In relation to your answer (I'm not a professional mathematician, thus if my comment isn't suitable here feel free to flag it to be removed) I think that it could be inspiring to refer in such talk/conference the efforts/works of some mathematicians, contemporaries or not, as refers 1) the Wikipedia Gösta Mittag-Leffler in its penultimate paragraph and also from last paragraph from the section Scientific impact of the Wikipedia Henrietta Swan Leavitt, or 2) the Wikipedia Jean-Marie De Koninck (winner of Canada’s Volunteer Awards 2019) in relation to the known as Opération Nez Rouge.

Great idea, if such a lecture were augmented by one in which an opposing viewpoint is expressed. Not so great if the lecture takes the form of bible catechism, where the trinitarian dogma of Diversity, Equ(al)ity, and Inclusion is rehearsed.

"A topic ... with no obvious other place to go ..." Are you actually serious? Attention to diversity is all over the place. Every university has diversity statements. There's entire conferences dedicated entirely to women or minorities. There even special job positions for 'diverse' individuals.

There are several viewpoints on this problem and different opinions on the possible solutions, the talk should not take sides and include them all.

Maybe I'll regret asking this later, but what are the "opposing viewpoints" in this context?

@RP_ I would also be interested in hearing what an "opposing viewpoint" is to what David wrote.

I can't honestly for one minute believe that you're actually under the conviction that everyone is on the same page on this subject. I am therefore filing both your questions under the rubric "bad faith" and I am not discussing them any further.

Of course I don't believe that, I'm simply unclear on what the details of the opposing viewpoints are. If you could fill me in or point me to a good source, that would be very helpful.

@RohilPrasad One obvious issue I know is about quotas. I prefer the talk to be impartial, just like a math talk :)

Politics of this particular topic aside, I don’t understand the requirement that all points of view be presented. If Alain Connes wanted to give a speculative talk on the connections between the Riemann hypothesis, noncommutative geometry, and physics, I don’t think anyone would say that he would need to yield the podium to someone else to rebut these speculations. All talks, but especially those at a prestigious conference like the ICM, are evidently putting forward a point of view (e.g., that the math in question is interesting, if nothing else).

I agree with Sam. The idea that all of the myriad viewpoints should be represented is a well-intentioned one, but ultimately difficult in just about any situation. Also, mathematics is never truly "impartial". There are a lot of societal factors at play in determining what research gets noticed, what problems/topics are considered "important" and so on. In any case, further discussion on this particular topic in the comments is likely to be unproductive. It would probably be best for anyone who is so inclined to post a good-faith "rebuttal" answer to David's of similar quality.

@SamHopkins My understanding of David White's suggestion is to present the point of view of the mathematical community on diversity, not a particular opinion. I would find much more enjoyable such a talk as it would give me a better understanding. Moreover there has been no talk on diversity before, so it would be quite logical to begin with a presentation.

@RP_ My question was genuine. I do understand that people are not on the same page. My exposure is mostly to what François Brunault calls "different opinions on the possible solutions" rather than "opposite viewpoint", so I took the chance of asking for clarification.

@Patriot To clarify: "No other obvious place in the ICM program." According to the fourth bullet point in the question, this kind of lecture could fit as an ICM special lecture. I stand by my argument that such a talk could help propel the math community to a better place, and could help mathematicians understand the current status, current initiatives, and research on what works. I do not think a "rebuttal" would be necessary, as the talk would be based on the published literature, and not what people feel we should or should not be doing. ICM can help legitimize this kind of work.

I would say an opposing position is MathOverflow. Granted within its scope there is variety, but this community is not as diverse as it could be. It does not handle all areas is mathematics, nor all levels, nor all interests. There are arguments for "keeping focus", which some may unfairly label as "good ol' boy network". Gerhard "Think Of Intentions, Not Outcomes" Paseman, 2020.08.06.

vade retro satanas

I think this is an excellent suggestion, which is unfortunately undercut by the fact that the 2022 ICM is being hosted in a country with deeply regressive anti-LGBT laws, which is of course a major impediment to it being an event inclusive to queer mathematicians.

@Gerhard, it is a little unclear to me to what extent MathOverflow can be described as a position. I would be more likely to describe it as a website.

@Sarah, I speak in vague terms, but hopefully those that Gjergji will understand. There is a position different from one of diversity, inclusiveness, and being equitable, and that position is held by some of the MathOverflow community. This is not without some justification, for MathOverflow was not meant to be global in scope and purpose. However, some of the reasons for restriction that apply to MathOverflow may also apply to David's thesis of this post. In any case there is a valid and distinct position to be held. Gerhard "Might Discuss It At ICM2022" Paseman, 2020.08.06.

"making math more diverse would also lead to better mathematics, as a diversity of thought and background will lead to new approaches to problems we care about" <--- this is a key point you should take away, even if you are someone with zero interest in the other (very worthy imho) aspects of this issue. How many great contributions throughout history have we missed due to the social aspects of being a mathematician? cf the well-known examples of Sophie Germain, Emmy Noether etc that are the lucky few not in the traditional 'Mathematician' box who managed to 'make it' (kind of).

@GerhardPaseman This vagueness and reluctance to indicate comprehensibly what is being discussed is uncharacteristic of mathematicians. I can't help but think it stems from a belief that the 'opposing viewpoint' under discussion, whatever it is, is considered so repellent as to invite professional or social consequences.

This comment space and this forum are not meant for general discussion. My intention was to give a brief answer to Gjergji's question prompted by RP's comment. Believe what you will; I am not going to contest David White's suggestion on this forum. I merely indicate such a contest is possible. Gerhard "Which Is Not Politically Favorable" Paseman, 2020.08.06.

@KamerynWilliams I am not sure if the country has deeply regressive anti-LGBT laws. Among the countries with similar GDP per capita or HDI the laws are quite reasonable (but of course looking from a purely Western perspective they are regressive). Maybe ICM should be exclusively hosted in developed countries to ensure its friendliness to LGBT-identified individuals. Russia's hostility to this demographic happens to be more well-known to the Western observer because they get more screen time in the news.

I am skeptical that (a) the diversity in terms of social circumstances (e.g. are you LGBT-identified or not, are you an ethnic minority or not, are you poor or not) is correlated with diversity of mathematical thought (b) that ensuring diversity of social circumstances will lead to better mathematics beyond the fact that some talent happens to be in the social minorities. To me, the reason to strive for social diversity is purely ethical (kind of like my opposition to slavery is purely ethical even though many economists claim that slavery is economically inefficient).

In fact, it seems to me that in some circles (usually unrelated to mathematics) which are especially dedicated to diversity of social circumstances diversity of thought is low and dissent is not tolerated. This is only an anecdotal observation though.

@DavidWhite "based on the published literature": we are beyond the field of mathematics here. People won't even agree on the meaning of "what works". Taking the opportunity of ICM to make people aware and discuss all these issues is a great idea and I don't want to go against that. I just feel that the format of talk is not ideal. Perhaps discussion tables are more adapted.

"As we know, women and minorities are underrepresented in math" - sorry, I do not know this. Some, say, ethnic or religious minorities are "underrepresented", some are "overrepresented". This follows implicitly from the pigeonhole principle, and is confirmed by evidence.

"Sadly, many mathematicians are unaware of this body of research" or too aware of its shortcomings. There is still no agreement on what the Implicit Association Test actually measures. Stereotype threat is not holding up too well against replication attempts. Sure I've picked just the most obvious cases from your references, but ...

It may also be relevant to note that what an American social justice activist means by an ethnic minority may not be the same as the statistical minorities on this planet (which is probably more relevant to ICM since it's a global event). I think it may very well happen in the future that the Whites will be less than 20% of the global population. Will we then organize inclusion program for the Whites? In other words, the levels of historical and current wealth and the distinction of local vs global distribution must be taken into account.

... when a field is still unable after some 40 years to separate the chaff from the wheat in its research findings, it might not be a good idea to have it present a keynote at a mathematics conference. We aren't inviting speakers to tell us how AI will save/destroy the world or to push their favorite cryptocurrency either -- if anything, we would host a debate with a balanced lineup. And keynotes hyping quantum computing or string theory routinely get dissected in widely-read blogs and twitter, as would the hypothetical Atiyah talk described by @SamHopkins be.

They had this at the last ICM just look on the ICM youtube channel. We all know the arguments at this point, and either we are convinced or not. This suggestion assumes that we all haven't had to write boilerplate l̶o̶y̶a̶l̶t̶y̶ ̶o̶a̶t̶h̶s̶ diversity statements. I also think we should stop rewarding the bad behavior of the people at Inclusion/Exclusion and QSIDE. The whole attempted 'canceling' of Abbie Thompson has left an extremely bad taste in my mouth. Maybe we should have Abbie come give the talk instead.

@FedorPetrov In this context, "underrepresented" has a very precise meaning. The proportion of female mathematicians is lower than the proportion of females overall. Same for people of color. If you want hard data on this (albeit for the USA, though the lack of diversity is a global problem), check out this talk by Dave Kung: youtube.com/watch?v=V03scHu_OJE

@darijgrinberg Actually, as Whistling Vivaldi makes clear, the stereotype threat effect has been replicated hundreds of times. I'll read the link you shared, but you should really read the book (it's short, very well-written, and has proven strategies teachers can use to mitigate the stereotype threat effect). I'd love it if newer studies struggle to replicate the effect, because that would be evidence that our efforts to make math more diverse are starting to bear fruit. But the task is far from done.

@crispr Check out the link above to Dave Kung's talk. Data is also presented about the massive wealth gap between white and black families, so that's one big reason we don't need inclusion programs for whites. Perceptions of what a mathematician "looks like" is another reason. Again, I recommend reading Whistling Vivaldi. Lastly, to Harry, "people are either convinced or not" has never been a successful strategy for achieving meaningful change. The Civil Rights movement wasn't easy either. If people are still not convinced, all the more reason for an ICM talk.

@DavidWhite: What do you consider to be the strongest studies that confirm stereotype threat? Whistling Vivaldi was written in 2010, which is some 5 years before "replication crisis" became a commonplace word. The Spencer/Steele/Quinn study its author lionizes is not as convincing as it sounds. The ...

... Steele/Aronson study had seriously low n's and confronted participants with racial stereotypes so aggressively that it's questionable how well it represents any academic environment. These are old studies, of course. What is the gold standard these days?

As to Francis Su, my impression is that, even though he has quite a list of research papers, he is being mentioned here as a TED speaker rather than a mathematician. But that's exactly not what ICM is for.

(Another remark about the relevance of stereotype threat, whether or not the effect is real: Very few lecturers at the university level stress emphasize the intelligence-testing aspects of mathematics. It is considered to be a bad idea even without equity coming into play, as it downplays the role of learning and habit-buildings and pushes a fixed mindset. It's K-12 teaching where there is still far too much group-based discrimination going on, albeit it's questionable whether the currently fashionable type of politics will improve this in any way.)

@DavidWhite I am not familiar with this terminilogy, who are "people of color"? If they are "underrepresented" in American mathematics in the described meaning, then Chinese obviously aren't.

I would like to remind David that there are with the ICM parallel or satellite sessions arranged by the IMU and other bodies involving women in mathematics and mathematicians from developing countries. Before instantiatng the kind of lecture you propose, it might be good to review the current efforts and see how they can be improved. Your post suggests to me that this reminder is needed. Either that, or I am misreading your intent. Gerhard "Agree In Principle, Not Practice" Paseman, 2020.08.07.

An ever-expanding comment thread is the wrong place for this, but the links shared by Darij do NOT take down the Steele work. Both link to a meta-analysis that actually shows the stereotype threat effect IS significant, despite a potential bias in which studies got published. To say this field cannot "separate the chaff from the wheat" is highly inaccurate. Further, note that the two links are on a non-peer reviewed blog, using the notion of R-index the blog author invented (also not peer reviewed), whereas papers on stereotype threat are in top journals and by highly respected authors.

On the topic of bias, it's wise to remember confirmation bias, where readers are more likely to accept research confirming what they already believe. A related bias seems to occur whenever privilege is mentioned, because humans don't seem to like to think that they might have had it easier than others. Lastly, a very old bias leads to papers by black authors (like Steele) being criticized much more than papers by white authors. I think we see a similar bias in math when women and minorities are refereed more harshly, and pushed back against while giving talks, more than white men.

Oh also, I forgot to add, the link about the reproducibility of Steele-Aronson only questions one small part of the fourth study: basically, whether or not a triggering phrase is needed to trigger students from stereotyped groups on a diagnostic task, or if knowledge that it's a diagnostic task is enough to already trigger them. That's why I said it's very far from taking down the main conclusions of the paper.

This comment thread makes clear that this is a topic that is on a lot of people's minds, which I think is a strong argument for finding some way to discuss it at important professional events like the ICM.

@DavidWhite: One point the first link makes is that the Steele/Aaronson study got published while a different study, also by Steele but failing to confirm stereotype threat, did not. This is a textbook example of publication bias. Let me repeat my question: What are the strongest studies showing stereotype threat in an academically realistic setting?

@R.vanDobbendeBruyn: Yes, but (as Harry Gindi said) it is a topic to discuss, not to give a plenary lecture on.

Something that I think is not getting enough attention in the entire thread: We are talking about a lecture (quite possibly plenary) at the largest mathematical conference in the world. An ICM typically has 20 plenary lectures, commonly meant to update listeners on the main advances in mathematical research(!) over the last 4 years. However worthy of discussion a topic is, before suggesting it for this format, it is worth keeping in mind that it would likely replace one of these lectures. (Unless it is merely a sectional lecture, but my impression is that this will ...

... leave more people unhappy than not scheduling it altogether.) For a topic that is mostly relevant to teaching to appear in a research conference (while there are myriad teaching conferences happening all around the world), something very unusual must happen. I don't see what this something is here, other than a media-orchestrated hype around diversity and racism.

On the topic of confirmation bias I have a belief, perhaps insulting to many, that when mathematicians enter the softer sciences the fact that a mathematician doesn't have to lit review/balance their reading (Paper A and Paper B won't/shouldn't disagree on the truth of Theorem X) leads them to being particularly susceptible to having an opinion and pointing at a paper to back them up.

@darij why do you say this topic is mostly relevant to teaching? While certain aspects are more relevant to teaching, in general, structural inequalities exist all the way from elementary school through to the tenure process and journal reviewing.

@JoshuaGrochow: Most mathematicians are not interested in acting out biases and rather correct for them when given the opportunity. So, when bias exists, it will likely materialize the strongest in first non-professional impressions and gradually even out over time as opportunity arises for it to be proven wrong. Thus I'd expect it to show up in student-teacher relationships much more than in (say) inter-colleague or purely research-related ones (such as tenure evaluations and refereeing). I don't look at an author's race when I referee a paper; by the time I've learnt of it, I've already ...

... have received much more informative signal about the author's competence from the paper itself. In tenure evaluations, I know the person for a year or two. In teaching, on the other hand, all I have most of the time are weak signals.

@darij You appear to be referring to implicit bias. I think I agree that implicit bias may have more effect in a teaching setting. But implicit bias is not the same as structural inequality (though each may feed off the other).

@JoshuaGrochow: I'm talking of both explicit and implicit bias. But yes, I'm talking of bias acted out by academics, not inequality accumulated from other parts of society. As to the latter, I see no good way of measuring it and no fair way of countering it on the level of academia.

@Harry: There are ways to (minorly) correct for, & at least not further contribute to, structural inequality on a dept (or perhaps university) level w/o discrimination as you say, e.g. by focusing on lack of opportunity/resources. E.g.: REUs that target students who would otherwise not have research opportunities. Making grad school more affordable (paying grad students more). Providing better medical/family leave policies. Actively dealing w/ explicit bias (even if it's only a minority who are explicitly biased, they can have an outsized effect on those they're biased against).

It's not just a teaching issue. Our lack of diversity starts with images of math in society, continues through education at all levels, implicit bias, microaggressions, harsher referee reports, harsher critiques in seminars, harder time winning a grant, fewer citations, more service work, lower salary, etc. All of this contributes to our lack of diversity, and as I wrote, this affects the quality of the math we produce. We should learn how to stop pushing talent out of the field. The sources I linked show ways to do this that do not violate the Civil Rights Act (e.g. stopping microaggressions)

Also, plenty of studies show that even if a reviewer (usually in a job context) doesn't know the race of the person being reviewed, if that person has a "funny name" then it implicitly affects how the material is read. Of course this leads to referees asking individuals with certain types of names (or from universities in certain (African) countries) for more details than white males in the USA. Of course, this leads to fewer publications and longer delays for those individuals. This is a real problem, and it's one we could at least attempt to fix. Shameful not to even try.

@FedorPetrov Are you serious? Did you try google (en.wikipedia.org/wiki/Person_of_color)? As I've written, lack of diversity is a serious problem in the USA, where I am based (hence the terminology), but as I've also written, it's a problem globally, e.g., with mathematicians in Africa having a harder road to publishing, grants, and jobs. So, if you're not familiar with the terminology, I'd suggest you try to GET familiar with it, so we can try to make the situation better. Ethnic minorities in Russia are also underrepresented. All the more reason to have an ICM talk about this topic.

Some minorities are overrepresented. Why did they ever stop numerus clausus?

@HarryGindi: Ah, now I think I understand better where you're coming from. It appears you're assuming that a particular class of "solutions" will be proposed ("put a finger on the scales in hiring decisions"), when no such thing was mentioned in the OP. If I am understanding you correctly, then - and I'm genuinely asking here - why not just say you're opposed to that type of solution, rather than opposed to having a talk on this at the ICM at all? (Also, it's not just about "outreach"; it's about removing systemic barriers to entry that have disproportionate effects on different populations.)

My previous comment was in light of things I believe are covered in the documentary Picture a scientist, though I wasn't aware of it until this morning, when a senior female colleague shared it with the department.

@HarryGindi do the suggestions in my earlier comment not fit your criteria? Here are a few more: all conferences should allow applications (no invite-only), provide funding for participants, and allow remote participation. REU-like program for high school students who otherwise lack such opportunities. Programs to help underprivileged students with their grad school application process. Remove PhD application fees and GRE requirements. I believe that if we improve the diversity of those who are well-prepared to, able to, and do apply (at all levels), that will result in huge progress.

@HarryGindi also, although it's true the underrepresentation is present already for first-year college students, the proportions don't then level off. As I recall, at least for some demographics, they get worse at literally every stage (HS, college, grad school, postdoc, Jr faculty, tenure).

The issues David raises in his post merit discussion, but not here. This topic is more fit for a panel discussion than a special lecture at the ICM. (A report on the progress, or lack thereof, might be more fitting at one of the satellite conferences. One of the remarks I heard from a satellite conference attendee at Seoul was why there wasn't a version of MathOverflow that was not in English. We as a forum community should examine ourselves when we discuss these issues.). Gerhard "People Are Thinking About This" Paseman, 2020.08.09.

@DavidWhite yes, I am serious and I had a look there. It says that Asian Americans are of color. You say that people of color are underrepresented in American mathematics. You obviously have some different definition in mind.

@JoshuaGrochow but do you think there is enough money in the world to do what you propose? I find the notion of a world where all conferences provide funding for all participants to be insane (unless there is an incredibly competitive application process). Is GRE biased as well? I think GRE is actually an opportunity for economically disadvantaged to prove that they are no worse than the rich. If you know how to properly use the Internet and you have the natural aptitude you can prepare for GRE on a very low budget.

Coming from a poor (but not a racial minority) background myself I would be concerned if people attempted to cancel GRE/SAT/ACT in the U.S. academic setting (and they actually have done something to that effect in California). And also, things you propose apply equally to say poor Blacks and poor Whites. I guess they would improve racial diversity because American Blacks are more likely to be poor than American Whites but painting it as being fundamentally about race is inaccurate in my opinion.

PhD application fees (among other things) serve the purpose of ensuring that the applicant is serious about their application and doesn't spam hundreds of universities. I guess there is the secondary effect of hitting the poor harder but you have to provide a replacement mechanism (otherwise the system will be completely drowned in noise). Alternatively, the entire application process could be reduced to objective numbers but I don't know how this could be done in e.g. pure math.

Also remote participation in a conference is unlikely to help in my opinion. You can find the videos for a lot of high-quality mathematical talks online. That is not the primary issue. The primary issue is having physical access (as in, being within a 10 m radius) to the top dogs in the field and listening to their informal conversations and possibly asking them questions. At the moment that can't be done via Internet.

In any case, the comments on this reply (and the voting pattern) appear to demonstrate the need for a balance between viewpoints/forum discussion.

@FedorPetrov Thanks for taking this seriously, and your careful read. I agree it's important to know what groups we're talking about. I like the term "underrepresented minority" (but others might disagree, and I'm not wedded to this term). If we follow the categories in the NSF data (ncses.nsf.gov/pubs/nsf20301/data-tables) then in the USA this includes Black, Latinx, Native American, and Pacific Islander. The tables also include White and Asian, but these groups are not underrepresented. Other tables focus on women, persons with disabilities, and international representation.

I'd argue that we should also consider LGBTQ status but this does not seem to appear yet in NSF data. And, I'm sure there are other types of diversity I'm forgetting to include, which is all the more reason to have an ICM talk on this so we can better understand the problem, what activities work (as supported by research), and what remains to be done. It's a complex, important, rich, and challenging problem; just the sort of thing mathematicians usually jump at the chance to work on.

@crispr Yes, the GRE, SAT, and ACT are biased by gender, race, and socioeconomic status, and several other fields have drawn attention to this so that GRE scores will not be given too much weight in applications. This was nicely covered by the Atlantic (rb.gy/hoj6ie). If you’re worried about spam applications, application fees could be a sliding scale of net worth, like how Finland handles speeding tickets, rather than a fixed fee for all (which is like a regressive tax). This way of thinking is the “equity” part. Access to conferences and experts is the “inclusivity” part.

@crispr GRE itself may not be the problem, but the cost is. Yes I think there's enough to fund conferences that way: reduce individual grants & increase conf grants (or just the latter). I believe Santa Fe Inst funds all their confs. I never said it was only about race. Completely agree about what makes confs valuable, but those who can't travel miss out on that value. Remote participation is getting way better, as just 1 great example: medium.com/@juniper.lovato/…. Postdoc and Prof apps don't have fees.

I think (based on my own recent experience) that remote participation in conferences does help getting otherwise "marginal" (in a very literal, geographic sense) people visible. But we aren't very likely to run out of remote-accessible conferences any time soon, even if Covid disappears tomorrow. I think the interesting questions here are of technological nature and best answered by people with experience organizing conferences.

@DavidWhite: I have never done any tallies, but I'd be surprised if LGBT++ (I can't even hope to keep up with this word as it grows longer each day) people are underrepresented in mathematics -- I'm pretty sure they aren't in my own circles at least.

@darijgrinberg Yes, you can keep up. It isn't that hard. And yes LGBTQ+ people are under-represented, at least among those who feel comfortable being out. There are also different levels of representation. For example, would you say that the LGBTQ+ community is well represented if the only people you know are white gay cis-men? Is that diverse?

@GerhardPaseman Maybe these efoorts could be improved by having more people pay attention to these issues by, I dunno, having a lecture on this at the main meeting and not some parallel or satellite session where less people participate.

@FrançoisBrunault by impartial do you mean like the part of the talk about the fields medal laureates where they talk about why the person SHOULD NOT win the fields medal?

@Sean, oh yes, let's do that. There is too much math going on at these events anyway. Perhaps we should add a few answers to this question giving more visibility to this issue, as they may be needed more than the answers to the original question. (Really, political and social topics need to be handled with more care at an event intended for mathematical topics, or in a StackExchange forum intended for something other than social or political issues. I recommend you research current efforts with the goal of your giving such a talk.) Gerhard "Really, There Are Better Ways" Paseman, 2020.08.10.

I think this thread has gone past its original purpose. I strongly encourage those who wish to continue to start a chat room and invite people to join on the topic of equity and inclusion. I promise to add a contribution regarding people outside traditional academia for inclusion. Gerhard "Let's Give Circle-Squarers A Voice" Paseman, 2020.08.10.

@SeanTilson: "would you say that the LGBTQ+ community is well represented if the only people you know are white gay cis-men?" Probably not, but that's a heavy "if"; I could easily name counterexamples to any single of your mentioned attributes, although not to all possible combinations (not surprising, seeing that these are some small probabilities getting multiplied there), and this is not counting in-the-closet cases or people that I just don't know enough to tell. Of course, if you keep piling more and more minorities, eventually you'll find one I haven't encountered; no surprise there.

@everyone talking of GRE, I'm wondering what your suggested alternatives are. It's not like the choice is "GRE or nothing"; the choice, at least currently, is "GRE or some other tools for weeding out applicants". And pretty much all the other tools I'm aware of are far more biased than GRE by definition, as you're replacing an exam that can be prepared and learned for by some exercise in self-presentation or in finding influential letter-writing patrons.

@darijgrinberg lol the GRE is biased towards tryhards :^ ]. Maybe I'm just biased because I never took it, but it looks annoying.

@SeanTilson We are not speaking of math here. A correct analogy would be: imagine an ICM where some fields of mathematics are banned.

@darij it really can be nothing in place of GRE. I don't know anywhere that uses GRE to admit, only that uses it as a bar to weed out, and then there's an admissions process for the rest. So simply remove that bar, and then use your regular admissions process. Presumably that takes into account things like transcript (what courses and grades), letters, personal statement, etc. Yes, there will still be bias in the process. But I think it'll be better. The GRE seems to serve little purpose in the process itself, but it does prevent people without money or time to prepare from applying at all.

I feel this comment thread has gotten too long. Folks who want to talk more about the issues raised can do so at the following chat room: chat.stackexchange.com/rooms/info/111634/…. If the conversation continues here, the moderators will probably move the whole thing to a chat room. I'd rather if existing comments remain here, as evidence of the multifaceted views of the community on my ICM talk suggestion, but future discussion should go to chat. Thanks!

I have removed a strongly worded comment of my own, but I want to repeat the essential content so as so set the record straight. At least two comments have been removed (both with 40+ upvotes), apparently by moderators, one comment by myself and one by Harry Gindi. Both of these were critical of David White's proposal, with my comment calling for the inclusion of alternative viewpoints on the matter. What disturbs me is that, from the remaining comments, it is considerably less obvious how controversial this subject is in the MO community (and presumably among mathematicians in general)...

... The comments as now they now are present a falsified picture of the original discussion, which is not something that moderators should do, imho.

This thread clearly shows how controversial this topic is, and there's no reason for it to be controversial. How many female mathematicians have been awarded the fields medal, how many active mathematicians are from the LGBTQ+ community? These questions will show you the underrepresentation. But instead, we are interested in defining "person of colour" and confirmation bias. For lack of a better term, this just seems like elitist jargon.

@RP_ the moderators have censored dissenting opinions of several users here. My comment, too, was removed. I made a meta thread about it and hope it garners some attention. meta.mathoverflow.net/questions/4671/…

@SayanChattopadhyay Fields medal is a measure that is not very sensitive to short-term changes (they've been awarded for several decades and at most one is offered per year if you average out). Defining people of colour is certainly important because different racial groups have been awarded Fields medals at different rates so the degree of bias seems to be different (so the solutions might be different as well). There is more Japanese Fields medallists than there is African Fields medallists for instance.

And I think mostly the issue was not about the existence of underrepresentation but the exact way in which we should address it (and the answer to your questions won't give an answer to that question).

Also regarding the LGBT statistics I am not sure how much of this can be pinned on the mathematicians. Many LGBT-identified people are reluctant to publicly state they are LGBT-identified so I don't know how accurate the statistics is. Who is responsible for creating a climate where people are reluctant to out themselves is a complicated question (and then should people even feel the need to out themselves?).

This conversation should be happening on chat. I'd ask people to stop commenting here, and even upvote my previous comment (that first linked to chat) so that it appears "above the fold" chat.stackexchange.com/rooms/111634/…

The active discussion is in the room below: