You talk a lot about experiences here, but all these answers express beliefs, not experiences. Beliefs can be arbitrarily biased -- just think about modifications of your method: Instead of asking black people about their "experiences" with the police, you could ask police officers about their "experiences" with black people. Or you could ask people without migration background about their "experiences" with immigrants. You could have asked Christian Germans in 1938 about their "experiences" with Jews, etc. What you will get is a bunch of opinions which could indicate real problems, or completely different sources of antipathy.
A far better approach is to actually test common sentiments by more reliable means (e.g. crime statistics). Doing standpoint epistemology seems no more valuable than doing arbitrary opinion polls.
By the way, I disagree with the assumption that Aumann's theorem vindicates any such "standpoint epistemology".
You talk a lot about experiences here, but all these answers express beliefs, not experiences.
Unless you have comprehensive measurement such as cameras etc., information about experiences are necessarily mediated by human beliefs.
However, human beliefs can be about many other things than experiences; e.g. about statistics/general societal tendencies, about options for change, about epistemology, about future trajectories, etc..
Beliefs about experiences constitute a specific form of beliefs, and experiences are in a sense the "native" way for humans to gain information about the world (i.e. hunter-gatherers have experiences too, but they probably don't work with statistical data), so they are probably the place where humans form their richest/most-detailed/most-informative opinions about.
Beliefs can be arbitrarily biased -- just think about modifications of your method: Instead of asking black people about their "experiences" with the police, you could ask police officers about their "experiences" with black people.
In the post, I specifically suggested that asking the police would make for a good followup:
We also shouldn’t forget that this is only a part of the story. I asked black people to describe their experiences, but I haven’t allocated time to ask police about their experiences. A functioning policy for society should presumably address both the problems black people face and the problems the police face.
When you say that my suggestion here is "arbitrarily biased", what do you mean by that?
A far better approach is to actually test common sentiments by more reliable means (e.g. crime statistics). Doing standpoint epistemology seems no more valuable than doing arbitrary opinion polls.
Could you help me understand how looking at crime statistics would help enlighten us about the problems black people face with the police? I think they would be more illuminating about the problems the police or society face with black people.
But also, I don't think statistics are all that epistemically different from surveys?
Help me out if I'm misunderstanding something here, but my understanding is that both my survey about experiences and crime statistics ultimately originate in experiences where people have interacted with the police. Cops and the people they interact with then process those experiences to remember them and make sense of what to do.
Where my approach differs from crime statistics is that my approach then just ends there, asking people about what they've figured from the interaction. Meanwhile in the case of crime statistics, police are given authorization to decide that some interaction is sufficiently harmful that they should arrest the person involved in the interaction. And then they kick off a governmental system of having people come out to investigate the scene of the crime and form their own experiences/beliefs from their observations, and having people to interrogate the experiences of the people involved in the interaction, and similar. And then based on this, records get created in the bureaucracy about the crime (which can then be abstracted into statistics), and punishments gets dealt to the person involved in the crime.
(Also, I've heard that since this whole process is very expensive and time-consuming, and potentially error-prone, they almost always use plea bargains where the person getting punished is offered a lighter punishment if they will state that they are guilty of a crime? Or something?)
So on the one hand, the crime statistics are based on information that has been investigated by multiple parties. But on the other hand, the crime statistics are very directly coupled to huge punishments and rewards, and are dependent on rigid and expensive bureaucracy, and they are disempowering in that you have no way of obtaining your own information about crime statistics, because it requires kicking off some huge bureaucratic machine, which you cannot afford. Plus crime statistics are very abstract. Both methods probably have their place, though.
As for opinion polls, when I think opinion polls I usually think of something like "How much trust do you have in the police on a scale from 1-4?". In such polls, there isn't really very much information in each response, e.g. if there are 4 response options then that makes for 2 bits of information. Contrast this with the responses I got, which seemed to have tons of bits of information.
In the context of qualitative interview questions like this, straightforwardly taking the answers to be about "the problems black people face" or "the problems the police faces", presupposes, about individual opinions on what these problems are, that these beliefs are neither incorrect, confused, or otherwise inaccurate. Again, imagine interviewing pre war Christian Germans to find out "the problems Germans face with Jews".
Qualitative interviews are even less reliable than opinion polls, since in those polls we get at least statistically significant results about which (possibly unjustified or incorrect) opinions are common and which are uncommon. But open ended questions can't be statistically analyzed in this way.
Qualitative interviews can have value to come up with questions for opinion polls in the first place, or to come up with hypotheses to test by other means. I would just warn against overestimating the value of such open ended questions by interpreting them to directly report actual experiences or problems.
Again, imagine interviewing pre war Christian Germans to find out "the problems Germans face with Jews".
I don't have any clear imaginations of what would happen in this case?
Like I know that antisemitism was rampant there at the time, so probably you would get a lot of angry negative opinions. But what would they be? "My pastor's friend's niece was killed by a Jew"? "Jews control the banking system which is evil and also they are breeding like rabbits"? "There's a group of child prostitutes downtown, and their pimp is Jewish"?
I would like to know what the results would be, to be honest. Probably there would be a need to take them with a heavy grain of salt, but I can't take them with a grain of salt without taking them in the first place.
In the context of qualitative interview questions like this, straightforwardly taking the answers to be about "the problems black people face" or "the problems the police faces", presupposes, about individual opinions on what these problems are, that these beliefs are neither incorrect, confused, or otherwise inaccurate.
I don't think we have to take everything as 100% dogma.
Rather, the important part lies in getting a better understanding for each other's perspectives. There are presumably a lot of things we don't know, but the information in this survey seems at least to me to be hugely valuable in getting an understanding of what black people face.
Qualitative interviews are even less reliable than opinion polls, since in those polls we get at least statistically significant results about which (possibly unjustified or incorrect) opinions are common and which are uncommon. But open ended questions can't be statistically analyzed in this way.
I think a qualitative question can reasonably well prove the existence of an opinion in a person. This allows lower-bounding the prevalence of the opinion, as one knows that it exists in everyone who expressed it. In particular, qualitative questions automatically weight the probability of an opinion being expressed by its prevalence in the population, because only those who have the opinion would express it.
However, I think a lack of mentioning an opinion in a qualitative question does not necessarily prove that one does not have that opinion, because qualitative questions are fairly loose and so one might simply not get around to saying it. I think this can probably be mitigated in a qualitative interview, as the interviewer can control the conversation to keep going back until the opinion has been confirmed or disconfirmed.
Okay, these points seem reasonable.
One other worry I forgot to mention however: I could be totally wrong here, but presumably most applications of this kind of "standpoint epistemology", in the last ten years, comes from researchers I would suspect of being far-left activists. If so, those people would of course be very eager to interview people they believe in their political worldview to be victims of oppression, i.e. especially black people and women. They would very rarely interview white men or Asians or police officers about their "experiences" or "problems". This can lead to publication bias, where a lot of "problems" and "experiences" of particular groups get published, but hardly any of other groups which fall outside the concern of the predominant political ideology of those researchers. Then the evidence is biased in virtue of the selection effects at place.
I'm not 100% sure about this, but from what I've heard a lot of left-wing academics don't even try all that hard to reveal black people's experiences, but instead mainly use black people as a tool to say that right-wingers are bad.
I agree that this sort of thing is a problem, but I'd think it is best addressed by doing more to map out different people's experiences in a publicly accessible way.
That is what I am getting at when I say:
And these principles don’t have to be disempowering. While “shut up and listen” may sound tiring, it is important to remember that going out to find people willing to educate you (possibly in exchange for payment, as in this post) is an active action you can take in order to improve your understanding of your world. Knowledge is power!
One of the benefits of Standpoint Epistemology is that it is very efficient. People naturally observe and remember many of their experiences as they live their life, and it is relatively quick to just ask them about it. This post only took me about a day’s worth of work, and less than $100 worth of money. If scaled up to be more comprehensive, it would presumably take more work, but presumably also be more informative.
My ideal outcome for this post would be if more people went out and mapped more groups' perspectives of more situations.
By the way, I disagree with the assumption that Aumann's theorem vindicates any such "standpoint epistemology".
That also stood out to me as a bit of a leap. It seems to me that for Aumann's theorem to apply to standpoint epistemology, everyone would have to share all their experiences and believe everyone else about their own experiences.
My take on Aumann's Agreement Theorem is Don't Get Distracted By The Boilerplate. Yes, it's usually phrased with certain technical conditions that might not fully apply in the real world, but the basic implications of the theorem constantly appear in everyday interactions, for spiritually similar reasons to the technical conditions required by the theorem, even if those technical conditions don't literally hold.
The argument you are linking to only shows that Aumann's theorem.works under convenient conditions.
Those convenient conditions are trust. The point I'm trying to get at is that while "common knowledge of each other's rationality, honesty and beliefs" is an excessively strong condition for it to commonly occur, sufficiently strong knowledge of rationality/honesty to obtain trust and sufficiently strong knowledge of beliefs to obtain information is a common occurence.
It doesn't literally appear constantly, because there are many situations where it doesn't, and where it doesn't the boiler plate does matter. You have a half-truth that you have stated as a truth.
I mean yes, people regularly interact with untrustworthy (crazy/liar) people, and it is a condition for Aumann's theorem that the people in question are neither crazy nor liars. I was interpreting Stephen Bennett to point at the stronger aspects of the condition (common knowledge technically being a very strong condition), but maybe that wasn't his intent.
It may be good to think of Standpoint Epistemology as an erisology, i.e. a theory of disagreement. If you observe a disagreement, Standpoint Epistemology provides one possible answer for what that disagreement means and how to handle it.
