Rating: 
Summary: Escape from predestination
Review: It seems highly appropriate that Douglas Hofstatder should re-release his epic work now. His central theme plays so eloquently in this place and time: Every system folds in on itself, be it physics, mathematics, or any form of language. All these systems are inherently self-referential, and as such, take on a life of their own. A life their creators could never imagine. Many reviewers have focused on the explicit messages of the book, their likes or dislikes, but the great beauty of this work lies within the realm of what it does not say. It is, no doubt, the most difficult book I have ever read, and I have to admit it took me several false starts to finally get through the thing. It is so incredibly deep - one cannot simply wade through it like a sci-fi novel. But if you take your time, spend, say about a year on it - work through the TNT exercises, discover the hidden messages the author has left, read the bibliography - and at some point it will strike you; the incredible richness of the message. The book, you, the world, all of it IS open. The pages of this universe are blank, unwritten. Dr. Hofstadter has woven a message of eternal optimism, one that transcends even the infinite depth to the tapestry of topics spread before us: The great freedom that we, nature's most remarkable matrix, are part of a future without destiny. Even if we were created, any purpose impressed upon us is lost in a cacophany of unexpected relationships. Deterministic, yet infinitely complex and unpredictable. We can never understand anything completely, and thus every life can experience the magic of observing that which cannot be explained. This is a book of wonders, and you will never regret the time you spent on it.
Rating:
Summary: A Book that belongs in an Atheist's Bible...
Review: ... but I am not sure if it would lie in the Pentateuch, the Late Prophets or the Gospels. Somewhere important, anyway. It attempts to answer the Q: If there is no God then how does this thing that I think I am do these things that I think I am doing? This book was an early answer, dwells on important basics (art, language, mathematics, logic) not ephemeral specifics, and blew my mind when I read it 20 years ago. It has not (yet) gone stale. Try and explain it to your children while the drones flock to churches and sports arenas on Sundays. If you can, if you dare.
Rating: 
Summary: Occasionally frustrating, but interesting as a whole
Review: Gödel's incompleteness thoerem is one of the most profound results of abstract mathematics, but unfortunately it is difficult to grasp its full meaning. GEB explains Gödel's thoerem in an interesting way by tying the mathematical result with Escher's paintings and Bach's music. The book can frustrating in some parts, but on the whole it's very interesting and useful to understand the incompleteness theorem. One interesting thing about GEB is the dialogues. Hofstadter puts fun dialogues between the chapters, usually to introduce the concepts in the next chapter from a different angle. Sometimes this makes the concepts easier to understand, and it can occasionally be confusing to read the dialogue and then wonder what parts of it were real and which were fictional, but mostly they're fun to read and a good break from the mathematics and philosophy. Gödel, Escher, Bach is worth reading if you're interested in philosophy, logic, artificial intelligence, or mathematics.
Rating: 
Summary: Innuendo is not a defence
Review: 'Anyone who disagrees with the contents of this book must be distraught because Hofstatder stepped on someone's philosophical toes.' This sort of sophomoric argumentum ad hominem (and perhaps ad baculum as well) is innuendo, not a defence of the ideas contained in this volume. Such nonsensical remarks about the *motives* of those who disagree with Hofstadter are more worthy of cultism than of genuine philosophical enquiry. Hofstadter's contentions about the origins of mind are highly debatable to say the least. Moreover, the 'picture theory of meaning' has indeed been 'discredited' in the only way philosophical ideas are ever discredited: by reasoned argumentation. It was revived earlier this century by the tyro Wittgenstein, who seems never to have read a work of actual philosophy in his life. Had he done so, he would not have offered his remarkable naive theory. The theory becomes no less naive when it is couched in the language of 'isomorphism'. Even if Hofstadter were using this term in its mathematically precise sense (which he most assuredly is not), it would still be false that the existence of a mathematical mapping from one set to another is sufficient to create *meaning* if mind is not already present. But this is the view to which Hofstadter is committed if he intends his work to support the contention a reviewer has quoted below, regarding the origins of consciousness and 'selfhood'. That reviewer was too kind. Hofstadter's reductionist 'explanations' of mind belong next to those of his comrade Daniel Dennett - in the trash bin.
Rating:
Summary: Better taste than a stale philosopher
Review: A number of people may bring out a number of axiomless 'proofs' in refuting this work, which is fine; GEB:EGB treads upon feet firmly planted in the grottoes of philosophy departments. (The word "discredited" is the alarm call of these philosophers, as if fads are beautiful enough to determine truth.) Whatever the case may be, this is a very interesting book. Take a look at it in the bookstore, then buy or order it from Amazon if you find that you like what 'interesting' can mean.
Rating: 
Summary: Standing by my remarks
Review: I thank my critic for his comments, but I am afraid I shall have to stand by my points. Here is Hofstadter himself on p. 709 of this very book: "My belief is that the explanations of 'emergent' phenomena in our brains -- . . . [including] finally consciousness and free will -- are based on a kind of Strange Loop, an interaction between levels in which the top level reaches back down towards the bottom level and influences it, while at the same time being determined by the bottom level. In other words, a self-reinforcing 'resonance' between different levels . . . . The self comes into being at the moment it has the power to reflect itself. This should not be taken as an antireductionist position. It just implies that a reductionistic explanation of a mind, _in order to be comprehensible_ [Hofstadter's emphasis], must bring in 'soft' concepts such as levels, mappings, and meanings. In principle, I have no doubt that a totally reductionistic but incomprehensible explanation of the brain exists; the problem is how to translate it into a language we ourselves can fathom." In short, Hofstadter has indeed done exactly what I said he did: speculated that consciousness and "selfhood" arise when a system acquires sufficient power to represent itself. And not because anything causal happens at that "higher" level; on the contrary, the "higher" level serves only the purpose of making a completely reductionistic explanation "comprehensible" _to us_. Hofstadter argues, then, that consciousness comes into being when a system becomes sufficiently complex to represent itself. (And yes, in strict consistency he is committed to believing this condition sufficient to render _any_ system conscious; as my critic has helpfully noted, Hofstadter does indeed argue that the "substrate" is irrelevant.) Shortly before this passage, Hofstadter has suggested that "Godel's theorem offers the notion that a high-level view of a system may contain explanatory power which simply is absent on the lower levels" [p. 707]. And the discussion immediately following this citation makes clear that Hofstadter has in mind not what Godel's theorem _shows_, but the _method_ Godel used to show it: the "Godel-numbering" technique by which the "undecidable string" is generated. In fact Godel's own understanding of the theorem in question would have precluded Hofstadter's speculation as cited above. Godel himself thought he had shown that mathematical "objects" were _real_ in some Platonic sense, and that the mind possesses a sort of mathematical intuition which is not reducible to formal operations (and therefore the mind itself cannot be). But, as I said, the _conclusion_ is not the aspect of Godel's theorem that is of interest to Hofstadter. I continue to see a bit more than "hand-waving" in this criticism. Again, though, Hofstadter's book is brilliantly engaging on some of the very points noted by my critic.
Rating:
Summary: Reviewer is mistaken
Review: I think Steve Keppel-Jones has missed the point of the previous reviewer. Steve says self-rep and self-ref depend not an outside observer but on the properties of the system itself. But that's an example of the same sort of hand-waving the earlier reviewer was talking about: an isomorphism (a mathematical version of 'resemblance') isn't the same thing as a reference. (In some respects Kilamanjaro resembles Everest, and Everest Kilamanjaro; do the two mountains therefore *mean* each other?) Hofstadter and Keppel-Jones are implicitly relying here on a thoroughly discredited 'picture theory of meaning' dressed up in mathematical language. Isomorphisms do indeed require minds in order to count as references, and - as the earlier reviewer said - there must be a mind in the system already in order for this to occur within the system itself. Otherwise the 'meaning' is imported from an observer outside the system. By the way, if Hofstadter claims that consciousness doesn't require simulation of the neuronal level, then doesn't that mean self-rep and self-rep *are* sufficient conditions for consciousness to occur? Or is he speaking only of 'simulated consciousness' (whatever that might be)? What, in short, does Keppel-Jones mean in saying that 'mind can be *represented* by symbol-manipulating systems'? Is this more hand-waving?
