What I read in 2005

12/31 Anne Krueger, Parenting Guide to Your Baby's First Year
Just what it sounds like: a guide written with the editors of Parenting magazine. I'm sure this book will come in very handy over the next year.
12/30 C. S. Lewis, The Horse and his Boy
12/27 C. S. Lewis, The Silver Chair
12/26 C. S. Lewis, The Voyage of the Dawn Treader
12/25 C. S. Lewis, Prince Caspian
12/24 C. S. Lewis, The Lion, the Witch, and the Wardrobe
12/21 Eric Linklater, The Wind on the Moon
12/15 Jack Vance, The Languages of Pao
A novel about the strong form of the Sapir-Whorf hypothesis.
12/12 Fritz Leiber, Ship of Shadows
12/10 Poul Anderson, No Truce With Kings
12/10 Ruth Reichl, Tender at the Bone
This isn't the book I had expected. I knew it was a memoir, but I had expected it to be about the life of the New York Times restaurant reviewer. No, it's a growing up story. (With the subtext: this is how one grows up into the kind of person who might become a major restaurant critic.) The other thing I hadn't expected was how funny it is! I suppose if I'd come from New York I would have known she was such a funny writer, because I would have been reading her for years.
12/5 Rudyard Kipling, Puck of Pook's Hill
12/3 Hannah Arendt, The Origins of Totalitarianism

I'm sorry the copy I read didn't have more information on publication history. Samantha Powers, who wrote the introduction, talked a little bit about the history of this book but didn't give all of the details. I think that what I read was the third edition, as revised in 1957. The first edition was published in 1951, and in the text I read Arendt mentions "the second (paperback) edition" a few times, so that would make this the third unless there's another edition that she didn't mention. In any case, whatever text I read is now somewhere around 50 years old. What does it look like in retrospect?

One thing I find striking is that Arendt makes statements about totalitarianism in general, even though she was quite explicit that she was dealing with a sample size of exactly two: Nazi Germany and the Stalinist Soviet Union. She was careful to distinguish between totalitarianism and ordinary tyrannies. Not every communist country is totalitarian, not every fascist country is totalitarian, not every country where racism is a foundational principle is totalitarian, not every absolute tyranny is totalitarian, not every country run by a genocidal madman is totalitarian. She does not regard the Soviet Union as totalitarian before the 1930s or after 1953. She mentions the possibility that communist China might be totalitarian, but says she doesn't know enough to give an answer one way or the other. (I suspect that if she had lived long enough to publish a fourth edition in the 1980s she would have said that China was totalitarian during the Cultural Revolution, but not before or after.)

So: a sample size of two (or maybe three), from the middle of the 20th century, each of which lasted a decade or two and flourished during the disruption of wartime. Does this really mean that totalitarianism is as unique, as radically different from other evil governments, as rare, as one would think from this book? Or does it mean that trying to discuss "totalitarianism" as a category, trying to say thing that are true of both Hitler's and Stalin's regimes but not true of anything else, is a category error?

11/27 Paul Hudak, John Peterson, and Joseph H. Fasel, A Gentle Introduction to Haskell 98
11/14 Walter Jon Williams, Dread Empire's Fall: Conventions of War
It is of course not a coincidence that the dissolute young nobleman who stays behind in the enemy-occupied city is named "Pierre". Not in a long multi-character book about tangled personal relationships in wartime.
11/13 Alison Bechdel, Invasion of the Dykes to Watch Out For
The latest collection of my favorite comic strip.
11/11 Walter Jon Williams, Hardwired
11/8 Jules Verne, Around the World in 80 Days (tr. George Towle)
11/3 Dawn Powell, The Wicked Pavilion
One of Powell's New York novels.
10/29 Simon Thompson, Haskell: The Craft of Functional Programming (2nd ed.)
Haskell is a strongly typed functional language, like ML. It differs from ML in that it's lazy, not strict, and in that it doesn't emphasize type inference so heavily. I'm mainly interested in Haskell because of its type classes and monads. The type classes are straighforward, but I'm sure I don't yet understand everything that monads are good for. And I definitely don't understand why monads in Haskell are called that; I don't see what they have to do with monads in category theory.
10/23 Kyril Bonfiglioli, Something Nasty in the Woodshed
The third and last Mortdecai novel.
10/18 John Crowley, Engine Summer
“It was as though a great sphere of many-colored glass had been floated above the world by the unimaginable effort and power of the angels, so beautiful and strange and so needful of service to keep afloat that for them there was nothing else, and the world was forgotten by them as they watched it float. Now the sphere is gone, smashed in the Storm, and we are left with the old world as it always was, save for a few wounds that can never be healed. But littered all around this old ordinary world, scattered through the years by that smashing, lost in the strangest places and put to the oddest uses, are bits and pieces of that great sphere; bits to hold up to the sun and look through and marvel at—but which can never be put back together again.”
10/8 Peter O'Donnell, Dead Man's Handle
Modesty Blaise.
10/8 Lawrence Osborne, The Accidental Connoisseur: An Irreverent Journey Through the Wine World

The book began by asking: what is taste, and how do you know whether you have it? What, if anything, is the distinction between being a connoisseur and being pretentious? After a while, though, there seemed a bit too much contrast between the faux-simple stance and the author's obvious wine expertise. Every place he went, whether California or Piedmont or Languedoc, he talked about famous wines and famous winemakers that everyone (in a very limited sense of "everyone") had of course heard of.

The most interesting thing I learned: the way reviewers and critics talk about wine now, identifying elements like "black cherry" and "tobacco", is quite recent. Before the 1970s, wine critics wrote about wine in other ways. This change in method of description may represent an increase in precision, but certainly ties in with changes in the social functions of wine.

10/5 Charles Dickens, Our Mutual Friend

This book is all about masks and deceptions. The reader is told the central deception, the identity of the main character, fairly early, and most readers probably guessed it even before they were told. Some of the other mysteries aren't revealed so quickly.

One of the deceptions that I find most interesting concerns a secondary character, Riah. The reader learns immediately that Riah is a mensch, that he's almost saintly; if he has any character flaw, it would be an excessive sense of gratitude. Most of the other characters in the book, however, are fooled into thinking that he is a rapacious moneylender. It's hard not to contrast this with Dickens's other important Jewish character, Fagin. Is this Dickens's way of apologizing?

