SELF-RECOMMENDING!

Against Lie Inflation

[Related to: The Whole City Is Center]

I.

I got into an argument recently with somebody who used the word “lie” to refer to a person honestly reporting their unconsciously biased beliefs – her example was a tech entrepreneur so caught up in an atmosphere of hype that he makes absurdly optimistic predictions. I promised a post explaining why I don’t like that use of “lie”. This is that post.

A few months ago, a friend confessed that she had abused her boyfriend. I was shocked, because this friend is one of the kindest and gentlest people I know. I probed for details. She told me that sometimes she needed her boyfriend to do some favor for her, and he wouldn’t, so she would cry – not as an attempt to manipulate him, just because she was sad. She counted this as abuse, because her definition of “abuse” is “something that makes your partner feel bad about setting boundaries”. And when she cried, that made her boyfriend feel guilty about his boundary that he wasn’t going to do the favor.

We argued for a while about whether this was a good definition of abuse (it isn’t). But I had a bigger objection: this definition was so broad that everyone has committed abuse at some point.

My friend could have countered that this was a feature, not a bug. Standards have been (and should be) getting stricter. A thousand years ago, beating your wife wasn’t considered abuse as long as you didn’t maim her or something. A hundred years ago, you could bully and belittle someone all you wanted, but as long as there was no physical violence it wasn’t abuse. As society gets better and better at dealing with these issues, the definition of abuse gets broader. Maybe we should end up with a definition where basically everyone is an abuser.

But a wise supervillain once said, “When everyone is super, nobody is”. In the same way, when everyone is an abuser, nobody’s an abuser.

Right now, if I hear that someone is an serial abuser, I would be less likely to date them, or I might warn my friends away from them, or I might try not to support them socially. The world is divided into distinct categories – abuser and non-abuser – and which category someone is in gives you useful information about that person’s character. I’m not saying that every abuser is an awful person who is 100% defined by their misdeeds and can never be redeemed. But I think the category contains useful information about a person’s character and likely future actions.

But if everyone used my friend’s definition, and we acknowledged that everybody is an abuser – the category stops being informative. “John is an abuser”. So what? Doesn’t mean you should worry about John, doesn’t mean you shouldn’t date John, doesn’t even mean you shouldn’t set your single friends up on blind dates with John. It just means John is a human. Maybe he cries sometimes. So what?

Broadening the definition of “abuser” this far doesn’t help fight abuse or make anybody nicer. It just removes a useful word from the English language. I can still eventually warn someone that John is cruel or violent toward people close to him. I just have to circumlocute around the word “abuser”, in order to find some other word or phrase that hasn’t been rendered meaningless.

(I’m cheating here by talking about “abusers” rather than “abuse”, since there is still a useful distinction between abuse and non-abuse actions. But although the abuse case is less clear, I think some of the same considerations apply – just because an action is abuse no longer means you can be sure it’s especially bad)

But it’s worse than this, because change to a definition doesn’t instantaneously propagate to all of its web of connotations in our minds. So probably some people will continue to use the new definition while still holding the connotations of the old definition. This means bad actors can stigmatize anyone they want:

1. We don’t tolerate abusers around here, right? Right!
2. John’s actions technically qualify as abuse by this incredibly broad standard that includes basically everyone.
3. Therefore we shouldn’t tolerate John.

I previously called this manuever The Worst Argument In The World. Maybe that’s a slight exaggeration, I’m just really tired of seeing it again and again.

II.

This is also my objection to broadening the meaning of “lie”.

The word “lie” is useful because some statements are lies and others aren’t. And although people may disagree on which statements are lies or not (Did OJ lie when he said he was innocent? opinions differ!) everyone agrees on a mapping between states-of-the-world and lie-vs-truth status. When I say “OJ lied”, everyone understands me as making a specific claim about the world, which they can either accept or reject. I don’t have a lot of leeway in how I use the word “lie”; if I’m calling you a liar, I’m making a specific claim about the world.

If “lie” expands to include biased or motivated reasoning, who’s going to throw the first stone? We’re probably all biased to some degree. Does that make us all liars? If everyone’s a liar, nobody is. I can accuse Donald Trump of lying constantly, and you can just nod your head and say “Oh, so you’re saying he’s not a perfect person free from all bias, whatever”. You’ll feel no need to decrease your opinion of him.

Maybe we should only apply the word “lie” to particularly egregious bias and motivated reasoning? But we’ve already abandoned the only defensible Schelling fence. So how will people decide where to draw the line? My guess is: in a place drawn by bias and motivated reasoning, same way they decide everything else. The outgroup will be lying liars, and the ingroup will be decent people with ordinary human failings.

This is my criticism of the original post that started this argument. It declared belief in a near-term singularity to be a “scam” and that believers were “being duped into believing a lie”. Its evidence was listing some reasons people might be biased to believe the singularity was near.

The obvious next step is that someone who believes the singularity is near writes a post listing some biases that singularity skeptics probably hold (for example, the absurdity fallacy). Having shown that skeptics are biased, they pronounce skeptics to be liars perpetrating a massive fraud on the scientific community.

There are a few ways this expanded-definition world becomes different from the world where people restricted “lie” to mean a knowingly false statement.

First, everyone is much angrier. In the restricted-definition world, a few people write posts suggesting that there may be biases affecting the situation. In the expanded-definition world, those same people write posts accusing the other side of being liars perpetrating a fraud. I am willing to listen to people suggesting I might be biased, but if someone calls me a liar I’m going to be pretty angry and go into defensive mode. I’ll be less likely to hear them out and adjust my beliefs, and more likely to try to attack them.

Second, bad actors can use The Worst Argument In The World to prove whatever they want. As long as you’re willing to equivocate and deceive people, you can prove anyone a liar, and then draw on now-obsolete connotations of “liar” to silence or ostracize them.

Third, the biggest beneficiaries are actual liars. Suppose some singularitarian claims that internal Google documents prove they have already created human-level AI. And suppose that’s totally false and no such documents exist. Usually I would accuse them of lying, and this accusation would be enough to alert people that, hey, something has gone terribly wrong here. But if both sides are constantly accusing each other of lying just for having normal human failings, then “you are a liar” no longer carries much weight. I have to come up with some complex circumlocution in order to let people know that someone told a mistruth.

And that complex circumlocution can only last until people realize it too is an exploitable signal. The whole reason that rebranding lesser sins as “lying” is tempting is because everyone knows “lying” refers to something very bad. But the whole reason everyone knows “lying” refers to something very bad is because nobody has yet succeeded in rebranding it to mean lesser sins. The rebranding of lying is basically a parasitic process, exploiting the trust we have in a functioning piece of language until it’s lost all meaning – after which the parasitism will have to move on to whatever other trusted functional piece of language has sprung up to replace it.

III.

I realize this is a kind of long post arguing against a weird thing that not many people are doing. But I think it’s an especially clear case of a broader thing that many people are doing. Words like “disabled”, “queer”, and “autistic” are also gradually shifting meanings, getting applied more and more loosely.

This isn’t always bad! Words are useful because as they separate the world into categories; this suggests a word should apply in more than 0% of cases but less than 100% of cases. Exactly where it should fall in between that range, I don’t know. The broader you make the definition, the better the word’s ability to name things that have even a small level of the relevant quality. But the broader you make the definition, the less power the word will have to separate strong examples of a quality from marginal examples.

I think of this as a sort of sensitivity-and-specificity statistics problem, setting a threshold to divide the population into two groups. If you have a very strict threshold for “abuser”, maybe only someone who inflicts serious physical injuries, then you can use it to separate the most abusive 1% of people from the other 99%. If you have a very weak threshold for “abuser”, so low that 99% of people qualify, then you can use it to separate the 1% least abusive people from the other 99%. If you set it in the middle, you can separate the more abusive half of the population from the less abusive half. If “abuser” picks out the most abusive 1% of people, it transmits a lot of information in a small number of cases. If it picks out the most abusive 99% of people, it transmits very little information in a large number of cases (and now “not an abuser” transmits a large amount of information in a small number of cases!). If the boundary is set at 50%, it transmits an equal moderate amount of information about everyone.

There’s no rule that 50-50 is always the best – for example, if the word “murderer” referred to anyone in the more murderous half of the population, that would be much worse than the system now, where it refers to a much smaller set of people, who you have much more reason to worry about as a discrete group. You’re going to have to find the right threshold for each individual concept.

But it’s never the right decision to draw the line outside the population, so that literally 100% of people fall in one category and 0% in the other.

A few months ago I told a fable about a city. The citizens worried that people living in the outskirts of the city felt unimportant and excluded. So they redefined “city center” to mean the entire city, including the outskirts. Nobody ended up feeling any more important because of this, because living in city center stops being prestigious when everywhere is city center. But it was now impossible to direct tourists to where they wanted to go, and people had to invent new phrases like “the part of the city where there are the most tall buildings” in order to discuss city center.

The moral of the story is: don’t set thresholds for category membership so far outside a distribution that they stop conveying useful information.

This entry was posted in Uncategorized and tagged . Bookmark the permalink.

325 Responses to Against Lie Inflation

  1. name99 says:

    Obvious analogy to California “This chemical is known to cause cancer” law…
    When everything is “known to cause cancer” no-one pays any attention to those notices.

    • JulieK says:

      Or car alarms that are so sensitive that they’re always going off for no reason, with the result that whoever hears it assumes it’s a false alarm.

  2. deluks917 says:

    There has been some discussion on rationalist facebook about ‘alarm raising’ and how rationalists do it too much. I wonder if this is part of the explanation for why people are using lying in this way.

  3. haroldedmurray says:

    Definitely seen this sort of thing coming from the social-justice left quite a bit. In particular, terms like “abuser” and others seem to have ever-expanding definitions, even legally (for example the expanding definition of “rape” to mean many lesser acts, which may not even require mens rea). I don’t believe that expansion is done on purpose, maliciously, and I don’t think the expansion is done consciously. It’s probably more of a memetic evolution, in that the people who make these subtle equivocations win more arguments, more people hear those arguments, and the boundaries get expanded.

    • Aapje says:

      I think that it is a logical consequence of that culture having a very strong slave morality. In excessive slave morality, there is too little concern for the needs of the capable, non-abused, ‘default’, etc; so to get your needs met, as well as prevent excessive burdens placed on you, it becomes imperative to express your needs in terms of disability, abuse, exceptionalism, etc & to deflect demands by presenting yourself as less capable, abused, exceptional, etc.

      So you don’t get your feelings hurt, you experience ‘violence.’ You don’t just have an unpleasant and undesired experience, but you were harassed or abused. You don’t just chafe against the gender norms, but you are non-binary.

      Excessive slave morality is a utility monster and not exaggerating your issues is a recipe for abuse, although ironically not the kind of abuse that the culture cares about.

      • Dragor says:

        Do you have any links to anything discussing “slave morality” in the way you have been using the term? I have never encountered the term how you use it and it interests me.

        • Luke the CIA Stooge says:

          Neitzsche’s on the geaology of morals is the core text, with Thus Spake Zarathustra being well spoken of, but I haven’t read it yet.

          But honestly I recommend just searching master and slave morality to get the basics though. Nietzsche is highly influential but not very accessible.

          A fun read though

          • Aapje says:

            Note that Nietzsche probably intended to write more to criticize master morality, but never got around to it, due to going mad, which is presumably why he was relatively easily to abuse by the Nazis to advocate master morality.

            However, the idea of the Übermensch is that he is neither master nor slave. The title of his book ‘Beyond Good and Evil’ suggests a nihilist outlook, but is actually intended as the opposite: a rejection of the definitions of good and evil that systems that espouse master and slave morality use to control people. Going ‘Beyond Good and Evil’ is to transcend to a level beyond ideology.

          • Protagoras says:

            Nazis weren’t master morality. Nietzsche actually tended to associate the German anti-semitic movement with slave morality; this is an oversimplification, of course, but then so is the whole master morality/slave morality picture. But blaming scapegoats is very much not a master morality move, for example.

            Though in his more nuanced moments Nietzsche did say that most modern thinking is a mixture of elements, he seems to me to have underestimated the role of master morality in the modern world; it is likely that he didn’t criticize it very much because he didn’t see it as worth the effort, when in his view nobody thought that way any more anyway.

      • niohiki says:

        There’s another layer to it, aside from the incentive to exaggerate abuse to remain “competitive” in the attention market of such slave morality. There’s an incentive, even while pretending to be a “strong” individual, to overreact to potential threats to the “weak” as a way of justifying arbitrary enforcement of repression. It’s the social justice version of “every day there are more terrorists, so you cannot bring water on a plane and we are going to backdoor your phones”.

        Excessive slave morality is a utility monster

        Also, I’ll keep this in my quote book.

    • MereComments says:

      This strikes me as exactly correct. As long as strategy A seems to be unstoppable, you will see more people trying to apply strategy A, even when previously it wouldn’t have applied.

    • AG says:

      It’s a case of slippery slope in action, which both sides gleefully use. Someone challenging USFG is labelled an un-American traitor. Questioning is rebellion. Opposition is persecution. Pro-life are misogynists and Pro-choice are murderers. Immigrants are invaders.

      Mountain redefined to cover molehills.

      It’s a side-effect of how we don’t grapple well with systemic phenomenon, always wanting to ascribe their effects to individual actors.

      • realitychemist says:

        It’s a side-effect of how we don’t grapple well with systemic phenomenon, always wanting to ascribe their effects to individual actors.

        Would you mind expanding on this idea as it relates to the current context? I think the phenomenon (people don’t grapple well with systemic phenomenon) is true, but I’m not exactly sure how lie inflation is a side effect of this. They seem like separate issues.

        • Pete Michaud says:

          One obvious application is “Bob is a racist” — do you mean that Bob specifically hates and tries to harm members of other races, or do you mean the global/national system that Bob was born into arbitrarily grants Bob certain advantages that he is not actively disavowing, but he’s an otherwise affable fellow?

          • nyc says:

            But that suffers the same criticism. You see the same expansion applied to actions as people. Charges of racism are no longer limited to Jim Crow laws and lynchings and Nazi genocide. It’s now labeled racist to take into account credit history when approving mortgage applications because of the racial disparity in credit scores, even if credit score is a good indicator against defaults. Yet that is a systemic choice made by a bureaucracy rather than an individual.

        • AG says:

          @realitychemist: The famous Moloch post kind of covers this dilemma, of how incentives build systemic phenomena, but our instincts for addressing these issues is “punish bad actors,” rather than changing incentives.

  4. SteveReilly says:

    Weirdly, the lexicographer Samuel Johnson used the word lie in an even more expansive way. From Boswell’s Life:

    Johnson had accustomed himself to use the word lie, to express a mistake or an errour in relation; in short, when the thing was not so as told, though the relator did not mean to deceive. When he thought there was intentional falsehood in the relator, his expression was “He lies, and he knows he lies.”

    ETA: Back to the OP. Doesn’t “gaslighting” also fit into this? I’ve heard people use it when they just mean “lying” (in the ordinary sense), rather than when discussing some nefarious plot to make a person feel insane.

    • Matt M says:

      Sounds like the inverse of the Costanza Method: “Remember Jerry, it’s not a lie, if you believe it.”

    • liskantope says:

      *sigh* The ridiculously broadened definition of “gaslighting” in popular use today is an example of a definition-broadening that is driving me up the wall. It’s not only used to mean “lying”; I now routinely see it used to mean “insisting something that contradicts my perception of reality” with the charge of deliberate dishonesty just kind of implied without justification.

      • Matt M says:

        I think 99% of common uses of “gaslighting” are incorrect.

        It’s getting to the point where, like “literally” before it, we’re probably going to have to just change the definition to encompass common usage, rather than attempt to educate people on the correct usage….

        • LCL says:

          Came here to shake my fist about the degradation of “literally” by this same process; I see I wasn’t fast enough to reply! Have we really given up though? What’s the replacement I can defend instead?

          a parasitic process, exploiting the trust we have in a functioning piece of language until it’s lost all meaning – after which the parasitism will have to move on to whatever other trusted functional piece of language has sprung up to replace it

        • Emperor Aristidus says:

          I don’t think “literally” is quite the right example, actually. People gripe about its use in sentences like “She literally killed me”, but I don’t think those who write such sentences misunderstand “literally” as a synonym of “figuratively”, as is commonly construed; rather, it seems to me that they use “literally” as a blank term that simply adds emphasis. It doesn’t actually expand the definition of the word, any more than the use of constructions such as “she f—king killed me” implies that whoever utters them is confused about what the f-word means, or that the original definition of it is being “lost” and “replaced”.

          • melolontha says:

            It’s not the same process, but I think ‘literally’ is a fine example of a useful word being ruined through definition creep. You’re right that it’s not simply a case of expansion — it’s more that the word has taken on an extra, separate sense — but the result is very similar. ‘Literally’ was useful for removing ambiguity in the telling of real facts or events that might otherwise sound metaphorical or exaggerated. But because literally-as-intensifier can be used in the exact same contexts, its surge in popularity stripped literally-literally of its usefulness. The old sense hasn’t yet been completely lost or replaced, but that’s where things seem to be headed — the more popular it becomes as an intensifier, the less useful it becomes as a disambiguator, and so the shift continues.

          • JulieK says:

            The same process happened to “very, “truly” and “really.”

          • Tom Chivers says:

            I get a bit annoyed by the idea that “literally” has been ruined. It’s been used for emphasis rather than simply to mean “to the letter” for centuries (off the top of my head, Dickens and Louisa May Alcott both used it, among others). More generally, insofar as the language needs a word or simple phrase meaning “in the non-metaphorical meaning of the word”, it will find one. There are well-studied processes for how words and meanings are lost and replenished. (And, to be honest, it’s usually pretty obvious from context whether someone means “literally” literally or not.)

          • And, to be honest, it’s usually pretty obvious from context whether someone means “literally” literally or not.

            Often, if someone quite literally means “literally” in the proper sense, they will say “quite literally” or add some other qualifier before the word “literally”.

          • melolontha says:

            I know that using literally as an intensifier is not new, but has it really not become much more common over the past, say, 20 years?

            I can’t think of an easy way to verify that, and it could easily be a misperception based on e.g. the internet exposing me to a wider range of styles and dialects. But I tend to suspect that the ‘it’s always been used that way’ side overstates its case when it claims/implies that nothing much has changed; my perception is that you used to be able to use the word ‘literally’ to reduce ambiguity, and it would be taken at face value in the absence of contextual clues to the contrary, whereas now that’s no longer possible.

            Of course this sort of shift is extremely common, and our language will adapt to fill this gap if necessary, but there’s no law of linguistic evolution that says no change can be for the worse. And in the period between losing ‘literally’ and gaining a good, concise, (temporarily) unambiguous replacement, I think we really are slightly worse off.

          • Nick says:

            The word literally has definitely become more common, per Google ngrams. But the data only goes up to 2008 (so missing eleven of the last twenty years), and that’s based on books, not on speech or unpublished material like texting or Facebook posts. ETA: To be clear, I’d wager that for those reasons ngrams is underestimating the shift.

          • AG says:

            As JulieK points out, though, literally is only the latest in a long line of words with that function. Other words going through the process include “actually” and “seriously.”

            So stop using “really” as a means of emphasis, y’all. It’s supposed to connote that it happened for real.

          • Don P. says:

            There are complaints about “literally” mentioned in the original 1920s edition of Fowler’s English Usage; it’s one of those things that’s always just about to ruin the language.

          • AG says:

            You mean, it’s literally always about to ruin the language, right

      • James Banks says:

        I think sometimes people are trying to talk about the way things feel and the things that happen inside them rather than about an act in the socially visible world. The effect is what they know, regardless of how it’s produced. Certain effects make you want to push back, cry out, etc.

        Gaslighting (the original definition) has a certain quality to it. If you feel like your memory is getting re-written, you want to call “gaslighting” (the new definition). It may not really be gaslighting, but there is a morphological resemblance between the feeling in your brain of someone driving you crazy over time on purpose, and this one act that starts to make you go crazy a little bit and puts you on edge toward the person who did it.

        If you accuse someone of gaslighting you, they can just deny it and then you can doubt yourself as to whether they were gaslighting you in the first place, whether it fit the definition. That’s not gaslighting, but like gaslighting it involves you not trusting yourself in a kind of visceral way.

        If you’re the kind of person susceptible to this abuse, you may be prone to learned helplessness. A certain amount of successful abuse trains you for more. Eventually you find yourself feeling really bad around someone, arguing with yourself whether they’re right or not, and you can’t really justify the bad feeling — until maybe you shut down, leave them, and slowly figure out what was going on the whole time. So you learn to go with your inner feeling even if it doesn’t connect to what’s socially provable, can’t be fit to a socially-established definition. (Ideally you would strengthen your mind so that you have a firmer grasp on your memories in the first place.)

        Ideally, there would be two different, well-accepted terms for the extreme version and the looser version of “gaslighting” (probably applies to other terms too, like “lie”, “abuser”, “rape”). The looser version is exactly as bad as it is and is not the stricter version.

        • caryatis says:

          That’s really interesting (and charitable!) It sounds plausible that there is a certain personality type whose response to someone disagreeing with her is to think, “Oh, I guess I must be wrong.” This sort of person would be both more vulnerable to manipulation by bad actors and more likely to find herself bulldozed by well-meaning people who themselves can easily maintain their own beliefs in the face of disagreement and don’t realize that the other person can’t.

          So this sort of person might repeatedly be in situations that she comes to see as gaslighting, merely because she is so very quick to question her own perceptions when stronger personalities nudge her to.

          • OxytocinLove says:

            Can confirm, I was such a person for a few years. It took significant mental effort to undo and I’m still prone to it in certain types of situations. (It’s telling that you used a female pronoun, I suspect this is actually quite common among AFAB people)

        • Peter Gerdes says:

          I don’t think that’s quite right. In particular, if that is all the content someone wished to convey they would say something like “it feels like gaslighting” or “it makes me feel like I’m being abused.” However, people very much don’t say that exactly because it doesn’t carry along the normative connotations.

          But I do think something similar is going on. People are saying “This is *bad* in the way that abuse is bad and should be seen as such. Perhaps coupled with a claim that this is sufficiently important/bad enough that we should condemn it the way we condemn abuse.

          This then explains the way people react when they say ‘but shouldn’t this be considered abuse’ or whatever. That’s a puzzling reaction if we are just taking abuse to define a certain range of behaviors. I mean you wouldn’t say “but shouldn’t a handshake be considered a wave.”

          In other words what is going on is that they think we ought to regard that behavior the way we regard abuse. While often it is effective to change language to force behavior change it can also backfire since, when others don’t agree, it risks undermining our norms against behaviors there is widespread agreement are unacceptable because.

          • James Banks says:

            Good point about the normative weight of what’s being said.

            If you say “It felt like X”, then a hostile interpreter will imply or say that maybe you were wrong. “Feels like X” is a weak statement, that could be wrong (and you know it just by saying “feels like”). If you say “It was X”, you’re making a definite statement, lacking the distance from reality of “felt like”. Also, maybe it’s *your* problem you felt that way. “Oh no! I’m so sorry you felt that way!” (aside, through gritted teeth: “I’m not responsible, you’re the one with mental problems that you should get over so you’re not blaming me anymore.”) But “is” is not about feelings. Saying “is” is a stronger way to speak, psychologically if not epistemically.

            People sometimes speak in order to point out facts. Other times the speech is like a lunge, or stab, or pushing away, trying to gain personal power, get an intention into the world, get heard on that level. Speaking in order to make an argument can go both ways.

            People (often) want power (in the words of a Swans song(*)) “because it feels good” — something unholy in that, but it’s also comfort for the afflicted.

            (*) (“Power and Sacrifice”, if you’re curious)

      • Jake Rowland says:

        Maybe I just haven’t been to the right (wrong?) parts of the internet, but the few times I’ve seen the word “gaslight” used it’s seemed broadly correct. I’ve never seen it used to just mean lying. It’s always someone lying about something they know the other person knows to be false. Like if someone lied to me about my own prior actions. This isn’t exactly a coordinated and extended campaign to make me doubt my own sanity, but it does seem broadly in the same vein if you squint. I think the liar is usually counting on people being non-confrontational enough to let the lie slide. This seems like a useful thing to have a word for.

        • Gerry Quinn says:

          I’ve been accused of it on internet arguments on Ars Technica (which actually has fairly intelligent commenters). I think they meant I challenged their preconceptions with effective arguments and/or examples.

        • SaiNushi says:

          I don’t know about the rest of the internet, or the world. But I lived with an sjw feminist who is in her mid 20’s for a few months this year, and she used “gaslight” as any situation in which she disagreed with another person about what happened, without any consideration that maybe it’s just a case of differing memories.

          • Simon_Jester says:

            This sounds like the kind of thing a lot of people do (“refuse to consider that maybe it’s just a case of differing memories”), only this specific person uses the word “gaslight” for it instead of just saying “I’M NOT CRAZY! YOU’RE CRAZY!” in a cruder form.

            If the English language even had other words for “consistently lie about past events so that a person starts to doubt their own memory,” those words would also get spammed by the kind of person who refuses to admit they’ve made mistakes.

      • Tarpitz says:

        My concern about gaslighting is the temptation to apply the charge to cases where in fact two people simply have differing (probably both mistaken) memories of the same event. Most people in such cases have a strong natural inclination to believe their own memory is correct, and the other person either lying or mistaken; a significant minority will viscerally doubt themselves and feel like they’re going mad. Neither reaction is healthy, but it’s very difficult to really on-board “we don’t know what actually happened and absent some external evidence we never will”.

        • Matt M says:

          Most people in such cases have a strong natural inclination to believe their own memory is correct, and the other person either lying or mistaken

          But even then, I think there’s a pretty big leap between “this person is stupid/crazy/wrong” and “this person is engaging in a deliberate tactic to try and undermine my own confidence in my own memory/assessment.”

          Like, I’ve occasionally had radical disagreements with say, a girlfriend, on the memory of specific events. Sometimes to the point where I’ve thought “She’s literally delusional here.” But never to the point where I’ve thought “She knows that my memory is correct but is trying to convince me to believe something untrue.” That charge requires such malice that you need some evidence to back it up which goes above and beyond simple disagreement.

          • Robert Beckman says:

            I think there’s a sizable cadre that don’t need much evidence to conclude that their opponent is indeed gaslighting them, as opposed to being simply in error.

            There’s been a change over the last decade (note: change may be in me – now I notice it, or it was always present at the same rate but is more visible, or is rising in frequency) in assuming that ones political opponents are acting in bad faith, largely among certain college age factions.

            If you know that your opponents only oppose you for improper reasons (they’re paid, they’re full of ill will, etc) then concluding that any singular example of them is gaslighting becomes easy: you know they’re acting in bad faith, and the bad faith version of a disagreement over history is gaslighting, so that’s what they must be doing.

            Whether your model of your opponents reflects reality is immaterial, of course.

    • Protagoras says:

      Hardly surprising. Even by the usual standards of celebrities for this sort of thing, Johnson was full of awful habits, and, again to an even greater degree than usual for celebrities, his fans seem to cheerfully excuse or rationalize them.

    • Doctor Mist says:

      “I remember back in ought six, there was this fellow Bill, “Big Bill” we used to call him, … no, I tell a lie, it was ought seven.”

      • Creutzer says:

        This is clearly a fossilised phrase from a time where perhaps “lie” still meant “falsehood”, as opposed to “knowingly asserted falsehood”. Cue: the present “I tell”, which in contemporary English would be ungrammatical outside this fossilised phrase, it would have to be replaced with the progressive “I’m telling”.

  5. Michael Arc says:

    But the thing under discussion is the central tendency of fraud, not an edge case. If “you” means “a particular homo sapiens”, not “a story you tell yourself” then “you lie” means “that homo sapiens says something other than what it knows to be true” hence phrases like “lie to yourself” being coherent.

    • Scott Alexander says:

      I agree that “says something other than what you know to be true” is a good definition for lie, but it sounds to me like it requires malice or conscious knowledge. A tech entrepreneur who believes the hype and repeats the hype isn’t saying anything he knows to be untrue.

      Re: lying to yourself – Shakespeare says Brutus is “at war with himself”, even though he has not passed a formal declaration of war against himself or send troops to invade himself. He means Brutus is experiencing conflict with himself; he’s pursuing many conflicting goals. But the US is experiencing some level of conflict with China right now and the two countries are pursuing conflicting goals. Still, it would be wrong to say “the US is at war with China” (without clarifications like “cold war” or “trade war”), because in the context of international relations, war already has a specific strong meaning and does suggest declarations and troops.

      It’s fine to use terms metaphorically in cases where it’s obviously a metaphor. Since you can’t (strong sense) lie to yourself or (strong sense) declare war on yourself, everyone knows these are metaphorical. But if you’re using it in a context where it could mean the strong sense, you need to be more careful so people don’t interpret it as the strong sense.

      • Steve Sailer says:

        “her example was a tech entrepreneur so caught up in an atmosphere of hype that he makes absurdly optimistic predictions”

        I actually had to give a deposition in a shareholder lawsuit against a corporation’s management that was exactly over that. The CEO believed strongly in his company and his own reality distortion field.

        Usually, management quickly settles in this kind of lawsuit, but the CEO felt it a point of honor and battled the lawsuit for years, ultimately winning in court.

      • Tarpitz says:

        Dangerous to anchor definitions on a word as slippery as “know”…

        It seems to me that there are inevitably going to be grey areas relating to the strength of the putative liar’s belief in their claim. I don’t think we can even model strength of belief as simply as assigned probability, because people’s thoughts are generally murkier and more complex than that at any given time, and their thoughts at times prior to the speech act at least plausibly also have bearing on whether we consider the speech act a lie. If I’m consciously and deliberately suppressing my doubts about what I’m claiming, am I lying? How about if I’ve deliberately temporarily brainwashed myself into believing what I’m saying so I can say it convincingly (which is not so far from what acting sometimes is)?

        Less importantly, I think it’s at least plausible that someone who mistakenly believes themselves to be saying something false is lying, which doesn’t work under your definition if we take true belief (in the technical sense) to be necessary for knowledge.

      • sharper13 says:

        A tech entrepreneur who believes the hype and repeats the hype isn’t saying anything he knows to be untrue

        Yeah, we already have a different word for that. He’s not a liar, he’s a fool.

      • Exetali Do says:

        I still recall back in college, a class called “Truth and Falsehood in History”, where our teacher led us to a definition of “lie” through the Socratic method. (The impetus was various ancient sources accusing Herodotus of being a “liar”.) The result was that we decided that a “lie” had 2 primary aspects and 2 secondary aspects.
        1) It has to actually be false.
        2) The person telling the lie has to believe that it’s false.
        3) The person telling the lie has to be trying to get you to believe it.
        4) The person telling the lie has to be trying to harm you, by getting you to believe it.

        Without 1, we’d have to say the person was *trying* to lie, and thought that they were lying, but failed to actually lie. Without 2, we’d say that the person was just wrong: factually incorrect. Without 3, the category expands to include “yo mama” jokes and suchlike, which might technically be lies, but they aren’t a central part of the concept. Without 4, we’d call it a “white lie”, along the lines of Jon Snow telling Robb Stark that Catelyn was nice to him, which again is technically a lie, but is different enough that it gets its own subcategory.

        YMMV, but ever since that one afternoon in the Classics department, it’s seemed glaringly obvious to me. It might not be an entirely complete definition, but that’s nothing that a few smart people can’t solve with a little co-operation. *shrug*

  6. liskantope says:

    Great take on something I’ve been concerned about for a long time: the broadening of definitions of terms (which are usually widely recognized as referring to Bad Things, because naively it seems advantageous to oppose even much milder versions of those Bad Things every more strongly) without paying due caution to the fact that this will cause a lot of people to take everything falling under that term’s definition less seriously (including the worst of the Bad Thing it referred to).

    But I think there’s another issue with the “lies” example, which in fact was the first issue that came to my mind when reading the post. It has to do with what I call “being able to separate out different layers of disagreement” (often my feeling is that these layers lie at different levels of “meta”, whatever that should mean) — this is rather closely related to the dangers in being a “low-decoupler”, to use a term that’s caught on with some in the rationalist-sphere. The idea is that if “lie” is defined to include “expressing a belief arrived at through unconscious bias”, then accusations of lying obfuscates each dispute in which it comes up by causing one or both parties to confound the issue of whether someone is deliberately telling an untruth with the issue of who might or might not be arriving at false beliefs through bias. The debate over who is lying should be treated as a discussion with two separate parts: (1) is someone deliberately saying something dishonest; and (2) is someone arriving at false beliefs through bias. But instead “lying” is viewed as one particular sin and the person making the accusation may oscillate between one definition and the other in the course of the dispute perhaps without even realizing it.

    It reminds me of a different example that I’ve brought up several times: defining “insult” to mean “wrongly claiming something bad about someone” as opposed to just “claiming something bad about someone” (which I believe is actually the correct definition). Defining “insult” in the former way invites confusion over whether we’re arguing about the rightness/wrongness of choosing to say nasty things about each other or whether we’re arguing about whether said nasty things are in fact true (which is typically much harder to demonstrate one way or another).

    I guess it’s not obvious that the broadened definition of “lie” would result in this type of confusion — perhaps each debate would only involve either disagreement over who is knowingly saying untruths or disagreement over who is arriving at wrong beliefs, but not both. However, my instinctual reaction to the broadened definition is to think that it would cause the issue I describe.

    • Dan L says:

      The debate over who is lying should be treated as a discussion with two separate parts: (1) is someone deliberately saying something dishonest; and (2) is someone arriving at false beliefs through bias. But instead “lying” is one particular sin and the person making the accusation may oscillate between one definition and the other in the course of the argument perhaps without even realizing it.

      Blurring that line is the speaker whose utterances are designed to serve some instrumental purpose orthogonal to truth. Effectively pushing back on that seems a much more salient problem than any definitional slide, IMO.

      “Reckless disregard for the truth” is a term of art and a high bar to clear, and if it isn’t captured by “lying” then I’m worried as to whether that word is being defined into impotence.

      • Error says:

        I’m pretty sure the word for what you’re describing is “bullshit“, as distinct from lying.

        • Dan L says:

          That is absolutely the work I was obliquely referencing. But I would argue that there is no reason not to apply the same affect for bullshit artists that we normally would reserve for liars.

    • liskantope says:

      I was starting to write another comment reframing my point in motte-and-bailey terms, because I’m afraid my parent comment may have come out a little garbled. Then in the course of writing, I realized that the “insult” example kind of works the opposite way: in that case, the definition that I claim is wrong (“to say something demeaning and untrue about someone else”) is strictly narrower than the definition I’m claiming is right.

      Hmm, I feel like there’s something in what I was trying to say, but I’d better stop commenting for now and come back when I’m less tired (which unfortunately may not be tomorrow…).

    • Peter Gerdes says:

      I agree with most of what you say but am going to nit-pick over the definition of insult.

      Insulting is a fundamentally social act. It’s hard to define without reference to the intent to insult. Frequently insults don’t even have factual content. For instance, consider boys on the playground calling each other ‘gay’. They don’t mean to make a factual claim about sexuality or even to suggest that the other person acts in an effeminate way. Indeed, even when everyone involved (e.g. everyone involved is actually homosexual in a social environment like SF) doesn’t see being homosexual as bad saying “You’re gay” can be an insult just because everyone understands it is intended as an insult.

      On the other hand if I truthfully testify at a trial that I saw Bob steal or hit someone I haven’t insulted him even though I said something bad about him.

      • liskantope says:

        Yes there is some social context and a certain kind of intent sprinkled in to the essence of what we call an “insult”. But that’s still kinda-sorta orthogonal to the rightness or wrongness of the charge against someone else: I can put someone down with the social intent of making them look/feel bad but still sincerely believe my putdown is true — the fact that I believe that thing may even be the main reason I dislike that person to the point of wanting to make them look/feel bad! Moreover, I think “insult” still has a somewhat broader definition than what you’re implying. If you testify at a trial that Bob did something awful, an ally of Bob’s who believes in his innocence and is incensed that you’d make such a horrible accusation might well consider your testimony to be an insult to Bob (or to his honor, or whatever). Or, to give a current example (and no I really don’t want to open a debate on what actually happened here, I’m jumping to this because it’s happening right now and seems too perfect an example to pass up), one of the House Democrats just made a motion to impeach Trump over his recent considered-racist-by-many remark. I haven’t searched through the opinions of many politicians who are defending Trump in this instance, but I imagine some are saying, “How dare he insult Trump by implying he’s racist!”, even though the motion was made with the explicit primary intent of getting Trump impeached rather than just scoring “playground put-down points” against him.

    • peterispaikens says:

      I feel that both the “lie” case and your examples share the same underlying social mechanic as the euphemism treadmill that’s probably an unavoidable natural order of how things work.

      Let’s use “dumb” as an example. It (at least originally) refers to muteness, however, as true muteness is not that common, the majority of its use is towards people who are clearly not mute, but have behaved as if they’re mute or insulted to be mute. This inevitably means that in the common usage the word “dumb” refers not to muteness but to generic inarticulate behavior and thus we get to the word now commonly, at least in NA, meaning “stupid”. And the word “stupid” has similar origins from originally meaning “stunned”.

      So there’s the natural tendency whenever there’s a name X for some bad minority condition or group with low status, that name will often intentionally be used (for example, as an insult or exaggeration) to refer to *something else*, something that’s clearly not-X, in order to associate it with X and its badness. In that instrumental usage, the goal of the communication is not an accurate representation but intentional stretching of boundaries as much as feasible to associate the not-X target with the low status X; and such usages are common enough so that we should expect that over the course of generations the meaning will migrate in the obvious predictable direction. Like the euphemism treadmill, like the many other examples mentioned here (such as the boundaries of the word “rape”), it seems like a single sociolinguistic phenomenon of how languages systematically change over time.

  7. Douglas Knight says:

    One man’s modus ponens is another man’s modus tollens.

  8. nelshoy says:

    Would like to point out that hyperbolic rebranding is already happening on accelerated discourse zones eg Twitter.

    Telling false statements is usually no longer described as mere “lying” if someone *really* wants to drive the point home. Even coming from different premises is described as “gaslighting”, or literally trying to drive someone insane as a form of abuse.

    • Conrad Honcho says:

      On the flip side, consider “fake but accurate.”

    • Reasoner says:

      Twitter is “insane”.

      • DinoNerd says:

        That’s what you get from a medium that explicitly excludes anything long enough to be more than a soundbite, even if some people use it by posting a series of tiny messages that together comprise a mini-article.

  9. Joy says:

    A great point about the informational content of a term!

    As for the lie, would be nice to have a term for “An utterance or a statement intended to deceive or mislead”, whether the utterance is technically true or not. Of course, this would cover a lot of consumer-oriented legalese, and most of the statements requiring fine print, but it might be a good thing.

    • thad says:

      Does “intentionally deceptive” or “intentionally misleading” not work for some reason?

      • deciusbrutus says:

        It includes statements that are true but intended to deceive or mislead.

        I for one reject any definition of “lie” that includes any action taken in a table game that complies with Hoyle’s Rules for that game-betting in poker can be intended to deceive or mislead.

        • Confusion says:

          If definitions need all kinds of caveats that obscure their general meaning, then they make room for malicious actors to start arguing semantics: “it’s not a lie, because politics is a game”. It seems better to me to have definitions that are understood to be suspended for very particular situations. So defining lie in such a way that one would be strictly be lying in a game of poker seems a better choice to me.

        • ARabbiAndAFrog says:

          Why not? I think you caught up with judgment tackled on “Lie is telling false information and also a bad thing, therefore a good thing, such as lying in a game that revolves around lying can not be lying”

    • The definition of “lie” I am familiar with is “an untruth intended to deceive.” A false statement that the speaker believes is true doesn’t qualify. Nor does a truth that the speaker believes is false. Nor an untruth intended to be taken as a joke.

      I think some people use “lie” to simply mean “untrue statement.” I can remember using it that way in my first novel, where the speaker says something, then follows it with “I lie–(corrected statement).” But to my ear, that’s a rhetorical play on the correct meaning—calling it a lie is a joke.

      • Confusion says:

        I’m not certain about English, but in Dutch it’s perfectly normal to say something and follow up with “No wait, I’m lying, it’s actually …”. You retroactively and earnestly declare the previous statement to have been an unintentional untruth.

        When someone asserts you said something that you believe to be untrue and you’re not interested in disputing whether or not you said it, saying “Then I lied” is fine and is understood as fessing up to a mistake/misunderstanding.

        But those cases are sort of colloquial, lack a certain gravity and do not reflect badly on the person saying it.

        • Martin says:

          Yes, but the word “lie” is used ironically in these cases. It’s not like these things are actually considered to be lies.

          (Maybe that’s what you meant, but that wasn’t clear to me.)

      • Martin says:

        I would even go so far as to say it doesn’t matter if it’s true, what matters is if the speaker believes it’s true. So e.g. when I tell you, with the intent to deceive, it’s sunny in San Fransisco while I think it’s raining, and it turns out that it’s sunny after all, I would still consider that a lie.

      • SamChevre says:

        Your definition of “lie” is the one I am accustomed to.

        My wife takes the stronger position that a lie is “a communication intended to deceive”–even if it’s true, or consists of not contradicting someone. In my historical social world (which is VERY strict about telling the truth), telling the truth without helping people have a clearer idea of the real state of the world is an art form–she dislikes that intensely.

    • Conrad Honcho says:

      This might be the difference between a “lie” and a “lie of omission.” A lie is a knowingly false statement intended to deceive. A lie of omission is a true statement intended to deceive.

  10. albatross11 says:

    It seems like this has to do with the incentives for fiery rhetoric. If I get rewarded by more attention/clicks for making more extreme claims, then nobody will ever be a little too authoritarian, they’ll always just be Hitler. Racism, sexism, fascism, sexual misconduct, all get more and more widely applied.

    • Radu Floricica says:

      Incentives also make fighting against it a bit futile. Sure, we’ll read the post, agree with it, maybe use expanded meanings a bit less. We’re 0.00001%. Just like an older article praising the virtue of ignoring a current event when it’s better off not being publicized rather than debated.

      I sometimes wonder if Gawker’s fall put even a slight dent into that business model. That was a very strong form of discourse correction: a direct legal attack to the heart of the problem that ended with complete victory and the bankruptcy of the perpetrator. I don’t know what it would mean if it worked, but if it didn’t… I’m pretty sure we’re better off learning to surf the wave rather than trying to do anything about it.

  11. John Schilling says:

    I am generally opposed to insult inflation, and e.g. strongly discount “Nazi”, “Fascist”, “rapist”, etc, unless there is clear contextual evidence that something approaching the classical and central meaning is applicable.

    But “liar” is a more difficult case, because anyone who remotely deserves it will almost certainly, well, lie about it. If you insist on a narrow, literalist definition where a person can only be called a liar if they can be proven to know that their statement is false, then “liar” becomes useless in the opposite direction – you can’t actually apply it to any person outside of a highly abstracted thought experiment, because they can claim to believe what they are saying and you aren’t a telepath.

    Both the too-broad and too-restrictive definitions of “lie”, become near-synonyms of “wrong” and make it difficult to describe degrees of culpability in propagating wrongfulness. There’s use for an intermediate term saying, in essence, “you ought to have known better than to say that, on account of it needlessly propagates falsehood or uncertainty, and we don’t actually care what combination of recklessness or malice is responsible, we’re not going to trust you any further”. This is what “lie” in practice usually means, and it is a useful term to use alongside e.g. “wrong”.

    So, on the one hand I agree that we don’t want “lie” to be inflated to include all forms of wrongness including honest mistakes, but I also don’t want it to be so narrowly defined as to include depraved indifference to the truth.

    • Nick says:

      But “liar” is a more difficult case, because anyone who remotely deserves it will almost certainly, well, lie about it. If you insist on a narrow, literalist definition where a person can only be called a liar if they can be proven to know that their statement is false, then “liar” becomes useless in the opposite direction – you can’t actually apply it to any person outside of a highly abstracted thought experiment, because they can claim to believe what they are saying and you aren’t a telepath.

      This is a bullet I’m willing to bite. I already follow this, personally.

      I also want to add that there’s a difference between concluding for yourself that a person is lying and actually accusing them of it. Of course accusations should be held to a standard of evidence if we think the purported action was a serious wrong—and likewise, if accusations mean anything at all, there should be a penalty for deliberately making unsubstantiated accusations too. That doesn’t mean you can’t privately conclude a person is probably lying and act accordingly.

      ETA: to be clear, by “a standard of evidence” I don’t mean beyond a reasonable doubt or anything like that. A weaker standard is fine most of the time, of course. So I don’t actually believe that as described the term would be useless.

      • John Schilling says:

        That doesn’t mean you can’t privately conclude a person is probably lying and act accordingly.

        So, “liar” is a word that you think but never say? I think I would prefer to have it as a word whose generally-understood meaning encompasses things that can actually be said.

        • Nick says:

          You may have missed my last edit—I don’t believe that as described the word is useless.

          Look, it’s pretty hard to prove someone has cheated on his spouse. I’m sure it’s easier to catch him in flagrante delicto than to catch him lying, but not that much easier. That doesn’t mean it’s a useless concept, or that using it to mean “being too flirtatious around other women” instead is going to be better. It just means public declarations that he’s a cheater are necessarily rare, because people will want evidence if they’re to take that sort of accusation seriously.

          If you want a general rule, watering down the meaning of something is a bad response to lacking sufficient evidence to persuade others that it applies.

        • JulieK says:

          Does it matter that much?
          You can still say “Bob’s statement is incorrect” and “Bob often makes incorrect statements, so you should be cautious of believing him.”

          • Watchman says:

            Theres a substantial implied difference between lying and being incorrect though, and that is about bad faith.

          • John Schilling says:

            What Watchmen says. We need to distinguish between people who are wrong in good faith, and people who are wrong in bad faith. And we need a word that carries a negative moral connotation for the people who are wrong in bad faith. It is much less important that we be able to finely distinguish between the specific ways that people are wrong in bad faith.

            Indeed, it can be actively bad to be overly-specific in how we assert that people are wrong in bad faith. Doing so, allows them to claim that they are not wrong-in-bad-faith in the very specific way we accused them of being, even if that is what is colloquially known as an outright damn lie on their part, and cast the broadly-correct accuser as the bad-faith actor even though the reverse is in fact true.

            We need a simple term that says “You are wrong in bad faith, you are a bad person because of it, we shouldn’t trust you because of it, and it does not matter whether your bad-faith wrongness was the result of malicious untruth or depraved indifference to truth”. We have such a term, in the common use of “lie”. I do not know of a superior alternative.

          • Matt M says:

            Fully agree with John Schilling.

            “Said something incorrect AND said it in bad faith” is the exactly right amount of specificity that we want the term “lie” to include. It allows us to distinguish between this bad behavior, and less bad behaviors such as “Said something incorrect but in good faith” (being mistaken) and “said something correct in bad faith” (being a jerk).

            But we don’t want it to be any more specific than that, because that allows liars to then nitpick over whether they technically qualify as liars or not.

          • Doesntliketocomment says:

            I would like to also chime in agreeing wholeheartedly with what John Schilling has said, that someone who is consistently wrong in a way that implies bad faith should be considered a liar. The importance of this definition is that it correctly places the impetus on the person who has made untrue statements to correct themselves, rather than on an outsider to determine their state of mind. If Bob is often incorrect and it hurts no one and gains him nothing, then we don’t need to label him in an accusatory fashion, but if his “mistakes” go hand in hand with a pattern of harm to others or self-gain, then we label him so that he must make more of an effort to be correct.

            If Bob absolutely cannot help but tell falsehoods due to some condition, then generally as a society will let up on the punitive morality aspect, but we take responsibilities away from that person to mitigate the damage they cause (frequently by putting them in a nursing home or institution)

      • Randy M says:

        Of course accusations should be held to a standard of evidence

        To do otherwise would be misleading, and you don’t want to be a liar yourself, do you?
        /not sure just quite how ironic

      • soreff says:

        >I also want to add that there’s a difference between concluding for yourself that a person is lying and actually accusing them of it.

        Agreed. I find a useful way to phrase an intermediate degree of certainty is:
        “He might be telling the truth, but that isn’t the way I would bet.”

        • Nick says:

          Yeah, I think there’s a lot of ways and reasons a person can be wrong, and it’s useful (to both parties even) to distinguish them. And incidentally, if I were distinguishing more I wouldn’t even necessarily use degrees of certainty. If I’d written a longer response to John, I’d have used an example like this:

          “The Parthenon is in Nashville, not Athens. I’ve visited it,” says Alice confidently.
          “No, there’s one in both places,” Bob replies.
          “What?!” she asks.
          “Yeah, the Nashville one is a recreation of the original in Athens. Why were you so confident there wasn’t one in Athens? Do you seriously believe Greek classical architecture originated in Tennessee?”
          “Well, no….” Alice says, embarrassed.
          “How many Scots-Irish settlers worshiped Athena Parthenos, by your estimate?”
          “I don’t know! I’m sorry! I’ll be more careful next time!” she says.

          Alice has every reason to be ashamed for her brash confidence here, but this has nothing to do with deliberately telling falsehoods. She made a false inference somewhere and attached too much confidence to it. But this is totally obscured if Bob were to accuse her of lying instead, and what’s more, Bob can teach her a lesson, pretty forcefully even, without ever using that accusation. I see no good use in extending the meaning of lying here.

    • Phigment says:

      I did some extended contemplation of lying a while back, for moral clarity reasons, and came to the conclusion that the important and objectionable thing about lying is not the literal accuracy, but the intent to deceive.

      That is to say, a liar is a person who tells you things designed to make you believe something false, not a person who tells you things which happen to be factually incorrect.

      This neatly resolved most things for me personally, in that it nicely encompasses the case of someone misleading people by telling them a carefully curated set of literally correct facts.

      It allows you to distinguish Bill Clinton splitting hairs over the definition of the word “is” from Barack Obama talking about 57 states; one of them is deliberately playing games, and the other is mis-speaking.

      It also allows you distinguish social lubrication from lying without sounding like an evil robot. When you go to a party and tell the host or hostess their house looks great, even though it’s objectively pretty average, that’s a compliment, not an attempt to deceive.

      From there, you can take “Do not lie” as a moral position to mean “I will not deceive people”, which is both easier and more useful as a code of behavior than “I will not say things that are factually false”.

      Unfortunately, as a standard, it’s even more subjective than not saying things that are factually false. Which is already the dominant standard; people in court swear to tell the truth, the whole truth, and nothing but the truth, but the usual expectation is they will tell the factually true, and the exact minimum of that which they can reasonably manage.

      • deciusbrutus says:

        Is a published study that doesn’t replicate a lie?

        • Not if the authors believed that what they wrote was true.

        • peterispaikens says:

          It depends on a lot of factors and intent, but clearly not replicating is not even sufficient grounds for even considering the results of that study being untrue (much less a lie), there are cases when studies don’t replicate because it later turns out that the results are highly dependent on some previously unconsidered factor.

          A study is a lie if either the data or the conclusions are fabricated and manipulated; A study is untrue (but not a lie) in case of experimental errors or simple statistic variation – I mean, p=~.05 is an explicit assertion that you believe that there’s a 5% chance that the observation you documented was caused by randomness and you believe that the study has (at least!) this chance of the conclusions being untrue; but a study might also fail to replicate without being either of these things – failure to replicate is a suggestion that it’s quite likely that it may be untrue, but not a certainty.

          • I mean, p=~.05 is an explicit assertion that you believe that there’s a 5% chance that the observation you documented was caused by randomness and you believe that the study has (at least!) this chance of the conclusions being untrue;

            (boldface mine)

            No.

            You are confusing the probability of the experimental result conditional on the theory being false (in a certain way–the null hypothesis) with the probability of the theory being false conditional on the experimental result.

            Consider the case where your experiment provides weak evidence for a theory which you already have good reason to believe is true.

        • Randy M says:

          But if the reporting on it is aware of a replication crisis, and reports the findings as fact–as in, “Lettuce reduces lifespan 10%!”, then even if those are valid representations of that study, I think that report is dishonest.

          That makes me wonder if merely stating the provisos in fine print is sufficiently honest if you suspect the audience will only read the headlines.

        • Phigment says:

          A published study that doesn’t replicate is not a lie, as long as the people who performed and published the study aren’t trying to mislead you with the results. It’s just a bad study. Lots of data collection is bad; science works that way.

          Intentionally using a study that you know doesn’t replicate as evidence to convince people of something is a lie.

          “Here are the results Kinsey reported when he studied human sexuality” is not a lie.

          “We should pursue policy X, because Kinsey’s study on human sexuality proves it’s better” is a lie.

          Relating to Randy M, I think putting the provisos in fine print is only honest if you actually expect that a person who read the fine print would still agree with the main article. If I wrote an article about how black rhinos are actually not endangered animals, there are tens of thousands of them in North America, having a footnote that reads “in this article, the term rhino is used to refer to labrador retrievers” is lying.

        • Mark Atwood says:

          The published study in the parable of [ https://www.explainxkcd.com/wiki/index.php/882:_Significant ] is a lie, because the researchers and their sponsors both all *knew* that the foundation assumption of the p calculation had been violated.

          My prof in my *undergrad* “Intro to Stats” course used the same parable, (with colored m&ms and indigestion across demographic slices), over 20 years before that particular XKCD was published. If that parable is taught at that level, then post-doc and lead researchers have no excuse, and I am quite comfortable with calling p-hackers “liers”.

      • John Schilling says:

        I did some extended contemplation of lying a while back, for moral clarity reasons, and came to the conclusion that the important and objectionable thing about lying is not the literal accuracy, but the intent to deceive.

        I would go half a step further, intent to persuade with indifference to truth. The tobacco-industry lobbyist who refrains from reading the research on lung cancer causality for fear of impairing his effectiveness in persuading people that tobacco is safe, ought to be considered a liar even though he’d honestly prefer that his persuasion be truly non-deceptive.

        If you believe something is true based on a contextually reasonable effort to determine the truth, and you communicate in a manner you reasonably believe will convey that truth, you aren’t lying even if it turns out that you are wrong.

        • Phigment says:

          I agree with that.

          Trying to persuade people to believe something false is lying, and you shouldn’t do it.

          And cultivating ignorance of what you try to persuade people about is not a loophole, any more than intentionally failing to read speed limit signs is a defense against traffic laws.

          • Trying to persuade people to believe something false is lying, and you shouldn’t do it.

            You shouldn’t do it (with rare exceptions), but it is only lying if you do it by saying things that are not true and that you know are not true.

            It’s worth making that distinction, because “false statement” provides something much closer to a bright line than “misleading statement.”

    • Scott Alexander says:

      But “liar” is a more difficult case, because anyone who remotely deserves it will almost certainly, well, lie about it. If you insist on a narrow, literalist definition where a person can only be called a liar if they can be proven to know that their statement is false, then “liar” becomes useless in the opposite direction – you can’t actually apply it to any person outside of a highly abstracted thought experiment, because they can claim to believe what they are saying and you aren’t a telepath.

      Not sure I understand. How is this different from eg “murderer”? I can always claim I didn’t murder someone, and you have to use evidence to assess my argument.

      Likewise, I can always claim I’m telling the truth – but if you catch me with cookie crumbs all over my face, then you can probably use evidence to assess that my “I didn’t eat the cookie” claim was a lie.

      A politician saying “my budget plan will make us all rich” is a more complicated case, since it’s hard to tell whether he’s misrepresenting the truth as he understands it or just too dumb to arrive at it. I still think it’s fine to accuse him of lying, as long as we understand that you’re positing that he’s misrepresenting the truth as he understands it. This seems like the same way we handle accusations like “Bob doesn’t care about the poor” – you don’t have to be a telepath, you just observe behavior and draw a conclusion.

      I think if a politician said “my budget plan will make us all rich” about an inexcusably bad plan, I would say something like “He’s either lying or an idiot”. Or if he’s shown other signs of not being that stupid, I would say “He’s lying” and feel okay about this as a hypothesis.

      • Dan L says:

        Not sure I understand. How is this different from eg “murderer”?

        The difference is that if it turns out someone lied about being a liar, we don’t throw them in a cell for a few decades.

        A difference in magnitude is a difference in kind. You’re right that similar evidentiary approaches can be used for “liar” v. “murderer”, but the latter is vastly less common and vastly more serious. That means noise is a bigger factor in relative terms, and there’s no mechanism (in common use) for forcing an authoritative decision. “We” don’t draw a conclusion.

        • deciusbrutus says:

          If they ONLY lied about being a liar, they either didn’t lie or generated a paradox where they lied IFF they didn’t lie.

        • Peter Gerdes says:

          Yes, they are very different in terms of CONSEQUENCES but that’s not the axis along which the distinction is being made. We are talking about distinctions in terms of truth-value and usefulness.

          Just because being a liar doesn’t get one thrown in prison doesn’t tell us it’s not useful to be able to say that you have such and such degree of confidence someone knowingly misrepresented the truth (rather than just fell victim to substantial bias).

          • Dan L says:

            You can’t separate usefulness from consequences. I’m not sure that’s even a coherent concept.

            As the outcome of a logical process becomes more important, the pressures on that process will change both in character and intensity. Evidentiary standards need to change accordingly.

            We could hold a trial to determine whether a false utterance lacked men’s rea. But we as a society are only going to do that in a few very rare cases. Barring that or a similar Schelling point, Scott’s argument is always going to be applicable in someone’s eyes – the principled objection dissolves back into a matter of opinion.

      • Peter Gerdes says:

        Exactly! A concept can be useful even if rarely apply it with 99% certainty, e.g., we might frequently want to evaluate if we have reason to believe it is 70% likely to apply.

        To explain why the concept is important even when we can’t conclusively prove it consider the following case.

        Suppose an acquaintance (Bob) calls me and says he’s in a foreign country with my friend who has just suffered some crazy weird accident and he needs me to wire him a bunch of money immediately to ensure urgent medical treatment/transit. Now if I have no reason to believe this acquaintance has ever been deliberately dishonest I won’t care if he’s convinced himself the singularity is imminent or the moon landings were faked because I’ll believe his first hand account of the situation is accurate even if I don’t trust his judgement.

        On the other hand if he’s given a bunch of other friends implausible stories about needing money for various emergencies which have never been disproven but are highly suspicious I’ll probably want to investigate further before sending the money even if I only assign a 70% chance he was being dishonest.

        So it’s important that I can tell the difference when my friends tell me they think Bob’s probably a liar if they are telling me Bob convinces himself of crazy things or Bob knowingly tells falsehoods even if they can’t prove it.

      • Randy M says:

        John is saying that the definition of lie depends on intent, and while you can’t always plausibly claim to be telling facts, you can always claim to have been deceived yourself.
        I think you can make a lot of inferences about actual intent based on how someone reacts when presented with contrary facts–“Oh, wow, I said we’d all get rich and now we aren’t. We need to rethink this plan, obviously we had mistaken assumptions somewhere” shows a level of contrition and humility appropriate for inadvertently misleading; “Oh, you wanted to know if I ate these cookies in particular. No, yeah, I totally ate those, sure. I thought you meant something else” is rather implausible.

        • Nick says:

          I think you can make a lot of inferences about actual intent based on how someone reacts when presented with contrary facts–“Oh, wow, I said we’d all get rich and now we aren’t. We need to rethink this plan, obviously we had mistaken assumptions somewhere” shows a level of contrition and humility appropriate for inadvertently misleading; “Oh, you wanted to know if I ate these cookies in particular. No, yeah, I totally ate those, sure. I thought you meant something else” is rather implausible.

          And notice that the reactions aren’t very strong grounds for shouting, “Aha! You are lying! You dirty, rotten liar!” but they’re pretty good grounds for ending the conversation, or warning folks privately that that guy doesn’t seem very trustworthy. Or telling yourself, “Wow, this guy seems pretty shifty. I don’t want to be like him.” Or teaching your children what not to do. And about a hundred other things.

          • Randy M says:

            And notice that the reactions aren’t very strong grounds for shouting, “Aha! You are lying! You dirty, rotten liar!”

            If I had used a more serious example, it may be.
            “When you asked if I had sex with your girlfriend, I didn’t know you mean oral. Sure, yeah, under that definition I did. You should have been more specific.”
            Pretty justified in shouting “You dirty, rotten liar!”

            Even in the cookie crumbs example, while the stakes may not warrant shouting, playfully responding “Man, you’re such a liar. Tell me the truth next time.”

          • Nick says:

            Yeah, okay, I for one wouldn’t believe them, in either example.

      • John Schilling says:

        Not sure I understand. How is this different from eg “murderer”? I can always claim I didn’t murder someone, and you have to use evidence to assess my argument.

        You can also claim that you didn’t “murder” someone because you honestly believed and intended that shooting them would just hurt them real bad, and that unintentionally killing someone you sincerely just meant to hurt real bad is only manslaughter. And then we can’t convict anyone of murder unless we have reliable telepaths to testify in court, but the people who accurately accuse people of murder are liable to be accused of slander.

        Except, the actual legal definition of “murder” is that the person has to have actually meant to kill someone, or to have acted with depraved indifference to life, so checkmate.

        So what I’m proposing isn’t actually different from murder, as murder is defined by lawyers who have put real thought into closing off the obvious loopholes that people would otherwise use to weasel-word themselves out of legitimate murder convictions. I’m saying that we need to apply the same standard to liars, to have a “depraved indifference to truth” standard in parallel to the “intent to deceive” one.

        That way, we can pin accused liars down to measurable facts and community standards of reasonableness in fact-checking. Things that can be reasonably proven by evidence in courts, of law or public opinion, and not trivially evadable by an “I’m innocent in my own mind and you can’t prove otherwise” defense.

    • kybernetikos says:

      > If you insist on a narrow, literalist definition where a person can only be called a liar if they can be proven to know that their statement is false, then “liar” becomes useless in the opposite direction – you can’t actually apply it to any person outside of a highly abstracted thought experiment, because they can claim to believe what they are saying and you aren’t a telepath.

      Agreed. The argument given in the post about the usefulness of words actually counts against ‘liar’ meaning deliberate deception – the word is almost unusable if we aren’t allowed to apply it without knowing the details of someone else’s mental state.

      In fact, in common use, ‘liar’ frequently seems to mean someone who tells an untruth, and who knows *or reasonably should know* that it’s an untruth.

      This slight expansion in meaning lets us use the word without always having to worry whether someone is engaging in double think to an extent that means they can’t even lie.

      • Nick says:

        the word is almost unusable if we aren’t allowed to apply it without knowing the details of someone else’s mental state.

        In fact, in common use, ‘liar’ frequently seems to mean someone who tells an untruth, and who knows *or reasonably should know* that it’s an untruth.

        ಠ_ಠ

        • kybernetikos says:

          I think you’re trying to indicate that knowing whether someone ‘reasonably should know’ that something is an untruth is equivalent to knowing the details of their mental state.

          But that’s how I’m saying I believe the official definition is typically extended – I’m using ‘reasonably should’ in the ‘man on the Clapham omnibus’ sense. It’s a way of avoiding the specificity of a particular person who may have all kinds of mental weirdnesses and instead using as a measuring stick a hypothetical ordinary, reasonable person.

          • Nick says:

            I think you’re trying to indicate that knowing whether someone ‘reasonably should know’ that something is an untruth is equivalent to knowing the details of their mental state.

            Not quite. My point was only that you said the definition of liar is useless because it requires knowing mental states, then contrasted it with a “common use” description which… includes requiring knowing mental states. But I didn’t bold the “*or reasonably should know*” part because it doesn’t actually apply to that—I agree you can stipulate what a person reasonably should know without reading their mind. It remains a bizarre definition as written, because the whole weight actually falls on your man-on-the-omnibus part. The bolded part contributes nothing.

            It may seem like a nitpick, but it matters. You’ve located your evidence that he’s a liar entirely outside of him. It’s a bit like saying, “Well, witnesses and fingerprints don’t really prove anything, so I can never know this guy’s a murderer… let me ask then whether the man on the Clapham omnibus would have murdered her in similar circumstances….” Maybe there are circumstances where it’s reasonable to think that way—you’re in some kind of Agatha Christie novel and the millionaire on the train has been murdered, and any ordinary person in this car could have done it!—but it’s far from what we most of the time mean by murderer. And I think by analogy it makes a mess of our notion of liar.

          • kybernetikos says:

            Sounds like I need to clarify.

            When I said “someone who tells an untruth, and who knows.. that it’s an untruth” I was describing what I took to be the standard meaning, and when I added, “*or reasonably should know*” I was describing what I took to be a typical, unspoken modification of the standard meaning precisely because the standard meaning is problematic for the reasons I stated.

            As far as I can tell, Nick’s look of disapproval is because he spotted the exact problem I was trying to address – the part of the definition that is ‘standard’ contributes nothing.

            As to whether it’s a bad thing to locate evidence for someone being a liar to things external to them, well, it’s necessary, since you can’t know what’s internal to someone else (even if they confess, they may be mistaken). It’s more that for maximum usefulness, you want to be able to call someone a liar based on observations rather than psychic ability.

      • Peter Gerdes says:

        Except the word ‘liar’ has never meant “someone who can be proven to have knowingly uttered a false statement.” We describe that via the phrase “proven liar.” Liar means someone who knowingly uttered a false statement (of substance…usually we exclude ‘white lies).

        Thus, we can meaningfully assert things like “Obama is a liar” to convey our belief that he knew it was false when he said “If you like your health care plan you can keep it” even if we don’t actually have objective proof it is true (imo yes he was…but it was a justified lie as may politicians tell are).

        It’s important to be able to understand what these claims mean so we can assign degrees of confidence in them even if those degrees of confidence might only rarely be so strong they rise to the level of ‘proof’.

        A word isn’t useful only to the extent we can prove the state of affairs it described really obtains. If that was true we wouldn’t have different words for criminal and convict.

        • Thomas Jorgensen says:

          … That particular case, the truth was pretty unutterable. I mean, what was he supposed to say:
          “If you like your current health insurance, I am very sorry but I do not believe in enabling the continued victimization of people suffering from stockholm syndrome”?

          • What he was supposed to say was:

            Even if you like your present health insurance it may no longer be available.

            That would have been an honest description of one consequence of what he was proposing.

            But he could have avoided lying by making no claim on the subject at all.

            Your comment leaves me wondering if you actually don’t realize that Obamacare, working exactly as it was supposed to work, could be expected to make some people worse off.

          • Matt M says:

            I mean, what was he supposed to say:

            To David’s point, he could have dodged the question and said nothing at all.

            Or, barring that, he could have significantly hedged and said something much more vague and nebulous such as “Based on our economic analysis, we don’t expect that this will have a significant impact for most people who want to stay on their current insurance plans.”

            Of course, since that clearly and obvious sounds vague and nebulous, the GOP would have pounced on it immediately, so he needed to say something much more clear and reassuring, confident that the media would side with him over anyone who dared suggest he might be uttering a false statement.

          • Randy M says:

            I mean, what was he supposed to say:

            “We’re not sure what will happen exactly, but clearly there will be winners and losers, and we are going to do our best to mitigate the harms and make sure the benefits accrue to those that need them most.”

            That would be an honest formulation of a charitable interpretation of the plans goals.

    • If you insist on a narrow, literalist definition where a person can only be called a liar if they can be proven to know that their statement is false

      We don’t usually expect people to limit their statements to ones they can *prove* are true.

      You can honestly call someone a liar if you believe he knew his statement was false. He can deny it. Third parties can decide for themselves whom to believe.

    • soreff says:

      >depraved indifference to the truth.

      That phrase sounds like it would make a great motto for a public relations department 🙂

  12. Randy M says:

    I got into an argument recently with somebody who used the word “lie” to refer to a person honestly reporting their unconsciously biased beliefs – her example was a tech entrepreneur so caught up in an atmosphere of hype that he makes absurdly optimistic predictions. I promised a post explaining why I don’t like that use of “lie”.

    “overhype because deluding oneself”:lie::negligent homicide:murder
    In some contexts it might be fine to elide the distinction, but it a non-central example at best.

  13. caryatis says:

    Let’s toss the whole concept of “emotional abuse.” Not sure if I’m echoing what Scott’s saying or going a step beyond, but the same arguments apply.

    Who among us can claim never to have been mean to a partner? Never raised your voice or criticized their character or tossed off a “fuck you”? Never questioned their competence or used a condescending tone or decided not to talk to them (because the silent treatment is “abuse”)? Cheating is definitely abuse and gaslighting, from what I’ve heard. Pressuring someone for sex is abuse, as is denying them sex.

    So everyone’s an abuser, and abuse is normal. One possible response is that doing these things rarely is normal, but doing them routinely or to a severe degree is abuse. But in that case the definition of abuse is so subtle, subjective and hard to pin down as to be completely unverifiable.

    The term “abuse” should be reserved for concrete criminal acts: physical and sexual abuse.

    When i say this, people often respond by stating that being treated badly by a partner can be very painful and even cause lasting psychological harm. I don’t question that, and I don’t advocate this sort of behavior. If you find yourself treating your partner badly routinely, you should look for ways to improve your circumstances, your reaction to them, or perhaps end the relationship. But the question is not whether this behavior is harmful, but rather whether it is productive to lump it in with actual criminal abuse.

    • Scott Alexander says:

      I feel like this has sunk so far in to the consciousness that un-defining it as abuse would leave people with the wrong connotations now.

      • caryatis says:

        What kind of wrong connotations?

        • Peter Gerdes says:

          I’d guess that by failing to label someone as an abuser you’re effectively saying they aren’t that bad. Stopping using the term would, at least unconsciously, cause people not to feel the same level of condemnation of abusers.

    • Ozy Frantz says:

      I think it is pretty normal throughout life to treat a consistent pattern of behavior differently from an isolated incident. Bob is unreliable if he comes in thirty minutes to an hour late every day; Bob is not unreliable if he comes in five minutes late once. Bob is depressed if he is sad all the time and constantly thinks about how he has failed at everything he is tried; Bob is not depressed if he is sad some evenings and occasionally is concerned that he has made a mistake. It is rainy in your city if it rains every day, but not if it rains one day a year.

      Of course, this means there will be judgment calls– is Bob an unusually sad non-depressed person or mildly depressed? Does this city count as ‘rainy’ or not?– but it would be very very silly to say “unreliable people don’t exist because everyone is late sometimes” or “depression doesn’t exist because everyone is sad sometimes” or “rainy cities don’t exist because it rains everywhere sometimes.”

      For that matter, I slapped my husband when I was in physical pain so bad that I was unable to think and repeatedly expressed the desire to die because it was preferable to die than to spend ten more minutes waiting for the painkillers– and I’m not terribly sure I was wrong in my conclusion. Your argument implies that either I am an abuser whom no one should date, or that physical abuse is not a thing and we should get rid of the concept.

      • Peter Gerdes says:

        I agree that there is definitely a moral difference between single instances and repeated behavior but even though it might not track the moral truth I think there is a lot of value in norms which draw a sharp line at a kind of conduct. To be clear I’m ultimately inclined to agree with you on this point but I think there is something to be said for the other side that’s worth saying.

        For instance, I think (as you suggest) morally there is a huge difference between the guy who gets angry once over the course of a marriage and slaps his wife during some deeply traumatic and unusual situation after an extreme provocation and the man who does so regularly. However, human nature being what it is if we allow those considerations into the rule for social condemnation there is a real risk every abuser will hide under that exception and every abused partner will tell themselves ‘ohh no, he’s not really an abuser…he’s just under exceptional pressure’.

        To make a legal analogy sometimes it’s best to have strict liability rules even if that sometimes means that we punish people who are in some sense less blameworthy because they cause the best outcome.

        The particular case with abuse is hard because we don’t have a seperate word for just the physical/sexual abuse. Ideally we’d have one word just for the physical/sexual abuse letting us create a bright line there plus a more flexible term for general practice of something we might call ’emotional abuse’.

        I’d also add that one of the things that makes `emotional abuse’ especially hard is that the very same behavior can be abusive or not depending on the psychology, expectations and responses of the other party. The very same kind of angry screaming, shouting and even insulting remarks (‘how could you be so stupid as to X’) could easily be abusive in one relationship where the partner is deeply affected by them and nothing more than an expression (perhaps unhealthy) of annoyance in another where both parties scream at each other and see at is little more than an expression of annoyance/criticism.

        • Clutzy says:

          Your point on emotional abuse is exactly why there isn’t an “eggshell skull” rule in intentional infliction of emotional distress, and other related torts. When it comes to emotional abuse, the First Amendment, etc we don’t tolerate “eggshell brain” arguments, because its 1) Impossible to prove in most cases; and 2) Subject to abuse consistently.

      • Doctor Mist says:

        Does this mean that only liars lie? I’m not sure I’m okay with that.

      • caryatis says:

        Yes, it’s true that repeated behavior is treated differently from isolated incidents in many cases.

        I still think, though, that drawing a bright line about physical abuse or sexual abuse is a lot easier than with emotional abuse. For instance, is criticizing your partner “emotional abuse”? Very hard to say–you have to bring about criticisms sometimes, but if it’s very severe and never-ending criticism, it could be traumatic. What about the silent treatment, or refusing sex, or asking for sex? All things that normal people do innocently that can also be “emotional abuse,” according to the kind of people that believe in emotional abuse.

        But is slapping your partner physical abuse? Yes, of course it is. That is not controversial. Don’t hit innocent people just because you’re in pain. I don’t endorse your idea that one act of physical abuse means no one should ever date you—that’s black-and-white thinking which doesn’t follow from what I said—but yes, you did commit abuse and you should try not to do that.

    • Radu Floricica says:

      I don’t know. I have a married friend who’s relationship with his wife deteriorated after the first kid. I’m not talking about abuse but I can certainly, 100% testify that the overall change in the quality of the relationship is a concrete thing. It’s the case where a quantitative change (large amount of small put-downs) adds up to a qualitative change. So I can definitely extrapolate from here and imagine couples where the added non-physical exchanges would add up to a result that’s objectively worse than some amount of physical abuse. In which case it makes sense to use the term “psychological abuse” and treat it as a Bad Thing.

      • AG says:

        “Hostile environment” has been held up in court, so the case of a large amount of small incidents amounting to a legal wrong is a thing.

    • Purplehermann says:

      Abuse as ‘action or pattern of actions causing damage, (usually) in context of a prior relationship which makes these actions particularly damaging’ (not definitive, just “pointing” to what i think is normal usage) seems useful to me, and including emotional abuse as part of that seems reasonable-same concept, the only difference between sexual abuse and emotional abuse being the medium- as well as useful, it makes emotional abuse more concrete.
      Deciding now that emotional abuse isn’t “really” abuse would probably cause people to take it less seriously.
      Overall changing this seems like a net minus to me

  14. Randy M says:

    Broadening the definition of “abuser” this far doesn’t help fight abuse or make anybody nicer. It just removes a useful word from the English language. I can still eventually warn someone that John is cruel or violent toward people close to him. I just have to circumlocute around the word “abuser”, in order to find some other word or phrase that hasn’t been rendered meaningless.

    Given that abuse already seems to encircle quite dissimilar actions, it may be better to let it die and use a more precise term or phrase.

    Except unfortunately violent already often means something other than infliction of physical force, and cruel suffers the same weakness as lying, that unintentional pain inflicted can be framed as cruelty.

    Maybe the only solution is to talk like a robot.

    • eric23 says:

      The word abuse is ripe for imprecision.

      Just look at its etymology: “ab-” (away, i.e. wrongly), “use” (use). We set up a standard of what is appropriate behavior, and “abuse” is by definition any behavior that violates that standard. Of course, the standard can change at any time, or from person to person, in which case the meaning of the word “abuse” changes accordingly.

      Any suggestions for a word or words which more precisely capture what exactly “abusers” in relationships are doing wrong? Possibly something to do with manipulation, control, or habitual cruelty?

      • Randy M says:

        Any suggestions for a word or words which more precisely capture what exactly “abusers” in relationships are doing wrong? Possibly something to do with manipulation, control, or habitual cruelty?

        My suggestion is to specify what they are doing. We can retain the word, even, but any conversation important enough to include the loaded word abuse is worth lengthening until you include the relevant actions. “Stay away from him, he is an abuser–he constantly insults people and I once saw him kick a puppy.”

        If you need to discuss the group “Abusers make poor partners”, add the relevant referents. “Abusers–that is, people who inflict physical pain or attempt to influence others via insults or threats make poor partners.”

        Yes, it’s longer; there’s a reason a lie can get halfway around the world before the truth gets its pants on.

  15. Nick says:

    Strong endorse.

    There’s a whole class of terms with recent redefinitions like this that especially irritate me. Lying is one of them, and so is bad faith. Scamming, as in the LW post, also falls under it. What unites lying, scamming, arguing in bad faith, etc., is that they’re deliberately done, they’re intended, they’re meant. “Murder” is the same way, and anyone who defines murder the same as “killing” is going to become very confused very quickly. Redefining lying to mean “honest reporting of unconsciously biased beliefs” and then relying on the longstanding condemnation of deliberately telling falsehoods is a pernicious form of equivocation because these are two things we really, really need to be able to distinguish.

  16. Furslid says:

    This is especially worrying when there is a push for zero tolerance. If the only option is to punish the behavior maximally, no one is able to say “Yes, that was abusive. I’m sorry. Let me make it right with some minor penance.”

    If the only sins are truly horrible things, then sinners can be cast out into the outer darkness.

    If impure thoughts are sins, the church needs to let some sinners admit guilt, say a few rote prayers, and get on with their lives.

  17. belvarine says:

    But if everyone used my friend’s definition, and we acknowledged that everybody is an abuser – the category stops being informative.

    I don’t think it’s less informative. It just makes it more difficult to analyze and address.

    Expanding the definition of things like abuse expands the scope and complexity of the problem. Upending society and instituting the Non Aggression Principle is a lot harder than voting for a DA who promises to be tough on domestic crime. Maybe after a certain point, from your perspective, the problem becomes too large for you to solve, and your focus shifts to smaller problems within your control. That doesn’t mean the newly introduced complexities aren’t worth addressing, it only means that you lack the willingness or resources to address them.

    It’s not that the terms have lost their usefulness, it’s that you no longer have use for them because you don’t think they’re actionable. I think you touched on this:

    But if both sides are constantly accusing each other of lying just for having normal human failings, then “you are a liar” no longer carries much weight. I have to come up with some complex circumlocution in order to let people know that someone told a mistruth.

    It seems like your satisfaction with the use of “liar” here is connected to how much labor you have to perform.

    See also:

    “John is an abuser”. So what? Doesn’t mean you should worry about John, doesn’t mean you shouldn’t date John, doesn’t even mean you shouldn’t set your single friends up on blind dates with John.

    You seem to be most concerned here with whether you’d have to take on additional load: worrying, being more selective when dating, assuming liability when screening blind dates.

    I do think this is a very relevant discussion though. Category membership is how we deny a lot of people access to food, shelter, medicine, care, etc.

    • Scott Alexander says:

      “It seems like your satisfaction with the use of ‘liar’ here is connected to how much labor you have to perform.”

      I mean, we’re talking about language here. The goal of language is communication. If we come up with a way to do language that makes true things very hard to communicate and act upon, then this is a bad way to use language.

      “You seem to be most concerned here with whether you’d have to take on additional load: worrying, being more selective when dating, assuming liability when screening blind dates.”

      I’m concerned about whether I (or anyone else) would correctly understand and act upon information. If I see Bob hit his wife, and I try to warn you by saying “Bob is an abuser”, and you misunderstand this warning and think that I just mean Bob is an average guy, and so you date Bob and get abused, then our inability to communicate properly has made the world worse off.

      • Clutzy says:

        I’m concerned about whether I (or anyone else) would correctly understand and act upon information. If I see Bob hit his wife, and I try to warn you by saying “Bob is an abuser”, and you misunderstand this warning and think that I just mean Bob is an average guy, and so you date Bob and get abused, then our inability to communicate properly has made the world worse off.

        I think this is the most important thing. By expanding the definitions (often without most of the population knowing) you are deleting information.

        Its like if there is a food truck and the menu lists Shrimp Tacos, Mushroom Soup, and a Club Sandwich. You order the club, and get a taco stuffed with ham, prosciutto, spinach, pineapple, and guacamole. That is not a normal club. It has similar elements:

        Bread> Taco
        Turkey> Ham
        Bacon>Prosciutto
        Lettuce>Spinach
        Tomato>Pineapple
        Mayo>Guacamole

        But its not a club. If you want it to be called “my club sandwich” you have to list all those new ingredients on your tiny chalkboard/whiteboard and all we’ve done is massively increase transaction costs.

        That is mostly what definition creep does, deletes info and increases transaction costs.

  18. Joseph Greenwood says:

    I posted this comment before, but wordpress appears to have eaten it.

    One important consideration in choosing where to draw the boundaries of membership for your words (we’ll use “murderous” as a toy example) is whether any other words draw their boundaries in similar places. I think, for instance, that it is genuinely useful to have a word for “at least as murderous as 50% of the population”–but we already have a very similar word! We call people like that “mean”. So if you were to redefine “murderous” to mean something similar, you would not be giving up the ability to “transmit[] a lot of information in a small number of cases” in order to “transmit[] a lot of information in a small number of cases”, you would be giving up the expressive power of the word “murderous” for nothing at all: the other use that you wanted to put the word to was already being achieved with other words.

    • Joseph Greenwood says:

      I first saw this point raised in the foreword to Mere Christianity, where CS Lewis argues against redefining “Christian” to refer to “virtuous people to whom God will grant salvation”:

      People ask: “Who are you, to lay down who is, and who is not a Christian?”: or “May not many a man who cannot believe these doctrines be far more truly a Christian, far closer to the spirit of Christ, than some who do?” Now this objection is in one sense very right, very charitable, very spiritual, very sensitive. It has every available quality except that of being useful. We simply cannot, without disaster, use language as these objectors want us to use it. I will try to make this clear by the history of another, and very much less important, word.

      The word gentleman originally meant something recognisable; one who had a coat of arms and some landed property. When you called someone “a gentleman” you were not paying him a compliment, but merely stating a fact. If you said he was not “a gentleman” you were not insulting him, but giving information. There was no contradiction in saying that John was a liar and a gentleman; any more than there now is in saying that James is a fool and an M.A. But then there came people who said – so rightly, charitably, spiritually, sensitively, so anything but usefully – “Ah but surely the important thing about a gentleman is not the coat of arms and the land, but the behaviour? Surely he is the true gentleman who behaves as a gentleman should? Surely in that sense Edward is far more truly a gentleman than John?” They meant well. To be honourable and courteous and brave is of course a far better thing than to have a coat of arms. But it is not the same thing. Worse still, it is not a thing everyone will agree about. To call a man “a gentleman” in this new, refined sense, becomes, in fact, not a way of giving information about him, but a way of praising him: to deny that he is “a gentleman” becomes simply a way of insulting him. When a word ceases to be a term of description and becomes merely a term of praise, it no longer tells you facts about the object: it only tells you about the speaker’s attitude to that object. (A ‘nice’ meal only means a meal the speaker likes.) A gentleman, once it has been spiritualised and refined out of its old coarse, objective sense, means hardly more than a man whom the speaker likes. As a result, gentleman is now a useless word. We had lots of terms of approval already, so it was not needed for that use; on the other hand if anyone (say, in a historical work) wants to use it in its old sense, he cannot do so without explanations. It has been spoiled for that purpose.

      Now if once we allow people to start spiritualising and refining, or as they might say ‘deepening’, the sense of the word Christian, it too will speedily become a useless word. In the first place, Christians themselves will never be able to apply it to anyone. It is not for us to say who, in the deepest sense, is or is not close to the spirit of Christ. We do not see into men’s hearts. We cannot judge, and are indeed forbidden to judge. It would be wicked arrogance for us to say that any man is, or is not, a Christian in this refined sense. And obviously a word which we can never apply is not going to he a very useful word. As for the unbelievers, they will no doubt cheerfully use the word in the refined sense. It will become in their mouths simply a term of praise. In calling anyone a Christian they will mean that they think him a good man. But that way of using the word will be no enrichment of the language, for we already have the word good. Meanwhile, the word Christian will have been spoiled for any really useful purpose it might have served.

  19. jonabar says:

    I think you’re underselling this idea by saying it’s a marginal issue. I think it highlights a really crucial and maybe productive tension in progressive social justice-y thought.

    On the one hand, a lot of progress was made in terms of racism/sexism/homophobia in the 20th century by singling out particularly egregious perpetrators and labeling them racist/sexist/homophobic. Those words acquired a moral force that made people want to really avoid the sort of blatant prejudice that would induce being called one of them.

    On the other hand, a central tenet of contemporary progressivism is the notion of structural violence, which underpins ideas like institutional racism, the patriarchy, etc. In this model, everyone who participates in a given social structure supports that structure, so it makes sense to say, for example, that “all police officers are racist” by virtue of all them upholding an institution that perpetuates racial disparities. This model also tries to have that statement not mean that “no police officers are racist.”

    Personally I think structural violence is a very useful way of understanding social problems, but practically it’s so hard to apply that model rigorously in day-to-day interactions, many of which devolve not into the nihilism of “everyone is X so therefore no one is X”, but rather into the more pernicious “60% of people decided that today X will be defined as what the other 40% of people do”.

  20. aashiq says:

    While we can try to police things like lie inflation, but the core problem is that words aren’t great for conveying meaning, because of their intrinsic vagueness and political value. When I want to be clear, I will include additional detail. Ie, Albert is at the 5th percentile in credibility. Better yet is to phrase descriptions as predictions of objective phenomena: prefer Tom will sleep in to Tom is lazy. Language is slippery, I’m often in awe that it works as well it does.

    • Scott Alexander says:

      Have you ever really said “Albert is at the 5th percentile in credibility”?

      • aashiq says:

        I would more likely say Albert is a 95th percentile liar, and yes I often say things like this, though I run in a weird crowd. Similarly, when I do statistics I prefer an effect size and confidence interval to p-values, and an out of sample prediction error to an r-squared. Reality is too complex to be summarized with only categorical variables (words).

      • Randy M says:

        A more human like circumlocution would be, for example, “You can trust Albert with money, but he won’t usually arrive when he says he will.” or “Albert has never let me down yet.” or “Albert says he’ll help us, but frankly I wouldn’t rely on that.”
        It’s rarely both useful and possible to actually rank people by credibility (although some people make a living doing it!) but you can specify the domains or frequency of the past deception/disappointment.

  21. Zamiel says:

    Scott,

    This post strongly reminded me of The Coddling of the American Mind by Greg Lukianoff and Jonathan Haidt’s (the 2018 book, not the 2015 Atlantic article), which I believe you have probably read, but didn’t mention it in the blog. Just in case, one of their main points in Chapter 4 (Intimidation and Violence) is very similar to what you lay out in this blog. Essentially, they notice that some parts of American society seem to be expanding the word “violence” to include speech, and they point out how expanding the definition of the term is hazardous. I guess the main difference is that you argue that expanding the definition of “lie” decreases its utility and has bad implications for civic society / discourse, and Haidt argues that expanding the definition of “violence” has bad implications for child fragility and their character/psychology. But perhaps the meat of the arguments could be applied to either case.

    PS: Love the blog.

    • Scott Alexander says:

      I haven’t read the book. I’m pretty skeptical of “this will make children fragile” type arguments generally.

      • Doctor Mist says:

        Because you don’t think children have been made more fragile? Or because you usually don’t agree about the alleged cause?

      • eightieshair says:

        I haven’t read the whole book, but based on excerpts and interviews with the authors, the point seems to be less “this will make the children fragile” and more “certain subcultures now make a practice of valorizing fragility and making ‘performing your fragility’ a desirable thing to do for the purposes of status and getting what you want”.

  22. Machine Interface says:

    Counterpoint: semantic broadening is a normal way words can and have changed since we have been able to record the evolution of language. Simultaneously, other words get narrower meaning. Sometimes a word for a part of the whole becomes the word for the whole, and reciprocally: late Latin “singularis porcus” (wild boar, literally “solitary pig” > French “sanglier” (wild boar); Latin “ponere” (to put) > French “pondre” (to lay an egg). Sometimes a word even completely reverses meaning through ironic usage: Old French “enerver” (to weaken, to leave without energy) > Modern French “énerver” (to anger, to upset, to make nervous).

    But semantic is a treadmill and language adapts; when a word no longer covers the semantic its users feel the need to convey, a new word is pushed in its place — or not, when it turns out the older semantic is no longer needed; “alcohol” no longer means “the product of distillation”, and no word has really taken on that meaning. Turns out that’s fine because alchemy is obsolute and having a single word for “the product of distillation” doesn’t serve much purpose in modern science.

    Edit: you and most of the commenters also seem to just assume that this is all done consciously and deliberatedly. That’s mostly not how language change works. No one controls semantics, it does what it wants.

    • Scott Alexander says:

      I’m saying this example is being done consciously because I’m referring to a specific argument with a specific person who said they were consciously doing this.

      Words have been broadening since the beginning of time. But since the beginning of time, people have been complaining about this and trying to stop it, like I am now. Why are you trying to mess with tradition? You’re literally killing me!

    • In addition to what Scott said, the arguments here for people to use words in certain way don’t really depend very much on whether people’s decision to use words a different way is deliberate or conscious. Rather, they can be read as plea for people to be more deliberate and conscious about their word choice, for those who are undeliberative and unconscious about these things.

      Also, just because something is natural, commonplace, and existed for a long time doesn’t mean that we shouldn’t oppose it. Lying is natural, commonplace, and existed for a long time. So is abuse. And the damage inflicted by lying and abuse usually fades away much faster than the time it takes language to correct itself. It might not be the most important thing in the world, but I believe that even though language change is natural it is good to try to deliberately direct it to improve the ease and clarity of communication.

    • anon1 says:

      > “alcohol” no longer means “the product of distillation”, and no word has really taken on that meaning. Turns out that’s fine because alchemy is obsolute and having a single word for “the product of distillation” doesn’t serve much purpose in modern science.

      The word you are looking for is distillate, and it’s often used in chemistry. Shorter than “the stuff that’s dripping out of the condenser”, and more accurate than [insert name of the chemical you hope you’re separating out]. Ex: “The expected product of the reaction had a low boiling point, so we tried distillation to separate it from the solvent. But the distillate’s NMR spectrum showed our synthesis had failed: it was almost all an unwanted side product.”

    • Randy M says:

      Language doesn’t have agency. It’s interesting consider the mechanism for how this happens. I doubt it’s random mutation and selection; more like it conveys a systematic benefit for the speaker to use the term that doesn’t quite fit but has more favorable (to them) connotations. It’s worth identifying and pushing back against this, even if it is futile long term.

  23. Sniffnoy says:

    Huh, I’m reminded of this recent post of Sarah’s about how where to place the line can vary based on context: https://srconstantin.wordpress.com/2019/02/27/the-tale-of-alice-almost-strategies-for-dealing-with-pretty-good-people/

    • Scott Alexander says:

      Thanks for linking that.

      I still don’t think these are exactly the same situations. If Alice is 90th percentile for kindness but you want people who are 99th percentile, we can debate whether to say “Alice is unkind” or “Alice is kind but doesn’t meet the high standards for kindness around here”, but I think it’s still flat-out wrong to say “Alice is a murderer”.

      In the same way, I think the tech entrepreneur in the first paragraph should be described as “not rational enough” or even as “biased”, but not as a liar.

      • AMT says:

        I think the best word to use in a situation like you described where you want to make it clear the person is telling a falsehood, but likely without the mens rea, is delusional. Or you can say deceptive. We have stronger words and weaker words with similar meanings, and it seems likely to me that they were using hyperbole by using “lie,” when other words can convey quite accurately what was actually meant. I agree that it is bad to try to expand the meaning of words. You don’t need to tell me someone is literally “lying,” when you can say “they have absolutely no idea what they are talking about,” “they are willfully ignorant,” “even stupider than [Politician of the opposite political party], and just after your money,” or whatever is appropriate in the situation.

        I recall a conversation with a female coworker who told me how a work conference went. She was having a good, fifteen minute conversation with a man she had met there, and then he asked if she would like to join him in his hotel room for a drink. It was perfectly clear that she (literally) couldn’t believe that a man would dare sexually harass her like that! After they were having such a pleasant conversation, too!

        I’m just glad that I’m now married and don’t have to risk committing the unconscionable crime of asking someone out, or after a date, ascertaining their interest. After hearing that, I would literally be scared to return to the modern dating world, unable to precisely define the expanding reaches of the term “sexually harass.”

        To concisely summarize: the problem is that the broader a term is, the less detail you receive from it. On twitter, and the internet generally (well, basically in any communication), details and nuances are often eschewed, sometimes maliciously.

        PS. I can understand how people might think his proposal was improper, overly forward etc., but remember, those are completely different terms!
        Yes, it also might be harassment in certain other situations like doctor-patient, employer-employee, or through repeated invitations, but those were clearly not the case here.

      • Tarpitz says:

        I in turn am unsure that your analogy is apt, even though I like you would not characterize the described behaviour as “lying”.

        Glossing “lie” as the high end of a scale of “culpable verbal or written deception” seems a lot more plausible than glossing “murderer” as the top end of a scale of “unkind person”. I do not believe such a scale is a particularly useful model of reality or language use in either case, but if it were, Jessicata’s scale would make a lot more sense than yours.

        I think there are two separate points on which you and she are disagreeing: the factual nature of the behaviour under discussion (which she believes to be much closer to deliberate deception than you do) and, separately, what constitutes a lie (which she believes to be less dependent on the beliefs of the utterer and more dependent on the state of the world than you do).

        The V-virtue scale discussion makes more sense in the context of skills or abilities than behaviours, and natural examples spring easily to mind there. Consider discussions of sport. I might say “I’m quite good at football” and “Tiémoué Bakayoko is terrible at football” without anyone who understands the context at all imagining that I think myself a better footballer than Bakayoko. Likewise an academic might call his dog clever and his colleague stupid, and be in no doubt that the latter was more intelligent than the former.

  24. Big Jay says:

    The rebranding of lying is basically a parasitic process, exploiting the trust we have in a functioning piece of language until it’s lost all meaning – after which the parasitism will have to move on to whatever other trusted functional piece of language has sprung up to replace it.

    Something similar has long been known as the “euphemism treadmill”. A simple word for an unpleasant concept (e.g. “moron”) gets euphemized into a more pleasant word (“special”). Ten years later the former euphemism is generally understood and no longer hides its meaning. A new euphemism is created (“exceptional student”).

    If you really want to see euphemism raised to an art form, your state’s Department of Prisons Corrections surely has a website.

    • Scott Alexander says:

      Do people really use “exceptional student” this way?

    • ana53294 says:

      I am more used to the traditional meaning of exceptional student, meaning a really bright and promising student. I guess it’s in the same way “genius” is sometimes used ironically. Although the usage of the word genius in both contexts does not seem to confuse people that much, because it’s pretty clear that when the word is used for a person in the bottom of the intelligence distribution, it is pretty easy to know the real meaning from context.

      So if abuse starts to include edge cases also, the badness of abuse becomes unclear. But if exceptional student is used for two clearly demarcated and distinct categories of people, we can guess the real meaning easily.

      So if a parent proudly tells you their kid is exceptional and gets lots of praise, you won’t even think about the second meaning, because it’s pretty clear.

  25. The most serious example of the pattern you describe, in my experience, is the inflation of “racist.” What it means to me, and what I think it meant more generally when I was growing up, was someone who hated or despised other people because of their race. Now it means, roughly speaking, someone who disagrees with the person using it on some issue having to do with race or someone who says something having to do with race that the person using the term may not disagree with but doesn’t think should be said.

    • len says:

      Other examples of this would be the inflating “murder” to include abortion, or “theft” to include taxation.

      On the other hand, gatekeeping aside I can’t think of many words that have their definition narrowed.

      • Peter Gerdes says:

        The abortion/murder thing isn’t an inflation. It’s a genuine factual disagreement about whether a fetus has the moral status which makes deliberate end of life wrong.

        It’s clear the pro-life crowd isn’t redefining the word. They just have different moral beliefs.

        I mean if, in the conversation Scott talked about, the individual in question said she believed that crying in a relationship fight was the moral equivalent of hitting her boyfriend then she wouldn’t be inflating the meaning of the term but, rather, conveying a mistaken moral belief.

        • AMT says:

          Agreed that it is not an inflation, but I wouldn’t say it’s even a factual disagreement, it is a disagreement over the proper definition of a human life. I think both sides agree on the fundamental “unjustified killing of a human is murder” moral belief, but they disagree about the classification of what is a human.

          I would agree that moral beliefs likely influence someone’s definition, but that makes it indirectly about morals. And honestly I’m not sure exactly what moral beliefs even differ between each side. I think both sides use practically identical morals, but end up with different conclusions. “And because of moral rule x, life begins at [conception/birth].” If a pro-lifer says “well the bible says a fetus is human” I would say that supports my argument, because there isn’t any moral logic needed to arrive at the definition, since it is already given, so it is purely a definitional disagreement. I also struggle to see what moral belief influences the pro-choice definition. “It would be immoral to consider a fetus a human because x?” Any other explanation I can think of seems tautological: “Humans have these criteria: [presence outside the womb, etc.]…therefore a human life only begins at birth.”

          Sorry if I am oversimplifying moral beliefs, but that’s how the reasoning seems to me.

          • Peter Gerdes says:

            It is a factual question presuming you are a moral realist which almost all pro-life and, indeed, almost all people implicitly are. So I think it’s reasonable to say it’s a factual question even though not an EMPIRICAL question.

            But yes, I was glossing over the realism/anti-realism distinction in morality since it’s not a factual question if you are an anti-realist.

          • ARabbiAndAFrog says:

            Unjustified killing of a human being is the definition of murder. Moral question is what kinds of killings are unjustified and what kinds of beings are human.

          • Tarpitz says:

            Presuming the participants in the disagreement are moral realists only gets you as far as said participants seeing it as a factual question. For it to actually be a factual question, moral realism would have to be correct.

        • len says:

          Definition inflation does seem to be mostly about moral beliefs though.

          I don’t see why Scott’s friend would choose define unintended emotional manipulation as “abuse” if she didn’t believe that it had at least some of the moral implications of abuse.

          Or why people who inflate definitions of lying if they’re not trying to make the point that doing X is just as bad as lying, so we might as well group it under lying.

          Even group definitions like “disabled” are about moral beliefs. Expanding the definition of “disabled” seems to be saying these people are just as deserving of sympathy and should be given the same moral weight as the disabled.

          I don’t see how separating definition inflation from morality is useful.

      • abystander says:

        Including abortion as murder might not be an inflation of the definition of murder, but instead a consistently held position held for 50 years when abortion was often illegal.

        I would consider the statement “taxation is theft” as a short hand method of indicating political philosophy in the same manner as the statement “property is theft”

      • My objection to “taxation is theft” is that it misses the distinction between theft and robbery.

        That aside, it isn’t an inflation of the term, it’s a consequence of the (debatable) belief that the government has no legitimate claim on the money it collects as taxes. Just like the abortion/murder case.

        The claim “failing to donate money for bednets is murder” is an inflation of the term, because it ignores the action/inaction distinction which is, I think, a central element of (all?) moral systems. As in the cases Scott mentions, it makes “murderer” meaningless since it implies that anyone who has anything he could have done that would have saved at least one life somewhere is a murderer.

        • Clutzy says:

          My objection to “taxation is theft” is that it misses the distinction between theft and robbery.

          That aside, it isn’t an inflation of the term, it’s a consequence of the (debatable) belief that the government has no legitimate claim on the money it collects as taxes. Just like the abortion/murder case.

          This is fairly correct from my POV. Taxation not being theft is a form of special pleading. Which I don’t think all special pleadings are wrong. I just have not yet been convinced this one is correct.

          The claim “failing to donate money for bednets is murder” is an inflation of the term, because it ignores the action/inaction distinction which is, I think, a central element of (all?) moral systems. As in the cases Scott mentions, it makes “murderer” meaningless since it implies that anyone who has anything he could have done that would have saved at least one life somewhere is a murderer.

          This is more than a mere inflation of the term. It lacks evidence that it is true from a utilitarian POV, which is the only POV that it can be argued from (that I know of). It is probably as likely that the average guy with disposable income, or his progeny/inheritors will save even more lives with the money in the future than a the nets he bought today. This is just compound interest and capital accumulation in action, whereas nets aren’t appreciable property. Most modern international charities, and large scale governmental “charity” projects show this. The evidence for positive dividends for these projects is scant.

          Of course it also has the problems you attach to it.

          • eric23 says:

            “Taxation is theft” is no more accurate than “taxation is debt collection”…

          • ARabbiAndAFrog says:

            Taxation is protection racket, everyone knows this. Biggest gang in town shakes you down, but if a smaller gang show up, they’ll probably step in at some point.

          • eyeballfrog says:

            Taxation is a user fee for the financial system the government runs.

          • John Schilling says:

            Taxation is a user fee for the financial system the government runs.

            The United States Government, at least, requires you to pay income taxes and in US dollars, even if your only voluntary economic transactions are done by pure barter. Try again.

          • eric23 says:

            The government built the “exchange” in which the barter took place.

          • John Schilling says:

            The government built the “exchange” in which the barter took place.

            If you and I are US citizens living on barren plots of land in New Zealand, and I walk over to your shack give you the wool from my sheep in exchange for a dozen of your potatoes, you and I are legally required to report this to the United States Government and pay them income tax in US dollars.

            I believe your argument would be more accurately rephrased as “Anyone who complains about paying the taxes I think they ought to pay is a Poopyhead; let’s think up some argument that makes me sound virtuous for saying so, and don’t bother checking if it’s true because that doesn’t matter”. Possibly I am missing some subtlety in your position.

          • eyeballfrog says:

            @John Schilling

            I am open to the idea that pure barter exchanges of ownership should not be taxed. It seems unlikely to have a large effect on tax revenues, given how much more convenient money is.

          • John Schilling says:

            OK, so if say Dr. Friedman and myself enter into an agreement to buy and sell goods and services to one another by physical exchange of Swiss Francs, you’ll agree that the United States government has no moral authority to tax it? Does it matter whether we do this at his house or in a boat off the coast, or can we just go do it through a Swiss bank already, and if any of these things morally exempt us from paying US income taxes do you seriously doubt that transactions so structured will have a large effect on tax revenues?

            Or are we really still at the United States government having the moral authority to tax everything you care about and we’ll tailor the rationalization to fit?

          • eyeballfrog says:

            Well, under this model the US would get to tax people using US dollars who aren’t citizens, which may offset the loss. It would also result in competition between countries w/r/t tax laws and currency systems, though I understand that for many corporations this already happens. It sounds like an interesting idea, though I can see issues with implementation on the whole “taxing non-US citizens” plan.

          • John Schilling says:

            It would also result in competition between countries w/r/t tax laws and currency systems, though I understand that for many corporations this already happens.

            It happens, but separately w/re tax laws and currencies. If the two are coupled the way you suggest, a small country like e.g. Switzerland whose government’s responsibilities are limited to A: defending policing and maintaining a small country like Switzerland and B: regulating a major banking industry that doesn’t care how people are earning their money and what they are doing with it and what other country’s taxes they are or are not paying, will be able to underbid pretty much anyone else. At that point, anyone who wants to e.g. engage in global peacekeeping or provide social services to thirty or three hundred million people or any other such thing, and tries to base their moral claim to tax revenues on “…but you’re using our currency so you have to pay our taxes!”, will come up rather short.

      • Gerry Quinn says:

        I would think of those as examples of rhetoric, rather than an attempt to inflate the generally accepted meaning of murder or theft. You are really saying ‘abortion is like murder in some ways” or “taxation is like theft in some ways”.

        • Conrad Honcho says:

          I think abortion is a bad example to use in this discussion. Pro-life advocates are not exaggerating: they literally believe a fetus is morally a human being, and killing it is literally murder. There is no distinction between murdering a human before birth and murdering a human after birth. I think a lot of people who are not pro-lifers do not understand this, and assume they’re being hyperbolic or engaging in rhetoric.

          • One possible way of getting the point across is to ask how they feel about infanticide.

            Most of the standard arguments for abortion being legitimate also apply to infanticide, provided you don’t prevent someone else from taking charge of the baby. They apply a little less strongly–but that’s also true of arguments for abortion at eight months vs arguments for abortion at four months.

            My guess is that most people who are sure that abortion should be legal, entirely at the discretion of the mother, are also sure that infanticide should not be.

          • Nornagest says:

            I’m not sure that line works too well. Birth is a hell of a Schelling point.

          • Secretly French says:

            I logged in to say this. Abortion is literally murder, because it is literally the considered planned deliberate knowing killing of a human being. The fact that that human hasn’t got as far in his life as being born, is simply not relevant to this point, any more than is one’s first haircut or losing one’s virginity.

            I would also like to push back against this use of the term ‘schelling point’ by Nornagest; didn’t we recently have a post about not abusing these technical terms? I understand a schelling point to be a decision made when agreement is necessary in the absence of communication – you have to pick the same card as your friend in the next room, so you pick the ace of spades thinking he will do the same. After which point to outlaw the deliberate intentional killing of a human being is not a decision made in the absence of communication – we talk about it all the time! In terms of arbitration, birth is an option, as is detectable heartbeat, earliest known survival following premature birth, conception, and possibly others.

            Crushing an acorn is not felling a tree, but it is ending the life of an oak. In this analogy, the human being is the oak, not the tree.

          • Conrad Honcho says:

            All I’m saying is that when choosing an example of “Morality Thing everyone agrees is of weight X inflated to be more like 2X so we can talk about the inflation” we should choose something everyone agrees the weight of Morality Thing is X. People assign very different values of X to abortion.

          • Dan L says:

            @ Secretly French:

            Abortion is literally murder, because it is literally the considered planned deliberate knowing killing of a human being.

            That’s the claim, yes. Just off the top of my head I can think of three different lines of objection, but if we’re arguing definitions probably the easiest is that you missed “unlawful” in your requirements for murder.

            Conrad’s right though, the object-level argument is a digression.

            I would also like to push back against this use of the term ‘schelling point’ by Nornagest; didn’t we recently have a post about not abusing these technical terms? I understand a schelling point to be a decision made when agreement is necessary in the absence of communication – you have to pick the same card as your friend in the next room, so you pick the ace of spades thinking he will do the same.

            Nornagest’s use is legitimate. Schelling points don’t exist only in a world of zero communication, they’re just the equilibrium solution points when mutual information is low-quality or unreliable.

            For the case of abortion, the relevant communication is on what people consider deserves protection and how this maps to human gestation. Birth is still a major milestone for a whole host of medical reasons, but it’s probably been passed in salience by “viability”. But it’s not clear, and you can still find an endless number of contexts that recognize one or the other.

          • Nornagest says:

            I would also like to push back against this use of the term ‘schelling point’ by Nornagest; didn’t we recently have a post about not abusing these technical terms?

            Do you have something better to propose, or did you just see something to complain about? Unlike the terms Scott mentioned, English doesn’t have a concise non-technical phrase that I can think of for a basically arbitrary but unusually salient place to draw a line. And inventing one would be worse, not better.

    • smocc says:

      I saw a sticker on a telephone pole recently that read “Beat the racism out of racists.”

      That worries me not because I’m categorically against violence, but because at the same time I’ve noticed people advocating violence against racists I’ve seen other people expanding the definition of “racist” to include almost anyone.

      • Matt M says:

        Right. A disturbingly large amount of the population seems to accept, independently of each other, both of the following claims:

        1. Everyone’s at least a little bit racist
        2. If someone is a racist, it’s okay to violently assault them

        Both claims seem “reasonable” enough on their own, but combine them, and it leads to some pretty downright horrible implications…

        • Secretly French says:

          I would love to see a defence of the reasonableness of (2), here of all places where people pride themselves on being intelligent rational people, not partisan ideology-drunk thugs.

          • Dan L says:

            I doubt you’ll get many takers until it’s substantiated in an intelligent, rational way, to borrow one of your descriptions.

        • Simon_Jester says:

          When directly questioned, do you find statistically significant numbers of people who actually believe the aforesaid implications?

          Or are these hypothetical implications that are rarely if ever actually believed by the people propounding the argument you’ve spotted?

          As in, are there significant numbers of people who DO believe that anti-racist beatings for everyone are a good idea?

          • LadyJane says:

            When directly questioned, do you find statistically significant numbers of people who actually believe the aforesaid implications?

            Of course not, because such people don’t exist in statistically significant numbers.

            Or are these hypothetical implications that are rarely if ever actually believed by the people propounding the argument you’ve spotted?

            Let’s call these “hypothetical implications” what they are: strawman arguments. Or at the very least, weakman arguments.

  26. ManyCookies says:

    But a wise supervillain once said, “When everyone is super, nobody is”.

    Syndrome is like the textbook definition of a high intelligence low wisdom character, a genius with a crippling obsession and a ridiculously dumb convoluted plan to satisfy it for the sake of his ego.

    (How’s that for some word deflation, wise+intelligent gaining some distinction thanks to DnD)

    • anchpop says:

      Can you think of any examples of low intelligence high wisdom characters? The only one I can think of might be HPMOR Dumbledore

      • Nornagest says:

        HPMOR Dumbledore isn’t stupid. He’s not very rational by Eliezer’s standards, but that isn’t the same thing at all. Hagrid would count, though, in most any Potterverse setting.

        Low intelligence, high wisdom isn’t uncommon in media. You probably see it the most in characters that’re cast as unusually spiritual: shaman types, animal lovers, hippies, elders (especially from ethnic minorities). Anyone highly religious, if the author doesn’t have a chip on their shoulder about religion. Among protagonists, you see it a lot in Western or noir stories: uneducated and kinda slow but stubborn and perceptive fits in well there.

  27. Zack M. Davis says:

    The later sections of my recent Less Wrong post, “Where to Draw the Boundaries?”, cover similar territory in the philosophy of categorization.

    I’m just really tired of seeing it again and again.

    Oh my god, me too! What’s really frustrating is when people who clearly understand the philosophical point you’re making (because, e.g., they’ve written extensively about it for years), suddenly seem to act as if they don’t understand when—and only when—understanding would be politically inconvenient. In the heat of frustration, it’s sometimes tempting to denounce such people as “liars”, but that’s actually a suboptimal use of categories, as you eloquently explain in this post. It’s a good thing people around here know that there are rules of rationality governing conceptual boundaries—it would be pretty absurd and shameful if someone were to explicitly deny this!

  28. Zephalinda says:

    C.S. Lewis has a good essay on this phenomenon, “The Death of Words.” He argues that precise descriptive terms under political pressure drift naturally toward becoming “terms either of mere praise or of mere blame.” According to that account, what we’re seeing might just be the natural movement of “lie” from meaning “a statement deliberately made in contradiction of the facts as the speaker understands them” to meaning “a statement that is Bad.”

    Especially interesting is his final point: “Men do not long continue to think what they have forgotten how to say.” “Lie drift” potentially implies some disturbing things about our eroding sense of any kind of objectively true shared reality.

  29. LadyJane says:

    Leaving aside the broader problem of shifting definitions, I think part of the problem in this particular case is that the Rationalist community overvalues Truth. That might seem like a strange claim; after all, we can all agree that Truth is good, right? But Truth is really quite a bit more subjective than people – especially Rationalists – tend to think, especially when it comes to human behavior.

    Let’s say Amy and Bob break up. Amy claims the breakup was Bob’s fault, and all of her friends believe her. Bob claims that breakup was Amy’s fault, and all of his friends believe him. Are either of their narratives True? There may be True events involved: Amy took an action that upset Bob, and Bob overreacted in a way that upset Amy, those are objective facts. But who bears more fault for the breakup? That’s an unanswerable question that likely doesn’t have any objectively True answer. A staunch reductionary materialist may argue that the parameters of the question do refer to some real physical things on some basic level (for instance, the precise neurological events occurring in their brains), but even if you agreed with that train of logic, the information required to answer the question is wholly inaccessible, and so the question still remains functionally unanswerable. For all intents and purposes, there is no True answer there.

    The same logic applies to arguments and debates. What does it mean to accuse someone of being racist? Some people may self-identify as racists. Others hold conscious opinions that most people would agree are racist. But others consciously oppose racism, while still being motivated by racist biases on a subconscious level. And the deeper you go, the blurrier the definition of racism becomes. Someone might simply have an instinctual revulsion to certain aesthetic traits that happen to strongly correlate to particular racial groups, without any kind of intellectual biases against those races; could they meaningfully be described as racist?

    What does it mean to say that someone is being intellectually honest in a debate? Many people hold different and oftentimes conflicting views in their mind. Someone might make an argument that they believe is possible, even if they’re not fully convinced of it. In a different conversation, they might make the opposite argument, without actually being dishonest in either case. Or someone might make a deliberately exaggerated version of an argument they genuinely believe to be true – perhaps as a way of testing the argument’s limits, or perhaps simply as a form of emotional catharsis.

    Our words and concepts and mental images are all rough abstractions at best, and these abstractions can be inadequate for describing the actual world around us. And, perhaps counter-intuitively, one of the things they are most inadequate for describing is the internal mental states of human beings.

    • Peter Gerdes says:

      But the failure in your example isn’t overvaluing truth. It’s making the mistake of believing either Amy or Bob’s explanation was true. If you really believe there is no fact of the matter then that would be the true account and it is false to accept either Amy’s account or Bob’s.

      • LadyJane says:

        @Peter Gerdes: I think you’re missing my broader point, which is that the internal workings of the human mind are so utterly incomprehensible and indescribable that any statements about them may as well be considered purely and totally subjective, and devoid of any objective truth value whatsoever.

        @ScottAlexander: I accidentally reported Peter’s post while trying to reply to it, please ignore. Sorry about that.

    • InvalidUsernameAndPassword says:

      The people who self-identify as racists also self-identify as anti-racists, even though it sounds like something the Ingsoc Public Relations department could’ve come up with. Meanwhile people who’re comfortable using racial slurs swear they’re definitely not racist. (They’re “race realists”, which everyone assumes means racist, another good example of word inflation and replacement by euphemisms).

      • Meanwhile people who’re comfortable using racial slurs swear they’re definitely not racist. (They’re “race realists” …

        How are you using “racial slur”? Does it mean any negative claim about a racial group, true or false?

        I would take the statement “African-Americans have a lower average IQ than other Americans” or “black Americans have a much higher than average murder rate” to be the sort of thing that would be said by someone who self-described as a race realist. But I wouldn’t describe it as a slur, or someone who made the statement as, on that basis, a racist.

        To me, a racial slur is something like “all blacks are stupid” or “all Japs are sneaky.”

      • Conrad Honcho says:

        A “race realist” who’s comfortable using racial slurs is a racist lying about being a race realist.

        • “Racial slur” defined how?

          Someone claims that the police obviously are racists, because they arrest more blacks, relative to the population, than whites. Someone else responds that blacks commit more crimes, and offers evidence for that claim.

          Is that a racial slur?

          • Conrad Honcho says:

            No, a racial slur is something like the n-word. “Look, I don’t hate n*****s, I’m just saying n*****s commit more crime. Just being a race realist!” The whole point of inventing the term “race realist” was so people could talk about sensitive racial issues and the differences between races while not motivated by or engaging in racial hatred. Once one starts using the racial slurs, they can no longer credibly claim to be not motivated by or engaging in racial hatred.

    • Aapje says:

      @LadyJane

      There may be True events involved: Amy took an action that upset Bob, and Bob overreacted in a way that upset Amy, those are objective facts. But who bears more fault for the breakup? That’s an unanswerable question that likely doesn’t have any objectively True answer.

      Who is (more) at fault depends on the explicit and implicit agreement that rules the relationship. For example, if the explicit agreement or cultural norm (that creates an implicit agreement) is that the relationship is monogamous and one person has sex with others, then that person bears fault.

      the information required to answer the question is wholly inaccessible

      This is nonsense, because there are objective facts that we can determine ourselves, people admit to certain things, etc.

      Your argument can just as easily be used to argue that (legal) justice is impossible. I’ll be the first to agree that perfect justice is impossible, but that is a far cry from arguing that the truth is utterly inaccessible.

      But others consciously oppose racism, while still being motivated by racist biases on a subconscious level.

      You forget the possibility that people may have a racist definition of ‘racism.’

      You also forget the possibility that people may have inconsistent definitions or apply their definitions inconsistently.

      • LadyJane says:

        Your argument can just as easily be used to argue that (legal) justice is impossible. I’ll be the first to agree that perfect justice is impossible, but that is a far cry from arguing that the truth is utterly inaccessible.

        Not necessarily, because a person’s actions can be objectively defined, so claims about a person’s actions can be objectively proven to be true or false. “Karl hacked the victim to pieces with an axe” is not a subjective statement, it’s something that either happened or didn’t happen. Likewise, it’s possible to prove with objective certainty that an individual knew something or planned something: If a killer wrote detailed notes about how he was going to murder his target, then it’s safe to conclude his crime was most likely premeditated. And if the killer’s accomplice helped to cover up the body, then it’s safe to conclude that the accomplice was aware of the murder, and was lying when he said he didn’t know the victim had been killed.

        Where it gets trickier is with distinguishing between something like second-degree murder (a killing that was intentional but not premeditated) and voluntary manslaughter (a killing that was unintentional but resulted from intentional violence). That’s a distinction that’s dependent on knowing the exact mind-state of the perpetrator at the time of the murder, and in a lot of cases, it really does seem to come down to what the prosecution feels like charging them with. When Karl was hacking the victim to pieces, was he thinking “I hate this guy, I want him to die” or “I hate this guy, I want to hurt him as badly as possible”? In all likelihood, Karl wasn’t thinking anything as coherent as either of those two statements, he was thinking something closer to “grrrr hate smash hurt angry bash red make-bad-thing-go-away keep-hitting raaaaagh.”

        It gets even trickier when it comes to more abstract crimes. There are clear-cut cases of investment fraud, where a con artist deliberately lied to investors and misled them, but there are also plenty of gray areas. An entrepreneur may genuinely believe his startup company is going to be a success, and might tell himself lies that trickle down to investors. His internal model of the world could be a messy storm of contradictory notions, all competing for the same truth value.

        I’m not saying legal justice – or psychology, for that matter – is impossible. But to use a metaphor, I do think the inner workings of the mind are probably less like a cartoon model of an atom with electrons floating around in a neat tidy clearly-defined orbit, and more like an actual atom with fuzzy electron probability fields in a state of constant flux.

  30. Peter Gerdes says:

    Great post but one small quibble. It’s not literally true that you never want to set the boundary of a concept outside of the population. For instance, if we managed to bring about a very very low crime society so that no currently living person was a murderer we might still want to keep the current boundaries of the word murderer so we could usefully oppose policies which might rollback progress and cause people to become murderers again or usefully talk about the probability that someone might become a murderer.

    Even if the concept isn’t a particularly live possibility sometimes it can be useful for understanding the world, e.g., explaining why it isn’t actual etc..

    • blacktrance says:

      Related quibble: it’s useful to have words for things 100% of the population currently does, because they include eating, breathing, etc.

      • Peter Gerdes says:

        Yes, though with those concepts I’d say the relevant ‘population’ is all objects not all people. I presume his use of population was meant for words that are implicitly restricted to people (we don’t call rocks or wolves murderers or liars).

        But now I’m nitpicking nitpicking and I should really work instead.

      • RandomName says:

        But would it be useful to have a word for an “Eater” or a “Breather”?

        The term “Eating” is useful, because people don’t spend 100% of their time eating (there are non-eating actions) but the term “An Eater” is useless, because it describes literally everyone. Scott makes this point when he says:

        I’m cheating here by talking about “abusers” rather than “abuse”, since there is still a useful distinction between abuse and non-abuse actions.

        The expanded definition of abuse renders the term “Abuser” useless in the same way that the terms “Eater” or “Breather” are useless. It just means human. But the expanded definition of “Abuse” is still somewhat useful, since it doesn’t describe all actions

    • n-g says:

      One useful concept whose boundary may be outside of the population on the other side: “biased”.

      Example from the above post:

      We’re probably all biased to some degree.

      It seems to me that the problem you’ve articulated is not that the scope of the redefined “liar” and “abuser” are too large in an absolute sense, but instead too large for their connotations.

      • Peter Gerdes says:

        I mean I agree we do have those concepts but biased is a complex case because of its context dependence. I mean when we say “Bob is biased about immigration” we don’t mean Bob isn’t perfectly rational. Rather we draw some cut-off that most people don’t pass and we are saying Bob is even more biased than that.

  31. randallsquared says:

    She told me that sometimes she needed her boyfriend to do some favor for her, and he wouldn’t, so she would cry – not as an attempt to manipulate him, just because she was sad. She counted this as abuse, because her definition of “abuse” is “something that makes your partner feel bad about setting boundaries”.

    Well, it’s definitely not as abusive as hitting someone, or even as abusive as consciously intending to make her boyfriend feel bad, but given that most people do not cry when someone refuses a favor, and presuming that her boyfriend does feel bad and then cave, or sometimes cave, I’m not quite so ready to say that this isn’t at all abusive in any way…

    • Bamboozle says:

      It is manipulative for sure. Though i wouldn’t say abusive. It’s not causing harm to the boyfriend, but it’s not nice.

      • Aapje says:

        It may be causing harm to the boyfriend if he is made to do things merely through coercion.

        Imagine that the girlfriend would threaten to kill herself if he didn’t do something. This is ultimately a milder version of the same basic mechanism, where she is getting her way not trough reason, but through showing pain.

        Of course, as Scott argues, we probably want to have a threshold before we call it abuse.

        • Deiseach says:

          It’s a little manipulative, right enough, but it’s probably not all that great for the girlfriend either if every time someone tells her “Sorry, I can’t do that right now” she bursts into tears because she’s so sad.

          Presumably she’s sad because he’s her boyfriend and they’re supposed to be special to each other and care about each other and not doing things she really really needs him to do isn’t caring about her. That’s not conscious manipulation and she probably does feel real hurt and sadness, but it’s also a line of behaviour she needs to stop right now (it’s too late to nip it in the bud before it gets established).

          She needs to grow a thicker skin and work on internalising “Sometimes people say ‘no’ not because they don’t care but because they can’t do it or it’s really not that important”, and let that help her avoid turning on the waterworks.

          • Aapje says:

            @Deiseach

            It’s actually very child-like, in that small children who are unable to either articulate their demands or to be persuasive, often resort to crying to get their way. This is not that great for those children compared to the alternative of getting their way with reason, although usually superior to not getting their way.

            The question is whether the behavior is due to an inability to get her needs met with reason, in which case, developing that ability would seem the best; or whether she actually wants more than is her due & gets it through crying. In the latter case, I would call it somewhat abusive.

        • ec429 says:

          It may be causing harm to the boyfriend if he is made to do things merely through coercion.

          Someone crying at you is not coercion. Please stop inflating that word; we libertarians need it.

          • Mark Atwood says:

            Someone crying at you is not coercion.

            I’m going to make a guess that you’ve never had an emotionally manipulative boyfriend or girlfriend.

            For most of them, most of the time, the tears are actually even genuine.

            It’s still abusive, it’s still coercive, and it is not ok.

          • Matt M says:

            Manipulation is not coercion. They are different words with different meanings, for a very good reason.

          • Aapje says:

            Compulsion, then.

    • Randy M says:

      I’m pretty firm in teaching my daughters that crying is not a way to get what they want; it is indeed manipulative when used to get somebody to change their mind. (Daughters are not infants, just to be clear, but preteens.)
      I’m still reluctant to call it abuse. Maybe if they sought out someone emotionally vulnerable, or realized it was a particular weakness of their partner, but even then, probably not.
      If we broaden the terms enough, the woman crying to get her way is abusive, as is her boyfriend walking away when she starts crying. Or vice versa. The term starts meaning “counter productive behavior” rather than sign of particular danger or even psychopathy.

  32. benquo says:

    The definitional boundaries of “abuser,” as Scott notes, are in large part about coordinating around whom to censure. The definition is pragmatic rather than objective.*

    If the motive for the definition of “lies” is similar, then a proposal to define only conscious deception as lying is therefore a proposal to censure people who defend themselves against coercion while privately maintaining coherent beliefs, but not those who defend themselves against coercion by simply failing to maintain coherent beliefs in the first place. (For more on this, see Nightmare of the Perfectly Principled.) This amounts to waging war against the mind.

    Of course, in matter of actual fact we don’t strongly censure all cases of consciously deceiving. In some cases (e.g. “white lies”) we punish those who fail to lie, and those who call out the lie. I’m also pretty sure we don’t actually distinguish between conscious deception and e.g. reflexively saying an expedient thing, when it’s abundantly clear that one knows very well that the expedient thing to say is false, as Jessica pointed out here.

    * It’s not clear to me that this is a good kind of concept to have, even for “abuser.” It seems to systematically force responses to harmful behavior to bifurcate into “this is normal and fine” and “this person must be expelled from the tribe,” with little room for judgments like “this seems like an important thing for future partners to be warned about, but not relevant in other contexts.” This bifurcation makes me less willing to disclose adverse info about people publicly – there are prominent members of the Bay Area Rationalist community doing deeply shitty, harmful things that I actually don’t feel okay talking about beyond close friends because I expect people like Scott to try to enforce splitting behavior.

  33. Cory Giles says:

    One of my hobbies is reading Hansard, the record of the UK Parliament, especially from 1937-1946.

    One of the many interesting things about UK parliamentary procedure is that it is always out of bounds to say, or suggest, that any other member of Parliament is lying. Yet obviously, being politicians, they lie frequently. So how does it work?

    If a MP wants to suggest someone is lying, they would say something like “I am sure the honorable member was mistaken when he said (X)” or, in very extreme cases, “It seems the honorable member misled Parliament when they said X.” The point is that a lie = a falsehood + intent. Since it is basically impossible to prove intent, it is not wise to accuse anyone of lying in any high-stakes situation. That is why they have banned such accusations entirely.

    In the end, it doesn’t really matter whether someone is lying or is wrong. Either way, the remedy is to show evidence proving them to be mistaken. If you don’t have such evidence, then you definitely can’t prove them to be lying, since one of the two requirements to be lying is to be mistaken. When it comes to prognosticating future events, like anything to do with AI, no one can be correct or mistaken because it deals with things in the future. In that sort of case, an accusation of lying amounts to an accusation that someone is predicting a thing they don’t believe will really happen. It is totally unfalsifiable. Therefore it should be left where it is and any predictions about the future should be taken with a large grain of salt.

    • Aapje says:

      Something similar is true in the Dutch parliament. The easy solution is to accuse the honorable enemy of the people of telling falsehoods, rather than lying.

    • peterispaikens says:

      It definitely does matter whether someone is lying or is wrong. Motive matters a lot when interpreting actions and statements of others, and deciding what to do about them.

      If MP Bob said that some project cost 5 money when it actually cost 50 money, then it makes a big difference whether that was an accidental mistake or an intential lie because it gives us highly relevant information on how to interpret other actions and statements made by Bob.

      An accusation of lying is also often not a generic “Bob is untrustworthy” but implying a particular specific intent and motive – there may be parallels with game theory; you asserting that one of the “players” has chosen a particular strategy is important information.

      If Bob says that we should do X because Y and I’m saying “Bob, you’re a liar” then that’s not equivalent to me saying “Y is not true” – it’s also me implying that everyone around me should consider that Bob is intentionally biasing the information he provides to advocate X, they should consider everything in the past and future that Bob did to favor X as untrustworthy, that they should consider what facts Bob is omitting because they don’t favor X, etc. If Bob is/was positioning that he doesn’t really care about X but is just doing it to get Y, then revealing that lie reveals Bobs preferences, and revealing that lie to others demonstrates Bob’s true preferences to others which is often also an important strategic/political result.

      • Randy M says:

        Yeah, in any particular instance, it doesn’t matter, but if you want to build a predictive model for the speaker it very much does.

