February 3, 2013

• Ahead of their time: A review of “Are We Getting Smarter?” by James R. Flynn

  JDN 2456327 EDT 19:20.

   

  We often say that a genius such as Leonardo or Newton is “ahead of their time”; I doubt most people realize just how accurate this is. In what is called “the Flynn effect”, IQ scores have been increasing steadily at a rate of about 0.3 IQ points per year since we have been keeping track of them. This means that in the past century we have risen an average of 30 IQ points, which is the difference between average and the top 2.5%, or the difference between the bottom 8% and the top 8%.

  We obviously haven’t been increasing this fast since Newton and Leonardo; if we had, Newton’s IQ of, say, 160 the year he died would be an IQ of only 74 today (borderline retarded). Leonardo’s IQ of, say, 180 when he died would become a mere 30 today (chimpanzee). That’s pretty ridiculous. So let’s assume that the increase started around 1850; in that case we spot them the same 50 IQ points, and Newton’s 160 becomes 110 while Leonardo’s 180 becomes 130. They’d still be smart, but not outlier geniuses. And indeed, if you tested them on standard IQ tests, I’m pretty sure that’s what you’d find.

  In Are We Getting Smarter? James R. Flynn (yes, the effect is named after him) charts this meteoric rise, and seeks to answer the question on all of our minds: Is this a real rise in intelligence, an artifact of bad tests, or something else?

  Flynn leans toward “something else,” but he also thinks “real rise in intelligence” is worth considering. He pretty clearly demolishes the idea that it’s just bad tests; we ask the same questions today we did 50 years ago, and people just plain do better at answering them. Moreover, these aren’t questions that require a lot of domain-specific knowledge; they’re things like pattern completion and recalling strings of numbers. The most domain-specific category is the vocabulary section, and it’s based on words that have been in common use for decades.

  Flynn’s theory is that the change has been abstraction, or what he calls “scientific spectacles”; we train people better now to think in terms of scientific categories and abstract relationships. Consider the question, “What do a fish and a crow have in common?” Most people used to say “A crow can eat a fish”, which is very concrete and won’t score well on an IQ test. Now most people say “Both are animals”, which is more abstract and scores higher. High-IQ people say “Both are vertebrates”, which is abstract, precise, and scientific.

  Flynn also thinks we could do even better at teaching scientific attitudes than we already do, and I’m inclined to agree, given the prevalence of Creationism and global warming denial. But even I have to admit, today’s Creationists use more scientific language than the Creationists of yesteryear. It’s bad science, pseudoscience; but it shows how highly valued science is in our culture. You simply can’t get away with saying “Science is of the devil” anymore, you have to say “Science reinforces the Bible”.

  I would hope that the steady rise in world IQ (which varies from nation to nation) would silence the sort of people who believe in an Idiocracy-style eugenic crisis where our population becomes dumber over time. But of course it won’t. In fact IQ is inversely correlated with birth rate, but it’s also inversely correlated with death rate, so it’s not clear that people with less IQ actually have any more children who survive to breeding age. But even if that turns out to be the case, the fact remains that overall IQ scores are rising at a remarkably fast rate due to environmental pressures. If our genes are making us dumber, our environment is making us smarter a lot faster.

  The book itself is full of data tables and statistics, which makes it a little dry. You can’t fault his scientific rigor, but beware that you will be bombarded with long lists of means and standard deviations. Also he has a tendency to report things like “3.284 plus or minus 1.261″, which is a really weird kind of precision. How about “3.3 plus or minus 1.3″, seeing as we already have a wide margin of error?

  The reader is kept awake, however, but Flynn’s fascinating insights and his occasional barbs at academic culture. He clearly doesn’t like a lot of things about how our academic community is organized. He goes through long lists of papers that contain basic factual errors, and spares no criticism for university administrators and funding agencies. It’s hard not to agree, given some of the examples of bad science he talks about that have gotten published. My favorite is this one: “The silliest piece of social science I have seen was not in psychology but in politics: a thesis on nonvoting in the Washington metropolitan area. The candidate was unaware that the Hatch Act banned the residents of the city proper, the District of Columbia, from voting in Congressional elections. [...] They constituted one-fourth of those sampled. The supervisor’s attempt to defend the merits of the thesis was fascinating.” (p.181) In short, why is voter turnout so low in DC? Maybe because people who live in DC aren’t allowed to vote? Who published that paper!? They should be fired.