Then why call it an epistemology? Call it Standpoint Erisology. But...
According to Standpoint Epistemology, people get their opinions and beliefs about the world through their experiences (also called their standpoint). However, a single experience will only reveal part of the world, and so in order to get a more comprehensive perspective, one must combine multiple experiences. In this way the ontology of Standpoint Epistemology heavily resembles rationalist-empiricist epistemologies such as Bayesian Epistemology, which also assert that people get their opinions by accumulating experiences that contain partial information.
This is already clearly epistemological. So calling it just a theory of disagreement seems already out of place. This also sets up a motte and bailey. Once the less standard claims, like "white people need to shut up and listen", get criticized, it will be possible to claim it's only saying that "people get their opinions and beliefs about the world through their experiences" which "heavily resembles rationalist-empiricist epistemologies such as Bayesian Epistemology", which begs the question why it needs a special name at all.
One important difference is that whereas rationalists often focus on individual epistemology, such as overcoming biased heuristics or learning to build evidence into theories, Standpoint Epistemology instead focuses on what one can learn from other people’s experiences. (...) As such, Standpoint Epistemology emphasizes that if someone tells you about something that you haven’t had experience with, you should take this as a learning opportunity
So it's a pointer to a useful source of information?? A technique for gathering it? That would be a much simpler and clearer argument to make than trying to cast this as an epistemology and comparing it with Bayesian epistemology.
it can be mathematically proven from the assumptions of Bayesian Epistemology, in a theorem known as Aumann’s Agreement Theorem
Aumann’s Theorem requires two honest Bayesians with the same prior. You definitely can't just assume that about everybody (and how much it applies to rationalists is also debatable). But really that's irrelevant, you don't need Aumann to make this point - every observation is Bayesian evidence, including what people say[1], which makes this point trivial, and again raises the question of why it needs to be a special epistemology.
10 participants
When you interview so few people you can easily get a biased sample out of pure randomness. This is the same ol' point of anecdotes vs data. The former lets you go deep, but risks missing the full picture and getting bogged down in noise (like dishonesty and outliers), and the latter lets you go wide and cut through noise, but risks a shallow understanding.
Really, they're best when used together. anecdotes help you learn what to study widely, and data lets you situate anecdotes inside a larger context. So instead of hearing the "experience of a black person" (Which is pretty abstract) you can hear the experience of someone in the X percentile of income, Y percentile of educational attainment, living in a Z percentile rate of crime city, etc.. By using both you can get a truly full picture. e.g, see how many blacks (compared absolutely and relatively with whites) have how much income, and then see in depth through interviews how it is to be a poor black person, a median income black person, and a rich black person. Same with racism, you can use data to approximate how many people experience racism, and then go deeper to see how it is to experience heavy racism, to experience light racism, no racism at all, or even 'reverse' racism (where people treat you better because of your race).
And I'd like to stress, that this is all still trivial inside a Bayesian framework (or even just a "commonsense" framework).
So let's think of how we should respond to hearing one response in the survey. As Bayesians, before taking the anecdote at face value, we should consider what information it actually constitutes. We should consider the sample (black people who use the survey website), and ask how much should we trust the sample (or people in general, if there's no reason to expect a relevant difference) to be honest, have good judgement, have good memory, and be equally likely to report both good and bad experiences. And then we should consider how surprising it is to hear a specific anecdote from someone in the sample. My point isn't to answer any of these questions, but to point out that Bayes doesn't allow you to just take the responses at face value, as standpoint epistemology may instruct you to do. if it does, it's false, and if it doesn't then it's trivial.
I asked black people to describe their experiences, but I haven’t allocated time to ask police about their experiences.
This is treating this like a conflict and going to hear the other side before even checking how different the experience of black people is to white people (unless you base off your experience, which I guess standpoint epistemology would approve of). Not that I disapprove of learning about police officer's experiences, but first make sure you have the experience of the first group situated correctly in the larger context.
To conclude, your post gives an extremely trivial account of "standpoint epistemology" (except the mention of "white people need to shut up and listen") which makes me disagree with the framing more than the content (I downvoted for the framing, but were it missing I would probably upvote for the content). "Standpoint Epistemology" is either trivial, like it's explained here, in which case the framing is bad and I suggest dropping it, or it's non-trivial, in which case it should be judged on the merits of its non-trivial claims, which from what I've read and heard, I believe are entirely mistaken (I might explore this further in a followup comment).
And as this is LessWrong, perhaps a good final question is, would you want to code standpoint epistemology into an AGI? Either there's something not contained in a bayes-like epistemology you would want it to do, in which case, what is it? Or it's fully contained inside such an epistemology, in which case it's too trivial for the framing.
- ^
it can also be evidence against what they're saying, if you believe they're more likely to say that in worlds where it's false
Then why call it an epistemology? Call it Standpoint Erisology. But...
