Glass Epistemologies

Richard Carrier seems to have gone on a new — or at least relatively recent — kick.  He’s convinced himself that he’s come up with a set epistemology, and now frequently casts his criticisms in the mould of criticisms of an epistemology.  Like he does in this “epistemology test”:

Your epistemology might be broken. Here is one test to find out. And if that’s what you find, you need to repair that broken epistemology; and I have some tips here on how to do that. But the broader skills you need to master for a reliable epistemology I have already covered in Advice on Probabilistic Reasoning, A Vital Primer on Media Literacy, and How to Successfully Argue Jesus Existed (or Anything Else in the World). The latter aims at general principles only using “the historicity of Jesus” as a test case, like a working example, in the same way as I will use certain claims about Anthony Fauci here: the objective is not really those specific claims (in the process, yes, we will get some clarity as to what is actually true about those claims—and what not), but actually the general principles you must absorb—because you need to be applying those general principles to every single belief you have about anything whatever in the entirety of your life, before you exhibit any confidence in it.

Rather large claims.  And rather large claims that he’s making despite the fact that, in general, right now it would seem that what he really wants to do is talk about how some claims people are making about Fauci are wrong, or at least aren’t justified, or at least are misinterpreting what was going on.  I’ll get into those later, but Carrier first wants to take on a comment from someone on a previous post about vegetarianism to show the danger of flawed epistemologies.  And it’s important to note that he believes this about epistemology in general:

I just recently encountered another example of this, which I’ll go through before getting to the Fauci test case, to illustrate by disparate examples a simple point: we need to always be focusing on the epistemology grounding a bad argument, rather than simply focusing on fixing or catching the bad arguments. Yes, we also need to do the latter. But we even more need to do the former. Because there is rarely a good reason for us to have even attempted these bad arguments. There sometimes can be—even people using reliable epistemologies can overlook things or make mistakes. But very often that is not what is happening; but rather, the fact that you made a particular mistake, a really obvious one you should not have made, indicates a much broader problem that you should be addressing: the broken epistemology that motivated you to use, and indeed even fall for, those bad arguments in the first place. Because if your epistemology let you do that in that one case, it must be letting you do that in every other case. In other words, catching yourself trusting a bad argument means you have no reason to trust that any of your beliefs, about anything, are well-founded. Since you haven’t been deploying a reliable epistemology, every belief it has generated for you is unreliable. And that means you need a complete revamp, a whole new system check, a repair regimen on your entire epistemology. Stat.

While he gives a very minor nod to the idea that someone can be wrong about something for reasons that, at least, don’t entirely invalidate their epistemology, he does take the strong stance that if you have a belief that follows from a bad argument then that means that your epistemology might be entirely wrong and so nothing that follows from that epistemology is reliable.  So you, at a minimum, shouldn’t accept anything that is generated by your own epistemology until you have analyzed what went wrong in that case and corrected it.  While he doesn’t say anything here about what you should do when it’s other people who have made the mistake, surely they’d have to be considered unreliable as well and so you couldn’t rely on them until you knew why they made the mistake and they corrected any flaws in their epistemology.  Remember that.

So he starts with this comment from CP 9:

I’m disappointed you mentioned Eat This Book, because it’s unreliable, as evinced by the fact that it argues the whackadoo unscientific premise that plants are sentient. Citing that book therefore makes meat eaters look bad.

Carrier replies:

Please take more care in reading what I write. I specifically did not endorse anything in that book. I merely mentioned someone else said it concurred with my position, a fact I explicitly said I had not verified. That you didn’t read what I wrote carefully leads me to doubt you read that book carefully. It is also suspicious that you seem to be picking a single argument (the one you found the weakest) and ignoring the rest of that book as though you had thereby refuted it. This is a huge red flag for a common cognitive bias. That you misread even a single sentence I wrote does not bode well for you having correctly read and absorbed a whole book. So at this point, I am not confident you have correctly apprehended what that book actually says about plants. [Note that I set aside their use of emotional-targeting language in the framing of their arguments, like talking about being “disappointed” and this making me “look bad.” But it is worth noting that as well: this is a broken epistemology at work, replacing facts and logic with emotional shaming and browbeating, and pretending the opposite has just happened.]

