Subscribe now

Even longer ago, I wrote a piece which just summarized some of the main points of Nick Shackel’s “The Vacuity of Postmodernist Methodology.” I do stand strongly behind that piece, which is just to say that Shackel’s paper is really good, and I just wanted to tell everyone about it. I still do. Read it, please! While you’ve probably never heard of Shackel or his paper before, this is the paper where the idea of a “motte and bailey” doctrine was introduced. That concept has become quite influential, and Shackel deserves credit for originating it.

For those who are perhaps unaware, the idea of the “motte and bailey” is a metaphor that comes from medieval warfare. A motte is an impregnable fortress, located in the middle of a bailey, some desirable territory that is otherwise difficult to defend. A motte and bailey defensive doctrine is to abandon the bailey in the face of any serious attack and retreat to the motte, where the defenders can rain arrows on the attackers until they have to run away, at which point the bailey can be occupied again. Shackel’s idea is that some people argue in this way. They have a set of easy-to-defend truisms, the motte, alongside a set of ambitious and indefensible claims, the bailey. The fact that the bailey claims are indefensible will invite criticism, but when the critics arrive on the scene to make trouble, the motte-and-bailey rhetorician retreats to the motte, and insists that they’re only asserting the most uncontroversial of truisms. Radical claims resume once the skeptic is out of earshot.

Put in terms that are this straightforward, it seems like it would be impossible for the motte and bailey to work, particularly in written scholarship. You’re not just asserting the uncontroversial truisms; the controversial claims are right there, on page 12! But motte and baileys do exist, because those who deploy them are clever. The motte and bailey are carefully constructed. The key ingredient in constructing a motte and bailey doctrine are sentences that Shackel calls “troll’s truisms.” These are sentences that are ambiguous, in the sense that they can be read in at least two distinct ways. On one reading, the sentence states an obvious truism. On the other reading, it says something much more radical. The motte and bailey happens when troll’s truisms are strung together. When the skeptic challenges them, the defender points to the obvious, uncontroversial reading. But it’s quite clear from the use that these troll’s truisms are put to that the more radical reading is intended.

There are several ways to construct a troll’s truism, but one central way is by the use of an equivocating fulcrum. An equivocating fulcrum is a key term that has two different meanings, and which the speaker will strategically vacillate between. Sometimes it is clear that one meaning is the only one that makes sense, sometimes it is the other meaning, and sometimes it is genuinely indeterminate. In this way, riffing on an equivocating fulcrum generates the set of troll’s truisms which enable the motte and bailey. Of course, if this ambiguity is brought into clear focus and terms carefully distinguished, the argument falls apart. This kind of call for disambiguation is viewed as deeply suspect by practitioners of the craft. We can illustrate this by providing an example, so I’ll cut to the chase and give the example which is my central focus here: the term ‘objective’ and its cognates (‘objectivity,’ etc).

Subscribe now

The term ‘objective’ can have two distinct meanings. One thing that ‘objectivity’ might refer to is a property of people. Objectivity, in this first sense, contrasts with biased; a person is objective just to the extent that they are not biased in their investigations or evaluations. The other thing that ‘objectivity’ might refer to is a property of facts. Objectivity, in this second sense, contrasts with relative or subjective; a fact is objective just in case its truth does not depend on any framework, perspective, or anything similar. People will sometimes talk (usually dismissively) about “capital-T truth.” I don’t know what capital-T truth is supposed to be, as distinct from lowercase-t truth (so far as I can tell, they’re both just truth), but if anything is capital-T true, it is objective facts.

Things are complicated further because the first kind of objectivity, the objectivity of persons, comes in degrees. People can be more or less biased, and thus more or less objective. ‘Objective’ is in this way a gradable adjective. But more than this, it is an absolute gradable adjective, because the scale of objectivity has an upper bound of being perfectly objective. ‘Flat’ is another term that works in this way. A surface can be more or less flat, and there is such a thing as being perfectly flat. Few things, if any, are perfectly flat; usually, when we call something flat, we just mean that it is flat enough for our purposes. But we can insist on a strict reading of the word ‘flat’ - we’re talking perfectly flat - and when we do that, it can be clear that nothing is perfectly flat. But this doesn’t mean that it’s false to say that the table is flat.

So similarly, we can think about what it would mean for someone to be perfectly objective. This would be someone who is not at all biased. So what are the possible sources of bias? Desires and interests, most obviously. Wanting something to be true has a strong effect on how we investigate or navigate the world. But beliefs are also a source of bias. Our senses don’t present reality to us pure and unfiltered; we interpret the data of our senses, and do so through the filter of our various preconceptions. So for someone to be perfectly objective, they would have to have no desires, no interests, and no beliefs. But of course that is impossible for human agents, we cannot have this kind of “view from nowhere.” So this kind of perfectly unbiased perspective is impossible.

This gives us our troll’s truism: “Objectivity is impossible.”

“Objectivity is impossible?” the skeptic asks. Surely not! “Well, of course objectivity is impossible,” comes the reply, and the argument for this is that everyone has interests and preconceptions, so objectivity is impossible. But when the skeptic drops this line of objection, the equivocation starts. ‘Objectivity’ (of persons) could mean perfectly objective or it could mean objective enough (for present purposes). Once we’ve established the motte that it’s impossible for a person to be (perfectly) objective, a certain kind of person will rush to claim the bailey, that all thinkers are hopelessly biased and thus systematically untrustworthy.

And it doesn’t stop there. Once we’ve established that objectivity is impossible, a further bailey awaits: that there is no such thing as capital-T truth, or that all such truth is hopelessly outside of our ken. The scientist might aim at discovering objective truths, but these aims are naive at best and pernicious at worst. Objectivity is impossible; haven’t you heard?

Subscribe now

This line of argument has been advanced by many people as a rebuttal to Boghossian and his coauthors over the last few days. Boghossian insists that objectivity is essential for scholarship. But is he really so ignorant to the fact that he has various interests and preconceptions? He is no more objective than anyone else, and his conclusions are no more objective than anyone else’s. He contradicts himself by demanding objectivity, because the standards that he attempts to apply in his demand for objectivity are themselves not objective. And how could they be?

That argument might sound compelling at first blush. But go through that last paragraph and, for each instance of the word ‘objective,’ try to pin down whether I was using that word to refer to a property of people or a property of facts (and, if a property of people, what kind of property; the absolute version or the “good enough” version). It’s not easy to do. There’s a lot of sliding, and a lot of genuine ambiguity in how the words were used. That’s how an equivocating fulcrum creates troll’s truisms.

The flaw with this kind of argument can be made clear by contrasting it with a standard example of objective knowledge. “Here’s a hand,” I say, as I wave my hand in there. “And here’s another,” as I wave the other. Those are objective facts. There are hands here, really, and the existence of these hands doesn’t depend on our frameworks or conceptual schema or anything like that. And not only are there those objective facts, but I know them to be true. I may be a biased individual in many ways, but my biases, in the form of both my interests and my preconceptions, in no way threaten my knowledge that here is a hand, as I wave it in front of my face.

So while some claims might be subjective or framework-relative, some are not. “Here’s a hand” is just true. And while I might not be perfectly objective, I’m objective enough to know that here’s a hand. So objective facts, and knowledge of objective facts, are both possible.

Of course, that there are some objective facts and some knowledge of those facts doesn’t mean that all facts are objective or that knowledge of them is just as trivial as my knowledge of my hand in front of my face. But once we’ve established that objectivity of these kinds is possible, it’s no longer relevant to point to general theoretical considerations to rule out objectivity in any particular case. If objectivity exists in some cases, why is this case not one of those cases? Answering that question involves engaging with particulars.

And so in the case of Boghossian, it’s not enough to say that he must be incoherent because objectivity is impossible. Boghossian has objective knowledge of the hands in front of his face. Why might he not also have objective knowledge of the possibility of objective knowledge in the humanities, or objective knowledge of the importance of scholastic standards which are directed at attaining such objective knowledge? “Because objectivity is impossible” won’t cut it. Nor will “because Boghossian has biases.” Perhaps the best answer to that challenge is “Because the humanities are not in the business of providing any kind of objective knowledge.” But then what are the humanities in the business of?

Subscribe now

Several scientists have alleged over the years that there is widespread misunderstanding and misuse of natural science in the work of prominent humanists. Philosophers have worried about the unquestioning embrace of problematic philosophical views, especially those concerning truth, evidence and knowledge. More recently, many different voices have suggested that humanistic disciplines have allowed background ideological values to distort the objective pursuit of knowledge in those fields.

To help us assess whether, and to what extent, there is a problem here, we charge a commission of eminent scholars from these disciplines to examine the state of scholarly work in their respective areas and to evaluate whether these allegations are justified.

And a commission of eminent scholars was, indeed, empaneled. 10 scholars from different areas of the humanities are listed as authors of the report, 4 of which are philosophers. And those four - Paul Boghossian, Kit Fine, Kwame Anthony Appiah, and Gideon Rosen - are among the most senior, respected philosophers working today. A murderer’s row, if you will. Boghossian in particular deserves mention, both because he served as the chair of the committee, and because some sections of the report clearly have his fingerprints all over it, as many of the arguments are brief restatements of arguments Boghossian made earlier in his 2006 book Fear of Knowledge.1

Bottom line: the committee concluded that there is, in fact, a problem with the humanities. The problem is not quite all-encompassing - there is no area of the humanities in which good work is not being done. Yet some disciplines (anthropology in particular is mentioned) produce an abundance of bad scholarship.

The introduction to the report likens the situation to one in which scholars in the astronomy department stop doing astronomy and start doing astrology. If it’s just one scholar going awry, the department can correct or fire him. If the whole department starts doing astrology, astronomy departments in other universities can call out the rogue department and marginalize it. But what if the entire field - all departments at all universities - is given over to nonsense? Well, then, some kind of outside intervention is called for, by serious and appropriately humble scholars from adjacent disciplines, working in tandem with administration. The report aims to advance such an outside intervention.

I think this report is great. I’ve seen a few criticisms of it online, and I’ll address those in a bit. Some of them make decent points - the report is not perfect - but on the whole, I love it. Let me explain why.

Subscribe now

The mission statement of the report shows that the focus of the report is more limited than just a broad “What’s up with the humanities?” It’s not a report that tries to show why wokeness or cancel culture (or anything like that) are bad. Rather, it’s about the quality of the scholarship in the humanities. From the top of Section 3 (the first substantively critical section of the report):

Our focus is rather the quality of scholarship: the research produced by professors employed by colleges and universities and published (for the most part) in academic journals and scholarly monographs. The critique we take seriously is that this scholarly enterprise has been damaged in recent decades, not just by a general erosion of standards, but also by a reconceptualization of scholarship as a form of political activity, answerable in part to extra-academic standards.

The concern is with whether or not scholarship in the humanities actually produces knowledge or understanding, or whether it is entirely bogus. Astronomy produces knowledge. Astrology does not. This makes works in astronomy potentially good scholarship. But astrology cannot be good scholarship. This is not to say that there are no norms for how astrology should be conducted. Among astrologers, there are doubtless standards for what counts as “good astrology” or “bad astrology.” But whatever merits good astrology might have, it does not generate knowledge or understanding. There are norms, but from an academic perspective, those norms are bogus. The central contention of the report is that the norms for scholarship in much (not all!) of the humanities are also bogus in this way. “We do find grounds for concern, not just in individual disciplines, but systematically.”

The problem is political - that scholarship is being done with the end of promoting certain progressive political goals related to “social justice.” The authors are quick to say that they do not think that social justice is necessarily bad or that it’s inappropriate for academics to promote it, either as part of their work or as activism related to their work. Rather, the problem is that work in the humanities has oriented around this social justice consensus, and in doing so have placed social justice above understanding as the guiding light of their scholarship. “The deeper problem is that an artificial consensus of this sort can only be maintained by distorting the scholarly ecosystem in ways that are profoundly damaging.” And here’s where things start to get philosophically interesting.

The entire conceptual framework thus far presupposes that there is such a thing as getting things right in humanistic scholarship and getting them wrong. There must be something that is correct - genuine knowledge or understanding - in order for distortions to genuinely count as distortions. The term “objectivity” always starts to get thrown around at this point, and the authors treat that term carefully, since it’s so vague and contested. But surely there must be some standard of correctness in humanistic research, or what are we even doing here? Moreover, many of the mechanisms by which the “artificial consensus” of social justice are enforced presuppose the existence of correctness standards. Controversial scholarship is almost never explicitly rejected for having bad politics, but rather for being bad scholarship, for not living up to disciplinary standards. Rejecting bad scholarship as bad is, of course, totally appropriate. But that just shows that we’re agreed that there are correctness standards of some sort. And this is what the report is worried about distortions away from.

So, whence these distortions? In Section 4, the authors identify three “tracks.” On the first track, scholarship is sometimes rejected as being inconsistent with “settled science.” On the second track, the goal of understanding is displaced by other political goals, with work having to clear a political bar in order to count as good scholarship. And “On the third track, the idea that there are genuine facts about the world or about what the evidence supports independently of our political commitments is rejected. On this view, good scholarship cannot be distorted by political values because it is, at bottom, irredeemably constituted by such values.”

This is where the report genuinely becomes the “Boghossian report,” for what follows next is a scathing evisceration of various relativist and constructivist views that are widespread in humanities scholarship. A list of quotations follows from humanists endorsing one version or another of relativism. Relativism is the entire basis of the “third track,” but it’s also a convenient tool for those on the “second track;” why pursue truth rather than political goals if truth is a mirage and political goals are so important? Relativism is then criticized as being self-undermining and incoherent, with the criticisms being largely a recapitulation of the central arguments made in Boghossian’s Fear of Knowledge: Against Relativism and Constructivism. (I’ll elide a discussion of the ins and outs of Boghossian’s argument here. That’s a bit more in the weeds than I want to get. Suffice it to say, I think he’s completely correct, and that Fear of Knowledge is a great book.)

And what of the first track? Well, once relativism is dispatched, the flaw with the first track is easy to see. There are facts, and we can attain knowledge of those facts through the use of rational methods. Once this is recognized, we can see that scholarship in the humanities should consist in pursuit of those facts using rational methods. Claims to “settled science” can be made, but only if they live up to those rational standards. The problem is one of making those claims without putting in the relevant work to establish a firm rational basis.

That’s the basic rundown. So, what makes this good?

Subscribe now

Much of the critical discussion focuses, in one way or another, on the fact that the report spends so much time debating relativism. Really? The problem with the humanities is relativism? It has not escaped the critics’ notice that Fear of Knowledge was published 20 years ago, and at the time was already a bit stale, since academic debates over relativism and postmodernism were at their hottest closer to 30 years ago. We have, by this point, long since moved on. Boghossian may be eminent within the philosophical community, but is this report little more than him grinding an old axe?

To understand the full context for this criticsm, we need to turn our attention back to the history of the debate over relativism and postmodernism and some attendant divisions within the social structure of the academy. We need, unfortunately, to talk about “analytic” and “continental” philosophy.

To briefly recapitulate part of my last post on the subject, philosophy experienced a kind of divergence in intellectual traditions starting around 200 years ago, post-Kant, with that divergence accelerating sharply about a century ago. It’s nearly impossible to give a characterization of analytic vs continental philosophy that neatly divides the two camps in a principled way; as such, it’s better to think of two different research traditions. Who is reading whom, and who is citing whom? Analytics (mostly) read and cite other analytics, continentals (mostly) read and cite other continentals, and there’s very little overlap. Consequently, there’s been a divergence in writing styles, in technical terminology and attendant conceptual apparatus, and in methodological assumptions. Again, it’s a mistake to try to draw the distinction explicitly in terms of writing style or any of these other factors. But there are two clear clusters of influence pattern.

This is relevant because postmodernist (and “poststructuralist”) philosophers are all within the continental cluster. And it is also relevant because, to the extent that scholars in other areas of the humanities have engaged with philosophy, they have done so by engaging with continental philosophers. So if we expand our influence-pattern analysis to look not just at the philosophy discipline itself, but all of academia, we’ll see that the humanities tend to cluster quite closely with the continentals. (You might expect the sciences to cluster with the analytics on the other side, and that’s partly true, but the connection here is looser, given scientists’ tendency to ignore philosophy.)2

And this, in turn, matters because of how things played out in that debate in the 90s over postmodernism. To oversimplify somewhat: postmodernist philosophy was developed in the 60s and 70s within the continental cluster. No big problem there yet, there are all kinds of philosophical movements that arise within philosophical communities all of the time. But as postmodernist ideas became widespread within the continental cluster, they began to expand into other areas of the humanities aggressively throughout the 80s. (These timelines are vague, of course, but should give you a rough sense.) The analytics (and their erstwhile allies in the sciences) had, to this point, been mostly content to let the continentals make their own silly errors. But with their influence spreading to other areas of the academy, perhaps it was time to take the problem more seriously. The postmodernists, for their part, thought that they had discovered some rather serious problems with science and analytic philosophy. This led to the heated “science wars” of the 90s. Continentals and their allies in the humanities accused the sciences of various political sins that scientists allegedly excused and covered up with their false pretentions to “objectivity.” Scientists and analytics responded in kind by attacking the postmodernist theories that were the basis of the humanists’ critique.

The ultimate resolution of these wars was a stalemate. Neither side convinced the other. The main effect was an increase in the intellectual divisions that existed between the camps. Both sides understood the other as being fundamentally mistaken, and retreated into doing their own work in their own mold, with the other side existing primarily to serve as a punching bag whenever cheap shots needed to be doled out to one’s intellectual enemies.

Things continued to evolve from there. Analytic philosophy has gone through its own convulsions, as foundational positivist texts of the early 20th century fall further and further into history and disrepute, and analytics debate the meaning and viability of a post-positivistic “naturalism” and what role (if any) metaphysics might play in such a picture. Continental philosophy, for its part, has evolved away from its foundational postmodernist texts and taken on a more earnest and stridently political stance that is more explicitly committed to social justice. But these developments occurred largely in isolation from one another, on either sides of a wall that had existed for a long time, and which got very high indeed after the science wars of the 90s.

This brings me to one line of response I’ve seen to the Boghossian report, the response from the continentals and humanists. This response is to absolutely sneer at the report. Boghossian is just relitigating the debates of the 90s. And we won those debates! This is nothing but sour grapes, relitigating old defeats. But this time, he’s hoping to appeal to outside authorities, to the administrators, to tear down our departments. He didn’t win on the merits - he never could! - and now he’s coming back decades later to try to work the refs. Obviously this shouldn’t work, and if it does, it’s approximately the cheapest and most intellectually dishonest thing anyone has ever done. The fucking gall!

But of course the humanists didn’t win those debates in any meaningful sense. Politically, the war ground to a stalemate. And due to intellectual silo-ing, each side is convinced that they won the debate on the merits. But on the merits, the continentals lost, miserably. (“Of course you’d say that, you’re Team Analytic!” Well, yes, I am, but I’m Team Analytic because the analytics were right and the continentals were wrong. I reject utterly the idea that any argument must be hedged and qualified by what side you’re on. The arguments are what they are, and they work or they don’t. Boghossian’s arguments work. I won’t try to show that because then I’d just be rewriting his book and I’m sure I won’t convince any continentals who might be reading this.)

Now, some 30 years later, a popular clamor has reached the ears of the leaders of American universities. Something is rotten in the state of the humanities, many people are claiming. The leaders of the universities are worried. Is this true? Let’s find out!

So they ask Paul Boghossian (and company; all very much Team Analytic, although Appiah is more fluent in the continental tradition) whether there is an intellectual problem in the humanities. And his response is that Of course there is, Jesus Christ! We told you about this 30 years ago and no one in the administration really listened and those putzes even think they won the debate and now look where we are. This is “I TOLD YOU SO: The Report.” And, yeah, fair enough. There’s nothing sweeter than an “I told you so,” and in this case it’s quite well-earned. That’s why I love it.

Subscribe now

But I said that there’s a good version of this objection, so let’s get to that. The good version of this objection is that today’s humanists are not the humanists of the 90s. Relitigating these debates over relativism is fighting the last war. The enemy they’re attacking is no longer the enemy as it exists today. The report is blind to the shape of the contemporary threat.

This is somewhat true. But recall, the remit of this report is narrow. The question the authors address is about scholarship, in particular, about “the research produced by professors employed by colleges and universities and published (for the most part) in academic journals and scholarly monographs.” That choice of question frames the entire report. There are good things and bad things about this. The bad thing is that it prevents them from considering larger structural concerns about, e.g., academic incentive structures. The people who think that everything that is wrong in academia is a product of adjunctification are mad that the report does not consider the role that adjunctification has played in speeding the degradation of the humanities. And perhaps adjunctification has played a role here, and it’s bad that the report is blind to this. It is also similarly blind to concerns about funding streams, partisan educational polarization, AI use, the relationship between the humanities and the larger “woke social justice” movement, and a hundred other factors that are playing a role. So the report is really quite narrow. But is it too narrow? I dunno. Maybe. But going over every problem with the humanities would require a report that’s much longer than 30 pages, with a much deeper bench of experts to analyze every aspect of the situation. For a report like this, though, you’ve gotta pick your battles. And if you get together a panel of 10 people, 4 of whom are philosophers, then yeah, they’re gonna focus on whether or not there are fundamental philosophical problems with humanities scholarship. Which is what they were explicitly asked to do in the directive to the commission! So while I agree that more is needed on the topic, it’s not clear that more is needed in this report.

In light of this, the better complaint is that humanities scholarship just isn’t that relativist anymore. Again, the continental/humanistic tradition has evolved. And the current version of the humanities is much less indebted to postmodernist relativism. The humanities are not politically hostile to conservatives or conservative-coded ideas because conservatives are so close-minded and it’s all relative and what even is truth, man? They’re hostile to conservative-coded ideas because it is objectively true that conservatives are human garbage and have no place in the academy. So if there is a flaw with humanities scholarship, relativism isn’t it.

But has humanities scholarship really moved away from relativism? The list of quotations that the authors provide in Section 4 is relevant here. Some of them are from foundational texts from the 80s, but many are very recent, with citations to other recent texts. (If I have a criticism here, it’s that they should have gone into more detail here and provided a true laundry list of citations in the footnotes to drive the point home.) It’s undoubtedly true that an endorsement of relativism shapes huge swaths of scholarship in the humanities even today. It is not consistently applied - indeed, one of the central points of Fear of Knowledge is that it’s impossible to be a thoroughgoing and consistent relativist - but it’s at least a background hum in a lot of humanities scholarship. Indeed, it’s this very inconsistency that makes the appeals to relativism so pernicious. When humanists talk about the kinds of moral, political and social concerns that get their dander up, they talk like stark, raving, moral realists.3 But when called on to provide a larger theoretical framework to justify their endorsement of these political projects as a scholastic endeavor, these demands are waved away with a glib reference to relativism.

Remember the “three paths!” There are some unreconstructed relativists who appeal explicitly to relativism to justify their activism - that’s the third path. But the second path is one where relativism is deployed selectively, as a kind of defensive cover fire, whenever it is suggested that there might be some other objective standard of correctness in scholarship that the work needs to live up to. “Of course, we know such standards are a myth,” and then you go on making strident moral claims to pursue your argument. If the third path is becoming less popular, the second is still very much the standard in much of the humanities. So attacking relativism as a source of great mischief in humanities scholarship remains relevant.

Subscribe now

The Boghossian report presents a certain theory of the case for the flaws in humanities scholarship. The theory is that there are two kinds of standards that one might appeal to in doing humanities scholarship, epistemic standards of objective truth/knowledge/understanding on one hand, and political standards on the other. While it is hard to characterize the epistemic standards in any kind of precise way, they serve an essential regulative function within all forms of scholarship, including the humanities. Some biased scholarship does acknowledge the importance of epistemic standards (the first track), but the very fact that those standards are acknowledged means that we’re playing the same game, and appeal to epistemic standards can help eliminate the bias. But when epistemic standards are rejected wholesale, either defensively (the second track) or offensively (the third track), the regulative function vanishes, and the whole enterprise quickly goes off the rails. The rejection of epistemic standards is the remit of relativism. Therefore, if relativistic scholarship is purged from the humanities, the second and third track will be foreclosed, and we’ll be in a much better position.

If there is a flaw with this analysis, it’s that it is too sanguine about the prospects of eliminating bias on the first track. Clever people can twist themselves into pretzels to say that any crazy view lives up to the relevant epistemic standards (this fact is one that relativists are extremely familiar with, although they draw wild conclusions from it). And the academy is nothing if not populated with clever people who are adept at twisting themselves into pretzels to support crazy views.

But if there’s a deep truth to this analysis, it’s that there is value in making people go through the pretzel-twisting effort. Scholars should be required to give epistemic reasons for accepting claims about the world, to try to show that their theories correspond in some way to the facts. It is deeply corrosive and academically unserious to respond to such demands with “objectivity, lol.” Yet that is precisely what much of humanities scholarship has become.

Is relativism the only problem with the humanities? No way. Is it the foundational problem with the humanities? Maybe, although I wouldn’t bet on it. Is it a major, central problem with the philosophical underpinnings of the humanities? Absolutely it is! And it has been since the 90s. Boghossian is entitled to his “I told you so.”

Will this report convince people in the humanities to abandon their relativism? No way. Will this report provide intellectual cover for administrators to more aggressively dismantle the humanities in many universities, particularly in departments like anthropology that embrace a relaitivistic and political methodology? Almost certainly.

Will that be an overall improvement to the intellectual life of the humanities, and to the academy overall? I don’t know. I hope so. I guess we’ll see.

Subscribe now

Now I am wary of linguistic prescriptivism. I don’t want to be the guy who insists “You’re using the term wrong!” when 99% of people use a term in one way, and I’m the holdout who uses the term in some other, more traditional way. If people have decided to use the term ‘moral relativism’ to refer to a kind of radical tolerance, then I won’t try to stop them. But I think that almost everyone knows that moral relativism doesn’t mean radical tolerance. Rather, they think that there’s a simple argument that takes us from moral relativism to radical tolerance. That simple argument would go something like this:

If moral relativism is true, then what’s wrong for me might not be wrong for you. Indeed, the fact that you’re doing something means that you probably think it’s ok. And if you think it’s ok, then, if moral relativism is true, it is ok (for you). So no one could be justified in criticizing you for your actions. After all, by your lights, you’re not doing anything wrong. By my lights you are, but my perspective is no more valid than yours, so I have no right to impose my perspective on you.

This can then easily lead into an argument for a form of moral realism, like:

But what you’re doing is really, really bad. You’re murdering babies! Of course I have a right to criticize that action. Therefore, moral relativism must be wrong.

But both of these arguments are very bad. Let me explain why.

Subscribe now

Let’s begin by thinking about what moral relativism is. I hope that everyone agrees that moral relativism is well-captured by the simple slogan, “Morality is relative.” Let’s pick that apart a bit. First off, note that moral relativism isn’t the view that morality doesn’t exist or isn’t real. That is a different view in metaethics (usually called moral error theory). If you think that some subject isn’t real, you say it isn’t real, you don’t say it’s all relative. Atheists don’t say that God is relative; people who look down on astrology don’t say that astrology is all relative. They say it’s wrong, or mistaken. So moral relativism is thus the view that there are moral facts, of a kind. These facts are just relative.

But relative to what? Our simple slogan doesn’t tell us. The most common versions of relativism say that morality is relative to either personal moral beliefs or to cultural norms. On the former view, the standards for morality vary from person to person; what’s morally right and wrong depends on what matters to me or to you. On the latter view, the standards of morality are social or cultural; what’s morally right and wrong depends on the dominant norms of some society. (Note that on the former view, people’s moral views are always, in a sense, correct, and trivially so, because the standards for evaluating a person’s moral beliefs are just their own moral beliefs. On the latter view, people could have incorrect moral views. An ethical vegetarian in a society full of meat-eaters would just be wrong about what morality requires of her because, in her society, eating meat is ok.)

This complicates our definition of moral relativism somewhat. Both the personal and the cultural versions of moral relativism are indeed versions of moral relativism, so we need to be somewhat schematic when defining moral relativism to capture the fact that there are different versions of the view. So let us say that moral relativism is the idea that moral facts are facts only relative to some framework. Different versions of moral relativism will say what a framework is.

But saying this leaves things too underspecified. What should we say about the view that moral facts are facts only relative to some framework, but the framework that they are facts relative to is the objective moral law? That doesn’t look like a kind of relativism at all. So we should say that moral relativism is the view that moral facts are facts relative to some framework, and there is more than one valid moral framework.1

Ok, so that’s relativism. Now what about tolerance? Tolerance is, in general, the idea that we shouldn’t judge/criticize/ostracize/whatever others. Perhaps, keeping an eye on how we defined relativism, we should add that we shouldn’t judge/etc others if they are acting in a way that is acceptable relative to some valid moral framework. Depending on how we feel about the debate I mentioned in the footnote at the end of the last paragraph, that might be a significant addition to the idea of tolerance. But either way, tolerance is a moral norm. It says what you should or shouldn’t do.

The central question we’re interested in is whether or not we should be tolerant if moral relativism is true. But if moral relativism is true, then there’s no objective fact of the matter about whether or not we should be tolerant. And if moral relativism is true, then there’s no objective fact of the matter about whether or not we should be tolerant if moral relativism is true. Whether or not we should be tolerant depends on our moral framework! That’s the whole point of relativism!

To illustrate, let’s consider the version of relativism where what’s right or wrong depends on your own particular subjective moral views. And you yourself are very committed to not tolerating baby murder (I hope). Indeed, according to your moral framework, baby murder is exactly the kind of thing we should be condemning. In this case, it would be wrong (for you) to tolerate baby murder. Now, of course, it would be wrong for you to condemn baby murder according to the moral framework of the baby murderer. But so what? That’s not your framework! Why in the world do you need to act in accordance with the baby murderer’s framework? He’s a sick fuck!

The argument that moral relativism implies radical tolerance assumes that we can and should be taking a kind of weird, free-floating position on moral norms. We shouldn’t be looking at what is true according to our own moral frameworks. Rather, we should be freeing ourselves from our own moral frameworks, and recognizing that all moral frameworks have their own kind of validity. And, recognizing that all moral frameworks have their own kind of validity, we should respect the validity of frameworks other than our own, and that means not judging people who are acting in accordance with their own moral frameworks. But this free-floating moral position on moral norms is not entailed by moral relativism; it is just another moral framework. It might be one that moral relativists are tempted to accept, but that doesn’t make that framework a part of moral relativism. It’s very much not.

And honestly, there aren’t too many relativists who are tempted to accept that free-floating framework, at least not in full generality. (Ask the stereotypical lefty moral relativist what he thinks about Gaza.) Tolerance is considered to be a virtue in contemporary liberal society, and perhaps a kind of tolerance is a virtue. But no one is really radically tolerant. And that’s a good thing! Moral relativism gives us no reason to be radically tolerant. Nothing does.

This isn’t to say that there is no problem with moral relativism. I’m not a moral relativist myself. But that’s because I have worries about the coherence of the notion of framework-relative moral truth that relativists are committed to. The idea that moral relativism entails some sort of norm of radical tolerance is a ridiculous brain worm that has infected the popular understanding of metaethics. It’s the product of sloppy thinking, and we are best rid of it.

Subscribe now

Philosophers are in a weird place about how they understand modality. The oldest accounts of modality, dating to Aristotle, attached modal terms to particular things, and say that things are necessary if they have the cause of their existence within themselves, and contingent if their cause is something external. (This is similar to debates within Buddhism and other Indian philosophies about “dependent origination.”) Many centuries later (and I’m skipping over a lot of intermediate developments here), Leibniz proposed a model where a claim is necessary just in case it is true in “every possible world.” This was something of a strange suggestion, and it didn’t really catch on at the time.

But this idea of “possible worlds” came roaring back into favor in the mid-20th century for two reasons. First, it was noticed that the notions of necessity and possibility have a logical property (duality; it doesn’t matter what that is) that quantification also has, and so it makes sense to understand modal terms as being quantifiers. But quantifiers require a domain, and so we could introduce “possible worlds” as the name for a dummy variable that modal terms quantify over. Second, work on modal logic in the early 20th century was concerned with a lot of questions like “If something is possibly necessary, then is it actually necessary?” and logicians proposed different logical systems for answering these questions that were all more or less unsatisfying. Then Saul Kripke (as a teenager!) developed a framework for integrating these different logics with one another. It didn’t solve any of the underlying problems, but it introduced a way to compare all of the different proposed axioms within a unified framework that utilized an “accessibility” relation between “possible worlds.” Again, the term “possible world” just refers to a dummy variable in a logical system. But it’s a powerful system that built on the successes of quantified modal logic. It quickly became widely accepted, and Kripke was offered a position at Princeton despite never getting a PhD or receiving any formal training in philosophy. He just nailed this one problem so hard that he earned everyone’s respect.

But this caused a lot of people to ask the question, “What is a possible world?” And there is no one more influential here than one of Kripke’s colleagues at Princeton, David Lewis. Lewis proposed that possible worlds are real, concrete, spatiotemporally disconnected realities. In effect, he was proposing a kind of “multiverse theory,” although a different multiverse from what we might find in the Everett interpretation of quantum mechanics or anything in popular science fiction. It’s a modal multiverse: when we say “X is possible,” what we’re saying is that there are an infinite number of different real, concrete universes, and X is true in at least one of them. Kripke thought this was silly, as did most other philosophers. But Lewis had arguments for his position, some people were convinced, and so this view remains a minor position in contemporary philosophy. It’s known as “modal realism.”

And this is where my gripe emerges.

Subscribe now

When philosophers use terms like “realism,” what they usually mean is something like “there is an objective fact of the matter.” So, for instance, moral realism is the view that there is an objective fact of the matter about what kinds of things are right or wrong. (What “objective facts of the matter” are is vague and contentious, so philosophers tend not to like to talk this way. But when philosophers talk about “realism” of various kinds, they’re basically trying to latch onto the same idea that non-philosophers are trying to latch onto when they use terms like “There is an objective fact of the matter.”) So here’s an obvious question: What would you call the view that there is an objective fact of the matter about what kinds of facts are necessary? Well, this is a kind of realism. And it’s a realism about modality. So the obvious term to use here is “modal realism.” But that’s not what “modal realism” means. “Modal realism” means David Lewis’s thing about the multiverse. We can’t use the term, because it’s already taken.

So what term have we come to use instead for this view? As best I can tell, we don’t have one. The view has become invisible because the most obvious term to refer to it got used for Lewis’s flight of fancy. It’s not entirely invisible, of course. People have views that reject it. Amie Thomasson does, for instance. But debate ends up centering on the particular virtues or vices of Thomasson’s view (or other anti-realist views), and not so much on the big picture question of realism vs anti-realism. Ask any philosopher what the debate over moral realism is, and they’ll be able to give a more or less apt characterization of what the debate is about and some major arguments on either side. Ask any philosopher what the debate over modal realism is, and they’ll give you kind of a confused stare. “The David Lewis thing?”

To illustrate, here are the results from the PhilPapers survey, which surveys philosophers on a wide variety of philosophical questions. There is no question for “modality,” but there is a question for “possible worlds.” Here are the results:

“Concrete” is the David Lewis view, sitting pretty at 4%. The majority “abstract” view is, for the most part, philosophers who think that possible worlds are sets of propositions. “Nonexistent” are philosophers who don’t like all this talk of “possible worlds.” But where do we place the philosophers who think that possible worlds are a dummy variable in a perfectly useful and powerful modal logic, but who deny that there is any objective fact of the matter about what kinds of claims are necessary? “Other?” Perhaps. But I’m not happy with that, because my view looks like it’s an answer to a fundamentally different kind of question. But that question doesn’t appear on the PhilPapers survey, because it’s not a visible question among contemporary philosophers.

And I think this actually matters. Just as first-order work in ethics uses moral notions, and generally proceeds as though some version of moral realism is true, first-order work in philosophy in general often uses modal notions. And philosophers generally assume that some kind of modal realism is true. This is not because they have arguments in favor of modal realism. They just use modal terms like ‘necessary,’ and don’t think too much about them, and so give them a naive realistic treatment. And this is a problem because many philosophical arguments that invoke modal notions would fall apart if modal realism is false. And there’s very good reason to think that it is false! My understanding is that most philosophers who have seriously considered the question end up being some kind of modal anti-realist. But that fact has difficulty penetrating because of the opacity of the terminology. The standard dialectic runs like this:

A: You’re saying here that this claim is necessary. What do you mean by that?

B (not an expert in modal metaphysics but went to grad school): You know. True in every possible world.

A: But what do you mean by that? What are possible worlds?

B: Oh, haha. We’re going there, are we?

A: Are you a modal realist?

B: The David Lewis thing? Oh, no way. that’s crazy.

A: So then what?

B: I dunno, not what I work on. Let’s say they’re maximal consistent sets of propositions. That’s the common view, right?

A: So your claim is true in every maximal consistent set of propositions? That’s not how sets work. Or propositions, for that matter. So this makes no sense.

B: Well, what do you want from me? I’m working on epistemology, not modal metaphysics. This claim is necessary. Bob down the hall works on modal metaphysics. Ask him what it means.

So the problematic assumption goes unnoticed.

And if you ask Bob down the hall, things will hardly be better. While Kripke wasn’t a modal realist in Lewis’s sense, he was absolutely a modal realist in the sense that I’m using the term. Lots of people working in modal metaphysics have sort of unquestioningly followed his lead, and so an underlying assumption of modal realism is just sort of in the water among metaphysicians. (Lewis and Kripke are no longer with us, but these days Tim Williamson is the God-Emperor of complacent modal realists, and he is insanely influential.) And modal realism’s influence is magnified by the fact that there’s no term for it. For most philosophers working today, it’s truly an invisible assumption.

I don’t want to make the case against modal realism here, for risk of running too long. Perhaps that’s my next post. But I do want to insist here that we need to reclaim the term “modal realism” and use it to refer to the idea that there is an objective fact of the matter about what kinds of things are necessary/possible/etc. Because whether modal realism is true is a centrally important question for many philosophical debates, and modal realism is ultimately no more plausible than Lewis’s “modal realism.” Those who accept it should be called on to defend it. And the first step in doing that is making sure that the view at least has a name.

Subscribe now

Subscribe now

Is philosophy truly perfect? Well, no, of course not. Cappelen admits straight up that the title is an exaggeration and a provocation. Still, it is not too much of an exaggeration of his central thesis of Hyper-Optimism, which he advances in opposition to the “flagellants” and “hedgers.” Flagellants think that philosophy is more or less utterly worthless, and their philosophy exists as a kind of self-flagellation for doing philosophy at all, combined with some encouragement to others to get out while they still can. Wittgenstein is, perhaps, in this camp. Hedgers are less militant, but still think that there are major foundational problems with the field, despite some occasional successes. Cappelen, by contrast, wants to show that philosophy is awesome and important, and that it’s working more or less the way it should.

The book is organized in four parts. The first part argues that philosophy is a foundational discipline, with all inquiry eventually leading back to philosophy. This is not to say that philosophical questions constitute foundations in the Cartesian sense of first principles or foundational premises. Rather, the claims of any discipline are subject to a battery of familiar challenges, e.g. to clarify key terms or elucidate the normative standards that justify the acceptance of those claims. Those working in intellectual disciplines don’t need to have complete and fully satisfying answers to all of those questions, but it is discrediting to not be able to say something intelligible in response. And, of course, this is just the first step into doing philosophy, as philosophy is the discipline that attempts to provide complete and fully satisfying answers to standard challenges.

In the second part, Cappelen argues against what he takes to be the strongest argument that philosophy is in bad shape, the argument from divergence. According to the argument from divergence, convergence on widely-accepted answers is the hallmark of a healthy discipline, while a failure of convergence is a sign of sickness. Philosophy demonstrates noteworthy divergence, and thus it is sick. In response, Cappelen argues that convergence is unhealthy, as it is often a sign of bias and groupthink. Divergence, then, can be a sign of health, as it shows that the discipline is safe for free-thinkers. This is particularly important for philosophy, which is essentially a critical discipline. It would be very bad if everyone within a critical discipline ended up accepting a slate of standard dogmas.

The third part argues that philosophy has largely succeeded at its goal of discovering the answers to the various “Very Important Questions.” This is not to say that philosophers have converged on the answers to those very important questions, for they have not (and a good thing, too!). Rather, the answers exist, somewhere in the philosophical commons. Someone found the correct answer, and you can find the correct answer too by doing a bit of reading (and these days, perhaps AI will help, although reliance on AI has its perils as well). He argues that we’ve found the answers to most of these very important questions because, over time, the logical space has been “mostly filled.” Most of these VIQs are, in one way or another, questions about whether or not something exists. And to the question “Does X exist?” Capellen argues that there are really only four answers, which philosophers have advanced in one way or another in relation to every VIQ. Those answers are (1), Yes, it does, in quite a robust sense; (2) Yes, it exists, but in a sort of reductive way, which might not be exactly what you thought; (3) No, it doesn’t exist; and (4) The questions “Does X exist?” doesn’t make sense. Of course, each of those answers invites further clarification, but that’s alright. Now we’re just filling in the details. There’s good, important, work to be done there. But we’ve got the right answer to all of our VIQs at at least some level of generality, and perhaps in many cases we have it to a high level of detail as well.

