An Ethics of Trust

An Ethics of Trust

Deontological ethics suffers from a lack of unity: we have a list of duties, expressed as moral imperatives, but no unifying principle behind them. Thus: don’t steal, don’t lie, don’t break promises, don’t commit adultery, don’t be late, don’t betray friends, etc. Is there anything in common to these prescriptions? Is there one that is more basic than the others and underlies them? One might think that promises provide some sort of more general analysis of moral duty: isn’t punctuality required because lateness is a form of promise-breaking, and similarly for adultery, lying, and betrayal? Aren’t these all types of implicit promise—leading other people to expect a certain type of behavior? But this idea doesn’t work well with stealing, and it stretches the meaning of “promise”. Also, it doesn’t ground promise-keeping on anything else but leaves it as morally primitive. But we can express the underlying thought in a different way—by bringing in the concept of trust. Suppose we say that the basic imperative is “Be trustworthy!”.  The OED defines “trustworthy” as “able to be relied on as honest, truthful, and reliable”. Now we start to see a more general imperative underlying the specific ones cited: all the prescriptions and prohibitions listed are examples of the virtue of trustworthiness. Stealing may seem the odd one out, but it is quite a complex case, because of what counts as stealing and whether it is always wrong. A paradigm example would be stealing from one’s friends and family (not an uncommon occurrence): here it is obvious that there is a violation of trust. Your friends admit you to their house and you abuse their friendship by stealing from them—wrong! Shop-lifting is an intermediate case in that you are being trusted not to steal while you pretend to shop, though security guards lessen the dependence on trustworthiness. Stealing from an orchard while starving because of unfair social conditions edges out of the zone of culpability. But stealing from people with whom one has a relation of trust is clearly immoral: it is a direct abnegation of trustworthiness. Trust, of course, is what makes social life—and hence human life—possible, so being trustworthy is vital to human well-being. Hence, we deplore violations of trust. The prime directive of deontological ethics is plausibly “Be trustworthy!” Once this prescription is absorbed the rest automatically follow, because they all involve questions of trust. The central concept of deontological ethics is therefore trust: be such as to be honest, truthful, and reliable—not dishonest, deceptive, and unreliable. Then you won’t be a liar, a cheater, a thief, an adulterer, a traitor, a promise-breaker, and habitually late for your appointments. But there is one type of misdeed that doesn’t fit easily into a trust-based ethics—causing harm. I think this case is special and shouldn’t be expected to belong with the other misdeeds mentioned. First, causing harm (including death) is not in itself immoral: doctors, dentists, boxers, wrestlers, football players, soldiers, chess players, tattoo artists—all cause harm to others as part of their calling. But this is not unethical because there are justifying reasons for the acts in question—the greater good, self-defense, voluntary competition, consent. So, causing harm is not as universally wrong as other wrongful acts. Second, in many cases wrongful harm is accompanied by violations of trust: wife-beating, child-flogging, murdering one’s spouse for money. In such cases we might try to put the wrongness down to failures of trustworthiness (not focusing too much on the actual harm caused). The really hard case is that of violence inflicted on people and animals where no issue of trust is involved—such as attacking strangers you have not encouraged to trust you. You are not being untrustworthy when you hit a stranger over the head with a brick from behind, though you are certainly being wicked. One might try to suggest that there is a violation of some general type of trust directed to mankind as a whole in such cases—you trust other people not to attack you in this way. But you may in fact distrust the person who attacks you and he has given you no reason to trust him, and the act is still wrong. Not all immoral actions involve failures of trustworthiness (a person can be reliably and honestly vicious). Trustworthiness will only get you so far in matters of morality (pretty far but not all the way). What this tells us is that morality consists of two components—two distinct and irreducible values. It revolves around a pair of basic concepts: trust and harm. Deontology concentrates on trust-involving action; consequentialism concentrates on harm-involving action. The two overlap each other, but they are not the same. We should not even try to include the wrongness of causing harm on the usual list of moral duties, because it doesn’t belong in that category. Roughly speaking, deontology deals with right and wrong acts considered in themselves, while consequentialism (as the name implies) deals with the effects of actions, often extending far into the future. We could even say that the former deals with act-ethics and the latter with result-ethics. These are different domains that should not be conflated or confused: ethics therefore cannot be monolithic, unified, and homogeneous. Ideally, they would have different names—say, “dutics” (duty ethics) and “effectics” (consequentialist ethics); but these are ugly and unnatural inventions, so will not be adopted. What has to be recognized is that harm-related ethical concerns belong in a class of their own; they are not a type of deontological ethical imperative (“Do no harm!”). Such an imperative has far too many exceptions that are difficult even to codify, and harm can extend indefinitely into the future and not even be predictable by the agent. It is better to accept that our moral outlook includes disparate components; certainly, we should not regard our theory of duties as inadequate because it can’t assimilate the wrongness of harming. We have a nice unifying account of the various moral duties and we shouldn’t reject it because it can’t include the wrongness of causing harm. Being untrustworthy is clearly a bad way to be, and causing harm is often (not always!) a bad thing to do: but these are separate spheres of wrongness. They are wrong for different reasons (roughly, social breakdown and the intrinsic badness of suffering), and they are conceptually quite distinct. Trust is about human relationships, a sense of personal security, while harm is about pain and suffering within the individual. Trust cannot exist in conditions of solipsism, but suffering can. We should be satisfied if we can devise a bipartite theory and not hanker after complete theoretical unity. It’s like sense and reference: we need both. Or belief and desire: complementary but inherently distinct. If we ask which of these two values is primary, the answer is neither: they are equally important and equally basic. But if I can put in a special plea for trust: being trustworthy is vitally important in human life, not some optional add-on, and violations of trust the most corrosive of failings. Oddly enough, it doesn’t seem to feature much in world religions—though the story of Judas and Jesus is etched into the hearts of all who know it. It really should be emphasized more.[1]

[1] There is far too much emphasis on loving one’s neighbor, on sexual morality, on humility, on purity, on self-denial, on detachment, on beneficence—and not enough on being a reliable and truthful individual. Punctuality is the paradigm: get this right and the rest will follow. More deeply, be the kind of person that others feel they can depend on, who won’t let them down, who won’t betray them for the proverbial forty pieces of silver. And beware people who look into your eyes and intone “Trust me”: they are often the ones who will let you down first. The con artist is among the worst of men (and women, of course). How can your neighbor love you if she can’t trust you an inch? How is respect consistent with distrust?