So Carrier starts with what really should be his main response:  Someone pointed out that the book aligned with arguments that Carrier had made, and so he advocated for it on that basis.  I’m not going to go look up the context to see if that’s accurate, and so will just trust Carrier here.  But, really, that first sentence generally indicates a response where someone is somewhat guiltily apologetic about referencing something that might make such a bad argument.  At most, they might point out that that wasn’t the argument they purportedly have in common and so that argument can’t in any way reflect on the actual argument Carrier is using.  And then that would be it.  But Carrier can’t do that.  Instead, he for some unknown reason wants to defend the book.  However, at this point he still hasn’t actually read it, nor has he actually read the argument in question.  So to defend the book he has to rely on claiming that CP 9 got his statement so badly wrong as to think that Carrier was endorsing the book that Carrier can no longer rely on anything CP 9 says about the book, so he has no reason to think that that is the argument they are actually making.  That’s … not a defense of the book, although it is consistent with his epistemology.  He correctly cites that it’s a cognitive bias to pick one argument and then ignore the rest of the book, but that actually has no relevance to Carrier unless he thinks that CP 9 was using that to refute Carrier’s argument.  Which he isn’t.  So this really comes across as a knee-jerk reaction to being challenged.  Intellectually, Carrier has absolutely no reason to defend the book in any way, and yet has to do it, even with arguments that don’t in any way reflect the actual content of the book and so can’t defend it.  That’s a pretty strong strike against his own epistemology, wouldn’t you say?  You shouldn’t defend things that you have no need and no ability to defend.

Also, it is rather ironic that Carrier makes a point about ignoring CP 9’s language, because his own comments use an awful lot of emotional brow-beating and shaming.  And in this case, he’s literally doing that instead of providing facts and logic because he doesn’t have any, since at this point he admits that he hasn’t read the book or even the specific argument yet.  So it’s good that he’s ignoring the sin that he himself is committing in spades.

So after this, the conversation continues:

CP 9: I just found it odd that you would mention it. As for the stuff about plants, you can read it yourself to check. I don’t know what else to say except that if someone is getting basic facts like that wrong, then it’s pretty hard to take them seriously from then on.

RC: It’s really your job to present the evidence for your position. So you should be the one quoting the book. But I skimmed it just now and found nothing corresponding to what you said. You seem to be mistaking the actual argument made in it about plants. But even if you didn’t, taking the dismissal of a single bad argument as a refutation of an entire book is simply not sound reasoning. You should never have attempted it. That you did, suggests a problem with your epistemology that you need to fix. “Making up excuses to ignore an argument” would be at the top of my diagnostic list.

Now, it would be fair to say that if CP 9 is going to claim that the argument is bad, he should probably be defending it himself with quotes.  But the problem here is that CP 9 also, at this point, doesn’t seem all that bothered by it.  While Carrier was oddly vigorously defending the book without reading it, CP 9 is perhaps oddly not bothering to attack the book or continue his attack at all.  So while CP 9 could be wrong, all he’s saying here is that he still believes that the book says that and that that argument is so ridiculous that he can’t trust the book on anything else.

But note that Carrier goes on to try to defend the book.  And in a very odd way.  He says that he “skimmed” the book and can’t find the argument.  But then he goes on to say that he thinks that CP 9 is mistaken about that argument.  And yet, he never actually gives that argument, or any argument that CP 9 might be misinterpreting.  He just says that he thinks CP 9 has misinterpreted the argument.  More importantly, he’s not doing it in a more inquisitive way, saying that he can’t find the argument and asking CP 9 what argument he’s referring to.  Instead, he’s insisting that there is no such argument and pretty much accusing CP 9 of misinterpreting the book.  A book, we must recall, that he has at best skimmed.

Even worse, he’s now calling out CP 9 directly for taking one argument and using that to dismiss the entire book.  We must recall that just previously he, in fact, dismissed CP 9’s argument that the book made that bad argument entirely on the basis that CP 9 supposedly misinterpreted Carrier’s statement about why he referenced the book.  So Carrier dismissed or at least called into question CP 9’s comment about the book based entirely on one misinterpretation of CP 9’s about something completely different.  And yet here he’s using CP 9’s statement about the book not being reliable because of that one ridiculous argument against him despite the fact that CP 9’s reference is actually far more relevant to what CP 9 is dismissing than Carrier’s was.  So he’s dismissing CP 9 for making a bad argument that’s actually slightly better than the pretty much identical argument that he made.