The fourth part ties up some loose ends by arguing that this model does constitute philosophical success, since simply finding the right answers is an accomplishment. You might object that simply having the right answer in a form that an LLM can find is not the goal of philosophy; Cappelen responds that it makes little sense to talk about the “goal of philosophy,” and to the extent that it does make sense, making correct answers available is as good a goal as any. He also argues that the perfection of philosophy is structurally immune to refutation, since any good refutation of the perfection of philosophy would itself be an example of good philosophical argument. This is the familiar point that arguments against the worth of philosophy are themselves philosophical arguments, but pitched at the level of methods rather than the level of particular claims. It shows that there must be at least one good method for answering philosophical questions, and that this method has been found (perhaps by one of philosophy’s deluded critics). And if we haven’t vindicated all of philosophy’s methods, that’s not a bad thing. It is perfectly sensible to criticize some ways of doing philosophy, but self-refuting to attempt to criticize them all. Criticism of philosophy is consistent with, and even evidence for, the idea that philosophy is perfect.

I have very mixed feelings about all of this. I find important things in each part of the book that I agree with vociferously, but also things that I disagree with very strongly. Let’s go through part by part.

Subscribe now

In the first part of the book, Cappelen is absolutely right that philosophy has the kind of foundational status that he says, and this is important. Because philosophy occupies this foundational status, philosophy is the kind of thing that academics in many different disciplines would benefit from knowing more of. One of Cappelen’s repeated complaints throughout the book is that philosophy of science has degraded over time. In the early 20th century, philosophers of science tried to develop grand theories of what science is and why science works. But as a result of the influence of figures like Kuhn and Feyerabend, grand theories of the philosophy of science have gone out of fashion. Contemporary philosophy of science has become more interested in trying to clean up conceptual confusions around the edges of contemporary work in the sciences. And while this work is hardly useless, contemporary philosophy of science is not interested in providing grand theories of the nature of science and how it relates to the world that might constitute complete and satisfying answers to fundamental questions about method in the sciences. Cappelen thinks that this is a major loss and an abidcation of philosophy’s role as a foundational discipline. I’m inclined to agree.

Yet when Cappelen says that this foundational status of philosophy means that skepticism about philosophy leads to a skepticism about all inquiry, he clearly overstates the case. Scientific progress continues, despite philosophers’ irrelevance to much work in the sciences. Philosophical awareness is something that can make a discipline better, and in some extreme cases a lack of philosophical awareness can impede further progress. (I suspect that fundamental physics is in a trap like this at the moment.) Yet in most cases, scientific disciplines can advance well enough without philosophy. Philosophers of biology and philosophers of medicine have much to contribute, yet we were still able to develop RNA vaccines without their influence. Philosophers of logic and mind have much to contribute, yet we were still able to develop LLMs without their influence. So while I think philosophy is important and can sometimes drive progress in other areas, philosophical success is not the sine qua non of all inquiry. This substantially lowers the stakes for Cappelen’s argument. Philosophy can help other areas. But if philosophy turns out to be largely bullshit, it is not all of human inquiry that is at stake, it is mostly just philosophy itself.

In the second part, I concur with Cappelen that convergence is not necessarily a sign of success, and divergence not necessarily a sign of failure. In particular, I think that he’s absolutely right to worry about groupthink as a cause of convergence. He also has an extensive discussion of the literature on peer disagreement, and (not to rehash my own work on that topic here), I think Cappelen is right that so-called “peer disagreement” is not always a defeater for our beliefs (although it sometimes is.) So it would be a mistake to reason in a shallow way from the existence of divergence to the conclusion that philosophy is sick.

Yet Cappelen’s argument misses an important point by focusing on extreme theses; he’s concerned with whether convergence is always good and divergence always bad. And they’re not; there are times when convergence is good and times when it is bad, and the same goes for divergence. Convergence can be caused by groupthink. But it can also be caused by having a progressive methodology, which builds on successes to push theories into greater alignment with the truth. Similarly, divergence can be caused by a healthy critical tradition and an openness to free thinking. But it can also be caused by a lack of mechanisms for ruling out insane theoretical non-starters, such that no theory could ever really be discredited, leading to an endless accumulation of theoretical cruft over time. The important question is why we generally experience divergence rather than convergence in philosophy. For myself, I think that the best explanation is typically the pessimistic one. But my own verdict aside, Cappelen doesn’t really make the case for his own optimistic verdict. His energies are dedicated to showing that an optimistic explanation of divergence is possible. And it is, which is a good point. But the important question is whether such an optimistic explanation is actual. Here he has little to say, and less that is convincing.

It should also be noted that philosophy has its own share of groupthink-driven convergence. While philosophy has not been quite as susceptible to the vogue for progressive social justice as some other disciplines, it was still largely swamped by that wave, and there is every indication that the situation will get worse rather than better in the near future given the culture among graduate students over the last decade and up until today. And for those who are optimistic that the critical tendency in philosophy will put things back on track eventually, it’s worth recalling that philosophy was dogmatically Catholic for nearly a millennium. The critical tendency in philosophy can operate extremely slowly, and we are not at all immune to groupthink.

The third part of the book is probably the one that I liked the least. There are still things I agree with here in broad strokes. Having discovered the right answer to philosophical problems is a genuine achievement, even if not everyone agrees with that right answer. And Cappelen is right to note that most answers to the Very Important Questions can be sorted more or less neatly into the “Yes, robustly,” “Yes, reductively,” “No,” and “That’s a bad question” categories. That’s an interesting observation. But noticing that we can easily divide the logical space in this way substantially mitigates the importance of being able to fill it. Compare: if you ask two people if God exists, and one person says “Yes” and the other says “No,” it would be insane to say “Well, one way or the other, we’ve got it. Time to pack it up and go home, theology is done.” As Cappelen notes, it’s not just important to get the right answer. We also have to provide sound arguments that identify genuine considerations that count in favor of our conclusions, such that understanding those arguments shows why our conclusions are true. But he doesn’t show that we’ve filled the logical space of arguments (and it’s not clear how we could, the logical space here allows for much more important and interesting subdivision than Cappelen’s four-fold division.)

Most importantly, Cappelen doesn’t take seriously the idea that philosophical arguments just aren’t the kinds of things that can provide the relevant kinds of genuine considerations. Philosophers have provided many arguments for their conclusions over the centuries, but from my perspective, this looks an awful lot like throwing a bunch of spaghetti at the wall to see what sticks. And the problem isn’t that none of it has stuck, it’s that all of it has, because we have no good way of sorting the good from the bad (and really, it’s all bad) and rejecting the bad. As I suggested above, much of the divergence in philosophical thought is explained by a combination of a lack of progressive methodology that converges on the truth and a similar lack of a method for excluding bad views from further consideration. So the cruft of bad arguments accumulates endlessly. This is hardly progress. Cappelen can be right about everything that he says in the book, and yet it could still be the case that philosophy is, in general, the practice of offering up answers to the VIQs (that all fit into four very broad categories) on the basis of absurd arguments that do not, and could not, provide any genuine understanding. Perfect!

This brings me to the final part of the book. Suppose that all of my criticisms of Cappelen have been right on the mark. Are these criticisms not philosophical arguments? And are there therefore good philosophical arguments? I have triumphed here, and I have triumphed as a philosopher, and thus this is a triumph of philosophy. This is Cappelen’s coup de grace. And I think he’s precisely right about this. There are good philosophical arguments. And perhaps the very best part of the book is how forcefully Cappelen makes the case that even the most broad and compelling skeptical arguments are, ultimately, vindications for philosophy.

But this goes very far from showing that philosophy is perfect. As I argued in my earlier piece on “anti-philosophy philosophy,” it is genuinely self-defeating to try to argue against all philosophy. But as sensible skeptics have long known, this means that the right way to proceed is to identify a narrow range where philosophy might succeed, and use that as a platform where we can launch attacks on the broad field of bad philosophy. A Wittgensteinian program of arguing that all philosophy, including this claim, is bullshit is self-undermining and cannot be taken seriously. But a Carnapian program of arguing that all philosophy outside of this narrow range is bullshit, and almost all philosophy falls into that bullshit category is not self-undermining and is likely correct. Cappelen will say that showing that this is so would be a massive philosophical undertaking, and if that undertaking can be successfully accomplished, that would be a huge achievement for philosophy. That is so. Yet it shows that Capellen’s final defense of Hyper-Optimism isn’t all that optimistic. If Cappellen’s thesis is consistent with the vast majority of philosophy being errant bullshit, then his title “Philosophy is Perfect” doesn’t look like an apt summary of his thesis. What his arguments ultimately support isn’t so much Hyper-Optimism but rather non-Hyper-Pessimism.

In sum, while Cappelen says that he’s arguing against both the flagellants and the hedgers, he’s really just targeting the flagellants. Here’s how Cappelen characterizes hedgers at the beginning of the book:

For them, philosophy isn’t exactly a wasteland, but something more like a respectable municipal park. Figures like Daniel Stoljar (2017) exemplify this view: they argue that, while philosophy hasn’t delivered the kind of triumphant, unconditional answers we might have hoped for, it isn’t the intellectual disaster area some critics make it out to be. On their view, we’ve made progress of sorts. We can clarify questions, refute really bad arguments, or map the logical space of possible views. This progress is usually negative, conditional, or piecemeal. Stoljar and others emphasize that many of philosophy’s “successes” are negative theses (what knowledge isn’t, what moral realism can’t be, what free will doesn’t require), defensive maneuvers, or careful taxonomies of conceptual possibilities. Lewisians, like Helen Beebee, argue that philosophy doesn’t yield robust knowledge or settled truth. The value of philosophy is, instead, that of cultivating an appreciation of the landscape of possible answers. In these more modest, half-hearted defenses, philosophy becomes less the engine of discovery and more the steward of uncertainty. It’s a discipline that, at its best, keeps us honest, clarifies our confusions, and ensures we don’t fall too easily into dogmatism or complacency. In such accounts there’s a kind of quiet dignity, but little in the way of exuberant affirmation.

While Cappelen spends much time responding to particular arguments by Beebee (and some by Stoljar as well), he doesn’t refute this overall approach to philosophy. Cappelen is in the business of exuberant affirmation, but in response to the many objections he considers, he ends up engaging in a lot of hedging. He offers a full-hearted rather than half-hearted defense of philosophy, but his arguments don’t reach to a defense of the kind of swashbuckling metaphysics we can find in the work of, say Timothy Williamson (who Cappelen cites approvingly many times). In the end, all that his arguments support is the hedger’s conclusion that philosophy makes “progress of a sort.”

That’s not entirely a criticism. Cappelen’s triumphal tone is inspiring, and drives home the important point that even a negative and skeptical philosophy is a difficult and worthy undertaking, and its success would be a genuine accomplishment, and a genuine accomplishment for philosophy, in particular. But no, philosophy is not perfect. Not even close. And Cappelen doesn’t really do much to show that it is.

Subscribe now

This annoys me because the questions of what evidence is, what it means to be in possession of it, and what claims are supported by evidence, are not easy. And according to many theories of evidence (and all of the most plausible theories), which claims are supported by evidence depends not just on the evidence itself, but also a huge background of assumptions. So there’s no easy way to say whether or not some piece of evidence supports some particular claim, and thus no easy way to say whether or not some claim is supported by any evidence at all. And so to say that a claim was made “without evidence” is itself to make a claim that is, uh, lacking in support.

Subscribe now

To back that up, let’s see why the plausible theories of evidence always end up saying that evidential support occurs relative to background assumptions. Here’s an example that I love. In The Matrix (the favorite movie of every epistemologist of a certain age), Neo is living inside a computer simulation. He just doesn’t realize it. Then he is freed from the simulation, and is taught how the world really works. He then later goes back into the simulation, and is surprised by how his perspective has shifted. His experiences now that he’s back in the simulation are substantively the same as they were back when he was ignorant of his situation. The noodle shop that he used to frequent looks exactly the same - although he now knows that the noodle shop isn’t real, it’s just an artificial construction in the computer simulation.

Before the events of the movie, when Neo goes into the noodle shop, he has a certain set of experiences. Those experiences provide evidence for a variety of everyday claims, like “I am in a noodle shop.” After learning the terrible truth, Neo could go back into that noodle shop, and have effectively the same experiences. Yet those experiences would support a completely different set of conclusions about his surroundings, since he now knows that those experiences are simply the product of a computer simulation. The evidence is the same, but the conclusions that are supported by that evidence are different.

(You might object to this argument by saying that Neo’s experiences support the same conclusion in both cases, but this would be bizarre. Beforehand, Neo’s evidence is the same as our evidence of the noodle shops that we frequent, and so if his evidence beforehand doesn’t support the claim that he’s in a noodle shop, our evidence also shouldn’t support our own beliefs about the noodle shops we’ve been in. This is a route to thoroughgoing skepticism. Or we might say that Neo’s evidence still supports the conclusion that he’s in a real noodle shop after he’s learned the terrible truth, but that’s also clearly wrong; Neo knows better now, and he correctly interprets his experiences differently.)

The most sensible thing to say about this case is that what the evidence of Neo’s senses tells him changes. Before, his evidence supports the claim that he’s in a real noodle shop. After, his evidence supports the claim that he’s in a different noodle shop. And what his evidence supports has changed because he’s got a different set of background beliefs about how the world works and how his experiences are connected to reality.

To give a more complete picture of how this works, we have to talk about explanations and about cause and effect. (Note: Here I’ll be saying some things that are more controversial.) We take evidence, including the evidence of our senses, to be an indication of how things are in reality on the basis of our understanding of relations of cause and effect and our total model of the way the world works. When we have some experience, object, state of affairs (or whatever) that we are using as our evidence, we proceed by asking ourselves (often implicitly) “Why are things this way? Why do I have these experiences? What would have to be going on in the world for things to be appearing this way to me?” The best answer to that question tells us what our evidence supports. And the best answer depends on our background understanding of the way the world is and the way the world works. Before, Neo thinks he’s in the real world, and so he believes that his experiences are explained in the normal way by real objects in the world; that’s why his experiences support the claim that he’s in a real noodle shop. After, Neo thinks he’s in a computer simulation, and so he believes that his experiences are explained by a computer simulation feeding electrical impulses into his brain through a jack at the base of his skull; that’s why his experiences support the claim that he’s seeing a simulation.

This kind of causal reasoning is deployed in every context, not just in the context of figuring out what our experiences are telling us about our immediate surroundings. When the detective finds the footprints in the mud under the window and the bloody knife in the bushes, she solves the crime by figuring out the causal story that makes the most sense. The footprints are there because someone fled from the kitchen. The size and shape of the prints indicates a man wearing heavy boots, running fast. The location of the knife indicates that it was thrown into the bushes when... All of this relies on a set of background assumptions about cause and effect. The detective can solve the crime, whereas we cannot, not necessarily because the detective is a better reasoner in some abstract sense, but because she has the experience and knowledge that helps her make sense of all of this evidence, to figure out how it all best fits together.

This is how things work in the case of testimonial evidence as well. In typical cases of testimony, the fact that the person told us something constitutes evidence that it is true. The testimony itself is evidence, and this is because people usually only testify regarding things that they know, where their knowledge has some suitable connection to the truth. Testimony that P provides evidence that P in precisely these cases where the most reasonable thing to believe is that this is a normal sort of case, where the person you’re listening to is saying what they’re saying because it’s true.

So that’s the theory.

Subscribe now

When we keep this in mind, that phrase “claims without evidence” starts to look rather bizarre for a couple of reasons. First, when someone claims something, they almost never offer any other evidence for what they’re saying. The claim itself is the evidence for the listener. If you ask me where the nearest subway station is, and I say “Down the street and left at the corner,” I have, in a sense, claimed without evidence that the subway station is down the street and to the left. But my claim is not in doubt for all that.

Now of course you might ask me for my evidence, and I might try to supply it. But second, when I try to supply evidence, I’ll do so by further testimony. “What’s your evidence that the subway is down and to the left?” “Oh, I live around here and take the subway all the time. I remember quite well where the nearest stations are.” That’s just more testimony! That further testimony might be helpful some of the time. But if we suspect that the person we’re talking to is just a liar, asking for evidence will just elicit more testimony, which we might just as easily take to be further lies.

What are we asking for if we’re asking for evidence? Charts? Graphs? The information on those charts might be lies as well! Ask yourself what it would look like for a politician to get up at a podium and make some claim with evidence, and the obvious absurdity of the exercise will clue you in to why it’s meaningless to say that someone is making a claim without evidence.

Perhaps “claims without evidence” doesn’t mean that the speaker failed to offer evidence at the time, but rather that they don’t have any access to evidence at all, and thus were making a claim that they hadn’t supported for themselves. But while this reading of “claims without evidence” makes more sense in light of how actual testimony works, it’s not the sort of thing that is at all easy for outsiders to discern. How we do know what experiences the speaker has had, what they’ve been told, what they’ve seen? Typically, we don’t. Unless they tell us. But if we don’t trust what they tell us, then we’re still getting nowhere.

The difficulties here are vastly enhanced by the points that I emphasized above, about how what claims are supported by evidence depends in crucial ways on the background information (primarily about cause and effect and the way the world works in general) that different people have. For even if you did have complete access to everything that the speaker heard and saw, that wouldn’t be enough to know whether their belief was supported by evidence, because you don’t know what their causal model of the world is. Knowing that Neo walked into the noodle shop and looked around is not enough to know what his evidence supports, because what his evidence supports depends on those deep background beliefs about what is causing his experiences.

And of course all of this applies to the question of whether to trust someone’s testimony. In Trump’s press conference a few hours ago talking about the operation to abduct Maduro, he mentioned that one of the goals of the operation was to secure control over Venezuelan oil. The “without evidence” crowd didn’t say that THIS was a claim made without evidence. Nor did they say “Well, Trump is a liar, he lies all of the time, so this is probably a lie as well.” No, the response was “Ha! He admits it!” That may, of course, have been the correct response. But it’s noteworthy that people will sometimes say about Trump that every word out of his mouth is a lie and at other times will say that he’s just admitted to his darkest impulses.

This is just a further example of the fact that what is supported by testimonial evidence is determined by our own background model of the world. If Trump says X, and you’re antecedently convinced that X is true and think that X is the kind of thing Trump will tell the truth about, you will regard him as telling the truth. If you’re antecedently convinced that X is false and that Trump might lie about X, you’ll think he’s a liar. There’s nothing inconsistent about this. This is normal and rational. This is how all parsing of evidence works.

All of this shows why saying that someone is making claims “without evidence” is just a sort of absurd editorializing that convinces no one. People always make claims “without evidence;” in testimony, the claim itself is the evidence. And whether people have evidence (whether or not they talk about it) depends on lots of complicated things that we don’t have access to. So really, if a journalist (or whoever) says that Trump (or whoever) is making a claim “without evidence,” they’re not really trying to make a serious point about evidence. They’re just saying “This is probably a lie, and so you shouldn’t trust it.” But whether or not something is probably a lie is the kind of thing that can only be assessed relative to a host of background assumptions.

This is not to say that how trustworthy someone is is unknowable or all a matter of opinion, nor is it to say that it’s impossible to evaluate the case to be made for or against the truth of some claim in a systematic and rational way. But evaluating evidential support is an incredibly complicated and difficult matter, and there is a strong subjective element involved.

That’s why I feel my blood pressure rise ever so slightly whenever I read the phrase “claimed without evidence.” Call the man a liar! That’s what you’re really doing anyway. And in the case of Trump, he really is a world-class liar, so you’ll often be correct to do so.

Please just keep your glib epistemology out of it. It’s not helping.

Subscribe now

Subscribe now

Philip Kitcher wrote a book, and Kieran Setiya didn’t like it. There’s a lot to pick over in this review about the nature of contemporary philosophy. But I’m going to focus on one claim that Kitcher and Setiya seem to be in agreement on:

[A]t least since Kripke, most philosophers have turned away from the analysis of concepts and towards the metaphysical investigation of things themselves: their subject isn’t words or what they mean, but the world they represent. Contemporary philosophers of mind, for instance, investigate the nature of consciousness and its relation to physics, not the meaning of the word ‘conscious’.

So says Setiya. To the extent that Kitcher disagrees, it’s because Kitcher thinks that philosophers haven’t gotten far enough away from conceptual analysis and haven’t embraced (socially-relevant) metaphysics enough. But they both seem to agree that metaphysics is a respectable philosophical pursuit, and it has been since Kripke. I strongly disagree. Metaphysics is a realm of pure “sophistry and illusions,” as Hume rightly argued nearly three centuries ago. It remains an alluring siren song for philosophers - who isn’t tempted by the idea that they can uncover the fundamental nature of the universe just by thinking about it? And so because philosophers want to do metaphysics, they’ll be tempted to accept any argument which purports to show that metaphysics is an intellectually respectable endeavor. This is why Kripke is so venerated.

The popularity and ubiquity of Kripke’s Naming and Necessity is perhaps a bit hard to understand for those coming to it for the first time. On its face, Kripke is arguing a technical point in the philosophy of language. Yet Kripke seemed to think that this technical point about language has sweeping implications for the possibility of doing a certain kind of metaphysical inquiry. Most philosophers have shared this assessment. I do not. Naming and Necessity does make genuine advances in the philosophy of language. But Kripke doesn’t vindicate metaphysics in Naming and Necessity. Not even close. To the extent that philosophers think he did, they’re just seeing what they want to see.

This next part should be familiar to anyone with a philosophy major; if you know the basis of logical positivism and the basic arguments from Lectures 1 and 2 of Naming and Necessity, feel free to skip ahead until you see bold text again.

To understand what Kripke was supposed to have accomplished (and to understand why he didn’t accomplish what many philosophers think he did), we need to go back to the era of logical positivism. Logical positivism was the dominant view in philosophy in the early-to-mid 20th century. The beginnings of the view had their origin in turn-of-the-century writing by philosophers like Gottlob Frege, Bertrand Russell, and Ernst Mach. By the 1930s, the view had gained massive popularity and wide acceptance. By the ‘40s and ‘50s, it had achieved a state of total dominance within the philosophy profession (apart from a few holdouts, of course), although cracks had already begun to appear in the foundations. Those cracks rapidly widened throughout the ‘50s and ‘60s due to the work of WVO Quine and (the later) Wittgenstein, among others. But the death blow to logical positivism was dealt by Saul Kripke. In three lectures delivered to the Princeton philosophy department in January of 1970, Kripke explained his dissatisfaction with some of the key tenets of the positivist program. Those lectures were later published as a book, Naming and Necessity. And the arguments in those lectures were so potent that, as the book spread, positivism died. Or so the usual story goes.

Kripke killed positivism; but what is positivism? Logical positivism is an anti-metaphysical philosophy that is rooted in a commitment to a descriptivist account of the meanings of words. According to descriptivism, all words, including names, have logically proper definitions, and those logically proper definitions determine what the word refers to. For instance, the meaning of the word ‘Paris’ might be something like ‘the capital of France.’ Because there is only one city that satisfies that description (i.e. there is only one capital of France), the word ‘Paris’ refers to that thing which satisfies that description. This is how words refer to things: by describing those things.1

Now, you might have noticed a problem. We just defined ‘Paris’ in terms of words like ‘capital’ and ‘France.’ But what do those words mean? We might try to offer up definitions of those words - perhaps ‘France’ means something like ‘the European nation across the Channel from England.’ But now we have to define ‘European,’ ‘nation,’ ‘England,’ ‘across’… We’re just making more trouble for ourselves. We seem to have found ourselves in an infinite regress. Do our definitions stop anywhere? If so, where? Are there fundamental terms in our language, terms that can’t be given logically proper definitions? And if so, what are those terms?

Bertrand Russell suggested that there are fundamental terms in our language, and that the fundamental terms are those that refer to our sense data. This was a very exciting suggestion. If he’s right, then we can engage in a certain kind of analysis. Begin with some sentence like ‘Paris is the capital of France.’ We can then analyze all of the words in that sentence by giving each word its logically proper definition. What we’ll have afterwards will be a more complicated but more fundamental sentence, which might still include words that need to be defined further. We can then define those words; this will then yield a sentence which is even more complicated but even more fundamental. We continue in this way until we have a very complicated sentence where all of the terms of that sentence are fundamental terms, i.e. terms about our sense data. Performing these analyses is the central task of philosophy. Philosophers take sentences of ordinary English (or whatever language you’re interested in), and analyze until we hit analytic rock bottom: a logical construction out of sense data.

This logical construction out of sense data is either guaranteed to be true by the logical form of the sentence itself, or it is not. If it is guaranteed to be true by the logical form of the sentence, then we call the sentence “analytic.” Analytic sentences are true by definition (as revealed by the philosopher’s digging into definitions). These sentences are are knowable a priori (because all of this analysis is an a priori investigation), and they are necessarily true (because their truth doesn’t depend on how the world is, they are true purely as a matter of logic and language). Claims that are not guaranteed to be true by the logical form of the sentence are “synthetic.” To know whether or not they are true, we can’t just reflect on the meaning of the sentence. We actually need to investigate the world. But what investigations of the world can tell us whether the sentence is true or false? Here’s the neat trick: Synthetic sentences are confirmed by the very sense experiences that constitute the definition of that sentence. The meaning of the sentence is exactly the same as the experiences which confirm that sentence. So in doing our conceptual analysis to uncover the meaning of the sentence, we also uncover what it would take to show, empirically, that that sentence is true. We can then investigate the world to see if we have those experiences (in the relevant circumstances). If we do, the sentence is confirmed, and thus probably true. If we don’t, the sentence is probably not true.

One immediate implication of this view is that if a sentence can’t be confirmed empirically, then it is meaningless. If the meaning of a sentence is just (a logical construction out of (claims about)) the sense data that confirm that sentence, then if there are no sense data that would confirm a sentence, that sentence is meaningless. The logical positivists loved this implication of their view. It shows why debates about the number of angels that can dance on the head of a pin are silly and pointless. Philosophers uncover analytic truths through conceptual analysis. Scientists empirically investigate synthetic truths. And everything else is sophistry, a pseudo-intellectual musing on claims that are literally meaningless.

There are lots of objections that could be made, and have been made, to this picture of philosophy. Kripke went for the jugular by attacking the description theory of language that is the basis of the positivist program.

Consider the sentence ‘Paris is the capital of France.’ If the positivist is right that ‘Paris’ means something like ‘the capital of France,’ then we can (partly) analyze ‘Paris is the capital of France’ as ‘The capital of France is the capital of France.’ This is a tautology; it’s analytic. That means that we can know a priori that Paris is the capital of France and that Paris is necessarily the capital of France. But both of those implications are crazy. We can’t know a priori that Paris is the capital of France; that’s something we can only discover by investigating the world. Nor is it necessary that Paris is the capital of France. If things had gone differently in world history, the capital of France could have been Avignon.

This second point, about necessity, is what really animated Kripke. For any name, like ‘Paris’ (or ‘Aristotle’ or ‘Nixon,’ to use two of Kripke’s favorite examples), if the meaning of that name is a kind of description, then the descriptivist theory of meaning implies that the person referred to with that name necessarily satisfies the description that is the meaning of the name. Aristotle is necessarily a great philosopher; Nixon necessarily resigned the presidency in disgrace. But this is crazy. Clearly, Aristotle or Nixon could have chosen to be humble farmers, or they could have died young, or something else. It’s not necessary that anyone satisfies any of the descriptions usually attributed to them.

Time to get into the details that matter for my argument. Philosophers can start reading again.

One under-appreciated aspect of Kripke’s argument is his discussion of possible worlds. Following what was (and still is) the standard convention, Kripke says that a claim is necessarily true just in case it is true in all “possible worlds.”2 This might give rise to a certain impression about how we can figure out whether or not something is necessarily true: we investigate all possible worlds in order to see if the claim in question is true in all of them. But Kripke points out that possible worlds aren’t real; they’re a kind of linguistic fiction. As a result, possible worlds are not investigated; they are described. In other words, a “possible world” is something like a description of the way the (actual) world could be. Possible worlds don’t exist independently of us, to be investigated in one way or another. Possible worlds are constructed in the act of description. When we say ‘Aristotle could have been a farmer,’ we’re not saying that there is a possible world out there somewhere where Aristotle is a farmer. We are instead just describing a scenario where Aristotle is a farmer. Kripke points out that the descriptivist is committed to a very weird sort of asymmetry in our act of constructing possible worlds. When we want to talk about farmers or being a farmer in some possible world, we can do so directly: think of a farmer. But when we want to talk about Aristotle, the descriptivist claims that we need to do so indirectly, by performing a weird referential bank-shot off of a description of Aristotle: think of someone satisfying the descriptions typically associated with Aristotle, e.g. being the author of the Nicomachean Ethics, and we can thereby come to think of Aristotle as the person satisfying that description. Kripke says this is overly complicated and just plain wrong. When we want to talk about Aristotle in some possible world, we can do so directly: think of Aristotle. Think of that guy. The descriptivist theory implies that we can’t do that; so much the worse for the descriptivist theory!

Kripke at this point introduces a bit of technical terminology that philosophers have since been quite impressed by: the idea of a “rigid designator.” A rigid designator is a term that refers to the same thing in all possible worlds. Rigid designators are Kripke’s answer to the descriptivist’s referential bank-shot. We have terms - names - that refer to things directly, not just to properties of those things. And because worlds are described, not investigated, when we use a name to describe a world, we are describing a world that contains the thing in question. Whenever we describe a world by using the term ‘Aristotle,’ we’re describing a world with Aristotle - with that guy. That’s why names are rigid designators. ‘Aristotle’ refers to the same thing (i.e. to Aristotle) in all possible worlds because the semantic role of ‘Aristotle’ is just to refer to Aristotle directly. So when we describe a world using the word ‘Aristotle’ we’re guaranteed to be talking about Aristotle.

Kripke then argues that any true identity claim between two rigid designators is necessarily true. In other words, if we have a claim of the form ‘A = B,’ and ‘A’ and ‘B’ are both rigid designators, then, if ‘A = B’ is true, it is true in all possible worlds. He’s correct about this, but it’s important to slow down a bit and realize why it’s correct. ‘A’ is a rigid designator just in case it is a word whose semantic function is to refer to a thing, not to some description of that thing or to the properties of that thing. Same for ‘B.’ And if it’s true that A = B, then A and B are the same thing. So if we describe a world using the word ‘A,’ we’re constructing a world with A in it. That world is also a world with B in it, because A and B are the same thing. So if we describe that world using the word ‘B,’ we’re talking about the same thing that we’re talking about when we use the word ‘A.’ This is true of any world we might try to describe using our words ‘A’ and ‘B.’ So, in all possible worlds (that we could describe), A = B. Thus it is necessarily true that A = B.

So… how does this all add up an argument that metaphysics is intellectually respectable?

First and foremost, it shows that positivism is wrong. Positivism is based on a description theory of language. The description theory of language has crazy implications when we’re talking about naming and necessity (hey, that’s the title of the book!), and so the description theory of language is wrong. Therefore, positivism is wrong.

But note that this is not a vindication of metaphysics! If positivism is correct, then metaphysics is sophistry. Positivism is not correct. But that doesn’t imply that metaphysics isn’t sophistry; denying the antecedent is a fallacy. It might be the case that metaphysics is sophistical for some reason that doesn’t have anything to do with the description theory of language. (This is my view.) Kripke demolished an argument against metaphysics that was popular at the time, but that’s not the same thing as showing that metaphysics is a worthwhile intellectual pursuit.

This brings us to the second way that Kripke was supposed to have vindicated metaphysics. Kripke claimed to have uncovered that there is a respectable metaphysical notion of necessity, and further claimed to have uncovered substantive truths about metaphysical necessity. Here’s a critical early passage in Naming and Necessity (emphases added):

…The second concept which is in question is that of necessity. Sometimes this is used in an epistemological way and might then just mean a priori. And of course, sometimes it is used in a physical way when people distinguish between physical and logical necessity. But what I am concerned with here is a notion which is not a notion of epistemology but of metaphysics, in some (I hope) nonpejorative sense. We ask whether some thing might have been true, or might have been false. Well, if something is false, it’s obviously not necessarily true. If it is true, might it have been otherwise? Is it possible that, in this respect, the world should have been different from the way it is? If the answer is ‘no’, then this fact about the world is a necessary one. If the answer is ‘yes’, then this fact about the world is a contingent one. This in and of itself has nothing to do with anyone’s knowledge of anything. It’s certainly a philosophical thesis, and not a matter of obvious definitional equivalence, either that everything a priori is necessary or that everything necessary is a priori. Both concepts may be vague. That may be another problem. But at any rate they are dealing with two different domains, two different areas, the epistemological and the metaphysical. (Kripke 1980, 35-36)

The positivists had their conception of necessity which was, in Kripke’s reckoning, essentially epistemological: A claim is necessary if and only if it is analytic, in the positivists’ sense. But Kripke wants to defend the coherence of a notion of necessity that doesn’t follow from positivist ideas about necessity. And he gives some examples of claims that are necessarily true even if not analytic, most famously the claim that water is H2O. This is necessarily true because ‘water’ is just a name for a certain kind of stuff, water, and ‘H2O’ is just a name for that same stuff. ‘Water’ and ‘H2O’ are both logically proper names, and therefore rigid designators. And, as we saw, any claim of the form ‘A = B,’ where ‘A’ and ‘B’ are both rigid designators, is necessarily true. So we’ve established that there’s a kind of necessity other than the positivists’ necessity.

But it’s not really that different from the positivists’ account of necessity! The positivists weren’t really talking about a metaphysical notion of necessity; their necessity was a product of their more fundamental commitments about language. So is Kripke’s. If ‘A’ and ‘B’ are rigid designators that refer to the same thing, then to describe a world containing A is ipso facto to describe a world as containing B. Thus, in all possible worlds, A = B, and so it is necessary (in some sense) that A = B. We might call that sense “metaphysical necessity” if we like, but that label would be extraordinarily misleading, at the very least. That A is necessarily B is not an expression of some metaphysical insight into the nature of A and B. It’s simply a product of the fact that the words ‘A’ and ‘B’ are labels for the same thing. The conclusion that ‘A = B’ is necessarily true when ‘A’ and ‘B’ are both names for the same thing is a product of Kripke’s fundamental commitments about language. Kripke has different (and arguably better!) linguistic commitments than the positivists, but his claims about necessity are still purely a product of linguistic commitments. Kripke’s necessity is no more vindicating of metaphysics than the postivists’ necessity.

Once again: Possible worlds are not investigated, they are described. To say that something is true in all possible worlds is to say something linguistic, about what our language may be used to describe, it is not to provide any sort of metaphysical insight into the nature of reality or “the distribution of facts across modal space” or anything like that. ‘Water is H2O’ is necessarily true, but that’s not a fact about water and H2O, that’s a fact about ‘water’ and ‘H2O.’ This isn’t the conclusion that Kripke drew, but it’s the conclusion that his arguments actually support.

We can see that Kripke misunderstood his own arguments by looking at how he misapplies his conclusions in Lecture 3 of Naming and Necessity. Lectures 1 and 2 are concerned entirely with highlighting the many insanely implausible implications of descriptivism related to naming and necessity. In Lecture 3, Kripke lets his freak flag fly and starts doing speculative metaphysics. He informs us, for example, that Queen Elizabeth II (RIP) couldn’t have had different parents than she actually did. If we try to imagine a world where Queen Elizabeth had different parents, we’d be imaging a world where the person who sat on the throne of England and was known by the name “Queen Elizabeth” would have been someone else. Thus, according to Kripke, ‘Queen Elizabeth II was the daughter of King George VI’ is a necessary truth. A metaphysically necessary truth.

But why?

In the case of ‘water is H2O,’ we have a nice little argument: ‘water’ and ‘H2O’ are merely labels for the same thing. But ‘Queen Elizabeth’ and ‘the daughter of King George VI’ are not both mere labels for the same thing. Indeed, ‘the daughter of King George VI’ is a description of Queen Elizabeth, in much the same way that ‘the capital of France’ is a description of Paris. Descriptions aren’t logically proper names, and so they’re not rigid designators. So we don’t have the same kind of argument that would show that ‘Queen Elizabeth II was the daughter of King George VI’ is a necessary truth. Kripke presents his discussion in Lecture 3 as though it is a simple extension of the kind of reasoning he’d been doing in Lectures 1 and 2. But it’s not. He’s up to something very different.

Kripke could - and does - try to clean this up by saying that ‘the daughter of King George VI’ is a rigid designator as well.3 But again - why? It’s not because ‘the daughter of King George VI’ is a name rather than a description - that phrase is clearly a description, if anything is. Rather, it’s because ‘the daughter of King George VI’ refers to a property which is an essential property of Queen Elizabeth. And descriptions of essential properties are also rigid designators, according to Kripke.

Note that Kripke is going way out on a limb here. Here, at the beginning of Lecture 3, he’s making assumptions that far outstrip anything justified by his earlier arguments. Grant for the sake of argument that if some property is an essential property of some thing, then terms that refer to that property will be rigid designators. But even granting this: Why think that essential properties exist? And how could we know whether or not a particular property is an essential property rather than an “accidental” property? And so how can we know that a term that refers to such a property is a rigid designator? Kripke doesn’t say; he just assumes it because he finds the claim that Queen Elizabeth couldn’t have had different parents to be intuitive.

Kripke is doing speculative metaphysics. There’s no other way to put it. He’s making speculative claims about the essence of things, and those claims are ultimately backed up by how intuitive Kripke finds them to be. Leibniz would be proud. Hume weeps.

This might strike some as uncharitable, but my read on the situation is that Kripke was lured by the siren song of metaphysics, as so many other philosophers have been before. Metaphysics had fallen into disrepute in the positivist era. The basic thought behind this disrepute remains fundamentally sound: we cannot know that P is an essential fact about the nature of reality by merely thinking about P and asking ourselves whether it is plausible that P is an essential fact about the nature of reality. But the positivists tried to make that basic thought concrete by backing it up with a flawed theory of language. When Kripke showed that positivism rested on a defective theory of language, the bonds that were tying philosophers to the mast of empiricism were suddenly and unexpectedly cut. Kripke dived immediately into the waters of metaphysical speculation, driven by pure intuition, and the philosophical community was only too happy to follow him overboard. Now we’re drowning.

Here’s Setiya again:

[A]t least since Kripke, most philosophers have turned away from the analysis of concepts and towards the metaphysical investigation of things themselves: their subject isn’t words or what they mean, but the world they represent. Contemporary philosophers of mind, for instance, investigate the nature of consciousness and its relation to physics, not the meaning of the word ‘conscious’.

It’s undoubtedly true that contemporary philosophers take themselves to be using reason to uncover truths about the world our concepts represent, investigating the nature of consciousness rather than the meaning of the word ‘conscious.’ In this, they are following Kripke’s lead. So much the worse for contemporary philosophy.

The better response to Kripke’s death blow to positivism was not to go running back to the sophistical comforts of metaphysics, but instead to seek to re-establish the problems with metaphysics on more solid ground. We need to go back to Hume: the problems with metaphysics are not linguistic, but instead epistemological and psychological. If I have a big picture project as a philosopher, this is it.4

Subscribe now

Despite the difficulty with giving a strict characterization, analytic philosophy and continental philosophy have very different vibes. They are, for the most part, extremely easy to tell apart for anyone with a bit of familiarity with the field. This is, in large part, because analytic and continental philosophy are really quite different fields of scholarship at this point. Analytics only read and cite other analytics, and continentals only read and cite other continentals, and so there has been a divergent evolution in the writing styles and the kind of jargon used. We can identify many common ancestors, from Plato up to Kant, but things began to drift apart starting in the 19th century, and that accelerated greatly in the early-to-mid-20th century. At this point, speciation has occurred. It’s unclear that there can be productive engagement between the two disciplines at this point. This is mirrored in how academic departments are structured. Philosophy departments at universities are staffed almost entirely by analytic philosophers, while continental philosophers are largely located in other humanities departments. Most philosophy departments will have one token continental philosopher, and these are generally nice people who get along well with their colleagues, although no one in their department would ever dream of reading or engaging with their work.

I’m deliberately being a bit vague because I don’t want to fall into the trap of saying something false or insulting about continental philosophy. This is a trap that I’m liable to fall into. I myself am a trained analytic philosopher, and as such have read very little continental philosophy. What I have read, I have loathed.1 Perhaps there is good stuff out there; I’m assured that there is. But I have no desire to read continental philosophy, and there is absolutely no professional expectation that I will have done so. I’ve never read a single word of Derrida, and I have no plan to. (“Wait, what, you’re a philosophy professor and you’ve never read Derrida?” Well, I’m an analytic philosophy professor, and we don’t read continental philosophy.)

Subscribe now

Still, if you’d like a more concrete characterization of what continental philosophy is, you may want to read this recent piece by pseudonymous philosophy blogger Bentham’s Bulldog. BB is very much Team Analytic, and while his piece is useful for picking up on the continental philosophy vibe - it is extremely distinctive, and many readers will be able to recognize it immediately - his criticisms of continental philosophy largely fall into the trap of analysis-that-is-really-just-insult. This is particularly so because many of his criticisms of continental philosophy work equally well as criticisms of analytic philosophy.

One of his main criticisms is that continental philosophy relies too much on jargon. Guess what? So does analytic philosophy. We just have our own jargon.2 For those trained in analytic philosophy, analytic jargon seems natural and readable while continental philosophy jargon seems obscure and convoluted. But that’s more a product of training than an intrinsic feature of the jargon. BB also complains that when you try to state the basic continental philosophy ideas in jargon-free terms, you’ll often end up with something that’s either false or trivial. That’s true, but it’s also true of much analytic philosophy.

