Three-Toed Sloth

Books to Read While the Algae Grow in Your Fur, July 2006

Attention conservation notice: I have no taste.

Joel Best, Flavor of the Month: Why Smart People Fall for Fads

An attempt at a sociological description and explanation of institutional fads, meaning things more like phonics or total quality management than, say, Rubik's Cube, or even blogging. Short version: our institutions always have problems, some of them real and remediable, and some of them simply because we can imagine them doing much better than they are, and believe in the perfectability of social arrangements. (E.g., he argues that pretty much any way of teaching children to read will always lead to some who aren't "up to level", and so create an audience for some new way of teaching them to read.) On top of this, there are various structural biases in media and social networks which will tend to propagate news of new ideas and their successes more more efficiently than failures or grounds for skepticism. (At some points he seems to be saying that some of these things are unique to America, or at least more pronounced here than in other industrialized countires, but of course presents no comparative data that we are more prone to institutional fads than, say, the French or the Japanese, or even than the Canadians.) Worth reading, which can be done in a weekend.

Naomi Novik, Black Powder War

Mind candy. Our heroes (see previous installments) cross Asia by the old silk road, only to help lose the Battle of Jena. (I refuse to regard this as a spoiler.) Continues to provide candyish satisfaction. Many hints are dropped as to plot developments in sequels; I hope Novik has the strength of auctorial character to resist following all of them up. — Sequel.

Andrea Camilleri, The Shape of Water and The Snack Thief

Mind candy. See earlier remarks on Camilleri. The series grows on me the more I read; fortunately for my productivity not that many books have been translated.

Clark Glymour, The Mind's Arrows: Bayes Nets and Graphical Causal Models in Psychology

What graphical causal models are, why they are such a good way of representing causal structure, why they might be good ways of representing causal knowledge, and ways various parts of psychology could benefit from using them. Much of Part III, on inferring mental architecture from lesions in neuropsychology, could almost equally well apply to functional imaging studies (except that lesion work is better grounded). Glymour spares few opportunities to point out just what a horrible idea it is to use linear regression and factor analysis for causal inference, concluding with a chapter on The Bell Curve. On the one hand, Glymour shows that, using the kind of evidence Herrnstein and Murray do, it is not possible to draw any reliable conclusion at all about the causal relations between IQ and social outcomes. On the other hand, an awful lot of mainstream social science uses the same kind of data and the same kind of method... That chapter in particular will make little sense to readers who haven't, at the least, forgotten how to calculate factor loadings; but I am not sure the argument could be honestly made in any more popular way, and I am impressed by how far Glymour has gone to simplify his rather technical work in this area.

On the down side, there are some really awful puns.

Kenneth Pomeranz, The Great Divergence: China, Europe, and the Making of the Modern World Economy

Gets its own review: King Cotton and King Coal Raise the West.

H. R. Trevor-Roper, The Hermit of Peking: The Hidden Life of Sir Edmund Backhouse

Droll and engrossing historical detection on a massively eccentric and bizarre literary (and commercial and espionage) fraud artist. I quote an early passage describing Backhouse's memoirs, to which John Burke drew my attention, to give some flavor of the whole: "I had not read far before I realized why the Swiss custodians of these volumes had preferred not to entrust them to the post. How, I asked myself, would a right-minded and conscientious customs officer react if he were to open and read these works? The text would surely be confiscated, and perhaps the law would inconveniently take note of the sender and the addressee. For the volumes were of no ordinary obscenity." (As John says, the word "ordinary" is a wonderful touch here.) — But this was in 1973; standards have, to say the least, changed somewhat, and I wonder if anyone has thought to publish Backhouse's memoirs, perhaps in conjunction with some queer/post-colonial effort at rehabilitation?

Alexandre J. Chorin and Ole H. Hald, Stochastic Tools in Mathematics and Science

A short (two-hundred-odd), highly non-rigorous, quick-and-dirty introduction to stochastic methods for applied mathematics, starting with basic probability and expectation, and ending up with some nice non-equilibrium statistical mechanics. (They do however assume quite substantial familiarity with linear algebra, operators, etc.) I am enough of a probability geek, at this point, to wish that they had done things a bit more rigorously and a bit more abstractly in places, because I think it would help the reader see, e.g., the Ornstein-Uhlenbeck process as part of a general pattern of stochastic ordinary differential equations, rather than a tricksy special case. Also, there are more places than I'd like where they don't lay out the motivation clearly, and a self-studying reader is going to say "wait, where is this going exactly?" — not so much of a problem in a classroom setting, which is where these notes began. I'd be very happy to teach from this text.

Kirk Mitchell, Sky Woman Falling and Dance of the Thunder Dogs

Mind candy. Latest (and possibly last) in a series of police procedurals set in Indian Country, broadly construed. Reasonably addictive; the earlier installments (Cry Dance, Spirit Sickness, Ancient Ones) are in some ways even grimmer, because they focus more on the complicated relationship of the detectives.

Books to Read While the Algae Grow in Your Fur