This is about I Am a Strange Loop by Douglas Hofstadter; my discussion is picky and pedantic and probably far too long for any but his devotees; but then, their number is many.

Context · I, like a whole lot of other people, bought Hofstadter’s Gödel, Escher, Bach in 1980 because Martin Gardner said to, and was dazzled by the virtuositic writing, erudition, cool typography, and playfulness.

Unlike most others, I was disappointed at the end of GEB. Granted, I enjoyed and, dare I say, was enlightened by the individual discourses on the genetic code and the Gödel proof and the structure of a Bach fugue and the flavors of recursion and the charming contrast between reductionist and holist explanation and all that other stuff. But at the end of the day and the end of the book, I just didn’t buy the Big Analogy. I don’t think the Gödel proof is much like the genetic code in any interesting way, nor are either much like Bach’s counterpoint.

So what? You don’t have to buy the conclusion to enjoy the experience of getting there, and I would say that if you haven’t read GEB carefully, you’re not sufficiently acquainted with some of the key notions that inform twenty-first century intellectual discourse.

Having said all that, I’m pretty sure that Hofstadter’s finest work, and the one most likely to have intellectual currency a few centuries from now, is Le Ton beau de Marot, which is at once stimulating, entertaining, erudite, beautiful, and tragic.

My Viewing Angle · It’s not unlike Hofstadter’s. I have a math degree and, partly because I play the cello, an emotional connection with Bach. At one point I buckled down with the Nagel book long enough to have convinced myself that I kind of understood Gödel’s argument.

On the other hand, unlike Hofstadter, I eat meat, enjoy watching football on TV, and like rock music in general and Elvis Presley in particular. Go figure.

Strange Loop · In his preface, Hofstadter complains that many who enjoyed (or despised) GEB failed even to notice what he considers its central point, and tells us that this book is another attempt to drive that point home. I worry that he’s talking about precisely that part of GEB where my suspension of disbelief ran out; oh dear.

In this book, I have big problems with his heroic analogizing and think he scoots past too much prior art, but he’s got a core argument about the size of the soul that really doesn’t need the apparatus he stacks against its sides in an effort to support it.

What’s Good · Before I get into nitpicking, I should say that there is some wonderful writing here, as well as two Big Ideas that may not be 100% new but are beautifully argued and deserve the evangelizing. First the writing, and if I may quote by way of example from the conclusion:

Poised midway between the unvisualizable cosmic vastness of curved spacetime and the dubious, shadowy flickerings of charged quanta, we human beings, more like rainbows and mirages than like raindrops or boulders, are unpredictable self-writing poems — vague, metaphorical, ambiguous, and sometimes exceedingly beautiful.

The first big idea is that one person’s consciousness may live, in part, in another person’s brain. The second is that there is not a binary consciousness/no-consciousness switch; some consciousnesses are bigger than others. Hofstadter uses the word “soul” in this context for consciousness. Following his usage: In souls, size matters. And among humans, those with the largest souls are arguably those who are especially gifted at hosting the the souls of others; this produces, among other things, the quality that we label “empathy”, but the argument is bigger and more beautiful than that.

But let me get back to griping, first over Hofstadter’s overambitious analogizing.

The Big Analogy · It seems to me that the most obvious deceased equine tasting the whip here is reductionist/holist tension. He argues over and over that you have to think about high-level problems (a basketball shot, the nature of consciousness, the design of a gas valve) at a high level, without worrying about bosons or neurons or the kinetic theory of gases.

I don’t see why this is controversial or, really, worth his effort. Having said that, if you know anyone who’s a hard-reductionist (I suppose one must exist), make ’em read this and if they don’t chill out, well then, you know they’re insane.

The Gödel Problem · I think Hofstadter has an unhealthy relationship with the Gödel proof; in particular, in obsessing over its mechanics. He really enjoys explicating Gödel’s fiendish cleverness in finding a mapping between the Principia Mathematica apparatus and integer mechanics, and then launching into the analogical wild blue yonder.

As someone who cares about mathematics, although without sufficient talent to make it a profession, it seems to me that Gödel’s conclusion is what matters: you can’t assume that something you can’t prove is false, and you can’t assume that something that’s true can be proved. Let me oversimplify: Truth cannot be mechanized; deal with it.

The two views of Principia Mathematica—Russell’s original and Gödel’s arithmetic mutation—are among the many phenomena to which Hofstadter attaches the “strange loop” label; to me, using this gap as an analogy, for example, to that between thinking about neurons and about cognitive symbols is at best forced. When, as in Chapter 16, Hofstadter takes his point as proved and starts routinely referring to his own consciousness (and implicitly mine) as a “Gödelian swirl”, I find myself actually offended.

Lacunae · While I Am a Strange Loop includes a substantial scholarly apparatus (end-notes, bibliography, index), it skates by some pieces of intellectual history that seem to me both relevant and important. So I’m going to indulge in some pedantry and fill a couple in.

Marx · Hofstadter quotes, and argues furiously in defense of, the argument by Roger Sperry that ideas and statements and mental symbols and in particular the notion of “I” have an objective reality in the world, and this without buying into dualism or religion.

Which is fine; I agree, and while the truth of this notion seems so obvious to me that I wouldn’t have invested the pages in defending it, I particularly enjoyed Hofstadter’s skewering of the sloppy and problematic dualist notion of the “soul”. Having said that, it seems at best rude not to have tipped any hats in the direction of K. Marx, whose influential if now widely-disbelieved writing on economics were heavily interleaved with weighty Teutonic excursions on metaphysics, much of which was concerned with showing that you could be a Materialist and still believe that ideas were Material in the sense of being real and having force in the world.

Theory of Mind · Perhaps the central interesting idea in this book that hasn’t been belabored previously by Hofstadter is the notion that in some meaningful way, the consciousness of one person can exist at least partially in another person’s brain. I won’t repeat his argument, which is eloquent and deeply personal, but it’s a fine piece of work.

But in modern neuropsychology, discussion of Theory of mind is routine; in fact, I recently read something (can’t put my finger on it) that presented evidence that the Theory-of-mind function can be localized to specific regions of the brain.

What is theory of mind? It’s the indispensible ability, while interacting with other people, to imagine what it is they are thinking and feeling. Real communication is (self-evidently, it seems to me) impossible without such a capability, and its lack is implicated in certain mental disorders. Once again, Hofstadter’s lengthy discussion of multiple “strange loops” in one mind, without at least touching on this work, is unfortunate.

The Doors of Perception · In chapter 15, Hofstadter writes “Put less metaphorically, my sense organs feed my brain directly” and goes on to show that in fact your brain is sometimes fed by the sense organs of others. It seems highly relevant to me (and to one of Hofstadter’s other core arguments) that, as Aldous Huxley argued powerfully in The Doors of Perception, your sense organs don’t in fact feed your brain directly.

As Hofstadter argues, your consciousness is in some large part a construct of the your lifetime’s aggregate inputs, obviously involving your sense organs. As he fails to acknowledge, and as Huxley showed, your perceptions are not of the world directly but are filtered through your “I”-ness, through your own structure of consciousness, and thus subject to heavy interpretation and censorship on the way in. Hofstadter is really missing a trick here, because the notion of the consciousness built from the accreted input stream then becoming a primary filter for that stream is probably what he would call a “strange loop”.

On the other hand, Huxley based his argument on his mescaline-eating experiments, which may prejudice some against it.

Read the Book? · I’d probably recommend it. But if you haven’t read Le Ton beau de Marot, go read that one first.


Comment feed for ongoing:Comments feed

From: ingo (Apr 06 2007, at 04:16)

One thing I find when looking at the GEB is that I don't get what all the fuss is about and my hunch is that is, because I'm -- as a late twen now -- arriving at the tail end of the all the discussions it chronicles. Lots of things it laboriously goes through I have already accepted at face value and don't particularly enjoy having laid down in such detail and in such a though-experiment-heavy way. A similar thing occurs to me with other books from the era such as Varelas "Tree of Consciousness", Daniel Dennets work and the like.

In a way, this is quite sad. Its not as if all of these questions have been settled satisfactorily. However, a lot of the details have changed, the association with early AI isn't exactly helping and I really wonder -- whats the modern day equivalent of these books? Is there one?


From: Sidharth Kuruvila (Apr 06 2007, at 04:45)

Aargh! 6 reads and it still doesn't makes sense to me.

I thought the whole strange loops bit was about our ability to comprehend self reference in a simple way, nothing more. I only managed to get through the first part of the book, so I may have missed a lot.

Anyway I'm doing a reread of the book, hopefully, your post will make more sense when I'm done.


From: Luke Kanies (Apr 06 2007, at 07:36)

I read GEB in the late nineties, as I would have needed to have been some kind of prodigy to read it in 1980, but the book has a strong place in my life because I had to work so hard to read it that it pulled me out of an intellectual funk and allowed me to get back on track.

You state that the main summary of Gödel's proof is "Truth cannot be mechanized; deal with it", but that is not how I (or Gödel, I expect) would summarize it. Instead, I would say: No representation of reality can be complete.