And we cannot forget that that precise sort of reasoning is fundamental to his own epistemology.  If someone comes to a ridiculous belief based on their own epistemology, then their epistemology is unreliable and has to be fixed.  One shouldn’t, then consider a source that has a similarly flawed or unchecked epistemology reliable as well.  And if our source is unreliable, by Carrier we shouldn’t trust it.  So all CP 9 is doing is actually applying Carrier’s own epistemology.  To then be called out for having an invalid epistemology either strikes against his epistemology, or reflects that Carrier is more interested here in defending the book than in ensuring that his beliefs and therefore that his epistemology actually work.

The comments go on:

CP 9: I don’t think I made the argument that the book is refuted based on that one argument. It’s clear you’re not going to read the book. I’m sorry for wasting your time.

RC: Yes, you did make the argument that the book is refuted based on that one argument. That was literally the entire point of your first comment: that I should not have even mentioned that book (and the only reason you gave was that single argument); and your subsequent comment: that its containing such an argument indicates none of what’s in the book can be trusted. So you have now completely abandoned your original argument and started pretending you never made it, and instead make an emotional excuse to bow out of this exchange, with the passive-aggressive statement about it being “clear” I won’t read the book and saying you are “sorry” for wasting my time—all rather than presenting any evidence or sound argument in defense of any point you are trying to make. This is all a bad sign. It suggests you are working with a broken epistemology. And you need to do something about that.

Technically, CP 9 is right here.  In his first comment, he merely claims that it isn’t reliable because of that argument, which would not imply that all of its arguments are refuted or wrong, just that we couldn’t rely on them and so at a minimum would have to check all of them ourselves, and in his second comment he merely says that it’s hard to take them seriously based on that argument.  Both of these are the exact things that Carrier has done to CP 9 in this very conversation.  And then Carrier goes on to call CP 9 out for being passive-aggressive while ignoring his own actual aggressiveness.  Tone does not come across that well on the Internet, but CP 9’s response is entirely consistent with someone who said something to someone they actually respect, had them react angrily, and who is now trying to back out of the discussion because they don’t want to get into that fight — and not discussion — with the person they respect.  Yes, CP 9 makes the comment about Carrier not wanting to read the book which is more of an accusation, but he also seems to simply want to let it drop, and Carrier won’t let it drop, and continues to aggressively badger CP 9 into accepting that his entire epistemology is wrong or unreliable.  That’s certainly emotional language, and note that again Carrier has not himself made any argument in defense of the book.  He’s relying entirely on saying that CP 9 is just unreliable and so sees no need to actually defend the book with facts and logic.

And that’s not the end of it yet:

CP 9: If I made a mistake, shouldn’t I be allowed to correct it? You’re not being fair. I used flawed logic, you corrected me, I eventually accepted the correction and I appreciate it. There shouldn’t be anything wrong with that. It’s not backtracking, it’s called being reasonable.

RC: The problem is not whether you can correct a mistake. The problem is that you made the mistake: that indicates a systemic problem with your epistemology. It is a symptom of a much bigger problem you need to recognize and fix. You should never have engaged in such reasoning to begin with. And until you address that, you will continue to make these mistakes in every area of your worldview and belief system. Here are your failure modes: you started by denouncing an entire book because of just one argument in it. I caught you out on that. Now you have retreated to just saying it has one bad argument people should be aware of, which is no longer a denunciation of the book; and not the argument you started with. And that one bad argument does not appear to even actually be in the book; you’ve failed to adduce any quotations attesting it, despite several attempts at making your point here. [Plus again notice more emotion-targeting language, e.g. saying I’m “not being fair,” implying I’m not “being reasonable,” and other efforts to trigger and shame; I continued to let that slide.]

So, technically, CP 9 hasn’t accepted the correction.  He hasn’t said that Carrier is likely right and that the argument isn’t that, or that the book can’t be refuted based on that one argument alone.  And yet Carrier doesn’t call that out.  At all.  Instead, he’s still focusing on CP 9 having made a bad interpretation — that he’s never shown that he’s even made that interpretation — and that that makes his entire epistemology unreliable and flawed.  You know, the exact argument that in the last sequence he said was invalid.  Heck, even here he references that as a bad approach while at the same time doing it himself.

