The idea of self as a strange loop makes sense of moments of self-awareness or baffled self-inquiry – but what about the rest of the time? The theory seems to imply that mostly we are not conscious at all, which may well be right, but Hofstadter does not discuss this. Then there are those profound moments of utter stillness or absorbed flow when the self is in abeyance. People describe these as being clearer than ordinary consciousness, but this cannot be explained if self and consciousness are as closely linked as Hofstadter claims. He also argues that the self loop is indispensable, and this might be challenged by those who have attempted, or even managed, to let go of the illusion of self. He quotes a Zen koan which seems to me beautifully to point the way out of strange loops and into awareness beyond self, but he dismisses it as “Just a bunch of non sequiturs”.

He realises that people will be dissatisfied, and provides a light-hearted debate between two numbered strange loops; himself and his opposing sceptic. I keep wondering whether I’m no better than sceptic SL #264 who really just doesn’t get it, though he (it?) comes up with some classic objections. I keep looking over at my black cat on the window sill, with her ragged fur brightly lit by the morning sun. Does it help to say that the experiencer of this vivid visual experience is a strange loop? Whatever the answer, this strange loop is enjoying the question.