Erisology is a very recently coined term, whereas the standard is to call it an epistemology. Also, as you point out, there are epistemological elements to it, and part of why I'm emphasizing the erisology angle is because I've been thinking of writing about a typology of erisologies that I have coined:
Mistake theory (disagreements originate in mistakes):
- ability erisology: some people are smarter (or otherwise better) than others and this leads to disagreements between the informed and the uninformed
- standpoint erisology: people have different experiences, leading to different beliefs, and if they cannot pool their info to come to a shared set of knowledge, then they end up persistently disagreeing
Conflict theory (disagreements originate in conflict):
- battle erisology: some people represent concerns that are in tension with your concerns and they are trying to promote their concerns by spreading lies and interfering in communication
- trauma erisology: different people have had different negative experiences that makes them avoid different sources of information, leading to them having different opinions due to their different sources of information
(Trauma erisology would be conflict theory because the "negative experiences" are usually related to conflict.)
This also sets up a motte and bailey. Once the less standard claims, like "white people need to shut up and listen", get criticized, it will be possible to claim it's only saying that "people get their opinions and beliefs about the world through their experiences" which "heavily resembles rationalist-empiricist epistemologies such as Bayesian Epistemology", which begs the question why it needs a special name at all.
"Sets up a motte and bailey" sounds like nonsense to me. Any time you mention multiple things, it is possible to ignore critiques of the less-defensible things and shift the discussion to the more-defensible claims.
If someone directly disagreed with the whole "white people need to shut up and listen" element, I wouldn't deny that I've linked to that, and would instead engage with that area.
So it's a pointer to a useful source of information?? A technique for gathering it? That would be a much simpler and clearer argument to make than trying to cast this as an epistemology and comparing it with Bayesian epistemology.
I've found it (or rather, theories adjacent to it which I learned before I got into this specific instantiation of it) to have profound implications for how I've been thinking about knowledge and evidence. 🤷 If you have already integrated the lessons for this, then maybe it is less profound to you.
Aumann’s Theorem requires two honest Bayesians with the same prior. You definitely can't just assume that about everybody (and how much it applies to rationalists is also debatable).
It definitely has assumptions, but also you shouldn't get distracted by the boilerplate. As long as we can agree that black people aren't generally liars or crazy, we should expect there to be a ton of value in listening to their experiences.
When you interview so few people you can easily get a biased sample out of pure randomness. [...]
Really, they're best when used together. anecdotes help you learn what to study widely, and data lets you situate anecdotes inside a larger context. [...]
I've definitely considered using these sorts of stories as inspirations for more rigid quantitative statistics. I didn't get into this as part of this study because I didn't have time to make it a huge comprehensive thing.
I agree that my study doesn't very precisely determine how big of a problem it is, but I think it is still quite informative. Obviously how informative it is depends on what conclusions you want to draw, but I've laid the stories bare rather than abstracting them into any sorts of problematic conclusions.
This is treating this like a conflict and going to hear the other side before even checking how different the experience of black people is to white people (unless you base off your experience, which I guess standpoint epistemology would approve of).
There's obvious tradeoffs in how well the police treat the black people they encounter vs how the police handle the black criminals they encounter, because the police do not necessarily know ahead of time who is criminal or not. Understanding what areas black people could get treated better versus what areas police face danger might help enlighten us on things like Pareto improvements to the treatment, or places where the treatment is inherently unfair to one side or the other.
It's not clear to me how comparing the differences of the experience of black people to white people ends up working out. For all I know, white people's experiences of the police might be "I haven't interacted with the police". (That's my experience, but 1. I don't live in the US, 2. I am 99.9% introversion, so my experience doesn't necessarily generalize to Americans in general.)
"Shut Up and Listen" can be good advice if framed in the right way "sometimes there can be a lot of value in just shutting up and listening, unfortunately, once you start speaking there's a good chance that this causes people to shape their views to be more palatable to you". Unfortunately, over the last few years, people have been using this as something of a power move to shut down opinions and I have to admit I find your framing of "there do seem to be some people" as somewhat strange given that this has been occurring all over the place.
Most public political discourse seems dysfunctional to me. While this also applies to people who appeal to the principles of Standpoint Epistemology, I think some people have mistakenly come to the conclusion that this means the philosophy underlying Standpoint Epistemology is incorrect. However I think most of the time the problem isn't with Standpoint Epistemology, but instead with people having a goal of undermining the discourse, which they can do with superficial similarity to any mode of interaction.
I think Standpoint Epistemology is an especially unfortunate target of this, because it seems to me that Standpoint Epistemology, applied properly, is the thing that is necessary to pass through the undermining of discourse.
I'd also like to see an identical survey but for white people. That way comparisons could be made.
This might be interesting if one is trying to figure out the extent to which white people face fewer problems than black people do. However I probably won't do such a survey because there are many other questions I am uncertain about and writings I need to complete which have higher priority.
