Friday, May 19, 2006
Monday, May 15, 2006
J. Phillipe Rushton's Race, Evolution, and Behavior covers a sensitive topic, racial differences, in a lot of detail, and with a heavy hand. Most of the book is overflowing with statistics and citations to an enormous number of references in quite a few fields by quite a few researchers. If the idea was to overwhelm the reader, I'll admit it was successful in my case. I eventually learned that I could skim the statistics, and look for prose summaries of the significance. Rushton didn't pay much attention to what it all means until near the end.
Rushton's main point is that there are several attributes on which there is systematic variation among the races, and his claim is that the consistency hints at a possible cause. He's mostly interested in convincing us of his pet theory of the cause of the differences. He's weighing us down with evidence to show us how consistent the inter-racial differences are. He mentions exceptions to the trends, but doesn't give them much emphasis, or provide plausible explanations within his framework.
The traits that Rushton focuses on include brain size, intelligence (as measured by IQ tests and other tests that correlate with g), maturation rate, personality, social organization, and reproductive effort. Other than intelligence (and brain size which is strongly correlated with it), and arguably social organization, these are not scales with a good end and a bad end, they're just differences. The consistency that Rushton focuses on is that whichever of these attributes you measure, you end up with whites in the middle, and orientals and blacks to either side. The story that Rushton wants to build out of that data is that the differences line up the way you'd expect if you tried to predict the direction from r/K evolutionary theory. (Basically, differential evolution in environments of scarcity and plenty; in this case, evolution in the tropics versus the northern latitudes in the Ice age.) He makes a fairly good case for the consistency; there appears to be something going on that puts orientals and blacks at two extremes and whites somewhere between them. But his favorite explanation doesn't seem to be strongly supported. It's roughly consistent with the data, but I'd need more explanation for the exceptions before I'll accept the case as demonstrated.
My biggest complaint about the book is about an important point, orthogonal to Rushton's argument, that he doesn't address at all. Rushton clearly knows that discussion of racial differences is a hot-button issue, and the reason that people care is not because of the difference the data will make to our beliefs about the evolutionary causes of the difference. It's a hot-button issue because accepting the data seems to tell us something significant about innate differences or tendencies to differ among people on aspects of behavior that matter to social policy. When Murray and Herrenstein addressed these issues in The Bell Curve, they were careful to give us the important caveat: the differences between the mean IQs are significant, but the overlap among the curves is large. That means that you can predict the average IQ of a group if you know the racial make-up, but you don't really know much about an individual from that detail. (Wikipedia has a nice graphic showing how much overlap there is among the curves. With the other attributes Rushton discusses (aggressiveness, impulsivity, law abidingness, life span), it's much harder to figure out what interventions might be called for if we knew the differences were significant and pervasive, but not knowing how much overlap there is between races on these measurements makes it hard for me to even tell whether they matter other than in discussions about evolutionary causes.
Another factor, relating to intelligence, that was brought up at the reading group, is fascinating and was completely unaddressed. Rushton says that American Blacks have an average IQ of 85, while in central Africa, the number is closer to 70. In the US, we have the feeling that we could immediately recognize someone with an IQ of 70 as subnormal, but if that's the average in sub-saharan Africa, then something else is going on, because you have to assume that 90% of the population is functional. So the IQ of 70 must mean something different than what we'd assume here, where it normally implies other deficits than just cognitive. It's not clear what an average IQ of 70 in a functioning population would mean. This casts some doubt in my mind on the usefulness of IQ for cross-cultural or cross-racial comparisons. The Wikipedia article makes similar observations.
As to Rushton's thesis, I remain unconvinced that r/K theory is the best explanation for the differences. r/K theory is well-established from studies of other animals, but Rushton isn't careful in marshalling his arguments to convince me that r/K with appropriate caveats for some unexpected cases provides a good enough match to the data. Wikipedia mentions several attacks on Rushton's thesis and methods, but no alternative explanations of note. That's fine; sometimes the facts you need for the right theory aren't at hand when you want them.
Wednesday, May 10, 2006
Strange, powerful book. I didn't notice till I finished it that it was 600 pages long. Very dark, foreboding imagery; China Miéville constantly describes how grimy the city is, and how poor the inhabitants are.
The story is a strange combination of fantasy, science fiction, and horror. Many kinds of magic run loose. The main character, Isaac Dan der Grimnebulin, is a proto-scientist, testing hypotheses, trying different approaches to control the forces he's beginning to understand, and consulting with other experts to share what he's learned and see what they know that might be relevant to his project. There are so many kinds of magic that anything is possible, but Miéville manages to keep most of it as a side show, and the magics that are crucial to the plot are kept under tight dramatic control. Every significant plot twist is plausible based on what we're shown of the powers of the magic beforehand.
The city and its inhabitants are endlessly fascinating. There are 4 or 5 major races, with multitudes of variations, since one of the magics is bio-thaumaturgy, the ability to remake living flesh into different forms. A short catalog of types: humans, garuda (half eagle, half man), vodyanoi (creatures of the water, who can mold living water so it will hold a shape), handlingers (paired parasites with psychic communication and flight), khepri (an arthropod/insect), mobile intelligent cactuses, an actual demon from hell, a meta spider (walking the world's causal web), wyrmen (diminutive, intelligent, communicative flyers), self-aware machines and the humans who worship them, and over them all, slake moths.
The slake moths are the scariest of all: nearly invincible; feeding on a victim's dreams and leaving a zombie behind; able to mesmerise practically anything that's aware. Perdido Street Station is the story of how the slake moths were brought to the city and accidentally released and the struggle to contain or kill them.
There are several back stories, nearly all crawling with depravity, but the one that drives the main story tells of Yagharek, a garuda, punished for the crime of choice theft (preventing another from making choices) who asks for help from der Grimnebulin. In his investigations, der Grimnebulin accidently unleashes the slake moths, and goes to great lengths to contain them. In the final struggle , after the crisis has passed, der Grimnebulin is confronted by Yagharek's victiim, who tells der Grimnebulin what Yagharek has done and asks him to uphold the verdict against him. It feels like a rude shock, pulling us back to an unexplored conflict from the beginning of the book. But the intercalary chapters have explored Yagharek's angst all along, hinting that this was the important substance of the story, and the starker danger of the slake moths was only a distraction.
Wednesday, May 03, 2006
The pointer I had on my home page to the bike races at Hellyer park velodrome was way out of date (1999, thanks Bob!), and it's nearly summer, so I went looking for a new link. They're currently tracking the racing schedule here. It's a site I haven't seen before. Perhaps it'll last longer than the previous ones. Every year I have to search for a schedule, since they never seem to be in the same place two years in a row.
The Friday Night series seems to start June 2.
Another bike race we often watch is the Cat's Hill Criterium. It takes place in Los Gatos on a course around a couple of city blocks, and features a pair of steep hills for the riders to climb and descend. This year's race will be on May 13. The races go from 9:30 - 5:40pm, with the pros starting at 4:10. More info here.