Verified by 
The best way to find out if you can trust somebody is to trust them.
~ Hemingway
Trust and its dark mirror image of betrayal are like life and death, one constitutes the other. Every life is lived under the shadow of death. Death, like taxes, is certain. What is uncertain is the timing. Every life also requires trust. A principled mistrustful person would hardly be able to interact with anyone. Hermits have never looked like heroes to me. Instead, they look like individuals (in the true sense of the word) that have forsaken sociality and thus many of the most profound joys life has to offer.
With trust, betrayal looms as a possibility; not just a remote possibility but a very real possibility that packs a punch when realized. Betrayal on the battlefield can kill; in ordinary social (or financial) life it just hurts like hell. Currently, social exclusion and ostracism are de rigueur as research topics, and trust also gets a fair amount of attention. But betrayal, oh boy, that is just too painful to look squarely in the eye.
If people must trust lest they become hermits withering on the social vine, their question is "Whom?" [or "Who?" in contemporary agrammatical American parlance]. If the right choice is not not to trust anyone ever, then it is the wise allocation of trust that is the riddle to be solved.
Quick check with our overpaid colleagues in the economics department: Econs say that the game of trust has an equilibrium of distrust (and hence not even the opportunity of reciprocity). Why? Once trusted, the trustee has no "real" [i.e., monetary] incentive to give anything back; hence she won't and hence the trustor, who can see it coming [or rather not coming], will not trust; hence nothing happens. Crafty as they are, economists do not concede that they "recommend" distrust or betrayal; oh no, that would make them look churlish. Instead, they assert that they just prove theorems. The fact that many people trust and that many honor the trust placed in them, is rather inconvenient for economic theory; but hey, why should human behavior get in the way of a beautiful theorem? Besides, many econs do trust others when they are not busy proving theorems and many also reward trust placed in them. This incoherence humanizes them. I like it.
Knowing that they must trust some people some of the time, trustors could opt for a "mixed strategy." They might decide, for example, that trusting 50% of the time is the right rate, and to achieve it, flip the proverbial mental coin [or, awkwardly, a real coin].
Alas, deploying a mixed strategy would be nerdy and it might suggest that you have given up. You are supposed to have some social intelligence and know whom [!] to trust and whom to avoid. You are supposed to have some Menschenkenntnis [knowledge of man and woman].
Imagine for a moment the trust problem were solved. We would live in a world comprising neatly recognizable trustworthy and untrustworthy people. Each would wear a big sign on her head saying honestly "trust me" or "do not trust me." After a generation or two, the untrustworthy folks would be gone, leaving a population of trustors and reciprocators. Until, that is, a freak mutant appears who exploits the trust placed in her. Evolution's next move would be to produce individuals willing and able to feign trustworthiness.
So there you have it: The logic of evolution dictates that [a] trustworthiness will be signaled more often than it is delivered [there are no strong incentives to feign untrustworthiness] and that [b] the signals of trustworthiness will have little validity. Trust must remain a dilemma.
Now, there are a few soft signals of trustworthiness. For what it's worth, here are the beginnings of a crude list. Bearing these cues in mind does not guarantee a safe journey, but it allows you to beat blind chance.
Trust people who
[1] are likeable or have expertise in the matter at hand
[2] have reciprocated trust in the past
[3] stand relatively little to gain from betrayal
[4] are interdependent with you (with the shoe getting on the other foot)
[5] have made promises [but beware; if betrayal comes, it will hurt even more]
[6] have not enrolled in economics courses [see above]
[7] belong to certain ethnic (Norwegian) or religious (Quakers) groups
[8] are similar to you [too bad if you are not Norwegian].
[9] have certain facial features studied by Alex Todorov and other lovers of FaceGen. [N.b., we don't know yet if these facial cues are valid.]
[10] themselves trust. Lao-Tzu says: "He [or she] who does not trust enough, will not be trusted." [See also the Dao of Hemingway as expressed in the epigraph]
In general, and as illustrated by the promise item, the sub-dilemma built into the basic trust dilemma is that if you have some semi-valid cue suggesting trustworthiness, the betrayal, when it comes, will sting more.
That leaves trust in god. Seriously? I can only hope that the religious among you who use this idea realize that it is a misnomer (they mean to say "faith"). If they truly mean "trust," they must reckon with the possibility of betrayal. Among humans, the betrayal of trust is universally condemned (Machiavelli dissenting, forse). If betrayal is not condemned in a deity, then there is no attitude of trust, but an attitude of "Whatever!" [Job had trust, not faith, in his god. He condemned his betrayal.]
Down here, we must live with the dilemma.
Note: The photo above shows Niklas Luhmann, eminent sociologist and author of many incomprehensible works, who wrote lucidly however about trust and who rejected the economists' pessimism.
