I would like to describe some thinking that I have been doing that is based on some rather remotely connected data points. I will briefly mention the data points and then on to the conclusion.

a) Many years ago I read something by Kant named “A metaphysics of morality” or “Toward a metaphysics of morality” or something like that. In it his basic thesis went something like this: We define moral behavior as being that behavior which is appropriate for everyone. The example that he kicks off with is the difference between telling the truth and lying. He points out that lying is only of benefit to the liar and more importantly, only works for the liar if everyone else is telling the truth and assuming that everyone is telling the truth. If lying was really morally correct behavior, then by his definition, everyone should do it and if everyone does it, then it ceases to work, therefore according to Kant lying is not moral behavior. That is a very simple overview of his thesis but I think you can get the basic idea from that.

b) Also many years ago in Scientific American I read an article by Douglas Hoffsteader (sp?) that involved some of the notions of game theory. Again without going into great detail, he mentions that for a certain class of games, where you can talk about the value of playing the game and where the game has certain symmetry properties you can conclude that there is one proper strategy for playing the game. He shows that for those games if everyone is smart and analyzes the game they will all play exactly the same way because if you deviate from the proper strategy you just lower your expected gain. Then he gave a puzzle, which involved a single decision, do you send a letter to Scientific American or do you not send a letter to Scientific American and he constructed his puzzle to be exactly of the sort that you could analyze with all the tools that he had just outlined. I think I actually remember the puzzle. He said that he would put up 1 million dollars in cash as a prize to be divided up among the people that sent in letters and asked for it. However, and This is the big catch, he would first divide the prize by the number of people that asked for it and then divide the resulting value among the folks that asked for it. Thus if only one person asks, he gets the whole million dollars, if 10 people ask for it, the prize is reduced to $100,000 which is then divided into 10 ten thousand dollar prizes. Furthermore, you could ask for the prize multiple times in a single letter. Just tell him how many shares you are in for. Clearly the more people that ask for it, the less there is to spread around, but if no one asks for it, then no one gets anything. I send in my analysis – using his everybody must do the same thing rule – if everybody sends in 0 letters they get nothing but if they send in 1 letter they get something, even though with the millions of readers of SA what they get will be reduced to virtually nothing. Imagine my surprise to find that my solution was wrong, as were the thousands of other submissions. The proper solution, according to Douglas, was to flip a biased coin based on the number of people that read SA, about 1 million and if you are the lucky one in a million you send in a letter, otherwise you don’t. This way, you reduce the number of submissions down to approximately one single submission, which minimizes the dilution on the prize and you have an equal chance with everyone else of winning the prize. My mistake is a common one in game theory. I solved the problem in the integers i.e. I reasoned that you must either send in zero requests or one request. I did not think about how to send in a fractional number of requests. So along with all the other folks that could not analyze the game properly I spoiled it for all the folks that did the right thing. I should point out, that from other game theory that I have read, it is not at all uncommon in a game where you must either do X or Y the optimal solution for the game is to choose X with a certain fixed probability p and otherwise do Y. The point being that just because the game requires a discrete integer response, X or Y, zero or one, the strategy can and often does involve a real number, namely p, the probability of choosing one of those integer values.

c) The third data point is nowhere near as specific. It is just the broad range of literature that I have read, mostly by Richard Dawkins, about Darwinian Evolution. I remember one book in particular, the Cartoon Guide to Genetics, that in describing the nature of a virus left me with the feeling that much as we may not want viruses with us, they will always be here, because they are so simple. It is much easier to create a simple organism that lives by robbing from a more complex organism than it is to build the complex organism that produces everything from scratch. The complex organisms only hope is to produce more complexity that is designed to lock out the virus but even that mechanism will have flaws that can be exploited. Viruses exist because they can. Any system that is able to produce a complex self-replicating object is capable of producing the much simpler virus that just borrows from the self-replicating object.

And now given those data points I haves started to wonder if Kant didn’t make the same mistake in reasoning about morality that I did in solving the game theory math puzzle. He is trying to solve in the integers rather than in the real numbers. He assumes that you must either lie or tell the truth all the time and argues that lying doesn’t work if everybody lies all the time thus the solution must be the other endpoint, to tell the truth

all the time. But what if the optimal strategy from a game theory standpoint is that you should tell the truth, say 95% of the time and 5% of the time you flat out lie. What does that system look like? You get much of the efficiency benefit that you would get from telling the truth all the time, you would occasionally slip a lie past (as would everybody) and get away with it accruing presumably some personal benefit and you would be perpetually a little wary of anything that anybody claims is true because you know that they may, with a certain probability be lying.

This has caused me to wonder if perhaps the game theory works as advertised, i.e. everyone is driven to the same optimal behavior and pay the consequences if they do not follow the optimal behavior. If they lie too much they are caught out and discredited to their disadvantage and if they tell the truth too often, they miss out on those few golden opportunities when they could better their payoff. The system is of course just a mixture of people playing the game, making their own independent and possibly flawed analysis of the game but that doesn’t change the game and it doesn’t change the optimal point.

Have we by means of evolution simply converged for the most part on that optimal point? Does evolution, which must deal in a world where viruses exist, where they must exist, in fact where evolution itself produces the viruses just as cheerfully as it produces the complex organisms that attempt to fight off viruses, simply find that optimal point where we all mostly tell the truth, mostly lean toward telling the truth, mostly encourage others to tell the truth and yet occasionally lie, and are always a little suspicious that we are being lied to. The suspicion may just be part of the immune system that helps us deal with a real world where liars can and do exist. Perhaps the heaven, where everyone is truthful all the time, is just like the fictitious world where there are no viruses. Perhaps the theological thinking that says we should always tell the truth is just a flawed argument attempting to solve the problem in the integers rather than in the richer set of the real numbers. Perhaps the religions that tell us to always be truthful are just part of the structure that drives us toward that higher probability of telling the truth. Perhaps that encouragement to always tell the truth is their one lie that allows them to take advantage of us. Perhaps we are already living in the kingdom of Heaven and are all playing the game in the real optimal sense, occasionally lying, and occasionally being caught out in our lies. Then again perhaps, I simply did not learn my religious training very well, and am merely justifying my own corrupt behavior by encouraging others to be equally corrupt. Perhaps I am the devil and THIS is the big lie.

You decide!