Ad Nausica

Let’s begin with a remedial lesson on betting. Consider a very basic bet. The bet comes with a certain cost, C. This may be the price of a lottery ticket, the wager you put on the roulette wheel, or some embarrassing task you agree to do if you lose. The bet also comes with a certain payoff if you do win known as the benefit, B. Suppose you were able to make the bet a very large number of times, N. You win the bet some of the time, for a total of W wins.

This means your total winnings are WB and total costs are NC. To get your average winnings and costs we divide both by the number of times you bet, which gives your average winnings as (W/N)B and average costs as (N/N)C = C. The ratio W/N is the fraction of times you win and is what defines the probability of winning, p = W/N. For any single bet then, we don’t refer to your average winnings but rather your expected winnings, pB.

A good bet is one in which your expected winnings exceed the cost to make the bet. That is, pB>C. Another way to write this is that your expected gain is E = pB-C>0. If you make many good bets over time, you will get richer, or at least continue to gain benefits. If you make bad bets, pB<C or E<0, you will tend to go broke and lose what you have. So, for example,

We’ve only looked at a simple bet. We can complicate the matter by adding multiple benefits with different probabilities, and multiple costs that also have different probabilities, q. For example, a lottery might have multiple prizes with different probabilities for the cost of a single ticket, or an activity may have a variety of potential cost risks.

The general form of the expected benefit is then

E = p_1*B_1 + p_2*B_2 + p_3*B_3 + … q_1*C_1 - q_2*C_2 - q_3*C_3 - …

For now we will stick to simple bets.

Hamilton Rules!

The math of betting has a surprising amount to do with the evolution of altruism. Let’s first identify the problem of altruism. It’s not hard to understand the genetic evolution of self-serving behaviours. A genetic allele that contributes to the survival and reproduction of the person it inhabits will tend to result in the host having more offspring than an allele that doesn’t contribute as well, meaning there are ultimately more copies of the first allele than the second in the general population. Over time, the first will tend become the dominant allele in the population and we’d expect to see the resulting trait or behaviour as the norm of the species. Such an outward trait of organisms that is an expression of the gene is called a phenotype.

Altruism does the opposite. Altruistic behaviour results in a proximate cost to the organism performing the act. For example, a genetic allele that tends to cause you to give your life to save somebody else doesn’t immediately make sense. Any gene contributing to that behaviour will tend to reduce the survival and reproduction of its host, and so we’d expect to see fewer offspring and hence fewer copies of that allele, and so intuition suggests think that it could never become common. This is true even for minor altruistic behaviours which may only cost a few calories, particularly when calories are precious. So how could altruistic behaviour evolve?

One of the solutions to this problem is reciprocal altruism. This is simply the case where an altruistic act only has a minor cost to the actor and there is an expectation of payback in the future. In other words, pB>C, where p is probability of payback, where B and C are both in terms of reproductive success. It is not my goal to discuss this solution, so I will point the reader to Richard Dawkins’ The Selfish Gene to read more. However, it is important to point out that reciprocal altruism can’t easily explain the extremes of altruism such as sacrificing your life for other people.

Another solution was formulated by William Hamilton in 1964. If the altruistic act aids other copies of the same allele then there is net value to the genetic allele even if it harms one of the hosts of that allele. That situation may be hard to comprehend. When we think of self-serving behaviour, we tend to think of it serving an individual person. But genes aren’t people; they are segments of DNA. They have no instincts for self-preservation because they have no instincts. The phenomenon we are trying to explain is not the self-interest of particular copy of the genetic allele, but the tendency toward more copies of a particular allele in the general population. If a genetic allele codes for behaviour that results in the demise of a particular copy but increases the number of other copies, that explains the phenomenon.

This is the context of the metaphorically selfish gene that Dawkins coined. It is not the selfishness of an individual copy. It is the apparent selfishness of the information coded by the gene. In effect, the math of copying frequency simply means that genes that contain information (e.g., particular molecular composition) that increase the copying frequency of itself tends to result in that information (molecular composition) being more common. It is not molecules themselves or a particular set of molecules that become more common, but the particular information corresponding to the pattern of the molecular structure. That is what acts in a metaphorically selfish way to increase copies of itself. But do please not that it is not actually selfish or “trying” to do anything. It is simply the math of a positive feedback loop: the information affects the rate of itself being copied. Hence if the information in question contributes to destroying one copy to save more copies, that serves this same “selfish” outcome of more copies of it.

While this can solve the problem of how genes for altruism can spread, it doesn’t address how a gene can “know” that there are other copies being saved by sacrificing its host. If the gene-coded altruistic behaviour was applied randomly to other people, the concept would have little hope except under conditions where most people around you are close relatives. Indeed that is plausible in small bands of close relatives but it would be diminish probabilities very quickly as the population grew. A large population cannot survive and reproduce where everybody is a sibling, child/parent, or first cousin. One generation later and the children of first cousins are now second cousins, then third cousins, and so forth.

For it to work, we can look at the math of betting again. The gene must conform to the rule pB>C. Here the cost, C, is the reduction in reproductive success to copy of the allele in the host performing the altruistic act, B is the benefit to the reproductive success of copies of the allele in the recipient of the altruistic act, and p is the probability of the recipient containing a copy of the same allele.

In the general population, p is pretty low. It gets higher the more closely related a person is to the host. The probability for related individuals has been understood since at least 1922 when formulated by Sewall Wright as the coefficient of relatedness, r. For siblings, parents, and children, r = ½. For grandparents and grandchildren, r = ¼. For first cousins, r = 1/8.

The explanation for the coefficient of relatedness is fairly straightforward. Every person has 23 pairs of chromosomes, one that comes from your father and one from your mother. When producing a gamete (sperm or egg), each parent’s pair of chromosomes separate and go into the gamete as a single chromosome. When the egg and sperm meet up, the two combine to make you. Hence, for any given gene each may have on a given chromosome, there is a 1/2 chance that your sperm or egg got that one instead of the other one. Since your (full) siblings also have the same two parents, there is also a 1/2 chance that the one you got is also the one they got. Hence r = 1/2.

There are many complicated relationships that offer a range of r values, including half-relatives (e.g., one common parent line). We can then replace p = r and reformulate the genetic bet then as rB>C. This formulation is known as Hamilton’s Rule.

To be fair, William Hamilton was not the first to recognize the math of this bet. Charles Darwin note the relationship in Origin of the Species (1859). J.D.S. Haldane famously joked in 1932 that he’d give his life for two brothers or eight cousins. Shortly after Hamilton, George Price provided more mathematical foundation to it and John Maynard Smith coined the term kin selection to describe it. The term applies to this concept because it immediately implies that such a gene could only propagate if it was more altruistic to closer relatives, as in Haldane’s quote.

Let’s look at Haldane’s math to see how this works. If a gene for altruistic behaviour contributes to sacrificing itself, we can approximate costs and benefits in terms of number of copies of the gene. Since the gene in the acting host ceases to exist, C = 1. If the host saves two siblings, r=½ and B=2, so rB = (½)(2) = 1, which is exactly equal to C = 1 If the host saves three siblings, rB = 1.5 which is greater than C = 1.

For kin selection to work, one must be able to recognize their close relatives. This seems like a lot of work for simple gene to code for both complex altruistic behavior and recognizing relatives. Thankfully, the gene doesn’t need to accomplish that on its own. It can piggyback on other genes that contribute to this whole process. Recognition of close relatives meets many other reproductive success needs that are more fundamental than kin selection. For example, feeding a child instead of somebody else’s child requires that a parent recognize its own child. In general, mothers have significant built-in mechanisms for this, including pheromones. This is an interesting topic on its own but not my intent here. For kin selection to work, it merely requires that its host has the capability to recognize close relatives.

One could then reformulate Hamilton’s Rule based on the expected relatedness of the recipient, E = (p_1*r_1+p_2*r_2 + p_3*r_3 + …)B - C > 0. Here each r is the coefficient of relatedness for a given relationship and the corresponding p is the probability that the recipient fits that relationship. The same formulation can combine multiple recipients, e.g., a sibling, a parent, and a cousin are saved by sacrificing your life for them. In that context, the Haldane example above should more correctly be re-written with B = 1 (rather than 2) but summed over the two siblings being saved as (½)(1) + (½)(1) = 1. That way you can add up all of the different relationships being helped. The reason we don’t do this is because Hamilton’s Rule isn’t really about the prediction of a recipient being related, but the outcome of the altruistic act as a result of how that person is actually related. That is, if in reality rB>C, then the outcome of the altruistic act is to increase the copies of that gene. What the person estimated their relatedness was before deciding whether to do the altruistic act, even subconsciously, is interesting but not the important point here.

We should also note that B and C are generally not in units of copies of the allele saved or sacrificed, but in terms of increased and decreased reproductive success, as the altruistic sacrifice need not be as extreme as giving up your life such that a complete copy is lost. But the examples are simpler if we can use whole copies as the proxy units in the extreme cases. Calculating reproductive success values for B and C could be quite difficult for smaller altruistic contributions, like giving a relative a piece of your food.

Epistasis

Here is where my Twitter argument started. A certain “independent researcher” (IR, for short) has been trying to demonstrate for years that Hamilton’s Rule is faulty since complex behaviours tend to result from epistasis. Epistasis is the phenomenon that the benefit of genes, in terms of reproductive success, can be dependent on the presence of other genes at other loci. We saw this earlier with the gene contributing to altrustic behaviour being dependent on the ability to recognize kin relationships to some degree of probability. It even more generally applies to the altruistic behaviours themselves as a single gene does not generally code for complex behaviours. Rather, such altruism would result from many genes working together.

IR’s position is that epistasis rules out kin selection and the concept of “selfish genes” in general, because it changes Hamilton’s Rule. In particular, for an epistatic group of gene alleles involving a number of genes, IR believes that Hamilton’s Rule should be modified to rªB>C. IR’s argument is that probabilities are multiplicative and uses the example of flipping coins. To get the probability of getting three heads, one needs to multiple the odds of getting one head three times, (½)³ = 1/8.

Indeed this is the correct calculation for getting four heads when flipping coins. In the context of epistatic gene groups, IR’s formulation calculates the probability of finding all of the genes in the group in the recipient of the altruistic act, thereby justifying the altruistic sacrifice from the perspective of the altruistic, epistatic group in the host making the sacrifice.

So is IR corrrect? At face value it could make sense. If a single gene were to code for the complex behaviour in one shot, the math of betting told us that the probability of the presence of a copy of that gene in the recipient is what warranted the altruistic act of the copy in the host. So why not apply the same logic to groups of genes. Should it not be that a probability of the presence of a copy of that group of genes similarly warrants the altruistic act of the group copy in the host?

To check if this makes sense, let’s look at it from the point of view of each gene in the epistatic group. Suppose three genes (G,H,I) contribute to a complex altruistic act of self-sacrifice, and their host has the opportunity to give its life to save three siblings. According to Hamilton’s Rule, the cost of losing 1 copy is C = 1, the benefit of saving three copies is B = 3, and the relationship coefficient to each one (aka, odds of copy being in the person saved), is r = ½, so rB = (½)(3) = 1.5 > C = 1, therefore it is a good bet and the act will tend to increase the number of copies. According to IR, (rª)B = (½)³(3) = 3/8 < 1, therefore it is a bad bet. What do the genes say?

Consider gene G. The probability of a copy being in any sibling is ½. That is independent of the fact that it is part of an epistatic group. Hence if the host were to commit the self-sacrificing altruistic act, gene G would lose 1 copy (itself) and save 1.5 copies of G on average. So gene G should vote for the host to do the act. The same is true for genes H and I. Each one should vote for the host to go ahead and commit the act since, on average it results in 1.5 copies surviving thanks to the cost of 1 copy disappearing, with an expected net value of 0.5 more copies than by not committing the altruistic act. That means all four would individually vote to do it.

Of course genes don’t directly vote in that way. What they do is code a little contribution to the act. Rather, the betting math describes what happens over time on average give a range of genetic behaviours. Remember that the probability p = W/N which is the number of times you win, W, out of N bets. Nature actually carries out the experiment of a large number of bets. Those genes that contribute to behaviours that increase W/N will, by definition, have more copies in the general population than those that do not. Hence, genes that contribute toward doing the altruistic act for close kin would tend to increase in copies and those that didn’t would have relatively fewer copies.

It is true that 3/8ths of the time this particular altruistic act may find all three genes (G,H,I) in the recipient simultaneously, and that recipient may have higher reproductive success than the siblings with only 1 or 2 of the genes, but that doesn’t change the effective votes, or statistical outcome, of the individual genes pertaining to the act.

Hamilton’s Rule still survives. The success of the group of genes corresponds to the success of all of the individuals in the group. The vote of an individual gene does not require the presence of the other genes in the group. The other genes are merely the environment of each gene that affects its overall contribution to reproductive success. If a particular gene is poor at reproductive success, it does not matter if it saves other copies in an act of altruism or saves itself by being selfish; it will generally not propagate because it doesn’t improve reproductive success. Even these genes would still prefer to save other copies, in the mathematical sense, but its future is still doomed.

Note that I said you may find the three genes only in 3/8ths of the siblings. This is not guaranteed, and perhaps not even likely. The 3/8ths calculation, and rª in general, assumes that the genes are copied randomly and independently from each other. This is not the general case. While it is true that any particular gene you have has a ½ chance of being in your full sibling, all other genes on the same chromosome go along with it. If your sibling does have this particular gene, then they have the same entire chromosome. If they don’t have it, they won’t have any of the genes on that chromosome.

Hence, if the epistatic group is entirely on the same chromosome, the chances that the entire group is in your sibling is the same as for one gene. That is, ½. It is only if all member genes are on different chromosomes that you get 3/8. If it is a mix, the odds of finding them together in your sibling will be somewhere between ½ and 3/8. To reiterate, the odds of finding them together is unrelated to whether the gene itself would vote to commit the altruistic act via whatever phenotypic compoenent it adds to the behaviour. In more direct terms, it means genes that code for committing such an altruistic act to save three siblings will tend to have more copies than genes that don’t.

So what exactly is rª then? Ultimately it is the minimum probability of finding the entire group of genes in the recipient of the altruistic act, assuming the genes are all on different chromosomes. Putting it into Hamilton’s Rule, as IR did, would correspond to an “all or nothing” case where the individual genes in the group actually reduce the reproductive success of the host unless they are found in the full group combination. To see why, let’s look at the betting math again.

Recall that rB>C means that any individual gene that gives up it’s own existence to save at least, on average, more than 1 copy of itself will tend to have more copies over time. To get this we used B and C in terms of number of copies. Really, B and C are in terms of reproductive success of the gene. It’s possible that more copies of a gene saved solely through this act of altruism could have lower reproductive success if saved than the copy in the host that is potentially performing the act. That is, not all copies are necessarily equal in their reproductive success.