      • Cory Giles says:

        I would say the difference is minimal at best because if Bob frequently lies or is just frequently mistaken, the response is the same from your end: don’t rely on anything Bob says. The predictive model could be trained just as well by noticing how frequently Bob is wrong. If you observe that he may have motives for being wrong then, well, especially in politics, everyone has motives.

        Keep in mind also that dealing with skilled players, it is not easy to prove even that they were mistaken. If Bob says the project costs 5 money and it really costs 50 money, Bob might well reply, “I meant 5 money this fiscal year” or “5 money in direct costs although indeed there are another 45 in indirect costs”, or “ah, but the 45 money difference will pay for itself through GDP growth”, or something similar. And even if Bob has a very good selfish motive for claiming a cost of 5 money, it could be that he was presented with different ways to state the cost and chose the lowest one through a combination of cognitive bias and self-interest, which I would suggest is a much more common situation than baldfaced lying.

        It is not that people don’t lie, or that there is no difference between an intentional and unintentional falsehood. Just that, in environments where accusations of lying are permitted, the accusation is thrown out far, far more often than it can be substantiated. If you accuse someone of lying, and can’t prove it, it’s only a suspicion (pretty much every day in US politics), then aren’t you guilty of lying as well? So it is better in my view not to go there.

        I see what you mean with the reference to game theory, but the trouble is that just as your opponent may have a motivation to lie, you also have a motivation to accuse them of lying when you don’t know for sure. If you are at loggerheads with someone you should just assume they will cherrypick evidence and combat it with the facts.

  34. arch1 says:

    Scott, I definitely get and agree with your point here, but it does prompt me to ask the following question about an earlier post that I liked even more: The one from early May concerning the 5-HTTLPR studies. When you said “ALL. OF. THIS. IS. LIES.”, were you using “lies” in the strict sense you discuss in the present post (which I assume refers to intentional untruths), or by “LIES” were you (as I assumed) indulging in a rhetorical flourish when your literal intended meaning was in fact closer to “FALSE”?

    • Evan Þ says:

      A good question.

      My interpretation, thinking back over that post and trying to parse out how I subconsciously interpreted that line, would be that Scott isn’t accusing the scientists of lying. Rather, he’s saying “When I was telling you that just a few paragraphs ago, I was lying to you. These claims are not true; I know they’re not true; I was stating them merely for the rhetorical effect of the contrast I am about to make.”

  35. rahien.din says:

    In a personal sense, I agree. For instance, it drives me kind of bonkers when my wife says, “Whoops I lied!” after realizing she’s made some rare factual error. I want to say no, you didn’t lie, you just made an unintentional error. It seems too harsh for her to call herself a liar over the mistakes that we all make.

    But I think this misses an important point. What we care about is whether a person is taking responsibility for their own trustworthiness.

    Consider the set containing three categories of untrustworthiness : {those who are intentionally misleading, those who intend to be trustworthy but fail, those who do not sufficiently intend to be trustworthy}. “Liar” could then serve one of two purposes.

    It could, as you suggest, cleave {intentionally misleading} from {intend trustworthiness but fail, do not intend trustworthiness}. In this case, when you say to your friend “You’re not a liar,” you are saying “You aren’t malicious, you might just be lazy or stupid.” This definition focuses on malice rather than effort, and implies that as long as you aren’t evil, you don’t have to try to be responsible.

    Alternatively, it could cleave {intentionally misleading, do not intend trustworthiness} from {intend trustworthiness but fail}. This definition focuses on effort. In this case, when your friend calls herself a liar, she is saying “It doesn’t matter that I’m not malicious, I need to try harder at being reliable!” This definition focuses on effort, and implies that as long as you are trying, you are meeting the standard for personal responsibility, even if you fall short.

    The second definition is by far the better response to universal human faultiness, because it separates the people who are trying to be reliable from the people who are not. Moreover, it is implicitly focused on how to help the person who had been harmed, rather than on how to absolve the one who does the harming.

    For instance, consider this concept applied instead to abuse. Three categories : the malicious person who intentionally insults their partner, the hapless person who puts their foot in their mouth, the inconsiderate person who insults their partner not intentionally but thoughtlessly. It is most appropriate to cleave {hapless} from {malicious, inconsiderate}, because the hapless person has the correct intention and thus is at least motivated to improve their performance. The other two have no potential for growth.

    • LadyJane says:

      It is most appropriate to cleave {hapless} from {malicious, inconsiderate}, because the hapless person has the correct intention and thus is at least motivated to improve their performance. The other two have no potential for growth.

      I disagree, particularly when it comes to relationships. If someone is inconsiderate to their partner, that’s a sign of apathy, which indicates a bad matchup (lack of common interests, lack of physical attraction, lack of chemistry). Whereas if someone is malicious, that’s a sign of abusive personality traits, which indicates that they’re just a bad person. The hapless person could learn from their mistakes, and the inconsiderate person would probably act differently towards a partner they were more compatible with, but I’d expect the malicious person to treat any other partner just as poorly. That makes a pretty huge difference when it comes to their future relationship prospects: I’d definitely advise a friend against dating someone who was malicious toward their previous partners. But I wouldn’t necessarily advise a friend against dating someone who was either hapless or inconsiderate toward their previous partners.

      I’ve known people who’ve been apathetic, inconsiderate, and thoughtless with their partners, sometimes to the point of offending them, simply because they weren’t really invested in the relationship. I wouldn’t consider any of them bad people; they were much more considerate and thoughtful in relationships with people they actually liked, and some of them even remained platonic friends with the exes they’d been apathetic about dating. I’ve also known actual verbal abusers, the types of people who would deliberately insult their partners all the time just to lower their self-esteem and/or social status, and they’re in an entirely different ballpark. Those people always act the same toward all of their partners, and I couldn’t imagine one of their exes remaining platonic friends with them.

      • rahien.din says:

        You clearly don’t disagree in principle. I know this because you are insisting that these designations must be leveled at contexts, but you are using that same cleavage within each context – {hapless} from {malicious, inconsiderate}.

        I would agree that a person might behave differently within a different context. However, we would apply that cleavage not only to single-relationship contexts, but also to larger more durable/reproducible patterns of behavior. If you knew someone who always acted inconsiderately toward all of their partners, you would consider it a durable pattern and warn your friends not to date them.

        One idea you bring up is prediction – is it easier to predict who will behave badly in subsequent relationship, based on our interpretation of their motives? Surely it is. That’s the whole point of having motives. But even malicious (or hapless) behavior may be driven by context. It might be too quick to slap that “evil” or “sociopathic” or “vicious” or “those people” label.

        • LadyJane says:

          But your own reason for cleaving {hapless} from {malicious, inconsiderate} was because the hapless person had more potential for growth (i.e. he might learn better social skills), and thus was likely to behave differently in future relationships. My point was that the inconsiderate person’s behavior is also likely to be different in future relationships, albeit for different reasons. So by the metric you originally posed – likelihood of behaving differently in the future – I would consider {hapless} and {inconsiderate} to be closer to each other than either is to {malicious}.

          And no, I wouldn’t go as far as to say that all abusers are completely irredeemable and can never change. Some of them can and do change their ways over time. But on average, going by my intuition and personal experiences, a malicious boyfriend/girlfriend is much less likely to behave differently in different relationships than either a hapless or an apathetic person.

          If you knew someone who always acted inconsiderately toward all of their partners, you would consider it a durable pattern and warn your friends not to date them.

          Sure. But the same is true of a hapless person who kept offending their partners out of sheer social ineptitude. If it was a consistent enough pattern, I’d think “well, they’re probably either unwilling to learn or incapable of doing so,” and warn friends against dating them.

          • rahien.din says:

            your own reason for cleaving {hapless} from {malicious, inconsiderate} was because the hapless person had more potential for growth

            False.

            This all comes down to motives. Remember? Scott and you are saying “The term ‘liar’ denotes malicious intent.”

            I am saying that this is a lower bar than we want.

            The threshold for “good person” is not simply “lacks malicious intentions.”

            You are not a good person if you are indifferent to the suffering of others.

            Instead, the bar for “good person” is “possesses and acts upon intentions to do good.” This is the threshold that cleaves {hapless} from {malicious, inconsiderate}.

            I also need to clarify : any of those designations implies a deficit in social performance, and thus some potential for greater skill acquisition. But the intention to do good is a stronger predictor of improvement than is apathy or malice.

    • LadyJane says:

      As for your first example, I have no idea what {those who do not sufficiently intend to be trustworthy} even means.

      If you’re referring to bullshitting, i.e. making a statement to get a reaction from people without concern for whether or not it’s true, then I’d consider that closer to {intentionally misleading}. However, bullshitting isn’t really “lazy,” and it doesn’t lack malice. It’s an active attempt to manipulate and potentially deceive someone, even if the bullshitter isn’t sure that their statement is false. It’s malicious in the exact same way that outright lying is, even if it’s not necessarily an actual lie.

      But if you mean something more like {Jack gets asked a question and just gives an answer that he’s 60% sure is true, because he can’t be bothered to look up the answer and find out for sure}, then I’d consider that closer to {intends trustworthiness but fails}, even if there are elements of laziness, apathy, and known uncertainty that wouldn’t be present in a purely honest mistake. I wouldn’t consider Jack deceptive in the same way as a liar or a bullshitter. If he did it often enough, I might consider him unreliable, but that would also be true for someone who was simply wrong often enough, even if their mistakes were all purely honest with no element of laziness involved.

  36. Jacob says:

    Where should we draw the threshold? Let’s putanumonit.

    One approach is to draw the line right between the extremes. Let’s look at “racist”: I think that the median American isn’t very racist, certainly not much more so than the most scrupulous progressive, and certainly a lot less than a Klansman. So I would draw the “racist” line at some point that would probably exclude at least 70-80% of Americans. The people who would draw the line to *include* 90% of Americans probably use the same heuristic – they actually believe that the median American’s perpetuation of the status quo makes them closer to a Klansman than to them.

    Another approach is consequentialist: how costly is a false negative compared to a false positive? Then we would balance the two errors out to minimize the total cost. But again, the costs of each are born by different people, who then have competing incentive with respect to where the line should be drawn.

    But at the very least, if we agree on which heuristic we’re using, people in the “most Americans are racist” and “most Americans aren’t racist” camps are one step closer to arguing about their actual cruxes and having a more productive discussion than just calling each other names. If you’re arguing about where to draw a line around a concept, try to specify which general approach you’re using.

    • DinoNerd says:

      *sigh* That seems like a good common-sense way to use language. But we don’t agree on out heuristics, and really can’t, because of context. Sometimes “racist” is a general purpose insult, to be flung around like “gay” or “dumb” or whatever. Sometimes it’s a description of behaviour/thinking (that was a racist argument). And sometimes it’s an attempt to categorize people (Joe the KKK grand wizard is racist; Bill who simply assumes that a black person in a lab coat is not in fact the doctor isn’t.)

    • tscharf says:

      I can think of about zero conversations in which the term racist was applied to a person in the past couple years that was intended to be constructive and enlightening.

      Racist is now a Scarlet Letter given out by a society that meets on Twitter. People who use this term are not interested in a stricter or better understood definition, they are almost solely interested in its destructive power. It is a weapon that now has a dull edge that will be rendered harmless when disagreeing with someone’s favored identity group is prima facie evidence of racism. The term prejudiced would be more accurate most of the time when there is a valid issue at hand, but this term doesn’t have the desired cancel power.

  37. Confusion says:

    For all things you want to express there are multiple words to choose from. The ones you choose are not only chosen for accuracy, but also for emotional impact. What we’re seeing is the result of people feeling they need to use stronger and stronger words, in terms of emotional impact, in order to be heard. They are willing to sacrifice accuracy to be heard. Calling that out is subsequently called out as deliberate derailing of the discussion, in ways that may involve accusations of ‘gaslighting’ them as to the meaning of words.

    We’re basically already seeing that ‘lying’ is not really a problem any more. The president of the US is a habitual liar. So isn’t this a lost cause and shouldn’t we indeed just come up with new words and use them in ways to give them strong emotional impact?

    • Bamboozle says:

      I agree entirely, i think some new word that essentially means “lie” but with a brand new emotional impact will come along and everyone will start using it to maximum effect for a while until they discover something else.

    • JPNunez says:

      If the president lies non stop and calling him a liar no longer moves people one way or the other, I doubt the problem is with vocabulary.

  38. Bamboozle says:

    How much of this is the fact that every generation takes the existing language and changes it so that some words don’t mean what they used to anymore (like gay going from happy to homosexual)?

    I take your point in all of the above, but 300 years from now “abuse” could mean something very different than it does today. Same with gaslight. To systemiser brain types like those that frequent this forum this might seem horrifying but language is something that is constantly changing and words don’t mean the same thing forever.

    Trying to get people across western civilisation to hold to some strict set of meanings for a word seems almost like a fool’s errand.

  39. JulieK says:

    In the restricted-definition world, a few people write posts suggesting that there may be biases affecting the situation. In the expanded-definition world, those same people write posts accusing the other side of being liars perpetrating a fraud.

    The problem is that while you prefer the first world, most bloggers prefer the second.

  40. Steve Sailer says:

    Here’s a suggestion I have: We currently have the terms Type I and Type II Errors. But almost nobody seems to apply the concepts to stories in the news, like all the articles condemning “BBQ Beckys” for calling the police over things that turned out to be False Positives.

    My guess is that Type I/Type II is just about the worst possible terminology for lodging these important concepts in people’s minds. Has anybody proposed an easier to remember alternative to Type I/Type II Errors?

    • Aron Wall says:

      False positive / false negative.

      (Or possibly the other way around since I can’t be bothered to look up which one is type I right now)

      • Oscar Sebastian says:

        Your order is correct, unless someone malicious or mistaken edited the Wikipedia page on these errors and no one’s noticed.

  41. YehoshuaK says:

    I realize this is a kind of long post arguing against a weird thing that not many people are doing.

    I think this is a common thing that lots and lots of people are doing.

    The word “Nazi” used to mean “an adherent of the ideology of the Nazi party,” or more strictly “a member of the Nazi party.” Now it means, roughly, “shut up.”

    The same is true of “racist.” “Racism” has gone from meaning “the belief that people are entitled to more or less rights and privileges because of their ethnic background” to meaning…oh, right, “shut up.”

    Someone mentioned the frequent warning “this product has substances known to the State of California to cause cancer.” It doesn’t mean that it has substances known to the State of California to cause cancer. It means “don’t sue me.” The same is true of “not suitable for children under the age of 5,” and “the contents of this cup may be hot.”

    All of these came about through the process you’re talking about. People taking a word (or phrase) that had some powerful meaning, and applying it to something else.

    The problem is easy to define–people tend to take words that have meaning and apply them to other things, thereby rendering the words fuzzy and unusable. Why do they do that? In my opinion, the most frequent reason is to fool people into category error. If I can apply the word “racist” to your opinions, or “Nazi” to your person, then people will reflexively shun your opinions and you, and I don’t have to actually argue or debate you. It is, in other words, a lie.

    And I mean “lie” with a restricted definition–the intentional conveyance of false information in order to deceive.

  42. An anecdote: My brother once called something “literally Nazi” or “literally Nazism” because it was nationalist and socialist. I commended him on his use of “literally”.

  43. Murphy says:

    “Racist” (similar for many other “ist”‘s ) : old meaning, someone who actually has something against other races.

    Meaning creep: everyone is racist all the time in everything they do.

    Result:

    https://www.youtube.com/watch?v=RovF1zsDoeM

    gaslighting:

    old meaning, misleading someone to torment them, mess with their head and convince them their perceptions are wrong.

    new meaning: “they said something I disagree with”.

    • March says:

      Interesting. I see the scope creep for ‘racist’ going the other way around. I’d say, in true SJW fashion, that everyone is racist in the sense that even if they have great working relationships with people from loads of countries and come from mixed families and what have you, they’re still going to have a xenophobic/negative stereotyping reaction every now and then.

      People who take offense at being called racist seem to think that people are accusing them of actively doing something racist all the time.

      • Murphy says:

        Well yes. That’s exactly the kind of thing we’re talking about.

        The term got stripped of all meaning by people enthusiastically embracing the idea that everyone is racist all of the time, no matter what.

        Like a bizarre version of original sin.

        Then their reaction when calling people racist stops having any effect because it’s equivalent to “this person is a homo sapien, shock horror.”:

        Who could have predicted that it would remove the bite from the term:

        https://i.kym-cdn.com/entries/icons/original/000/027/475/Screen_Shot_2018-10-25_at_11.02.15_AM.png

        The language was devalued due to their free choices and those actions and choices had consequences.

        If “racist” still had any meaning it would mean someone actually doing something racist.

        not the idea that if you flicker images faster than they can consciously perceive they’ll react with slightly higher heart rates for images of black people … in a way that correlates with none of their conscious decisions.

  44. njnnja says:

    This insight is quantified by the information content of a message. It can be shown that a more unexpected/rarer message (e.g. “lie” when that applies to 1% of the population) actually contains more information than a message that is less unexpected/more common (e.g. “lie” when that applies to 75% of the population). Of course, you can always send more information by including more bits (e.g., “and by the use of the term ‘lie’ I mean that Bob has knowingly and willingly made a statement that is verifiable about some observable condition but the statement made does not accurately represent the observable condition that an objective party would measure, and Bob said that even though he knew or had reason to believe…”), but A shorter message is a clear benefit in a world of finite capacity.

  45. thomasbrinsmead says:

    That’s a lot of words to point out that if you foolishly define the threshold for the motte so broadly that it includes the bailey and beyond, you could well end up dancing so recklessly outside of the actual bailey that the actual motte is effectively useless.

    It’s also a lot of words to point out that non-agreed semantics hamper communication, and some semantic interpretational choices are not very helpful for either communication or analysis.

    Even if your co-conversationalist redefines a word, there’s nothing to stop you from keeping the old concept – just give it a new word. (though its equally sensible to use a new word for the extended concept – and Alex rightly points out that the choice has a significant impact on the convenience of interpreting texts that weren’t written following your new convention.)

    The feeder to wolves of troll’s scarf tree sail slayer carves truth songs with his mouth’s battle twig, a wave steed’s foam path Alex draws quite nicely, don’t you think?

    • Conrad Honcho says:

      The feeder to wolves of troll’s scarf tree sail slayer carves truth songs with his mouth’s battle twig, a wave steed’s foam path Alex draws quite nicely, don’t you think?

      You took the words right out of my mouth.

  46. midjji says:

    Seems like a big piece of the problem is the insistence/heuristic on extrapolating actions/events/properties to classes.

    If instead of saying “john is an abuser”, you say “john refuses requests sometimes”, or “john struck his partner”. The problem is alleviated, it also becomes easier to reason with e.g. john as rather than arguing with a forced or chosen identity/group membership you are arguing about events. One downside is that it is less efficient i guess, a bigger one is that people feel free to force identities on people, and us vs them comes up alot if you try this in practice -_-‘

    I’d think a good/common idea of lying is something like “A believed/knowingly false statement made with the intent to deceive” The last part excludes a great deal of stuff people dont think of as lying e.g. stories, entertainment, sarcasm, social protocol/white lies, examples etc. No doubt to broad and too shallow, but closer perhaps.

    But there is a problem with this if you use it to attempt to never lie/ only speak truth. Lets say I make a statement is true, and I know it is. However I make this statement knowing that the recipient will draw a false conclusion and remember me having said something different. Is that a lie? What if they will draw the specific false conclusion which I desire? If i didn’t know that they would misunderstand, but I should have as it would only have required minimal work to find out on my part? If I seek to only speak the truth, am I obliged to find out?

    This has long created a conflict for me as I am strongly inclined towards to speak only truth, but if such a statement is a lie, truth becomes subjective which I find detestable. On the other hand its fun to jokingly play a fey like character, seeing exactly how much I can twist someone’s understanding only using strict truths and “honest” answers and that is certainly deceptive. I am tempted to call this misuse of truth, but it is commonly called lying and most people react to it the same way.

    The consequentialist differences between a true statement which will lead to misinformed action and a false one is bothersome, similar to a false statement leading to informed action. It’s not just a hypothetical though, try as I might, false statements are extremely useful in teaching, as the nuance and complexity of the truth, or even just the qualifiers required to make the statement true, impede progress towards understanding. Am I a liar?

  47. aristides says:

    Another point in your favor, we already have a word for moral failings everyone is guilty of and you should feel bad about, but not treat others who commit it differently, and you used it several time. You can just call them sinners or say that they sinned, and move on. Lier doesn’t need to absorb everyone like racist almost has.

  48. bagel says:

    Boring possible alternative: most people who “lie inflate” do so when they are embarrassed.

    Consider the following hypothetical sequence:
    1) Highly motivated person makes a hyperbolic (or even fraudulent) accusation, using the parasitic accusatory power that Scott describes.
    2) Some poor sap, who doesn’t know any better, believes them.
    3) Poor sap confidently repeats the accusation to someone who does know better.
    4) This someone corrects the poor sap.
    5) Poor sap is now embarrassed.
    6) Poor sap panics and tries to save face by moving the goalposts, in some cases by inflating “lie”.

    Bonus panel: do this to the same poor sap a few times and you get learned epistemic helplessness. Aren’t consequences fun?

  49. Corey says:

    I had an interesting argument with my manager at a smallish company about this a decade or so ago.

    My contention was that managers have to lie in the course of their duties; e.g. you know a layoff is coming but have to coach employees as though none is, or a product is getting killed but you have to keep people working on it, etc.