He also references the emotion-targeting language again, despite a) his using slightly different emotion-targeting language himself, b) that language being reasonable if CP 9 had corrected his reasoning and Carrier was still badgering him about it (as a model that properly takes criticism and corrects its flaws would be a valid one) and c) that he in fact didn’t let that language slide in the previous comment.

And the final exchange:

CP 9: As far as the plant argument, I didn’t quote the book because I didn’t think you wanted me too. But here goes: the title of a subsection is “Plants, Too, Are Sentient Beings” and in that it says “the ethical vegetarian sincerely believes that the plants he consumes in such good conscience do not suffer and have no interests of their own, but his conviction is neither as rational nor as empirically grounded as he supposes” because the scientific evidence entails “like all other living beings, plants have interests and actively pursue them” [and also that they react to damaging stimuli, etc.]. These are deepities: to the extent that they’re true, they’re trivial; to the extent that they’re profound, they’re false.

RC: Those are not deepities: the author is making a correct point about the circumstantial equivalence of having needs and pursuing goals without consciousness. They are not claiming plants are cognitively sentient. You keep making this mistake. Why? You are confusing two different things, and thereby misrepresenting an argument in that book. If moral value attaches to having needs and pursuing goals, then it attaches to plants. Q.E.D. Ergo a vegetarian cannot appeal to those properties as a reason not to eat animals. They may have other reasons for their conclusions, but that’s the point I made before about how you cannot rebut a whole book by cherry-picking a single argument: that book does not assume that’s the “only” argument it has to address, so you were acting irrationally when you acted like it did. They are merely there addressing one particular argument, not “the entire case.” You acted like it was the latter; and then completely misrepresented what that one argument even was, and even got wrong its salience. These are fundamental epistemological failures. Which returns me to my first point: you should not be making any of these mistakes. So you seriously need to examine yourself to answer why you are making these mistakes, and what you can do to change your epistemology so you stop making them—not just in this case, but in every subject of belief in the entirety of your life.

So, finally, Carrier receives the purported terrible argument and tries to refute it.  And provides no quotes on his own, and doesn’t give any quotes or anything to show that his interpretation is correct.  Meanwhile, just taking CP 9’s quotes Carrier’s interpretation … doesn’t seem that likely.  Mostly because the subsection specifically says that plants are sentient, and because the simple argument about interests and reactions to stimuli that is supported by the evidence cites is not one that vegetarians, at least in general, actually make.  So referencing that could only be to clear out some philosophical underbrush before getting into the real arguments.  And that that is what they’re doing is more credible if they were a neutral or vegetarian advocate rather than an opponent to vegetarianism, because it would be aimed at ensuring that no one makes the mistake in thinking that that is what vegetarians actually mean.  No, they don’t give moral values to animals based on that meaning of having needs and pursuing goals.  They mean it entirely in the sense of consciously having needs and goals that they can be frustrated and so feel misery in being blocked from satisfying and achieving those goals, and in reacting to damaging stimuli consciously and being miserable if they can’t prevent it or stop it.  So to use those terms in the way they exist in plants as if that was how vegetarians use those terms in their moral arguments would be at best equivocation (I dislike the term “deepity” because it’s mostly meaningless in most arguments, like in this one).  And since Carrier seems to use it that way, he at least risks that equivocation, if not himself then in those who read that comment.

So Carrier insists that CP 9 needs to fix his epistemology, when it does seem that a major flaw in Carrier’s epistemology is needing to defend his beliefs and actions, even when he could easily give it up and even when it’s pointless to do so.  It also seems like he takes these sorts of challenges personally and so ends up defending his own self-image, which is why he does so so aggressively.  As I just showed, he’s arguing for and against his own epistemology in the same comment, and is blissfully unaware of that, and spent a lot of time defending a book that at the beginning he expressed that he had no need or desire to defend.  That certainly seems like a flaw in his epistemology.