A simple example would be a young person in the prime of their fertility saving copies of their genes in relatives who are on their death bed. Perhaps the altruistic act saves the lives of those close relatives, and saves more copies inside them than is lost in the altruistic host, but if they die a few minutes later from natural causes then the net effect of the altruistic act was effectively to remove one additional copy.

Hence, in principle if there were circumstances in which a gene is best to vote against saving more copies than itself under those circumstances, it would tend to reduce the altruism in general. IR's formulation requires this be the baseline case, otherwise Hamilton's principle would apply to each gene and they would vote for the altruistic act. IR's forumlation then effectively creates an exception case to this anti-altruism baseline, the exception being when all members of the epistatic group are present in the recipient, at which point the altruistic act then becomes a net benefit to the genes. Then, and only then, would IR’s modification apply. It requires *all* genes in the group be present or else no individual gene would vote for the altruistic act, i.e., it's all or nothing. Keep in mind that this simply defines a case where IR's fomulation could *possibly* apply. For it to fundamentally replace Hamilton’s Rule, this set of circumstances would have to be the only type of circumstances possible, which simply doesn't fit the mathematical possibilities. The chance of such a case happening at all are small and certainly doesn't eliminate the more common case where genes do vote for the altruistic behaviour following Hamilton's Rule.

Also note all of the circumstances required to replace Hamilton’s Rule with IR’s Rule, so the list of requirements for IR's formulationto apply include:

All genes in the group must be on different chromosomes.

The gene should generally vote against more copies of itself for a given altruistic behaviour it codes for.

The gene should vote for more copies of itself if, and only if, all other members of the group are in the recipient of the altruistic act.

No other cases exist except the above 3 circumstances.

Zuk explains well several important problems with this reasoning. First she points out that "paleo" doesn't describe an exact time or location. What our ancestors ate 10,000 years ago pre-agriculture was vastly different from what they ate 100,000 years ago (or even 30,000 years ago for that matter). Second, which humans? Are we talking about those who lived in Northwestern North America, Europeans, Asian, or African humans? On the coast or inland? Diets of our ancestors varied significantly over geography and time.

Another problem Zuk addresses is that we haven't stopped evolving and we are quite different from our ancestors in that respect. Lactose tolerance evolved only 5000-7000 years ago, for instance, and a large portion of the world population have such tolerance and gain nutrition from milk throughout their lives. Our microbiome has also changed significantly. These are the micro-organisms that live on or in us that affect our digestion.

There is also the problem of determining what they ate. While Paleo tends to shun grains, evidence suggests some of our ancestors ate a lot of it, but this brings us back to which early humans we're meant to mimic. Even if we could figure out their diet well, Zuk points out the problem that those foods generally don't exist anymore. Humans aren't the only ones who have changed over thousands of years. Most of the plants and animals they would have eaten either no longer exist or have changed significantly, possibly because we ate them. (See below for explanation.)

The interview also includes a few side-bar analyses that include comparing Paleo to other diets and noting that, while it is an improvement over the average North American diet, it is far from the healthiest and is left lacking in a few areas such as too much cholesterol and too little vitamin D. In particular, there is little to no evidence that Paleo helps you lose weight and certainly none for the claim disease cures or preventions. Ultimately she points out that if we want to know how to eat healthier then we should be looking at how our bodies are now, not how things were in some vague time and location in the past.

I find Dr. Zuk's approach to the Paleo diet very refreshing and highly complementary to my own approach of looking at "natural" and "organic" food through the lens of the food itself (rather than the humans) in my 2011 article, Jus Naturale (a play on Latin for "natural law" and French for "natural juice"). My primary point was that plants and animals didn't evolve to be healthy for us; to the contrary it was the plants and animals that were not healthy for consumption by others that survived and reproduced better as per the basis of natural selection.

It was only with the invention of artificial selection (domestication, breeding) that we were able to turn the tables and artificially select the healthiest foods to reproduce the most (by our hand, on farms) while killing off the least healthy variants. We still have some troubles with some foods like potatoes that have a high content of natural pesticides, but we're still getting better at it. Direct genetic modification can improve upon this by taking out the random variation part of breeding.

I may just have to pick up Dr. Zuk's book. It sounds like quite a read.

(2) He offers no means to determine who is a "good" person or any practical means of doing so, other than obvious restrictions on known criminals which exists already and does almost nothing.

(3) He confuses proximate choices to own a gun under current circumstances with ultimate solution of changing the current circumstances. You can argue for banning guns while owning one until it happens.

(4) He essentially admits to being irrational about protecting his family, that he'd rather put them in greater danger of being killed as long as it means they are killed under his preferred circumstances.

(5) He still makes one-sided hypothetical circumstances where only the good people have guns and the bad don't.

(6) He builds his beliefs on contrived thought experiments and cherry-picked anecdotes while completely ignoring the reality of how violent crime actually happens statistically. Ironically, he accuses others of not addressing how it actually happens. It essentially comes down to his error in thinking only of the worst case at the expense of the likely case. This is the very definition of irrational fear.

(7) He believes himself to be above the statistics with no evidence to support this. While he makes arguments about why that might be, he neglects all of the ways in which being rational and competent do nothing for the risks and just doesn't accept that he could be irrational about being rational.

(8) He completely fails to address his contradiction for the dangers in a world without guns with his belief they should be kept safely at home. In fact, he deepens the contradiction by giving great detail on the risks of carrying a gun in public.

(9) He continues to promote a more expensive and less effective option for solving just school crimes over the option that would be the less expensive and more effective option of significantly solving the greater violent crime problems.

(10) He doesn't address how culture of violence changes, the norm cascades that I mentioned in my last criticism as described by Steven Pinker.

(11) He never realizes that he is part of the problem; that his belief that significant change can't happen is a self-fulfilling prophecy because people like him are making such change harder.

(12) He doesn't address Steven Pinker's description of how violent culture changes via norm cascades.

Here are the issues he addresses one by one with my criticisms:

1. Other countries have solved the problem.

In my prior criticism this topic was a main thread throughout, and particularly critical of him not addressing it at all. He has redeemed himself a little by addressing it, sort of.

Harris takes a defeatist attitude that it can't be done in the U.S. because of there are too many guns already and the Second Amendment requires amending to get rid of them, both for which there isn't sufficient political will in America.

The problem with this line of thinking, as I pointed out last time, is that it is circular reasoning. There isn't political will because people like Harris argue that it is futile and so if you can't beat 'em, join 'em. And people like Harris argue that it is futile because there isn't political will. It is a self-fulfilling prophecy.

When called to task that this defeatism is hypocritical with his promotion of atheism, which also has a huge uphill battle in America, Harris suggests

With respect to guns, I need to make a practical and ethical decision about whether or not to own one, given my specific security concerns and the level of violent crime in the society in which I live. This is not the same as deciding whether or not to write a book criticizing religion.

This is a strawman argument. It is one thing for Harris to own a gun for protection and quite another to argue that they are a good idea in general. Harris could write about how important it is for politicians to deal with the gun problem, the value of a massive buy-back plan and of amending the constitution, and he could do this while owning a gun and be entirely consistent and not be a hypocrite. Doing what is best under the current conditions and arguing for those conditions to change are different things.

This is common in a race to the bottom, like I have demonstrated many times with the Prisoners Dilemma. You can act on your proximate (local) best interests now while recognizing, and arguing, that everyone's ultimate (global) best interest is served better forcing everyone (including yourself) to chose another option. This may be unintuitive, but it is not contradictory.

So does Harris argue for the politicians to do this? Does he try to convince the public to support such efforts? No. He tells them all it is futile and it is better to give up and just arm and train everybody, especially schools. He makes it very clear (especially in answer to #3 below) that he does not support banning guns. It isn't just pragmatism of his current situation, he actually believes a world without guns is a worse place, despite the overwhelming evidence to the contrary.

Harris is a proponent of the options that costs far more money and lives. (As I pointed out in my prior article, armed guards in every school amounts to the order of $10 billion per year in perpetuity, yet he finds it unrealistic for a $100-$150 billion buy-back program over a decade or so.)

He then repeats his mistake of last time, arguing that some crimes are worse in other countries. He points to higher rape rates in Sweden and Australia and assaults in other countries. Yet those are not statistical arguments since, as I pointed out last time, the U.S. is consistently within the top two or three on rapes, and is by far in first place in assaults leading to death. Understanding why Sweden and Australia are high in rapes is a good case study, but not statistically representative of a correlation for guns versus rapes.

Personally, I suspect that it has more to do with the definitions of rape in those countries and/or the comfort and support that women have in reporting them, as many have pointed out. (Sweden is known to have very wide definitions of rape. Generally, there is an inverse correlation between cultural misogyny and reported rapes.) But again, that is all irrelevant since the statistics are clear that more guns doesn't lead to less violent crime.

In fact, Harris repeats his single-sided delusions. He suggests that "we might want to give Aussie and Swedish women some guns". Great, but that also means their rapists also have guns. More likely it means more raped and dead women.

2. Rarity of violence, even in U.S.

Harris here address the criticism, one I also brought up, that he contradicts himself by arguing the rarity of such violent acts means gun control laws aren't necessary, yet argues for the benefit of arming oneself and preparing for the worst.

His counter-argument is that preparing for the worst, even if unlikely, can be beneficial in the sense of confidence. He goes about this by describing the value of learning Brazilian Jiu-jitsu (BJJ) and then extends it to gun ownership.

The problem with this line of thinking is that it ignores the cost-benefit analysis. The benefits of BJJ may very well be as he describes, and may have many other benefits that include health, focus, and discipline. The question of whether the cost is worth the protective power, but whether it is worth all of the benefits.

Guns, on the other hand, do not have the same costs and benefits. Certainly there can be a confidence in having a gun (though a false confidence), but that is about it for benefits. Whether talking societal or personal, more BJJ does not put you at greater risk whereas more guns do. Even personally, no other person can make use of your BJJ but they can of your gun. There are no game theoretic problems with BJJ like there are with guns. A burglar who sees you gun, who might not have otherwise harmed you, now has to kill you or else be killed. This is not a problem with BJJ. If BJJ were to spread it is certainly possible that a burglar could know BJJ, but that doesn't make them any more likely to kill you. In fact, learning the BJJ is more likely to provide them a discipline and mental focus that reduces their likelihood of even being a criminal.

Harris admits at least part of this problem with his argument, but doesn't incorporate as any sort of invalidation of such an analogy even though it does invalidate it. Instead, regarding the increased risks that guns bring he claims:

I don’t think these broader statistics apply to me (and I don’t think this judgment is the product of a reasoning bias). Just as I can say to a moral certainty that I’m not going to open a meth lab or start a dog-fighting ring, I can say that I’m not going to commit suicide or murder my family. There are people who experience much more chaos in their lives who cannot honestly say the same. Such people should not own guns. 

I'm in awe of the arrogance and stupidity of such an argument. This is a re-statement and an extension of his first article in which he says he sees nothing wrong with judging himself to be rational. OK, fine. Let's unpack these statements to see what they include.

(1) Sam Harris is rational by his own judgment.

(2) Sam Harris is not irrational about being rational. In other words, he declares that he is not subject to the common problem whereby "people fail to realize the irrationality of their actions and believe they are acting perfectly rational, possibly due to flaws in their reasoning".

(3) The statistics of the risks of guns are actually separable into "those with chaos in their lives" and "those who are rational with less chaos", that these statistics only apply to the former, and that Sam Harris falls into the latter.

(4) Statements #1 and #2 above will not change for Sam, or if they do he will be aware of the change and will discard the guns.

(5) Statements #1-4 also apply to anybody who has potential access to his guns.

(6) Sam never makes mistakes when under pressure of rapid decision-making, such as deciding whether the person breaking in is a burglar (with a gun), or his spouse, children, friend, or even somebody that fits similar categories but is at the wrong house.

(7) Sam never makes mistakes as to keeping his guns under safe control such that somebody, especially his daughter, might access them.

(8) Sam is omniscient about who may have access to his guns and their applicability to all of the above statements as well.

Harris has had his turn at cherry-picking stories for fear-mongering so let's try a few on him. How about the story of Michael Griffin. He was shot dead by his father, a retired police officer with 42 years of experience. Michael lived there but apparently his father thought he was upstairs asleep and so assumed he was a burglar. Is a 42-year police veteran well enough trained and rational enough in the use of weapons? Did he have "too much chaos" in his life?

How about Tyler Giuliano, a 15-year-old shot by his father after sneaking into his house. As it turns out, Tyler was apparently wearing a black mask and nobody knows why. Seems his dad did everything rational in the situation, yet he killed his son.

Or how about the completely rational people doing completely rational things that get killed solely because of their rational use of guns. Andrew Lee Scott, 26, simply answered his door with a gun ready and was shot by police. The police were incompetent in not announcing themselves and the mistakes that lead them to Mr. Scott's house, but that doesn't make him less dead. Was Mr. Scott's life full of chaos? What rationality or training could have saved him from this outcome? What would Mr. Harris have done differently from his own training that would have saved him?

How about John Adams, 61, shot by police after they mistakenly broke into his house and he went for his gun to protect his wife, thinking they were burglars.

Then there is Jose Guerena, the former Marine and Iraq War vet, gunned down when police mistakenly busted down his door, entered his house, and shot him 60 times after finding him with a gun, presumably for protection against the home invaders he thought were coming in. Was this Marine not sufficiently trained? Was he irrational? Was his life in chaos?

These "death by incompetent police" problems are so wide-spread, particularly with no-knock warrants, that Indiana has even had to legalize shooting police who invade your home, as a means of self-protection. The irony is, perhaps, that this just makes it more likely they will kill you because now you can shoot without first determining if they are even police, and so the police are even more likely shoot back.

Of course none of this applies to Sam Harris. These cases don't apply to him. He claims to keep his many ready-to-shoot guns near entries and in lock boxes. (See answer 7 below.) There couldn't possibly be a risk with those, like perhaps some gun safes that can be opened by a three-year-old. No, Sam is rational and well trained, unlike police officer Ed Owens whose three-year-old son managed to open his gun safe and accidentally kill himself. If only Ed had been more rational and better trained, like Harris.

Ah, but these mistaken cases too are rare, right? I'll let the rational Mr. Harris answer it himself:

I do not believe it is irrational to prepare for very low-probability events which, should they occur, would produce the worst suffering imaginable for oneself and those one loves.

By Harris' own point, it is rational to prepare for mistaken shootings, mistaken police invasions, and kids accessing guns by keeping those guns the hell away from your house.

As to whether the broader statistics apply to Mr. Harris or not, I welcome his citation of references that demonstrate that his situation does not apply, except that he notes there aren't any. He has simply assumed they don't apply.

3. Non-lethal weapons

Although not part of my criticisms, others have brought Harris to task on the use of non-lethal weapons like tasers and pepper spray to address some of his points. His answer is that they have their uses but aren't good enough:

But their limited range and cartridge capacity, along with other vagaries of their operation, makes them (in my view) inadequate for home defense.

Just to clarify, Harris thinks that a range of 15 to 35 feet is insufficient for home defense, so he expect to make life-or-death decisions from even further away and not make any mistakes about whom he is killing. This is sheer delusion. As to the "other vagaries", that is a bit vague.

At least in this answer he does fully admit that he would fully support a ban on guns if there were a non-lethal option that worked as well. Just to be clear, from a utilitarian perspective, Harris believes that the benefits in the incredibly rare case of danger that only a gun can help more than offsets the risks associated with both owning that gun and in allowing others to access guns. This too is highly delusional. There isn't a lick of evidence to support this. Even the cases where a gun can help don't even make up for the risks statistically, yet he believes the small subset of those cases where other means are insufficient somehow does. I recommend remedial probability and statistics courses for Harris.

4. Violence against women

Harris responds to criticisms, mostly from Sean Faircloth, that guns are disproportionately used in violence against women. (This was not one of my original criticisms, but is a valid point from Faircloth).

Harris' answer returns to his prior appeal to emotion and fear from his first article:

As someone who was raised by a single mother, and as the father of little girl, I tend to view all questions of self-defense through the lens of what will enable a woman to protect herself from a man who is bent upon raping and/or killing her.

Oddly, the idea of making guns rare and thus reducing the danger to women seems to escape him. He proceeds to reiterate the argument that guns are an equalizer, even directly calling it true:

I will be accused of peddling NRA propaganda about guns being “an equalizer.” But it’s not propaganda if it’s true.

Except that it isn't true. Yes, of course you can construct a scenario where a woman being attacked by a man can defend herself with a gun. It is most convenient if you make sure that he doesn't have a gun in the scenario, but assuming he does and they are both equally trained you have now equalized her chances of survival to about 50%. Unless, of course, he was never going to kill her but has to now in which case she's made her chances of survival much worse.

But contrived scenarios aren't truth. They are hypothetical. Truth is statistical. Do readily available guns put women at greater risk or less risk. The statistics are clear that it puts them at greater risk. That is truth. As I pointed out last time, they do not equalize; they escalate.

5. Slippery slope of firearms

Another criticism I missed was the "why not a tank" argument. Harris' response is that only sufficient stopping force is necessary for crime and nobody needs a tank, rocket launcher, or mass destruction for such purposes. I agree, which is why I never brought it up myself. But it is with great irony that he words it as he does:

Once again, the fault lies with an unwillingness to think about how violent crime actually occurs.

For him to say this while he is continually unwilling to understand how violent crime actually occurs is incredibly sad. The key word change says it all. Harris thinks his thought experiments trump reality. The problem is that his thought experiments are limited in scope and not at all representative of what really happens most of the time, including violent crime, accidents, and mistakes causing death.

6. Why does he have more than one gun for protection

Apparently some critics suggested he shouldn't need more than one if just for protection as he claimed. This is not one of my criticisms. His answer is that he keeps them at locations around the house so that one is reasonably nearby should the need arise.

That is at least consistent with his home defense arguments, though following his prior article he might want to carry one around to save those women being raped he might witness and/or multiple attackers while he is out.

Of course, having more around the house significantly increases the risks of somebody getting access to them, including his daughter, or of using them mistakenly and getting killed for it. But since Harris is oblivious to how to calculate risks, apply probability and statistics, or perform cost-benefit analysis, I see no reason for him to address it for this circumstance either.

7. How can he be both secure and responsible?

Harris originally pointed out that he is responsible with guns and but most of his original arguments relied on having a gun ready to go. Some critics took him to task on this contradiction. His answer is a lockbox he can open in seconds.

The criticism is somewhat related to my original criticism on Sam's Imaginary World Without Guns where people running home to get their guns isn't helpful to his hypothetical victims. However, the lockbox doesn't solve that critcism, so I'll leave them unrelated.

As far as the lockbox, he'd better hope that no unauthorized people, like his daughter, ever learn the combination (or find the key). And to repeat myself, surely he couldn't possibly have one with a defect that could be opened by a three year old who then might die from it. (Watch the video on that first link if you want to be scared about how easy it can be to open these.) Of course Harris is a well-trained expert on lock boxes too so he couldn't be mistaken about the safety of his system.

8. The "swimming pool fallacy"

This one refers to the Sean Faircloth criticism, that the outcome is so rare that it should not be dealt with, referring to the risk of drowning in a swimming pool. Faircloth correctly points out that you need to do a cost-benefit analysis, as I have suggested throughout my two articles critical of Harris, where the costs are the risks of harm (probability of each consequence) and the benefits are the values gained, such as enjoyment of a swimming pool (or gun). This is very similar to #2, and a criticism I brought up as well.

Harris' response is that he meant the effort should be proportional to the problem, or of deaths. This is close to the reasonable answer for prioritization but misses the mark, which I suggested it a function of cost-benefit, risk, and uncertainty.

The problem with his answer is that he doesn't apply it to gun control or to his own arguments. He says gun violence is low risk, implying that the effort to fix it should be smaller than our other efforts. But gun control generally isn't a big effort. It costs little to ban certain types of guns. It costs little, relative to what is spent on other problems, for background checks and registries.

Even the suggested one-time $50-$150 billion buy-back program to get rid of the 200-300 million guns in the U.S. is miniscule compared to the estimated $100 billion per year cost that gun violence causes in the U.S. Granted, it is unlikely that it will all go away and pay for itself in only one year, but realistically the savings in economic costs would pay for such a program on the order of a decade or perhaps less.

Now how about applying the same "proportional" effort to Harris' suggestion of mass arming and training or his support of the idea of armed guards in every school. As I showed last time, and he agrees in this article, that alone would amount to about $10 billion per year just for the schools, every year in perpetuity. What savings would it generate? OK, assuming it worked completely as suggested and stopped school massacres completely (but not theaters!), there would be some economic savings (not to mention lives). But given how little those rare school shootings are to the total economic costs, such a program will never pay for itself.

In other words, the massive buy-back and banning of guns (except for hunting or other acceptable reasons, as in the Australian rules), would be much cheaper and pay itself back. It is a proportionately small cost compared to, say, addressing the swimming pool problem. In fact, it is a net earner over the order of a decade or so.

9. Trading murder for assault

Harris was criticized for pointing out other countries that had greater assaults and implied that guns would have helped there, even at the cost of more homicides.

His answer here is that he doesn't know the value of trading, say, two homicides for 400 assaults in terms of human suffering, but that there should be some scaling. That might be a valid argument if there was a clear, statistical trade-off. But as I pointed out last time, there is no such correlation. The U.S., with all of its guns, is near the top in assaults as well, is far on top in terms of assaults causing death, is at or near the top in rape, and is at or near the top in most violent crimes. Adding guns doesn't trade one violent crime for another. At best it increases violent crimes in a few areas and does nothing in others; at worst it generally increases all violent crimes.

To achieve his argument, Harris cherry-picks the one Western country, the U.K., with a worse assault rate. Outside the U.K., the Western countries with fewer guns tend to do far better than the U.S. The statistical correlation is actually in the positive direction: more guns leads to more assaults. The U.K. is the exception, not the rule.

Harris again pulls out irrelevant anecdotal evidence for his case, quoting an interview with convicted assaulters Steve and Nev who do not have to worry too much about their victims having guns. So based on this cherry-picked anecdote comes his belief that

I would rather that these brutes be obliged to worry that their next victim might have a gun.

As I've just shown, that doesn't make any statistical difference, and in reality it more likely means more assaults, and more deadly assaults (where the U.S. is 3-4 times everybody else). Harris is generally aware of the problem of also arming Steve and Nev though but claims it can be done:

I wouldn’t want Steve and Nev to have guns, of course—and gun control advocates will insist on the impossibility of arming good people without simultaneously arming the thugs. But I’m not entirely sure this is true. Starting from scratch in a country like the U.K., it seems that it should be possible to keep guns out of the hands of violent felons while allowing responsible people access to them.

Where is Harris' evidence for this? Who has succeeded in doing this? It's not like people walk around with signs saying "I'm a bad person". How does Harris suppose we find out? What would Harris add to the list already included in the Gun Control Act of 1968 that bans the sale of arms to people previously sentenced to more than a year (essentially all felonies), fugitives, drug users (whether convicted or not), known to have a mental defect, illegal immigrants, dishonorably discharged from the Armed Forces, has denounced their U.S. citizenship, has been served a court order declaring not to own one, or has been convicted of even misdemeaner domestic violence.

If this hasn't stopped the "bad" people from getting them, what will? Let's also not forget that a large segment of violent criminals had no prior record of them being a "bad" person. In some states, up to 50% of felony murderers are first-time offenders. Department of Justice statistics show that about one third of all violent offenders with a gun are first-time. (Table 7 of the linked report shows this, with 31.1% of 155,195 state + 31.8% of 3952 federal first time offenders convicted of violent crimes with a gun, compared to 28.4% of 360,564 state plus 38.4% of 9866 federal recidivists currently convicted of a violent crime with a gun.)

The first-time offenders can't be stopped by "good guy" checks, and the recidivists aren't stopped by existing "good guy" checks. So what does Harris propose? So far, nothing.

Perhaps more embarrassing for a trained philosopher like Harris, his assertion relies on cartoonish essentialism, the idea that people have an "essence" that is good or bad and it is not hard to determine which is which. Where is the evidence for that. I'm not saying that there isn't some component of probabilistic or causal relationships from genetics and/or environment, only that you can't just lump people into universally "good" or "bad" for determining if they should get a gun. Most people fall into the grey area between and change over time. If Harris' means something other than past records then his belief is passing uncomfortably close to Minority Report style "pre-crime" psychic prediction.

Perhaps an even harder problem than determining who is good or bad is how to keep all of those guns from getting into the thugs hands via indirect routes. The other two thirds of violent criminals with guns who are recidivists have no problem getting those guns despite it already being illegal.

Making guns rare goes a long way to solving that problem by both making them hard for thugs to find and by making them very expensive due to scarcity combined with high legal penalty for being caught with one and/or selling one. Harris doesn't even address this issue in allowing "good" people to have guns, even if we could figure out who they are.

He finished off by repeating his "defeatist" claim:

In any case, all of this is beside the point in the U.S. Here, Steve and Nev already have guns, and no one has a plan for taking them away. Given that fact, we must decide how difficult we should make it for law-abiding people to have guns as well.

To which I have the same response as in #1, that he is confusing two issues: immediate need for guns under the current circumstances and general solution to the problem of guns. If the U.S. is going to continue to allow easy access to guns, then there is a case (though statistically wrong) for a personal value of guns. But that does not preclude also promoting and supporting methods to fix the problem of ubiquitous guns by banning most of them and removing most from circulation, thereby changing the circumstance so that there is no case for the personal value of guns. Harris has fallen into the Prisoners Dilemma trap and can't see it.

10. Contradiction of rape witness and Seal Team Six

This was a criticism of mine that was so obvious that I turned it into open mocking. He argued that in a world without guns a rapist could keep a crowd of witnesses at bay with a knife yet argued even a trained Navy SEAL would need a gun to fight off more than one attacker.

His response is no better. It essentially comes down to every criminal being better trained and more courageous than everyone in the public sphere, hence the (well-trained) rapist could keep the (cowardly) crowd at bay whereas the (cowardly) Navy Seal could not fight off the (well-trained and willing to die) criminal crowd. His answer is that a gun gives you range and therefore courage.

As to why it doesn't also give the criminals range, he is mute. Let me spell it out for you, Sam. Guns escalate violence; they don't equalize it. A gang with guns will more likely kill you, especially if you have a gun too. Showing people beating each doesn't change that. A rapist will a gun will kill you faster than you can determine if you should shoot him. (Harris himself explains why this is so in #11.)

Following Harris' argument, since these are all career criminals they are better trained with guns than you. They are a better shot and less hesitant at pulling the trigger and killing somebody. Guns do not solve this problem; they escalate it.

In case you were thinking of arguing this is why people should get trained better with guns, so that they aren't at a disadvantage to the criminals with guns, the same argument applies to Harris' point about criminals knowing how to fight with knives. In a world without guns, if knives are a risk then getting better training on them is the solution, or self-defense in general. Making guns more ubiquitous makes you less likely to survive due to your lesser skill, not more likely.

11. Delusions that statistics don't apply to Sam Harris

He was called to task for suggesting statistics don't apply to him, as I have done in #2 above. His response has three parts. First he separates out the unstable elements and claims to be among the stable side, yet can quote no statistics to support that even the "stable" aren't still at greater risk.

Second, he admits that he actually could be delusional, but he doesn't see any reason to believe so. This is, apparently, as opposed to the delusional and irrational people who do see the reason to believe they are.

Third, he states that even if he actually is putting his family at greater risk of being killed then it is still rational:

I am willing to incur some additional risk to be better able to respond to a very low-probability, worst-case scenario.

Under what line of reasoning is this considered rational? That is no different than the rationality of spending $10 in gas to drive to a further gas station to save $1 in gas. I can't find a single definition of rationality that this would fit. He would rather increase the odds of having himself and/or his family die or be harmed as long as those harms of the more likely sort of deaths by guns, and not the rarer sort. WTF?

I've already linked to the definitions of irrationality above and in my prior article, so I won't repeat them other than to say working against your own best interests is a key part, as is completely failing to notice that you are being irrational.

What is Harris' explanation? Well, he does other things that are riskier, like taking them skiing. Again this fails the cost-benefit test, also described as the "swimming pool fallacy" by Sean Faircloth. You get enjoyment out of skiing and the risks are low, so the benefits outweigh the costs. The benefits of keeping guns for the worst case scenario do not outweigh the costs of the added risks to their lives; that is the very meaning of "greater risk" in the statistics. In an effort to keep his family safe he has made them less safe. In an effort to keep them from being killed, he has increased the odds they will be killed. These are net odds, already accounting for the cases where guns actually do save lives.

Mr. Harris has essentially admitted here that he is being completely irrational, without using that word. He just seems to think there's nothing wrong with being irrational because he judges himself to be rational.

12. Contradiction between protection outside the home and being against ordinary citizens carrying guns

This contradiction was one of my criticisms as well. Harris' answer is, well, he doesn't answer it at all.