    (Companies of nontrivial size get around this by keeping front-line managers in the dark about everything.)

    He reacted as though I’d called him a (pre-meaning-creep) Nazi.

    The big thing I learned from this: because of the bad connotations of “lie”, nobody wants to think of what they do as “lying”. They will euphemize or otherwise rationalize it, even if it meets a pretty narrow definition. Of course often lying is socially necessary (e.g. in this case, to keep secrets, or consider the category of “white lie”).

    • Randy M says:

      Of course often lying is socially necessary

      If the manager says “I can’t discuss lay-offs” do the employees assume that they are coming? It seems simply stating that you can’t discuss a topic now should often be able to take the place of the lie, and, as socially nice as the lie would be in the moment, the truth has less chance to harm your credibility when you have to announce the lay-offs the next day.

      • Matt M says:

        If the manager says “I can’t discuss lay-offs” do the employees assume that they are coming?

        It probably depends on past behavior. If the manager always answers questions about layoffs with “I can’t discuss that” then fine. But if the manager frequently says “No lay-offs this year!” every year, then suddenly in an economically down year he starts saying “I can’t discuss layoffs” employees will probably assume layoffs are either imminent, or at least being very strongly considered.

      • Nick says:

        Nearly every lie that folks will say is necessary isn’t. This is a really common rationalization. And white lies, for that matter, are only ‘necessary’ if the social good in question is absolute. But how many social goods are absolute?

        Even when folks try to justify lying on the consequences the analysis is often suspect, e.g. “Alice will get really hurt if I tell her her breath smells bad, so if she asks me I’d better lie.” Totally stupid thing to do. Be honest with her. Alice and everyone in her general vicinity will be better off if you are. Granted, this can happen with any sort of action up to murder, but lying is one that has proliferated in our day for just this reason while murder hasn’t.

        Of course, I’m not married, so my position here may be naive. 😉

        • Randy M says:

          Of course, I’m not married, so my position here may be naive. 😉

          I think you can tell the truth about how the effect of a dress on butt size while still showing concern for feelings. And then your honest compliments will be all the more meaningful.

          It helps to be honest even when it works against you.
          ex: “Did you get a chance to do the dishes?”
          “Yes, I had a chance, but neglected to actually do them.”

    • DinoNerd says:

      I’m with you on decribing this behaviour as lying, but when I see it, I will mistrust the person doing it on everything else they say and do. If I were to see it as required by their job, etc., I’d mistrust everyone in that job.

      I *might* be able to figure out what types of things they lie about, and what they do not, and teach myself not to expect a random manager to lie about everything, or at least everything where they might benefit from the untruth. But I’d probably require explicit coaching (e.g. from someone with normal ‘soft skills’) to find the boundaries. *If* there are boundaries; I’m not convinced that a person who knowingly tells untruths in one area won’t do it in every other area as well.

      FWIW, I do make an exception for “but you have to say this”. So when a senior manager at my employer left “voluntarily”, “to pursue other opportunities”, with great heaps of praise in both the internal and external announcements, I hoped the bastard – whose decisions I never liked – had in fact been pushed out because of those decisions, i.e “fired for cause”. And I didn’t tag the company CEO as a liar for announcing this departure in this way – instead, I hoped it meant that he had more sense than I’d previously believed.

      So your example might fall into the “have to say it” category.

      But frankly, coaching employees as if they aren’t a probable target of the coming layoff is just plain wrong. If there’s no point in them trying to fix their problems at their current job, and should begin job hunting ASAP, you are causing them actual harm by misleading them about this.

  50. Virriman says:

    Ever the contrarian, I seem to be naturally inclined to engage in the opposite behavior. Whenever I’m in a conversation where anyone is accused of lying, I insist that the accused is merely mistaken, or exaggerating, or my personal favorite – “telling stories”.

    When I was a child I had a stuffed rabbit named Happy and during play time I’d talk about how Happy went and ate vegetables in our neighbors’ gardens. Whenever my parents or siblings pointed out that was stealing, I’d insist that it wasn’t stealing; Happy was just “helping”. Now that I’ve grown up, I’m more likely to apply that kind of logic when defending unconventional politicians.

    In my book, nobody is a liar unless a refutation is literally hopping up and down making a racket right next to them at the exact moment they are speaking an untruth, and even then, they might earnestly believe the refutation is some kind of distortion..

  51. Corey says:

    “Socialism” is a good example – on the right it has long meant “left of Friedman” or just used as a snarl word, and on the left the “left of Friedman” definition is taking hold.

    (Present company excepted; the SSC commentariat would never be that imprecise)

    • MostlyCredibleHulk says:

      Pretty much all mainstream Republican politicians are left (at least economically, social left-right scale is excluded I presume) of Friedman. This means all Republicans consider themselves socialist? Or is “the right” does not include Republican politicians – and who does it then include? Does not seem to match the observations.

      • Corey says:

        True, Milton F is probably not a good example, and definitely imprecise. Maybe common usage is better modeled by “to the speaker’s left economically”, and that makes sense as the expansion is largely driven by the snarl-word usage.

    • “Socialism” is a good example – on the right it has long meant “left of Friedman”

      Which Friedman?

      I used to accuse my father of being a socialist. After all, he wanted teo important industries—rights enforcement and dispute resolution—to be run by the government.

      • Corey says:

        The one everyone knows 🙂

        As Hulk points out, my choice of Milton as an example is imprecise, but you provide a good example – if “socialist” means “not libertarian enough to privatize courts and law enforcement” then it covers approximately everybody, and we need a term for “wants state control of all means of production”.

  52. JPNunez says:

    While it’s true that saying everyone is an abuser is weakening the word abuser, saying that everyone abuses someone else at some point is still useful information.

  53. Paper Rat says:

    A few months ago, a friend confessed that she had abused her boyfriend. I was shocked, because this friend is one of the kindest and gentlest people I know. I probed for details.

    Wouldn’t you probe for details if the universally agreed definition (if such a thing was possible) of “abuse” was as narrow as you want it to be?

    Isn’t it how a normal conversation supposed to go? First a general theme or situation is described with broad terms (“I’m an abuser”, “he is a liar” etc.), then after a prompt by the listener, which basically says “I’m interested in this theme”, follows a more detailed explanation, that allows for much more nuance, but also takes way more time.

    Never in my life would I respond to a “he is a liar” line with “ok, so I won’t trust him in the future”. In most cases it would be “what did he lie about”, in some cases it would be “I don’t care about that person” (meaning whether he’s a liar or not is irrelevant to me). So the examples you provide, where someone makes a judgement based on one word description like “liar” or “abuser”, in my opinion, don’t show the problem with having “liar” or “abuser” mean a lot of different things. They show the problem with people making judgements of other people’s character based on one word descriptions.

  54. A mostly excellent post.

    But describing motivated reasoning as a “lesser sin” seems only half right.

    I think some of the disagreement that triggered this discussion came from the belief that stereotypical motivated reasoning causes roughly the same harm as lying, so that there’s nothing clearly lesser about it. I have sometimes been tempted to conflate motivated reasoning and lying, because they have pretty similar effects and root causes.

    Whereas you’re more focused on the idea that motivated reasoning implies very different predictions about the person’s behavior in other contexts, and about what would cause them to make more accurate claims in the future.

    So I’d like to have “lesser sin” replaced with more specific terms, that distinguish between something sort of consequentialist versus something like the virtue ethics meaning of lesser.

  55. DinoNerd says:

    I rarely see lie inflation – instead, I see lie deflation, except that it look to me like lieing about lieing 🙁

    I’m on the autistic spectrum. A lot of things people say around me appear to me to be both false, and known to the speakers to be false. This would include perhaps 90% of all advertisements, but also most other corporate “messaging” [do they use that word because they know it isn’t true?] both internally and externally directed.

    Some of it is conventional – there’s no expectation that non-naive observers believe what is being claimed. I’ll agree that isn’t actually lieing – if I recognize it. (In general, I have to be told about categories of conventional falsehood. If I’m not told explicitly, I start out “naive” [unless the claim is also impossible, which is common] and move on to disillusioned/feeling betrayed when I discover the untruth.)

    I’m not sure what I’d call the broad category of statements intended to motivate someone else’s behaviour, that actually have no truth value whatsoever. This would be trying to convince people to buy product X because cool people buy X (which might or might not be true) and thus buying X will make the purchaser cool (conceivable, because of group and status signalling effects – but not at all intuitive or natural to me). Or it would be using standard perceptual distortions to get people to do what you want – presenting a 2% failure vs 98% success rate. Or just showing pictures of actors pretending to some social affiliation in association with the product, policy, etc.

    But basically, my expectation of the behaviour of whole classes of people (executives, politicians, salespeople) is formed by observing them producing some statements that they must know to be falsehoods, and assessing both their honesty and (often) the rest of their characters accordingly.

    I try to account for my own social skill limitations – maybe some of these falsehoods wouldn’t be believed by any normal person, and the speakers know it, and hence are producing metaphor, or conventional behaviour, or fiction, rather than lieing. But I probably err on the side of seeing as lies falsehoods that aren’t in fact intended to be believed, and assessing the speakers’ character according to what I see.

    • Corey says:

      I think a lot of claims, corporate and advertising included, are *technically* true if you squint right. The amount of mis-leading-ness people tolerate before calling it a lie (vs., say, “spin”) varies.

      My favorite example of spin: the claim that the U-in-a-circle on food products is the mark of having paid a “Jewish tax” to avoid a protection racket, and Jews will refuse to buy your products unless you pay up. (Snopes)

      It’s a kosher certification, so this is technically true. Certifying things costs money, and Jews who keep kosher will probably not buy your product if it’s not certified; it might not be obvious whether it has e.g. insect-derived ingredients.

    • MostlyCredibleHulk says:

      A lot of things people say around me appear to me to be both false, and known to the speakers to be false.

      A good theory would be in that case to consider that words may actually be intended to mean not the same thing as they appear to mean. As a Freudian would say, sometimes people say one thing but mean their mother. More seriously, human communication is almost always multi-stream, and people on the autistic spectrum sometimes miss one or several of the streams. Eric Berne wrote several books on transactional analysis deconstructing those streams – I am not sure how scientifically sound it is but his main ideas seem pretty plausible to me.

      Yes, that means that some statements – in advertising, probably most of the statements – are devoid of “truth value” or even have negative one (i.e. state facts that aren’t true and can not be – like promise that if you switched to insurance company X, you’d always save tons of money – on a free market, this is pretty much impossible to sustain). That’s not some omission or mistake, this is by design – their message is different stream than factual statements.

      And that’s intentional because behavior of most people is not ruled by facts – or at least not by facts alone. Most people just don’t have enough facts and enough training and enough discipline to make decisions without resorting to heuristics. Often it’s plain impossible – if you have 1000 brands of shoes, how do you know which one would provide you with optimal experience? Solid factual approach would take you years of research, probably. Heuristic would take you 2 seconds. No wonder people use the latter. And ads influence heuristics. Which sometimes makes them very bad heuristics – e.g. you really shouldn’t drink that sugar/caffeine water, it will poison you, ruin your metabolism and may ultimately kill you – but that’s how it works. So whether these things are “believed” by “normal person” is a very tricky question – the fact is that ads influence heuristics, but how, which ones and how much – that’s tough to say, probably depends a lot on a person and circumstances.

  56. ajfirecracker says:

    Are you just assuming that concepts cannot refer to something that could (at least in theory) refer to something objective? We have a lot of words that seem to be somewhat useful despite being universally applicable or inapplicable, such as “mortal” and “immortal”.

    I think the ideal percentage of the population for the word “murderer” to describe depends on what proportion of the population has killed another human being, not on something about the selectivity of the word “murderer”. Yes, this means you have to argue about whether a policeman shooting a person he thinks may have a gun constitutes “murder”, but you have to do that regardless of whether you use the word “murder” or some complex circumlocution that avoids this word.

  57. MostlyCredibleHulk says:

    this definition was so broad that everyone has committed abuse at some point

    Maybe that’s the point? There are whole religions built on the principle that everybody is guilty beyond redemption, but if you do this and that and listen to the right people and obey them, then maybe, just maybe, you, a miserable worm, have a chance. But if not…
    Looks like pretty good business, so no reason why only people talking about deities can get into it and others can’t.

  58. Jiro says:

    I think that people are using the term lie to mean “an intentionally or recklessly false statement”. And this is quite reasonable.

    Yes, it’s always possible that using the word like that can make it expand to cover more things or be used as a weapon, but we have a recklessness standard in the law and in a lot of other places. It’s clearly not an unworkable concept.

    Also, suppose I tell you that some random vegetable extract I’m selling will cure cancer. Have I lied? I just picked a substance and I have no reason to believe it does cure cancer. But I haven’t lied, in the sense of “I know it doesn’t cure cancer but I said it does”. Rather, I have no good reason to believe that it cures cancer. I didn’t actually do any research that said it doesn’t, so when I claim that it does I’m not literally lying. I just made a reckless claim.

    I think most people use “lie” that way and if you don’t want to sound like an evil robot, you need to understand how they talk. Calling the singularity a scam and a lie doesn’t mean “they know there is no singularity but said there is one”, it means “they are recklessly claiming there will be a singularity without proper evidence”.

    (Note, by the way, that the expanding definition of “abuser” is not this.)

  59. Erl137 says:

    In his poem Dulce et Decorum, Wilfred Owen refers to the titular sentiment, “dulce et decorum est pro patria mori”, (“it is sweet and fitting to die for one’s country”) as “the old lie”.

    Does Owen abuse the language? Does he specifically charge Horace with personal dishonesty?

    I don’t think so. And I don’t think Owen is using some special license only available to poets. Rather, I think this use of “lie”—a common, motivated falsehood—is well within the common understanding of the word. And I think a definition of lie that includes such utterances and permits such accusations carves reality at its joints, as the rationalists say.

  60. yildo says:

    I’m unsympathetic to the argument if only because I find that not enough rather than too many things are labeled lies. The news media is unwilling to call anything a lie because one can never truly know what is in the mind of someone else. Anyone making a false statement after having been told repeatedly it is a false statement, or making a false statement after having previously made the opposite true statement is never called a liar in news media. 0% of people are called liars in conventional news reporting, which to me is unuseful.

  61. tayfie says:

    I agree cheapening powerful words is bad, but disagree with the two main examples.

    There needs to be more distinction between an action and a character trait. Calling someone an “abuser” is more complicated than meaning “one who has abused”. Ditto for “liar”. It’s about the extreme and habitual nature of the action especially while knowing the consequences.

    Notice your friend did not call herself an abuser. She said she had abused her boyfriend. The Less Wrong post did not call anyone a liar. Jessicata accused them of telling a lie. I view both of these things as correct uses of language in describing the consequences of actions. Did your friend distress her partner by using emotional manipulation to gain favors? That is abusive, but it doesn’t make her an abuser. Did an AI organizations make claims about AI progress that are untrue? That is dishonest, but it doesn’t make them liars.

    This is important because, while we need “abuser” and “liar” to distinguish people, we need “abuse” and “lie” to distinguish actions. If you are overly strict in definitions where the only way to say someone lied is to call them a liar, it becomes impossible to call out bad behavior without passing judgement on a person as a whole.

  62. Eponymous says:

    While I agree with your conclusion, I think your argument fails in that you are conflating issues of intent and severity. I think this is a very serious error.

    The standard usage of “lie” implies intentional deceit. It does not depend on how egregious the deceit is (hence concepts like “little white lie”). I think that intent matters, and so I like keeping the requirement of intent to use the word “lie”.

    Most of your argument seems to hinge on the idea that “everyone” engages in arguments rooted in bias and motivated reasoning, and thus expanding the notion of “lie” to cover these cases will render it useless. I think this is a weak argument, and beside the point. First, as I observed above, we already include fairly innocuous cases under the word “lie”, without causing the problems you cite. But more importantly, not all cases of biased and motivated reasoning are the same — yes, we probably all engage in them to some degree, but some are much worse than others. And we need language to condemn extreme and egregious cases.

    Consider the argument you are responding to, namely: “Rationalists are engaging in egregiously motivated and biased reasoning to argue for a near singularity, which is lying”. I think that your response is essentially reversing the worst argument in the world — essentially you’re saying, “Actually, technically this is not lying, and if we expand the notion of lying to include cases of biased and motivated reasoning, where does that leave us?” The implication is that rationalists are not doing anything so bad — since “lying” has a very negative connotation.

    But note that you haven’t actually addressed the argument at all! At best you’re arguing for keeping the notion of consent within the concept of “lie” — but then you use this to argue that in fact the arguments advanced by rationalists are (in some sense) “not so bad” — even if they are an extreme case of motivated and biased reasoning!

    Your example actually makes this point for me. Someone who cries in order to emotionally manipulate their partner into doing what they want is engaging in something akin to abuse, whether or not you want to use that word in particular. My judgement on it depends on the severity of the case, not whether it technically is abuse (again, reversing the worst argument in the world! “see, technically this isn’t ‘abuse’, so…”). If this person engages in this behavior routinely, and uses it to get their way to an unreasonable degree, then I would judge it quite harshly — even if it was unintentional!

    Though I agree intent matters, severity matters too. There is a notion of what a reasonable person would/should do. If someone says something that is factually incorrect because they did not exercise an appropriate standard of epistemic caution, this is bad — whether or not you want to use the word “lie” or save that only for intentional deceit. And someone who routinely does this is untrustworthy, whether or not you want to call them a “liar”. Likewise, if someone routinely fails to control their emotions in a way that emotionally manipulates others, this is bad, whether or not you want to use the term “abuse” (I would at a minimum clarify to “emotional abuse”). This is because we expect certain standards of emotional self-control from people, just as we expect certain standards of epistemic awareness.

    I apologize for the long comment, and also apologize if someone made the same points above. It is unusual for Scott to make such a poor argument, and I think it’s important to correct it, plus point out the general pattern of the flaw (as I see it).

  63. embrodski says:

    True story: not long before I stopped dating this particular person, she told me (infront of a gather of friends) that a mutual friend was a rapist. I was appalled, but also confused. She wanted to go on tour with his band, and went to his Christmas parties, and generally acted like a friend of his. I got clarification. Turns out he would use his position as the band’s frontman to have one-night trysts with young (but legal aged) fans who wanted to sleep with him while he was touring.
    I no longer felt safe knowing someone who would tell all my friends that I’m a rapist for basically being a standard young male.

  64. Garrett says:

    I spent a bunch of time trying to find a good definition for “lie” during the “Bush lied. People died.” phase of political discourse.

    The best definition that I found was something like:
    1) Believing one thing
    2) attempting to convince someone of the contrary
    3) without their consent

    1) Exists because in many cases it’s difficult to determine the ultimate truth value of something. Likewise, telling someone what you believe to be true but being incorrect unintentionally should not have the same moral connotations.
    2) This gets into (1). If you are trying to convince someone of what you believe, you aren’t lying. You might be mistaken.
    3) There are times when we tell each other un-truths consentually, such as in works of fiction. Yet producing such a quality work of fiction for the enjoyment of others when they know that is what they are getting is highly regarded.

    There’s also a separate issue which involves the idea of “lies to children” where a simplification of what you truly believe is provided because it provides sufficient provision to answer a question without having to go into deep subject matter.

  65. teageegeepea says:

    The concept you’re discussing is known to other psychiatrists as “concept creep“. The Atlantic did a story on it a couple of years ago.

  66. dark orchid says:

    I realize this is a kind of long post arguing against a weird thing that not many people are doing.

    I know it’s been hinted at in the comments already, but isn’t this one of the main battle grounds in the culture war?

    My own take on the culture-war side is that both the 5% and 95% categories of ‘abuse’ that you mentioned convey useful information, but completely different kinds of information. The same with the two different categories labelled racism, sexism etc.

  67. benquo says:

    It seems like there’s substantial tension between the supposed reluctance described here to apply the term “lie” to someone whose conscious intent you don’t know, and the definition actually used in In Favor of Niceness, Community, and Civilization.

  68. meh says:

    So probably some people will continue to use the new definition while still holding the connotations of the old definition.

    This post can also be read as a reason we shouldn’t redefine a word like ‘truth’.

  69. devils_rights_advocate says:

    In your view, is there any place for the idea of someone “lying to themselves”? This seems to me like a useful concept, a real thing that people are trying to point at when they say someone is “lying”, even though that person may just be using highly-motivated reasoning. There are occasions when I have said people are lying about something that could perhaps more aptly be called a result of unconscious bias, but only when I’d also be confident enough to claim that said people are lying to themselves as well.

    Maybe that’s worth abandoning, but I think the idea is something more than just claiming “you’re biased”. It seems connected to the idea of “belief in belief”. If someone clearly doesn’t actually believe in something, if all their expectations and excuses are set up in advance as if they don’t believe something, is it a lie when they say they believe it?

    I know people who heavily exaggerate stories, for example. And I think part of them believes it while they do it, while another part of them must surely know that what they’re saying is not true, and they’ve decided to ignore the cognitive dissonance. Sometimes I can *tell* when people are exaggerating parts of a story to spin a more self-serving narrative. Can I call them liars?

  70. jbslattery says:

    To state the obvious, this is exactly what has happened with the word “racist.” Now there’s a huge chunk of the population that thinks that every accusation of racism is just an attempt to smear someone, and who are more predisposed to try to give the benefit of the doubt to people who are accused of “being racist,” since that can include all sorts of behaviors and beliefs outside of the traditional “person who discriminates against people based on their skin color” definition.