Interesting post. Are your friends Jasper fforde fans? (Wednesday...)
Atheism tends to be treated as an open-and-shut case here and in other intellectually sophisticated venues, but is that fair? What about Wednesday? What would have to happen to her to get her to give up those beliefs? Well, for starters, she’d have to dramatically change her opinion of her family. Her parents care enough about honesty that they are already planning not to deceive her about Santa Claus—should she believe that they’re liars?
That would be very uncharitable of her. She should continue loving them while evaluating their beliefs for herself, and criticize their beliefs to the extent that that is a good idea under the circs. That’s what we all do with loved ones whom we consider wrong, isn’t it? (Are other LWers ruining Christmas dinners with atheist ranting?)
Note, the whole reason we are having this conversation is because some beliefs are badges of identity, especially religious beliefs, so criticizing them is similar in emotional impact to saying “your identity sucks.” This is unfortunate, but I do wish to gently point out that it is not the fault either of LessWrong or of Atheism that the rhetorical temperature is so high when such beliefs are questioned.
Is it really essential that, as a community, we exclude or dismiss or reflexively criticize theists who are good at partitioning, who like and are good at rational reasoning in every other sphere—and who just have higher priorities than being right?
I think the criticism should be done tactfully, and on their terms. But it would disrespectful of them as fellow epistemic agents to refuse to criticize their beliefs, where relevant. A person who has experienced such kid gloves “respect” will tell you that it is very unpleasant to be condescended to.
I have priorities that I’d probably put ahead of being right, too; I’m just not in a position where I really have to choose between “keeping my friends and being right”, “feeling at home and being right”, “eating this week and being right”. That’s my luck, not my cleverness, at work.
Wait: “being right” != “always blurting out what you believe.” As a quote I saw recently has it, “you keep two books, not no books.”
What does this mean? (I had a look on the web and found only what I take to be your source—a blog entry at “The Last Psychiatrist”—and a few other references to that. Its meaning was no clearer to me in that context. It looks as if TLP is also quoting but if so there’s no indication of where from.)
“Books” refers to accounting (e.g. the records you would use to keep track of your business transactions). A common euphemism for manipulating your records, e.g. to lie to the IRS for tax purposes, is “cooking the books.” “Keep two books” means “have two sets of records of your business transactions, one which is the actual set and one which you use to lie to people.” “Keep no books” means “don’t keep track of your business transactions.”
The metaphor, as I interpret it, is that you should maintain a distinction between what you believe and what you signal believing (or else you run the risk of losing track of both, mixing them up, etc.).
Yes, I’d thought it was probably “books” in the accounting sense. However, I was either too dim or too ethical or both for the idea “a person or company might deliberately not keep accounts, in an attempt to make fraud easier” to have occurred to me. Thank you.
Really? I interpret this to mean “signal believing exactly the things I actually believe,” which strikes me as a terrible idea in general. If you’re determined to believe true things, some of the things you believe will end up being things you can’t say, and saying them would not be instrumentally a good idea. Michael Vassar once pointed out that a commitment to saying what you believe disincentivizes believing things you can’t say, which is, y’know, bad, and his advice was that rationalists should become more comfortable with lying.
(This is of course distinct from “signal that I signal believing exactly the things I believe,” which is a great idea. If that’s what you were doing, then great! Carry on.)
I am determined to believe true things. I don’t believe there’s anything I can’t say (though I have a little trouble with “heteroskedasticity”) but indeed there are things it’s usually better not to say. So I usually avoid saying them. If this requires a load of extra bookkeeping then I’ve failed to notice so far. Lying requires that second book; not saying things when saying them would have bad consequences, even if you’d otherwise feel like saying them doesn’t appear to.
I don’t make a strong distinction between lies in the colloquial sense and lies by omission. “Not saying things when saying them would have bad consequences” still requires that you keep track of what things it would have bad consequences to say.
I find extremely few occasions when there’s any need to be actually deceptive by not saying things. For the rest, no keeping track is required; a policy of the form “tell the truth, but don’t say things that will cause too much trouble” suffices.
A difference between this and actually lying (in which category I include “lying by omission”) is that in order to lie credibly and not get caught, you need to remember just what lies you’ve told to whom (in the best case, I suppose you can get by with just two “books”, keeping track of the truth for yourself and a single set of lies for everyone else) and make sure it’s all coherent. But any two subsets of the truth are consistent with one another.
Fair. I think I shouldn’t have used the word “lie” because it seems to have primed you into a direction other than the one I was headed, but I don’t know a good substitute. One kind of lie I have in mind is things like explaining Newtonian physics to physics students instead of quantum mechanics or relativity. The implicit claim that Newtonian physics accurately models reality is in some sense a lie, but it’s a good enough approximation for many purposes and also useful for understanding what comes after.
An analogous kind of lie in interpersonal relationships is the following. Suppose I’m on OKCupid and it asks “are you a feminist?” Before a few months ago, my answer would have been an unhesitating “yes.” Now I’m not so sure. I’m trying to keep my identity small, and “feminist” is a term that comes with a lot of baggage. I don’t know if I want that baggage in my identity. I’d at least like to taboo “feminist” by default.
Nevertheless, very little has changed about how I actually treat women. I want people on OKCupid to know that. They’ll have a better understanding of me, or more precisely how I treat women, if I answer “yes” to this question than they will if I answer “no.”
One kind of lie I have in mind is things like explaining Newtonian physics to physics students instead of quantum mechanics or relativity.
I have heard this sort of thing referred to as “lies-to-children”. Your average junior school is full of them. (All numbers are on the number line! The atom is like a very tiny solar system!)
I’ve been thinking that “keep your identity emergent” or “keep your identity honest” might be better advice than “keep your identity small”. That is, people should let their identity emerge as a consequence of their individual object-level views, instead of deriving their individual object-level views from their identity. That reversal of causation seems to me the problem with identity, not identity in itself. So instead of deleting almost all of my identity (and how would I know which little bits to keep?), I should figure out my object-level beliefs first, and then summarize them as aspects of my identity.
Using feminism as an example, if I notice one day that I’m identifying as a feminist, I stop and ask myself about each of the individual object-level issues that feminism touches upon. If my views on those object-level issues really & truly align with those connoted by the “feminism” label, I might as well identify as a feminist; the identifier arises organically from the beliefs. If my views don’t align with it, then I should stop identifying as a feminist. (My views could of course change over time, in which case I adopt/drop the identification accordingly.)
Your comment shows one advantage to this approach: it’s less liable to mislead people than simply keeping one’s identity “small”. If I agree with the X-ist cluster of beliefs and behave accordingly, other people might well have a more accurate model of me if I self-identify as an X-ist than if I stoutly refuse to identify as such. (Of course, if I want to taboo “X-ism” in a conversation, it can make sense to avoid identifying as an X-ist. But doing so indiscriminately can increase confusion & exasperation rather than reduce them.)
people should let their identity emerge as a consequence of their individual object-level views, instead of deriving their individual object-level views from their identity. That reversal of causation seems to me the problem with identity, not identity in itself.
I’d expect it to be extraordinarily hard to keep the causation one-way, even if you’re trying hard and are aware of all the consequences. In order for something to be promoted to conscious attention, it has to make it through a set of perceptual filters which include some coherence checks with your existing identity: it’s quite possible to believe earnestly that you’re taking into account all the data even as you silently drop half of it from your consideration.
To make matters worse, I’d also expect it to be extraordinarily hard to keep identity criteria stable. For example, the kids in the famous Robber’s Cave experiment (Sherif et al., 1954) readily generated stereotypes for themselves, all to support a more or less fabricated image of a distinct identity group; and this certainly isn’t limited to the laboratory, as the behavior of whatever political group you like the least should demonstrate! The lesson seems to be that identities aren’t static classification functions; justifications and superstitions accrete around them like nacre in the guts of an irritated oyster, growing and feeding back into an increasingly tangled complex of beliefs.
I suspect that if identity is as sticky & accretive as you suggest, trying to purge my identity could prove at least as hard as wearing my identity loosely on my beliefs. But that is just a guess on my part — I ought to chew on what you’ve said for a bit.
One of the benefits I’ve found from keeping my identity small doesn’t seem to be reducible to keeping my object-level views honest. Namely, I’ve recently identified areas of my life in which my identity was preventing me from trying new things, e.g. I thought of myself as the kind of person who didn’t care about nutrition or exercise. I wasn’t mistaken about any property of the world but I was supplying myself with excuses for not expanding my comfort zone. (Edit: Academician describes this happening to him in this post which I think is a useful follow-up to Keep Your Identity Small.)
With the first sort of kinda-lie, there’s again little keeping-track needed. You have some particular not-quite-right theory that you’re putting forward; it’s basically coherent and matches the world reasonably well, because otherwise you wouldn’t be using it. And even if you slip up and mention some quantum or relativistic stuff to your students, no serious harm is done.
With the second sort, surely it’s universally understood that all you’re saying when you answer this sort of question is that “yes” is a less misleading answer than “no”. So again I don’t see any particular need for keeping-track on this account. If I were making an OKCupid profile and had to answer that question, I too would answer yes. If asked in a context that allowed for a more detailed answer, I would give one. No lying or other deception required. (Of course that might discourage possible partners who don’t like detailed answers, but that’s a feature, not a bug.)
I have encountered very few situations in interpersonal relationships where deliberate deception is required.
For the avoidance of doubt, I am not claiming that I absolutely never lie. But not lying is my goal, I seldom deviate from it, I strongly suspect that most such deviations I make are actually not in my best interests, and I have not found it necessary to keep track of anything much about The Lies I Tell Others.
You may of course believe me or not, as you please :-).
Interesting post. Are your friends Jasper fforde fans? (Wednesday...)
That would be very uncharitable of her. She should continue loving them while evaluating their beliefs for herself, and criticize their beliefs to the extent that that is a good idea under the circs. That’s what we all do with loved ones whom we consider wrong, isn’t it? (Are other LWers ruining Christmas dinners with atheist ranting?)
Note, the whole reason we are having this conversation is because some beliefs are badges of identity, especially religious beliefs, so criticizing them is similar in emotional impact to saying “your identity sucks.” This is unfortunate, but I do wish to gently point out that it is not the fault either of LessWrong or of Atheism that the rhetorical temperature is so high when such beliefs are questioned.
I think the criticism should be done tactfully, and on their terms. But it would disrespectful of them as fellow epistemic agents to refuse to criticize their beliefs, where relevant. A person who has experienced such kid gloves “respect” will tell you that it is very unpleasant to be condescended to.
Wait: “being right” != “always blurting out what you believe.” As a quote I saw recently has it, “you keep two books, not no books.”
What does this mean? (I had a look on the web and found only what I take to be your source—a blog entry at “The Last Psychiatrist”—and a few other references to that. Its meaning was no clearer to me in that context. It looks as if TLP is also quoting but if so there’s no indication of where from.)
“Books” refers to accounting (e.g. the records you would use to keep track of your business transactions). A common euphemism for manipulating your records, e.g. to lie to the IRS for tax purposes, is “cooking the books.” “Keep two books” means “have two sets of records of your business transactions, one which is the actual set and one which you use to lie to people.” “Keep no books” means “don’t keep track of your business transactions.”
The metaphor, as I interpret it, is that you should maintain a distinction between what you believe and what you signal believing (or else you run the risk of losing track of both, mixing them up, etc.).
Yes, I’d thought it was probably “books” in the accounting sense. However, I was either too dim or too ethical or both for the idea “a person or company might deliberately not keep accounts, in an attempt to make fraud easier” to have occurred to me. Thank you.
(Personally, I prefer to keep exactly one book.)
Really? I interpret this to mean “signal believing exactly the things I actually believe,” which strikes me as a terrible idea in general. If you’re determined to believe true things, some of the things you believe will end up being things you can’t say, and saying them would not be instrumentally a good idea. Michael Vassar once pointed out that a commitment to saying what you believe disincentivizes believing things you can’t say, which is, y’know, bad, and his advice was that rationalists should become more comfortable with lying.
(This is of course distinct from “signal that I signal believing exactly the things I believe,” which is a great idea. If that’s what you were doing, then great! Carry on.)
I am determined to believe true things. I don’t believe there’s anything I can’t say (though I have a little trouble with “heteroskedasticity”) but indeed there are things it’s usually better not to say. So I usually avoid saying them. If this requires a load of extra bookkeeping then I’ve failed to notice so far. Lying requires that second book; not saying things when saying them would have bad consequences, even if you’d otherwise feel like saying them doesn’t appear to.
I don’t make a strong distinction between lies in the colloquial sense and lies by omission. “Not saying things when saying them would have bad consequences” still requires that you keep track of what things it would have bad consequences to say.
I find extremely few occasions when there’s any need to be actually deceptive by not saying things. For the rest, no keeping track is required; a policy of the form “tell the truth, but don’t say things that will cause too much trouble” suffices.
A difference between this and actually lying (in which category I include “lying by omission”) is that in order to lie credibly and not get caught, you need to remember just what lies you’ve told to whom (in the best case, I suppose you can get by with just two “books”, keeping track of the truth for yourself and a single set of lies for everyone else) and make sure it’s all coherent. But any two subsets of the truth are consistent with one another.
Fair. I think I shouldn’t have used the word “lie” because it seems to have primed you into a direction other than the one I was headed, but I don’t know a good substitute. One kind of lie I have in mind is things like explaining Newtonian physics to physics students instead of quantum mechanics or relativity. The implicit claim that Newtonian physics accurately models reality is in some sense a lie, but it’s a good enough approximation for many purposes and also useful for understanding what comes after.
An analogous kind of lie in interpersonal relationships is the following. Suppose I’m on OKCupid and it asks “are you a feminist?” Before a few months ago, my answer would have been an unhesitating “yes.” Now I’m not so sure. I’m trying to keep my identity small, and “feminist” is a term that comes with a lot of baggage. I don’t know if I want that baggage in my identity. I’d at least like to taboo “feminist” by default.
