Worse Is Better

Reinhard P. Gabrieli

Lucid Ethics Working Paper, 1989

Reprinted in The Journal of Applied Moral Philosophy, Vol. 3, No. 1, Spring 1991

I am going to describe two approaches to moral conduct that lead to fundamentally different kinds of person, and I am going to argue, with some reluctance, that the worse approach produces the better result.

The Problem of the Wallet

The argument began, as so many do, over lunch. I was eating with a colleague from MIT — a moral philosopher of considerable reputation, whose ethical framework I admired without wishing to adopt — when the waiter brought us a wallet that had been found under our table. It contained two hundred dollars in cash, a driver’s licence, a library card, and a photograph that was, by any conventional standard, embarrassing. Not illegal. Not harmful. But the kind of photograph that a person would not wish to have returned to them by a stranger, because the return would imply that the stranger had seen it.

The MIT philosopher — I will call him the Right Thing, because that is what his approach demands — proposed the following: we should return the wallet to its owner in person, complete with all its contents, and we should acknowledge having seen the photograph, because to pretend otherwise would be dishonest. Honesty, he argued, is not optional. The moral agent does not get to select which truths to share based on social convenience. The wallet must be returned whole, the truth must be told whole, and if the owner is embarrassed, that is the cost of living in a world where people sometimes lose their wallets. The moral agent is not responsible for other people’s comfort. He is responsible for his own integrity.

I proposed an alternative. We should return the wallet by mail, with a brief unsigned note: “Found under a table at Rosario’s. All contents intact.” The owner would recover his property. He would never know who had returned it, and therefore never need to wonder what we had seen. The truth, in this formulation, was not suppressed — all contents were intact — but it was managed. The interface between us and the wallet’s owner was kept clean.

“That is a lie by omission,” said the MIT philosopher.

“That is a kindness by design,” I said.

He was right, of course, in the narrow sense that I had not told the whole truth. But I was right in the broader sense that no human being in the history of social interaction has ever told the whole truth in every situation, and the ones who have tried are remembered not as moral exemplars but as clinical cases. The MIT approach to the wallet produced a morally complete solution and a deeply uncomfortable lunch. My approach produced a morally adequate solution and a wallet returned without incident. The question is which outcome is, on the whole, preferable.

Two Schools

I call these the MIT approach and the New Jersey approach to moral conduct, partly because I encountered them in those places and partly because the geography is apt. MIT is a place where people build systems of extraordinary formal elegance in buildings of extraordinary aesthetic ugliness. New Jersey is a place where people get things done.

The MIT approach (The Right Thing). Every moral situation must be resolved completely and correctly. The agent’s conduct must be consistent across all cases. No edge case may be left unhandled. If a principle applies, it applies universally, regardless of social cost. The resulting moral framework is beautiful, rigorous, and internally consistent. It is also, in practice, unusable. The agent who applies it discovers that most social situations contain edge cases that the framework handles correctly and catastrophically — correctly in the sense that the logic is valid, catastrophically in the sense that the friendship is over.

Kant is the patron saint of this school, and the fact that Kant is reported to have had almost no close personal relationships is treated by his admirers as irrelevant and by his critics as evidence.

The New Jersey approach (Worse Is Better). The moral agent should be decent to the people in front of him. Local consistency matters more than global completeness. If a principle produces a result that is technically correct but socially ruinous, the principle should yield to the situation, not the situation to the principle. The resulting moral framework is imprecise, inconsistent at the edges, and uncomfortable to examine too closely. It is also livable. The agent who applies it has friends, is invited to dinner parties, and transmits his ethical sensibility not through argument but through example — which, as it turns out, is how most moral education actually works.

Hume is the ancestor here, though Hume would have been too polite to claim the role. Aristotle’s phronesis — practical wisdom, the ability to judge what a situation requires rather than what a principle demands — is the closest classical analogue. The New Jersey approach does not deny the existence of moral principles. It denies their sovereignty.

The Problem of the Interface

The deepest difference between the two schools concerns what we might call the moral interface — the boundary between the agent’s inner ethical life and the social world in which that ethics must operate.

