Books Read in 1998

The Demon-Haunted World is a hodge-podge. Sagan is a fairly good writer, but many of his targets here seem like straw-men. Of course, I'm probably overestimating the average intelligence of Americans, but does a significant portion of the population really believe in alien abductions and psychics? Maybe so. For me, the best part of this book was toward the end, when Sagan tells the story of Maxwell's breakthrough in electromagnetism and how it has affected scientific thought. In the final analysis, Sagan makes strong cases for revamping scientific education for schoolchildren and for public funding of basic scientific research.

This was Vonnegut's first novel (it was published in 1952), and it's pretty obvious that he hadn't yet found his voice. The ideas are great, however, and it's a worthwhile read for Vonnegut afficianados.

This is perhaps the best time-travel novel ever written. The episodic pastiche of Billy Pilgrim's life is very carefully constructed, and Vonnegut's prose flows with deceptive ease. When I first read this (in high-school), I didn't recognize the references to Vonnegut's other books, but several characters have cameos, including Kilgore Trout (of course), Eliot Rosewater, Howard Campbell, and the city of Ilium. There may even be others, but I haven't read all of Vonnegut's novels yet, and I may have missed some.

This novel (actually a collection of ultra-short sketches) reads like something from Calvino. It is reminiscent of Invisible Cities, for example. The prose is poetic and engaging, and the whole thing doesn't feel contrived, even though it is. This novel is short enough to be read in a single sitting, but it is worth savoring.

This was a fun, light read while I sat around in hotel lobbies and airport terminals. Varley crafts ingenious universes and many of the better elements are cycled through most of his novels and stories. This is the best of his novels that I've read (it's a lot tighter than Steel Beach, but perhaps still not as good as his short stories).

This novel is a lot of fun, due in large part due to Ruff's considerable powers of imagination. It reads a little bit like Vonnegut, but the prose is more conventional and the fictional world more thoroughly developed. Ruff nods several times to Pynchon, and Pynchon supplies a nice cover blurb, but I disagree with the reviewer who said that fans of Pynchon and David Foster Wallace would enjoy Ruff's novel. Pynchon and Wallace write considerably more complex (and interesting) prose. I think Ruff is more of a cross between Vonnegut and Neal Stephenson. In the final analysis, any writer who beats up Ayn Rand the way Ruff does in this novel (that is, thoroughly and intelligently), earns good marks. I enjoyed SG&E much more than Fool on the Hill, and I hope Ruff's next novel isn't quite so long in the making.

Richard Powers has clearly established his place as one of my very favorite writers. Although Three Farmers doesn't exhibit all of the mastery evident in The Gold Bug Variations, it has many of the characteristics that distinguish Powers' mature writing. I eagerly await the publication of Gain in paperback.

Vonnegut is just brilliant. He has an amazing ability to make the most horrible aspects of human life gut-wrenchingly funny. While this novel isn't as tight or as funny as Cat's Cradle, it's still much better than most modern novels.

This was a pretty huge disappointment for me. The novel is told in the voice of a nine-year-old girl, and the character development is too subtle; on the surface, pretty much nothing happens during the entire novel. At the moment, I'm of the opinion that Baker's novels have been on a steady downward slide since the excellent The Mezzanine.

If only it were that easy. Bolker is a psychologist, and this book is basically about getting out of writing ruts. Speaking from my own experience, it sure is hard to get started writing the dissertation document.

This was cited in Wiener's Cybernetics and in Marr's Vision, and the pun in the title was enough to make me seek out this obscure monograph in MIT's unbelievably great library system. Unfortunately, the book itself isn't terribly interesting. It's a criticism of Ayer's views on sensory-data and perception.

Dennett is fun to read. He seems to think of himself as simultaneously a philosopher and an outsider to the philosophical community. This book is great when he's talking about his own ideas, but I found it quite boring where he analyzes the relations of his ideas to those of other philosophers (what did I expect out of a philosophy book?!). I recommend Consciousness Explained somewhat more highly (I got more out of it).

I read the first third of this. The philosophical implications of the ideas just aren't that interesting to me right now, but I may read Parts II and III someday. The first third, however, is quite good, particularly for pointing out some of the finer points of Rosch's Basic-level Categories research.

Picard tackles a lot of difficult issues related to computers and emotions (both giving computers the ability to sense and react to human motion and the more controversial idea of "giving computers emotions"). To me, this book reads somewhat like a bloated doctoral proposal. There's a sense of: here's a topic, and here are some tools that might be useful for attacking it, but we haven't really done any of the work yet. So it works as a manifesto of sorts, but there aren't any real results in the book itself. Working in the same lab as Picard and her students (and knowing many of them personally), I have a fairly good idea of what they're up to now, and they seem to be making some progress toward the goals set out in this book. This area is something to be taken seriously.

This purports to be a biological theory of mind. In it, Edelman touts his "Neural Darwinism" theory and makes wildly exaggerated claims about how it explains consciousness and human thought. There is no doubt that his low level ideas are a useful direction for research, but his conclusions seem wholly unfounded. We need better theories to bridge neurophysiology to human thought. In my opinion, Dennett's Consciousness Explained offers better ways to think about consciousness, and Minsky's The Society of Mind offers better ways to think about the complexity of mind. Your mileage may vary.

I bought this sight-unseen, and it turned out to be written at a much lower level than I had hoped. This might be appropriate for an undergraduate introduction to musical sound, but that's as far as it goes.

This book is a philosophical essay on the differences between recognition, identification, and classification. In quite a lot of the scientific literature, these terms are confounded, so Sayre's careful thinking about the semantic differences is worth considering, even if his presentation is extremely drawn out. Sayre's application of these ideas was to handwriting recognition; I don't know the historical context (this was the early 60s), but it appears to have been very forward-thinking.

Martin Cooke's dissertation is one of the first attempts to build a sound-analysis system based on Bregman's Auditory Scene Analysis framework. Some of the ideas from his "synchrony strand" representation translate to the correlogram representation that I'm using in my research, and that is my main interest in his work.

This is the AI book they used at Cornell in the early '90s. I wanted a general overview of knowledge representation, so I read those sections and skimmed the rest.

This is the classic work on musical sound, written in the mid-19th century by one of the greatest scientists of that era. The early sections on the physical acoustics of musical instruments and on the components of instrument "timbre" are very much worth reading. I did not find the later sections on beats and consonance/dissonance to be very useful, but I'm happy to have read this historically very important work.

Professor Grimson is a member of my doctoral committee, so I thought it would be a good idea to make myself familiar with some of his research in vision, even though it may not apply directly to my own research in hearing. Indeed, the concentration in this book is on geometrical constraints for visual object recognition (as is fairly obvious from the title). It is not clear whether similar constraints are useful in hearing, or whether his approach, which is not based on modeling human perception, is directly relevant to my own work. Nonetheless, the presentation is rigorous and the results impressive.

There is a lot of good material in this book about Bayesian BeliefNetworks. I didn't find the early parts of graphoid theory to be very interesting, but the later material is very practical. I'm particularly interested in applying the parts about taxonomic reasoning to my work in musical instrument recognition.

What I read was a partly-written draft of the forthcoming update of the classic pattern-recognition text. The chapters that are fleshed out are wonderful, and I eagerly await the completed text, which covers a broad range of important topics.

I don't have much to offer by way of comments here, other than the obvious "how surreal" and such. I would find it interesting to read a psychological evaluation of Joseph K.

There are a couple of passages in here that I think are references to The Stranger, but I'm not quite sure. Overall, this book is about coming to terms with the fact that there is no God, and that we are meaningless bits of matter who need to think that we have a purpose in order to be happy. Camus seems to think that love between people is the only thing people can somewhat control that can make them happy. I'm drop-dead tired as I write this, so I'm not going to bother trying to write an intelligent review.

I started this a day early, apparently, but I don't think that explains why I didn't care much for this book. Winter's Tale is something like a John Irving novel, written while tripping on acid after having read juvenile fantasy books. Okay, so it's not that bad. The prose is clean, but mostly uninspired; there are no descriptions that jump off the page like some of the ones in Memoir From Antproof Case did.

Some people adore this book, and I can in part see why. It's a breezy read with a large number of fairly interesting characters and quite a lot of creative reality twists. On the down side, I didn't find the overall plotting of the book to be very coherent, and there are far too many characters that aren't quite "minor", even for a near 700 page novel. My main frustration stems from the fact that there were so many promising elements in this novel that never became coherent.

The opening chapter of Players is brilliant. DeLillo sets the scene in an airplane, where the in-flight movie (a terrorist thriller) is accompanied by a lounge pianist in the manner of the silent films of years past. The rest of the novel is vintage DeLillo, with the theme being people who are morally detached from the world. The novel is somewhat unsettling, perhaps because it doesn't quite make sense.

In this novel, Vonnegut uses Borges's technique of writing summaries of nonexistent novels and stories written by a fictitious author (in this case, of course, the author is Kilgore Trout). One of the main themes of this novel is that humans are automatons having no free will. The idea that people are characters in the mind of an unknown author is visited in several ways. Breakfast of Champions is more "vulgar" than the average novel (by the end of the novel, the reader is aware of the penis measurements - both length and girth - of every male character), and I'm not sure what narrative purpose Vonnegut's technique serves.

Dan Simmons is an amazing writer. His prose is consistently readable and fluid (though most of the time unremarkable); it is his imagination that stands out. I don't read very much science fiction, but I won't let myself miss any that this genre-hopping scribe produces.

It is a much stronger novel than Endymion, but not as good as the first two novels in the series, which can be read as a pair ( Hyperion and The Fall of Hyperion). Many of the loose ends from the first three novels are resolved in the final novel of the series — most of them satisfyingly — but the main flaw is that too many of the plot twists have a deus ex machina quality to them. Raul Endymion is somewhat awkward as a hero, and both he and Aenea remain remarkably shallow characters for the 1000+ pages devoted to them. My main complaint is that Simmons devotes more energy to developing the planets that the characters visit than to the characters themselves. On the plus side, some of the characters from the first two novels do gain in depth during the final installment of this epic.