Nevertheless, very little has changed about how I actually treat women. I want people on OKCupid to know that. They’ll have a better understanding of me, or more precisely how I treat women, if I answer “yes” to this question than they will if I answer “no.”
I have heard this sort of thing referred to as “lies-to-children”. Your average junior school is full of them. (All numbers are on the number line! The atom is like a very tiny solar system!)
Yes, so I guess what I’m saying is that I also don’t make a strong distinction between children and adults.
I’ve been thinking that “keep your identity emergent” or “keep your identity honest” might be better advice than “keep your identity small”. That is, people should let their identity emerge as a consequence of their individual object-level views, instead of deriving their individual object-level views from their identity. That reversal of causation seems to me the problem with identity, not identity in itself. So instead of deleting almost all of my identity (and how would I know which little bits to keep?), I should figure out my object-level beliefs first, and then summarize them as aspects of my identity.
Using feminism as an example, if I notice one day that I’m identifying as a feminist, I stop and ask myself about each of the individual object-level issues that feminism touches upon. If my views on those object-level issues really & truly align with those connoted by the “feminism” label, I might as well identify as a feminist; the identifier arises organically from the beliefs. If my views don’t align with it, then I should stop identifying as a feminist. (My views could of course change over time, in which case I adopt/drop the identification accordingly.)
Your comment shows one advantage to this approach: it’s less liable to mislead people than simply keeping one’s identity “small”. If I agree with the X-ist cluster of beliefs and behave accordingly, other people might well have a more accurate model of me if I self-identify as an X-ist than if I stoutly refuse to identify as such. (Of course, if I want to taboo “X-ism” in a conversation, it can make sense to avoid identifying as an X-ist. But doing so indiscriminately can increase confusion & exasperation rather than reduce them.)
I’d expect it to be extraordinarily hard to keep the causation one-way, even if you’re trying hard and are aware of all the consequences. In order for something to be promoted to conscious attention, it has to make it through a set of perceptual filters which include some coherence checks with your existing identity: it’s quite possible to believe earnestly that you’re taking into account all the data even as you silently drop half of it from your consideration.
To make matters worse, I’d also expect it to be extraordinarily hard to keep identity criteria stable. For example, the kids in the famous Robber’s Cave experiment (Sherif et al., 1954) readily generated stereotypes for themselves, all to support a more or less fabricated image of a distinct identity group; and this certainly isn’t limited to the laboratory, as the behavior of whatever political group you like the least should demonstrate! The lesson seems to be that identities aren’t static classification functions; justifications and superstitions accrete around them like nacre in the guts of an irritated oyster, growing and feeding back into an increasingly tangled complex of beliefs.
You and Qiaochu_Yuan raise good points.
I suspect that if identity is as sticky & accretive as you suggest, trying to purge my identity could prove at least as hard as wearing my identity loosely on my beliefs. But that is just a guess on my part — I ought to chew on what you’ve said for a bit.
One of the benefits I’ve found from keeping my identity small doesn’t seem to be reducible to keeping my object-level views honest. Namely, I’ve recently identified areas of my life in which my identity was preventing me from trying new things, e.g. I thought of myself as the kind of person who didn’t care about nutrition or exercise. I wasn’t mistaken about any property of the world but I was supplying myself with excuses for not expanding my comfort zone. (Edit: Academician describes this happening to him in this post which I think is a useful follow-up to Keep Your Identity Small.)
With the first sort of kinda-lie, there’s again little keeping-track needed. You have some particular not-quite-right theory that you’re putting forward; it’s basically coherent and matches the world reasonably well, because otherwise you wouldn’t be using it. And even if you slip up and mention some quantum or relativistic stuff to your students, no serious harm is done.
With the second sort, surely it’s universally understood that all you’re saying when you answer this sort of question is that “yes” is a less misleading answer than “no”. So again I don’t see any particular need for keeping-track on this account. If I were making an OKCupid profile and had to answer that question, I too would answer yes. If asked in a context that allowed for a more detailed answer, I would give one. No lying or other deception required. (Of course that might discourage possible partners who don’t like detailed answers, but that’s a feature, not a bug.)
I have encountered very few situations in interpersonal relationships where deliberate deception is required.
For the avoidance of doubt, I am not claiming that I absolutely never lie. But not lying is my goal, I seldom deviate from it, I strongly suspect that most such deviations I make are actually not in my best interests, and I have not found it necessary to keep track of anything much about The Lies I Tell Others.
You may of course believe me or not, as you please :-).
That doesn’t feel particularly tongue-twisty to me, for a word of that length. Try “red lorry, yellow lorry”. :-)
Oh yes, there are other worse things of similar length. I confess that I chose that example partly because I like the word.
I think homoskedasticity has more intriguing possibilities as a desired-for attribute that begins with ‘homo.’
Aaaarghdvghhv