Scholars asked Harvard President Claudine Gay for her data. She said no. | Hacker News Reader

Scholars asked Harvard President Claudine Gay for her data. She said no.

dossier.today

94 pointscs70289 comments

I'm finding this thesis kind of pathetic, it doesn't seem PhD material to me.

The US is so ¬ DEI and sensitive about it that even mentioning anything DEI related gets you in a position of power at Harvard now?

Fun fact for me, I beat Harvard in an ACM north east programming competition many years ago (but got destroyed by MIT). To be fair, I don't think Harvard is known for its comp sci.

She attended Princeton and Stanford as an undergrad and got her PhD at Harvard. She sounds more like a legacy hire than a DEI hire.

I lost faith in academia when I saw so many fudge data and copy pasta. This wasn’t even secret, it was quietly acknowledged.

I once inherited code that “worked” but operated on uninitialized values and out of bounds data unintentionally. I also couldn’t get it to complete a run without crashing. Even when I fixed bugs, I couldn’t reproduce the results. My own mentor told me to look the other way and move on as if this was known and totally usual.

It was also exceedingly hard to reproduce results from papers and authors seldom responded. Few would open data and even fewer would detail enough about their methods in the paper.

I’m sure others saw exceptional levels of fraud and other unethical behavior. I don’t, for one second, doubt that Claudine could have forged her way to the top. In fact, I think it could be the norm.

Nobody ever read or utilized President Gay’s “research” anyway, or that of the overwhelming majority of social “scientists.” What’s the point of adhering to some charade about analytical rigor? It’s not like bridges are falling down as a result.

It's still sadly fashionable in engineering circles to bash social science without reason.

Please list objections objectively * .

'* Was Kant a social scientist? Perhaps. Odd trivia: his work reversed the notion of subjectivism and objectivism. But what's in a word?

Good reasons:

1) “Social science” isn’t science.

2) “Social science” doesn’t enable you to build bridges, computers, weapons, power plants, or anything useful.

The situation at Harvard is strong evidence of the above. Do you think the other folks in President Gay’s department are much more rigorous about their work? But it doesn’t matter. Nobody relies on that research. Gay’s predecessors (Faust, Bacow), may not have plagiarized their papers and books, but nobody read those either. If folks don’t have an axe to grind about President Gay over her comments about Israel, nobody would have even paid attention to this story. But people noticed when, e.g., Boeing engineers were lax on the 737 MAX!

I may somewhat agree with your point but method of getting there makes me doubt yourself and myself :-)

1. "Social science isn't science" is a claim, but there's no proof or detail or evidence or logic put forward to back thay claim.

2. Social science doesn't build computers or bridges,sure. But that's not definition of science (maybe engineering?). If for example social science could reliably predict social group dynamics and behaviours that would make it a science. Or if it could help a group of construction workers reliably work better together to build that bridge that also counts.

3. All sciences sadly have at times both real whopsie and repeatability crisis.

If we want to make a (fairly worn out) claim that social science is lacking, we really need to out forward a considered claim.

> 1) “Social science” isn’t science.

Have you ever read anything by Eric Kaufmann? He's a political scientist (like Claudine Gay), but one whose conclusions I think you'd find have a lot in common with your own thinking (unlike Claudine Gay).

I wouldn't call Eric Kauffman a "scientist," and I don't think Kauffman would call himself one either. His bio page says:

> I am a Professor of Politics from Vancouver, British Columbia, I was born in Hong Kong and spent eight early years of my life in Tokyo. In winter, my favourite pastimes are (ice) hockey and skiing. In summer, cooling off in the ocean or by the lake. I am principally interested in cultural politics: ethnicity, national identity, left-wing ideology and religion.

Lots of people study things, without properly being called scientists. Historians, archeologists, etc.

> I wouldn't call Eric Kauffman a "scientist," and I don't think Kauffman would call himself one either.

Back in October, during an interview with Tara Henley, [0] he said "I'm a political scientist..." And, the topic of the interview is how he's leaving the University of London to start up a new "Centre for Heterodox Social Science" at the University of Buckingham.

"Political science" is the standard term for Kauffman's academic discipline, and so not only does he call himself a "political scientist", it is standard for journalists reviewing his books (etc) to call him that too, see [1]

> Lots of people study things, without properly being called scientists. Historians, archeologists, etc.

Well, Merriam-Webster defines [2] "archaeology" as "the scientific study of material remains (such as tools, pottery, jewelry, stone walls, and monuments) of past human life and activities" (my emphasis), and very many archaeologists do consider their discipline to be a "science": for example, the University of Wisconsin-La Crosse's Mississippi Valley Archaeology Center defines archaeology as "the science that seeks to learn about past humans and their culture from the material remains that were left behind" [3]

Likewise, the British philosopher and historian R. G. Collingwood's definition of history as "the science of res gestae, the attempt to answer questions about human actions done in the past" is frequently quoted.

[0] https://tarahenley.substack.com/p/transcript-eric-kaufmann

[1] https://www.newyorker.com/news/q-and-a/a-political-scientist... https://www.abc.net.au/sundaynights/stories/s3003131.htm https://www.theatlantic.com/ideas/archive/2019/06/eric-kaufm...

[2] https://www.merriam-webster.com/dictionary/archaeology

[3] https://www.uwlax.edu/mvac/process-of-archaeology/

Are you seriously arguing that fabricating data should be allowed for social sciences?

Because otherwise it might just look like you are you just going off topic to grind some personal axe.

I’m arguing it’s pointless to criticize that because the “data” and results aren’t the point anyway.

Gay should be fired. Meanwhile: almost every comment you make on HN is a social science argument of one sort or the other. You're a social science nerd. It's fine if you think social science at Harvard is being done badly; that's a colorable argument. But I don't think you of all people can put yourself above the whole endeavor of social science.

I think firing Gay would be indulging in a fiction that any of this matters.

You’re correct to identify the similarities, but my point would be that none of us are doing science. My arguments in social issues are rooted in culture and tradition. Pointing to data and observation doesn’t turn those arguments into scientific ones. Culture and tradition also relies on experiential knowledge—that doesn’t make it a science.

Indeed, my problem with the social sciences is that they intrude on the province of culture, tradition, and religion. It fills the vacuum of authority resulting from the secularization of society. But it’s not any better, it has no greater claim to authority because it’s not science. You’re better off asking your grandma how to raise kids rather than the Harvard faculty.

Gay isn't the President of Harvard because she's widely viewed as one of the most effective social scientists of her generation; if anything, her hiring puzzled social scientists, because of her scant track record. So calling out her whole academic discipline is a non sequitur. By the letter of your arguments, she's close to the ideal social scientist for the role: one who does very little social science.

Culture, tradition, and religion are all objects of study by social scientists. People do social science one way or the other. You can't have a world without it; you can only have good social science or bad social science. Again: that Harvard produces bad social science is hardly an outré argument. Just make that argument.

You’re begging the question. Just because people who study the same objects as culture, tradition, and religion, call their field a “science” does not mean we have to accept their terminology. I would call their field something more akin to theology, with the use of numbers being a sort of ritualistic exercise. Gay’s presidency is consistent with that—it simply abandons the pretense that those trappings actually matter to the field.

It's called "science" because it uses the tools of science to address questions about its subjects. Like all sciences, it makes predictions and tests them. As any social scientist will tell you, it is much more difficult to model and predict social phenomenon than basic physical phenomenon. You know that, which is why, in the message on the other branch of this thread, you couch your objection in "precise" predictions --- not an actual qualifier of science, but a decent rhetorical hedge.

Ultimately though, you're just quibbling with the name. You're as intensively interested in this field of study as its practitioners; from economics to developmental psychology, it is almost all you talk about. That's fine! But really, you just object to the conclusions you perceive social scientists to be arriving at.

> you're just quibbling with the name.

I'm not quibbling with the name, I'm objecting to classifying them together with real scientists. That classification has immense implications for how society views them and their assertions. I'd be fine if they called themselves scientists, if we called real scientists something else. The point is that society should recognize that their aping of scientific methodology confers no special validity to their conclusions.

> I'm objecting to classifying them together with real scientists.

Do you regard meteorology as a science? As a group, meteorologists aspire to make the best predictions they can with the available tools AND to develop better tools. As near as I can tell, social scientists do basically the same (as exemplified by, e.g., Kahneman and Tversky).

> Culture and tradition also relies on experiential knowledge—that doesn’t make it a science.

What's "science"? In mediaeval Europe, theology was viewed as a "science", indeed the "queen of the sciences", more noble than all the others. It may be objected that "science" was used in a broader sense in the Middle Ages than today; maybe that's true, but on the other hand, it is unclear exactly what the contemporary definition actually is - nowadays, people talk of "science" or "sciences" which are variously described as "natural", "empirical", "observational", "experiential", "experimental", "formal", "theoretical", "mathematical", "psychological", "medical", "clinical", "social", "legal", "human", "historical", "philological", etc - what is it that all these things actually have in common?

