Sunday, January 21, 2007

On Trust

A recent article in the New York Times Magazine called "Knowing Noreen" (7 January 2007) muses about the character and consequences of trust. Noreen Mulholland, a nanny once employed by New York writer Lisa Belkin, was subsequently convicted of assaulting and poisoning patients at an Irish hospital, provoking Belkin to reconsider the wisdom of placing her children in the hands of someone she realised she did not truly know.

Belkin raises provocative questions about the ways we represent ourselves and read others, and wonders aloud whether knowing more about a person is likely to make us more generous or merely more judgmental. Certainly these are important if perhaps unresolvable questions. But in reading the article, I was struck by what I thought of as its remarkably naive approach to trust, one leading almost inevitably to confusion because it reduces trust to a material thing.

Belkin writes, "as well as you ever know anyone, you can know only what he or she allows you to see." I disagree. In my experience, people reveal themselves quite reliably, and what we know about others is limited by what we allow ourselves to see.

I am not a person who trusts easily. This is not to say I am especially distrustful, but my relationships with people are filtered through an evidence chamber in which people's words are distilled alongside their actions, and I pay as much attention to the sedimenting lenses as I do to the particles that remain suspended. I read them like entrails. As a result, people do not often surprise me, and on the occasions when I have been betrayed, what surprised me most was my own failure to have seen what had been obvious all along.

In her own commentary, Belkin considers, long after the fact, the ways Noreen had revealed herself throughout their relationship: the attempt to dissuade Belkin from contacting previous employers, the victim narrative that held up only as long as only fragments of the history were known, the selective disclosures designed to distract attention from more troubling silences, the refusal to accept responsibility even for minor misunderstandings, the willingness to blame a five year old child for broken trust. As shocked as Belkin might have been by the things ultimately revealed during Mulholland's criminal trial, she might well have been more dismayed by her own eagerness to trust prematurely even in the face of troubling suggestions that Mulholland was not as advertised. Moreover, it seems to me that Belkin's naive willingness to trust Mulholland impaired her ability to see even when the evidence became overwhelming. In the concluding paragraphs, Belkin surmises that Mulholland might have hidden and then misrepresented her past because she was afraid of judgment, an inference that seems much less compelling than the conclusion that Mulholland doled out narrative fragments to suit each current story, revising the history whenever doing so suited her purposes.

I do not mean to suggest here that I think selective or incomplete disclosure is tantamount to untrustworthiness, nor that I believe exhibitionism makes a person easier to know or trust: it is also an error to think that people who tell the longest stories have the least to hide. But as someone who writes and interprets narratives for a living (and not inconsequentially because my father was once an intelligence agent and my mother is a retired forensic consultant) I am acutely aware of the ways evidence is constructed.

In my view, the threshold of trust is marked not by the size or shape of people's narratives but by their consistency. A more prosaic way of putting this is to say that people's words must match their actions. And this is something that can only be known over time and by reading the accretion of evidence that accumulates like sediments beneath a still lake. And if the waters remain muddy, this too is a reason to defer trust. Trust, in this sense, is not a thing or condition: it is better understood as a process, or perhaps more appropriately as a horizon.

I have written about trust previously as part of a talk prepared for a management course at the University of Toronto. At the time I spoke sharply against fuzzy and romanticised notions of trust as something beautiful in its simplicity. I argued, instead, that trust was beautiful in its complexity. I should have gone further and said that trust is not a thing at all. And it is for this reason that I object to 'trust-building' exercises such as the one where a person falls backward into the (presumably) waiting arms of colleagues, accompanied as this act is by the delusion that trust is a thing you can give and receive through such a limited demonstration, like exchanging a bouquet or a business card.

As someone trained extensively in negotiation and who used to bargain collective agreements union-side, I think trust is overrated. In my experience a negotiation is much more likely to succeed when the parties admit they don't trust one another. Reducing this kind of delusion also reduces opportunities for deception.

Unless there is a good reason to trust you, I will not do so. This is not to say I will avoid you or hide behind a veil of paranoia and suspicion, or that I shrink more generally from worldly engagements. But my dealings with you and with the world as a whole will be informed by a risk analysis, and when I expose the more vulnerable elements of my person or affairs, it is not because I trust you but because I am prepared to risk certain kinds of damage to them. And in this sense, Lisa Belkin can hardly be criticised for making a very reasonable risk/benefit analysis in seeking child care (one which, it must be noted, appears not to have led to any harm to her children and which almost certainly benefited both the Belkin family and Mulholland herself in the short term). But it seems to me it was a significant error for Belkin to mistake that calculation for trust.

Afterthought: It is possible to identify a number of kinds of trust errors (which might be used, ultimately, to develop a more coherent model of trust than exists currently). Taking Peter Fruchter's modalities (also rendered here) as a conceptual framework, we can observe at least three kinds:
  1. Being overly definitive about what trust entails (and what entails trust): assuming that trust is a necessary precondition for (or consequence of) a relationship (i.e., instead of treating a relationship as an expression of acceptable risk);
  2. Being overly descriptive about the character of trust: mistaking trust for (only) a thing; and
  3. Relying on imperative or authority claims: "Trust me ..."
Second afterthought: Until we develop a more coherent and useful model of trust (the philosophical literature on trust, including Annette Baier and Cheshire Calhoun's otherwise exemplary contributions to moral philosophy, strikes me as manifestly confused), it seems fully reasonable to rely on the more coherent concept of risk (which is easier to apply and test in everyday settings such as the one Belkin confronted, and furthermore is a subject whose theorists are at least honest about [or aware of] their materialism).

["Trust" image by purplejavatroll and used here under the aegis of a Creative Commons license.]

No comments: