Tuesday, July 15, 2014

Trust and the prisoner's dilemma

This is pretty obvious, but I never quite thought of it in those terms: The prisoners' dilemma shows the need for the virtue of trust (or faith, in a non-theological sense). In the absence of contrary evidence, we should assume others to act well, to cooperate.

This assumption perhaps cannot be justified epistemically non-circularly, at least not without adverting to theism, since too much of our knowledge rests on the testimony of others, and hence is justified by trust. Our own observations simply are not sufficient to tell us that others are trustworthy. There is too much of a chance that people are betraying us behind our backs, and it is only by relying on theism, the testimony of others, or directly on trust, that we can conclude that this is not so.

It seems to me that the only way out of the circle of trust would be an argument for the existence of a perfect being (or for some similar thesis, like axiarchism) that does not depend on trust, so that I can then conclude that people created by a perfect being are likely to be trustworthy. But perhaps every argument rests on trust, if only a trust in our own faculties?

3 comments:

carlos said...

Assuming every argument depends on trust and we neglect strongly skeptical possibilities: may the non-theist point to the epistemic possibility that pro-social behavior like trust increases fitness, and thus is likely to be (group-) selected for?
Such epistemic possibility could then be established by an appeal to current evidence from Biology.

Alexander R Pruss said...

Sure, but I think you're not going to be able to argue for the fitness of pro-social behavior without making use of massive amounts of empirical evidence which depends on the testimony of others, and hence rests on trust.

astridnova said...

Dawkins pointed out (in The Selfish Gene) that tit-for-tat only comes out positive in viscous populations, noting that viscous populations (in humans) used to be the rule, but are no longer. The 'trust' element relies on whether you are likely to run into the people involved in tit-for-tat soon. Obviously if you live in a small community, you won't do very well if everyone knows that you cannot be trusted. However, if you live in a modern, atomised population, like a big city with many people going in and out of it, then you are unlikely to be held to account. I'm writing about this as Sheila Newman in the third in series of Demography territory law, to be published by countershock press.