Summary: I criticize David Thorstad's reply to me regarding his citation errors. I summarize his supporters' credentials as evidence about problems with academia, think tanks, and the world’s current thought leaders. There's a serious problem with scholarship standards, including misquoting, among credentialed intellectuals.
Introduction
In Criticizing "Against the singularity hypothesis", I criticized the content of David Thorstad's paper. In Checking Citations from David Thorstad, I criticized quoting and citation errors in the same paper. Thorstad tweeted replies about the quoting and citation issues but didn't reply to the criticisms of his ideas.
For context, Thorstad has a philosophy PhD from Harvard and works at an Oxford think tank (source). The main reason I’m writing this is because I think it reveals a lot about what kind of world and intellectual climate we live in.
The reason I wrote about Thorstad initially was because I let people at the Effective Altruism forum submit literature (that they liked a lot) to me for criticism. I checked three quotations and their citations from the paper partly because some people at the Effective Altruism forum denied that misquotes were a widespread or common problem.
Thorstad’s Tweets
Thorstad wrote four initial tweets in a row, plus one tweet replying to a supporter.
"Is Thorstad just one bad thinker" because ... I made two typos and said one word people don't like. Seriously? https://criticalfallibilism.com/checking-citations-from-david-thorstad/
This is a misquote. It suggests that I called Thorstad a bad thinker because of three criticisms which are mischaracterized (straw manned) as two typos and a disliked word.
Here’s what I actually wrote:
Is Thorstad just one bad thinker, while most intellectuals do better? Should we blame Thorstad personally? I don’t think so. Based on doing this kind of thing many times, I think Thorstad’s mistakes are pretty normal. There’s a widespread problem related to intellectual culture and norms. The attitudes of many people need to change, not the actions of a few.
Reading Thorstad, if you figured out that I was writing a question (which isn’t very clear), you’d think my answer to the question was yes. But it wasn’t. My answer was no. I said Thorstad is not “just one bad thinker”, but Thorstad is misleading his audience to believe that I claimed Thorstad is just one bad thinker.
Thorstad misquoted my article that pointed out his misquoting problem, and this misquote substantially changed the meaning of what I said. What I was actually saying is that Thorstad’s errors are representative and illustrative of what many other scholars are like. I was denying that the problem is Thorstad personally.
I also didn’t accuse Thorstad of making two typos and saying one disliked word. That’s an egregious mischaracterization. I accused him of making three errors related to quotes and citations. I didn’t say any of the errors were typos and I didn’t think any of the errors could be explained away as merely typos.
The disliked word comment is vague but I think Thorstad means that the paper noted something, but he said they “lamented” it. As the lead in to a quote, that misrepresented the meaning of the quote. The people Thorstad was talking about expressed themselves using a neutral word. Thorstad misrepresented their neutral writing as highly negative. Thorstad is welcome to form his own opinion, but not to speak for others and abuse quotations to mislead readers about what they said.
Back to quoting Thorstad tweets:
Correction, one mistake. The year on Good isn't wrong: https://www.sciencedirect.com/science/article/abs/pii/S0065245808604180
Here, Thorstad is reiterating his error about the year that Good’s paper was published. It was 1965, but Thorstad mistakenly cited it as 1966. (This is one of the errors he mischaracterized as a typo earlier. But if he’d merely made a typo, he wouldn’t still be defending 1966 as the correct year.)
Instead of learning from my correction, Thorstad has refused to change his mind. He still thinks it’s 1966. Why? He found a modern source that says 1966 and cherrypicked that information. It looks like he really wanted to be right and get me back for accusing him of making errors, so instead of doing unbiased research to find the truth, he just found the first thing he could that backed him up and then he claimed he was right all along. (Or maybe, rather than it being bias, he just doesn’t know how to do effective research on an issue like this.) I don’t think he noticed the text “Copyright © 1965” on the page he linked, which is a hint that 1966 might be an error.
It’s contradictory to, at the same time, claim it’s not a big deal but also be defensive enough to incorrectly claim your error was actually true. If it doesn’t matter, why go find some false information to try to defend yourself with?
How can we settle this issue definitively? Previously I linked a scan of an old reprint showing the date 1965. I thought that was convincing but apparently Thorstad didn’t. I didn’t expect this point to actually be disputed or I would have given more evidence. It’s pretty easy to do a better job of looking it up now that there’s a dispute.
The Internet Archive scanned the original document. It says 1965 on the title page.
It’s one thing to make a factual mistake. It’s much worse to refuse a correction and reiterate the mistake. Even if the original mistake wasn’t a big deal, Thorstad’s reaction to it matters.