9/28 J. Roger Hindley, Basic Simple Type Theory

Despite the title, this is not an introduction to the field of type theory. It probably could be used in a first course on type theory, but it's a specialized book: it's a short and mathematically rigorous treatment of a specific type system, TAλ.

The "simple" part is because TAλ is one of the two simplest ways of adding types to the pure λ-calculus. It is still a purely functional system, and has no atomic constants (hence no atomic types): it consists of the λ-calculus with the addition of a logical framework for deducing that a type may be assigned to a term. Some terms, such as λx. x x, aren't typable. (Given the fact that type theory started with Bertrand Russell, this isn't surprising.) Other terms, such as λx. x, can be assigned an infinite number of types. So TAλ exhibits polymorphism—a specific type of polymorphism, which most people other than the author of this book describe either as ML-style polymorphism or Hindley-Milner polymorphism.

One of the nice things about this book is a historical discussion. The two ways of adding types to the λ-calculus go back to Alonzo Church and Haskell Curry. In Church's approach, terms are tagged with types. In this approach there is a different identity function for each type. In Curry's approach, by contrast, there is a single identity function that can be assigned the type σ→σ for arbitrary types σ.

Probably the most interesting part of this book is its proof of the principal type theorem, saying that for any typable term there is a single type, unique up to renaming, from which all of its other types can be generated. The principal type algorithm is what makes languages like ML possible.

This book devotes a fair amount of space to the Curry-Howard correspondence, which establishes a mapping (an isomorphism, if you squint appropriately) between type deductions for closed terms in TAλ, and deductions of tautologies in intuitionist logic. This correspondence has a practical benefit: it means that type theorists can use theorems and techniques that originated with logicians, and vice versa. But beyond that practical use I'm not completely sure whether this shows a deep connection that gives insight into these two fields, or whether it's just a purely formal similarity that doesn't mean anything particularly interesting. I suspect the latter. In that respect it reminds me of the structural similarity between the partition function in statistical mechanics and the generating function in the path integral formulation of quantum field theory. The similarity is interesting, and it allows you to use similar computational techniques in different fields, but I've never seen any way in which it gives you any real new physical insight.

9/27 Vicki Iovine, The Girlfriends' Guide to Pregnancy: Or everything your doctor won't tell you
The author has no medical training and doesn't pretend to. This is not intended to be one's only pregnancy book. The idea is: These are the sorts of things your friends who have been through it before ought to tell you. It's part advice (including things like fashion advice), part reassurance, part entertainment.
9/25 Iain Banks, The Business
In some ways this book reminded me of Gibson's Pattern Recognition. It's not science fiction but it shares some of the concerns of science fiction, such as the contingency of history and the importance of social and economic change.

I didn't like this as well as many of Banks's other books. The central conceit of the Business itself was fun, but I don't think the book did anything very interesting with it. The setting and the characters seemed too detached from each other.

9/24 Ann Douglas, The Mother of All Pregnancy Books
Yes, I am reading this sort of book for the obvious reason.
9/18 Susanna Clarke, Jonathan Strange & Mr Norrell
9/18 Lois Bujold, Diplomatic Immunity

I picked this book up again because Janet has just finished reading the Vorkosigan books, and it was lying around the house. Like many of the Vorkosigan books Diplomatic Immunity is largely about reproduction, including the social effects of advanced reproductive technology. I don't know of any other writer who has done as much with this subject as Bujold.

I've been thinking a lot more about reproduction lately than the last time I'd read this book; the opening scene, where Miles is cooing over the medical imaging showing "baby's first cell division", cracked me up.

9/10 William Shakespeare, The Tragedy of Julius Cæsar

Marc Antony's funeral oration is often quoted as an example of great rhetoric, and with good reason. The contrast with Brutus's speech is especially striking. An observation that I haven't seen quite as often is that Antony's speech is superior not just in the rhetoric, but also in the arguments. It's not just that Brutus's arguments were weak, but that they're practically nonexistent. Listening to his speech you would know that Brutus thought it was necessary to murder Caesar to preserve the republic, but you wouldn't know why he thought so.

There were arguments to be made, of course. Other characters in the play make them. It is true (at least in the context of the play; I have no idea whether the play is historically accurate in this respect) that Caesar was to have been crowned king, at least outside of Italy. And the play begins, after all, with a reminder that Caesar's "triumph" is for defeating a Roman army, not a foreign one, and with admonishing the people for ingratitude toward Pompey; this is an interesting echo of the fact that ingratitude is also Antony's strongest charge against Brutus.

It's also interesting that there isn't a clear central character. Not the title character, surely, who dies less than halfway through. Brutus perhaps comes the closest, but there are large chunks of the action that he isn't central to. It's also not clear how admirable he is. He isn't venal or envious, but he is at least in part motivated by an exaggerated sense of self-importance. On several occasions he shows himself to be weak and indecisive, and he was manipulated into the conspiracy. It's not at any clear that we should take "This was the noblest Roman of them all" any more at face value than "Brutus is an honorable man."

The Roman Empire is a symbol that has been used for many purposes. Today, to a 21st century American, Octavian's subversion of the republic, preserving its outward forms while assuming unchecked power, making himself a king in all but name, seems like a sickening betrayal. In 13th century Florence, on the other hand, Dante selected Brutus as one of the three traitors at the absolute center of Hell. To Shakespeare and his audience, living in the Tudors' powerful centralized monarchy, the Roman Empire probably meant something different yet again. Probably the ambiguity in this play is intentional.

9/4 Gene Wolfe, Exodus from the Long Sun
Now I know who and what the gods are, who built the Whorl, and (in part) why. There is a real connection with The Book of the New Sun, but this is in no sense a continuation of that story. It seems pretty clear, however, that The Book of the Short Sun will have to be a continuation of this story.
9/2 Gene Wolfe, Caldé of the Long Sun
8/31 Gene Wolfe, Lake of the Long Sun
8/28 Adobe Photoshop CS2 User Guide
8/27 Stuart Gilbert, James Joyce's Ulysses

If this wasn't the first book-length literary study of Ulysses, it was certainly one of the first. For the most part, this book can be considered an exposition of Joyce's famous schema where each chapter is assigned a Homeric correspondence, a time of day, a color, a technic, an art, a symbol, etc. Again, the first edition of Gilbert's book might have been the first place that schema was actually published. (Joyce told Gilbert and others about the schema, but never published it himself.)

