March 31, 2012

Books to Read While the Algae Grow in Your Fur, March 2012

Attention conservation notice: I have no taste.

Elizabeth Bear, Range of Ghosts
Suppose that a very good fantasy novelist — someone who gets the attraction of heroic fantasy at a bone-deep level, and has a core a conviction that everything valuable involves a painful sacrifice (with luck, only proportionately painful) — decided to dig back to, or through, the roots of the genre in writers like Tolkien, Howard and Leiber. Suppose then that she also took inspiration from the medieval history of Central Asia, and especially from books like Grosset and The Tibetan Empire in Central Asia and Beckwith. We might then find ourselves with a fantasy trilogy opening with a war between the sons and the grandsons of the Great Khagan for dominion over the empire of the steppes. If we are fortunate (and we are), this will go on with: shamanistic sorcerers trained in isolated mountain kingdoms; hungry ghosts; butterflies of ill omen; the peculiar beauty of fertile valleys in the high desert and their towns; rocs and their taming; Baluchitherium; death cults building their fortresses on desert mesas; bipedal tiger-demons; haunted kurgans; lunar yet vivid landscapes; necromancers drawing all the peoples of the world into war; hair-breadth escapes; painful wounds; plausible dynastic politics; wrenching choices; remarkable horses; yurts; heroes who have learned "to speak the truth and to handle bow and arrow well"; many very different big skies; polyandry; colored salt from the roof of the world; a Song monk wandering the western lands on a mysterious errand; and (spoiler follows) n phefrq evat bs tbyq juvpu gheaf vgf jrnere vaivfvoyr.
I realize that some of the buttons this pushes for me are rather arcane, but honestly it has been years since I read a novel in this genre with the same enjoyment and the same "but how does it the story go on?" feeling at the end.
(More spoilers: V guvax gur Pneevba Xvat vf zber be yrff Na Yhfuna/Ebuxfuna, juvyr Qhancngv vf Nggvyn, ohg V nz ernyyl abg fher nobhg rvgure vqragvsvpngvba, naq vg'f n avpr dhrfgvba jung rvgure jbhyq zrna. Naq: Grzhe, lbh gjb-gvzvat pnq.)
Author's self-presentations: 1, 2, 3
Orhan Pamuk, My Name Is Red
A re-read; fortunately I'd forgotten the solution to the murder mystery, which of course is not the point. The point is rather: art, memory, ambition, longing, melancholy, transience, eternity, tradition, style, individuality, imperfection, perspective, vision, blindness. And then of course there are the games following from a novel about the Ottoman heirs of the Old Masters of Herat trying to learn the methods of the "Frankish and Venetian masters", being written by a Turk who has obviously mastered the methods of masters like Calvino. (It is not clear to me if the story-within-a-story technique is really a nod to the 1001 Nights, or rather to The Castle of Crossed Destinies and Invisible Cities.)
I rather disliked the first book of Pamuk's I read, but this is wonderful, and you should go read it if you haven't.
Ideally, however, there would be an illuminated edition.
(The one thing I would change, and I realize this is petty so I put it in the end in small type, is the disquisition about the meaning of oral sex, which is not at all indecent but just inadvertently funny, like the worst bits of Updike.)
Clark Ashton Smith, Zothique
Mind candy. Smith was a talented early fantasy writer, overlapping with science fiction and horror, of the same vintage as Lovecraft, largely remembered, nowadays, as the latter's friend. This is a bit unfair. Smith didn't have the same power of vision that Lovecraft did, but he was a much better story-teller, and an actually-good stylist. (Smith was also much more pervy and directly influenced by turn-of-the-century decadence, which may be a feature or a bug, depending.) Zothique is a collection, edited by Lin Carter, of Smith's fantasy stories set in the far future, named after Earth's last continent, when the sun has dimmed and weird magics haunt a world full of ruins. If they weren't a direct inspiration for Vance's The Dying Earth (and so much else), I will spend a night reciting "The Empire of the Necromancers" in a Pittsburgh cemetery. This are good stories of their kind (ObDisclaimer: casually racist and misogynist author was casually racist and misogynist), but the truth is, Vance was just better.
This particular anthology is long out of print, but his complete stories are online.
I see that I bought my copy from Moe's Books in Berkeley on 23 April 1993; I am indeed a sloth.
Martha Wells, The Cloud Roads and The Serpent Sea
Mind candy, but, like all of Wells's books, very high quality mind candy. These are romances of caste and ecology among the social lizards. (More exactly, shape-shifting social lizard-men, who mercifully owe nothing to Anunnaki mythology.) I gather that there will be a third book in the series in 2013.
ObLinkage: Author's self-presentation for The Cloud Roads
Sequel, apparently concluding the series.
Marius Iosifescu and Serban Grigorescu, Dependence with Complete Connections and Its Applications
Full-length review: Memories Fading to Infinity
— I used to joke that my nightmares included giving a talk and having an member of the audience announce at the end, in a thick Eastern European accent, that everything I'd said was a special case of a theorem his adviser had published in the Proceedings of the Academy of Agro-Technical Sciences of Outer Yajanistan in 1962. (I am under no illusions about being funny.) Reading Dependence with Complete Connections is a bit like wandering into that joke. I have been dealing with chains with complete connections since my first paper, though for most of the time I didn't realize that's what they were. I can salve my pride by saying that the problems I'm interested in (e.g.: given a stochastic process find the smallest random system with complete connections which generates it, for a particular value of "small") are not the ones solved by the old masters of Bucharest. But when I think of all the times I've said "they're like HMMs, only not exactly", I feel very low.
Tony Judt with Timothy Snyder, Thinking the Twentieth Century
Partly Judt's autobiography, partly Judt and Snyder conversing about the intellectual history of the twentieth century, intellectuals in politics, and about how intellectuals ought to behave, and just how far that has been from our actual conduct. At least as presented here, Snyder's main contribution to the book was making it possible at all — a truly moving story, so I will just refer to Thinking the Twentieth Century as Judt's.
Little in this book will surprise those who have read Judt's previous books, especially Postwar and Reappraisals, but this is rather more concentrated and systematic than his other works, and perhaps more accessible than the vast Postwar. As I've said before, I find a lot of Judt's views sympathetic and generally well-argued, and his prescriptive ideas very attractive. I'm glad we have this.
But, following a proud tradition, I am going to mostly quibble with a minor point, or rather some absences. It was very striking this time just how Eurocentric and literary-ideological Judt's perspective on intellectual life was. Intellectuals are authors from Europe or North America who write on politics, morals, or the arts. Anything about the natural world, technology, or math is right out. (There is a partial exception as to math in favor of economists, but even then the only two treated at length, or even I think by name, are Keynes and Hayek, who are, of course, unusually non-mathematical economists. [Actually, the discussion of how Hayek's road-to-serfdom ideas relate to Austrian politics in the '30s is very interesting.]) The image of a Central European thinker with massive influence from the middle of the 20th century onwards is Martin Heidegger*, not John von Neumann; the image of an intellectual in politics in Léon Blum, not Jawaharlal Nehru. (Even if one only cares about intellectuals as ideologists, von Neumann has a lot of claim to our attention.) It is not at all obvious that this is the best way to look at the century which saw the dissolution of the European empires, or the enterprise of science and technology assuming such vast size and consequence. One could defend these choices of perspectives as selections ("I am interested in this corner of the whole panorama of human intellectual life") or as judgments ("this is the most important history, for such-and-such reasons"), but I don't think Judt (or Snyder) realizes that they are choices.
(I will say nothing about Judt's pronouncements on American feminism in the last chapter, out of respect for his memory.)
*: In saying this, I don't mean for a moment to suggest that Judt agrees with Heidegger about, well, anything.
D. R. Cox and Christl A. Donnelly, Principles of Applied Statistics
Full-length review, originally in American Scientist.
Shorter me: There are two great traditions of applied statistics. One is what we now call "data mining" or (distastefully) "data science". The other is aims at solving scientific problems, and is what Cox has been contributing to for longer than most working statisticians have been alive. Short of apprenticing oneself to a master of the art for a few years, there is no better introduction to how one translates between scientific questions and statistical problems.
Nota bene, D. R. Cox, the eminent real-world statistician, is not to be confused with the fictional Dr. Cox, despite what some search engines might suggest. Indeed, so far as I know, no one has ever suggested that the actual Prof. Cox is a "bastard-coated bastard with bastard filling".
Amanda Downum, The Kingdoms of Dust
Mind candy. Suppose that Lovecraft's "Colour out of Space" had been what happened to Iram, City of Pillars — what then? (Previous adventures of our heroine.)
Jack Knight and James Johnson, The Priority of Democracy: Political Consequences of Pragmatism
Full-length review: Dissent is the Health of the Democratic State
Shorter me: This is actually a very deep book about democracy, and why, exactly, it is so awesome, but it's written so obscurely it will have no impact, which is a shame.
Seanan McGuire, Discount Armageddon
Mind candy. Between the community of monsters living among us in the big city, the generations-old organization dealing with same, the heroine's day-job as a scantily-clad waitress at a dubiously-themed bar, and (most of all) the chapter headings, if this isn't descended from (high quality) Middleman fanfic, there's been a lot of lateral genetic transfer. This, to be clear, is a good thing.
ObLinkage: Author's self-presentation.
Lauren Willig, The Garden Intrigue
Mind candy. By this point my commitment to the series and characters is slightly disturbing.
Franklin E. Zimring, The Great American Crime Decline
A nicely written summary of the established facts about the huge and enduring decline in crime in America during the 1990s, as well as just how little we understand about its causes. (Zimring has some malicious, but entirely justified, fun at the expense of those who, in the early and even mid 1990s, confidently predicted a massive crime surge*.) The "usual suspects" --- demographics, unemployment, and locking people up** --- all pointed towards a decline in the crime rate, but nobody, before it happened, would have been able to predict the massive scale of the decline. Even retrospectively, it is hard to make these account for more than about half of the decline***. Causal theories advanced after the decline had become obvious have obvious selection-bias problems, and when one tries to cross-check them by looking at what they would imply for phenomena other than national crime totals, the results are not happy****. (I am a little surprised that no one has tried to argue for the benevolent influences of e-mail, first-person shooters and hypertext.) Going over these facts and the evidence for and against putative causes involves a lot of examination of data and methodological criticism, but Zimring is good at conveying this clearly.
Zimring puts a lot of stress on two aspects of the crime decline. First, crime rates fell much more in New York city than in the rest of the country --- roughly twice as much. Something had to be very different about New York, even compared to other large cities in the US. What is still stranger is that, as Zimring says, in many ways New York was much the same city in 2000 has it had been in 1990, when it had several times as much serious and violent crime — there had been no vast social or moral change.
Second, and I find these even more interesting than the stuff about New York, the crime decline in the US was paralleled by a crime decline in Canada of similar timing and magnitude — but not in other developed countries. Yet Canada had no massive surge of incarceration, no huge expansion of policing, not even the same sort of economic boom in the 1990s... In fact, the most astonishing thing for me in the whole book is Figure 5.23 on p. 132, showing US and Canadian murder rates tracking each other with eerie precision from 1961 to 2002. Unless this is the consequence of a massive exercise in juking the stats, no explanation which focuses on causes only working in either country can be very plausible.
The Great American Crime Decline was published in 2006. I'd love to read an update, but even so I learned a great deal from it, and recommend it.
(Thanks to M. R. for telling me about this book and lending me her copy.)
*: Considering the prominent place in this scare-mongering of scholars like James Q. Wilson and John "Super-predators" DiIulio, there is an interesting essay to be written about the selective skepticism of conservatives and neo-conservatives towards social science forecasting.
**: Except, of course, for the crimes those in prison perpetrate on each other, as they notoriously do.
***: The major demographic variable here is the number of adolescents and young adults, especially the number of teenage boys and young men. As Zimring nicely puts it, however, the fact that young males are always disproportionately likely to be criminals doesn't help us predict the number of crimes from the number of young males. To do that, we would have to know not just the crime rates among different demographic groups, but also be able to extrapolate those rates into the future. If anyone has figured out how to do that, they're not telling.
***: Zimring also has some nice, tart examples here of "the cross-sterilization of the social sciences" (a phrase he attributes to an unnamed judge), especially when it comes to economists — i.e., Steve "Freakonomics" Levitt — writing about crime and demographics.

Books to Read While the Algae Grow in Your Fur; Scientifiction and Fantastica; Enigmas of Chance; The Progressive Forces Central Asia; The Collective Use and Evolution of Concepts; The Commonwealth of Letters; Writing for Antiquity; Commit a Social Science; Pleasures of Detection, Portraits of Crime

Posted by crshalizi at March 31, 2012 23:59 | permanent link

Three-Toed Sloth:   Hosted, but not endorsed, by the Center for the Study of Complex Systems