The MIT approach insists that the interface must be transparent. The agent’s internal moral state — his reasoning, his principles, his conclusions — must be fully visible to the other party. To conceal any part of the reasoning is to compromise the integrity of the act. If you return the wallet, you must say why. If you saw the photograph, you must acknowledge it. The interface between you and the world must faithfully transmit everything that is true about your moral position, regardless of what the world wishes to receive.

The New Jersey approach holds that the interface is the ethics. The relationship between you and the wallet’s owner is not a channel through which moral truth is transmitted; it is the moral reality itself. To manage the interface — to choose what to transmit and what to withhold — is not a compromise of the moral act but a constitutive part of it. Tact is not a dilution of honesty. Tact is a recognition that honesty operates between persons, not in a vacuum, and that a truth delivered without regard for its reception is not more honest but merely more destructive.

This is the crux. The MIT philosopher believes the interface should be a perfect conductor: zero resistance, total fidelity, complete transparency. If the signal is unpleasant, that is the signal’s nature, not the conductor’s fault. The New Jersey pragmatist believes the interface has impedance, and that this impedance is a feature. Some signals should be attenuated. Some should be filtered. The resulting transmission is less faithful to the source and more faithful to the relationship, and the pragmatist believes the relationship is what matters.

The Virus

The MIT approach has a problem: it does not spread.

A moral framework that demands complete consistency, total honesty, and the resolution of every edge case is a framework that requires extraordinary personal resources to maintain. It is exhausting. It is socially costly. It is, in a word, difficult, and difficulty is the enemy of adoption. The MIT philosopher converts his students through argument, and argument converts slowly, and the converted tend to drift back to their prior habits once the semester ends, because the prior habits are easier and nobody likes the person who insists on resolving every dinner-party disagreement to its logical conclusion.

The New Jersey approach spreads like a virus. Not because it is argued but because it is demonstrated. The person who returns a wallet with a kind note, who tells a useful truth and omits a useless one, who handles a moral situation with grace rather than rigour — this person is imitated. Not because anyone decides to imitate him, but because his approach works, visually and socially, and humans are imitative creatures who adopt what works. The virus is not transmitted through the intellect but through the nervous system. You see someone handle a difficult situation well — well meaning adequately, not perfectly — and something in you adjusts.

This is how Worse Is Better conquers the world. Not by being right — the MIT approach is, by any formal measure, more right — but by being viable. The adequate moral framework that a million people actually use is more effective, in aggregate, than the perfect moral framework that twelve people admire and one person practices. The virus wins not through virulence but through ease of transmission.

I find this troubling. I was trained in the MIT tradition. I believe, in my bones, that the Right Thing is the Right Thing, that principles should be applied consistently, that edge cases matter. But I have watched, over the course of a career, as the New Jersey approach has colonised the moral landscape of every institution I have worked in, not through conquest but through convenience, and I have begun to suspect that the convenience is the argument.

Conclusion

Aristotle, who was from neither MIT nor New Jersey, would have recognised this problem. His phronesis is not a compromise between rigour and laxity. It is the claim that moral knowledge is practical knowledge — that it inheres in the act, not in the principle, and that the wise person is not the one who knows the most rules but the one who judges best in the particular case. This is, I believe, a more sophisticated version of Worse Is Better, dressed in a toga and therefore more respectable.

I remain ambivalent. The MIT approach is noble and impossible. The New Jersey approach is adequate and everywhere. A philosopher is supposed to prefer the noble. A pragmatist is supposed to prefer the everywhere. I am both, and neither, and I suspect most people are, and that this ambivalence is itself the New Jersey approach operating at the meta-level — the refusal to resolve a question that does not need to be resolved in order to get on with lunch.

The wallet was returned by mail. The owner never called.

Reinhard P. Gabrieli is the founder of Lucid Ethics, a moral philosophy consultancy based in Palo Alto, California, and the author of Patterns of Virtue: A Case-Based Approach to Moral Reasoning (Oxford University Press, 1996). He teaches occasional seminars at Stanford on the ethics of compromise and the compromise of ethics.