Besides the schema a good deal of this book consists of explaining esoteric doctrines of metempsychosis, and discussing possible historic origins of the Odyssey: identifying a geographic feature from which a legend of Scylla and Charybdis might have originated, pondering where to find Helios's island on a map, and so on. I don't find any of that worth the effort. Yes, there are occasional allusions to metempsychosis (beginning with "met him pike hoses"), and there are a few scraps of Sanscrit or pseudo-Sanscrit here and there, but I just don't see how a detailed explanation illuminates Ulysses. Same for the supposed historical origins of the Odyssey: Gilbert provides no argument for why such an astonishingly literal-minded reading of Homer tells us anything interesting about Joyce's book. I don't think it does. To me, what all that stuff smacked of was someone who did his homework, collected a lot of background notes, and couldn't resist the temptation to include all of his notes in his final published work.

The subtext of a lot of this is a defense of Ulysses against the charge of obscenity. That's the main reason that I think this book is outdated and that I need to read a different literary study instead. United States v. One Book Called "Ulysses" was more than seventy years ago, and it's time for us to get over it. Judge Woolsey's opinion is still important to First Amendment law, but there's no reason for it to be a focus for a literary analysis of Ulysses.

8/27 Gene Wolfe, Nightside the Long Sun

First book of The Book of the Long Sun. Despite the similarity of names, it doesn't appear to have any important connection to The Book of the New Sun (although there do seem to be some hints that it might be set in the same universe). Patera Silk couldn't be more different as a character than Severian. He reminds me more of Father Brown, which is quite possibly deliberate.

The book ends with political intrigue, but the question I'm wondering most about is: exactly who and what are the gods, and is Silk right to worship them?

8/21 Christopher Moore, Island of the Sequined Love Nun
8/20 James Joyce, Ulysses

Yes, it really is just as great, and just as difficult, as they say. (Like most people, I find "Oxen of the Sun" the most difficult chapter by far.) This was the third time I've read it. Each time I'm getting more out of it.

Ulysses has the reputation of being an example of stream-of-consciousness, or interior monologue. This isn't completely false, but it also isn't completely true. Every chapter is written in a different style; some are stream of consciousness and some aren't. What I find most interesting about Joyce's use of this technique is that we see enough of Stephen's interior rhythm in early chapters like "Telemachus", and of Bloom's in early chapters like "Lotus-eaters", so that in later, more complicated chapters, like "Circe", we can immediately recognize their thoughts even in tiny fragments.

This time the main thing I noticed about this book was the way in which both Stephen and Bloom were haunted by the past, by losses that they still haven't reconciled themselves to, and by the vague sense of having betrayed the dead.

8/11 Sara Paretsky, Windy City Blues
A collection of V.I. Warshawsky short stories. By far the silliest was "The Maltese Cat", which is just what it sounds like.
8/10 Frank Conroy, Stop-Time
A memoir and coming-of-age story.
8/4 Lois Bujold, The Hallowed Hunt
8/2 W. Richard Stevens and Stephen A. Rago, Advanced Programming in the Unix Environment (second edition)

The first edition was published in the early 90s, and it was a book that every serious Unix programmer needed to read. Once you'd gotten used to the ISO C library, it was the place to go to learn how to write C programs using Unix system calls and other direct interfaces to the OS: I/O and file manipulation (open and close and creat, read and write, stat, fcntl, mmap), process control (fork and exec and dup), signals, file record locking, and so on.

In the time since the first edition came out, both Unix standards and the real world of Unix programming have changed. Network programming is more important than it was, and we're starting to see more use of IPv6. Almost all nontrivial programs use dynamic libraries and threads. We're all more security-conscious than we were.

Unfortunately, Stevens is no longer with us to do an update—hence the coauthor.

One of the most striking differences between the two editions is that the commercially significant Unix systems are different than they were a decade ago. The second edition focuses on Linux, OS X, FreeBSD, and Solaris. I think it's reasonable to treat those OSs as the versions of Unix that matter.

7/30 J. K. Rowling, Harry Potter and the Half-Blood Prince
Yes, it's a lot better than Order of the Phoenix. What a very strange open-ended ending, though!
7/28 Eoin Colfer, Artemis Fowl
7/22 Karin Lowachee, Cagebird
7/4 Carl Hiaasen, Stormy Weather
6/21 Terry Pratchett, The Thief of Time
6/21 Doug Henwood, Wall Street
This is not a book on how to choose investments. It's a careful explanation of how the American financial industry works, written from the perspective of someone who is skeptical of capitalism. (The author founded the Left Business Observer.) It asks, for example: what are the different kinds of financial instruments in use today, how did they develop historically, what are the actual social and economic functions performed by the financial industry and how do they differ from what that industry says are its functions, and how many of the typical left critiques of the WTO and the Federal Reserve are coherent? This is the first book I've read by someone who's very familiar both with the technical details of American financial analysis and with Marxist thought. It's a fascinating perspective.
6/20 Alan Furst, Night Soldiers
A historical novel set in the world of espionage in Europe in the years preceding and during World War II. The author's books are often described as "panoramic", and I can see why: we see the Balkans, Spain, Paris, Moscow, and so on. The general tone of this book is, I think, more typical of today's novels than of spy novels from a few decades ago. The main character isn't a dedicated agent fighting for a cause. For much of the book he works for the NKVD, but his loyalties are always complicated and conditional. Probably realistic. I imagine it would be hard to be completely loyal to a system that undergoes another senseless and bloody purge every other year.
6/18 David R. Butenhof, Programming with POSIX Threads
The standard introduction to pthreads, and as far as I know still the best. I'd never read a pthreads book before, but much of this was familiar anyway; multithreaded programming is now the norm. Parts of the pthread model are elegant, and it's good to see why they were designed the way they were. Other parts are ghastly, the result of trying to retrofit threading into a language and OS that already existed. The C version of synchronous thread cancellation is just so obviously a poor substitute for C++ exceptions, and fork handlers strike me as too convoluted for anyone to use correctly.
6/15 Hayao Miyazaki, Nausicaä Of The Valley Of The Wind 7 (tr. Matt Thorn)
6/14 Hayao Miyazaki, Nausicaä Of The Valley Of The Wind 6 (tr. Matt Thorn)
6/12 Charles Stross, The Hidden Family
The Family Trade isn't really a novel, but the first half of a novel. The Hidden Family is the second half.
6/11 Hayao Miyazaki, Nausicaä Of The Valley Of The Wind 5 (tr. David Lewis and Toren Smith)
6/10 Hayao Miyazaki, Nausicaä Of The Valley Of The Wind 4 (tr. David Lewis and Toren Smith)
6/9 Charles Stross, The Family Trade
Charlie describes this as fantasy, but to me it reads more like science fiction in the way that it builds its world: more extrapolation than wholesale subcreation, exploring the consequences of a single "what if". What's I find especially interesting is that, as in so much of Charlie's other fiction, one of the most important sciences that this book explores is economics and one of the most important technologies is finance. I can't think of any other writer who takes innovation in corporate governance or financial structures as seriously as innovation in rocket engines or biotech.
6/7 Burton Malkiel, A Random Walk Down Wall Street: The Time-Tested Strategy for Successful Investing
An introduction to investment, for individual investors. I read an earlier edition a decade or so ago; this one has a 2003 copyright date, which I think makes it the 8th edition. It's still a good book, and I'm about as convinced of the efficient market hypothesis as Malkiel is: I don't believe financial markets are rational all of the time, but I do believe it's rare for there to be deviations from rationality of a form that can be exploited as investment opportunities.
6/4 Joe Haldeman, Camouflage
Co-winner of the 2004 Tiptree Award. As promised: it's a good science fiction story that wouldn't have been out of place in Analog, and it also says interesting things about gender. [Update: a friend pointed out that "wouldn't have been out of place" is an understatement: this book was originally serialized in Analog.]
6/3 Patricia C. Wrede and Carolinie Stevermer, Sorcery and Cecelia