What he does do is explain in greater detail why he thinks carrying a gun in public is much more problematic and prone to error and danger than using it for protection at home. Agreed. But this doesn't address the contradiction.

Let me make it clear what the objection is. In his first article, Harris claimed that a world without guns is a worse place because rapists with knives can hold off a group of witnesses, which he claims is solved if the witnesses have guns. He also claimed a gang of attackers could easily overpower an individual, even a member of Seal Team Six, unless that individual had a gun. Yet that same problem exists if those rape witnesses and Seal Team Six member leave their guns at home. Harris' world with guns is no better that his world without guns as far as the very examples he used to demonstrate what is wrong with the latter.

In fact, his world with guns is worse because he's increased the odds that the rapist and gang now have guns. He has created worse world, not a better one, following his beliefs. (This too fits the definition of irrationality.)

And he has no answer for this contradiction.

13. Calls to ban guns

Harris originally claimed that gun control advocates were not calling to ban guns, but just assault weapons. His critics have pointed out that many people do suggest a ban or heavy restriction on most guns except for hunting rifles and a few exception where demonstrable need exists. The Australian gun laws of 1996 are a good example of this and many have suggested it as a model for the U.S. I too suggested such restrictions and buy-back in my criticism.

Harris clarifies that he was talking about politicians and pundits. He also makes some important clarifications of his position. He does support most efforts on gun control that make getting a gun as difficult as getting a pilot's license, even more than many public gun control proponents ask for, and this makes him different from the NRA.

That clarification should be the main theme of his articles. Instead, both articles spend the vast majority of their effort arguing for the value of guns and ease of access for "good" people. Where are his detailed arguments for such laws? I request Mr. Harris to provide a detailed article explaining the reasoning behind those laws and why he supports them. That would actually be helpful. But he still doesn't acknowledge that many people do want an Australian-style ban and buy-back.

Politicians are the wrong measurement stick. They have constraints that keep them from openly taking such a position. If they said they wanted to amend the Constitution and ban guns they'd only get votes from people who support that. If they dial it back a bit they still get those votes as well as the votes of people who generally want gun control but think it is unrealistic to go further (like Mr. Harris). What policies they support behind the scenes is really up to them, but we can't ignore political limitations in vocalized policies.

Harris also repeats his error of confusing the issue of a proximate solution in current circumstances versus an ultimate solution to change the circumstances. He points out that although Gabby Gifford and her husband, Mark Kelley, are strong gun control proponents they own guns, implying a contradiction. As I have pointed out several times, this is not contradictory since the two are independent decisions. If the circumstances change due to policy change, that you support, you can revisit your personal decision to own one.

Harris finishes with a false dichotomy whereby he says the only options he sees for school massacres are either doing nothing or putting armed guards in every school. He doesn't even entertain the notion that a less expensive buy-back of guns and a comprehensive ban could work.

At this point the conversation has gone full circle as now he is back to his self-fulfilling prophecy. If he, and other public intellectuals, laid out the logic of how such a change would solve the problem, as it has in other countries, and convinced enough brave politicians, then it could happen just as in the norm cascades that brought about the end of racial segregation, voting rights and equal rights for women and minorities, and the near elimination of capital punishment in the Western world. It is Harris' defeatism that justifies his own defeatism. I wish he addressed the norm cascade point, or at least how cultural violence does change.

The situation is very much like the saying, "Whether you believe you can do it or believe that you can't, you are probably right." That applies culturally as well. Some people are ready for the hard work of convincing and bringing about valuable change. People like Sam Harris get in their way.

It would be more helpful if Mr. Harris wrote about his support for those gun control laws he does support. It would be more helpful if Mr. Harris wrote about how getting rid of the guns and changing U.S. gun culture, if possible, would make the U.S. a stronger, safer nation. By helpful I mean "leading to an increase of well-being", which by his own moral landscape beliefs would be a better moral position. Instead, he just perpetuates the problem.

Only a few more hints before the reveal. She got married January 10th, 2009 and is still very happily married. OK, so now you might be thinking she's not likely a Hollywood celebrity. You are getting warmer. She has two beautiful children with her husband. Final clue: Her name is the same as the emotion she brings to her husband's life.

Give up? Her name is Joy, and she's been my wife for four years today. She is a real Emergency Room nurse. (Technically it is the Emergency Department. They have many rooms.) Joy is a beautiful person and the love of my life. I am incredibly happy I found her.

As I said at our wedding, I don't consider myself to be lucky to have found her. Luck had nothing to do with it. I simply set my standards to the highest they could possibly be and then I kept searching for somebody who could meet them, never settling. It was therefore inevitable that I'd find her.

So how is it that I have joined this exclusive group that consists of Channing Tatum, Bradley Cooper, and now me? No, it is not because Channing + Brad = Chad. Am I sexier than Ryan Gosling because she turned him down and married me? No, she turned Ryan down when they were tweens/teens in Ottawa, Ontario, and in the same dance class together. (He lived in nearby Cornwall. Oh, and that magazine ad? She was about seven years old.)

The contradictions in Harris' article come from selective application of his arguments. He argues that a world without guns isn't such a great place and selectively uses examples where a gun would help to demonstrate this.

A world without guns is one in which the most aggressive men can do more or less anything they want. It is a world in which a man with a knife can rape and murder a woman in the presence of a dozen witnesses, and none will find the courage to intervene. [...] The fantasies of many martial artists aside, to go unarmed against a person with a knife is to put oneself in very real peril, regardless of one’s training.

In the very next sentiment he directly contradicts this:

A world without guns is a world in which no man, not even a member of Seal Team Six, can reasonably expect to prevail over more than one determined attacker at a time.

So let's get Harris' arguments straight. In Sam's Imaginary World Without Guns a single rapist can hold off a dozen witnesses, and presumably his victim, while raping his victim, with just a knife, but even a member of Seal Team Six can't hold off more than one determined attacker without a gun. Do you see a little bias in this baseless assertion?

From his rapist example you might think Harris is a big proponent of citizens walking around with their guns. After all, it wouldn't do any good for those dozen witnesses to have their guns locked up safely at home when they unexpectedly came across that rapist. Guess again:

Carrying a gun in public, however, entails even greater responsibility than keeping one at home, and in most states the laws reflect this. Like many gun-control advocates, I have serious concerns about letting ordinary citizens walk around armed.

You've lost me, Sam. If Sam's Imaginary World Without Guns is such a bad place because a rapist can hold off a dozen witnesses with a knife, then doesn't that also apply to Sam's World Where Ordinary Citizens Don't Walk Around Armed? Harris claims "I have never had any illusions about how quickly the police can respond when called." but he does seem to have illusions about quickly rape witnesses can run home and get their safely locked up guns, load them, and return. Hopefully Seal Team Six carries theirs around at least, otherwise they too would have to run home before taking on their gang of determined attackers.

Harris' obliviousness to his contradictions here seems to be a result of having the delusional belief that guns are a great equalizer:

A world without guns, therefore, is one in which the advantages of youth, size, strength, aggression, and sheer numbers are almost always decisive.

Sure, this makes sense as long as you have the gun and they don't. If a rapist with a knife can hold off a dozen witnesses, a rapist with a gun can hold off more and do so more effectively. The fantasies of action heroes aside, no man with a gun, not even a member of Seal Team Six, can reasonably expect to prevail over more than one determined gun-wielding attacker at a time.

In reality you get the worst of both worlds. The rapist can keep more witnesses at bay, the group of determined attackers will now shoot you instead of breaking your leg, and you have the added bonus of increased risk from your own gun at home from accidental shooting, access by children, and, as Harris admits, "resorting to deadly force in a circumstance that would not otherwise have called for it". Contrary to Harris' original assertion, a world with guns is one in which the most aggressive, and mentally unstable, men can do more or less anything they want and take out dozens of people, including children. That doesn't happen so easily in a real world without guns.

Guns don't equalize the problem; they escalate it. It becomes a literal arms race where the most aggressive, unstable, and criminally minded dictate the terms and any individual who doesn't have as good a gun or training is in greater danger than if they were in a real world without guns (as opposed to Sam's imaginary one).

Harris is fully aware of that inequality even in the gun world but is completely obliviousness to how it destroys his equalization arguments. He spends much time discussing different types of guns, bullets, and their usage including this point about rifles:

There is, in fact, no marksman on earth who can shoot a handgun as accurately at distance as you would be able to shoot a rifle fitted with a scope after a few hours of practice. This difference in accuracy between short and long guns must be experienced to be understood. Having understood it, you will in no way be consoled to learn that a madman ensconced on the rooftop of a nearby building is armed merely with a “hunting rifle” that is legal in all 50 states.

Once again this is a wonderful argument for the utter uselessness of arming everybody. Even if everybody carried around personal handguns and could stop that rapist (who didn't have a gun for some reason) or that gang of determined attackers (who didn't have guns for some reason), Harris argues here exactly why they'd be useless against a madman with a hunting rifle on a rooftop.

It is utterly futile to try to stop madmen determined on taking people out by making guns more ubiquitous. The best that we can do is to make it as difficult as possible for such a person to acquire that gun, to increase the likelihood of catching at-risk people by having greater review on of who does buy them, and reduce the likelihood of anyone thinking of doing it.

A real world without guns is actually far safer than Sam's Imaginary World Without Guns. And here's the difference in my argument versus Harris': you don't need to take my word for it. Such worlds essentially exist in reality. Other Western nations have solved this problem.

Some Statistical Facts About Guns

Let's look at Harris' rape scenario. Rape statistics put the U.S. consistently among the top 2 or 3 of all Western countries. It was #3 in 2010, the latest year in the data, and #2 in 2009. In terms of guns per capita the U.S. is by far in first place with double the rate of the second place, the small peaceful country of Switzerland with mandatory military training of citizens, and triple that of the other Western countries. The point here isn't the inverse correlation from what Harris proposes. Quite the contrary, the point is that there isn't really a correlation. Harris' rape scenario is a fictitious contrivance. More guns don't reduce rape.

It's a similar story with Harris' assault on Seal Team Six. The U.S. is by far the #1 country for assaults leading to death of all OECD countries, about three to four times that of other OECD countries. (Keep in mind too that this is assaults from any kind, not just guns. Other countries don't compensate with assaults from knifes or other weapons.) If there is a correlation, it is the opposite of Harris' assertion. Seal Team Six is better off in a world without guns. You just have to be talking about the real one and not Sam's imaginary one.

Yes, of course these other countries aren't without guns. That is not the point. The point is that Harris's arguments about what happens in a world without guns are utterly delusional. At best the guns have no effect and at worst actually make the rapes and assaults worse.

What about at the personal level then. Clearly in the greater society a world with guns is worse off but that's including all of the criminals, mentally unstable, and madmen. What about if we just look at those good, upstanding citizens. Harris suggests

it seems to me that there is nothing irrational in judging oneself to be psychologically stable and fully committed to the safe handling and ethical use of firearm.

The truly stable people are personally better off, right? Think again. The statistics are clear:

Having a gun in your home significantly increases your risk of death — and that of your spouse and children.

And it doesn’t matter how the guns are stored or what type or how many guns you own.

If you have a gun, everybody in your home is more likely than your non-gun-owning neighbors and their families to die in a gun-related accident, suicide or homicide.

Furthermore, there is no credible evidence that having a gun in your house reduces your risk of being a victim of a crime. Nor does it reduce your risk of being injured during a home break-in.

Perhaps Harris should look up the definition of irrational. In economics it refers to the state in which "people's actual interests differ from what they believe to be their interests." In psychology it is defined as "the tendency to act, emote and think in ways that are inflexible, unrealistic, absolutist and most importantly self- and social-defeating and destructive." Yes, Sam, it seems to me that there is indeed something irrational in what you suggest. It's also interesting in the study of irrationality that "people fail to realize the irrationality of their actions and believe they are acting perfectly rational, possibly due to flaws in their reasoning".

Some Facts About Irrational Fear

So why does Harris bring up rape and assault at all if they are uncorrelated or reversely correlated from his assertion? As with most gun apologist arguments, he is resorting to the appeal of fear. The antithesis of a "bleeding heart liberal" is a "fear-mongering conservative", and Harris brings it out in spades. In fact, (stealing one of Sam's lines) I have never seen an gun apologist conform to left-wing caricatures of himself with such alacrity.

Harris starts by putting you in the place of a helpless rape victim or witness, hesitating without courage to do anything. If only you had a gun. Next, you are assaulted by a gang and are helpless to do anything about it. If only you had a gun. Then he shows us a Houston PSA video re-enacting an office shooting scenario and asks us to "Imagine being one of the people in the Houston video trapped in the office with no recourse but to hide under a desk." If only you had a gun. Heck, why not just show everybody how a gun can save you in those scary scenarios. Sure enough, Harris pulls out a video of a motel clerk shooting an armed robber

He swiftly dismisses worries about crossfire with ease by appealing to what should be your real fear,

If you found yourself trapped with others in a conference room, preparing to attack the shooter with pencils and chairs, can you imagine thinking, “I’m so glad no one else has a gun, because I wouldn’t want to get caught in any crossfire”?

Feel the fear. Let it grow. Now think about your daughter:

As the parent of a daughter in preschool, I can scarcely imagine the feelings of terror, helplessness, and grief endured by the parents of Newtown. But when I contemplate atrocities of this kind, I do not think of “gun control”

Are you feeling it yet? Are you one of those horribly bad parents who would let their child die by supporting gun control? Think of your terrified child. If only there were guns to stop this guy. Feel how the fear disappears by imagining holding a gun in your hand and what power it gives you to stop atrocities, to stop rapists, to stop gangs, and to save your daughter.

If you listen closely you can just hear the ghost of Charleton Heston exlaim, "Good, I can feel your anger. I am defenseless. Take your weapon. Strike me down with all of your hatred and your journey towards the dark side will be complete!"

I've always wondered why "conservatives" are always so scared. (I use the quotation marks here because modern conservatives have little to do with actual social, fiscal, or political conservatism.) America used to believe in things like "live free or die" and "It is better to die on your feet than live on your knees." But now a large potion cowers in fear, ready to get their gun if somebody steps on their lawn. Forget how socially corrosive it is. Forget how it increases your real danger. Get your gun because it give you power and relieves your fear.

Apparently that's what matters. American conservatives are driven by irrational fear. Harris seems to be falling in this direction. I use irrational here as defined above and with multiple applications. It is irrational to worry about the circumstances Harris describes. As even he points out such cases of violence are very rare and are decreasing even in the United States. If it's so rare then rational people don't need guns to protect themselves. It is even irrational to think that personally owning a gun in your home actually makes you and your family safer. It is also irrational to think more guns in general lead to a safer world, or that a world without guns is worse, when international statistics clearly show that is not the case.

