Book Review: The Blank Slate : The Denial of Human Nature and Modern Intellectual Life by Steven Pinker
Pinker makes a good case that there’s a widespread bias toward a blank-slate world-view. But when dealing with serious scientific literature, his attempts to find clearcut enemies seem mistaken.
Pinker’s claim that “The second scientific defense of the Blank Slate comes from connectionism” is pretty puzzling. This “defense” consists of modeling the mind as “a general-purpose learning device”. But the books that Pinker references (Rethinking Innateness, and Parallel Distributed Processing), are both careful to point out why their models are completely consistent with the kind of genetic influences on behavior that evolutionary psychologists are talking about. Their disagreements with Pinker seem to be at most about how those influences are implemented, and even there I can’t find anything in Pinker’s arguments that clearly rejects what the connectionists believe.
Pinker’s attacks on Gould’s quasi-defense of the blank slate mainly convinced me that Gould didn’t want to think clearly about the subject, probably because he considered that any mechanistic explanation of the mind (genetic or environmental) was demeaning.
Pinker’s arguments that it’s silly to believe in the tabula rasa and noble savage world-views are eloquent and compelling, but his response to the “it’s demeaning” attitudes will convince fewer people, because he ignores the very real benefits of holding an unrealistically high opinion of one’s self (overestimating one’s abilities seems to be an effective means of advertising one’s strengths). To those who want to portray themselves as angelic or as wiser than software of the future, an accurate model of the mind is genuinely demeaning.
Pinker seems somewhat inconsistent about how important it is to know whether the mind is a blank slate.
On pages x – xi he says “the conviction that humanity could be reshaped by massive social engineering projects led to some of the greatest atrocities in history.” But in the chapter on fear of inequality, he claims (more convincingly), while defending his views from the charge they will encourage Nazism, that the differences between Nazi beliefs in genetic superiority and the blank slate viewpoints of Stalin and the Khmer Rouge didn’t have much effect on whether those tyrannies engaged in genocide – it was the greater tendency to divide people into in-groups and out-groups that best distinguishes the worst of the genocidal tyrants.
Pinker exaggerates the importance of finding the correct answer to the nature-nurture debate in other ways as well (I suppose we shouldn’t be surprised that an author overestimates the importance of what he’s selling).
He gives examples such as forcing people to live in drab cement boxes (as if their taste for a more natural surrounding could be reversed by social engineers), or releasing psychopaths (because societal problems caused their insanity).
But a genetic component to these behaviors doesn’t prove that they can’t be altered (I have genes for brown hair – does that mean I can’t dye my hair blue?). It only gives hints as to why they might be difficult to alter.
It sure looks like careful scientific studies of whether we knew how to alter these behaviors would be a more reliable way of debunking the faulty conclusions.
Science and Technology
There’s a report that the Flynn effect has stopped fairly abruptly in the industrialized countries. The new data suggest a more sudden halt than nutritional theories would predict. I’m uncertain what to make of this.
Book Review: The Emperor of Scent : A Story of Perfume, Obsession, and the Last Mystery of the Senses by Chandler Burr
This is an entertaining and informative story of how a failed class of theories that tried to explain smell as purely shape recognition have been challenged by a theory that involves detecting the vibrational frequencies of bonds.
Burr does a very good job of making some rather technical discussions of chemistry readable and accurate without intimidating amateurs. The story is enjoyable except when reporting the ugly side of academic and corporate politics. The extent to which he claims that researchers who have based careers on the failed approach to smell refuse to think about the vibrational theory seems rather extreme. Yet I can only find one instance in which Burr unfairly criticizes one of those researchers (on page 233, when he says it’s strange not to know what “the same vibration” means, and implies that “same” simply means an identical integer number of nanometers. Yet it would be strange if vibrations had an integer number of an arbitrary unit such as nanometers, and on page 64 Burr implies that vibrating at 2550 nanometers is the same as 2500 nanometers).
One other incidental comment the book makes about the peer review process is worth repeating. One of the reasons that big name research labs continue producing good results is that they have an advantage comparable to insider trading as a result of seeing papers at the peer review stage, while the average lab has to wait longer to get the same ideas.
Bryan Caplan writes about Robin Hanson’s contrarian views on the effectiveness of medicine and parenting.
Caplan’s conclusion about medicine involving a mix of beneficial and harmful practices is probably correct (and probably consistent with Robin’s views), but some of his reasoning is bogus:
Start with medicine. Modern techniques have clearly saved a lot of lives. If memory serves me, survival rates for premature babies have skyrocketed from 10% to 90%.
Part of Robin’s point is that we can’t tell from the improved survival rates that medicine was responsible. It may be that improved maternal nutrition has made babies better able to withstand premature birth. There is no easy way to distinguish the causes, and there’s some reason to think doctors are more effective at biasing consumers to credit them with improving peoples’ health than farmers are.