That said I would encourage you to do a similar survey if you are interested in the topic.
In their book on social dominance, Sidanius and Pratto make a relevant point: first, they cite a bunch of audit studies where researchers send fake resumes to employers, and find a marked bias against employing African American. Then, they point to Gallup polls asking people whether African American face any discrimination, and almost half of African-Americans themselves say they don't. Same for discrimination in justice or housing. So, when the book was published back in the 90s, many black people didn't believe in racial discrimination, even though it affected them personally in their life.
This means that people's perception of discrimination is not so much influenced by lived experience, but by what the dominant ideology is at a particular time. Back in the 90s, racial discrimination wasn't emphasized in the dominant discourse, so people thought it wasn't very important. In that case, standpoint epistemology just entrenches the dominant beliefs.
Another example: today, there's a vast body of research showing large discrimination against men when applying for housing. If you've ever applied for a place to rent, this has affected you personally favourably or not. But were you aware of it? If you did an online survey asking men what discrimination they face, how many of them would bring up housing discrimination? They don't know about it, because the dominant ideology doesn't talk about it.
I see this as a critical failure of standpoint epistemology (and the "lived experience" approach in general). Here is a 2011 survey where white Americans claimed that white people face more discrimination than black people. I don't think this gives us any valuable information about how much discrimination white people actually face.
(I do see the value of lived experience for hypothesis generation, of course.)
In their book on social dominance, Sidanius and Pratto make a relevant point: first, they cite a bunch of audit studies where researchers send fake resumes to employers, and find a marked bias against employing African American. Then, they point to Gallup polls asking people whether African American face any discrimination, and almost half of African-Americans themselves say they don't. Same for discrimination in justice or housing. So, when the book was published back in the 90s, many black people didn't believe in racial discrimination, even though it affected them personally in their life.
I don't think the Gallup polls are very well-designed (at least for this purpose). It seems like they ask about only a few bits of information about very abstract general questions that aren't straightforwardly related to people's personal experiences, and then they aggregate the data together into averages, destroying most of the information about who sees what. I don't feel I get a picture of what people mean by their claims of discrimination when reading those polls.
This means that people's perception of discrimination is not so much influenced by lived experience, but by what the dominant ideology is at a particular time. Back in the 90s, racial discrimination wasn't emphasized in the dominant discourse, so people thought it wasn't very important. In that case, standpoint epistemology just entrenches the dominant beliefs.
Another example: today, there's a vast body of research showing large discrimination against men when applying for housing. If you've ever applied for a place to rent, this has affected you personally favourably or not. But were you aware of it? If you did an online survey asking men what discrimination they face, how many of them would bring up housing discrimination? They don't know about it, because the dominant ideology doesn't talk about it.
So I agree that when a problem appears very diffusely, without it being reliably observable in the instances where it does occur, then it is hard to identify it by asking people about their experiences.
However I am not sure social science is the solution to this. Social science typically also seems driven by ideology more than factual matters, and also social scientists usually seem plain bad at their job. Furthermore these sorts of studies tend to strip away so much context that they are hard to make sense of.
I haven't specifically read the literatures you point at. If you think they are much better than other social science, feel encouraged to write it up.
I see this as a critical failure of standpoint epistemology (and the "lived experience" approach in general). Here is a 2011 survey where white Americans claimed that white people face more discrimination than black people. I don't think this gives us any valuable information about how much discrimination white people actually face.
Again this is a really abstract study which strips away most detail. I'm not sure how I'd quantify discrimination in such a way that one can about who faces "more" discrimination, and I'm not sure how the participants quantify it.
But I feel like there could totally be some sort of experiences behind it which explains their views. In fact the paper itself speculates that maybe what white people have in mind is that the institutions designed to rank and teach people have policies prefer to admit lower-skilled black people to higher-skilled white people, or that the American government has policies to require the contractors it works with to prefer hiring black people.
Does this mean white people are """more""" discriminated against than black people? Dunno. It seems like the discrimination is in different areas that are hard to compare.
Do you think the original proponents of Standpoint Epistemology would agree that it's simply a logical consequence of Aumann's agreement theorem?
I haven't talked with the original proponents, and I don't think the original proponents have heard of Aumann agreement. It's also kind of complicated because I think it's not some monolithic thing that was created by the original proponents and then treated as a settled issue, but rather something that evolved over time.
Also I didn't mean to imply that it is simply a logical consequence of Aumann's agreement theorem. Rather, it is a consequence of Aumann's agreement theorem combined with other things (people are usually neither crazy nor liars, talking with people about their experiences and concerns is cheap and easy, etc.). However Aumann's agreement theorem seems like a key thing to point at since it is rarely discussed by rationalists.
I had never heard of Standpoint epistemology prior to this post but have encountered plenty of thinking that seems similar to what it espouses. One thing I can not figure out at all how this functionally differs to surveying a specific demographic on an issue. How, exactly, is whatever this is more useful? In fact to me it seems likely to be functionally worse in that for a survey the sample size is small and there is absolutely no control group, as someone else pointed out, we don't get any sense of what any other group responds with given the same questions.
I don't really have an issue with the proposition that there is value in considering different groups experiences, what I do have an issue with is why it seems bound to devolve into a myopic consideration of a very small number of people's experiences.
I don't really have an issue with the proposition that there is value in considering different groups experiences, what I do have an issue with is why it seems bound to devolve into a myopic consideration of a very small number of people's experiences.
I would definitely find it interesting to survey people-in-general too. However, that seems quite difficult. First of all, the site I'm using to survey people mainly has people from USA and Britain. Secondly, most people don't speak any languages that I speak, so I cannot design the questions for them myself, nor can I interpret their answer myself. It would also be a much bigger project as I would need to put even more effort into understanding their local cultures in order to ask relevant questions.
I had never heard of Standpoint epistemology prior to this post but have encountered plenty of thinking that seems similar to what it espouses. One thing I can not figure out at all how this functionally differs to surveying a specific demographic on an issue. How, exactly, is whatever this is more useful? In fact to me it seems likely to be functionally worse in that for a survey the sample size is small and there is absolutely no control group, as someone else pointed out, we don't get any sense of what any other group responds with given the same questions.
A control group is mainly relevant if one is interested in differences, e.g. if one wants to know how Caucasian-American problems differ from African-American problems. That may very well be a topic of interest, but I think African-American problems are also interesting in and of themselves.
The necessary sample size to understand something depends heavily on the amount of variance in that thing and the precision to which you want to map out the variance. For instance, if you want to estimate the mean value for a variable with standard deviation , then typically the accuracy (standard error) of the estimate will be proportional to . If is low - that is, if there is broad agreement, where there is a lesson that matches with all of the different narratives - then the needed sample size to get a small standard error is also low.
This seemed to happen to a great degree in this survey: while the exact content of the different participants' responses differed a lot from participant to participant, the updates to my beliefs that seemed to be suggested by their experiences didn't differ hugely. So I feel relatively safe making those updates. (That said, ideally I should still cover the various biases I mentioned in the end of my post.)
As for "surveying a specific demographic on an issue", do you mean something like opinion polls? Opinion polls tend to use questions that are more rigid/less open-ended, have lower information content, and focus on more "processed" parts of things (e.g. attitudes to specific policies, which presumably depend on all sorts of odd factors).
This is quite abstract and cites difficult to access sources, so I can't easily engage with it. It looks to me like it is citing people applying Standpoint Epistemology as if those applications were the arguments for Standpoint Epistemology.
However I notice it is on a website by James Lindsay. Overall I don't have a good impression of James Lindsay, as he often seems to be misrepresenting things when I dig deeper. For instance part of what spurred this post was the various arguments my interlocutor gave for Standpoint Epistemology being bad, and among those arguments were James Lindsay's book, Critical Theories. This is one of the main things I was referring to when I wrote:
the resources that criticized it seemed to me to either be misrepresenting what Standpoint Epistemologists were saying
It was also hard to find great examples in the Critical Theories book, since he kept things fairly abstract. However one of the concrete 'problems' I found Lindsay to point to was this:
The problem is that as far as I can tell, the 2014 paper "Tracking Epistemic Oppression" does not exist, and he was mixing together a 2014 paper titled "Conceptualizing Epistemic Oppression" with a 2011 paper titled "Tracking Epistemic Violence". Based on his description of the paper, it is likely that his critique is directed towards the former paper.
The paper in question isn't great - it's convoluted and poorly written, and it seems to use a strained Plato's Cave metaphor in order to avoid spicy political examples. However, while the paper is pretty bad, it's not bad in the way Lindsay is describing it.
For instance Lindsay claims that Dotson claims that schemata have been specifically set up to work for dominant groups and exclude others, but really the way Dotson characterizes schemata is:
Organizational schemata, as I understand it, are a shared epistemic resource like language that enables goals and pursuits to be shared collectively.
This schemata concept is really similar to the popular LessWrong post of Shared Frames Are Capital Investments in Coordination.
The problems she describes in terms of including oppressed people's experiences into the shared epistemic resources are mainly problems of trust rather than the knowledge deliberately being organized around excluding others.
This post is also available at my substack.
This post started from a bit of a weird place. I was in a Discord chatroom, and someone started complaining that Standpoint Epistemology had been “taken way past its carrying weight”.
I didn’t know much about Standpoint Epistemology, so I asked for various examples and resources about it. The resources she gave me that were written by Standpoint Epistemologists seemed relatively reasonable, and the resources that criticized it seemed to me to either be misrepresenting what Standpoint Epistemologists were saying, or to be criticizing people for something other than excessive Standpoint Epistemology.
At some point I got to the conclusion that in order to evaluate these things, it would really be useful for me to apply some Standpoint Epistemology myself. Specifically, since a lot of the discussion in the Discord server was about black people’s experiences with racism, I thought I should apply Standpoint Epistemology to this. In this post, I want to detail how I went about this, and what my results were, so that others can learn from it, and maybe usefully apply Standpoint Epistemology themselves.
Disclaimer: As you will see, this is not a thorough investigation into what African-Americans want. Rather, it is a brief initial investigation, which suggests places for further investigation and further learning. This is probably more a practical tutorial into how I would apply Standpoint Epistemology than an article on race issues per se.
What is Standpoint Epistemology?
It may be good to think of Standpoint Epistemology as an erisology, i.e. a theory of disagreement. If you observe a disagreement, Standpoint Epistemology provides one possible answer for what that disagreement means and how to handle it.
According to Standpoint Epistemology, people get their opinions and beliefs about the world through their experiences (also called their standpoint). However, a single experience will only reveal part of the world, and so in order to get a more comprehensive perspective, one must combine multiple experiences. In this way the ontology of Standpoint Epistemology heavily resembles rationalist-empiricist epistemologies such as Bayesian Epistemology, which also assert that people get their opinions by accumulating experiences that contain partial information.
One important difference is that whereas rationalists often focus on individual epistemology, such as overcoming biased heuristics or learning to build evidence into theories, Standpoint Epistemology instead focuses on what one can learn from other people’s experiences. There is only one underlying reality, but different people observe different aspects of it. As such, Standpoint Epistemology emphasizes that if someone tells you about something that you haven’t had experience with, you should take this as a learning opportunity, rather than concluding that they must be irrational, biased, or crazy.
This notion that one should listen to and believe what others say does not contradict the mathematical underpinnings of traditional rationalist epistemology such as Bayesian Epistemology. Instead, it can be mathematically proven from the assumptions of Bayesian Epistemology, in a theorem known as Aumann’s Agreement Theorem. However, while Standpoint Epistemology follows from Bayesian Epistemology, I feel like we don’t necessarily see rationalists being as positive towards it as they could be.
In the specific case of racism, one article that the person in the Discord server shared with me as an example of Standpoint Epistemology was The Part about Black Lives Mattering Where White People Shut Up and Listen. This article, taken literally, is a good introduction to how to apply Standpoint Epistemology to areas of race that I wasn’t familiar with. However, the article is written very abstractly, so let’s instead take a concrete application.
What problems do African-Americans face?
I had originally thought of asking black people what their experiences with the police were. However, as a white person who doesn’t even live in America, I thought it was a bit presumptuous to just assume that I knew what the relevant issues for black people were. So rather than assume this, I decided to first ask some black people what they considered most important.
Of course as a white person who doesn’t live in America, I don’t know all that many African-Americans that I could ask. And even if I did, I could worry that I had already preselected them for ideological agreement with myself. Alternatively, you might be worrying that I was pressuring them into agreeing with me as a condition for my friendship or support.
To bypass these problems, I recruited black people from a website where people get paid to take surveys. At first, to get an overview I just asked a broad question, “What do you think about politics? Feel free to mention whatever is important to you; I am broadly interested in knowing what politics you care about.”.
In response to this question, pretty much all the respondents critiqued the way discourse works these days. For instance, a 49 year old black man wrote:
A different person, a 43 year old black man, wrote:
In my experience, a lot of other people also feel like political discourse has become very bad. I think probably we could learn more about the problems these people face by drilling deeper down into their answers, e.g. asking them for examples of the problems they have in mind.
I didn’t drill deeper because this was just meant to be a brief investigation, and my investigation went in a different direction. You probably already have your own detailed impression of how well political discourse is going, but if you want a more complete perspective, talking with black people about the problems they face in discourse might teach you something new.
That said, there wasn’t a 100% focus on complaints about the discourse, for instance a 22 year old mixed-race woman wrote:
Overall, there wasn’t much tendency to explicitly mention Black Lives Matter and the relationship between black people and the police. The one exception was a 22 year old black woman who wrote:
One thing I was concerned about is that maybe by asking “What do you think about politics?”, I was suggesting to people that they should comment on the meta-level, about how political discourse should work, rather than about object-level policies such as education, health care, and justice. This is one of the downsides with surveys, that it can be unclear whether the questions are understood correctly, but that problem can be mitigated with iteration. Still, clearly there was some strong agreement that the current political discourse is a big problem.
Another thing I was concerned about is that maybe the people in question didn’t have much personal experience with the police (perhaps some sort of selection bias was going on), so in order to investigate both whether the participants had experience with the police, and whether I had asked about their political views in a suboptimal way, I decided to run another survey.
In this second survey, I asked people “What do you think is the most important political question/topic/area?” and “Have you ever been stopped by the police?”. Furthermore, in order to not be implying to the participants that the survey was uniquely about police, I also asked about random unrelated questions, such as “Have you ever tried to start a business?” or “Are you satisfied with your education?”.
5 out of 10 participants reported having strong political opinions, and 7 out of 10 answered the question about the most important political question, with the responses listed below:
Furthermore, 8 out of 10 of the participants reported having been stopped by the police, so clearly there wasn’t some crazy selection bias against participants having had interactions with the police, and probably the original focus on political discourse was in part driven by the question phrasing.
Overall, I would say that these results suggest that police are not the sole important issue for black people, and that different black people have different concerns. In fact there is a lot of overlap between the policies that black people are most concerned with and the policies that white people are most concerned with. These facts probably don’t surprise anybody, but they are nice to see confirmed by my own investigations.
I think it would be massively beneficial to dig more into the problems with discourse, education, social and job equity, etc., which the participants mentioned. However, again, this is just an initial illustrative investigation, so I will restrict myself to digging into a single thing:
What about black people and the police?
Part of what originally inspired me to do this survey was discussion about police racism. In order to not be presumptuous, I think it was good that I did the investigations in the previous section. But still, I am curious about what is going on with respect to the police.
So I fired off a third survey, asking the following question:
What did they respond?
A 46 year old black man:
A 50 year old black woman:
A 52 year old black woman:
A 25 year old black woman:
A 35 year old black man:
A 46 year old black man:
A 27 year old black man:
A 24 year old black man:
A 42 year old black woman:
A 52 year old black woman:
Given all these experiences, it makes a lot of sense to me why the black trust in the police in America is so low. It seems like there are quite a few problems in these interactions. However, there are also positive interactions, and I don’t doubt that having the police is important for society.
Conclusion
I have sometimes seen people criticize Standpoint Epistemology for being inflammatory or irrational. There do seem to be some people who use language superficially similar to Standpoint Epistemology in order to shut down discourse. However, it seems to me that the underlying principles in Standpoint Epistemology are sound enough.
And these principles don’t have to be disempowering. While “shut up and listen” may sound tiring, it is important to remember that going out to find people willing to educate you (possibly in exchange for payment, as in this post) is an active action you can take in order to improve your understanding of your world. Knowledge is power!
One of the benefits of Standpoint Epistemology is that it is very efficient. People naturally observe and remember many of their experiences as they live their life, and it is relatively quick to just ask them about it. This post only took me about a day’s worth of work, and less than $100 worth of money. If scaled up to be more comprehensive, it would presumably take more work, but presumably also be more informative.
We also shouldn’t forget that this is only a part of the story. I asked black people to describe their experiences, but I haven’t allocated time to ask police about their experiences. A functioning policy for society should presumably address both the problems black people face and the problems the police face.
When applying these methods, it is probably important to think about sampling. For example, I asked ordinary black people about their ordinary experiences with police; meanwhile, police probably interact disproportionately with hardened criminals, so they probably have very different perspectives. This means that seemingly contradictory experiences (such as black people feeling they are usually acting reasonably and police acting unreasonably, versus police feeling that they are usually acting reasonably while black people are acting unreasonably) may be due to differences in selection bias (such as police disproportionately getting called to the scene when someone is acting unreasonably).
The issue of selection bias becomes especially important in the context of politics. Politicians, activists, journalists and commentators may choose to specifically amplify the black voices that agree with them, while disregarding the ones who disagree. I think for any view there will almost certainly be someone who has some rare or extreme situation that causes them to endorse it, but often that situation will not directly be relevant for the applications you would have for that view. As such, if someone has an opportunity for doing heavy selection to find a standpoint that agrees with their own biases, I would be skeptical and would try to get a broader selection of opinions.
It is also worth being aware of self-selection. Plausibly, when the participants were scrolling through the surveys to find one of interest to them, those who had more negative experiences (or unusually positive ones) with the police were more likely to click on my survey. There are ways of mitigating this problem, such as placing the questions as just one smaller part within a bigger survey. I mostly did not do this because I just wanted to quickly get a rough expression, even if it was potentially somewhat biased in one way or another. This (and other biases in various directions) does mean that my results should be taken with a grain of salt.
I think these investigations can be taken much further. One could ask more questions or deeper questions of the people one is interviewing. One could ask more demographics about more subjects of interest. It seems like it would be good for society to better map out the problems that different people are facing, so that these problems can be fixed, and it seems like people describing their experiences would be a good step forward with respect to this.
One potential problem with the current investigation is that despite the initial open-endedness, it was still fairly rigid and controlled. I had no real back-and-forths with the participants, and I did not talk much about the purpose of my questions or do much (beyond having open-ended questions) to directly allow them an opportunity to redirect my attention. Someone who has time for a deeper investigation should probably do a more egalitarian and interactive approach, e.g. setting up interviews and informing participants about the deeper purpose of their investigation.
Thank you to Justis Mills for proofreading and feedback.