> Indeed, my problem with the social sciences is that they intrude on the province of culture, tradition, and religion. It fills the vacuum of authority resulting from the secularization of society

As well as the traditional Christian understanding of "theology as a science", I might also point to the phrases one sometimes encounters in Islamic contexts, such as the "science of hadith" or "science of fiqh". The idea you have of culture/tradition/religion and science as opposed is in itself rather non-traditional, not how mediaeval thinkers thought.

I would say that science requires being able to formulate internally consistent theories that generate precise predictions of reality that can be falsified by experiment. Nuclear physics is a science. Medicine is barely one. Simply drawing rough conclusions from data isn’t science—that’s something everyone does.

> I would say that science requires being able to formulate internally consistent theories that generate precise predictions of reality that can be falsified by experiment.

Is astronomy a science then? Astronomy is often said to be an observational rather than experimental science, but if your definition of "science" requires falsification by experiment (as opposed to just observation), then astronomy might not be a science at all.

And social science can make experimentally falsifiable claims. For example, one theory in political science (with some data to support it) claims that parliamentary systems (in which the executive is subordinated to the legislature) produce superior outcomes (greater political stability, superior policy choices, etc) to presidential systems (in which the legislative and executive branches are independent and roughly equal powers). One way to test this experimentally: convince a few US states to adopt the parliamentary system, wait a few decades, and then look for any statistically significant differences in outcomes between those states which adopted a parliamentary system versus those which retained a presidential (or should I say gubernatorial) one.

I mentioned before Eric Kaufmann. His book Shall the Religious Inherit the Earth? suggests that religion will triumph over secularism in the long-run due to a higher birth rate. If, over the next 500 years, the world never gets significantly less secular than it is today, I think he'd agree his theory would be falsified.

A friend of mine did her PhD in a prestigious lab in basic medical sciences. A postdoc in the lab was quite clearly fabricating data, but the PI refused to believe it. The postdoc got a tenure track position at a top ten department, from where he was quietly let go a couple years later when they figured out what was going on. The real tragedy is that, as the PI of my friend's lab was about to retire, the postdoc took decades of work--cell lines, antibodies, etc.--with him to start his new lab, and appears to have just thrown it all away when he lost his job.

My own experience in academia was not nearly so bad, but I was pressured by my advisor to exclude some less-flattering results from a paper we wrote. I wanted to leave them in because I thought they illustrated some interesting trade-offs between our work and other approaches, but I wasn't brave enough to stand up to someone with the power to seriously fuck my life up.

Not entirely the same, but at each step of the semiconductor manufacturing process, the process step is run on a pool of available tools from the fleet. (Think photolithography machines to make the pattern, etch or deposition tools, and SEMs &c for inline measurement.) These things being hardware for controlling physical and chemical processes, they're of course fickle, so even though two tools should be expected to produce the same results for the same step running the same configuration ("recipe") they unsurprisingly don't always do that. So in a given fleet maybe only 3/4 of them will be cleared to run a given recipe, and all the tools in this pool are supposed to be running the same version of the recipe. Discovering that several tools have not been running the same recipe and are producing noticeably diverging results going back weeks or months is an every-other-week sort of discovery and raising the alarm was always met with the same response—cover it up; don't make waves.

This tends to be more of the norm for the ones at 'good' institutions, which is aligned with the common knowledge / heavily replicated studies that 'top performers (best students, etc) are the most likely to cheat.

There can still be good academic works (and the alternative in industry has even worse incentives, such as intentionally hiding the ability to replicate) but unfortunately the structure as of now rewards bad behavior.

Your comment reminds me of a mini rant I put in a blog post of mine about my irritations in academia. Maybe a tad over-emotional but I stand by the lack of openness as an issue.

>Regarding headbang.py and beat tracking, I tried to find some "open" beat tracking datasets used in ISMIR and MIREX, and actually couldn't! I even emailed the task captain. Ludicrous. Hidden algorithms with hidden code using hidden datasets peer reviewed by the same hidden cabal that peer-review each others' beat tracking papers. Good job, you solved beat tracking .

        Dear Sevag, I am sorry to tell you that I don't have/know those datasets. I was the captain on the downbeat estimation task. For the beat estimation, you can email <redacted>. Maybe he can help you. Unfortunately, even as task captains, we don't have access to the datasets used for the evaluation. I am not sure if <redacted> will be able to give these to you but he might be able to guide you toward other datasets. Good luck, <redacted>

    Hello <redacted>,

    I was able to find the SMC12 dataset (hosted by SMC), but I'm having a harder time finding any copies of the MCK 06 and MAZ 09 datasets available for download.

    Are those available?

> I once inherited code that “worked” but operated on uninitialized values and out of bounds data unintentionally. I also couldn’t get it to complete a run without crashing. Even when I fixed bugs, I couldn’t reproduce the results. My own mentor told me to look the other way and move on as if this was known and totally usual.

20+ years ago I worked as a research assistant, on a software engineering research project (at an Australian university). A defence industry consultant had built his own requirements database, in Microsoft Access, which (apparently) was being used on some military project. There was also an Italian academic who had written some tool for parsing English requirements into logical statements and feeding them into a theorem prover. The tool only ran on Linux, and he wouldn't give us the source code, only a binary. One programmer was working on porting the Microsoft Access database to VB6, whereas my part was building an RPC system using Perl 5 (I used some CPAN module, might have been RPC::Simple) which the VB6 code on Windows could use to call the tool on Linux.

Somehow we got it all to work, for a highly unrealistic contrived demo; and then the funding ran out and I left the project. Some months later, it had stopped working, they couldn't work out how to fix it, and they wanted to hire me back to fix it-but I declined. I always had the feeling the project was designed to maximise research funding (international collaboration got one type of grant, collaboration with the military got another) rather than deliver something with real world value, or even serious value as basic research. But it was my first ever programming job, I hadn't even finished my bachelor's degree, so I took what I could get. I'd moved on to something more obviously valuable (working on the university's course catalog database), and I didn't want to go back to that project. I never found out if they got the code to work in the end or not.

Question to many of y'all: What does this have to do with DEI? Can she not just be a bad person...

She can just be a bad person, but my sense from scanning the comments is that a bad person artificially propped up by a social framework is worse because in theory it made it easier for the bad person to achieve a position of power where otherwise it’s less likely they would have achieved such a position. At least that’s my take on what people are arguing without diving in too deeply.

But what of her past indicates she's propped up by a social framework like DEI? She's related to NYT best-sellers, she lived in Saudi Arabia with a father that worked with the Army Corps of Engineers. She attended a private boarding school in New Hampshire, completed her undergrad at Princeton and Stanford, and earned her PHd at Harvard. If anything, she sounds like a privileged rich kid who's a legacy hire. Now some DEI-supported "token".

I’m curious too! I’m not making the argument, just distilling a little of what I’m sensing from others.

You've used "legacy hire" in a couple comments here. I'm not familiar with the term, and searching doesn't find it used much: https://www.google.com/search?q=%22legacy+hire%22

Are you consciously trying to create it as a new term, or is it in common use in some circles that I'm unfamiliar with? I don't think it's a good parallel with "legacy admission" unless it were to refer to hiring the child of an a previous employee, but this doesn't seem to be what you mean.

It's used in many circles for higher-class employment. But also "Legacy Preferences" may be the better term here: https://en.wikipedia.org/wiki/Legacy_preferences

Granted, her legacy is that she received her PhD at Harvard, then worked as a Dean of the Faculty of Arts & Sciences, on top of many other positions with various organizations.

My main point is that people claim that she's terrible because she was just a DEI hire and is not qualified for a position as Harvard President. When her resume shows she's more than qualified for the role.

This isn't a DEI issue, this is an issue of Harvard not doing the proper research on her dissertation & research and attempting to act like it's not a big deal for a President to be caught plagiarising works.

> https://en.wikipedia.org/wiki/Legacy_preferences

The secondary term here, "legacies" and "legacy students", are probably where most people have heard it, it's used in movies that involve sororities, when the snobby characters are gossiping: "How did she get in?" "She must be a legacy." and such.

She would never be in the position that she is in if it wasn't for DEI. DEI systematically helps bad persons into power.

We have a couple dozen years of modern time to show that bad people got into power way before DEI was ever a thing.

I don't think the parent implied that DEI is the source of all bad behavior, just that it's another avenue for bad people to get in.

Positions that are too important to be filled on merit.

Not even bad, just... merely average. An entirely ordinary person person who may have done sloppy work in the past and when called before congress, did what she was told.

I don't care who runs Harvard, a school I have zero connection to, and I can't scapegoat one person for sudden widespread spike in antisemitism over the last few months. They just look totally dumbfounded.

I imagine they will become a lot less dumbfounded in short order. DOJ investigations and a few million dollars in civil suits can have a near magical clarifying effect on a bureaucracy.

Claudine is just an ordinary person who just has had a lot of good positions under her belt at Harvard over the years. And from the meeting with Congress, it sounds like Harvard just royally messed up how their media relations should go.

Between this, plagiarism accusations and antisemitism, it’s remarkable that she still has her job.

Sad watching prestigious institutions fail.

The author asks many questions - typically in conservative publications.

> what happens when such crypto-communism is allowed to creep into our once hallowed institutions? What happens when Marxists become our central bankers? [1]

> Is Claudine Gay a Plagiarist? (Harvard President) [2]

[1] ref: https://www.karlstack.com/p/the-federal-reserve-goes-woke

[2] ref: https://christopherrufo.com/p/is-claudine-gay-a-plagiarist

Other writings: https://www.theamericanconservative.com/author/christopher-b...

https://freebeacon.com/men-of-the-year/2022-man-of-the-year-...

Questions about the author: https://www.poliscirumors.com/topic/why-was-chris-brunet-dis...

Yeah, and dossier.today is Jordan Schachtel's substack, who is also an interesting person:

https://thefederalist.com/author/jordanschachtel/

https://brownstone.org/author/jordon-schachtel/

https://www.aier.org/people/jordan-schachtel/

Thanks for sharing this info. I posted the OP, and didn't know anything about the author until now. A friend shared the link with me, and on a first pass it seemed accurate, so I posted it here. It stills looks accurate to me. No matter how poorly we think of the author, we should evaluate the OP on its merits, no?

To the moderators: Maybe there's a better, less-tainted link to this story?

I'm not from U.S., so I didn't look into this story closely. Just the name was familiar, and I also recalled this tweet from Ben Collins:

https://twitter.com/oneunderscore__/status/17378879685708392...

I mean this looks a bit like conservative PR campaign, but maybe plagiarism accusations are also true, I don't know.

Thanks. I don't know either. You may be right: Maybe it's both true and a PR campaign.

IF this is true: Shame on Harvard, shame on Harvard's administration, and shame on Harvard's Board of Overseers, for failing to live up to the University's ideals.

I mean, IF this is true, the implication is that no one at the university ever bothered to perform due diligence with a skeptical eye on Gay's track record.

IF this is both true AND representative of the state of leadership at our leading universities: They need major organizational change to clean up the rot.[a]

---

[a] Stanford's President had to resign for similar reasons: https://www.npr.org/2023/07/19/1188828810/stanford-universit...

---

To the moderators: I'm not sure why the OP is flagged. While the author is very ideological and controversial (see here: https://news.ycombinator.com/item?id=38782112), on a first pass the facts about Gay in the OP appear to be accurate. Maybe there's a better, less-tainted link to this story?

Should just let the fraud keep her position. It paints Harvard in exactly the light it deserves.

Great point. These institutions are being eroded from within. Some say DEI is a failing practice, I say it's achieving the exact results it set out toward.

I agree with her position, because what they're trying to do is a fishing expedition to find excuses to take her down. Why is everyone now trying to prove her academic honesty, only when she has become president and especially after the current events regarding the Middle East? Every lowlife is now trying to find some fault with her previously accepted work.

Shouldn't people in power be subject to more scrutiny? Politicians get taken down by past comments/tax evasion/affairs all the time. In those cases should we also dismiss those because they were "fishing expeditions"?

Tell me how many white male Harvard presidents suffered this kind of close scrutiny of everything they ever wrote? This is not a fair comparison. Every chauvinist, racist, and other kinds of lowlifes who want to take her down are joining in this fishing expedition.

"Controversial figures get more scrutiny" explains this without having to get race involved and implying that all scrutiny are from "lowlifes". Furthermore, I'm not sure what your policy prescription is here. Should we give people in disadvantaged groups a pass on scrutiny because we can write off any attempts as bad faith attempts from "lowlifes" to discredit them?

If you don't understand why she is being closely scrutinized now, just check the front pages of major newspapers in the last weeks. That is not an academic integrity check, it is a campaign by lowlifes to take her out of a position where they don't want her to be.

We're going in circles here.

>If you don't understand why she is being closely scrutinized now, just check the front pages of major newspapers in the last weeks.

So far as I can tell, she's being scrutinized because she was involved in some controversial stuff. That is perfectly consistent with the "Controversial figures get more scrutiny" explanation from my previous comment.

>That is not an academic integrity check, it is a campaign by lowlifes to take her out of a position where they don't want her to be.

You didn't answer my previous question. What should we do in cases like this? I'm sure a non-zero amount of the scrutiny is from "lowlifes to take her out of a position where they don't want her to be", but the same could be said for people digging up dirt on politicians.

I will tell you what could be done. Don't accept any kind of "academic integrity check" at this point. However, put all these questions on a file and check again at least a year from now when the issue is not on the newspapers anymore. Then I want to see who is really interested in the academic issues per se.

Reply on news.ycombinator.com