The whole goal of Russell's Principia was to prove that the mathematical constructs known as the "real numbers" perfectly mapped to reality, and Gödel shattered that goal by proving that not only was this not the case, but that any sufficiently complex system was bound to be subject to Gödel's method and thus be found incomplete.

While in some ways I agree that Hofstadter's focus on mechanization can get in the way, a significant amount of his point is that this mechanized process can consistently be used to find these sufficiently complex systems and break the mapping between our conception of these systems and their reality.

I, too, was slightly disappointed in his Seed article, although the issue as a whole was spectacular, I thought.


From: MBA (obfuscated for various reasons) (Apr 06 2007, at 13:09)

A speedy way to experience theory of mind and marvel at the ways you didn't realize your mind could interact with others: Ketamine in a small social group or with one other person. Enlightening. It is unlikely to happen the first few times. Make sure you're safe and in trusted company.

A speedy way to experience the loop: Acid, alone or with a partner less affected but preferably experienced, someone who can feed ever so slightly an exploration of any random but preferably positive subject matter. Safe company, and depression-free before you start is the only way to go.

Too bad more intellectuals and thinkers don't experience these things. (Or do they?)


From: John L. Clark (Apr 07 2007, at 05:47)

With respect to the theory of mind, you mention that it recently pinged your consciousness from another source. I ran across it recently in a New York Times article published on 2007-03-04 entitled "Darwin's God", which, sadly, has passed into the elitist "pay to read" portion of the web, but which referenced a smörgåsbord of potentially interesting resources on the topic. These include "The God Delusion" by Richard Dawkins, "The End of Faith" by Sam Harris, "Breaking the Spell" by Daniel Dennett, "The Varieties of Religious Experience" by William James, "The Descent of Man" by Charles Darwin, "In Gods We Trust: The Evolutionary Landscape of Religion" by Scott Atran, "Why Would Anyone Believe in God?" by Justin Barrett, "Descartes' Baby" by Paul Bloom, "Pigeon Feathers" by John Updike, "Religion Explained" by Pascal Boyer, "Tragic Sense of Life" by Miguel de Unamuno, and "Darwin's Cathedral: Evolution, Religion, and the Nature of Society" by David Sloan Wilson. Might it have been any of those? If so, what were your thoughts after having read it?

(Granted, one very large reason for submitting this comment is to keep - and share - a byteful trail of these resources.)


From: Dan Davies Brackett (Apr 07 2007, at 14:47)

I read GEB a few years ago, and loved it. I picked up Ton Beau shortly thereafter, and couldn't get past the first third. Two major things stood out: (1) Hofstadter doesn't see translation/transculturation and quoting as fundamentally different, even though I *really* think they are; and (2) Ton Beau has an extremely irritatingly self-congratulatory tone. Part of the charm of GEB was that although it was dazzlingly clever, it wasn't particularly aware of that fact; it went about its ideas, meandered and played, even referencing its previous content -- but never referencing *itself*. Whereas Ton Beau was full of "oh, aren't I clever! look at that joke I just made wasn't it the funniest one you ever heard!! I AM SMART LOOK AT ME BE SMART SMARTNESS RIGHT HERE YOU'RE SEEING IT" and that's when I put it back on the shelf. Again.


From: John Faughnan (Apr 07 2007, at 20:16)

The neuronal basis for modeling another's consciousness has been most recently alleged to be "mirror neurons" (Sci Am did an article a few months back.) I think that's what you were trying to recall.

Wild speculation has it that autism is a mirror neuron disorder, hence a primary disorder of modeling consciousness. My own wild speculation is that the injured brain scavenges this network to supplement injured subsystems -- so mirror neuron deficits may be secondary to another brain injury.


From: Darren Chamberlain (Apr 08 2007, at 08:27)

Re: Darwin's God: I happened to save a formatted copy of this story on a private Writeboard; email me for access to it (dlc at I saved it because I knew I wouldn't get a chance to read it before it expired (and I was right; I've still not read it).


From: Martin Probst (Apr 09 2007, at 07:31)

<blockquote>On the other hand, Huxley based his argument on his mescaline-eating experiments, which may prejudice some against it.</blockquote>

If it makes you feel any better, I think you can find all perception-reasoning you'll ever need in Kant and Schopenhauer.


author · Dad · software · colophon · rights
picture of the day
April 04, 2007
· Arts (11 fragments)
· · Books (120 more)
· The World (126 fragments)
· · People (91 more)

By .

The opinions expressed here
are my own, and no other party
necessarily agrees with them.

A full disclosure of my
professional interests is
on the author page.