"We were talking about Kant's categorical imperative. And that's basically the Golden Rule, right?"

That's how my philosophy professor began class one morning.

"No," responded a student. (OK, it was me.)

"Good, you didn't fall into my trap."

Unfortunately, Errol Morris, the acclaimed documentarian, falls into the trap in his piece for the New York Times about lying — "Seven Lies About Lying."

Morris's lie-about-lying #4 is, "Lying can never be justified" — "one should always tell the truth." He correctly attributes this view to Kant. Unfortunately, he adds:

It was linked to his "categorical imperative," Kant's version of the Golden Rule. Would you like others to lie to you? Then don’t lie to others.

That part in bold is the classic mistake about Kant's ethics. Morris tries to support it with a footnote quoting Kant's Critique of Pure Reason:

"I cannot wish for a general law to establish lying be-cause no one would any longer believe me, or I should be paid in the same coin."

The key word here that refutes Morris's interpretation is "cannot." This should be taken literally: it's about whether it's possible for you to want everyone to follow this general rule, not about whether you would actually like for everyone to lie. Kant thought it's impossible for everyone to follow a rule of lying for personal gain. After all, if everyone followed that rule, no one would be able to trust anyone's statements. Thus, lies would become ineffective, since lies only work if people generally trust other people's statements. The idea of a world in which everyone follows a rule of lying for personal gain isn't merely unsavory; it's self-contradictory. Since you can't conceive of something self-contradictory, you cannot wish for a world where everyone followed the rule. Consequently, you shouldn't follow this rule; in other words, you shouldn't lie.

(That's my off-the-cuff rendition of Kant. I haven't recently read the primary sources, so it might not be perfect. If you'd like to read a more rigorous explanation — using the more traditional Kantian terminology of "universal maxims" and so on — you could try this blog post.)

In fact, the whole foundation of Kant's theory was that people should be guided by reason, not by their personal preferences. The Golden Rule — "Do to others as you would like them to do to you" — directly refers to your personal preferences. It's not surprising, then, that Kant actually criticized the Golden Rule.

The Golden Rule is more self-centered than Kantian ethics. In the standard formulation, the Rule refers to "you" twice in one short sentence: it's about what you would like to have done to yourself. Since different people have different desires about how they'd like to be treated, this implies a relativistic moral code. Taken literally, the Rule may provide wildly different advice to different people based on their idiosyncratic traits.

But I also have a deeper problem with the Golden Rule's invocation of what-you'd-like-done-to-yourself. Even if we put aside concerns about whether it's too relativistic or unstable, there's still the unanswered question of where these desires come from. Why do you want anyone — even yourself — to be treated a certain way? The Golden Rule seems to take this as a given, but the question of what people want — or should want — is hardly simple to answer. It would seem that an explanation would need to come from something beyond the Golden Rule itself. And perhaps that something is actually more fundamental to ethics.

There's another problem with equating the Golden Rule with the categorical imperative: the Golden Rule is, at least on its face, just about how to treat others. Kant saw ethics as including how you should treat yourself. (For instance, one of his most famous examples of the categorical imperative is his argument that suicide is morally impermissible. While suicide does hurt others, Kant was more concerned with the wronging of oneself.) I don't subscribe to Kant's ethical theory, but I at least give him credit for trying to address profound moral questions that the Golden Rule doesn't even touch.