One of the reasons this book is so fun, I'm sure, is that the authors had fun writing it. It's an epistolary novel, an exchange of letters between two young ladies in Regency England. And that's just how it was written! Each author took a character, and they wrote letters to each other. The authors didn't discuss the plot with each other, except through the letters themselves. And then when they were done, they realized they had written a book.

As the title might suggest, It's a slightly altered version of Regency England: one of the pieces of country gossip that Cecy mentions in her first letter is that her neighbor Sir Hilary Bedrick has just been named to the Royal College of Wizards.

5/31 James Surowiecki, The Wisdom of Crowds
5/30 L. Timmel Duchamp, The Grand Conversation
A collection of essays on feminist science fiction. First in the "Conversation Pieces" series from Aqueduct Press.
5/30 Alain de Botton, How Proust Can Change Your Life
A curious mixture of literary criticism, biographical snippets, and self-help. I certainly learned things I hadn't known before about Proust's life and how it related to his work. Now I need to read a real biography.
5/25 Scott Meyers, Effective C++ (Third Edition): 55 Specific Ways to Improve Your Programs and Designs
Big improvement over the second edition. This book was revised in light of the last decade's worth of experience with C++ programming; it now contains useful advice about templates and design patterns, for example, and takes use of the standard library (and even parts of TR1) for granted. There's no space wasted on workarounds for buggy compilers with end-of-life dates in the previous century.
5/24 Diana Wynne Jones, Conrad's Fate
A new Chrestomanci book!
5/23 Sandy Sturges (ed.), Preston Sturges by Preston Sturges

This book was published in 1990, decades after Preston Sturges's death. It's told in the first person from his point of view, but I'm not at all clear on how many of the words are his. The forward, by his son Tom Sturges, said that his mother Sandy "put together everything my father wrote about himself—from letters, diaries, and an unfinished autobiography—to create what you hold in your hand: a seamless narrative that tells the story of his life." So, Preston and Sandy share the authorship somehow.

This book wasn't at all what I was expecting. I was expecting a Hollywood story, since, after all, Preston Sturges was a famous screenwriter and director. But the part of his life where he made masterpieces like The Lady Eve was only about fifteen years long. We don't even get to Hollywood until page 267, more than three quarters of the way through the book.

Not that the first three quarters of the book were boring! Sturges spent most of his childhood in various parts of pre-War Europe, and it was an extraordinary childhood. His mother's best friend was Isadora Duncan, and he also got to know everyone from Cosima Wagner to Aleister Crowley.

What's also extraordinary is how good both Sturges and his mother were at acquiring and losing vast sums of money. Before Sturges was a famous writer-director he was a composer of pop songs, a perfume manufacturer (his mother started the business, which at one time was fabulously successful), an inventor and mechanic who sold boat engines, and a stockbroker's clerk. It's an unlikely story of rags to riches to rags to riches to [repeat n times]. Nobody would have believed this story if it had been a movie.

5/21 Satyajit Ray, The Complete Adventures of Feluda (vol. 2) (tr. Gopa Majumdar and Chitrita Banerji)
More Feluda! The stories in this volume date from 1978 through 1994 (the last one published posthumously). Reading all of these stories in chronological order, it's interesting to see how the characters change. One of the conceits of these stories is that the publication takes place in the same world as the action, so characters in the later stories have read the earlier ones.
5/18 John Milton, Paradise Lost

Somehow I'd never gotten around to reading any of Milton's major works before now. He's well worth reading; the poetry really is majestic. My favorite parts, I think, were Adam's laments after the Fall.

I hadn't realized just how explicitly this poem was modeled on classical epics: complete with an invocation of the muse, and classical alusions and similes. References to events involving Greek gods might seem odd in such a Christian poem, but the content and style match. It's such an audacious project, reviving a classical form for Christian themes, that it ends up working. It's more Virgil than Homer, but Virgil, too, was reviving an ancient form for the needs of his time.

Lots of modern readers find Satan the most interesting and sympathetic character of Paradise Lost. To some extent I did too. I think there are three pretty obvious reasons for it. First, the poem begins from Satan's point of view. He's the protagonist for roughly the first half. Second (and again, this is a common observation), he matches the idea of a grand hero struggling against the universe—that is, the idea of a hero that we get from the Romantic era. If people once thought of Lord Byron as Satanic, it's equally true that Milton's Satan now seems Byronic to us. And third (and this is an observation I haven't seen anywhere else, but I'm sure I wasn't the first person to think of it), Satan is a psychologically complex character. He has inner struggles and doubts; he can feel compunctions, and regret, and remorse. That is, he matches the idea of a well realized character of modernist literature.