I’m not going to go through all of BB’s points one by one, because I just don’t care that much. But I do want to zoom in on one point in particular, because it very much rubbed me the wrong way. After giving a few examples of bad forms of arguments that he finds in much continental philosophy, BB writes:

But hopefully I’ve given a picture of how continental philosophers can go about writing very long books that purport to make arguments without ever actually providing reasons that should convince someone who doesn’t agree. There are genuine patterns connecting their premises to conclusions; the only problem is those methods don’t involve giving reasons.

This is true of continental philosophy, but again, it’s also true of analytic philosophy. It is true of philosophy.

There’s just an immense amount of confusion and willful blindness packed into those two short sentences. BB is concerned with the act of giving reasons to believe something. But what are reasons to believe? Are they, perhaps, considerations that will convince someone who doesn’t agree? This is what he suggests. But that sort of characterization of a reason to believe runs into well-known problems. Many times people will be convinced by stupid bullshit arguments because they’re idiots, and will not be convinced by perfectly compelling rational considerations because they’re stubborn. Arguments that aim only at convincing are sophistry.

Very well, then. What we want is not arguments that will convince, but arguments that should convince. We want arguments that provide genuine evidence for their conclusions, which can result in knowledge if we follow those arguments. We want to give real reasons to believe our conclusions. But that’s very hard to do, and I don’t think that any philosophers regularly succeed in the task.

Let’s draw a distinction from Logic 101. Deductive arguments are arguments where the premises entail the conclusion, where if the premises are true, then the conclusion is logically guaranteed to be true. Inductive arguments, on the other hand, are arguments where the premises support the conclusion, although the support in question is weaker than the guarantee of logical entailment. For any philosophical argument, then, we may ask whether the argument is supposed to be deductive or inductive.

Suppose it is deductive.3 If the premises entail the conclusion, then in an important sense the information in the conclusion is already obtained in the premises. A paradigm deductive argument is something like:

P1. The soul is immaterial and 2+2=4.

C. Therefore, the soul is immaterial.

If P1 is true, then C is guaranteed to be true because C is part of what P1 says. But then it’s not clear that P1 provides a real reason to believe C. Anyone who believes P1 already believes C, in some sense. So the premises can’t really be reasons to believe the conclusion. At best, they’re a way of showing you that you already do believe the conclusion.4

Analytic philosophers often like providing deductively valid arguments for their conclusions, but a moment’s reflection on what a deductively valid argument even is shows that this isn’t an effective way of providing reasons for the conclusion. The true substance of any philosophical argument comes in evaluating the arguments for the premises. For once you accept the premises, you already accept the conclusion. But what kinds of arguments do we give for the premises? More deductive arguments? That just puts the problem off for a minute, it does not solve it. The case then comes down to the arguments for the premises of the arguments for the premises. You just can’t provide reasons by providing premises in a deductive argument.

This is not to say that it’s bad to give a deductive argument in a philosophy paper. Giving deductive arguments can often be a useful way of organizing your thoughts, and giving a bit of readable structure to a complex set of considerations. But the real juice isn’t in that deductive argument. It is in the inductive arguments for the premises of your deductive argument.

So, if deductive arguments don’t really give reasons to believe our conclusions, we should be giving inductive arguments. But what makes for a good inductive argument? Here, there is an easy answer: No one knows, and it is not for lack of trying. The analytic philosophy tradition was founded by philosophers who wanted to give a general account of what makes an argument inductively strong, to provide a “logic of induction” which was just as rigorous and comprehensive as deductive logic. That project was the consuming passion of a huge group of brilliant philosophers who worried the problem for decades before they eventually gave up in frustration and despair.5 In sum, then, we provide good reasons to believe things by giving good inductive arguments, but we have no idea what a good inductive argument is.

This is not to say that there is no such thing as a good reason to believe something. I hold out hope that there is. But it’s prohibitively difficult to give a general characterization of what good reasons are. So there’s no easy or straightforward way to show that my arguments are proving good reasons while theirs aren’t.

Analytic philosophers are familiar with all of this, of course. The limitations of deductive logic and the failures of the program of inductive logic are the stuff of first-year graduate seminars. But this is largely taught as a lesson about the limitations and failures of one particular program that was of consuming interest in the first half of the 20th century. Ask a typical philosopher about what the methodological upshot is for their own work, and you’ll be met with some embarrassed mumbling.

Continental philosophers are familiar with this as well, of course, they just (often, not always) come to the conclusion that philosophical methodology is a free-for-all, we can argue for whatever we want by using whatever means we can find easy to hand, and if that means arguing for radical left-wing politics by offering a string of shitty metaphors, then so be it. This is bad. But analytic philosophers have not responded to the situation in a better way. The analytic philosophers (often, not always) wear the trappings of formal logic like a skinsuit, pretending to mathematical precision and proof, but instead offering something quite different.

And what is that something different? It’s intuitions, baby. The standard argumentative move among analytic philosophers is “It seems to me that P, therefore P.”6 Once you’ve established a few different propositions by appeal to nothing more than your own gut, the next step is to propose some general principles that can systematize those intuitions into something that looks a bit like a general metaphysical framework. Write it up (using tons of abbreviations) using LaTeX (to handle the mathematical notation), cite all the right people, and you’ve got yourself a publication right there. (Just hope a peer reviewer doesn’t have different intuitions, or isn’t angry that you didn’t cite them.) Feel free to pat yourself on the back for your rigor and clarity. Of course, the standard person on the street couldn’t understand a word of what you’re saying. But you can’t expect the hoi polloi to share your refined understanding of the literature or your well-honed intuitions, can you? Your peers can understand you, and that’s good enough. Good enough for tenure, at any rate.

So yeah, I’m an analytic philosopher, but I’m a pretty cynical one. I remain in the analytic camp because I loathe all of the continental philosophy I’ve ever read and I only loathe most of the analytic philosophy I read. Analytic philosophy is, I think, closer to getting it right than continental philosophy. But not, like, a lot closer. This is why manifestos like BB’s rub me the wrong way. I can take people being wrong - most people are wrong about most things most of the time. But I cannot abide those who cloak their wrongness in arrogant self-regard.

Analytic philosopher, heal thyself.

Subscribe now

That this point has become so widespread makes me uncomfortable, because I think the issue of trans inclusion in women’s sports is actually quite complicated, and it’s not at all clear to me which position should win out in the end. Trans inclusion in women’s sports is certainly not transparently absurd, as many people seem to think.

Before I begin, let me hasten to state that this is a philosophical argument, not a political argument. I want people to think with a bit more depth and clarity about this one hot-button issue. But if, God forbid, this article were to somehow make its way into the inbox of anyone involved in Democratic electoral politics, let me state up front: YOU SHOULD ABSOLUTELY BE AGAINST TRANS WOMEN IN WOMEN’S SPORTS. This is an 80-20 issue, and Democrats do themselves no favors by being on the losing side of issues like that. And I suspect that this will be perennially good advice, since I don’t see public opinion pivoting on this issue any time soon.

But maybe it should? I don’t know. Let’s go through the exercise and make the case for trans inclusion in the strongest possible terms. I’m not sure that I fully believe this argument. But I’m happy to defend it in print.

Subscribe now

The first point to make is that excluding trans women from women’s sports is not victimless. This is a policy that hurts trans women in some pretty concrete ways. The most significant harm isn’t that trans women can’t participate in the sporting leagues that they want to or that they feel they should be able to; you can’t always get what you want. Rather, the problem is that it puts a certain group of trans women, specifically those who have fully medically and socially transitioned, in an extremely awkward position. The stereotype of the trans woman in women’s sports is a male individual, who is very obviously male, in a competition where all the other competitors are females. Anyone observing the competition will think “That person doesn’t belong here.”

But this scenario has a flip side. Imagine an athletic competition where the men’s event features a lineup of obvious males, together with one individual with long hair, feminine features, noticeable breasts (supported by a sports bra)… in short, someone who everyone in attendance will immediately see as a woman. “What’s she doing competing with the men?” everyone will think. And then, “Oh…” A policy that forbids a fully transitioned trans woman from competing in women’s sports will both be incredibly awkward for the woman in question, and have the effect of outing her every time she competes. Very few would be willing to put themselves through that ordeal. Accordingly, a blanket ban on trans women’s inclusion in women’s sports would effectively ban fully transitioned trans women from all athletic competition. That is a real harm, and so we need a good reason to impose it.

To be clear, I’m not arguing for trans inclusion simply on the basis of self-ID. If a male fails to qualify for the men’s team on Saturday, he shouldn’t be able to declare “Actually I’m a woman” and try out for the women’s team on Sunday. But policies like the NCAA’s old policy on trans inclusion, which allowed males to compete in women’s sports following a prolonged period of hormone therapy, are not open to abuse, and will help avoid the kind of embarrassing scenario that blanket bans on trans women’s inclusion would make inevitable.1

So that’s the argument for trans inclusion. What is the argument against? There is really only one: that male athletes have a huge competitive advantage over females, and that while this advantage may be somewhat attenuated by hormone therapy, it is not eliminated. Going through male puberty creates a variety of physical changes in a human body which provide obvious and ongoing athletic advantages. So it is unfair to allow males to compete in women’s sports.

The problem with this argument is that the kind of unfairness that comes from having different body types is utterly rampant in all dimensions of sport. There’s nothing even approximately fair about who has a chance to be an elite athlete, and we shouldn’t pretend that there is.

Athletic capability is the result of two things, inherent biological gifts and physical training. Physical training is important: those who work hard will have a huge advantage in athletic competition over couch potatoes like me. But a world in which everyone works hard, following a rigorous and intelligent training schedule, is not a world in which everyone is an equally good athlete. There will still be large differences in athletic capacity, and those differences will be biological and often innate. The traits that make someone a good athlete will vary somewhat from sport to sport, but certain things, like height, strength, speed, hand-eye coordination, etc., will be relevant in most sports. And on a sport-by-sport basis, certain traits will have surprising relevance. Michael Phelps is the greatest swimmer of all time in part because he has abnormally large hands, which help him push his way through the water more easily. Victor Wembanyama is arguably the best basketball player in the world because the dude is 12 feet tall and his body is approximately 80% arm.2 In a world where everyone trained hard, only the very most physically gifted would be top athletes.

Now, in our world where not everyone trains hard, people who are not at the utter pinnacle of physical gifts still have a chance to compete at a high level, because many who could have been great athletes if they trained hard never put in the effort. But rather than radically democratizing sport, this fact just lowers the bar somewhat for how physically gifted you need to be in order to have a shot at participating at the elite level. Instead of being in the top .1% of the naturally gifted, you can have great athletic success if you’re in, say, the top 20%, but work really hard. But this means that people in the bottom 80% have no chance of competing at the highest level.

None of this should be controversial; indeed, it is the very rationale for the creation of women’s sports leagues. Because of the physical advantages of going through a male puberty, that top 20% is overwhelmingly populated by males. In an open competition where everyone competes against everyone else, few if any females would be in a position to compete, much less to win, at the highest level, no matter how hard they work.

In response to this situation, we’ve taken the step of creating specialized sports leagues to allow people outside of that 20% a chance at high performance in an athletic competition. The most well-known of these specialized sports leagues is women’s sports, but the basic principle is applied in many different ways. We have age-restricted sports leagues, so that children or the elderly can compete against their peers with respect to age. The Special Olympics and the Paralympics give those with developmental disabilities or physical disabilities like partial paralysis a chance to compete.

There are two things to note about this practice of creating specialized leagues. The first is that we don’t have a separate league for every single group that disproportionately falls out of that elite top quintile. There are no special sports leagues for the anemic, the asthmatic, the chronically undernourished, or the merely scrawny. Nor should we have separate sports leagues for every such group. There are costs associated with establishing and maintaining a sports league, and challenges involved with policing who is eligible for membership in specialized leagues. Only some groups get their own leagues, and which do is a political matter.

The second is that even within specialized leagues, it’s not the case that everyone can compete. Not every woman has a chance to be an elite athlete within a women’s sports league. Instead, it is only the most physically gifted women who will have a chance to compete. Specialized leagues are not a way of making sure that everyone can be an elite athlete. They are a way of providing opportunities for elite competition to those members of a restricted class of people who are physically gifted relative to that class. The familiar point about Phelps’s or Wembanyama’s remarkable physical gifts also applies to Simone Biles or Caitlyn Clark. They are the physical elite among women, and most women have no chance of competing with them, no matter how hard they work.

Put these two points together, and attend closely to what follows, because it is absolutely the most important thing to realize about sports competition: For any particular group, it is only the members of that group who are in the top quintile (or so) with respect to physical gifts who have any chance of competing at a high level if we establish a sports league for that group. The idea that “everyone deserves a chance to compete at a high level” is feel-good nonsense. We can establish more and more sports leagues in order to increase the total number of people who are in the privileged top quintile with respect to the eligibility criteria for some group or another. But society will not bear the costs of an indefinitely large number of sports leagues, nor should it be expected to. For any plausible policy regarding the establishment of sports leagues that we might cook up, the majority of the population will not be in the privileged top quintile of any eligible group. Some people just weren’t bound to be athletes. Sad but true! So what we have to do is make a decision - a political decision - about how many sports leagues we’ll have and what the eligibility criteria for those various leagues should be. These decisions will be sharp-elbowed. People care a lot about who will be on the winners’ podium at the end of the day, and changes in eligibility criteria will shuffle around who is in a position to be on the podium. But all appeals to “fairness” in this context aren’t really about fairness, because there’s no truly fair way to set eligibility criteria for specialized sports leagues. It’s just the dirty politics of working out how to draw the lines, with everyone trying to do so in a way where they end up on the podium at the end of the day.

Here’s an illustration of the basic point. Height, and the closely associated metrics of arm and leg length, are vitally important factors in athletic competition for almost every sport. Any person who is under 5 feet tall has little or no chance of being an elite athlete in practically every sport. Yet there are no sports leagues for the under-five-footers out there. If you’re short, you’re simply out of luck. Of course, we could establish a sports league for short people. But we haven’t, because we haven’t decided that the interests of short people matter in the relevant way. It’s unfair that people who were born short have no shot at being elite athletes, but we haven’t decided that it’s unfair in the way that means we should establish a short person’s sports league. Why not? Who knows; it’s arbitrary. (But imagine if we did establish such a league how fiercely height-measuring techniques would be debated and how much lawyering would go into establishing exact eligibility criteria, for slight deviations one way or the other would have an impact on who stands on the winners podium!)

Subscribe now

With all of these lessons in mind, let’s go back to the issue of women’s sports. Let’s first set aside an annoying side issue (which people absurdly insist on treating as relevant): Women’s sports are for *women.* So are trans women really women? This is not the question. The language that we choose to use here is arbitrary. The word ‘woman’ has two different possible definitions. Let us say that ‘woman1’ is the term that maintains the biological definition of the word ‘woman,’ while ‘woman2’ is the term that is associated with the trans-inclusive definition of ‘woman.’ With this in mind, the question about women’s sports is the question of whether “women’s sports” should be women1’s sports, or whether they should be women2’s sports. What should the eligibility criteria be?

Note that the group of women1 and the group of women2 are largely the same group. And notice also that, however we define the group, the vast majority of people will not be in the top quintile in an open competition. (Estrogen therapy does not eliminate the advantages conveyed by a male puberty, but it does substantially mitigate them, such that trans women who have been on extensive estrogen therapy are at a distinct competitive advantage relative to their cis male peers.) Being on estrogen therapy does not place competitors at a disadvantage relative to cis male competitors in the same way or to the same extent that being a cis woman who goes through a regular female puberty puts one at a competitive disadvantage. That’s so. But why does that matter? Women1 are, in the vast majority of cases, outside the top quintile and would be at a competitive disadvantage in any open competition. Women2 are also, in the vast majority of cases, outside the top quintile and would be at a competitive disadvantage in any open competition. Why exactly is it better to establish a women1’s league than a women2’s league?

One reason might be that, while a women1’s league and a women2’s league would both only be open to people who are almost entirely outside of that top quintile for the population as a whole, it’s better to have alternate leagues that are open to fewer people in the top quintile. So if more women1 than women2 are outside of the top quintile, we should have a women1’s league and not a women2’s league. But by this logic, we shouldn’t have a league for all women1. We should have a league for short women1, since more short women1 than women1 are outside of the top quintile. Insisting that justice requires us to minimize the number of eligible people who might compete well in an open competition takes us to absurd places.

Another reason might be that leagues for women2 or for short women1 would be leagues that have gerrymandered eligibility criteria. Only women1 is (what philosophers call) a natural kind. And for that reason, we should have sports leagues where for all and only women1. But why should eligibility criteria for sports competitions be determined by the metaphysical naturalness of the properties that establish those eligibility criteria? The suggestion is obviously ad hoc.

Another reason might be simple conservativeness. We’ve had women1’s sports leagues for a while, and haven’t had women2’s sports leagues, so we shouldn’t change. This hardly seems like a strong reason, though, particularly given the harms to trans women that come from having women1’s leagues rather than women2’s leagues that I mentioned at the beginning of this article. You might argue in response that women1 are harmed by the policy of having women2’s leagues, namely those women1 who would have made the team/won medals if the eligibility criteria were different. Yet every way of drawing eligibility criteria has harms of this kind, so this is not a reason to prefer one set of eligibility criteria to another.

The best reason I can think of for insisting on women1’s leagues rather than women2’s leagues is that the existence of women1’s leagues is part of a larger feminist project of countering the pervasive misogynist subjugation of women in society, and that a move from women1’s leagues to women2’s leagues would erode that project at least somewhat. This is a real cost to including trans women in women’s sports. But excluding trans women in women’s sports harms trans women! Contemporary society is at least as anti-trans as it is anti-woman. So if we want to take these kinds of systematic considerations seriously, we have to realize that they cut both ways.

The upshot is not that there is no reason to prefer women1’s leagues to women2’s leagues. It’s that there are reasons that push in both directions, and we’ve failed to find a consideration that makes one set of reasons clearly stronger than the other. We’re confronted with an irreconcilable conflict of interests here, and navigating that is a tricky political matter. There’s no way to solve this without someone’s feelings being hurt, and hurt quite badly. That sucks.

Yet we must resolve the question in one way or another. And if we do so by opting for women2’s leagues rather than women1’s leagues? Well, that’s hardly an absurd decision to make.

Subscribe now

Subscribe now

1. In your view, does an external world exist as a collection of objects, states of affairs, or facts that are independent of perception, reason, and consciousness? Can we prove, justify, or rationally confirm the existence of the external world; if so, how exactly? Does overcoming skepticism regarding the external world hold any philosophical value or significance?

This is a big question, of course, and to answer it fully would require a book of its own. So, briefly: Yes, an external world exists independently of ourselves and our minds. This is not something we can “prove” in a conclusive sense, but we do have evidence that this is the case. This evidence is provided by sense experiences. Specifically, the fact that the evidence of different sense modalities are in agreement with one another, and the fact that what we perceive remains stable over time, are strong indicators that there is an independently existing external world.

This much is relatively uncontroversial! The most famous skeptical hypotheses concerning brains in vats or evil demons do not deny the existence of an external world, they simply hold that the external world has a different nature than we might have thought. The computer hooked up to the brain in a vat or the evil demon would, after all, be objects existing independently of our perception, reason, and consciousness. Yet these hypotheses regarding computers or demons add elements to the external world story that we have no evidence for, so I do not believe that these more complicated stories are true. In short, our evidence that there is an external world is the nature of our sense experiences, and those sense experiences support the hypothesis that there is an external world because the truth of that hypothesis is the best explanation of why we have those experiences.

As for whether this conclusion has any philosophical significance, I think it does, but here we must be careful. For in a sense, all that the basic step of dispelling skepticism has accomplished is that it validates something that everyone already believes as a matter of common sense. So this is not a conclusion that we really needed any philosophy to reach. Philosophy got us into trouble by introducing the skeptical worry, and then got us back out again by dispelling that worry. Perhaps we would have been better off, on this topic at least, by not even getting started. But there is still a worthwhile lesson to be learned here. For while skepticism about the external world is a mistaken view, other kinds of skepticism, including skepticism about the divine, about ethics, and about metaphysics, are quite correct. And this is because the anti-skeptical argument that I outlined in the previous paragraph cannot be successfully adapted to these other areas. So anti-skeptical philosophy is in this way limited. Common sense is not always correct; there are some things we should be skeptical about. The external world just isn’t one of them.

2. What do you think about G.E. Moore’s proof of an external world and, more broadly, the Moorean arguments against skepticism?

Moore’s proof of an external world is fascinating, and much more interesting than his critics give him credit for. As I read Moore, all of the work is being done by what I call “Moore’s meta-proof,” his proof that his first-order proof of an external world is in fact a good one. Moore argues that for a proof to be a good proof, it must satisfy three criteria. First, the premises must be different from the conclusion. (He means this in the minimal sense that the conclusion is not just a literal restatement of the premises; P therefore P is not a proof.) Second, the premises must entail the conclusion. And third, we must know the premises to be true. And he argues quite convincingly that his proof satisfies all three of these criteria. His premises “Here’s a hand” and “Here’s another” are different from the conclusion that there is an external world. They do entail that conclusion, because he is very clear in the full lecture that by “hand” he means an “object that can be met with in space,” i.e. an object in the external world. So if there are hands, there are external world objects. And he does know that he has hands. So all three of his criteria are satisfied.

If we are to object to Moore, then, we are probably best advised to do so by attacking his adequacy conditions for a good proof and saying that something more is required for a good proof than just satisfying those three criteria. But what more is required? The obvious answer is to say that his third criterion is too weak: it is not enough to know that we have hands, we must be able to prove that we have hands. But if we insist that for a proof to be good, we must also have in hand another proof, viz. a proof that the premises are true, then we kick off an infinite regress, for all of our proofs would require other proofs ad infinitum. So the requirement that we be able to prove our premises is too much. There are things we can know that we cannot prove, and one of the things that we know but cannot prove is that we have hands. But if we know that we have hands, we can prove that there is an external world.

I think that all of this is basically correct, and obviously so. I have only two quibbles. First, when Moore says that he has proved that there is an external world, this implies for many readers that we may be rationally certain that there is an external world. But that is a stronger conclusion than what is supported by Moore’s argument, and probably a stronger conclusion than Moore ever intended. At best we can say that Moore’s conclusion can be known with the same degree of confidence as his premises, which is very high but not maximally high. In this way, Moore’s proof is still consistent with falibilism, which makes it unlike mathematical proof. Moore’s proof would be better if it were clearer on this point. Second, Moore gives very little in terms of a positive account of how we can know that we have hands if not by proving it. And in other work, he appeals to common sense as the great philosophical arbiter. As I mentioned in my previous answer, I think appeals to common sense are almost always misguided. So there’s a huge lacuna in Moore’s proof that he does a poor job of filling in in other work. But others have done a better job of filling in that lacuna, so I consider this a mild criticism of Moore.

3. Consider the statement, ‘There exists some planet N, located beyond the observable universe (in such a way that the existence of this planet is fundamentally impossible to confirm or refute empirically, either now or in the future), and on N there is silver’. Can this statement be considered meaningful, or is it merely a nonsensical collection of words? Does this statement possess a truth-value (i.e., either it is true or it is false), given that it cannot be empirically confirmed or refuted?

This is a meaningful statement with a truth value. The idea that something is true only if it is empirically testable is a dogma of logical positivism, and philosophy is glad to be rid of it. Claims represent the world as being a certain way, and they are true if the world really is that way, whether or not we are in a position to confirm that the world is that way.

I also take issue with the framing of the question. We know that there are planets in the observable universe on which there are silver (Earth, e.g.). We also know why that is the case. Planets are formed from heavier elements which are ejected from exploding stars, the fusion cores of large stars is capable of generating almost any heavier element, including silver. And large silver-generating stars are widespread throughout the observable universe. So unless the unobservable universe (i.e. areas of the universe outside of our light cone) is fundamentally unlike the observable universe, there is probably silver on planets in the observable universe. And we should not think that the unobservable parts of the universe are unlike the observable parts of the universe unless we have some particular strong reason to think so. To do otherwise is to reject the assumption of the uniformity of nature, which is a fundamental presupposition of all science. So by any reasonable scientific standards, we do have evidence that planets outside of our light cone contain silver. The idea that we can’t have evidence about any state of affairs that’s not directly observable is another dogma that we are better rid of.

4. Is skepticism about the external world exclusively a product of Early Modern European philosophy, or does it possess universal relevance?

External world skepticism is not at all exclusive to Early Modern European philosophy. There are external world skeptical arguments in western antiquity as well as in ancient Indian philosophy, particularly Buddhist philosophy. Indeed, these other traditions seem to take external world skepticism much more seriously than the early moderns did. Descartes was not a skeptic, nor was Berkeley. Even Hume and Kant only endorse a very limited form of skepticism. But Pyrrho famously (albiet implausibly) was such a committed external world skeptic that his students had to lead him around for he so lacked confidence in his own senses that he wouldn’t move anywhere of his own accord. And Buddhist philosophy takes the inter-dependence of all things (and thus the dependence of the world on our minds) and the resulting impermanence of all things very seriously indeed. It is more accurate to say that Early Modern European philosophy uses the threat of external world skepticism in distinctive ways to motivate some distinctive epistemological conclusions.

5. How do interpretations of quantum mechanics that posit consciousness as a factor in wave function collapse influence debates about the independent ontological status of the external world?

To the extent that interpretations of quantum mechanics that posit consciousness as a factor in wave function collapse influence other debates, this is a deplorable state of affairs and everyone involved should cut it out.

I’m sure many readers will be somewhat mystified by this question, so let’s back up a bit to talk about quantum mechanics and what the whole idea of “wave function collapse” is about. The underlying issue is related to what is known as the “measurement problem” in quantum mechanics. The issue is that the fundamental equation of quantum mechanics, Schrodinger’s equation (known as the “wave function” because the math is modeled on the kinds of math used to describe waves) is indeterministic. It does not give exact values for, e.g., the location of subatomic particles. Rather, it gives values that can be interpreted (via the “Born rule”) as probabilities for the location of any given particle. In this way, the wave function deals in fuzzy probability fields, rather than in determinate locations for any given particle. And yet, when we set out to measure the locations of particles, we are capable of doing so, and we always find the particle in one determinate location. We just can’t predict what that location will be in advance by using the wave function, although we can say something about the probability of finding it in any particular place.

For about a century, there has been a rather heated debate about what to make of this. On one side of the debate, we have figures like Einstein and Schrodinger, who held that the indeterminism of the wave function is fundamentally an epistemological matter. The particle has a determinate location, the wave function just doesn’t tell us what it is. (And we can’t get any more specific information about that location because of the prohibitive difficulty of measuring subatomic particles; it is physically impossible to generate tools that are subtle enough to measure the subatomic without altering the underlying reality so much that we wouldn’t know what it was like before we measured it.) On the other side of the debate, we have figures like Bohr and Heisenberg, who held that quantum mechanics is fundamentally complete. There are no states of affairs other than those given by the wave function, so if the wave function doesn’t assign a definite location to a subatomic particle, then there is no definite location for the subatomic particle.

The Bohr/Heisenberg position became known as the “Copenhagen interpretation” of quantum mechanics, named after Bohr’s Institute of Theoretical Physics at the University of Copenhagen. Many great physicists of the early 20^th century spent time at the Copenhagen Institute (including Heisenberg), and, under the influence of Bohr, all accepted the “Copenhagen interpretation.” Yet there is not any one view which we could consider to be the Copenhagen interpretation. All the members of the Copenhagen group thought that it was fundamentally mistaken to talk about the location of a particle before it was measured. Yet they disagreed about why. Bohr thought it was mistaken because he was of a positivist or Kantian bent: there is no reality other than what we observe, for Bohr, and so it made no sense to talk about the state of the physical system when it is not being observed. Others disagreed, saying that the underlying physical system was itself indeterminate in some way, but that the act of measurement caused the physical system to become determinate.

Yet that is a strange thing to say in its own right. What is “measurement,” exactly, and how can the act of measuring cause the underlying physical system to assume a determinate state? Different Copenhagen members said rather different things in answer to that question. And no wonder! The question concerns something that is fundamentally not scientifically decidable. We know the wave function works, in the sense that it has been immensely theoretically fruitful and we can use it to build new technologies. But there is a strange disjuncture between what the wave function describes (probability fields) and what we actually observe (determinate locations for particles). Explaining that disjuncture is known as the “measurement problem,” and it is a thorny question precisely because it is a metaphysical question in the sense that its answer lies outside of the bounds of empirical science.

So to quickly sum up what we’ve said so far, we have looked at three responses to this measurement problem. One is to follow Einstein in holding that there is a fact about the location of particles, we just don’t know what it is. The second is to follow Bohr in saying that questions about the location of particles are metaphysical and therefore meaningless, and so we should simply ignore them. “Shut up and calculate.” The third is to follow Heisenberg and other of Bohr’s colleagues and say that subatomic particles alternate between two states, an indeterminate state and a determinate state.

The main problem with this third option is that it posits a process whereby particles move between two distinct states, determinate and indeterminte, where the move from indeterminate to determinate is known as “wave function collapse.” But this process is not described by the wave function itself. It’s a kludge put in place to reconcile the indeterminacy of the wave function with the determinacy of measurement. What do we make of that? One approach, the “multiple worlds” approach, denies that the wave function ever collapses. The universe is just endlessly propagating into infinite branching possibilities, and the determinacy of measurement is an illusion created by the fact that any particular measurement occurs only on one branch. A second approach, the “spontaneous collapse” approach, says (to oversimplify a bit) that the wave function collapses by itself all of the time, and that the universe is constantly fuzzing into indeterminacy and then snapping into determinacy an immeasurably small fraction of a second later. A third approach says that the wave function does indeed collapse, and that measurement is what causes the collapse. But again, what is measurement, and how does measurement cause the collapse?

The traditional answer, put forward in one form or another by most of the Cophenhagen group, was to say that some systems are classical and others are quantum. Measuring devices are classical systems; scientific instruments are not subatomic particles described by the wave function. So we may say that “measurement” consists any interaction between classical systems and quantum systems; thus all interactions with scientific instruments will count as “measurements,” and this explains what we want to explain. With this rough account in hand, the Cophenhagen interpretation became hegemonic within physics by the mid-20th century. Yet it has an obvious problem: “classical” systems are composed of quantum systems. If we bounce a photon off of a mirror, the photon isn’t just interacting with a classical mirror, it’s interacting with the various subatomic particles that compose the mirror at the point of impact. So this interaction between the quantum and the classical is also an interaction between quantum and quantum. Realizing this makes the distinction between quantum and classical unclear, and this degrades the intelligibility of saying that measurement consists in the interaction between quantum and classical systems.

And so Eugene Wigner introduced a kludge to save the kludge of wave function collapse. He said that measurement is a conscious awareness of the quantum system. Thus consciousness itself is the cause of wave function collapse! Among working physicists, this suggestion was regarded with a fair amount of derision. Not only is it a silly suggestion on its face, it attempts to engage in a metaphysical project that no one saw the use of anymore. The heated metaphysical debates of the early 20th century had cooled substantially. Quantum mechanics was now established to work extremely well. The Copenhagen interpretation dominated not because it was a good solution to the measurement problem, but because it seemed to license indifference to it while physicists went about doing the real work of applying and extending quantum theory. But while Wigner’s suggestion was met with amused indifference by the physics community, it was embraced enthusiastically by the mystics of the world. And that brings us to today.

I rehearse this history in order to make the case that the idea that consciousness causes wave function collapse is a silly theory. It’s a kludge on top of a kludge that is relevant only within one approach to quantum mechanics, a metaphysical theory in the most pejorative sense of the term “metaphysical.” To the extent that this view has had an impact on any other debates, something has gone very wrong, and everyone involved should take a good hard look in the mirror.

6. What is your stance on the simulation hypothesis, the idea that our world is a computer-generated construct?

To the extent that the “simulation hypothesis” is just another way of describing the “brain in a vat” hypothesis, it’s another skeptical hypothesis that we can reject on the basis of lack of evidence, as I said in my response to the first question.

However, when people talk about the “simulation hypothesis,” they are often pairing the standard “brain in a vat” hypothesis with a further argument for the conclusion that it’s actually likely that we’re in a simulation. That argument says that any sufficiently advanced civilization will be interested in creating simulations of intelligent life, and indeed a vast number of such simulations. Those simulations of intelligent life will themselves be intelligent. And so, in any universe containing a sufficiently advanced civilization, the vast majority of intelligences will be simulated intelligences. Therefore, we are probably simulated.

Every single premise in this argument is highly dubious. The idea that sufficiently advanced civilizations would create vast numbers of simulated intelligences is speculative at best. The idea that simulated intelligences would themselves be conscious intelligences like us relies on a very strong version of functionalism about the philosophy of mind that is, again, speculative at best. And the idea that a large number of possible simulated intelligences makes it likely in some epistemically relevant sense that we ourselves are simulated is a horrible abuse of probability theory.

7. What texts would you recommend as essential reading on skepticism about an external world?

When it comes to “essential” reading on skepticism about an external world, I’d just list the classics. Descartes’s Meditations, particularly Meditation 6 where he dispels the skeptical worry that he cooked up in Meditation 1. Berkeley’s Three Dialogues. Hume’s Enquiry Concerning Human Understanding. GE Moore’s “Proof of an External World.” Chapters 1-4 of Bertrand Russell’s The Problems of Philosophy. Hilary Putnam’s “Brains in a Vat.” They’re the classics for a reason. But in addition to these, I’d also recommend Kevin McCain’s new book, Explanatory Solutions to Skeptical Problems. That is the most well-developed version of the anti-skeptical strategy that I find to be most promising.

I do not recommend reading Kant.

Subscribe now

Given this observation, you might now be expecting to turn things around a bit and say that this is bad. It’s not. I’m an atheist myself, and this state of affairs seems entirely good and proper. Theistic views in philosophy are, on the whole, laughably misguided. That is my professional judgment. Perhaps I’m just a product of my socialization here (although I was raised Christian; it just never stuck). But religious approaches to philosophical problems all seem to me to be hilarious non-starters. This is not a distinctive problem for religious philosophy. Whenever I read a paper where people start talking about what occurs in various “possible worlds,” I know I’m in for some utter horseshit. The same goes for papers that take the notion of “credence” literally. And so also for papers that go on about “God.” What’s interesting to note is just that there are many philosophers who take possible worlds and credences seriously, but precious few who take God seriously.

This is something of an anomalous situation. Until the 1700s or so, it was absolutely de rigeur to use the notion of God to solve philosophical problems. Indeed, the dominant approach was to assume that God was obviously part of the solution to all philosophical problems. The only trick was to explain how. And even within contemporary (non-academic) culture, religious ideas are widespread, and the idea that religion is part of a solution to philosophical problems is equally widespread.

Nowhere is this disjunction between popular philosophy and academic philosophy more pronounced than my own particular area of specialization, metaethics. Metaethics involves a cluster of (primarily) metaphysical, linguistic, and epistemological questions regarding “first-order” ethics. Where first-order ethics asks what kinds of things are good, metaethics asks what goodness even is, what the word ‘good’ means, and whether (and how) we can know that something is good. Contemporary metaethics is fully atheistic. The idea that goodness has something to do with Godliness (etymology aside) is simply not contemplated in the contemporary metaethics literature. To which I say: good! But for many non-academic philosophers, the idea that good and God are inextricably linked is just common sense. It is widely held that atheists simply can’t believe in ethics, that atheism just implies nihilism.

I don’t understand why anyone thinks that. This is particularly mystifying to me because the vast majority of philosophers are not nihilists. Many are some form or another of anti-realist. But the majority position among contemporary philosophers is that there are objective moral facts but that there is no God. I don’t understand why so many people think that this majority position is incoherent.

This is not for lack of trying. I think it’s very important to try to give a charitable reading of your opponent’s views. I’ve found that, most of the time, if some philosophical view is simply baffling, that’s because you haven’t adequately taken the time to think through it and understand its appeal. So I’ve tried to understand this argument, and I keep on bouncing off of it. This is, in large part, because no philosophers think this argument is any good. There are some theistic philosophers, and even a few theistic philosophers who work in metaethics, but none that I can think of who think that atheistic metaethics is flatly impossible. And if I try to get theists who passionately affirm this argument on social media to explain their thinking, it goes poorly. Talking through a complicated argument step-by-step is a skill that is not possessed in abundance by the average Twitterite.

So I’ve tried to work through the argument myself from plausible first principles, and I’ve ended up in kind of a weird place. I feel like this can’t really be what most common-sense theistic moralists are on about when they claim that it’s obvious that atheism implies nihilism. But since I can’t make any better sense of the position, I figured I’d present this to see how much it resonates. And if I’m clearly wrong, will someone please give me a better understanding of the argument?

Subscribe now

Begin at the beginning: why isn’t theistic metaethics popular, even among philosophers who believe in God? It really all goes back to a dialogue by Plato, the Euthyphro, which has Socrates debating the nature of piety with the priest Euthyphro. Socrates poses to Euthyphro the following question: do the gods love what they love because it is pious, or is the pious pious because it is loved by the gods? In a more contemporary context, we might render this: Does God love what he loves because it is good, or is the good good because it is loved by God? The common sense answer, which Euthyphro promptly gives in the dialogue, is that God loves what he loves because it is good. God knows the good, and he is good, and his goodness consists in his responding correctly to the good that he knows.

But Socrates points out that this is inconsistent with saying that goodness is identical to being loved by God. For if God loves what he loves because it’s good, and goodness = being loved by God, then God loves what he loves because it’s loved by God. But that’s gibberish. Things have the property of being loved by God because God loves them, not the other way around. (When Socrates points this out to Euthyphro, Euthyphro gets very angry and tries to change the subject.)

Of course, this all only follows if we say that God loves what he loves because it’s good. What if we give the other answer and say that God doesn’t love what he loves because it’s good? But then why does God love what he loves? For no reason at all? Then God’s commands are entirely arbitrary, and morality is merely divine whim. For some other reason? That “other reason,” whatever it is, would again be arbitrary unless that other reason were itself a moral reason. But then we’re back to saying God loves what he loves because it’s good.

So there are two choices: either morality is arbitrary or goodness has a reality independent of God’s love. Arbitrariness seems like a bad thing to say, and inconsistent with the idea that God is good. So the obvious thing to say - as Euthyphro realized - is that goodness has a reality independent of God’s loves/commands. Thus, it’s not incoherent to say that goodness is independent of God’s commands; if anything, it’s incoherent to deny this! And if it’s not incoherent to say that goodness is independent of God’s loves/commands, it’s not incoherent to believe that goodness exists but God does not. So atheism does not entail nihilism; if anything, theistic ethics is itself incoherent.

What’s the counter-argument?

There’s a large literature on this “Euthyphro dilemma,” where the mainstream theist response typically involves drawing some fine distinctions about what it means to say that God is good. It’s not the case that God is good in the sense of God’s having the property of goodness. Rather, God himself is identical with goodness. This is supposed to make the case that God’s desires are arbitrary less troubling. But (a) it’s not clear whether this really does help with the arbitrariness concern, and, more importantly (b) the common sense theistic view is that God has the property of goodness! “God is great, God is good, let us thank him for our food” doesn’t mean that God, though he be great, is not himself good, but is instead the very property of goodness. This is vague because the English word ‘is’ is ambiguous between the ‘is’ of identity and the ‘is’ of predication. But I imagine most people would be surprised to learn that God doesn’t have the property of being good. Perhaps God is good in both senses of ‘is,’ but this raises even more tangled metaphysical questions about whether properties can have themselves as properties. More damningly, it makes God’s goodness trivial in a way, since his goodness consists in him simply living up to the standard of himself.

I’m not saying there’s no a rich vein to mine here - philosophers have been mining it for thousands of years - but I’m unsatisfied with this whole approach because it doesn’t look like it does a good job capturing the common-sense thought that atheism implies nihilism. Whatever that common-sense thought is, it’s not one that can be best described by careful parsing of different senses of the word ‘is’ in an attempt to rescue divine morality from a concern about arbitrariness.

Subscribe now

I think the thought is much simpler. The problem with saying that God isn’t responding to an independent standard of goodness is that it renders God’s goodness incoherent or else weirdly trivial. But the problem with saying that God is responding to an independent standard of goodness that is independent of his own existence is that it posits something whose existence is independent of God. But everything's existence is dependent on God - God is the creator of everything! So the response to give to the Euthyphro dilemma (if a response must be given) is that while goodness is independent of what God loves or commands (and God is good in light of the fact that he loves what is good), goodness is not independent of God’s creation. So goodness still depends on God. And that’s why it’s incoherent to be an atheist who believes in goodness.

I think that’s the argument.

But if that’s the argument, there’s an obvious problem with it. The fundamental flaw that the atheist makes, according to this argument, is that the atheist believes in something that was created by God without believing in God. But according to the theist, God created everything. So if you believe in anything without believing in God, you’re making the same mistake as the atheist who believes in goodness.

But this isn’t how people argue. I see theists regularly scoff at the idea that anyone could believe in ethics without believing in God. Yet I haven’t seen anyone ever scoff at the idea that anyone could believe in tables without believing in God. Why not? I feel like if I understood the answer to that question, I’d be closer to understanding why theists think that atheism implies nihilism.

Chandler: Kathy and I got to talking, and one thing led to another, and…
Joey: And what?! Did you sleep with her?
Chandler: No! No! No! I just kissed her.
Joey: What?! That’s even worse!
Chandler: How is that worse?!
Joey: I don’t know! But it’s the same!

The last three lines of this exchange have popped into my head repeatedly over the last two weeks with the right’s full-on embrace of cancel culture following the murder of Charlie Kirk, and the subsequent cancelation of Jimmy Kimmel’s late night show. Because it seems like I’ve seen some version of the following exchange at least 100 times in recent days:

Free Speech Defender: It’s short-sighted to have the FCC lean on ABC to get Kimmel canceled. It sets a dangerous precedent. Next time the left is in power, they’ll use that power to censor conservatives.
MAGA Supporter: Oh, come on! The left has been using their power to censor conservatives for years now. What about the censorship of the Hunter Biden laptop story? That’s even worse!
FSD: How is that worse?
MAGA: I don’t know! But it’s the same!

Criticisms of “whataboutism” are ubiquitous, and I basically agree with the criticisms. If you do something bad, it is no defense to point out another time that the other side did something bad. Two wrongs don’t make a right, after all. And the bad thing the other side did is never exactly the same as the new bad thing the whataboutist is trying to defend. Lesser wrongs can’t possibly justify greater wrongs.

Still, I want to dig in a bit here, because I think the whataboutist dialectic is not really well-understood. And that’s a problem, because that whataboutist dialectic is, in many ways, the defining characteristic of our current hyper-partisan age.

Subscribe now

Let’s talk about the cancelation of Jimmy Kimmel, and let me begin by saying, in unequivocal terms, that this was bad. The FCC leaning on ABC to get Kimmel canceled for his speech on the show earlier this week is indeed a very bad thing, and should be roundly condemned by everyone. FCC commissioner Brandon Carr’s comments on a podcast that his approach to ABC was “we can do this the easy way or the hard way” is the kind of cartoonish thuggishness that would be funny if it wasn’t being stated in earnest by a top administration official.

But it’s important to keep in mind the exact nature of the pressure that was brought to bear on ABC. There wasn’t an official action brought by the administration against ABC, and the “we can do this the easy way or the hard way” line was not said by Carr to ABC, but rather his summary of his negotiating stance on a sympathetic podcast. What actually happened is that the FCC contacted ABC and said that Kimmel’s speech was a problem, and they should probably do something about it. And ABC, out of a combination of agreement that Kimmel was a problem and fear of what the FCC might do if ABC did nothing, canceled Kimmel as a result. That is, the government asked a private entity to censor on its behalf, and the private entity agreed. This is known as “jawboning.” And again, it is unequivocally a bad thing. The courts have been somewhat equivocal on whether it is strictly unconstitutional under the first amendment. But whether or not it’s unconstitutional under current doctrine, jawboning is an abuse of power, and shouldn’t be done under any circumstances.

And that’s an important conclusion, because Democratic administrations engaged in extensive jawboning. That was bad, and an abuse of power. The most recent, blatant example of this was the Biden administration’s jawboning of social media companies to change their moderation policies to combat right-wing misinformation. That this jawboning occurred is a matter of public record; the so-called “Twitter files” document the Biden administration’s efforts. (A lawsuit was filed by the states of Missouri and Louisiana to prevent this jawboning, but the case was eventually dismissed by the Supreme Court, 6-3, on the grounds that the states of Missouri and Louisiana didn’t have standing to sue.) This jawboning occurred against the background of increased pressure on social media and other “big tech” companies by Biden administration officials, like former FTC head Lina Khan, who wanted to break up or heavily regulate those companies on anti-trust grounds. (Those efforts led to the rightward swing of Silicon Valley in recent years.) And so social media moderation policies were changed, out of a combination of agreement and fear. The jawboning worked.

I would venture that most of the people who are (correctly!) incensed about the Kimmel cancelation didn’t immediately think, “Hmm, another case of jawboning, just like with the Twitter Files.” I didn’t think that. I had to be reminded of the connection by conservative whatabouters. And that gets to the first part of the dynamic I want to highlight: the bad things that are done by your side get memory-holed. The right has spent the last few years constantly reminding each other about social media censorship and marinating in anger about it. The left forgot the story the moment that it fell off the front page. So the Kimmel cancelation feels entirely unprecedented, even though it’s not. We just forget the precedents that look bad for our own side. Next time, when the shoe is on the other foot and the Democrats are attempting to censor the people on the right, they’ll be shouting “How can you complain after what you did to Jimmy Kimmel?” And the people on the right will respond, with complete sincerity, “We did what to Jimmy who?”

Another reaction that I expect, now that I’ve brought up social media censorship under Biden and compared it with the Kimmel cancelation, is that people will say this is an inapt comparison. The two cases are, after all, not exactly the same. And in the differences, we can find exoneration for Biden. Right wing misinformation is a real problem on social media. And might it not be the case that social media companies realized this, and worked together with their partners in government to try to create a better social outcome? Yet we could say the same things about the Kimmel firing. Kimmel’s offense was to state baldly that Kirk’s killer was on the right, when that was always at best speculation and now seems to have been definitively disproved. Was Kimmel not spending left-wing misinformation? That’s a real problem! So might it not be the case that ABC realized this, and worked together with their partners in government to try to create a better social outcome? The justifications line up quite nicely. And indeed, we see the right offering up these same justifications now. They claim that (A) Kimmel’s remarks really were beyond the pale, and so (B) ABC really ought to have fired him no matter what the FCC said, and (C) ABC would have gotten to that conclusion even without FCC intervention, and so (D) what intervention there was had no real impact and (E) whatever impact they had was a good thing anyway. That’s more or less the precise line of defense that was offered of Biden’s efforts to combat right wing misinformation by jawboning social media companies.

Yet there are differences, and I’ll allow that they make the current actions of the Trump administration worse than the prior actions of the Biden administration. Lina Khan wasn’t going on podcasts to say that social media moderation would happen “the easy way or the hard way.” So yes, Biden was bad, but Trump is worse. But this is just how the ratchet happens. One side does something bad, the other does something slightly worse next time they’re in power, the first side does something even worse the next time, and so on and on.

This isn’t how it feels, though. And it doesn’t feel that way because your own side’s malfeasance felt justified in the moment and then was immediately forgotten the next moment. So when the other side engaged in the ratchet and did something slightly worse, it felt like a whole new thing, an unprecedented attack on liberal norms. The whataboutists start whining in response to your outrage, harping on about the last bad thing your side did rather than try to justify their action on the merits. (Well, they do try to offer some justifications, but they’re transparently flimsy, and they’re clearly motivated by revenge and the will to power.) That just shows how morally and intellectually bankrupt the other side is. Hearts harden, partisanship increases, and next time you’re in power, you feel emboldened to take the next step, since unilateral disarmament is simply surrendering to the hostile other. And when you take that next step, the other side is shocked. They’ve done nothing to deserve this! But of course they have done something to deserve this, as you’re all too happy to point out. “What about…?”

It’s a bad cycle, and I don’t see any way out of it.

I mean, there is a way out. We should all agree to “unilateral disarmament,” sincerely embrace liberal norms, recognize when those on our side violate those norms, and be willing to harshly punish them for doing so because having principles is the only alternative to this ratchet. But I’m not holding my breath.

Subscribe now

For those with training in philosophy, this quick back and forth is extremely well-known. (Those without philosophical training often find themselves playing the role of the scientist in that exchange; I see some version of it play out once a month or so on social media.) The lesson that follows is simple and devastatingly compelling: philosophy is not all bullshit. If philosophy is all bullshit, then the claim that philosophy is all bullshit, which is itself a philosophical claim, is therefore itself bullshit. If philosophy is bullshit, then it’s bullshit that philosophy is bullshit. Therefore, philosophy is not bullshit. Case closed, philosophy wins.

Or does it? For certainly some philosophy is bullshit. The idea that we can understand important facts about the nature and structure of reality just by sitting back in our armchair and thinking about it really hard has always had at least the slightest whiff of the incredible about it. And many of the theories that philosophers have come up with over the ages have the overwhelming odor of bullshit about them, clearly untrue and perhaps so confused and incoherent that they don’t even rise to the level of falsity.

But how can one prosecute the case? For in trying to spell out what makes the clearly bullshit philosophy count as clearly bullshit, you immediately run the risk of drawing a line that your own theories fall outside of. Anti-philosophy philosophy has an overwhelming tendency towards being self-contradictory and self-undermining. We’ve learned that lesson well. But yet we’ve learned it too well, for the sad state of affairs is that much of modern philosophy is clearly bullshit.

The solution to the problem, at least in outline, is to try to draw a distinction. There’s two kinds of philosophy: light, good philosophy and heavy, bad philosophy. Once we’ve drawn that distinction, we have the ability to attack philosophy without paradox. Heavy, bad philosophy is bullshit. Now, the claim that heavy, bad philosophy is bullshit is a philosophical claim, to be sure. But it’s a light, good philosophical claim. And light, good philosophy isn’t bullshit. Only the heavy, bad stuff is. This is not a self-contradictory position.

But then we face the problem of how to draw distinction between the good philosophy and the bad philosophy. How much philosophy counts as good, and how much philosophy counts as bad? Any way of trying to draw this distinction faces a version of the same problem all over again. If we are very generous about what counts as good philosophy and confine the label of “bad philosophy” to only a small region of the most esoteric metaphysics, then our thesis lacks bite and we risk allowing in much errant bullshit under the label of “good philosophy.” But if we are stingy about what counts as good philosophy and allow only a select few claims to enjoy that privileged label, we better be damn sure that our reasoning that supports our line-drawing exercise only involves those select few good philosophical claims that pass the test, or else the our anti-philosophy philosophy will fall into incoherence. So to be a coherent anti-philosophy philosopher means walking a rather fine line.

No attempt to walk it has met with much success. The last sustained attempt was logical positivism, which was a dominant approach in the early 20th century before it was pretty decisively refuted as incoherent and self-contradictory in the mid-late 20th century. (The seeds of its demise were planted as early as the 1940s but they didn’t fully sprout until the 70s.) Philosophy today is mostly divided between two camps. On one hand are those who follow (the later) Wittgenstein in embracing incoherence; if we’re all just screwing around with nonsense theories, it hardly matters if they’re logically incoherent. It’s not even clear that these nonsense theories could even be logically incoherent — applying logic to philosophy is a kind of category mistake — and so logical criticisms like worries about being self-contradictory always fall wide of the mark. On the other hand are those who think that incoherence is genuinely bad, yet anti-philosophy philosophy is always incoherent, so philosophy is basically fine; that’s not to say that all philosophical theories are true, of course, just as not all scientific theories are true. But there’s no “good philosophy” and “bad philosophy,” there’s just true philosophical claims and false philosophical claims, and we decide between the two by using the methods of philosophy.

I’m not in either camp. I hold out hope that we can draw the line between a narrow good philosophy and a broad bad philosophy, and that the resources we need to do so can all be found on the good side of the line. I won’t try to say how here - that’s a big project, and one that I can’t hope to begin to make clear or compelling in a short blog post. Hopefully this will be my next book. All I’m confident to say now is that the current state of affairs is unacceptable. Philosophers are (by and large) divided between those who are self-consciously incoherent and those who wrongly think that they are coherent. The right balance between confidence in the success of a philosophical project and skepticism about the prospects for philosophical success is extremely difficult to strike. But wrestling with this tension and trying to resolve it is in the only way to approach philosophy seriously.

…I hope I can prove that.

Subscribe now

I’ve always loved that answer. While this may not have been his intent, I took away from the exchange the idea that the standards for belief and the standards for defending something in print are different. You should believe something if the evidence supports it, and it seems highly likely to be true. But it takes something different to defend something in print. You should defend something in print if there are cogent arguments in favor of it, and the strength of those arguments hasn’t been sufficiently appreciated. If something is worth taking seriously, but isn’t currently being taken seriously, it’s worth defending in print, even if you’re not sure you really believe it.

So let me defend affirmative action for conservatives in academia in print.

Subscribe now

The question of affirmative action for conservatives in academia became a hot topic of conversation for a few days when Donald Trump threatened Harvard with severe consequences unless it gave in to a list of demands that went well beyond the demands that were made of Columbia a few weeks previously. One of those extra demands was a demand that Harvard engage in affirmative action favoring conservatives. This caused a massive uproar. Among liberals, this was lambasted as an assault on academic freedom, which it was. But even conservatives were largely opposed. I mean, the Trump cult loved it, but they love anything Trump does, so they can be safely ignored. But conservative troll-cum-intellectual Richard Hanania took to the Economist to make the case that affirmative action is always bad and merit in hiring is good, and that didn’t change just because Trump wanted it turned to conservatives’ favor. Most conservatives that I follow on social media seemed to agree with this assessment.

My own reaction was more conflicted. I had just argued that academia could not survive Republicans holding political power unless the institution was remade to be something that Republicans could support, however grudgingly. Part of that remaking, I argued, would have to involve affirmative action for conservatives. That was not really an argument for affirmative action, but rather a descriptive analysis of the dilemma facing academia: the choice is to implement affirmative action or face draconian funding cuts. I left it to the reader to decide which of those alternatives was more unpalatable; the point was just that that is the choice academia faces. My mood soured further on the topic once Trump started making these demands explicit. The choice remained the same, but the bitter pill of reform now came served with a side of even more bitter cowardice and capitulation to petulant authoritarianism. Resistance and financial ruin are all but assured.

But these political conflicts are largely a distraction from a genuine ethical question of whether affirmative action for conservatives is justified. Trump is pushing Harvard to implement the policy out of bitter, foolish vindictiveness, and Harvard is righteously resisting. But if none of that were happening, would it be good for academic institutions to implement affirmative action? I think there’s a strong case to be made that it would. My argument, in short, is that the case for modest forms of affirmative action are stronger than conservatives realize. And the rationale that supports modest forms of affirmative action absolutely obtains with regards to the situation of conservatives in academia (although liberals are largely in denial about this).

Subscribe now

Let’s begin by saying what I mean by “modest affirmative action” and by examining what conservatives get wrong in their arguments for merit and against affirmative action. By “modest” affirmative action, I mean affirmative action as a sort of tie-breaker. Once all applications are in and the applicants are ranked according to merit, if there is a tie at the top, consideration should be given to other factors. This approach is opposed to more radical forms of affirmative action involving quotas, or where considerations like race or gender (or political affiliation) are used as a “first cut” to narrow the pool of applicants to only those from a targeted demographic. Affirmative action is, instead, a last cut.

I would also insist on rather stringent criteria for what it takes for two applicants to count as “tied,” such that affirmative action criteria might be brought to bear to break the tie. I’ve heard fellow philosophers argue that really the only criterion that matters when it comes to applying for a philosophy job is having a PhD, and so every applicant in the pool (apart from the one or two cranks who always end up applying) is effectively tied with respect to qualification. That being the case, we can make selections more or less entirely on the basis of demographic characteristics, secure that merit has been adequately taken into account. I reject that idea. When it comes to hiring for a research position, for instance, the quantity and quality of publications that an applicant has produced is clearly relevant. Someone with a dozen publications in top journals is more qualified than someone with no publications at all. This is obvious.

And yet things are rarely so cut and dried in actual hiring decisions. Sometimes there is one candidate who is a clear stand-out, with substantially more publications in substantially better journals than any other applicant. But often that’s not the case. Often, there are two or more candidates who are very strong, but in ways that it’s hard to meaningfully compare. One applicant has three publications, one in a top 5 journal, the other two in top 20 journals. Another applicant has six publications, 2 in top 20 journals, 4 in more niche journals, and none in the top 5. A third has only one publication, but it’s in Philosophical Review, the consensus best journal in the field. Which applicant has the strongest publication record? You could make a case for any of them. And that’s assuming we’re just looking at publication records. There’s another candidate in the pool whose publications are clearly (but marginally) less impressive, yet her teaching portfolio is outstanding. Teaching is a very important part of the job for almost all academic positions! And another candidate in the pool who is an accomplished conference organizer with a strong record of success in obtaining grants. Departments need people like that in order to thrive in the bureaucratic culture of contemporary academia. Which of these five candidates is the most qualified? Who comes out on top in terms of merit?

Decisions must be made, and so they are made. Hiring committees are a mess of squabbling (often congenial, but not always) about exactly what weights to place on each of these criteria. There will be the one person who’s really impressed by the Phil Review pub, the one who insists that we’re teachers first and foremost and that this is the most important thing to take into account (assuming that the publication record is strong enough to indicate a good candidate for tenure), and on and on. So it goes, and there’s nothing wrong with that. But to say that all of this squabbling is besides the point, for what really matters is who has the most merit, would be patently ridiculous. The squabbling results because each candidate has a plausible claim to being the most meriting. Given that each candidate has a plausible claim, what’s wrong with making the decision to hire the first black woman the department has ever employed? The objection is that this would put decisive weight on what is essentially an arbitrary factor. But that objection overlooks the fact that, in many cases, any decision will amount to putting decisive weight on what is essentially an arbitrary factor. If some further social good can result from doing so, then, honestly, that just sounds great. Let’s do that.

And all of this applies to the question of affirmative action for conservatives. If there is a genuine tie with respect to merit, what’s wrong with ensuring that there is conservative representation on the faculty? It’s no worse of an arbitrary factor than any other. And if some further social good can result from doing so, then let’s do that.

This just raises the question of whether some social good would come from making an effort to hire conservatives. And here it seems the answer is clearly yes.

Subscribe now

Let us set aside the purely pragmatic case for affirmative action for conservatives that I referenced earlier. There are two standard justifications for affirmative action. One is that diversity is in the interest of institutions, and affirmative action promotes diversity. The second is that affirmative action is an appropriate remedy for past (and perhaps ongoing) discrimination. Both of these justifications clearly apply in the case of affirmative action for conservatives.

Let’s take the diversity issue first. This point has been made so many times over the last decade or so that I feel kind of silly trotting out the old talking points, but these arguments are no worse for the wear. Intellectual diversity is healthy in any institution, but particularly in an academic institution. A university where people come from every country of the world, with every conceivable skin tone and sexuality proudly represented, and where everyone thinks exactly the same thing is not a healthy academic institution.

Progress comes from disagreement. If you’re wrong, you need someone who disagrees with you to correct you. “Hmm, maybe I’m wrong about this” is just not the sort of thought that will occur to anyone who is surrounded by a chorus of people agreeing with you all day. And even if the thought does occur to you, the chorus of agreement will discourage you from pursuing the thought further, particularly if there are social consequences to voicing that disagreement. And even if you’re right, there’s value in holding onto your beliefs as a living truth rather than a dead dogma. Maintaining a living truth requires deep understanding, and it is deep understanding, rather than superficial correctness, that is the key to further discovery.

These points are all so obvious that it’s almost surprising that they’re not universally accepted. That these are controversial points at all becomes intelligible once you realize that if intellectual diversity is important, then that suggests that some sort of affirmative action for conservatives might be appropriate. And the idea of affirmative action for conservatives is seen as so laughably misguided that valuing “intellectual diversity” (always put in scare quotes) is taken to be equally misguided. My friends, it is time to seriously consider that we run the argument in the opposite direction and take some serious steps towards remedying academia’s political monoculture.

This point has been debated over the last few days after an OpEd by philosopher Jennifer Morton in the NYT which has elicited substantial controversy. I don’t want to engage directly with Morton here, since I’ve seen many others do so (this piece sums up my thoughts pretty nicely). But the main objection I’ve seen to intellectual diversity in the wake of the controversy is that we, as an academic community, have already considered conservatism and rejected it as being wrong, and part of the whole idea of an academic discipline is that we reject wrong ideas once we’ve sufficiently established them to be wrong. Here’s Brian Leiter:

Diversity, including viewpoint diversity, is not an academic value. …[V]iewpoint diversity is irrelevant in serious academic disciplines; in less serious ones, it may be relevant, but there is no way to impose it without violating core academic freedom. Academic disciplines presupposes the unequal worth of different viewpoints, and the job of scholars is to assess those viewpoints, and discount the unworthy ones.1

There is a good point here. We no longer teach phlogiston in chemistry class, nor flat earth in astronomy or geology. Our understanding of the world progresses in large part by discarding things that we have found not to work. The idea that intellectual diversity should require us to teach or engage with flat earth theory is laughable. Might we say the same thing about conservative thought?

We might not. Let’s begin by disentangling, to the extent possible, empirical claims about the way the world is from moral claims that tell us how we should respond to the way the world is. When it comes to empirical claims, I hesitate to call any empirical claim “conservative” or “liberal.” The relevant distinction is between empirical claims that are true and those that are false. Truth and falsity are often hard to discern, but we have our methods for sorting the one from the other, those methods seem to work pretty well, and we should apply them impartially. Now there are some claims that are false and unsupported by evidence and which tend to be believed disproportionately by conservatives nonetheless. Young earth creationism, for one. I’m not saying that we should be inviting young earth creationists into the academy, at least not as geologists or historians. Of course, this point cuts both ways. Homeopathy is a bizarre empirical system of beliefs that is left-coded, and it has no place in medical schools. This is as it should be.