So now let’s move on to Fauci.  Carrier picks up three recent comments where people are calling out Fauci, and tries to deal with all of them.  This is the first:

The first of those is an obvious nothingburger, in which Fauci probably just expresses his interest in Facebook considering doing something to combat misinformation about the pandemic that threatens public health and safety. He does not there endorse anything Facebook eventually actually did; just that he thought doing something was a good idea. Which is entirely correct: Facebook should have done something to prevent its platform being used as a public health threat. One can certainly criticize whether what Facebook did actually succeeded or was crap; but Fauci had no control over that, and no email regarding his opinion of that has been found; so no opinion you have of Fauci can be based on that. So if you made that false equation, your epistemology is broken. Fix it. A reliable epistemology will take care to get correct what the evidence actually is (e.g. Fauci did not command or even comment on any specific Facebook policy or activity; all he did, so far as we can tell, was endorse their concern to benefit public health and safety) and reason without fallacy from that evidence (e.g. nothing negative can be inferred about Fauci’s honesty or competence from this email; it simply evinces him correctly doing his job).

A huge red flag for me is trying to oppose a view without ever actually clearly stating it, and Carrier doesn’t do that here.  He simply says that it’s perfectly reasonable for Zuckerberg to talk to Fauci about clearing up medical misinformation on Facebook, and insists that that doesn’t mean that Fauci was involved in what specifically was done.  However, one big problem with his defense here is that he’s assuming that even after Zuckerberg specifically solicited Fauci on it, and even after Fauci expressed interest, that there would be no further conversations on the topic.  So, Zuckerberg would have asked Fauci about this, and Fauci would have expressed interest, but they’d never talk about it again.  Yes, in this E-mail can’t show that, but it’s reasonable to think that there might have been more E-mails, or phone calls, or video calls, or all sorts of other things where they talked about it.  It’s not a smoking gun, but it would seem to make more sense that Fauci was involved in discussions about the specific details of what Facebook wanted to do than that he remained blissfully unaware of any of these things and Facebook just did stuff.  The original E-mail existing at all doesn’t really make sense under that idea (although it is possible, but that would show that at least one of them is incompetent or shady).

However, from Carrier’s own source, that’s not the bigger issue and bigger evidence.  The bigger issue is that something was redacted from that original E-mail, and was redacted because it would expose trade secrets of Facebook:

Zuckerberg, in a March 17, 2020, email to Fauci, offered his platform to help in disseminating information about the virus and mitigation measures. At issue here is another redaction of something a National Institute of Allergy and Infectious Diseases aide labeled as Zuckerberg’s “even bigger offer.” Fauci responded to the email saying he was “interested” in Zuckerberg’s ideas.

Critics noted that the redaction was deemed necessary not for the more-standard reason of “deliberative process,” but because of “trade secrets” — i.e. information related to a private business that is privileged or confidential.

“What’s the offer Zuckerberg made to Fauci?” Fox News host Laura Ingraham asked. “The redaction references ‘trade secrets.’ Must be challenged.”

While I’m not certain of the reliability of the “bigger offer” comments, something being challenged because it protects trade secrets suggests that it’s talking about the details of how Facebook does things, which in an E-mail like this would only be relevant if Zuckerberg was suggesting something that he was thinking about doing.  Again, maybe it isn’t, but it certainly seems more reasonable that that “trade secret” was a “Facebook has the ability to …” line than pretty much anything else.  So Carrier misses what the opponents are really concerned about.

I agree it’s not important, though, not because there’s nothing there, but because this is the exact sort of thing that Fauci should be getting involved in in detail.  It’d be more reasonable to call him a fraud if he wasn’t involved, not because he was.  He may or may not have approved what Facebook did, but unless it broke the law no one should care about how involved he was there.

So, now, the second one, which is about Fauci getting an E-mail from someone in a lab who suggested that there might be attributes of the virus that suggested that it was a modified lab sample, and offering to look into it.  Later, that person concluded that it wasn’t a modified lab sample.  However, while that was going on Fauci commented that there was no evidence that the virus was made in a lab, despite his actually having access to at least potential evidence that it was.

Carrier first tries to defend him by pointing out that the first theories were that it was deliberately released, and that seems false, and so now the later theories are that it was accidentally released, and while that’s still possible it’s not likely.  And Fauci never said that it wasn’t accidentally released.

