I think the normal way we think about lying is all wrong.
Here’s a thought experiment (paraphrased from memory) that I’ve heard Penn Jillette describe: A man’s wife comes to him and says: “Look, you’ve been really distant lately, and it feels like there’s something going on. You say you’ve been staying late at work every other night, but I’ve called your office and they said you weren’t there. You’ve been cagey about letting me look at your phone, like you’re worried about messages I might see, and also other examples of suspicious behaviour. I don’t know what’s going on, but I want you to look me in the eye and tell me: Are you seeing another woman?”
The man answers no.
The truth is that he’s cheating on his wife with another man.
Did he lie to her?
By a common interpretation of “lying”, the answer has to be that technically, no, he answered her question truthfully. But the fact that you need to insert that “technically” qualifier in there, for that to be an answer you’re comfortable with, should be a hint that it’s not an answer that’s good enough. Even if he gave a strictly accurate answer to the precise question posed, that’s less important than that he deliberately led her to a mistaken understanding.
A question like this doesn’t exist in a vacuum, outside of any cultural or interpersonal context. A wife who asks whether her husband is seeing another woman is seeking a clarification of a broad situation, not a single isolated data point.
Penn’s stance is that yes, the guy is lying. He might not actually be seeing another woman, as he was asked, but what she was getting at is clearly a more general issue of his marital fidelity. Hiding a same-sex affair and feeling like you haven’t been dishonest, because it technically doesn’t come under the scope of the exact question posed, is a perversion of the actual purpose of language: to share ideas and understanding, such that everyone involved can acquire a more accurate view of the state of the world.
And personally, I’d go further. I think as soon as he’s even having an affair, in the context of a relationship where fidelity and openness and sexual honesty have been agreed, the lie is already present. If you fail to mention something which would undermine a person’s implicit assumptions, when you know those assumptions are there, and when you’ve been instrumental in letting someone use those assumptions as part of their model of the world, then you’re playing a leading role in their deception.
And that’s a lie, in any meaningful, important sense. Saying you “haven’t lied” sounds like it means you’ve been honest and truthful. But if the guy in the above conversation is being honest and truthful, then our vocabulary for dealing with these things is badly letting us down.
This is why I’m not persuaded by the idea that words like “deception” already describe the kinds of misleading behaviour and omissions that I’m expecting “lie” to cover. Lying gets so much more press than implicit deception by cunning and deceitful wordplay. It’s a nice short word, easily and commonly used, and as such it seems to have largely become the yardstick for what meaningful deception is. But restricting its definition in such a sharply demarcated way tricks us into thinking that there must be some substantial, absolute difference between what does and doesn’t count as a lie.
If you haven’t technically lied then, according to the way we often talk about lying, whatever deception might actually have taken place was probably just some relatively minor, trivial, pernickety business. Which is often bullshit.
The guy in my example is cheating on his wife in every meaningful sense that she’s interested in. When confronted with her suspicion and her justified reasons for it, he doesn’t volunteer facts which would absolutely be relevant to her interests, he hides the information she’s seeking entirely for reasons of self-preservation, and he answers precisely within the technical limits of her spoken question – and he probably feels like he’s got away with it without lying. He probably feels in some way relieved about not compounding his guilt with a lie, in a way that he wouldn’t have been able to if she’d phrased her question slightly differently, as “Are you seeing someone else?”.
In that case, perhaps he would have given the same answer, in which case he technically “lied” in a way he didn’t above. But the difference between the two situations is totally unimportant. The phrasing of her question doesn’t make the deceitful nature of his answer and his actions any more or less morally wrong. Or perhaps he might have felt compelled to answer differently; maybe if he were confronted with a direct question like that, he’d feel compelled to answer “honestly”, and ‘fess up. In which case his moral compass rests on such a dumbass set of foundations I don’t even know where to start.
As is often the case, this comes back to the Less Wrong sequences for me, specifically this one, which may be the one area of the rationalist community which has most strongly and noticeably influenced my day-to-day thinking. The map is not the territory; whether or not some given statement of fact can be called “a lie” is not some concrete feature of reality. Words are a means for humans to provide each other with a more accurate mental conception of some feature of the universe. Deliberately steering someone anyway from that more accurate conception – whether through false statements, silence, or “technically true” assertions – should be called what it is.