i recently watched
the eagle huntress, a documentary about a girl in mongolia who wanted to be the first woman to hunt with golden eagles. it's a pretty good movie, especially if you like rugged scenery, large raptors, and/or people living under what urban sissies like me regard as insane conditions. not as insane as the people in
happy people, nor as anywhere as self-reliant as them -- the girl's family's ger has a solar panel, and some of the semi-nomads do ride motorcycles into town rather than horses. (the nearest city is
ulgii, variously anglicized in the movie. they have an eagle hunting festival. it's apparently a stop on the exotic-places tour, since the only people in the movie with cameras were white tourists taking pictures of the contest. even i do a little better in interesting places.)
the movie's one weakness is that the director seems to be trying too hard to find villains for the plot. while there's a series of talking-head interviews with four old get-off-my-steppe guys who are absolutely against women learning the manly art of eagle hunting, by and large, everybody in the movie shrugs and says, "let her try".
still, i recommend it, but i like rugged scenery, large raptors,
and insane living conditions.
on the other hand, i bravely read
scale: the universal laws of wide-eyed, self-deluded social-science types, that angels feared to read. it could have been worse.
the author may be on to something -- as opposed to just on something -- but it's impossible to tell from the evidence. too much of it looks cherry-picked, some of it has problems with the curve fitting -- today's $64 word is heteroscedasicity: uneven or systematic variation in the scatter of data -- and the rest may be explained by hypotheses other than the author's pet one. for example, he has a nice set of graphs showing that the number of gas stations in european cities grows with population with a particular set of exponential laws. he attributes the sub-linear scaling to the physical efficiency of large cities, but it could also be explained by bigger cities being more likely to have public transit, and simply needing less gas per person.
but! he does correctly use metropolitan statistical areas (MSAs) for talking about US cities. yay! too much stuff i've read on the wonders of cities uses only figures for the cities proper, not their vast suburbs/conurbs. that renders the results nearly worthless, because the economic effects of modern¹ cities no longer conveniently stop at their borders.
more importantly, this implies there's data on MSAs, and i now know how to find it, so i can poke at the wonders of cities myself. while it's clear that (for example) average salary grows quickly in population, i wonder about
median salary... or total compensation. latte-sipping city boosters seem to like to forget that their easy lifestyles are supported by legions of poorly-paid people, many of whom have been pushed out of their now-gentrified lifestyle-enabling areas.
so, while the book may be a loss, it's not a total loss, and i don't discount the possibility that the universal scaling laws the author wants exists. but i'd like some reputable data to support it.
1: modern being since the 19th century development of streetcars, subways, and elevated trains, which allowed cities to grow far beyond their previous sizes.