This doesn’t absolve him from the issue above, so this is how Carrier defends that:

Which brings us to the Fauci email. Per The Washington Post, “In a Feb. 1, 2020 email,” which as they importantly note was very early “in the virus’s life in the United States,” the “immunologist Kristian G. Andersen wrote to Fauci stating that the virus had limited ‘unusual features’ that might suggest manipulation in a lab,” and offered to research whether that was the case. Fauci made no objection. Andersen subsequently found the evidence didn’t pan out that way; it turned out to be highly unlikely that the features she was looking at were engineered—the result was that Nature Medicine article I just mentioned. Until that study came out, publicly Fauci always framed the matter as “there being no real evidence” the virus was engineered, rather than there being evidence showing it wasn’t engineered—which was then true. Subsequently it was also true the evidence did indeed show the virus wasn’t engineered, and Fauci then correctly said so.

The problem is that we know that saying “There’s no real evidence” in the context that Fauci was in really does come across as a denial, not simply being careful.  It’s basically weaseling here to say that “Well, he never really said that there was no evidence, but that there was no real evidence”.  But in both words and intent he clearly wanted people to not think that that was the case and so believe otherwise.  So, yes, that would be misleading people given that at that point Fauci did have reason to think that it was engineered.

Still, I agree that this isn’t a problem, because Fauci was also doing politics there and there would have been nasty consequences for suggesting it, and he didn’t know that it was true.  So minimizing that may not be entirely honest, but it’s a reasonable political move.

Finally, masks.  Fauci early on made a statement that people didn’t need to wear masks, and then said later that he knew from the start that masks would be useful — when he was trying to defend mask mandates — but wanted to ensure that people didn’t buy them all up and deprive medical workers of them, and then there’s another source that points out that he told an official that she didn’t need a mask and that it wouldn’t do much anyway.  Here’s Carrier’s defense of that:

This “doesn’t really show Fauci saying anything privately that he wasn’t saying publicly.” Hence “it would probably be more concerning if he had been telling health officials like Burwell something different from what he told the general public. But he didn’t.” Indeed, pay attention to the contextual details: “particularly since you are going to a very low-risk location” means Fauci is not giving Burwell general advice that would apply in a pandemic zone (he’d thus have advised she wear a mask if she went to a high-risk location; which not long after became the entire United States), “drug store” masks weren’t known then to be effective against particulate virus (which was at that time true), and wearing a mask is for “infected people to prevent them from spreading infection” (ditto). Later, scientific studies and case study evidence proved that that accepted science was wrong, or not as applicable to the peculiar nature of Covid-19, finding instead even ordinary masks not only reduce infection in the uninfected but even reduce the severity of infections (thus resulting in fewer hospitalizations and deaths among those who are infected). At the same time, it was soon discovered that Covid-19 had unusually long periods of asymptomatic infection, and unusually large numbers of asymptomatic infected. Thesefactschangedeverything.

I think Carrier’s right here that it was later evidence that suggested that universal mask wearing would actually provide benefits rather than at the beginning, and that the reason was because we have more asymptomatic cases and so need more protection.  I also definitely agree that at the time regular masks like the cloth masks we use now weren’t seen to provide much benefit.  However, the idea that he even mentioned wanting to keep masks for medical professionals seems odd.  If he really didn’t think they would help much, then why care about them being taken away from medical professionals?  So the most charitable interpretation is that he didn’t think that simple masks would work but professional ones would, but knew that suggesting mask wearing would get people buying professional masks and causing a shortage there.  But then at least publicly he should have simply said that at the time universal mask wearing didn’t seem like it would help, but then with new information it became clear that it would, without even hinting that early on he knew or suspected that wearing masks would help.  As I recall, that’s exactly what happened in Canada, where Theresa Tam said just that.  In saying what he did and implying that one of the reasons he didn’t call for a mask-wearing mandate or suggestion was because he didn’t want people buying up all the masks, it implies that he knew that they would work and lied to people — or at least misled them — in saying that they wouldn’t so people didn’t need to wear them.  Again, that’s not really a concern — no one really thought masks would be that useful at the time he said they wouldn’t be — but if people get the impression that he’s being inconsistent in what he says he really only has himself to blame.

So I agree with Carrier that these aren’t big deals.  However, Carrier doesn’t seem to get what the concerns are and still, as is normal for him, aggressively declares his opponents wrong.  That seems to be a common flaw in his own epistemology that he might want to address before he goes around telling others to fix their epistemologies.