The evidence that medicine is less effective than most believe has fewer practical implications than a superficial glance suggests. It implies that you shouldn’t choose an expensive health plan over a cheap one, but leaves open the possibility that you should still see your doctor fairly often, and the possibility that you can “buy” health care that will slightly increase your life expectancy by moving from, say, Havana to San Francisco.
The argument (started by Judith Rich Harris) that parenting styles have little effect has a stronger conclusion. Caplan claims:
The same goes for parenting. We all know kids who let their parents plan their lives for them. Maybe it’s 100% genetic, but that’s a stretch. It’s more plausible to acknowledge that these pliable kids exist, but point out that they’re only half the story. We also all know kids who heard their parents’ plans for their future, and did exactly the opposite just to spite them.
I do not know kids who come close to fitting the first pattern after puberty. Essentially all kids need to demonstrate to their peers by about puberty that they are mature enough to be somewhat independent of their parents. And if you think about the sexual selection pressures on children around that age, you should expect that to be just one symptom of the pattern that Harris points out. Their reproductive success is heavily dependent on their ability to compete with and to impress people who are sufficiently close to their age to become a mate or to compete for a mate. That implies that it is important for them to adapt their personalities in ways that respond to evidence about their peers, and to treat parental opinions as much less relevant.
Unlike the arguments about the ineffectiveness of medicine, the evidence against the importance of parenting styles appears to show that all attempts to improve parenting styles (except for those, such as choosing the best school, which influence whom the child can have as peers) have failed to show benefits.
We have a large industry devoted to convincing parents to buy its advice on parenting styles. This creates a nontrivial incentive to provide evidence that some parenting styles work better than others. That includes incentives to distinguish children that will be helped by style X from those who will be helped by the opposite style. And unlike the evidence that some medical practices work, the evidence for the value of advice on parenting styles consistently fails when subjected to close scrutiny.
It is still possible that parenting styles are sometimes helping and sometimes hurting, but theory and the breadth of the evidence suggest betting against that. Eventually, given tools as drastic as manipulating the child’s genes, parents will someday find ways to manipulate their kids minds. But since there’s little reason to think that children are currently suffering from negligent parenting styles, and there are moderately good reasons to guess that youthful rebellion is mainly the result of children pursing their (gene’s?) interests, it’s hard to see why parents should be trying to alter their children’s behavioral strategies rather than ensuring that they have the resources to do what they want. (Unless, of course, parents have good reasons for pursuing different goals than their children. I’m having trouble analyzing that possibility.)
Temple Grandin’s latest book Animals in Translation has a couple of ideas that deserve some wider discussion. (The book as a whole is disappointing – see my reviews on Amazon for some of my complaints).
She reports that Con Slobodchikoff has shown that prairie dogs have a language that includes nouns, adjectives, and verbs, and they can apparently combine words to describe objects they haven’t seen before. This seems sufficiently inconsistent with what I’ve read about nonhuman languages (e.g. in Pinker’s books) that it deserves more attention than it has gotten. I can’t find enough about it on the web to decide whether to believe it, and it will take some time for me to get a paper version of Slobodchikoff’s descriptions of the research.
Grandin has an interesting idea about the coevolution of man and dogs. Domestication of animals causes their brains to become smaller, presumably because they come to rely on humans for some functions that they previously needed to handle themselves. It seems that human midbrains shrank about 10% around 10,000 years ago, about when dogs may have become domesticated. That is what we would expect if humans came to rely on dogs for many smelling tasks.
Book Review: FAB: The Coming Revolution on Your Desktop–From Personal Computers to Personal Fabrication by Neil Gershenfeld
This book brings welcome attention to the neglected field of personal, general-purpose manufacturing. He argues that the technology is at roughly the stage that computing was when minicomputers were the leading edge, is good enough to tell us something about how full-fledged assemblers as envisioned by Drexler will be used, and that the main obstacle to people using it to build what they want is ignorance of what can be accomplished.
The book presents interesting examples of people building things that most would assume were beyond their ability. But he does not do a good job of explaining what can and can’t be accomplished. Too much of the book sounds like a fund-raising appeal for a charity, describing a needy person who was helped rather than focusing on the technology or design process. He is rather thoughtless about choosing what technical details to provide, giving examples of assembly language (something widely known, and hard enough to use that most of his target users will be deterred from making designs which need it), but when he describes novel ideas such as “printing” a kit that can be assembled into a house he is too cryptic for me to guess whether that method would improve on standard methods.
I’ve tried thinking of things I might want to build, and I’m usually no closer to guessing whether it’s feasible than before I read the book. For example, it would be nice if I could make a prototype of a seastead several feet in diameter, but none of the examples the book gives appear to involve methods which could make sturdy cylinders or hemispheres that large.
The index leaves much to be desired – minicomputers are indexed under computers, and open source is indexed under software, when I expected to find them under m and o.
