In his famous essay Keep Your Identity Small, Paul Graham argues that the reason arguments about religion, politics, and programming languages get so heated is that they’re parts of a person’s identity. It isn’t just about the merits of this or that policy or this or that programming language, it’s about who you are as a person. Therefore, he suggests, to be less stupid, avoid identifying as anything: be “a person who uses Lisp sometimes” rather than “a Lisp programmer.”
The problem is that, as Sarah Constantin writes:
Dasein [identity] is what you do when you assert what it means to be human, what it means to be you, what it means to be a member of your community. Dasein is self-definition. And, in particular, self-definition with respect to a social context. Where do I fit in society? Who is my tribe? Who am I relative to other people? What’s my type?
These are really basic psychological needs for a lot of people. As Sarah writes later, identity is about self-expression, self-definition, self-assertion. Of course, not everyone shares these needs: not everyone needs to know who they are and where they fit into the social fabric. There are a lot of different kinds of minds. But having a self-model, a sense of yourself, is a legitimate need for many people.
Of course, if you have that need, and you try to eliminate your sense of identity, your brain isn’t just going to let you go about not having a basic psychological need filled. Probably you’re going to have an identity anyway, you’re just not going to be consciously aware that you have it. And that means you can’t take any steps to compensate for the ways your identity might be biasing you. You can’t take a step back and say “hey, I know I get emotional about programming languages because I identify as a Lisp programmer, so I should take a step back and deliberately notice the good parts of this other heathen programming language.” Instead, you just say “Lisp is the best language for all purposes! You should believe me because I don’t have an identity as a Lisp programmer so I’m objective!”
It is always dangerous to have a self-image as being objective, because all too often one acquires the self-image before one acquires the objectivity.
Of course, the criticisms Graham says still apply: if people’s identities are getting wrapped up in an argument, it’s really hard to have the argument in a remotely civil and enlightening way. So how do we minimize the harm of having an identity while still getting the benefits?
Identify as things that aren’t based on empirical claims. If you identify as a libertarian, then you’re signing on to a bunch of empirical claims, like “the minimum wage is bad for poor people” and “environmental regulations cause more harm than good” and “police officers abuse people’s civil liberties.” If it turns out that actually police officers are totally nice to black people everywhere, this might be a threat to your identity as a libertarian, and then you’re going to have flamewars about the subject.
Compare this to an identity as a baker. I am really having a hard time thinking of empirical claims related to being a baker. “Bread is yummy”? “It is a good use of one’s time to knead things”? “Being paleo is maybe a bad idea and maybe a reason for creativity in the use of gluten-free flours”? But “baker” serves all the same purposes of self-definition (you are a person who can make really delicious cakes) and understanding oneself within a social context (you are a person who feeds people such as Ozy delicious cake). A baker self-identity is free epistemically.
Identify as things you want to preserve. Many effective altruists are concerned about values drift. They start out valuing altruism, become an investment banker so they can donate more money, and wind up valuing $300 sushi dinners and designer ties. Fortunately, identity tends to keep things fixed. That’s awful for facts– you want to be convinced by new evidence– but great for values– you want to keep your own values and not other people’s. Therefore, it can be useful to identify as an altruist, or an artist, or a family man, or an ambitious person, or a hedonist, or some other value you wish to preserve.
There’s a failure mode here. It’s very easy to have an identity as a good person, and then when people threaten your identity as a good person you lash out at them in the same way that Lisp programmers lash out when people suggest Lisp isn’t the best programming language. This can lead to results even more disastrous than bad arguments about politics. So if you’re going to identify as a good person, it’s really really important that your definition of “good person” isn’t “a person who always does the right thing” but “a person who notices their mistakes and tries to fix them.” A similar concern applies to artists, family men, and so on: it’s important to make sure that your definition of artist is not “a person who makes great art” but “a person who values their art and puts a lot of effort into it.”
Keep your identity large. Don’t just identify as a Lisp programmer. Identify as a poet, a mother, a Communist, a Jew, a weightlifter, an altruist, and a Lisp programmer.
If the only identity you have is Lisp programmer, then it’s terrifying to think about not being a Lisp programmer. How will you know who you are? How can you relate to other people? On the other hand, if you have a lot of identities, you have something to fall back on. Even if you have to switch programming languages, your children still exist, you can still benchpress twice your body weight, you still light Shabbat candles, and you still spend an absurd amount of time explaining that your preferred economic system wouldn’t have any gulags at all not even a little bit and capitalism has killed more people than socialism anyway. Your sense of self in a social context remains secure. You can admit the flaws in Lisp or that prices are an elegant means of solving coordination problems, without threatening who you are as a person– because who you are as a person isn’t grounded in just one thing.
The baker approach doesn’t really avoid the problem. What if baking stresses you out a lot for little reward? What if baking causes a lot of harm to others? Or what if you can’t be a professional baker anymore because you’re put out of work by technological change? If that’s part of your identity, you’re likely to be resistant to change. If you identify with beliefs, they might be false, but if you identify with activity, it might be something you can’t or shouldn’t do. It also ties you to the community of bakers – for example, if someone says that they’re disproportionately likely to commit violent crimes, you’re likely to feel defensive and deny it, even if the accusation is true.
LikeLiked by 6 people
I think that the example would be more clear with an extra ‘n’: bankers instead of bakers.
Also, to expand on your point, professions tend to come with a culture that goes beyond the mere mechanics of the work. Bankers of today don’t have the same culture of the bankers of the past. My beef with the profession is not with the basic mechanics of lending, but this culture.
LikeLiked by 1 person
Lets take for example the person who is a poet, a mother, a Communist, a Jew, a weightlifter, an altruist, and a Lisp programmer.
This person will need to be familiar with a large body of literature, tend to children, be familiar with 1900s cultural history, eat kosher, exercise regularly, live modestly, and only work at the 0.01% of companies that use Lisp. If the person fails at any of these, then they have failed at their identity.
Identities have benefits and drawbacks. The benefit is that they make the person with the identity feel good. The drawbacks are mostly felt by others who do not abide by the identities as the drawbacks are seen as benefits by the person who abides by the identity.
I am a believer in the law of conservation of identity, but simply more identities is not the solution.
The normal way of dealing with an externality is to keep it contained and consistent. A similar argument could be made about identities. However, in some cases, the identities being created at this moment foster a much smaller burden on others. I don’t think a general principle can be deduced.
For the record, moderate socialist (AKA Sanders, single payer, higher taxes, etc) who thinks robots will 50% take over and can go with out eating for a day because eating is a nuisance.
LikeLiked by 1 person
The idea is that it’s OK to fail at one of those identities, because you can afford to drop it, while maintaining enough ego-affirming identities that you succeed at.
If you put all your eggs in one basket, you cannot afford to drop that basket as you lose all eggs. But if you merely drop one egg, you can just bake a slightly smaller cake.
Evan Þ. said:
Is this actually the case in the real world, though?
Yeah, it’s the case.
The most extreme case is the battered wife whose controlling husband prevents her from having any social existence outside of being his wife. Isolated to a single role, the wife cannot imagine ever leaving the husband. It would be negating her whole person.
Only when she gains new friends, new perspectives, new social circles can she gain the power to leave the controlling husband, because then she has those new people and new roles to fall back upon, rather than basically negating her entire existence (as a wife).
It’s obviously an extreme case, but it totally happens.
Regarding the baker, it may help to identify more as a person who *does* things than a person who does *well* at them, depending whether you’re more prone to low self-confidence or laziness. A baker who identifies as being able to bake *well* will perceive criticism as an attack; however a baker who identifies as one who bakes a lot will be prone to procrastination-based spirals of self-doubt. Those two factors and probably more will need to be balanced depending on your personality.
This confuses me. Are all American libertarians (supposed to be) anti-racist?
1) American libertarians seem to be largely anti-police:
I don’t see it as a core belief of libertarians, but the commenter community at Reason.com is strongly biased against police. (I mean that they’re hugely suspicious of police good faith, not that they think we shouldn’t have any, although some might think that second thing as well).
Pretty much any claim that any police officer was acting in good faith then he or she detained, questioned or injured someone is going to get a lot of push-back.
2) Anti-racism is a closer question:
AFAICT, most libertarians (and here I’m still using Reason commenters who are supportive of libertarians as a probably not very good sample) think that (a) racism is morally wrong, including racism against white people, but (b) “SJ” attempts to reduce racism through coercion are also morally wrong and probably poorly targeting, and (c) the government shouldn’t be in the business of fighting racism except maybe for the most extreme cases. There are a strain of libertarians that are pro-HBD and strongly skeptical of factual claims about the prevalence of racism, but there are also libertarians who are anti-HBD. Hardcore libertarians believe the government shouldn’t be in the business of stopping racism at all, and that if someone wants to open a restaurant or a social club that doesn’t admit Swedes, that’s their business.
LikeLiked by 1 person
I took Ozy’s statement as implying that libertarians believe that the police are nice to white people, otherwise Ozy could have left out ‘black’.
My perception of anti-racists is that they tend to believe that white police officers act mostly in good faith towards white people and with bad faith to black people. If you believe that police of any ethnicity act badly against everyone, then the common demand for more ethnic diversity of police officers makes little sense.
But perhaps Ozy’s statement was based on the common ground between SJ libertarians and not-SJ libertarians, where all tend to believe that the police harms black people and the latter group tends to believe they also harm white people.
The more I see other people describe libertarians (in the libright context) the more I realize I am pretty much a libertarian, which is weird, because ‘unrestricted capitalism is the worst thing possible’ is one of my most consistent and strongly held political opinions.
(Also, I’m not sure how HBD quid pro quo leads to not believing in racism. Racists are overrepresented amongst HBDers, but the cause and effect there goes the other way around.)
LikeLiked by 1 person
I think that by “empirical claims”, you meant “value judgements”. “The minimum wage is bad for poor people” is not an empirical claim in the usual sense of being a directly observable fact.
The most interesting part of this post is the idea that having many independent identities reduces the risk that we’ll hold on too hard to any particular one, by effectively diluting it into the others. If we follow this to its conclusion, it’s a call to generalize your life’s skills and activities, as opposed to specialize in one specific, narrow field.
I think that generalization versus specialization is actually a very difficult trade-off. Generalizing more really does have the benefit of not putting all our eggs in the same basket, and is probably better at averting personal catastrophes. It puts things into perspective. It’s probably healthier overall.
But I don’t know, I also really like the fact that some people are *super-devoted experts* at one particular specialty. Specialists that devote their entire life to one field, whose identity really is all-in on that. If such a specialist ends up being wrong or fails, then yeah it can be absolutely soul-crushing. But if they are right, then they can be the single most shining force behind human progress, pioneers of entire new fields? So it’s kind of hard to tell people to avoid this path.
Machine Interface said:
I don’t quite get the starting point of this article.
The quoted bit is confusing and the answers to most of the questions it asks seems to be “nothing, it means nothing, this question has no truth-value”
Who goes around saying things like: “Lisp is the best language for all purposes! You should believe me because I don’t have an identity as a Lisp programmer so I’m objective!”?
Why is “avoid identifying as anything” equatable to “have a self-image as being objective”?
It doesn’t seem the problem you point out proceeds logically from Paul Graham’s advice at all. Rather, this seems to be the age old problem of people applying advices in letter but not in spirit – either because they don’t understand the spirit, or because they understand it too well and are trying to “game the system”.
veronica d said:
It strikes me odd that Paul Graham made that claim considering he was instrumental in constructing the modern “tech nerd” identity, including countless essays and at least one book on how freaking great nerds are.