Wait a moment ... my "typo" was correcting a grammar error someone else made and not writing [sic] like a jerk!?
What is “typo” a quote of? If it’s me, it’s a misquote. If it’s of himself in his earlier tweet where he wrote “typos”, that’s odd and unclear. It seems like he’s saying that I accused him of making a typo, and using quote marks, but I didn’t say that.
This writing is ambiguous. By “and not” does Thorstad mean “without” or “instead of”?
If Thorstad means “without”, then he’s wrong. You use [sic] when you don’t make edits to fix typos, not when you do.
If Thorstad means “instead of”, then he’s presenting a false alternative. He could have made the correction using square brackets or he could have given the exact quote without writing [sic]. Those were both reasonable alternatives. His choices weren’t just using [sic] or stealth editing a quote.
Also, writing [sic] is fine and doesn’t make one a jerk. [sic] is a scholarly tool used as part of literal, accurate quoting. Having a negative attitude towards using [sic] shows a bad attitude towards quotation literalness and accuracy.
Thorstad seems to be implying/confessing that he edited the quote intentionally. He’s denying it was a typo, or in other words denying it was an accident. I believe intentionally editing quotes (without using one of the few allowed exceptions or properly indicating the edit) is an ethics violation that is against the honor codes at many universities. I actually think the written rules about quotes and cites are often reasonable. The existence of those rules is then sometimes used to deny there’s a problem with intellectual culture. Surely whatever the rules say to do is what most people do, right? Sadly, no. Systemic reform is needed so that most intellectuals actually want to follow those rules and value them instead of considering them overly pedantic.
I didn’t think it was just an accidental typo. I thought it was due to some sort of bad attitude or other problem with Thorstad’s ideas. Thorstad has confirmed that I was right.
@curi42. Sad face.
I appreciate that Thorstad tagged me. Without a notification, I wouldn’t have seen these tweets.
I don’t appreciate the “Sad face” comment, which I read as unserious and mean. There are many similar comments – in terms of both style and content – in the replies to Thorstad.
As context for Thorstad’s fifth tweet, Dan Carey replied to Thorstad:
Odd for them to be so nit-picky about citations. This is from their criticism of your piece. Arn't all of these ideas addressed explicitly in Bloom et al (2020) which u cite in their quote of u right above?
I’ve left out an uncropped phone screenshot showing mostly the Error One section of my Criticizing "Against the singularity hypothesis”. Thorstad wrote back agreeing with Carey:
I know, right? I cite my sources like a normal person
Calling me an abnormal person is a social insult. One of its purposes is to pressure me to increase my conformity. Instead of arguing that his way of using citations is good or criticizing my way, Thorstad instead calls his way normal. But calling something normal isn’t a rational argument that it’s good. And I already said Thorstad is normal in my article! He doesn’t seem to comprehend that I’m criticizing something I consider normal instead of trying to fit in and be normal myself.
My article made statements like “I think Thorstad’s mistakes are pretty normal”. My point was that there’s a widespread problem with intellectual culture, not a problem with Thorstad individually. I was saying that what’s currently normal is problematic and should be reformed. In that context, Thorstad asserted something is normal and assumed that makes it good, which is the logical fallacy called begging the question (which basically means assuming a conclusion about one of the points currently under debate, like you don’t understand that it’s being disputed).
Carey’s and Thorstad’s claim is factually false. My points in that section are not “all” addressed “explicitly” in Bloom et al (hereafter just Bloom). I’ll give one example. I said “The healthcare industry, as well as science in general (see e.g. the replication crisis), are really broken”. I text searched the Bloom paper for “crisis” and “replic” but found no discussion of the replication crisis to address my argument. I also skimmed but didn’t see it covered.
Also, even if Bloom had “explicitly” addressed “all” my points – which they didn’t – there was no way to know that from Thorstad’s writing. Thorstad cited Bloom as one of three sources for the claim “As low-hanging fruit is plucked, good ideas become harder to find”. Thorstad provided no information about Bloom addressing specific counter-arguments to a claim (about healthcare research productivity) that Thorstad brought up in a later paragraph with a different citation. In general, you need to repeat a citation each time you use it for a different purpose or else give some kind of explanatory comments.
The Credentials of Thorstad’s Audience
Intellectuals commonly tweet with their real names and state their credentials and employer in their public profiles. I want to review their credentials to show what kind of people make or side with scholarship errors. I’m leaving out their names and refutations of what they said, but if any of them wants to debate me, I can elaborate.
First I’ll share credentials of people who wrote reply tweets. None of their tweets were significantly better than Thorstad’s tweets that I analyzed above. They generally seemed similar to Thorstad although some were worse.