  A large section of the book is dedicated to showing that the measured differences in IQ between White males and others are due to sociological, rather than genetic, causes. Flynn chides other IQ researchers for leaping to genetic conclusions; but he also chides others for not being willing to face the facts and do the actual research at all. Rather than assuming that genes are not the cause, we should be able to marshal evidence for that; and this is what Flynn tries to do. His case is pretty convincing; the sociological evidence is compelling, and the opposing arguments are pathetically weak. At times I felt that he actually went too far in the direction of being charitable to his opponents, some of whom clearly are racists and sexists. If you’re going to be so harsh against papers with bad statistics, how about you direct some of that toward blatant racism?

  One argument that keeps turning up is that because IQ is highly heritable, it must be largely genetic. This argument is not just wrong; it’s ludicrous, as can be seen from a simple example. The heritability of the trait “speaking Japanese” is quite high, as you might imagine; most people who speak Japanese had parents who spoke Japanese, and most people who speak Japanese have children who speak Japanese. However, speaking Japanese is entirely environmental; as long as you have the basic genetics required for speaking any human language (you are not so retarded as to be nonverbal, your ears and mouth work normally), you can learn Japanese just as well as anyone else if you are exposed to it from a young enough age. Japanese parents who adopt an American child will have no problem raising that child to speak Japanese.

  Yes, smart people tend to raise smart children. This because they live in smart environments, provide smart opportunities, and live in smart nations. It could also be because they carry smart genes, but you need to actually show that; it doesn’t by any means follow from the heritability coefficient itself.

  My favorite chapter of the book is chapter 7, “The sociological imagination”, in which Flynn derides other researchers in psychology for being greedy reductionists who ignore cultural and social causes and try to make everything about brain physiology. I’ve been saying this for years, and I will continue to say it; trying to understand human behavior purely in terms of the brain would be like trying to understand the Internet purely in terms of quantum mechanics. Humanity really needs to get over the long-standing mistake that things are defined by the stuff of which they are made, and not by the function that they do. You are made of neurons, yes; but that does not mean you are neurons. A house is made of wood, but a pile of wood is not a house. A vat full of neurons would not be Mozart.

  It’s the failure of sociological imagination that leads to another ludicrous argument Flynn’s opponents make, which once again he gives more credit than it deserves: Women and Black people couldn’t score lower for sociological reasons unless there were some mysterious “factor X” that only affected them and not White males, something highly implausible and impossible to verify. What mysterious phenomenon harms Blacks and not Whites, women and not men, gays but not straights? I wonder what it could be? It’s called discrimination, you idiot. It’s not mysterious at all. (And indeed, as Flynn shows, countries that discriminate less show smaller differences in IQ scores between various groups.)

  It’s particularly amusing how he shows that the fact that women in college have a lower mean IQ than men in college doesn’t mean that women are dumber in general; instead, it just means that women have an easier time getting into college without being as bright. On average, women are the same as men in IQ; it’s just that women can get into college with an IQ of 95, while men need an IQ of 100. Flynn uses data on dropout rates and conscientiousness scores to argue that the reason is basically that women work harder, and as a result can pass the thresholds into college without having as much raw intellect. This certainly meshes with my own experience; everyone I know who is brilliant but lazy is male, while everyone I know who is of average intellect but works their ass off to succeed is female. Personally, I’m working on being brilliant and industrious both… but how the laziness tempts.

  The chapter I feel most ambivalent about is chapter 4, “Death, memory, and politics”, in which Flynn tries to argue that we should adjust our IQ thresholds for the death penalty in light of the Flynn effect. I did not know this before, but apparently it is illegal in the US to execute someone whose IQ is below 70, on the grounds that below that level they must be mentally retarded and unable to understand the consequences of their behavior.

