Trusting Strangers
I am currently working on a paper about deep fakes in which I argue that the concept of vouching for someone or for a source enables us to construct a horizontal form of trust, i.e. trust without authority relations. This is an anathema to many ways of thinking I encounter, from the basic structures built into our IT systems, which are of their nature authoritarian, to the model of liberal democracy as requiring authoritative institutions.
Working on this has me reading various things about trust and trusting and I'm struck how it seems obvious to most philosophers and probably others working in the area that trusting strangers is a problem. Here is one example:
Strangers, by contrast, are comparatively difficult to trust, even if we are able to rely on them in certain ways. We have no prior experiences with them, and there is no particular reason to assume they will do especially well by us. (Paterson, 2021, 524)
Paterson here offers what appears to be an explanation for why it is harder to trust strangers than acquaintances. What she says is true, if you are unacquainted with someone, you have no information which distinguishes them from any other stranger, then of course there can be no particular reason to assume anything about them. However, for this to be relevant to the question of whether to trust them or not, there needs to be a certain presupposition. If one needs particular information about a person in order to assume they will do well by one, then:
Another person doing well by one is the exception rather than the rule.
If the general rule was that people do well by each other when they can, then one would need particular information to distrust someone.
Before discussing that issue, I should deal with a likely response from Paterson, namely, that I have overlooked the importance of that word 'especially'. If 'especially' here means to treat one better than they would treat a stranger, then the point is trivial. There is of course no reason to assume that a stranger will treat one better than they would treat anyone else who is a stranger to them.
There are many cases of trust when the trustor is expecting special treatment from the trustee. Nearly always this is because of a particular relationship between them, such as parent, child, friend, or lover. However, most trusting is not about asking for special treatment. Paterson gives the example of trusting a taxi driver to take her to her destination and treat her with respect. She does not want special treatment, she does not want to be treated differently from any other passenger. It seems likely that the overwhelming majority of situations where we trust another person are like this.1
We need to take care here to distinguish specialness from directedness. Even when one is not seeking special treatment from the trustee, one is seeking directed treatment. One is seeking out a certain behaviour towards oneself in particular. Paterson trusts the taxi driver to take her to her destination, and in that sense she is placing trust in the driver performing an action directed at her. But it is directed at her in virtue of the relational fact that she is currently the passenger, and not in virtue of any intrinsic fact about her which distinguishes her from any other potential passengers.2
Rules, Generalizations and Principles
So let's return to the assumption Paterson seems to be making in the passage quoted.
Another person doing well by one is the exception rather than the rule.
Suppose we ask "Is this true?" If we think it is an empirical claim, then with enough patience, ingenuity and, let's be honest, funding, we could come up with an answer about the proportion of strangers, who would "do well by" a stranger. However, treating it as an empirical claim seems to be misguided. Knowing the percentage of trustworthy strangers out there would only make any difference if there was some percentage such that if it was greater than that, then it would be rational to trust this particular stranger. That way of thinking about the individual interaction shifts us from trusting to relying upon the stranger. Suppose the traffic lights have turned red and you decide to cross the road even though there is a car approaching. If your reason for doing that is your knowledge that 99.99% of drivers stop at a red light, then you are relying on that driver. If your reason is, in contrast, that the driver is a human being who would not want to injure, or even merely seriously frighten, another human, then you are trusting them, and the percentage of drivers who stop at a red light is irrelevant.
Perhaps then, it is an a priori truth. The philosophical thought experiment of how people would behave in a "state of nature" might be understood as an attempt to give an a priori argument for this assumption. But this is inconclusive because, for each Thomas Hobbes who thinks a state of nature would be a state of war of all against all, there is James C. Scott who thinks a state of nature would be a life of freedom, dignity and mutual respect. Hume thinks the difference turns on whether we imagine this as a state of abundance of natural resources or one of scarcity. In a state of abundance, he allows that:
Avarice, ambition, cruelty, selfishness were never heard of: Cordial affection, compassion, sympathy, were the only movements with which the mind was yet acquainted. (Enquiry Concerning the Principles of Morals, III.i)
Rather than interpreting this debate as a matter of a priori argument for the truth or falsity of the assumption, it would be better to understand it as the acceptance or rejection of a moral principle. Someone who accepts the assumption starts their moral reasoning from a position of distrust in strangers; whereas someone who rejects it starts their moral reasoning from a position of trust in strangers. Instead of asking which is correct, as if there were some truth of the matter, let us ask instead which is better: which way of doing your moral philosophy is itself morally preferable?
Like any other moral principle, we each have to make an individual decision to adopt it or not.
It is worth being clear that these are genuine cases of trust rather than mere reliance. When I took a robotaxi in Shanghai earlier this year, I relied upon the autonomous driving system to get me safely to my destination.I did not trust it because it was not able to treat me as a person. From its point of view, the passenger could have been a sack of potatoes. Whereas when I trust a human taxi driver, I am expecting them to recognize my humanity and respond appropriately to that humanity in any given situation. Trusting a human taxi driver means trusting that they will notice if their passenger is upset or frightened and respond in an appropriate way.↩
The reader might note that being one's child is equally a relational fact, but it might merit special treatment. But we need to distinguish the directed duty to, for example, feed one's own child, which one does not hold to someone else's child, and a duty, if there is one, to treat one's child in better ways than one would treat any other child, should one be in a position so to do. The parent handing out cake to a group of children, who gives their own child the larger slice, is behaving in the latter way.↩