(I’m probably only saying that because I’m a Lisp programmer.)
LikeLiked by 3 people
The verbiage ecstatic said:
I’d say it makes a LOT of sense. People who really go down the path of reflecting on and discarding identities til they whittle it down to a core tend to end up being singular individuals who strike out on their own, unique paths. They end up being very much themselves, rather than one of the crowd. People like that attract imitators, and if they’re successful in life, they often accidentally become identity entrepreneurs. This is the classic guru problem: one person goes exploring, comes back with some wisdom, everyone else cargo cults their behavior and values and forms a cult / religion / movement that loses the insights of the founder. So I think Paul Graham both making a statement like that and being a founder of a new identity is actually quite plausible
I’ll accept that there are people who feel better for having an identity, but I seriously don’t get why it’s satisfying.
I am who I am. I’m more libertarian than I am anything else, but that’s descriptive – I don’t feel compelled to agree with any particular libertarian idea unless I think it’s right.
There are communities that I have access to because of my identities – Catholics, other people in my profession, other people who went to my college, etc., but again I don’t feel that my identity requires me to believe anything. If I say that I went to a game at our hated rival school and the students there were a blast to watch a game with, that’s just a statement of fact. The absolute worst case is some people might not see me as part of the community, but that’s not the same thing as who I see myself as.
I guess I wouldn’t wear hated rivals’ colors to a game, and I don’t eat meat on Fridays during lent, but that’s definitely the exception – most of my identities are fully descriptive – they describe the kinds of things I already believe and do, not prescribe them.
LikeLiked by 1 person
Apologies to Ozy, but I think I’d like to strenuously object here. Graham’s advice is *most* important for those who feel a strong drive to maintain an identity, because they are the people most at risk for the dangerous and harmful failure modes he identified.
I’d also like to clarify a point that was somewhat blurred in Ozy’s opening paragraphs: Graham doesn’t advocate for NO identity, but merely to make it as small as can be done to minimize irrational attachment to large groups of behaviors and beliefs. In his essay he’s pretty open about people still maintaining some (smaller) level of personal identity.
As others have pointed out, even Ozy’s innocuous baker example can easily be twisted under circumstances to be a harmful identity- so there really is value in keeping one’s identity small, even if you can’t *immediately* see what the harm in self-identifying as something would be.
I would also add, as a supplementary point, that keeping one’s identity small helps one to empathize with others who wouldn’t share your identities- if “man” is a very weak part of my identity, then it’s easier to empathize with people who are not men. If I don’t feel strongly “white”, it’s easier to empathize with nonwhite people, etc. In essence, I’m trying to avoid dividing people into groups where one group instinctively feels like “my group”, because I know that can trigger all manner of irrational and unhelpful instincts.
Of all things in life to base a personal identity on, your profession is probably one of the best ones. Just from the basic way that a specialized skill works, your skill at your profession is something that (hopefully) you really do want to strongly defend as if it were part of yourself, to grow as you grow, and to keep for a very long time. If you don’t at least somewhat identify as a baker, then you’re just not going to be a very good baker.
Sure, we can contrive examples where it’s bad to attach yourself too much to a professional identity, like “baker”. But given that we need some identity anyway, then your professional identity should rank really high among your choices, I think. It’s a special case among personal identities and should probably be regarded as such. To a certain degree, we *want* professionals to strongly identify with the quality of their work!
A somewhat related concept I like to keep in mind is an *ideal* to aspire to. Like, if I’m a baker, then I keep constructing this conceptual *ideal baker*, who is a paragon of bakery, who always does the best bakery practices. And probably this mental ideal baker really is an amalgamation of a bunch of teachers, mentors, peers, and celebrities I’ve seen or interacted with.
Then whenever I’m feeling lazy and want to cut corners, to be sub-par as a baker, I imagine this ideal baker being disappointed in me, or telling me to do better. “What would the ideal baker do in this situation?”
In my own case (coder), I have a mental group of people who play this role. Whenever I’m about to cut corners by copy-pasting code around, I feel the disappointment from one specific mentor who I disappointed in the past by copy-pasting too much. Or I have a former colleague who, in my mind, will berate me when I try to do something manually and I should just use automated tools to do it way faster. Or a third friend who will tell me to always simplify my sometimes-convoluted explanations. Together, they play the role of the ideal coder, whose validation I crave!
It sounds kind of stupid, but as a mental tool that speaks directly to my brain and inspires me to do better, it’s surprisingly effective! It hacks the deeply social parts of my brain, that crave external validation, and gives them something to chew on, even if only indirectly. It’s like the secular version of “What would Jesus do?”, I suppose, and it’s not quite about identity itself but it sure has a similar flavor.
I’m not sure I agree with the part about keeping an identity small being a good way to empathise with those who don’t share it. I think identities can be good ways of empathising with those who hold identities of a similar category type but which are different at the object level- e.g., someone who identifies strongly as an American might find it easy to empathise with someone who identifies strongly as a Spaniard, but not with someone who weakly identifies as a Norwegian. I think this intuition is behind the failure of some of the attempts to generate empathise for trans people (‘Imagine you woke up as a gender you don’t identify as’ won’t work as well on cis-by-default people as it does on the other type of cis people, because they don’t identify as firmly that way).
And this relates back to Ozy’s point. Compare ‘I am a chef and I identify very strongly with this, and this makes me different to most people’ with ‘Wow, other people are all terrible chefs. I don’t even identify as a chef that strongly, they must suck’.
LikeLiked by 1 person
I agree with you that having a weak national identity perhaps makes it harder to empathize with those who hold strong national identities- and indeed, as someone who’s probably cis-by-default, your trans example is a good one as well. But for this to work, you really only need to feel strongly about ONE identity, and then learn how to construct an analogy. Still good to feel strongly about as few things as possible, I think!
It seems like this is all definition-wrangling here. Paul and Ozy are advocating for the same consequentialist result and similar mechanisms, dictionally and tonally tailored to the bravery debate that their targeted different populations need.
Which is, essentially “use whatever concepts (large, small, whatever) you need to be chill about your identity.”
The complex part of this, of course, is that it’s a bit of a “everyone needs to get on board for this to work.” Alice who is completely chill about their identity could have one imposed upon them by Carol who hasn’t got the message yet, and have to defend their chosen lifestyle against Carol even if Alice personally doesn’t even identify with the label Carol slings against them. After all, that’s the very situation of how nerds become socially outcast in the first place.
Tons of people who had no direct stakes in Gamergate or its context whatsoever picked a side/identity only after someone came after them. “Well, I guess I’ll just have to embrace the insults you tarred me with, out of spite!”
An aside: If you identify as a libertarian, then you’re signing on to a bunch of empirical claims, like “the minimum wage is bad for poor people” and “environmental regulations cause more harm than good” and “police officers abuse people’s civil liberties.”
Pretty sure following Hayek doesn’t require me to believe any of those other than perhaps #1, and even that, while unavoidable from those first principles, doesn’t require me to think it harms poor/low-skill people very much in the forms it’s in right now.
“Reason’s target-market Libertarian” ain’t the same identity, though I realize from the outside such subtleties aren’t obvious.
(I also tend to Andrew’s analysis above.
I mean, in my own example, I label my economics libertarian because that’s the best label for them.
But it doesn’t mean I view that as foundational to my identity, as a label, such that I somehow feel the need to defend, oh, Rothbard’s stance on government.)
Good point about the description vs identity label, Sig. I might say I’m a “math student”, which is a true descriptive label of my academic status. But it isn’t an identity- I don’t *feel* like I’m something important that is “a math student”, nor does it lead me to feel the urge to defend math students, or show group loyalty beyond what I think is useful in a direct way, etc.