  It seems to me that this system is really founded upon an outdated concept of free will, in which you must be “really responsible” for your actions before it’s legitimate to punish you for them. We can’t execute someone who is too stupid to understand why they murder people! Well, why not? After all, they murder people, and we don’t know how to stop them. These, for me, are pretty much the reasons we’d execute someone. I’m not terribly concerned about the question of whether they murder people rationally; indeed, it’s pretty clear to me that almost no murders are rational, because it could only be rational to murder someone in a society where murder wasn’t reliably punished severely. Even then, it would still be wrong, but at least maybe it could be rational. (And don’t try to pull out self-defense or just war exceptions: Those are by definition not murder, and wouldn’t carry a death penalty. In fact, they are usually not prosecuted at all.)

  If you can find a better way to stop the murders, well, let’s do that instead. But then, this raises the question: Why not always do that? Why kill some murderers and not others? If imprisonment works just as well as execution, why use execution at all, knowing it is costly, controversial, and irrevocable?

  It would make some sense to me to execute psychopaths and not others, as we know that psychopaths are especially dangerous and especially likely to charm parole boards into releasing them. And I suppose it’s probably true that murderers of higher IQ are more likely to be psychopaths (this could certainly be tested easily enough). But we have standard tests of psychopathy, which are not dependent on IQ and don’t show the Flynn effect over time. (Oddly enough, tests of narcissism do show a Flynn-like effect, in that certain narcissistic traits have become more common in the population over time. We’re not really sure about the causes or the significance of this; it could well be another book worth writing.)

  Also, Flynn’s argument that we should adjust scores to the current averages doesn’t quite sit right with me. It seems to rest upon a notion that moral responsibility is a purely relative measure, which isn’t true. Homo sapiens are smarter than Pan troglodytes, which are smarter than Canis lupus. As such, we grant more rights and responsibilities to humans than to chimps, and more to chimps than to dogs. But if the world were changed somehow so that all the really stupid animals died out (all the insects and worms are gone, but somehow we survive without them), we would not thereby conclude that dogs deserve fewer rights because they’re now closer to the bottom. Likewise, if human beings died out, it would be wrong to say that chimpanzees should vote and fly airplanes because they are now the smartest apes (Planet of the Apes notwithstanding).

  If IQ is really important to moral responsibility, then rises in IQ should also entail rises in moral responsibility. And if it isn’t, why are we using it to decide who gets executed? Flynn tries to argue that what we really care about is retardation, which (he presumes) isn’t decreasing over time, always representing a steady 2.5% of the population. He then makes the additional assumption that retarded people would always score lower on IQ tests than non-retarded people, even as all the IQ scores rapidly improve over time. Both of these assumptions strike me as entirely unsupported; maybe the IQ gains are actually due at least in part to improved treatment and prevention for mental retardation! In any case, if we’re not going to base our decisions on IQ tests, how about we come up with a new test that actually measures what we really care about?

  In fact, Flynn himself hints at the idea that rising IQ should entail rising moral responsibility, though not in so many words. He briefly analyzes the history of political debate, showing that the complexity of arguments used has increased over time, especially in arguments used by politicians and economists toward one another. He particularly compares William Jennings Bryan’s “Cross of Gold” speech, which is rhetorically brilliant but almost content-free, with more modern economic commentators like Paul Krugman, who regularly cites tables and graphs. Apparently, the arguments that politicians give the masses aren’t changed as much, which is rather disappointing but not terribly surprising. (“Change we can believe in!” I like Obama for the most part, but come on; what does that even mean? “It’s not a coincidence that you use R to go back and D to go forward!” Yes, actually, that’s exactly what it is.) Also, there’s a study saying that the Tea Party caucus displays a lower level of cognitive complexity than mainstream Republicans or Democrats, which is amusing but also not surprising.

  With this in mind, the future is bright: We can predict that a smarter population will also be a morally better one. As support, I offer Pinker’s observation in Better Angels of Our Nature that violence is decreasing over time. Science shows people are getting smarter and nicer! How’s that for good news?

   

  • 10:22 pm
  • 5 Comments