56 of 68 people found the following review helpful
A bit disappointing,
This review is from: I am a Strange Loop (Hardcover)
Hofstadter revisits a number of topics from his earlier books, centered around his concept of a "strange loop." All rests on a few basic observations about multi-scale systems: (i) the higher (aggregate) levels can often be described more succinctly and profitably with their "own" sets of laws; (ii) information flows both up and down between the various levels (essentially through boundary conditions, although Hofstadter can be trusted to come up with more flowery terms like "downward causality"); (iii) whereas the lower levels involve a modest set of different entities with relatively simple rules, the higher levels tend to allow for a huge variety of entities behaving in complicated ways; (iv) hence the higher levels are endowed with representational power and can accomodate a representation of the system itself; (v) such a self-image would have to be abstracted relative to the real thing, meaning that the lower level is in some sense inaccessible at the level where representations interact.
Hofstadter makes these points using a very neat pedagogical example, called the "Careenium" (which I believe was first introduced in an Achilles-Tortoise dialogue in "Metamagical Themas"). Especially observations (i) and (ii) are brought out very nicely by the Careenium. Hofstadter spends a lot of time discussing observation (iii), which is really not such a hard idea to come to terms with. This is a pity since general popular science books often make the reader feel clever in a cheap way by banging on endlessly about a simple notion; such tactics are generally beneath him. I would have welcomed a Hofstadterian analysis of the technicalities surrounding (ii), which, after all, are his research speciality. Instead, we get a discussion of Goedel's construction, which is fine, even if it is just an abridged reprise of GEB. The most tenuous (and tedious) part of the book is where Hofstadter connects the Goedel construction with multi-scale systems by insisting that observation (iii) holds in both cases. I am not so convinced of the strength of this analogy. Does Goedel's construction really mean that at the level of fantastically long PM strings, PM is thinking about itself? What is lacking is perhaps that we can construct meta-mathematical statements as number-theoretical statements, allowing them to "talk" about themselves or other statements, but they just "sit there" (not exactly in plain view, but that is beside the point). They do not interact much. I make rather a lot of this point since it seems to me that the general usefulness of the concept of "strange loop" is riding on this. My impression is that, after discarding throwaway examples like Escher prints and the like, the human mind (with its self) is the only instantiation of a strange loop that Hofstadter is really serious about.
At any rate, Hofstadter seems to be aware of the weakness, since this is where he resorts to italics and talk of the system "engulfing itself": signs that words are failing him and he hopes that we will "get it". I am reluctant to go along, irked as I am by the assertion, repeated several times over, that Russell himself never "got it" while I am not given detailed pointers to the literature in support of this unkind appraisal, true as it may be (the only reference to work by Russell is to the Principia itself; annoying in a reference list more obsessed with cute references to fictional works).
Perhaps so as not to stress the multi-scale vs Goedel analogy to the breaking point, another earlier example of Strange Loops is not or barely discussed. This is the soi-disant "isomorphism" between the Goedel construction and molecular biology, which received a lot of emphasis in GEB. Good riddance, since this analogy is not all that great either. On the other hand, it is a pity that relatively little attention is paid to biological systems besides the brain which actually are genuine examples of multi-scale systems, like social insect societies or the immune system. I would be very interested on Hofstadter's take on the immune system.
Hofstadter's resolution of the mind/body problem is based on observation (v): his claim is that thinking about our minds in terms of the neurological processes that form its physical correlate naturally comes very unnatural to us. This is basically the "there is no real problem" argument advanced by, among others, Hofstadter's friend Dan Dennet.
The book offers the usual entertainments of a Hofstadter book: parables making technical points, clever neologisms (for one's memory-analogue of things that never happened, or might yet happen). There is also a dollop of self-indulgence (do we need footnotes explaining the jokes?), which is more annoying here because it is not tempered with intellectual rigour as it was in earlier books, which always distinguished clearly between fact and opinion and opinions were usually closely reasoned. This book feels sloppy and rushed by comparison, and could have done with proof-reading by friends who know their biology.
Sort: Oldest first | Newest first
Showing 1-2 of 2 posts in this discussion
Initial post: 2 Nov 2007 11:21:22 GMT
Carno Polo says:
perceptive review but too long...
Posted on 15 Jul 2008 14:41:01 BDT
Olly Buxton says:
Well I thought it was just the right length. Great review.
‹ Previous 1 Next ›