The irrationality can be well-hidden though. Consider if everybody else had guns and your chance of being brutally murdered is, say, 30 in 100,000. If you get a gun you might be able to stop a few attempts, perhaps reducing it to 25. Therefore it might seem rational that more guns improves safety. Except it is relative to a baseline where everybody else has guns. If you got rid of most of them your risks might drop to 2 or 3 per 100,000. Arguing for a local minimum in the face of a much lower global minimum is irrational. ("Global" and "local" here are referring to the mathematical relationship of guns and violence, not geography.

Harris and other gun apologists can't use rationality to justify the availability or personal choice of guns. This is why they must resort to contrived and cherry-picked scenarios where guns have been, or could conceivably be, helpful. That is a problem with the politics of fear. They make you chose options that put you, your family, your co-workers, and especially your children at greater risks, not less.

Some Facts About Irrelevance

Hypothetical and anecdotal stories about how guns might help are worse than just biased cherry-picking, they are completely irrelevant. They serve only to spread irrational fear that cause people to make choices that put them in greater danger. The only thing that matters in discussing control is whether there are policies that actually reduce crime and risk, and are the costs of such policies worth the improvement. That's it. Anything else is a misdirection.

This is where Harris fails terribly. He spends much time creating Sam's Imaginary World Without Guns stories. He spends much time telling scary stories to raise your fear and then how much better you'd feel with a gun in your hands, even when in reality it'd statistically put you at greater risk. He spends much time discussing different types of guns, bullets, and equipment that only serve to undermine his other arguments about how arming people can help. But he never really addresses policies, their real-world (statistical) effectiveness, and their costs.

When Harris does use statistics he uses them selectively:

Fifty-five million kids went to school on the day that 20 were massacred at Sandy Hook Elementary in Newtown, Connecticut. Even in the United States, therefore, the chances of a child’s dying in a school shooting are remote.

He continues by describing how violence is decreasing and how the types of guns that some gun control advocates are chasing are a negligible portion. As I've pointed out already, it's interesting that he only applies this to why gun control activists shouldn't be so worried, but not to why arming and training everyone and posting armed guards at schools is foolish. But this isn't just trivia; he has a point to make:

Of course, it is important to think about the problem of gun violence in the context of other risks. For instance, it is estimated that 100,000 Americans die each year because doctors and nurses fail to wash their hands properly. Measured in bodies, therefore, the problem of hand washing in hospitals is worse than the problem of guns, even if we include accidents and suicides.

Harris makes a huge, but common error here. He confuses the relative size of the problem with its priority. If this were reasonable, it would be foolish to address any problem other than the worst one. But that's not the correct metric. Priorities are a function of cost-benefit-risk analysis, perhaps easiest understood by utilitarian mathematics. You need to address the costs of solutions with their expected benefits, both scaled by their uncertainty and risks. You also hedge your bets by diversifying -- addressing multiple problems in parallel.

A small problem with a simple, inexpensive, and guaranteed solution should be a parallel priority to a large problem with a complex, expensive, and uncertain solutions. Heart disease is the biggest killer. Does that mean we shouldn't address car safety until we've cured heart disease? Let's not put safety railings on bridges because only an insignificant number of people accidentally fall off bridges.

Of course that line of thinking is ridiculous. You put in cheap and simple solutions where you can. Let's look at that hand washing problem. I can't find any reference supporting Harris' claims of 100,000 deaths due to improper hand washing in hospitals. The closest I can find in the literature is references to about 80,000 deaths due to nosocomial (hospital acquired) infections. That is, infection from all sources in hospitals, for which improper hand washing by doctors and nurses is just a small fraction. Putting that aside, there actually is appropriate effort at implementing policies and solutions to improper hand washing, including "wash-control" policies like a five-point checklist and alcohol gels that show under UV light where you've missed to help educate workers.

There is one important difference to note about the efforts to implement "wash-control" policies: there aren't any dirty-hand apologists getting in their way and arguing that there are worse problem so people should stop over-emphasizing clean-hand policies. They simply move forward by testing various policies and implementing those that work and where the benefits are worth the costs.

Harris makes no effort whatsoever to look at the costs and benefits of any solutions. He goes to great length to describe how armed guards and well-trained citizens can actually protect us from madmen with guns. It's not just a transient thought; he means it as a viable solution:

This leads me to believe that if we care about minimizing the harm caused by the next school shooter, we should focus on stopping him at the doors of the school. To be sure, hiring enough guards to protect our nation’s schools would be a daunting task. The security industry is notorious for poor quality control, and there is even reason to worry that some police officers have insufficient training with their guns. But it is clearly possible to hire as many competent guards as we want, should this become a national priority. This is entirely a question of money, not of whether it is possible to enlist, train, and equip 100,000 highly qualified men and women to protect our children.

Ignoring how he showed that we actually can't stop them if they have a rifle on a rooftop (another contradiction), what are the costs of such a solution? The National Center for Education Statistics reports that in 2009-2010 there were 98,817 public schools, or roughly the 100,000 that Harris mentions. The average salary of an armed security officer is $31,000 as of this date. Just in salaries alone this amounts to $3.1 billion per year. Add in training costs, equipment, management, and other overhead and you are probably talking well over $10 billion per year. This is just for the armed guards at public schools. We haven't even begun to account for guarding any other place or the training or arming of individual citizens which will likely dwarf this. I would not be surprised if it was on the order $100 billion per year.

For comparison, there are somewhere between 200 and 300 million guns in the U.S. Quality guns average about $500 to $1000 new depending on the type. Let's say about half that amount for the average used gun. That means American taxpayers could buy back all of the privately owned guns for a one-time cost of about $50-$150 billion, or somewhere between about one to ten years worth of costs of the Harris-approved NRA suggestions, depending on whether we're just talking schools or including greater arming and training in general. In the long run, arming everyone is far, far more expensive and far less effective.

Yet Harris is at a complete loss here:

Any effective regime of “gun control,” therefore, would require that we remove hundreds of millions of firearms from our streets. As Jeffrey Goldberg points out in The Atlantic, it may no longer be rational to hope that we can solve the problem of gun violence by restricting access to guns—because guns are everywhere

So spending on the order of $10-$100 billion per year in perpetuity is "just a question of money" when it comes to protecting our children, but spending a similar amount once, amortized over a few years to a decade, and to make them a lot safer as part of a permanent solution, is somehow not rational. I can't be the only one who finds problems with such baseless and biased assertions.

Granted, just buying them back won't solve the problem outright. Several million new guns are bought every year, though that number would likely drop organically simply from the reduction in danger. Although the risks already outweigh the benefits of ownership, with most guns removed from circulation the irrational fears that drive the behaviour would be more transparent.

Such an "all-in" solution would likely have to include legislation strictly controlling the production, sale, and purchase of guns of all kinds, and the massive buy-back would probably have to be made mandatory. Gun apologists would probably argue that criminals would not give back their guns. Of course that relies on the a caricature of the reality. People who use guns in serious crimes aren't usually career Mob hit men; they are otherwise normal people who snap, drift into criminal behaviour, or have mental problems, and have easy access to guns that were never intended for criminal activity. The Newtown shooter stole his guns from a family member who wasn't a criminal. He wasn't a career criminal himself. Furthermore, significantly reducing the number available in circulation or harder to get approval means it is harder for criminals to get them and raises their prices in an illegal market.

Of course this level of effort would require significant political will and possibly constitutional amendment should any of the inevitable Second Amendment objections succeed.

Is this realistic? In principle, yes, but in practical terms is seems unlikely only because of America's irrational political atmosphere and gun culture that Harris is contributing to, not because it can't be done. Australia had success in this area. After the Port Arthur massacre in 1996, where 35 people were killed and 21 injured by a mentally unstable man with two semi-automatic rifles, the government implemented a buy-back program that got about 640,000 guns out of circulation of all kinds. It lasted one year. (Given Australia's population of about 22 million, that's about the equivalent per capita of the U.S. buying back 10 million guns in a single year.) These reforms also removed semi-automatic and pump-action shotguns and rifles from civilian possession.

Furthermore, the laws put strict requirements on the acquisition of firearms. The purchaser must be at least 18 years old, hold a firearms license, must have secure storage for it, and must obtain a permit to acquire. That permit has a 28 day waiting period with background check (for the first permit) and requires an approved and demonstrable reason for the purchase. Acceptable reasons relate to pest control, hunting, target shooting, and collecting. Self-defense is not an acceptable reason.

How did it do? According to published review of progress, in the 18 years prior to the legislation there were 13 mass murders or about one every year or two. In the 14 years (and possibly 16 years as of this writing) following the legislation there were none. As a possible bonus, firearm-related deaths (particularly homicides and suicides) appeared to decline faster, though it is debated whether or not the overall declines in homicides and suicides can be attributed to these laws directly.

So here is a country that had a much, much smaller problem than the U.S., had the political will to solve it, and did so. Was it the buy-back program that eliminated the massacres? Did the acquisition restrictions do it? Was it the social outrage that did it? Or was it just ongoing social change?

There are two problems with this line of questioning: it doesn't matter, and it assumes they are independent and separable. Imagine if we could show that the laws didn't actually have any contribution to the end of massacres. Would you change them back? Why? Do these laws not make sense in their own right? When it comes to weapons, we have to have limits on the destructiveness that any individual citizen can perform, and in the gray areas we need some means to demonstrate responsibility of ownership. We don't let any person own any weapon.

A good measure of the destructiveness is the amount of power and energy output for the effort input by the user. Sure, a knife could be used to kill somebody, but it only has an destructive ratio of 1. All of the energy and power come from the human operating it. In a gun, the output energy comes from the chemistry of the gun powder, and the output power (energy per unit time) comes from the firing rate. The input is the effort to pull, and hold, the trigger. The input efrot is negligible to the output energy an power.

Even a small child can kill an adult with a gun. I can pull out a video too. (Warning, the video shows exactly what I say it does.) Sure, stupid parent, right? That's the point. A small lapse in judgement or memory, like taking the clip but forgetting to check the chamber, is very easily fatal. With guns it takes little effort to kill, and with semi-automatic and automatic weapons it takes little effort to slaughter many people in seconds, perhaps even before those armed guards are anywhere near the scene.

Sure, a madman can still use even a knife to try and kill many people, but it is much more difficult, far less likely to succeed, and far less likely to happen in the first place. It is probability that matters, not possibility. This doesn't stop Harris from falling into the cognitive trap; he pulls out the Chinese knife attacks on schoolchildren as a counter-example. He does pay lip service to admit that guns are more efficient, but dismisses the objection by simply saying "but the truth is that knives are often lethal enough".

What does that even mean? Does Harris not believe that many more people would have been killed had they used guns, especially semi-automatic? How rare are these knife attacks compared to gun massacres? How easy is it to stop somebody with a knife compared to a gun? Apparently saving lives doesn't matter to Harris. What matters to him is the binary condition of whether a weapon can kill or not.

The arguments of possibility are irrational in that they work against our own interests. The problem of violent deaths is a statistical issue. Reducing the number of deaths is the goal. The inability to make that number zero is irrelevant. Cutting the number is what matters. Even ignoring the emotional arguments, a purely utilitarian consideration says even one statistically reduced death is of value.

Of course the reduced deaths need to be put in the context of the lost benefit of having easily available, highly dangerous weapons. So what are they? What are the benefits that society, or even an individual, gets out of having easy access to semi-automatic or fully-automatic weapons, or any of the kind banned in Australia? How do those benefits in aggregate compare to the lives lost due to allowing them? Now what about the same question for handguns? Remember, the statistics are pretty clear that you and your family are at greater risk by owning one, and certainly the statistics show that the society at large is worse off. From a purely utilitarian cost-benefit analysis of risks, what are these net benefits we'd lose? The NRA and Harris have none to offer. That is why they both rely on anecdotal stories that are unrepresentative of the statistics.

Hence even if these Australian laws didn't directly cause the drop in massacres to zero, that is not itself a justification for having them repealed. You need to identify a net cost, or a benefit that is lost that exceeds the benefits gained.

Some Facts About Symbology

Harris worries that gun laws would be merely symbolic. If that were true, so what? Who says that symbols are not of value? Symbolism doesn't stand alone; it represents the values of a society and acts to facilitate social change. Many laws are symbolic and difficult to enforce in practice. Do oaths and perjury laws keep people from ever lying? Of course not. But the statement of values in these rules and the threat do affect people's behaviour. People don't lie and cheat just because they can find loopholes. Even if it were just symbolic, which likely isn't true, symbolism has real effects.

Can Australians actually buy guns for protection by lying about it? Sure, but it means admitting to yourself that you are a liar who skirts the law for personal gain, and you have to maintain the false appearance of your honesty, making sure you never slip up. That alone affects how people behave.

More importantly, legislation is an important part of changing culture. Harris notes, as others do, that America's gun culture is the primary problem. I think everybody can agree on that. I'm unaware of anyone who argues a few simple laws alone will solve the problem. But it also doesn't happen by magic.

As I mentioned earlier, Harris points to Steven Pinker's recent book The Better Angels of our Nature that effectively demonstrates that violence is decreasing even in America. I too am an avid reader of Pinker's and highly recommend his books and speeches. Let's let Steven tell us how that violence decreases. We can get it from his own mouth in a video of the Global Empowerment Meeting 2012 put on by the Center for International Development at Harvard University. Although I do recommend the whole video, I've started the link at 11:41 as this is where he starts to talk about how such social changes happen. He starts by summarizing Adam Smith and other Enlightenment thinkers on how reason works to bring about social change.

Of particular interest is the segment starting at 16:20. Pinker describes the process of social change by norm cascade and describes, as an example, the detailed analysis by Andrew Hammel on the decline of capital punishment in his book Ending The Death Penalty. Here is blueprint for how a norm cascade works to put an end to some behaviour X:

1. Intense controversy, with the majority favoring [X].

2. Elites, influenced by rational argument, defy popular opinion, push through abolition [of X].

3. Nothing terrible happens.

4. People and press get bored.

5. Politicians realize issue is no longer a vote-getter.

6. Political inertia: No one wants to re-open the issue.

7. People get used to it, favor the status quo.

8. Alternative becomes unthinkable except ...

9. Among radical fringe groups, whose extremism only cements popular consensus.

That is how reason overcomes irrational cultural norms. Other examples given include racial segregation, use of nuclear weapons in war, criminalization of homosexuality, women in the workplace and military, and we're even now seeing it in areas like same sex marriage and universal health insurance. Heck, the same process works with addressing traffic congestion using nominal tolls.

The problem is that people are naturally risk averse. It is easy to come up with imaginary reasons why things won't work (like Sam's Imaginary World Without Guns). Change is an unknown; better to live with the devil we know. Except that once the change is implemented and the benefits realized, people en masse tend to change their opinion even to the point where they don't think they originally objected to the change. Change is a feedback loop. The status quo can change as a result of changes in what people believe in, but you can change what people believe in as a result of changing the status quo. People aren't psychic and people aren't all statisticians or scientists; sometimes we need to see the change before realizing our imagined outcomes were fictional fears. Luckily we have other countries to look at, if gun apologists would only recognize that.

This is how a large part of human progress works and does so efficiently. We all benefit from the best ideas of a few. We don't all need to cure cancer. We only need one person to cure cancer and 7 billion people can benefit by copying it. It's the very basis of how scientific progress works. When reason shows that gun violence is greatly solvable, supported by statistics and the fact that the rest of the Western world has solved this problem, we can all benefit by copying it. The problems are the public inertia to do nothing (other than express outrage one way or the other) and the active resistance to doing that which reason shows.

This doesn't mean that it is a short-term or easy solution. Norm cascades can take decades. Nor does it have to an immediate all or nothing. U.S. gun legislation need not start as comprehensive as what Australia did, but the framework of unifying laws and compounding power of large gun buy-back programs can help start a norm cascade of further laws and social change.

Gun laws also are not mutually exclusive from other solutions. Certainly mental health needs to be better addressed. At-risk youth programs can help. Often overlooked is journalism reform, which may or may not involve legislation. Part of the reason for massacres such as Newtown is the anti-hero infamy offered by 24-hour news programs. Many of these deranged individuals taking out innocent people, especially young children, are lonely and suicidal nobodies who want to stop being nobodies. Going out in a "blaze of glory" and being at the top of "worst" lists is being a somebody.

Even the sane can understand that somebody committing suicide alone, feeling that the world thinks of them as a pathetic loser and being remembered by nobody, may be less attractive than committing suicide while feeling powerful and being remembered as a powerful person making a mark on the world, even if bad. This is especially true for young men with an innate, but overactive, need to demonstrate their status in society, even when such drive conflicts with the cognitive consequences. There is no such thing as bad press, as they say.

Far more people can name the Newtown and Columbine killers than can name a even a single victim. The Aurora theatre shooter even has his photo as an internet meme. (I refuse to name the killers, itemize their arsenals, or rank their body counts. They're just all pathetic losers.)

Some Facts About Gun Apologists

I know that Sam Harris is the wrong villain here. There are far worse gun apologists. Harris' article at least notes some of the absurdity of extreme NRA arguments and the validity of some points of gun control proponents, and even the limits in some of his arguments. As I've noted, he only tends to pay lip service to those objections, as if only to cover his ass, without actually incorporating them into the balance. He also isn't against gun control laws entirely, suggesting basic laws are sensible (without going into much detail). And to be entirely fair he does aim some of his arguments specifically at ignorance within the gun-control community, including significant errors in the press.

Caveats aside, the bulk of Harris' arguments are exactly as I've presented here, suggesting that guns are good for protection and safety (statistically they are not), that a world without guns would be worse (it would not, by his own arguments), and that putting armed guards in schools is perfectly rational and a good idea (it is not) whereas it is probably irrational to try to tackle the number of guns and gun culture problems (it is not).

The difference with Harris is that NRA spokesmen, right-wing pundits, and politicians are easy to dismiss. They have an agenda and are fundamentally ideological by nature and by duty to their employers and core voters. By contrast, Sam Harris is a respected intellectual. Critical thinkers like me look to him for rationally thought out arguments when it comes to public debates such as this. It is therefore especially disheartening to see him do such a sloppy job with such blatant bias and irrationality, and not once address the most important point that undermines almost all of his arguments: the rest of the Western world has solved this problem already, and more guns isn't it.

[Update]: Harris has responded to many of his critics including many of the criticisms I have given here. Unfortunately he just compounds his mistakes and biases. I'll address them soon, but the epitome of his irrational bias here is summed up in this line:

But I don’t think these broader statistics apply to me (and I don’t think this judgment is the product of a reasoning bias)

and this line:

There are people who experience much more chaos in their lives who cannot honestly say the same. Such people should not own guns. 

The problem with top-down planning is if you only use top-down planning. Likewise there is problem when you only use bottom-up trial and error. The details matter. There is no general rule regarding bottom-up versus top-down. The quality of information, experience, and the complexity of the system in question are all important details as is your flexibility in adapting to changes including updating the top-down plan.

The problem with Taleb's thesis, as presented in Ridley's article, is that it takes an extreme all-or-nothing perspective. Sure, one can demonstrate some successes of trial and error without planning and some failures of top-down planning. But there are also cases of bad outcomes with trial and error (hence the "error" part) and successes of top-down theory. Bottom-up trial and error is even sometimes impossible or expensive compared to top-down planning, especially in cases where you only get one shot and failure is catastrophic. For example, good policy on climate change doesn't come from trying out different scenarios in the real world and finding out which ones lead to environmental and economic catastrophe and picking ones that don't. We don't get do-overs.

Ridley comments on Taleb's biological examples, including the bottom-up trial and error approach of evolution by natural selection:

Biological evolution, too, is anti-fragile. The death of unfit individuals is what causes a species to adapt and improve.

Indeed, that is exactly how natural selection works, at least the "survival of the fittest" component of it, and as long as you define "improve" to mean in terms of the maximizing the local reproductive success of genes. (The results of sexual selection and genetic drift may not be considered "improvement" in some respects, for instance.) But that isn't the whole story. Dinosaurs may have been very robust (anti-fragile?) to their environment but they couldn't avoid being wiped out by an asteroid. We can avoid them because we have something they didn't: top-down planning. We can apply theory to find them. We can apply theory to predict which ones will hit us. We can apply theory to divert them.

Of course these are all fairly simple systems with robust and accurate models. Things like climate change and hurricane paths are more complex and only predictable in bulk properties. Yet still top-down applied theory and planning are quite successful. We know days in advance of hurricanes where to evacuate and prepare our resources, and what level of damage to be prepared for. This ability is new by applying scientific predictive theory. A century ago we had no warnings.

He even gets much of modern medicine wrong, claiming that it is a " 'craft built around experience-driven heuristics' that had to fight against entrenched, top-down theorizing from Galen and other wise fools." But almost all modern medicines derive from applied theory of various processes in the body and principles of evolution, and the new treatments are then tested to see if they do as expected (from theory) and if they have side effects that were not planned.

Even Galen is the wrong enemy here. While his peers were mainly divided into Rationalists (theory) and Empiricists (experimentation), among Galen's many contributions to early medicine was his regard for medicine "as an interdisciplinary field that was best practiced by utilizing theory, observation, and experimentation in conjunction." He was truly a valuable influence on the concept of science and optimizing the use of information from all sources.

Similarly Taleb seems to pick on Nobel Prize winning economist Joseph Stiglitz largely because of Stiglitz is a Keynesian economist who builds and applies theory versus Taleb's "street smarts" trial and error. Taleb's complains that "Globalization creates interlocking fragility, while reducing volatility and giving the appearance of stability." Among Joseph Stiglitz's complaints about globalization is that "the IMF has often called for policies that conform to textbook economics but do not make sense for the countries to which the IMF is recommending them." Stiglitz, like Taleb, is very much anti-textbook as far as economic policies, and very much against the destabilizing effects of finely planned tariffs, subsidies, and complex patenting systems. He is a big proponent of understanding the particular circumstances and doing what the empirical evidence suggests, unlike free-market economists (or "free-market fundamentalists", as he calls them) who have blind faith that simple ideological ideas and free markets will magically solve problems.

From my own perspective, the problem of fragility and destabilization is a function of purely maximizing efficiency, which markets often do well (but not always), largely by optimizing comparative advantage (specialization and trade) such that much of the needs of the world are compartmentalized and accomplished by those who are best at it. This optimization of efficiency comes at the cost of robustness to the very unexpected destabilizers that Taleb worries about. If all of the world's food is produced by a single region or group who is most efficient at it, a single failure of that region or group can starve the world. Robustness typically means redundancy which is contrary to efficiency.

As an example, many of the systems I worked on as a prime contractor for NASA were Criticality 1, meaning missions could fail or astronauts could die if the equipment or processes failed. This invariably meant triple redundancy such that catastrophe was at least "three failures deep", and that these redundant systems were independent of one another. We designed these systems by applying theory and demonstrated them through rigorous empirical stress testing, including random, unexpected events as Taleb's work warns against (and seems to think he invented).

Taleb seems to misunderstand this process. Certainly theory comes from experience, consistent with the second half of Taleb's thesis statement, but we also most certainly do put theory into practice, and with much success, contradicting the other half. This fact is rather obvious for we engineers and applied scientists. I don't mean that in a "common sense" perspective; I mean it is how we do our jobs. Essentially every bit of modern technology and medicine comes from applying theory top-down to generate the very things that we test via trial and error bottom-up. We don't just randomly try things. Heck, just recently Nate Silver repeated his smack-down demonstration of applying top-down theory to election prediction, kicking the ass of "street smart" pundits, as has sabermetrics in perfecting "Moneyball".

So how could Taleb be so far off? Is it because he has little understanding of science, engineering, medicine, biology, and even economics as "a former trader and expert on probability" and "a self-taught philosopher steeped in the stories and ideas of ancient Greece"? This is where it gets confusing, because I'm not entirely sure that Taleb says what Ridley is suggesting.

Taleb is more than just some former trader with self-taught Greek philosophy. Nassim Nicholas Taleb has been a professor at several universities (including Oxford), a hedge fund manager, a practitioner of mathematical finance, an adviser to the IMF, and writer of the books Fooled by Randomness and the best selling Black Swan which has been described as "one of the twelve most influential books since World War II".

Black Swan theory, the basis of Taleb's thesis, does not contradict with applying theory or top-down planning. Rather, it is complementary to such planning. Applied theory works on known behaviours and risks. Black Swan theory works on unknown behaviours and risks. Further confusing where Ridley notes

If trial and error is creative, then we should treat ruined entrepreneurs with the reverence that we reserve for fallen soldiers, Mr. Taleb thinks.

In fact, this sentiment is the subtitle of the WSJ version of the article. Yet this is in direct contradiction of Malcolm Gladwell's version of Taleb's views in the New Yorker:

We associate the willingness to risk great failure – and the ability to climb back from catastrophe – with courage. But in this we are wrong. That is the lesson of Taleb and Niederhoffer, and also the lesson of our volatile times. There is more courage and heroism in defying the human impulse, in taking the purposeful and painful steps to prepare for the unimaginable.

So which is it? Does Taleb think we should treat failed entrepreneurs with reverence for their risk-taking courage, or with disdain for not taking purposeful and painful steps to prepare for the unimaginable?

On policy the contradictions stand out even more. He hates top-down applied theory planning in lieu of bottom-up trial and error, even in the face of progress that is clearly a result of applying good theory and planning to drive such trial and error, such as is common in medicine and engineering. Once again, the dinosaurs would still be alive had they been able to do top-down planning for avoiding asteroids, since their trial and error genes had not supplied them with the capability of surviving such catastrophe. And yet it is this very sort of unexpected dinosaur-killing asteroid that Taleb thinks we should plan for by applying his theories.

When I compare Ridley's review with any other reporting of Taleb they seem to describe two very different points of view. Ridley's Taleb is filled with childish strawman reasoning like "You don't need a physics degree to ride a bicycle," or that he "systematically demolishes what he cheekily calls the 'Soviet-Harvard' notion that birds fly because we lecture them how to-that is to say". This version of Taleb is a simpleton who says tripe like

Planning is inherently biased toward delay, complication and inflexibility, which is why companies falter when they get big enough to employ planners.

or this gem of stupidity:

A law that bailed out failing restaurants would result in disastrously dull food. The economic parallel hardly needs spelling out.

That's because there isn't an economic parallel, at least with the recent bailouts. It is a bad analogy. There are thousands of independent restaurants and the ability to make meals at home, among a variety of alternatives. Of course there's no reason to bail out a failing restaurant. Now imagine if there were only a small handful of restaurant chains and in addition to food preparation they also controlled most of the national food production and distribution, and they are interlinked. Now a complete failure of one or several restaurant chains means starvation of millions. Now would you consider a bail out? Would you consider a set of regulations for reducing reliance for so much on so few? That is the "too big to fail" problem as applied to large and few banks that control the vast majority of lending and investment. Perhaps This Taleb needs some remedial lessons.

Yet on further review, he knows this. In the Black Swan he wrote

Financial Institutions have been merging into a smaller number of very large banks. Almost all banks are interrelated. So the financial ecology is swelling into gigantic, incestuous, bureaucratic banks – when one fails, they all fall. The increased concentration among banks seems to have the effect of making financial crisis less likely, but when they happen they are more global in scale and hit us very hard. We have moved from a diversified ecology of small banks, with varied lending policies, to a more homogeneous framework of firms that all resemble one another. True, we now have fewer failures, but when they occur .... I shiver at the thought.

Indeed, that sounds like the wise version of Taleb. Even he agrees that concentration and interconnectedness are the problem and we should avoid them. This is exactly the problem of instability by over-efficiency I describe as my view above. Yet Taleb seems to generally oppose regulations to keep this from happening, as Stiglitz and other Keynesian economcists recommend, and is opposed to bailing them out when they do fail. All his suggestions amount to us getting used to shivering.

There is so much contradiction in policy, seriousness, and level of intellectual content that I've thought either Ridley has completely misunderstood him or Taleb has gone off the deep end of bipolar disorder.

Sadly, these sorts of over-simplified comments appear to be consistent with his general views on economics and intellectualism. He actually does seem to simultaneously hate applying theory while providing recommendations on how to apply his theories. He simultaneously "believes that universities are better at public relations and claiming credit than generating knowledge" while being a Distinguished Research Scholar at Oxford University as well as affiliated with many other universities and performing much academic research.

Taleb comes off as either an eccentric genius or a narcissistic simpleton troll. The reality seems to be somewhere between the two. Just when he makes a good point for a specific circumstance he seems to follow it up with a non sequitur generalization aimed at a strawman caricature of how things are done.

I think I get what Taleb is trying to say. Relying only on theoretical models without regard for how they fit reality is unwise. It's just that he doesn't seem to realize this is not how people generally use models. We do usually make good use of them for top-down decision a directions and guidance, not gospel dogma. We do usually test them. We do usually limit their application to the assumptions built into them. We do usually change plans when the circumstances call for it. Perhaps the trading world that Taleb got his experience is a little different. If so, perhaps he should limit his criticisms to those areas. Generalizing them as he appears to be doing just makes him look ignorant of what people actually do, and a little self-contradictory at that.

He really does appear to want us to embrace trial and error while avoiding top-down planning and applied theory, unless it is his theories that we are applying for top-down planning purposes. If it's the Federal Reserve that sees the economic asteroid coming and has the engineering plan to divert it, it seems Taleb wants us to just accept our fate and suffer the consequences. At least he'll be able to profit from it using his Black Swan scheme.

What Ridley is suggesting in the article is much more. The WSJ title on civilization, focused on reference to Jason Collins' work, suggests this process of seeking greater wealth is what created civilization via innovation. That suggestion is, of course, far too over-simplified for the title question and not my current focus for comment other than to say there is a significant difference between causally related and explanatory value. Adolf Hitler's parents having sex was causally related to the Holocaust but their amorous activities offer no explanatory value to why it happened. Explanatory value requires predictive power. Sexual selection certainly contributed causally to the progress of civilization, but how much it can contribute to explaining civilization, and which parts of civilization, requires very detailed and thorough analyses.

The inequality question is a little more interesting and directly related. Combined with the article content, the question extends the conspicuous consumption hypothesis a step further to disliking inequality. Ridley does not spell out the exact hypothesis. He implies that dislike for inequality is a product of jealousy and envy brought about by this tendency for a few males to succeed over others. From the article:

Back in the hunter-gatherer Paleolithic, inequality had reproductive consequences. The successful hunter, providing valuable protein for females, got a lot more mating opportunities than the unsuccessful. So it's possible that men still walk around with a relatively simple equation in their brains, namely that relative success at obtaining assets results in more sexual adventures and more grandchildren.

Ridley summarizes this principle of relative envy over absolute utility as classic sexual selection:

It isn't the peacock with the big-enough tail that gets to mate; it's the peacock with the biggest tail.

There is certainly argumentative merit here. However, it only works internal to the facts brought up in the article. In the grander context is has some significant problems to overcome as well as competition for explanatory value.

Sexual Selection Mismatches

Ridley's hypothesis is based on male jealousy over not having the "biggest tail". The main problem here is that this cannot explain why women also dislike economic inequality. In fact, it appears that women dislike economic inequality more than men. For example, Millionaire Corner, an online magazine published by Spectrem Group, the "premier research and consulting firm in the wealth and retirement industry", performed a survey on the perceptions of income inequality. They posed four statements about feelings of differences in wealth levels among Americans. On the two statements that declared the difference in wealth levels was not a problem, women agreed less than men. On the two statements that declared it was a problem, women agreed more than men. The differences were consistent and significant on a relative scale.

Ridley's hypothesis seems to fail this twice: women should not dislike income inequality and women should not dislike it more than men. Sexual selection works by allowing women to differentiate males to pick the "better" ones (in terms of likely reproductive success via proxy signals). It predicts that women should want more differentiation, not less, and for it to be harder to fake. This is the standard explanation of why the peacock's tail is so long and ornate. Size, cleanliness, and symmetry are easy to differentiate and hard to fake.

An obvious first step in analyzing Ridley's sexually selected hypothesis is to evaluate it with respect to indicators and models of sexual selection, something Ridley failed to do.

A classic sign of sexual selection is sexual dimorphism - the tendency for large differences between the sexes such as very large males and small females or peacocks with long, ornate tails and peahens with short, plain tales. Looking at those survey results above we see that, although women consistently have a relatively bigger dislike of income inequality than men, they both have generally similar levels of distaste on an absolute scale. There are no answers with 10% women versus 90% men, for instance. Such a large difference would be a potential sign of sexual selection. Their similar absolute level is a contra-indicator of sexual selection and, for what difference is there, it is the opposite of what sexual selection would predict.

Another indicator of sexual selection is that the trait in question is used during mating rituals or other indications of attraction in the opposite sex. The peacock's tail is unfurled and displayed during mate selection rituals. Male bowerbirds build intricate bowers - basically their decorated houses and front garden - and females select mates based on the impressiveness of these bowers. Male birds of paradise often perform intricate,dance-like displays. 

When we're talking about humans, a pretty good sign of sexual selection in this respect is whether or not people of one or both sexes actually find a trait "sexy". That men find young women with soft, clean skin and certain body shapes attractive is a good sign that these may be sexually selected. Women's protruding breasts are a focus of men's arousal and sexual play (and often to the sexual enjoyment by women), a trait and behavior not found in other apes. (While breasts do become engorged in female apes when breastfeeding, it is temporary and occurs after impregnation, not as part of mate selection.)

Similarly, that women tend to find wealth, social status, humour, confidence, competence, and similar traits sexy is a good sign that these may be sexually selected.

So ask yourself: when making a dating profile, on a first date, or just promoting yourself to the opposite sex, is it normal to emphasize your position on wealth inequality? Is disliking economic equality something either or both sexes tend to find really sexually attractive in the opposite sex? I don't mean just alignment of interests such as religious beliefs, position on organic farming, or enjoyment of rock climbing; I mean is the dislike of inequality itself something we all tend to emphasize to attract the opposite sex? If not, this might be a sign it isn't sexually selected.

Another indicator of a sexually selected trait is its use in competition between members of the same sex. The big horns of various species of deer and goats offer good examples. Like the peacock's tail, these large, ornamental traits can become quite intricate and costly but, unlike the peacock, these are also used in direct physical competitions between males who, quite literally, butt heads to demonstrate their superiority directly.

The competition indicator is the basis for Ridley's hypothesis, that since conspicuous consumption is used by males to compete with each other for female attention and, indeed, that women tend to select men with wealth and resources. However, this is just standard sexual selection of wealth-seeking and display. Ridley's hypothesis is that the dislike of economic inequality is sexually selected, not just seeking it. Do males compete their level of dislike of economic inequality against other males? If there isn't a reality show emphasizing it, I have my doubts.

Finally, there are subsequent, secondary indicators of sexual selection of the above indicators. One example is juveniles practicing the behaviour or emphasizing the trait outside of selection rituals, such as those dances of birds of paradise, play-fighting among males, or caring for pretend babies (e.g., dolls) in young females. I'm unaware of either sex instinctively practicing the behaviour of disliking economic inequality when they are young. I certainly don't remember playing a game of "Occupy Classroom" as a youngster.

As far as sexual selection indicators go, Ridley seems to be zero for all of them.

Competing Explanations

Assuming the dislike of income inequality is even explainable via natural selection at all (see top of the article about explanatory value), there are better alternatives. The free rider and exploitation problems are much better predictors of disliking income inequality than sexual selection. These evolutionary and economic problems describe characteristics that allow one to exploit the work of others, benefit the same as others without paying the same cost, and fake personal superiority cues without earning them through competitive merit. (For simplicity I will refer to all of these problems as free riders.)

If exploiting and faking are easy then free riders can acquire more assets with less work and hence have much greater opportunity to mate. The exploitation problem in the context of sexual selection predicts that both men and women should dislike suspected free riders, and that women should dislike them more. Men will dislike them because, while they personally work hard to demonstrate their own status, free riders both exploit this hard work for personal gain and do so at the expense of opportunities for the honest men. However, this dislike for free riders is countered by men's interest in the ability to steal or fake it a little to improve their own status, making them a little less honest. (We all tend to cheat a little on occasion.)

Women, on the other hand, will generally dislike free riders because the very signal they are sexually selecting on requires honesty. Faking superiority or gaining it through non-meritorious means is not of value. If peacocks could easily steal the tails grown by other peacocks and wear them as their own, long ornate tails would no longer suffice as cues of meritorious superiority. They would not demonstrate that the peacock is free of parasites or is so superior that it can handle such a long tail and still not worry about predators. (The latter is the handicap principle of sexual selection.) Rewarding that parasitic behaviour means that more cheats will reproduce and so the signal loses its true value over time and hence would be selected less.

Women won't universally dislike free riders though, because the ability to acquire assets, even if stolen, still provides resources to their offspring. But it does so with risks of revenge by those males who were exploited so its value is limited compared to earning assets by honest merit. The equilibrium between value, non-value, and anti-value of cheats should tend to be strongly against it since truly superior males offer net better value than fake superior males.

We see this exploitative, free-riding behaviour throughout history. Slavery and private governance (monarchies, dictators, warlords, hierarchical churches) all work by individuals gaining personal value off the work of others and using their power of position to keep out competitors rather than by out-competing them.

Capitalism and democracy are not immune to such exploitation. Monopolies are one version wherein a business that manages to achieve dominance at some point, perhaps early in a new market, uses anti-competitive means to stay in that position. Such means can include predatory pricing, cross-market bundling, exclusive agreements across supplier and distribution chains, defamation of competition, espionage, protection rackets, and even basic threats such as burning down competing stores and factories or breaking the legs of sellers in your territory. This is why we have laws and regulations against such actions and anti-competitive measures are closely monitored. Superiority by out-competing in terms of cost, quality, and value are good for citizens. Superiority at anti-competitive behaviour is only of value to the exploiters and free riders and not to the general public.

Even within a well-regulated society there are means to free ride and exploit. The Ultimatum Game is one such mean. In the Ultimatum Game there is an opportunity for two players to earn a reward, say $100. Player 1 gets to propose a split of the money and Player 2 gets to either accept or reject it. If rejected, neither earns a cent. It is in the best interest of Player 2 to accept any non-zero offer; the small amount offered is still better than nothing. Hence Player 1 is best to offer very little and keep as much as possible, say offer $1 and keep $99.

To be clear, Player 1 has done nothing to earn the $99. They have not demonstrated 99 times the value of Player 2. Yet Player 1 gets rewarded 99 times the amount of Player 2 (or 9999 times if they offered only 1 cent). This is, essentially, free riding. They receive benefit disproportionate to their cost compared to others. Now imagine some work was involved in getting that $100. Player 2 could do all of the work and still get only $1 while Player 1 got $99 for doing nothing at all other than having access to the $100. It is the structure of the transaction that determines benefits, not contribution. (The dynamics of this game can get much more complicated depending on the circumstances. I discussed these variations in a prior article related to unions.)

The Ultimatum Game overlaps with the reality of labour markets. Corporate owners and managers are similar to Player 1 and employees are similar to Player 2. This is not to say they are identical. Owners and managers do contribute much to the success of businesses, as do labourers. However, the relationship also has the overlaid structure of the Ultimatum Game, especially taking into account the larger labour markets. (The linked prior article covers this in more detail.) It is impossible to separate the parts of wealth compensation to owners and managers that are due to the value of their contribution from those of the Ultimatum Game structure itself. It is really a mix of the two.

This explains the dislike of economic inequality in terms of differences perceived to be out of proportion to contribution. Few people, if any, seem to perceive a problem of relative wealth inequality when the comparison is between a person who works very hard and one who is lazy. The Atlantic surveyed a wide demographic of people on income distribution and found they were all fine with a reasonable level of meritorious inequality, but which is much less than the inequality seen today in America.

People inherently value rewarding merit. What they dislike is inequality that seems to be far out of proportion, such as when a CEO and senior managers receive large bonuses despite a company failing. The managers' jobs are to see that the company on a whole succeeds and despite failing at it they receiving compensation that is hundreds to thousands of times higher than labourers who are performing their jobs successfully.

The free riding problems also explain the dislike of economic inequality in terms of non-competitive tactics. While many anti-competitive activities are regulated, some regulations are removed or ignored either via ideological beliefs (e.g., libertarianism) or by influence of those who benefit such as quid pro quo campaign contributions and lobbying at the political level. They earn money not by competing for it but by influencing the rules and public beliefs that redirect public assets to themselves such as rent seeking.

Ridley's hypothesis fails in this respect too, then. If it is envy over the biggest peacock's tail that explains why everyone hates inequality then they should also hate merit. The peacock with the biggest tail actually has the biggest tail. But people by and large support a true, competitive meritocracy. For the most part, people support a level playing field. It is not a tied score they seek; but they want that score to be on merit and not because some players control the rules or the referees.

Then there is the fact that too much relative wealth concentrated with too few people collapses an economy. This is true even in evolutionary terms, and is especially true if the relative wealth is achieved through free riding. Free riders can only survive as long as most of the population are "honest" workers. If free riding takes off then there are more free riders fighting for less honest work to exploit. While zero free riding is unstable since a first free rider will have plenty of effort to exploit, the stable equilibrium will generally include only a small percentage of free riders, and there will be an evolutionary arms race of people getting better at detecting and punishing free riders while free riders get better and foiling detection.

Ultimately, I think the answer to Ridley's question is likely, "No, sexual selection does not explain dislike for inequality." It doesn't fit the predictions of sexual selection, there are other evolutionary drivers that do much better, and the dislike for income inequality may not have any genetic component anyway (beyond the trivial) in which case selection pressures have nothing to do with it. 

Upate [2012-11-19]:

Matt Ridley responded by tweet to suggest that sexual dimorphism isn't always true in sexual selection and it could be mutually selected, as he suggests at the end of his article. While this is true in principle it negates the analogy used about the largest peacock tail and requires significantly more detailed explanation, particularly for why women simultaneously do prefer the most successful males (which is sexually dimorphic) and dislike excessive economic inequality.

In the article, Ridley argues that women are attracted to wealth and conspicuous consumption, men instinctively seek it to maximize opportunities as a result, and yet "They [people] dislike (and envy) conspicuous consumption". That is a contradiction and sounds more like self-loathing. (Though, as a friend pointed out, it is possible to "hate the game" and not "the player".)

There seems to be a lack of explanation for the sexual selection of disliking inequality itself in either sex. Do men select it in women or do women select it in men? If men select it in women then it doesn't seem to actually have the intended effect of distracting women from the males with the "longest tails". Where is the selection pressure then? If women select it in men, why would that be? Where is the benefit to women, and where is the selection pressure since men actually do instinctively seek wealth.

Mutual sexual selection requires additional explanation on top of this. What gives sexual selection its name is that it is a trait in one sex that is selected by the other for choosing a mate. To be mutual sexual selection requires that it work in both directions and it doesn't even seem to work in either direction here. Ridley's suggestion is one of co-parenting in humans (unlike peacocks and peahens).

This certainly puts us further out on the limb of explanation, quickly approaching a "just so" story, but let's entertain it for a minute. The rare cases of monomorphic mutual sexual selection have been attributed to:

Non-adaptive genetic correlation, meaning one sex selects it in the other and the selecting sex carries the trait themselves since it doesn't affect the choices of the opposite sex. Here there is no an explanation for which sex has selected dislike of inequality in the other and the explanation given does make it clear that the interests of the sexes diverge: women should want more differentiation in men and men should generally want less of it.

Selection pressure for sexual indistinguishability, usually attributed to avoidance of competition in monogamous flock species. No argument has been made for this, particularly given the premise is based on dimorphic competitive differences of attraction to, and seeking of, wealth.

Weak or absent sexual selection, which is my argument here.

Simultaneous sexual selection resulting from similar male and female parental roles. This appears to be Ridley's suggestion. The problem is that it only applies if the trait acts as a signal for high quality parenting such as health (e.g., free of parasites), strength (e.g., handicap principle) or nurturing capability. Disliking wealth inequality doesn't fit this at all. In fact, it should arguably the opposite where parenting improves in both sexes the better each is able to acquire resources. If parenting was equal in hunter-gatherer days, both should be attracted to high wealth status in the other, not dislike it, as long as it is an honest signal.

The alternatives I propose seem to do much better. Men and women don't hate wealth or conspicuous consumption. They dislike it when it appears to be out of proportion to merit which you would expect from mutual selection pressure against allowing gaming of the "wealth as a competitive merit signal" system. This has nothing to do with jealousy or envy and everything to do with making sure the merit system sends honest signals and penalize free riders, something both sexes should want and women arguably should want slightly more (which they do) because men would want to keep some minor ability to cheat a little for their own sake. This is true even if it is a reasoned, cognitive signal rather than an innate, genetic instinct. It make no cognitive sense either to reward out of proportion of value given.

One detail that we should take from Ridley's monomorphic suggestion, both in the article and response tweet: it is clear he is suggesting that the dislike of economic inequality is, itself, sexually selected. He could have just meant that dislike is just a form of jealousy and jealousy is a general response of those who don't compare well to "top" examples of traits: women with small breasts may be jealous of women with bigger breasts, men of low status are jealous of men with high status, and so on. But that jealousy itself is not sexually selected and is cross-gender, so the article would be pointless. Men are not jealous that women have larger breasts than them. Both sexes dislike income inequality as a general principle, not specific to whether it is in men or women. Ridley's suggestion of monomorphism makes it clear that he proposes the dislike itself is a sexually selected trait.

[Update 2: 2012-11-19]:

A potential confusion occurred to me. The dimorphic behaviour whereby women prefer men with wealth and men instinctively seek it, as Matt Ridley describes in the artlcle (testing with photos, David Petraeus) is clearly a sexually selected trait discussed for many years. The proposed hypothesis in Ridley's article and blog title refers to the largely monomorphic behaviour of disliking economic inequality and whether it is a sexually selected trait. That is the one I believe doesn't work.

[Update 3: 2012-11-20]:

Something occurred to me when discussing this topic with a friend. None of what I'm proposing is reliant on my personal opinion of what is meritorious or "fair" compensation, nor on the economic and game theory tactics used to obtain it. I certainly have opinions on that topic. Rather, I'm suggesting that people using their own evaluation of non-meritorious wealth leads to disliking excessive economic inequality, and that such dislike itself is based on innate distaste for gaming the "assets-as-merit" signal.

The U.S. President is selected by electoral votes rather than the aggregate of individual votes (commonly referred to as the popular vote). Each state gets a number of electoral votes equal to their number of Senate seats, which is two per state, plus their number of House of Representatives seats, which is proportional to population.

Most states have a winner-takes-all policy where the candidate that gets the most votes (a plurality) within the state takes all of the electoral votes of the state. Maine and Nebraska are the exceptions where the two Senate seat electoral votes are winner-takes-all and the House seat electoral votes are distributed by Congressional districts. This doesn't really complicate matters since Maine only has two House seats and leans very Democrat and Nebraska has only three and leans very Republican, so they rarely differ from winner-takes-all and a few electoral vote errors wouldn't add up to much given there are 538 in total.

This makes modeling U.S. elections fairly straightforward. If you have polls in each state, or that are separable by state, you can simply apply the poll results to find the likely winner for those state electoral votes and then add them up. To get better estimates with more confidence you average multiple polls, correct for measurable biases, account for turnout rates, and weigh polls differently based on their own confidence values and how recent they are. Overall confidence in predictions depends on the confidence of each poll and how close the split is in each state. If the poll accuracy is much better than the gap between parties then you can be very confident in the prediction for that state.

Canada is a different beast. We don't elect our Prime Minister. We elect Members of Parliament (MPs) in ridings and the party with the most MPs governs. Their party leader is the Prime Minister. It is a two-tier first-past-the-post system; the first tier selects the MP based on a plurality of individual votes and the second tier selects the governing party and Prime Minister based on a plurality of MP seats. (The nature of this format causes some significant problems that I've discussed before.) This would be essentially the same as if the U.S. President was selected as the leader of the party that has the most seats in the House of Representatives, which in this case would mean John Boehner would be President because he is the leader of the party with the most seats in the House. (Yes, he is Speaker of the House but in Canada that role is different and isn't the leader.)

To get accurate predictions, you need to poll at the level of aggregation. In the U.S., that is at the state level so state polls work well, even in Maine and Nebraska due to their partisanship. In Canada, polls need to be at the riding level to provide accurate predictions. That is prohibitively expensive. Whereas the U.S. only needs 50 polling regions and has the population and money to support that, Canada would need 308 polling regions. That level of effort, precision, and cost quickly approaches that of the election itself.

A second significant problem is that we have more viable choices. We have five sitting parties in Parliament plus independents (Conservatives, NDP, Liberals, Bloc, and Green). The U.S. is essentially a two party system. Other parties can get votes but the nature of plurality voting means that third parties siphon votes from their most similar party and hence increase the chances of the polar opposite party getting elected, typically referred to as vote splitting. (Analysis shows this was a significant component of how the Conservatives won a majority in 2011 as voters moved leftward from Liberal to NDP.)

These choices in Canada mean more variables to estimate and smaller thresholds for voters to switch parties. In the U.S. the gap between Democrats and Republicans means few tend to jump across the lines. (Those that do are likely picking the lesser of two evils.) In Canada, it is unlikely to switch from NDP to Conservative or vice versa, but the policy gaps between NDP and Liberal or Liberal and Conservative are much smaller. Regional issues also play a bigger role, especially in Quebec. In 2011, we saw a significant portion of voters jump from Bloc and Liberal to NDP leading up to the election.

The added viable choices and vote splitting also create another roadblock to prediction: strategic voting. Despite naysayers, strategic voting mathematically optimizes your preferences in a plurality voting system. In simplest terms, given three or more options, you are best to vote for your top choice of the two parties most likely to win in your riding. Like in the U.S., if your favorite party is a distant third for most voters in your riding then your vote won't get them in and you will end up helping your least preferred party. It would serve you better to ensure your preference of the two leading parties in your riding beats your lesser preferred party or candidate.

The problem with strategic voting is that you need a good estimate of voter preferences in your riding. Polls measure preferences, but not at the riding level. Past riding votes measure who people voted for, but not their preferences. Perhaps many of them voted strategically. Perhaps few did. It can be very difficult to account for strategic voting or what that will mean for the next election.

A final complicating matter is modeling the principle upon which people are voting. In the U.S. Presidential Election it is clear that people are voting for the President. In Canada, people might be voting based on parties, who they want as Prime Minister, or based on the local candidate. If a lousy leader or candidate is selected, people might vote for a different party than they did last time, or they might always vote along party lines.

All is not lost, however. We can still do analytical model projections in Canada. Per riding estimates may be best based on national or provincial popular vote polls but scaled by historical support within each riding, or preferably generating trend lines over time within a riding and projecting that forward. Note, however, that this will not be even close to the accuracy achievable by recent polling within the riding as Nate can do with U.S. states.

You can also adjust for changes in candidates based on statistical modeling. For example, incumbents tend to win more often. A brand new unknown candidate even for a previously winning party might tend to do worse than expected. Conversely, a star candidate such as a child of a known politician might increase their chances. Statistical analysis of such categories can estimate their effect.

It's the Math, Stupid

The problem is, U.S. President's aren't selected by popular vote; they are selected by electoral votes. Each state gets a number of electoral votes equal to their number of Senate seats, which are two per state, and House of Representative seats, which are based on population. This means that the distribution of support for each candidate is more important than aggregate popularity, which is what the popular vote polls directly measure.

One can win a Presidental election quite handily even while losing the popular vote. To see this, imagine an election with ten electoral votes each representing ten people for a total of 100 voters. Suppose the first three electoral votes come from districts with 100% Romney support and the other seven electoral votes have 30% Romney support and 70% Obama support. Out of the hundred voters, that means 51 voted for Romney (30 from first three + 21 from the other seven), and 49 voted for Obama. Romney wins the popular vote. But Obama easily gets seven of the ten electoral votes, more than twice the electoral votes Romney gets and more than twice the support within each electoral district that he won. It isn't even close.

Now imagine the polls show this distribution, and are accurate to within 10% in each electoral district. That margin of error only represents one vote per electoral district so there is essentially no way Romney can lose the first three (at worst he'd get 9 or the 10 votes in each district), and no way Obama can lose the other seven (at worst he'd get 6 of the 10 votes in each district). This pretty much guarantees an Obama win, with high certainty, even with Obama losing the popular vote 51-49. In cases where the polls have 5-5 or perhaps 6-4 in a given electoral district, then the outcomes are less certain even if they come to the same prediction. In fact, while Obama did win the popular vote in the real 2012 election by about 2.4%, it looks like he could have done with only about 48% of the popular vote.

This is essentially all Nate Silver does. Sure, you might say, but polls don't tell you how many people will actually vote and they can be biased. Yes, and you can measure that based on historical data and methodological information. Those complicating details aside, it's just a matter of understanding how the votes add up to elect a President and measuring that as best as possible.

Animal Instincts

So why pick on him, or model-based projections? I think the answer is because political pundits tend to be hedgehogs and Silver is a fox. No, I don't mean he's attractive (at least not according to Dean Chambers, the chief poll skewer at UnSkewed Polls, whose primary argument against Silver, and can be ignored, is because he "is a man of very small stature, a thin and effeminate man with a soft-sounding voice that sounds almost exactly like the "Mr. New Castrati" voice used by Rush Limbaugh on his program.") Rather, I'm referring to the terms Philip Tetlock used in his 18+ year study of political pundit prediction accuracy. (There is a great video lecture of his study available at Fora.tv.) He, in turn, was referring to the terms used in Isiah Berlin's 1953 essay "The Hedgehog and The Fox".

A hedgehog is a person who views the world "through the lens of a single defining idea". Nowadays we tend to call them ideologues. A fox, by contrast, collects information from many sources and experiences and can't boil it down to a single idea. Nowadays we tend to call them indecisive or "wishy-washy". (I'd be a very wealthy man if I got $1 for every eye roll or drifting attention I received when starting an answer with "It depends." or ending it with caveats about uncertainty or probability.) Scientists tend to be foxes, reaching conclusions through aggregation of evidence and always careful to note the uncertainty of results and provisional nature of the conclusions. Science doesn't just work in spite of this feature, it works because of it.

Tetlock demonstrated that over the 18+ years and 28,000 predictions of his study that political pundits (aka, political "experts") were lousy at predictions. Hedgehogs are worse at it than the coin-flipping monkeys. (This refers to random guessing. No real monkeys were used in the study.) They were much worse than simple, extrapolative trending algorithms. Foxes are the only category examined that performed better than random. In fact, they performed quite well. This shouldn’t be a big surprise as hedgehogs are prone to simple, thoughtless bias essentially by definition. Foxes essentially come up with a most probabilistic case. Hedgehogs work by faith and intuition, foxes by evidence and calculation.

The fundamental problem is that the public generally prefers to listen to hedgehogs. We like confident, assured, assertive statements. Foxes are boring with their uncertainties, caveats, aggregate statistics, and probabilities. From Tetlock:

What we find, essentially, is that there is an inverse relationship between what makes people attractive as public presenters and what makes them accurate in these forecasting exercises.

This is unfortunate, but not shocking. Popularity often has little to do with utility. Take The Learning Channel for example. Over 40 years, through the wonders of privatization and free markets, this channel morphed from a publicly funded NASA science and technology channel that informed and educated, to do-it-yourself instructive home renovations (Hometime, Home Savvy), to pseudo-edutainment (Junkyard Wars, Trading Spaces), to docu-dramatic "Adrenaline Rush Hour" (Trauma, Life in the ER, Paramedics), to emotional "overcoming adversity" stories (Wedding Stories, Baby Stories, Flip that House), and (hopefully) has finally reached bottom with vapid train wrecks like the pedophile-approved Toddlers in Tiaras and everybody's favorite fartastic whoopie cushion, Here Comes Honey Boo Boo. When maximized public attention is the basis for choosing who to put on TV it is not surprising that we end up with the lowest common denominator.

Hedgehogs fit this scenario perfectly. They are assertive and confident. They titillate and shock. They disagree assertively. It's like watching an old fashioned duel, waiting to see who can win the "I can talk over you more" battle. Think about the following description:

He attains his high-ranking position through intimidation, strength, and intelligence, often exhibited in "dominance displays." Such a display often begins with spectacular charging, during which a male hurls himself along the ground, sometimes upright, slapping his hands, stamping with his feet, dragging branches, or hurling rocks.

Ask yourself if I'm referring to alpha male displays in apes or Scarborough's and Chambers' attacks on Silver. Do we watch these displays because they provide value or because as a tribe we're interested in who the dominant apes are?

Nate Silver is clearly a fox. Chambers, Scarborough, and most of the mainstream TV pundits are hedgehogs. It is their hedgehoginess that makes them interesting to watch. They're just mostly wrong.

People Talk, Data Speaks

I think their reasons for picking on Silver are then twofold. First, he doesn't view the world the way they do. They work by talking out simplistic ideas and using gut feelings on what it means, and are very sure of their gut. Conversely, Nate works by gathering cold, boring, objective, uncertain data and applying boring math to it. This is a foreign language to hedgehogs. Just listen to Joe's emphasis and confidence that he knows what's what and the data-modeliers "are a joke". He truly can't conceive that their math actually works.

Second, his success threatens their jobs. Hedgehogs don't have anything else but their (over)confidence to live off of. They are lousy at predictions. Their only skill is yammering on. If they constantly get shown up by a boring, nerdy, "thin and effeminate man with a soft-sounding voice" then who will pay them to fling their poo?

This conflict isn't new. It's the same conflict seen in the movie Moneyball (or preferably the book). It is the story of the 2002 Oakland A's who managed to make it to the American League playoffs with a tiny budget and no star players, thanks to using statistical models of players and how games are won (called sabermetrics). Hedgehog scouts, executives, and media had a tough time accepting this for the same reason as the political pundits have a hard time with Nate Silver. If a mathematical algorithm could pick the players, what value was there for their gut feelings.