There’s a PhD student at the philosophy department of the London School of Economics (which was founded by Karl Popper), a grantmaker in global priorities research and international policy at Longview who has a philosophy PhD from Rutgers, an assistant professor of philosophy at University of Sheffield who is a fellow-in-residence at a Harvard center, a Ph.D. student at Pardee RAND, an associate professor at University of Pennsylvania, an employee at OpenAI and a PhD student at Oxford.
Second, here are some credentials from people who clicked the like button on Thorstad’s tweets. There’s a Harvard medical doctor PhD student, a University of Minnesota philosophy professor, an executive research coordinator at Rethink Priorities, a philosophy professor who wrote a book, a professor of mathematical statistics, two AGI safety researchers, a philosopher working at Rutgers with a PhD from Ohio State, a philosopher of science at University of Groningen, and a philosopher with a PhD from Cambridge. There’s also a philosopher at University of Bristol, whose name I recognize because he wrote a Bayesian textbook that I’ve looked at. I was trying to find a book explaining Bayesian epistemology premises and reasoning in a way I could engage in debate with, but it wasn’t suitable.
(I gathered these credentials a few months ago. Some could have changed before publication.)
Thorstad Reinforced My Point
Thorstad and the people who replied to his tweets reinforced my point (and they seem unaware of this). My point was that many scholars are OK with misquoting, which is an example of how intellectual culture in general is broken and should be reformed. The people tweeting all seem to think misquoting isn’t a big deal, and that people (like me) who consider it a big deal should be shunned. So they are examples for my claim.
The typical response I get when I criticize misquoting in general is being ignored or being told that everyone already agrees with me that misquotes are bad so there’s no point in talking about it. But the typical responses I get when I criticize specific examples of misquoting are hatred, denials that misquotes matter, etc. I’ve written various things about this including Misquoting and Scholarship Norms at EA and EA Misquoting Discussion Summary.
Rather than say his mistakes were accidents, Thorstad reiterated them. He didn’t retract his inaccurate “lamented” paraphrase. He repeated his false claim that 1966 was the true publication year. He ignored the issue that he changed the type of quotation marks used within a quote. And he implied that he edited the Chalmers quote on purpose, not as an accidental typo, because he thinks that correctly using [sic] makes one a jerk. Thorstad also agreed with a commenter falsely claiming that Bloom had answered all my arguments about a particular issue already.
Who has bad attitudes to scholarship including low standards for quotation accuracy? A lot of the problem is PhD students and people who already have PhDs. These tweeters and the people who like the tweets are (not exclusively) a bunch of graduate students, professors and think tank employees. Many of them went to or work at good universities. Thorstad got his PhD at Harvard and works at Oxford.
In articles like Ignoring “Small” Errors, EA Should Raise Its Standards and “Small” Errors, Frauds and Violences, I explained my view that dismissing errors as “small” can result in large problems like incorrect conclusions. In general, you can’t actually judge which errors are large or important until after you find solutions. In retrospect, you can see how much work they were to solve and how much difference the solution makes. But before you understand the issue, you don’t know how big a deal it is.
Also, I don’t think all small or picky points matter. I say an error is a reason an idea will fail for a goal (that someone involved has). I just think some particular errors, like misquotes, matter to goals like having productive discussions. Lots of “little” errors in discussions add up to the current, widespread problem of most discussions becoming inconclusive messes. I think higher standards for some fairly objective issues like quotation and citation accuracy (as well as logic, reading comprehension, factual accuracy and using arguments instead of insults) is relatively easily achievable and would actually help a lot.
I’m aware that some people will respond to this article by thinking I’m even more unreasonably pedantic than they thought before. I wrote this anyway because I disagree and I think I’m commenting on some important, widespread social-intellectual problems that are making society, philosophy and science worse.
Conclusion
I wrote about how misquotes and citation errors are a widespread problem. The response – primarily from credentialed intellectuals – was basically to mock me for caring about details like those. In other words, people agreed with my position that many intellectuals, like them, think inaccurate quotes and citations are OK. Credentialed intellectuals aren’t even close to as rational as the general public views them as. This is one of the factors making it really difficult to have productive discussions about intellectual issues.
While it may be tempting for many academics (who aren’t Thorstad’s Twitter buddies) to place the blame on Thorstad, I believe it’s a widespread issue and Thorstad is merely a typical example. I’ve seen similar attitudes in many other places, and I’ve failed to find better attitudes anywhere. If you know of a group with better attitudes, please tell me.
Do you disagree? I’m open to organized, serious debate following explicit methods and aimed at reaching conclusions. See my debate policy.