And despite the lip service he pays to open source software, the CAM software he wrote comes with a vague license that doesn’t meet the standard definition of open source.
Book Review: On Intelligence by Jeff Hawkins
This book presents strong arguments that prediction is a more important part of intelligence than most experts realize. It outlines a fairly simple set of general purpose rules that may describe some important aspects of how small groups of neurons interact to produce intelligent behavior. It provides a better theory of the role of the hippocampus than I’ve seen before.
I wouldn’t call this book a major breakthrough, but I expect that it will produce some nontrivial advances in the understanding of the human brain.
The most disturbing part of this book is the section on the risks of AI. He claims that AIs will just be tools, but he shows no sign of having given thought to any of the issues involved beyond deciding that an AI is unlikely to have human motives. But that leaves a wide variety of other possible goals systems, many of which would be as dangerous. It’s possible that he sees easy ways to ensure that an AI is always obedient, but there are many approaches to AI for which I don’t think this is possible (for instance, evolutionary programming looks like it would select for something resembling a survival instinct), and this book doesn’t clarify what goals Hawkins’ approach is likely to build into his software. It is easy to imagine that he would need to build in goals other than obedience in order to get his system to do any learning. If this is any indication of the care he is taking to ensure that his “tools” are safe, I hope he fails to produce intelligent software.
For more discussion of AI risks, see sl4.org. In particular, I have a description there of how one might go about safely implementing an obedient AI. At the time I was thinking of Pei Wang’s NARS as the best approach to AI, and with that approach it seems natural for an AI to have no goals that are inconsistent with obedience. But Hawkins’ approach seems approximately as powerful as NARS, but more likely to tempt designers into building in goals other than obedience.
Robin Hanson has another interesting paper on human attitudes toward truth and on how they might be improved.
See also some related threads on the extropy-chat list here and here.
One issue that Robin raises involves disputes between us and future generations over how much we ought to constrain our descendants to be similar to us. He is correct that some of this disagreement results from what he calls “moral arrogance” (i.e. at least one group of people overestimating their ability to know what is best). But even if we and our descendants were objective about analyzing the costs and benefits of the alternatives, I would expect some disagreement to remain, because different generations will want to maximize the interests of different groups of beings. Conflicting interests between two groups that exist at the same time can in principle be resolved by one group paying the other to change it’s position. But when one group exists only in the future, and its existence is partly dependent on which policy is adopted now, it’s difficult to see how such disagreements could be resolved in a way that all could agree upon.
This book provides a moderately strong argument that the production of cheap oil is peaking, although it isn’t as conclusive an argument as I’d hoped for, and is only a little bit better than the brief summaries of Hubbert’s ideas that I’d previously seen on the net.
Much of the book consists of marginally relevant stories of his career as a geologist. He occasionally slips in some valuable tidbits, such as that Texas once had an oil cartel.
He does a mediocre job of analyzing the consequences of scarcer oil. He provides a few hints of how natural gas could replace oil, but says much less about the costs of switching than I’d hoped for. His comments on how to protect yourself are misleading:
In the past, a useful way of insuring major producers and consumers against the effect of a price changes was purchasing futures contracts. However, the ordinary futures contracts extend for a year or two. The oil problem extends for 10 years or more. The oil problem extends for 10 years or more. Anyone who agrees to supply oil 10 years from now, for a price agreed on today, very likely will disappear into bankruptcy before the contract matures.
At the time the book was first published (2001), crude oil futures contracts extended about 7 years out. They weren’t liquid enough to hedge a large fraction of consumption, but if a desire to hedge had caused them to say in 2001 that crude would be at $60/barrel in 2008 rather than saying it would be in the low twenties, that would both have signaled a need to react and reduced the risks of doing so. The idea that bankruptcy would threaten such futures reflects his ignorance of the futures markets. An oil producer who sold futures as a hedge will almost certainly not sell more futures than it has oil to deliver on. Speculators might lose their shirts, but futures brokers have the experience needed to ensure that the defaults are small enough for the brokers to absorb (see, for example, what happened in the gold mania of the late 70s).
Patri Friedman asks why websites often require users to deal with annoying pulldown menus such as those listing 50 states. I expect that the main reason is that users who are allowed to type in text will enter it in nonstandard forms. For example, Massachusetts will be entered as Mass or MA, or if limited to 2 characters the user might not remember the correct 2-letter code. Sites that need to calculate sales taxes differently for different state, or who think (not necessarily with good reason) that they need to analyze customers by location for marketing reasons, need either standardized input or a good deal of imagination to predict all variants they will get. Imagination isn’t cheap.
I suspect there’s also a desire by some designers to show their status over users by preventing users from entering unexpected input.
I doubt these factors are enough to explain all examples of annoying pulldown menus, but I’d guess they explain at least half.