The latter two of those reasons are anachronistic, an artifact of reading Milton with the knowledge of literary conventions that postdate him by centuries. The first is not. We got a very dramatic and deliberate shift in viewpoint midway through the poem. I suppose one point was dramatizing the link between the causes and consequences of sin. Probably another was to explore the distinction between true freedom, and being ruled by one's passion. Milton's version of psychology is Aristotelian; to him, freedom requires self-government and can't exist without following one's reason. Satan is an illustration of a false version of freedom.

Famously, Milton wrote this poem "to justify the ways of God to men." Probably I'm the wrong person to evaluate Christian theodicy, but I have a lot of trouble swallowing something like this:

For man will hearken to his glozing lies,
And easily transgress the sole command,
Sole pledge of his obedience; so will fall
He and his faithless progeny. Whose fault?
Whose but his own? Ingrate, he had of me
All he could have; I made him just and right,
Sufficient to have stood, though free to fall.

It comes across as petty blame-shifting, like Adam squabbling with Eve after the Fall and blaming her for everything. But there we're hearing the words of fallen man, so pettiness is to be expected. "It's not my fault" seems beneath anything I would describe as omnipotence and omnibenevolence.

Fundamentally, I suppose what I can't accept is the central moral narrative of Paradise Lost: original sin, the perfect sacrifice, and the salvation by grace of those who accept it. And when I say I don't accept it, I don't just mean I don't think it's a factual description of the physical world; I mean I can't see it as a description of what morality and justice are. Which I suppose is another way of saying that I'm not a Christian.

5/11 Satyajit Ray, The Complete Adventures of Feluda (vol. 1) (tr. Gopa Majumdar and Chitrita Banerji)

Yes, the author is the great Bengali director, but this book has nothing to do with film. (Except for a few disparaging comments about Hindi films, and one story set in Bollywood.) This is a collection of young-adult detective stories, most of them originally published in the YA literary magazine Sandesh.

I suppose if I'd known more about Satyajit Ray, I would've been a bit less surprised to learn that this book existed. Ray came from a literary family (his grandfather founded Sandesh), and he had a longstanding interest in mystery and science fiction. He read Sherlock Holmes stories as a boy, and one of the stories collected here, "The Royal Bengal Mystery", struck me as a pretty explicit homage to "The Musgrave Ritual".

Since these are YA stories, they're explicitly didactic. The main characters have adventures all over India, and there's lots of history and geography thrown in. A good thing for me, considering how ignorant I am! There were still some words and concepts I was unfamiliar with. The translators assumed that their readers were Indians, not Americans.

4/29 Alex Martelli, Python in a Nutshell
Tutorial and reference for the Python programming language. Python targets roughly the same design space as Java, Smalltalk, and Objective-C: it's a garbage collected object-oriented language with dynamic typing, and where "everything is an object." I haven't ever used Smalltalk, but for designs where the dynamic object-oriented style is appropriate, and where maximum run-time performance isn't a major issue, I find Python cleaner than Java or Objective-C. I like its uniformity.
4/29 Laurie Marks, Earth Logic
4/23 Georgette Heyer, A Civil Contract

One of the best known Regency romances, written by the creator of the genre. A little bit unusual as romance novels go, though, at least in my limited experience. I suppose I would characterize romance as novels where the main plot is about a couple coming together, after complications that make it seem as if they couldn't or didn't want to. And this novel does fit that description, more or less, except that instead of ending with a marriage this novel begins with one. Basically, this is a book about a man coming to realize that he has underestimated his wife.

So the other half of "Regency romance" is that as well as being a romance novel it's a historical novel, set in a very particular period. Every event of the novel can be nailed down to a specific year, and in some cases you can nail it down to the exact day. (The date of Bonaparte's abdication, for example.) I thought the romance aspect of this novel was successful, the historical aspect less so. I've been trying to but my finger on exactly why. I suppose the best one-sentence answer I can give is that its portrait of early 19th century England is too filtered by the way we see that era, too little like the way people who lived then saw themselves.

The characters in this book read Jane Austen, debate the merits of Lord Byron's poetry, talk about the events of the war, attend operas by Handel Gluck. That's all authentic, of course, but in another way it seemed slightly off. After a while it started seeming like name dropping, or like an inside joke; every name that was mentioned, whether it was a politician or an author or a theater star, was someone whose name is still known. People did read Mansfield Park then, yes, but they also read less distinguished novels by Austen's now-forgotten contemporaries. It's only the filter of time that gives us the names that now stand out, and the people who lived then had no such filter.

I had the same reaction in the novel's treatment of economics and technology. As in the Hornblower books, we see the beginnings of the industrial and transportation revolutions and the beginnings of modern capitalism. The trouble is that both Heyer and Forester are focusing on the larval stages of things that later became important, but what we know from our own lives, and from contemporary novelists like Austen, and from better historical novelists writing about the same era, like Patrick O'Brian and Lev Tolstoy, is that the larval stages of a phenomenon don't seem particularly important at the time. It's just like what Stephen Jay Gould observed about the organisms found in the Burgess Shale: you can see the distant ancestors of phyla that are now extremely important, like the arthropods, but there's nothing to distinguish them from the myriad of other organisms whose phyla died out long ago.

In this book the Regency era carries too much historical weight: it's the time of transition to modernity, the time of social uncertainty because of tensions between the aristocracy and the new wealthy bourgeoisie. But to the people who lived then, it was just the world as they knew it.

4/16 Richard Morgan, Broken Angels
Sequel to Altered Carbon
4/16 Mary McCarthy, Intellectual Memoirs: New York 1936-1938

Possibly I would have appreciated this book more if I'd read more of McCarthy's other books first; I've read only two of her books before this one, and none of her novels. This was McCarthy's last book, and her friend Elizabeth Hardwick, who wrote the forward, thinks it was originally supposed to be continued. Among other things, it discusses the people and events in her life that made it into her fiction.

But it was still fascinating. Part of the book was about McCarthy's first two marriages and her various other friendships and love affairs, but most of it was about exactly what the title would suggest: the intellectual life McCarthy lived in her mid 20s. So we've got the conflict between the Stalinists and Trotskyites (which I think I understand intellectually, but which at a gut level seems about as distant to me as the Dreyfus affair), and left-wing theater (the first performance of "Waiting for Lefty", for example), and coming to grips with Eliot and Joyce, and the founding of Partisan Review.

It ended up making me envious. I don't think we have the institutions today that made it possible for that sort of public intellectual to exist.

4/16 Ursula K. Le Guin, Changing Planes

A collection of linked short stories, some of which really are stories (i.e. narrative fiction), some of which are more like anthropological notes on imaginary societies, and some of which (I'm thinking especially of "The Nha Mmoy Language") are philosophical speculation that reminds me of Borges. The framing device for these stories is flippant, but most of the stories aren't.

A few of the stories were published separately. I read "Seasons of the Ansarac" my Tiptree year, and I liked it just as much when I reread it this time.

4/10 Kyril Bonfiglioli, After You With The Pistol
The second Mortdecai book. The third hasn't yet been reissued in the US.
4/8 Hayao Miyazaki, Nausicaä Of The Valley Of The Wind 3 (tr. David Lewis and Toren Smith)
No, I didn't skip a volume; I changed editions. The "perfect collection vol. 1" that I read before contained the first two volumes of the original Japanese edition, but the Viz edition I'm reading now is in the same format as the Japanese. This means, among other things, that both the pages and the panels are read right to left. It's interesting to see how automatic the order of panels in American comics has become to me, even though I didn't think of myself as a particularly experienced comics reader.
4/5 Erich Gamma, Richard Helm, Ralph Johnson, John Vlissides, Design Patterns: Elements of Reusable Object-Oriented Software

This book was published in 1995, and I must have first read it that year; I see that my copy is a second printing, and there must have been many more printings than that the first year alone. This has been a very successful book. Lots of programmers have read it, and it's spawned paterns conferences and a whole patterns community. Wikis, among other things, came out of that community.

There are a few obvious ways in which the book is slightly dated, most of which aren't tremendously important. Lots of the examples are from GUI programs, since that's a domain area where pretty much everyone uses object-oriented programming; the specific windowing systems the authors discuss include X, NEXTSTEP, Presentation Manager—and not MS-Windows. Nowadays that would be impossible. The authors present examples in Smalltalk and C++. Nowadays I'm sure they would have dropped Smalltalk; they would have presented examples in Java, maybe in C++, and maybe in C#. If they had decided to include C++, they would probably have pointed out that the C++ community has now evolved alternative paradigms other than object-oriented programming, and they might have discussed what patterns mean outside the confines of OOP.

The more interesting question: ten years after the book that introduced patterns to software engineering, to what extent has it achieved the authors' goals? They don't pretend to have written about new algorithms or new design principles. Their goal was to discuss patterns that have been used many times before and have stood the test of time. (And they succeeded in that; I have seen all of these patterns in real code, usually for sensible reasons.) They were trying to write a catalog, to establish a common vocabulary, and to establish the idea that patterns were something worth cataloging.

I would say that in that sense the book has been a partial success. Object-oriented programmers have certainly become more conscious of structures of collaboration at a higher level than the individual class. A few of the patterns the authors named really have entered most programmers' vocabularies: Visitor, Singleton, Iterator, Factory (although most programmers use that word to include a whole spectrum of techniques that abstract creation), maybe Flyweight, and maybe Proxy. (Although again, most people use the word "proxy" in a vaguer and more general sense.) Others haven't. Possibly that's because in lots of programs the lines are just too blurry: several of the patterns in this book have enough variability, and are similar enough to other patterns, that figuring out which of the Design Patterns names to apply to a particular design wouldn't always tell anyone anything very useful. (Is this an Observer I see in my program, or a Mediator? Or maybe something in between?)

One thing I do find a bit surprising is that I have yet to see a second catalog of general-purpose patterns. I've seen a book on "antipatterns" to avoid, and a book on implementing patterns with C++ template metaprogramming, and at least one book on patterns in a particular application domain, but, in the decade since Design Patterns came out, I haven't seen any books that try to extend its catalog by discussing more general-purpose patterns. That's one of the things the authors of Design Patterns expected to come of their work, but either it hasn't happened or it just hasn't filtered out from specialized patterns conferences to the more general community of programmers.

4/4 Dashiell Hammett (ed. Steven Marcus), The Continental Op

I don't know if these stories were collected under this title during Hammett's lifetime. They were originally published in the 1920s in pulps like Black Mask.

These stories are told in the first person, and the anonymous main character is an employee of the Continental Detective Agency. (Fictitious, but I'm sure based on Pinkerton's.) It's interesting to see how different these stories are from the usual hard-boiled cliches, especially since Hammett created a lot of those cliches. The main character isn't a loner; he's a member of a large organization, and relies on its support. He's on good terms with the police. He doesn't spend much time struggling with ethical issues; he follows his employer's rules. He's tough, but other than that he couldn't be more different from Sam Spade.

I need to read a biography of Dashiell Hammet. Basically I know the three things about him that everyone knows: he once worked for Pinkerton's, he was involved with Lillian Hellmann for many years, and he was a leftist and at one time a communist. I don't know how to reconcile those things. When I think of Pinkerton's in the first decade or two of the 20th century I don't think of solving murder mysteries, but of union-busting. Not the sort of organization I'd expect your average communist to have much to do with.

3/31 Charles Stross, Iron Sunrise
Set in the same universe as Singularity Sky, with a couple of recycled characters, but it's not a sequel in any meaningful sense. It's a very different book, and so far it's my favorite of Charlie's novels. Parts of it reminded me of Vinge's A Deepness in the Sky.
3/29 Virginia Woolf, Orlando

This is the second time I've read Orlando, I think. It's a seriously strange book: lighthearted, consciously self-referential, fast-paced, modernist. I've only read two of Woolf's other novels, and this couldn't be more unlike them. The other two of her novels I've read are stream-of-consciousness explorations of a single character's inner life, and, while Orlando has some of that, it's also full of external events and adventures. Lots of the events are fantastical or self-contradictory, and I suppose it would be possible to read this book as science fiction, but that would be a stretch; it make more sense to read the absurdities as obviously absurd.

One of the most obvious things this book is about is gender. This is obviously a feminist novel, and it's easy to see that it's written by the author of "A Room of One's own". I'm thinking especially of two passages: one that makes fun of D. H. Lawrence, and one that makes fun of a couple of writers (if I'd been more familiar with early 20th century English literature I suppose I would have been able to name them) who asserted that friendship between women was impossible.

