September 30, 2007

Books to Read While the Algae Grow in Your Fur, September 2007

Jonathan Chait, The Big Con: The True Story of How Washington Got Hoodwinked and Hijacked by Crackpot Economics
There is nothing here which should surprise anyone who's read Hacker and Pierson two years ago, or Krugman four years ago, or John Judis seven years ago, or Michael Lind eleven years ago, or even just kept their eyes open over that period.
But these truths need to be hammered at again and again, until they are driven securely and irrevocably into the frame of the national debate. So, to repeat: the modern Republican Party is a paranoid, anti-democratic sect, one of whose primary objectives is to help really, really rich and powerful people become even more rich and powerful, and what they say about why this will help the rest of us is some much bullshit, invented by cranks and peddled by a massive infrastructure of lying. It is simply untrue that the Democrats are equally radical (I wish they were, Chait very much does not). The institutions — in Congress and the executive branch, within the parties, in the media and in the country at large — which are supposed to prevent extremists from implementing a radical, unpopular, and in fact crazy agenda are all failing us. This does not speak well for either our institutions or, in aggregate, our character. It is this completely true story of greed, deceit and culpable credulity that Chait tells, and he tells it very well. Because it omits bigotry, fear and war, it is not the complete story of our politics, but it's a brief, easy, persuasive read, and I hope it does a lot of good.
(The only error of fact I could find was in the discussion of the Senate: in showing how unrepresentative it is, Chait points out that the number of voters per senator in Wyoming is vastly smaller than in California; unfortunately a decimal place must've slipped, because the real ratio is 1:70 against California, not a mere 1:7.)
I. J. Parker, Island of Exiles
Our Hero experiences life on the lowest possible rungs of Heian-era Japan's social ladder (Judge Dee never dreamed of enduring such indignities in the course of upholding the Confucian Way), while investigating the murder of an exiled prince. The climax is worthy of Kurosawa. While the latest book published in the series, narratively it falls between The Black Arrow and The Hell Screen.
Zellig S. Harris, Language and Information
Harris — see Wikipedia, or this shrine — was a major figure in the later part of structuralist linguistics, and had a big influence on the development of the field, not least through a certain student who has over-shadowed him. Having now read him, I feel like the relative eclipse was unfortunate.
I am naturally pre-disposed to anyone whose approach to linguistics is to regard it as a massive exercise in statistical language learning, especially if they approach it as a matter of discovering hierarchical structure by repeatedly testing for violations of conditional independence assumptions. If their account of grammar itself has a certain algebraic flavor, well, I can like that too, even if he makes it seem more self-referential than it needs to be. And to see a general theory of linguistics put to work on non-trivial real-world problems, like trying to understand how scientific papers in a particular field actually work and evolve, and getting results, is very nice.
All of which said, this was also a somewhat frustrating book to read.
I was never able to get clear in my head exactly Harris meant by "information". I think it was something close to, but not quite the same as, what an information-theorist means by it, but exactly where the difference lies, or how to formalize his notion, I could not tell you. Since "information" is half the subject matter of the book, I feel he could have taken the trouble to be explicit.
The other frustrating part was this: It's plain that Harris was a forbiddingly smart guy, with a deep knowledge of language who had thought long and hard about these matters. It is also rarely clear, when he makes an unsupported assertion, whether this is because (1) he regards the supporting argument as trivially obvious from what has gone before, (2) he has given a supporting argument in a more technical work, or (3) he is merely sure of himself. (Since at least some the more prominent of his intellectual progeny share this trait, as shown e.g. here, we have a nice question of selection versus influence: did they learn to write like this because of they're in that lineage, or did they join that lineage because of the things which lead them to write like this?) It would have been easy enough to fix this, too.
Added to my to-read pile on account of this post by Fernando Pereira, which made it sound like someone I know was re-discovering some of the same ideas in the course of their own work on analyzing networks of scientific papers. (They should publish!)
Walter Bagehot, Physics and Politics: Thoughts on the Application of the Principles of "Natural Selection" and "Inheritance" to Political Society [PDF at McMaster, plain text at Project Gutenberg]
A query of Mark Liberman's over on Language Log prompted me to finally take this book off my shelf and read it, a mere eight years after I bought it at a library booksale. (It is now on-line, so I will be donating my copy to the nearest library, unless a local reader wants it.) Since it was published in 1867, a few years' delay in reading hardly seems to matter.
The reading itself is a very curious experience. This is, so far as I can tell, the first attempt to apply genuinely Darwinian, that is, selectionist, thinking to social development. It has interesting, and indeed sound, observations on how selectionism can be implemented through the imitation of those who are successful, which anticipate a large chunk of the work in evolutionary game theory over the last few decades. It has, in places, the analogy between an idea, or practice, as an object of selection and a contagion, which Dawkins made famous under the catchy label of "memes". It has the basic observation that human groups whose customs and traditions enhance social power, especially its more coercive forms, will tend to expand at the expense of those whose traditions do not enhance social power, which has been spun into a whole (not-bad) book on The Parable of the Tribes. Underlying it all is a defense, or perhaps better embrace, of classical liberalism in a form which perhaps only the editor of The Economist in the 1860s could provide, and an attempt to under-write it by means of this evolutionary theory of society and culture (a term he does not use, at least not in the modern sense which was about to be invented by anthropologists). And I like liberal evolutionary naturalism, I (pretty much) believe it; this book is the ancestor of things like Nonzero and Cultural Software (the latter via William James and American Pragmatism), and at least close to the lineages of The Social Animal, Guns, Germs and Steel and even The Open Society and Its Enemies.
It is also horribly, mind-bendingly racist. What is most striking about reading it — what strikes me whenever I read the really respectable Victorians — is how obvious racism seemed, how much it was something taken for granted. For instance, the following, offered without any support whatsoever (and which someone like Amartya Sen should use as an epigraph, if he hasn't already): "To offer the Bengalese a free constitution, and to expect them to work one, would be the maximum of human folly". This, mind you, is in the context of arguing that "There then must be something else besides Aryan descent which is necessary to fit men for discussion and train them for liberty". (Less fallaciously, he also points to the many Phoenician republics of antiquity, most notably Carthage.) A huge portion of the book is devoted to the problem of explaining why non-European societies were static and arrested, a problem which goes away once your knowledge of the history of those societies is not, in fact, a load of horribly self-serving rubbish. (For instance, the history of democracy in ancient India.) So while I can comfortably look down on Bagehot for this sort of thing, thereby reaffirming the reality of the moral progress in which he himself believed, I am left uneasily wondering about what of the things I accept unquestioningly are also wrong, and as repulsive as Bagehot's thoughtless racism.
(The stuff about the genetic inheritance of acquired characteristics, on the other hand, doesn't bug me at all, because it's a simple error of relying on then-current science, and actually not needed for his arguments, which work equally well through cultural inheritance.)
Carrie Vaughn, Kitty and the Midnight Hour
Mind-candy, well-suited to improving a beautiful late-summer weekend marred by a miserable cold.
Conjecture: The "contemporary fantasy" sub-genre, of werewolves, vampires, etc., etc., trying to lead more or less ordinary lives in early-21st-century America, is, in some sense, a reflection of the recent, and rapid, relative acceptance of people whose sexual preferences aren't straight-and-vanilla. Query: How could we make "reflection" precise here? Query: how on Earth could we test this idea? Query: Should this idea be written off to cold medicine?
— Sequels: 2, 3, 4, 5, 6, 7.

Books to Read While the Algae Grow in Your Fur

Posted at September 30, 2007 23:59 | permanent link

Three-Toed Sloth