Rating:
Summary: Brilliant and thoroughly engaging.
Review: I can't add much to the other 5-star reviews presented so far (it's a great book, life-changing, read it again and again, etc.) but I thought the comment by SandGRyan needed some followup. He appears to have misinterpreted Hofstadter's thrust on AI (as presented in GEB). To wit, Hofstadter never claimed that mind arises "simply from self-ref and self-rep", nor that a "formal system complex enough to represent itself [automatically] becomes conscious". Hofstadter's main point about AI is that he believes mind can be represented by symbol-manipulating systems without having to simulate the level of neurons and neurotransmitter molecules. Such symbol-manipulation systems, if they are to approach the complexity of a human mind, must *necessarily* be capable of self-ref and self-rep, just like a human mind; but those abilities are not *sufficient* conditions. Note that Hofstadter has not "distracted from the crucial fact" that self-rep and self-ref must be provided by an external agent, because this "fact" is far from "crucial". A solitary symbol-manipulation system (e.g. a human mind or a complex AI) is perfectly capable of generating and manipulating symbols which are, in fact, references to itself, independent of any other party's judgment. Whether any given symbol refers to the system in which it is being manipulated is a property of the symbol and the system in question, and has nothing to do with the opinions of any outside agencies. SandGRyan's conclusion that GEB:EGB is a "tremendous begging of the question", then, amounts to nothing more than the hand-waving of which he accuses Hofstadter. SandGRyan's discussion of Godel's Theorem also completely misses both Godel's point and Hofstadter's use of it. Godel's point is that any sufficiently complex system must be *incomplete*, which applies not only to mathematical formal systems but also to minds and brains. Hofstadter explicitly agrees with this, and gives examples of how human minds (and, necessarily, AIs) are just as incomplete with respect to truths about themselves as are mathematical formal systems. Hofstadter has not "reinterpret[ed] Godel's work in favor of strong AI", because Godel's work does not refute strong AI in the first place. For my own final comment: read the book! Lend it to your friends! Enjoy!
Rating:
Summary: Imaginative, but watch out!
Review: Douglas Hofstadter's imaginative and engaging GEB:EGB asks the question, "Can machines be conscious?" and answers in effect, "Certainly, because we ourselves are such machines." And there is no doubt that he earned his Pulitzer Prize for this fascinating book. But watch out! His reliance on imagination actually masks the real problem. The "real problem" is this: mind can't arise _simply_ from self-reference and self-representation, because reference and representation presume the existence of a mind to begin with. Only minds refer and represent; _resemblances_ (even fancy ones like "isomorphisms") aren't references/representations. And in Hofstadter's undeniably well-presented examples, his reliance on imagination serves to distract from the absolutely crucial fact that the reference and the representation are always provided by a mind _outside_ the system in question: the reader. A formal system complex enough to "represent itself" doesn't become conscious; it takes a mind _outside_ the system to "see" the isomorphisms in question as references/representations. The system _itself_ can't do so unless mind is _already_ there -- so Hofstadter's bootstrapping "explanation" fails. As an _argument_, then, GEB:EGB is a tremendous begging of the question. Invoking Godel's Theorems and waving one's hands about "strange loops" doesn't alter the fact that Godel's Theorem itself delivers a killing blow to "computational" theories of consciousness: semantics is _not_ reducible to syntax; truth is not reducible to provability within a formal system; reason is not reducible to purely formal logic; meaning is not reducible to isomorphism; and mind is not reducible to computation. (And indeed, this reading has much more in common with what Godel himself thought he had shown than does Hofstadter's attempt to reinterpret Godel's work in favor of strong AI.) But GEB:EGB is still a remarkable intellectual accomplishment and a joy to read. Just be careful!
Rating:
Summary: Mandatory reading
Review: Absolute, unmitigated genius. Hofstadter takes difficult concepts in several different fields (mathematics, philosophy, psychology, music, biology, art, and even religion), explains them lucidly and even humorously, and brings them all together into a single unified whole. This book is an absolutely indispensible part of a modern liberal arts education.
| 