Also obviously, this is a book about transformations. Before the first hundred pages are over, Orlando has been boy, soldier, courtier, lover, ambassador, country gentleman, and frustrated author. As a physicist, when I see transformations I look for invariances. And there are some, chiefly literary: the bald man Orlando glimpsed, who might or might not have been Shakespeare, and the poem "The Oak Tree". But I wouldn't quite say that those things can be called the center of the book. It doesn't quite have a center; it's deliberately written so that it doesn't have any neat shape.

I suppose it would be possible to read this book as an allegory about the evolution of English literature. (Digression: I realized a couple days ago that, for mathematical folks, there's a very precise way to define allegory: an allegory is an isomorphism.) And I think that reading would probably work, but it's a little too ponderous. To the extent that this allegorical reading is present, it's probably best to regard it as another of Woolf's jokes.

3/27 Marvel, Silver Surfer: Communion
3/24 Dave Abrahams and Aleksey Gurtovoy, C++ Template Metaprogramming: Concepts, Tools, and Techniques from Boost and Beyond
3/19 Don Foster, Author Unknown: Tales of a Literary Detective
3/16 Sylvia Louise Engdahl, The Far Side of Evil

My storyreading group just finished reading Enchantress from the Stars, so I went ahead and read the sequel.

I first read The Far Side of Evil at something like the age it was written for: probably when I was about 12. That would mean that I read it only a few years after it was published. That would also mean that it's been at least 25 years since I last read it. But it had a strong impact on me at the time, so a lot of it stuck with me. I remembered the main character, of course, and her basic mission, and the notion of a world's "Critical Stage". I remembered the totalitarian state, and the main character's interrogation and torture. I remembered how the book ended, even if I hadn't remembered everything about how it got there.

It's odd to say this about a book that's dominated by totalitarianism, torture, and the threat of a civilization-ending nuclear war, but this is an intensely hopeful book. Its basic message is that human history isn't meaningless despite all of the horrors of the past and present, that the future can be better; humanity isn't evil, just immature. I very much want to believe that.

The "Critical Stage" in this book is the stage at which a civilization has acquired the technology that it can use either to explore space or to destroy itself by nuclear war; the premise is that those are mutually exclusive options, and that once a society has committed itself to space travel the threat of destruction is over. Even in 1971, when this book was published, that probably counted as wishful thinking. Now, 30 years later, I find it impossible to believe. I understand that the author wrote a second edition a few years ago that takes the events of the last few decades into account. It would be interesting to compare it to the original.

3/13 Hayao Miyazaki (tr. David Lewis and Toren Smith), Nausicaä of the Valley of Wind: Perfect Collection (vol 1)
I knew that Miyazaki's film was adapted from a graphic novel, but until I picked this up I hadn't noticed that Miyazaki was adapting his own work! It's a fascinating example of taking a story from one medium to another. The art in the film is quite faithful to the original comic, and so are the overall themes, but Miyazki wasn't afraid to be ruthless in simplifying the plot, eliminating some characters and societies altogether, and merging other characters together. Now I'm looking forward to seeing what he does with Howl's Moving Castle!
3/10 Scott Meyers, Effective C++: 50 specific ways to improve your programs and designs (2nd ed.)
A little old; it seems strange to see features like bool described as recent. Fortunately, there's a third edition coming.
3/6 Philip K. Dick, The Three Stigmata of Palmer Eldritch
I think maybe I've finally read enough Dick that I'm beginning to understand his books. Beginning...
3/3 Patricia Highsmith, The Talented Mr. Ripley
2/27 Warren Ellis, Darick Robertson, and Rodney Ramos, Transmetropolitan: One More Time
2/27 Jared Diamond, Collapse: How Societies Choose to Fail or Succeed

The subject of this book is just a little more specific than the title would suggest: it's not about societal collapse in general, but about collapses that involve environmental factors. Most of the book consists of case studies of past societies that died out in whole or in part (Norse Greenland; Easter Island; the Anasazi; Mangareva; high Mayan civilization), past societies that faced terrible environmental challenges but survived in a sustainable way (Inuit Greenland; Norse Iceland; Tikopia; Tokugawa Japan; the New Guinea highlands), and present societies with environmental challenges similar to those that past societies dealt with well or poorly (Australia; Haiti; the western United States; China).

One of the main lessons I got from this book is that a society can collapse at the height of its power and success, and after many years of stability. It would be correct in a sense to say that Norse Greenland was always ecologically marginal and that the settlers made decisions that could only have worked temporarily, but it would also be misleading; in this case "temporarily" meant half a millennium. And that's undoubtedly one of the reasons that it's so hard to make social and political changes that can prevent societal destruction. A "temporary" spell of good weather, like the wet years that water policy in the American west were based on, might last decades. A "temporary" policy of unsustainable resource extraction might last a century. Humans aren't used to thinking of time scales like that as temporary. If something has always worked for as long as anyone can remember, it's hard, no matter how good the evidence is, to accept the idea that it won't continue to work forever.

Probably the starkest examples in this book, and the ones that have been mentioned most in reviews, are Norse Greenland, which destroyed itself by trying to maintain a Norse lifestyle in an environment ill suited to it, and Easter Island, which destroyed itself by deforestation. One of the most widely quoted passages is a response to people who are incredulous that Easter Islanders could really have cut down the forests that their society depended on for survival:

I have often asked myself: "What did the Easter Islander who cut down the last palm tree say while he was doing it?" Like modern loggers, did he shout "Jobs, not trees!"? Or: "Technology will solve our problems, never fear, we'll find a substitute for wood"? Or: "We don't have proof that there aren't palms somewhere else on Easter, we need more research, your proposed ban on logging is premature and driven by fear-mongering"?

The lesson of this book is: it can happen here, partly because lots of decisions are perfectly rational for individuals or in the short term, but terribly destructive for society as a whole or on a longer scale. There are plenty of decisions our societies are making right now that could be as momentous for us as forest management policies were for the Easter Islanders.

Our environmental issues will be resolved, one way or another: as the truism goes, "If something can't go on forever, it will stop". But these issues may or may not be resolved in a way we'll like. Some past societies have made good decisions about environmental issues, and some have made bad decisions. The costs of bad decisions have sometimes been catastrophic.

2/21 Andrew W. Appel with Jens Palsberg, Modern Compiler Implementation in Java (second edition)

The first edition of this book came in two flavors, ML and Java. Only the Java flavor got a second edition.

