Three-Toed Sloth

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

C. J. Sansom, Dissolution

A hunchbacked lawyer shuts down an English monastery for Thomas Cromwell and Henry VIII; soul-searching and homicide ensue. There is an amusing, but mericfully passing, allusion to The Name of the Rose.

Yuji Ijiri and Herbert Simon, Skew Distributions and the Sizes of Business Firms

One of the depressing aspects of the power laws project, for me, was that it proded me into finally reading this collection of papers by Herbert Simon and collaborators. By now, everyone interested in the subject is (or should be) aware that the "preferential attachment" mechanism for generating power law networks is a re-discovery of a 1955 result by Simon (building on earlier work by Yule), showing that distributions with power-law tails are produced by certain kinds of multiplicative growth processes obeying a weak form of Gibrat's law. What is more depressing is to see: how Simon was able to modify his basic model to produce other log-normals, cut-off power laws, and other heavy-tailed distributions; how he could handle correlated growth; how he developed the connection to Bose-Einstein statistics (no I am not exaggerating); etc. Strictly speaking, Simon generally works with what he calls a Yule distribution, and some people call a Yule-Simon, which approximates a power law in its tail; his growth models don't seem to produce pure power laws, but neither do the data support pure power laws.

The really depressing aspect is that Simon realized the thing to do wasn't just to draw straight lines, or even particular kinds of curves, on log-log plots, but to see where the data disagreed with the model's predictions and to use that to learn about the mechanisms at work, especially by checking their other predictions and assumptions. Shockingly little of this has been done over the last thirty years.

I used to think that when I described complex systems as a series of footnotes to Herbert Simon, I was joking, but now I wonder...

Warren Ellis, Chris Sprouse and Karl Story, Ocean

A deliberate hybrid of 2010 (a space station named "Clarke's Walk", no less!) and Revelation Space, though with fewer squicks than the latter.

Warren Ellis and Ben Templesmith, Fell, vol. 1, Feral City

Collection of the first eight issues of the comic book; each issue was deliberately designed to work as a self-contained story, and they do, but they also nicely integrate into a cohesive tale, about unhappy homicide detective Richard Fell trying to maintain order in the "feral city" of Snowtown. (I wonder if that name, and the bit about the well in the first issue/chapter, are a reference to John Snow and the Broad Street pump?) The protagonist is transferred to Snowtown, under an unspecified cloud, at the beginning of the story; we get to discover it along with him. Readers of Ellis's weblog will remember that many of the crimes here — especially the most grusome and implausible-sounding ones — are culled from actual news reports.

Homicide seems like a definite influence; there's something of Frank Pembleton (pre-stroke) in Fell. This does not seem to help explain why Richard Nixon is repeatedly sighted wandering around Snowtown in a nun's habit, however.

Andrea Camilleri, The Smell of the Night

Ponzi schemes — not called that — gone sour. Thanks to "Uncle Jan" for the gift.

John Mueller, Overblown: How Politicians and the Terrorism Industry Inflate National Security Threats, and Why We Believe Them

Brief comment: Word.

Longer comments: Terrorism, Mueller says, is not an existential threat to our civilization or our societies or polities, which are very, very strong. A much worse threat to our freedoms is in fact over-reaction to terrorism, and the (ridiculous, cowardly) impulse to advocate authoritarianism and the abandonment of the rule of law in the name of "security". The readiness of so many supposedly serious people in positions of authority to advocate this is far more worrying to Mueller, and I must say to me, than is terrorism itself. If, in the incredibly implausible worst-case scenario, someone manages to nuke the Capitol, the White House and the Pentagon, that would be an incredible disaster and a tragedy. One should, obviously, take reasonable steps to prevent that. (Let me add that many of my family works around there, so I'm not one to treat this lightly.) But I utterly fail to see how that would be in any way helped by abandoning the Constitution, much less how that would threaten the survival of the United States of America...

There isn't enough, for me, on the political economy of the situation here. Mueller hints at, but doesn't really explore, a notion which sometimes seems only too plausible to me, that we have created a permanent class of defense/intelligence/security contractors, linked in very hard-to-unravel ways to military and political figures, all of whom look for threats to justify themselves (Mr. Cunningham and friends beyond only a particularly flagrant example), creating a mess which it is hard for the ordinary political process to fix.

Diarmaid MacCulloch, The Reformation: A History

Vast, global history of how Latin Christendom came to tear itself apart into so many distinct, mutually-intolerant chunks, going from the late 1400s through about 1700. It's well-written, and does a very good job of explaining just enough of the theological issues at stake — and I use the word advisedly — to make the disputes comprehensible. One nice feature is the sustained attention to the reformations in eastern Europe (Poland-Lithuania, Hungary, Transylvania...), part of a broader sense on MacCulloch's part of it-could-have-gone-otherwise. Another running theme — often no more than a sub-text — is that almost all the major participants on all sides were viciously intolerant theocrats, who would have regarded freedom of conscience, if it had been explained to them, as a flat invitation to heresy, and a policy promoting it as the civil power failing in its God-given duties, probably with Satanic inspiration. Modern ideals of liberty grew out of the locales where the different strands of Reformation and Counter-Reformation fought each other to draws, and so failed — places like, for a while, some of the eastern European commonwealths, and more permanently and momentously, Amsterdam and London. The modern world was born from the 17th century abandonment of the ideal of "Christendom", in which all full members of society belong to the same Church, whose writ is enforced by the State which it in turn legitimates.

(MacCulloch is too good a historian to waste time on silly analogies that say that "what Islam really needs is a Reformation", but I'm not a historian so I'll bite. If you take this seriously, then what you are prescribing is that multiple transnational ideological movements compete violently for the authority to reward virtue and punish vice, leading to an endless series of civil and regional wars. The Reformation was not the Enlightenment.)

Books to Read While the Algae Grow in Your Fur