This is a survey book, an introduction to compilers. It has two parts, "Fundamentals of Compilation" and "Advanced Topics". Part I would do nicely as a textbook for a one-semester undergraduate compiler course; by the end of the course the students would essentially have written a compiler for a toy language, one at roughly the level of sophistication of the Dragon Book or pre-3.0 versions of GCC. This part of the book is organized very straightforwardly: one chapter on lexing, one on parsing, one on instruction selection, one on register analysis, and so on.

So what's "modern" about this"? Well, one answer is in the other half of the title, the "in Java" part. Much of the discussion of implementation techniques is in terms of OOP: using base classes to hide the difference between different target architectures, using the Visitor pattern for tree walking during some phases of analysis and translation in the front end, and so on. I found some of those uses of OOP more compelling than others; I'm not convinced that the Visitor stuff would survive contact with a real-world front end that was actively maintained over a period of years.

In any case, the other answer to what makes this book "modern" is Part II. That, too, is a survey, but the vast majority of topics in it are ones that the Dragon Book doesn't touch: garbage collection, compiling functional and OO languages, and advanced optimization topics like SSA form, software pipelining, and cache-based optimizations. In this part of the book the Java implementation largely disappears; algorithms here are are presented as graphs or pseudocode.

One puzzle: I know where to go if I want a book that gives a deeper view of modern optimization techniques. (I'd probably start with Muchnick's Advanced Compiler Design and Implementation). But what about the poor people who write compiler front ends? Where are the books that go into parsing and semantic analysis at more than an introductory undergraduate level?

2/19 Iain M. Banks, The Algebraist

Space opera on the grand scale: thousand-ship interstellar invasion fleets, billion-year-old secrets, galactic empires and semi-empires, a backstory that includes an anti-AI war and something a lot like Uplift. This book makes it clear that the sizes and times we're dealing with are beyond our imaginations.

This isn't a Culture book, but it does have an interesting echo of the Culture: one society in this book, the Dwellers, is another take on what a post-scarcity civilization might look like. As with the Culture, outsiders find them incomprehensible and/or decadent. But post-scarcity is about the only point of similarity. The Dwellers are truly ancient; they're aloof, not evangelistic; they have no Minds and no Contact; they're not human in any sense. Their version of a post-scarcity society reminded me more of Cory Doctorow's Down and Out in the Magic Kingdom than of Banks's Culture books.

2/18 Elizabeth A. Lynn, Dragon's Winter
2/15 Philip K. Dick, VALIS

Part of my project of trying to learn Dick better in time for Potlatch 14.

I definitely don't understand this book. But then, understanding probably isn't the point. Partly it's about what it's like to have an incomprehensible transcendent experience, partly it's about what it's like to be insane and think you've had a transcendent experience, partly it's the exposition of a version of mysticism, one that's a sort of syncretic version of lots of different mystical and esoteric traditions. (Gnostic Christianity! The Dogons! Buddhism! Parsifal! Three-eyed aliens from Arcturus!) It's also about a dead cat.

What I find especially fascinating is the idea that one can be insane, and know it, and and also think that you might have learned something real and important from a mystical experience.

2/12 Herman Melville, Moby-Dick
1/22 Pauline Kael, I Lost It at the Movies: Film Writings 1954-1965

The bulk of this book is film reviews—not, to my surprise, from The New Yorker. At this point in her career Pauline Kael was living in the Bay Area, and many of the reviews in this book were originally broadcast on KPFA. There are also a few extended essays on what film criticism should mean, some of which are written in the form of justifiably negative reviews of film theory books.

Kael was a smart reviewer. She pointed out things I hadn't noticed in movies I'd seen, and made me aware of other movies I haven't seen and should. (Meaning largely foreign and art-house films. If she thought a mass-market film was junk, she usually just ignored it. She only wrote negative reviews for films like La Dolce Vita and Hiroshima Mon Amour, films that she thought had an undeserved reputation for being intellectually deep.)

I found this book completely fascinating as a period piece, irrespective of what Kael had to say about movies. She lived in a different world than the one I do, one in which it was perfectly natural to be a public intellectual writing for other intellectual but not academics. She's writing in a context when she can expect her audience to be conscious of modern sexual research (she describes a particular line in a movie as making a point that referred specifically to the Kinsey report), when she can expect her audience to have read Freud and to be generally sympathetic toward the anticommunist left.

1/16 Terry Pratchett, Small Gods
More Discworld
1/15 Jamie O'Neill, At Swim, Two Boys
A novel: gay Irish youth around the time of the 1916 Easter rising.
1/8 Francis Glassborow with Roberta Allen, You Can Do It!: A Beginner's Introduction to Computer Programming

One of only two introductory programming books I've seen that teach C++ to people who have never done any programming before. (The only other one I've seen is Andy Koenig and Barbara Moo's Accelerated C++.) The author information isn't because Francis hired a ghostwriter, it's because he had a student! Roberta Allen, who had never done any programming before, was a β reader; she worked through all the material in this book and made extensive changes.

Francis deliberately uses very simple C++ in this book (he doesn't mention inheritance or virtual functions; he shows how to use templates, but not how to define them), because his goal isn't to teach C++ but to teach beginning programming. I'm sure that's the right choice.

He also makes another interesting choice: almost all of the programs in the book, starting with the very first one, are graphical. (The programs use a simple graphics library, provided on a CD, so that students don't have to worry about all the overhead of event-driving GUI programming in a complicated system like Win32 or Carbon or Gnome.) I'll have to think about this a bit more. Something along these lines is certainly the right pedagogical choice, and Bjarne Stroustrup is doing something similar in his introductory classes. Programs that do nothing but read and write from the console just aren't interesting or realistic anymore; it's not an important form of I/O these days. It's the details that I'm less sure of. We don't want to be teaching about event loops on the first day of class, but I'm not sure if the model of graphics in this book is the best possible one either.

1/4 Robert McChesney, The Problem of the Media
Media criticism. I didn't learn much from this book that I didn't already know from Bagdikian, I'm afraid. Mostly, what I learned is that I wanted to read a different book: one about the change in newspaper ownership patterns in the last half of the 19th century and the first decade or two of the 20th century. I knew that there were important changes in that period, in technology, in government media policy, and in the economy as a whole, but some of the details hadn't occurred to me before. Now I'd like to read a close study.

(Dates are the date I finished reading the book)


Matt Austern