But moral beliefs are a different story. The sharpest controversies over conservative representation concern the inclusion of people with right-of-center beliefs about moral and political questions. These controversies are most pronounced when those questions concern race and gender, but these same controversies arise around every other moral and political issue - war, faith, political economy, liberty, equality, and on and on. For each of these issues, it’s not difficult to identify what would count as “conservative values” vs “liberal values.” And the academy just doesn’t have many people with conservative values. This matters the most when it comes to subject matters that either touch on moral issues or are themselves fundamentally moral or political disciplines. In most cases, it doesn’t much matter what a chemist’s attitudes towards the morality of war and peace are. But it is precisely in the disciplines that are moral or morality-adjacent that the bias against conservatives is most pronounced:

STATISTICAL bias, that is. I’m not talking about deliberate discrimination (yet). For now, our focus is on the question of whether or not it is sensible to discriminate against conservatives because we have determined that their beliefs are just wrong. Regarding moral beliefs, then, the claim would have to be that we, as an academic community, have come to the conclusion that conservative moral beliefs are just wrong, and so we are justified in rejecting conservatives in moral disciplines, just as we are justified in rejecting young earth creationists from the geology department.

This is a pretty incredible argument, since it presupposes both (A) that there are objective moral facts, and (B) that we can know those moral facts with something like the same degree of rational certainty that we have that young earth creationism is false. Are (A) and (B) true? Well, I certainly don’t think so. When I’m writing papers for publication in academic journals and not just tinkering on this here blog, my primary project is to argue that (A) and (B) are false. (Check out my book with Spencer Case on the matter!) So fully making the case that (A) and (B) are mistaken is a task that falls far outside the scope of this one blog post that’s already running far overlong. For now, I’ll content myself with the following three points: First, it’s hard to say what even could count as evidence for any moral claim. Most moral philosophers end up endorsing some version of “intuition” at one point or another, which is precisely as dubious as it sounds. Second, (A) and (B) as metaethical claims are themselves extremely controversial, and so it would be remarkable if a whole discriminatory hiring apparatus was based on them. Third, the idea that there are objective moral facts is widely rejected by the very people in the humanities who are so keen to discriminate against conservatives. When Brian Leiter isn’t arguing that it’s fine to discriminate against conservatives for having bad moral beliefs on his blog, he’s writing academic papers where he argues against moral realism on Nietzschean grounds.

Now, this isn’t necessarily inconsistent. Moral anti-realists don’t generally eschew first-order morality. They just understand their moral commitments as nothing more than their own personal attitudes. I’ve argued that this is consistent and sensible in my academic work. But that puts the idea that conservative values have been “discounted” as “unworthy” (to use Leiter’s words) in a different light. The argument, it appears, is not that we know that conservative moral beliefs are false. Rather, we as an academic community have just gotten together and decided that we don’t take conservative values seriously, and so, as a result, we’re happy to actively discriminate against conservatives in hiring. They’re not wrong, in any objective sense. We have just decided that we don’t want to have to deal with them.

Subscribe now

This brings us to the final question: Do academics discriminate against conservatives? Uh, yes. Obviously. First piece of evidence: that graph above. 5.5:1 ratios of liberal to conservative in the hard(er) sciences, 20:1 in the soft sciences and the hard(er) humanities, 40:1 or worse in the arts and critical disciplines. That doesn’t happen by accident. Second, direct evidence:

That chart is from a 2012 study by Inbar and Lammers.2 That’s 2012, before Trump, before the “Great Awokening.” Things have not improved. For those of you who go a bit cross-eyed when asked to read charts and tables (my sympathies!), this says that, for hiring decisions, most academics rated themselves at least somewhat likely to discriminate against conservatives, with 37.5% say that they themselves would be more likely than not to discriminate against conservatives, and 44.1% saying that of their colleagues.

Now you might not think that’s so bad. “Less than 50%…” But think about how hiring committees actually work, as I outlined above. There’s a debate over candidates, and a consensus needs to be reached. If the decision has to be unanimous, then a 40% chance of having a committee member biased against conservatives means that any committee of any decent size will have at least one biased member on it. (And then the recommendations of the committee go to the department as a whole, which can question the decision.) And even if unanimity isn’t required - as it usually isn’t, else no one would ever get hired - the existence of faculty members who are vocally opposed to hiring a conservative will count strongly against offering the job to the conservative for the rest of the committee. Because it is obvious that hiring the conservative will introduce a source of conflict into your nice, collegial department. Some people will go out of their way to make the conservative feel unwelcome; and even if that’s not what you will do, will you defend them? Then you’re on their side, battle lines are getting drawn, people stop talking to one another at department colloquia, and who really wants all of that? 40% of faculty being willing to discriminate against conservatives in hiring decisions translates into a much, much higher chance that no conservative will ever be hired.

And a third piece of evidence that academics discriminate against conservatives: In response to the recent controversy over intellectual diversity, one of the main lines of argument against intellectual diversity has been “But it’s good to discriminate against conservatives, because they’re wrong” (as we just saw above).

The common response to this worry is that conservatives are just less worthy of being hired. This response comes in three varieties. The first is the bald-faced “discrimination is good actually, provided it’s against conservatives, because I don’t like conservatives” line that we looked at a few paragraphs ago. Suffice it to say that I think this is a pretty loathsome attitude, and that one of the best justifications of affirmative action is that it can push back on bigoted nonsense like this. The second is the idea that conservatives are inherently less intelligent than liberals and that they are therefore less likely to even pursue careers as academics. I take this to be little better than the first variety. “There’s just not that many of them in academia because they’re intrinsically stupid” is the sort of line that appears easily in the mouths of bigots and segregationists of all stripes. “Of course they’re dumb, they’re conservative!” That plays well as a laugh line, but it’s got little basis in research on intelligence. It also seems to assume that in order for one to have the best values, one must simply be intelligent. Liberals are right and conservatives are wrong - just think about it, and that will become clear! Again, this assumes a certain perspective on moral epistemology that is highly controversial, to put it lightly, and most likely to be condemned in the abstract by those who are most eager to put it into practice.

It is true that conservatives are less likely to pursue careers as academics, but the better explanation for that is that every conservative with academic inclinations knows that the academic job market is highly competitive and that they are likely to be actively discriminated against on that job market. Why take the years to get a PhD when it will all go to waste? It’s true that universities can’t just turn around and fill up their departments with conservatives because there is a massive pipeline problem in conservative academia. But the traditional solution to pipeline problems that result from past (and potentially ongoing) discrimination is… affirmative action.

The third variety of “conservatives are less worthy to be hired” is the most defensible, since it comes from simple examination of CVs. Conservatives have fewer publications, and thus are less likely to be hired on the strength of their research. Accordingly, something more than the moderate affirmative action that I talked about at the top of this post would be required to increase conservative representation. This is often true (although there are exceptions).

There are two points to make in response here. First is that this CV deficit is itself often the result of discrimination. In the Inbar and Lammers study, 34% of academics said that their colleagues would be more likely than not to discriminate against a conservative paper if they reviewed it for a journal. This is very likely to make a downstream difference. (Unlike hiring committees, journal editorial decisions often do require unanimity for publication among the editorial board, since acceptance rates are kept low because journal pages are limited.) This discrimination is partly because of active anti-conservative bias, and partly because of other incentives. As one conservative academic recently recounted, he submitted a book proposal on conservative thought to a respected academic press only for that proposal to be rejected because it didn’t want to get a reputation as the kind of press that publishes more than a token handful of conservative books. That reputation would prevent it from attracting manuscripts from left-leaning authors, which all the stars in the field are. So conservatives are actively kept out of mainstream left-leaning journals and presses. Could conservatives just found their own journals? Eh. That takes a lot of time and money, and the journal would lose money because libraries wouldn’t pay to stock them. (Because no academics at the institution would request access from the library, because the established academics are all progressives with no interest in reading conservatives writing for other conservatives in the Journal of Conservatives.) The resulting journals would inevitably be considered “low quality,” of course, and thus publications in them would not really be useful in hiring decisions.

The second point is a concession. Yes, “moderate affirmative action” would not address the problem of anti-conservative bias, in publication or elsewhere, in light of all of the above. But that just means that moderate affirmative action is the least we could do in order to increase conservative representation in academia.

Subscribe now

Yikes, this went on longer than I intended. Sorry. I’ll sum up the main points:

• Hiring decisions are often inevitably made on more or less arbitrary grounds.

• Moderate affirmative action is therefore not worse than other rationales for hiring on arbitrariness concerns, and can therefore be justified if other goods can be achieved through those means.

• Affirmative action can remedy discrimination and increase diversity.

• Intellectual diversity is good.

• Conservatives have been discriminated against in academia, and remedying that is good.

• So moderate affirmative action for conservatives in academia is justified.

I’m willing to defend all of this in print. Now the argument has been made, in as much detail as I can muster. It cannot be dismissed with a laugh anymore. Where does it go wrong?

A month or so ago, Scott Alexander wrote a post entitled “Come On, Obviously The Purpose Of A System Is Not What It Does.” Alexander leads with a number of examples in which a system is set up to do something, but then, through bad luck or incompetence or as an inevitable side effect, does something else:

Consider the following claims

• The purpose of a cancer hospital is to cure two-thirds of cancer patients.

• The purpose of the Ukrainian military is to get stuck in a years-long stalemate with Russia.

• The purpose of the British government is to propose a controversial new sentencing policy, stand firm in the face of protests for a while, then cave in after slightly larger protests and agree not to pass the policy after all.

• The purpose of the New York bus system is to emit four billion pounds of carbon dioxide.

These are obviously false.

So clearly, the purpose of a system is not what it does.

Why does this matter? Well, because “The Purpose of a System is What It Does” is a slogan, typically used by the far left, which is frequently used to argue for radical social change. If something in society is having bad results, then its purpose is to have those bad results, and thus it must be totally destroyed. If police kill some number of unarmed black people, then the purpose of the police is to kill unarmed black people, and so police abolition is the only just response. Alexander replies, quite sensibly, that the purpose of the police is to protect and serve, and sometimes it just does its purpose badly.

Case closed? Not quite. Because there is a sense in which the purpose of a thing is what it does.

Subscribe now

The problem is that word, ‘purpose.’ As anyone who has studied Aristotelian philosophy knows, the word ‘purpose’ is incredibly slippery, and has a number of closely related meanings that we can disentangle. And when we’re dealing with words with multiple meanings, there’s a risk of mutual incomprehension, of talking past one another. (Alice: “Banks contain money.” Bob, standing next to a river: “No, banks contain the water.” Bob isn’t wrong, but his claim doesn’t make contact with what Alice is saying, because his use of the word ‘bank’ means a different thing than hers does.)

So what does ‘purpose’ mean? Two different definitions are relevant here. One is the one that Alexander has focused on, which is something like “the more-or-less explicit intention of the creator of the thing.” In this sense, the purpose of the NY bus system is to move people around in buses. That’s what the people who founded the NY bus system got together with the intention of doing, and set up the organization to further that end.

But consider the claim that “the purpose of the heart is to pump blood.” I think that everyone will agree that this is true. But what does “purpose” mean here? It doesn’t mean the more-or-less explicit intentions of the creator. I mean, a theist might say that it does, but (with all due respect to theists) come on. The usual definition for this use of the word ‘purpose’ is something like ‘the function that this part of the organism performs which promoted the survival of that organism and its ancestors in its evolved environment.’ That stuff about promoting survival in an evolved environment was added after Darwin to make sense of this idea of “purpose” in a Darwinian biological paradigm. The earlier, Aristotelian biological paradigm basically said that the purpose of a thing is what it does. The heart pumps blood, so that’s it’s purpose.

This is a point that someone raised to Alexander in response, and so in a follow-up post he tried to address this objection with the brief, pithy response: “Oh! I agree this makes sense if you need to talk about the “purpose” of an un-designed system with no humans in it.”

But this misses the point. For we have in biology (and other areas; the commenter that Alexander is responding to mentions the ersatz field of Systems Thinking) a completely coherent notion of a purpose that has nothing to do with design and intention. Can we not apply that notion within social analysis?

It seems quite obvious that we can. Take the modern, Darwinian definition of purpose, because that’s actually more or less exactly what we want. Biological organisms are extremely complicated physical systems, and that complexity evolved over time by a process of natural selection. There are some substructures within that system that are (partly) responsible for that organism and its ancestors being able to survive and reproduce within their evolved environment. The purpose is the survival-promoting function. Similarly, human societies are complex systems, and that complexity evolved over time by broadly Darwinian mechanisms. (Dysfunctional societies tend to die out while well-ordered societies survive and expand.) There are substructures within human societies, i.e. particular institutions, and some of those are partly responsible for the survival and expansion of the flourishing society because they play particular roles in the society. The institution’s purpose is the survival-promoting function.

Now institutions also have a “purpose” in Alexander’s sense, having to do with explicit intentions of the founders. But they may also have an institution in this other, Darwinian sense, and those two “purposes” may not be the same. The explicit purpose of a religious institution might be to glorify God, but the Darwinian purpose might be to promote a certain kind of social cohesion. So simply knowing the explicit purpose of an institution doesn’t rule out that it might have a different Darwinian purpose. And if it does have a different Darwinian purpose, it would be good to know that since it would help us better understand the nature of the society that we live in.

Of course, the Darwinian notion of purpose is more complex than just “the purpose of the system is what it does.” Emitting billions of pounds of carbon is neither the explicit purpose nor the Darwinian purpose of the NY bus system. And this sort of analysis doesn’t necessarily support leftist radicalism; it is neither the explicit purpose nor the Darwinian purpose of the police to kill unarmed black men.

But this notion of Darwinian purpose is still a useful one to have in our toolkit because it might well be an impetus to social change. It is possible for a stable society to be in a kind of sick equilibrium, where individual institutions survive by doing bad things (while professing to do good things), and where surrounding institutions have bent to accommodate these bad institutions so that the overall society can still thrive.

Perhaps the military-industrial complex is like this. We shouldn’t end our social analysis of the “purpose” of Lockheed Martin at the fact that Lockheed Martin proclaims on its website that it exists to secure the national defense, and that that is what its executives will say, in full honesty, if asked. If Lockheed Martin survives and thrives because of political corruption and its promotion of militarism (with all of the death and suffering that inevitably implies), that’s a good thing to know about Lockheed and a good reason to engage in some sort of social reform.

That sort of analysis is more complex than the simplistic “the purpose of a thing is what it does” slogan suggests. But it’s not that much more complex, and it’s both intelligible and useful. It certainly has a role in our analytical toolkit.

Subscribe now

It feels safe to come out of the closet as a libertarian because the insane authoritarian populism of Trump has made those on the left (my natural political tribe in a number of ways, and the probable majority readership of this blog) more sympathetic to my politics. I’ve seen a number of left-ish people online recently say some version of “I’ve never been a libertarian, but free minds and free markets sound pretty good around now.”

Particularly interesting to me was a recent post by popular progressive(-ish) econ blogger Noah Smith, who recently wrote that he had a newfound appreciation for libertarianism. While he got his start in blogging by raking libertarianism over the coals (as any good progressive-ish economist would do), he failed to recognize that libertarianism is good, actually. More specifically, while he harped on a number of ways in which the economy could be improved by being slightly less libertarian, he had failed to appreciate the various ways in which having libertarianism be a sort of default position in American politics (particularly right-wing American politics) kept the lid on some rather awful political tendencies that have come to assume power now that that libertarianism has been kicked to the curb. If you didn’t like the libertarian right, wait until you see the post-libertarian right.

And yet Smith, as a good non-libertarian, felt the need to open his piece with a list of ways in which he still disagrees with libertarianism. The second item on the list is:

By treating all of society as an interaction between a government and the individuals it governs, libertarians tend to ignore the threats to liberty from non-governmental institutions (“local bullies”), and from foreign governments. This led some libertarians to oppose the Civil Rights Act, and to underestimate the threats from illiberal powers like China. And these omissions led to some unsavory people grafting themselves and their oppressive ideas onto the libertarian movement.

So ok, let’s talk about the Civil Rights Act.

Subscribe now

I know, I know! I feel you reaching for your pearls again, ready to clutch. So let me begin by saying that I am not against the Civil Rights Act. I think it has been, on the whole, good. But as with much else in politics, there are good arguments to be made on either side. And I think that there are good arguments to be made against the Civil Rights Act. It is, on net, good. But that doesn’t mean there’s no bad to it. And the bad is, basically, the argument that libertarians have long had against the Civil Rights Act.

The libertarian argument against the Civil Rights Act is that it represents a dramatic (and arguably unconstitutional) expansion of government power to interfere with the way that private institutions and organizations conduct their affairs. Prior to the Civil Rights Act, if a business wanted to refuse service to black people, that was their right. They were utter dicks for doing so, but, as they say, it’s a free country. If you start a business, you can do with it as you please. And if you’re a racist asshole, you can conduct your business in a racist, asshole-ish way. That changed with the Civil Rights Act. The Civil Rights Act said, in so many words, that if you want to use your business (or other institution) to discriminate on the basis of race (or sex, or any other number of protected categories), then you are, at that point, the government’s plaything. They can fuck with you in any way that they want in order to get you to stop your discrimination.

This was a good thing (on net)! The system of Jim Crow apartheid in the south over the century leading up to the passage of the Civil Rights Act was put in place in part by official acts of government discrimination, but in larger part by the panoply of institutions of civil society - schools, businesses, churches, etc. - that made up southern life engaging in routine discrimination. Simply ending government discrimination would have hardly made a dent in Jim Crow. Something more sweeping and energetic was needed.

The right protested. Some protesters were racists, of course. Bull Connor’s objection was that segregation was good, and so the government forcing an end to segregation was bad. But libertarians like Barry Goldwater had a different objection. Of course segregation was bad, but empowering the government to end segregation would be to give the government too much power over private institutions.

There are basically two lines of response to this libertarian critique. The first is that the libertarians argued in bad faith. They were really segregationists like Bull Connor, and simply hiding it with high-minded rhetoric. Doubtless that is true of some self-described libertarians. But it wasn’t true of all. So the second response is that this is anti-authority objection is fundamentally silly. Government power was being wielded in novel ways, to be sure. But it was being wielded to end discrimination on the basis of race! Discrimination on the basis of race is bad. Very bad. And if federal government power is the only way to end discrimination on the basis of race, then government power is the appropriate medicine.

And so the Civil Rights Act was used in a variety of ways to force various private institutions to stop discriminating on the basis of race. Most famously, of course, the various shops, restaurants, and other businesses across the Jim Crow south could no longer refuse service to black people, as they’d long been accustomed to doing.

The remit of the Civil Rights Act is not limited to businesses. Even academia fell under the jurisdiction of the Civil Rights Act. Bob Jones University, a conservative university with regrettable views on race and race relations, was an active practitioner of race-based discrimination. They didn’t admit any black students until 1971, and even after maintained a policy forbidding inter-racial dating and marriage. In response, the IRS (citing the Civil Rights Act) stripped Bob Jones University of its tax-exempt status, forcing BJU to pay millions in back taxes.

Was this a good ruling? Well, it imposed a penalty for racist discrimination, and I can’t say I’m opposed to that. And it was entirely legal. If you engage in racial discrimination, you are, after all, the government’s plaything, and they are fully entitled to fuck with you until you shape up. That’s what the Civil Rights Act says (in so many words). And here it seems to have been used to bring about a good outcome.

But.

If you engage in discrimination on the basis of race, the Civil Rights Act says that you are now the government’s plaything. And affirmative action is discrimination on the basis of race. So, according to the Civil Rights Act, any institution that has a policy of affirmative action is the government’s plaything. And the government is, at the moment, largely controlled by one Donald J Trump, who is positively delighted to use the power of the government to fuck over any liberal controlled institution. And since approximately every liberal-controlled institution has had a policy of engaging in affirmative action for the last few decades, the government now has a right to fuck over every liberal-controlled institution. The Civil Rights Act says so.

This includes academia. Trump has already (famously) moved to strip Harvard of its tax-exempt status, on the basis of Harvard’s discrimination against whites, Asians, and Jews. It has not stopped there. The news today is that the Trump administration is launching a Civil Rights investigation into UCLA’s medical school for practicing affirmative action policies. They are entirely within their legal rights to do so; the Civil Rights Act says so.

Is this bad? I sure think so. But on what basis might one object? A first objection might be that this is a mis-application of the Civil Rights Act. But it is not. While proponents of affirmative action might make a distinction between good discrimination on the basis of race and bad discrimination on the basis of race, the Civil Rights Act does not. According to the Civil Rights Act, any discrimination on the basis of race is forbidden and may be stamped out with the full might of the federal government. This is both the plain meaning of the text of the law, and also how the law has been consistently interpreted by the Supreme Court.1 That liberal institutions who engage in affirmative action have not found themselves beneath the boot heel of the federal government thus far has been an exercise in forbearance by the executive. Trump is unique in that he is not particularly given to forbearance.

So on what other basis might one object? Here’s a tempting reply: the government has no right to trample on academic freedom. Ah, but they do! Ask Bob Jones University. The government has a legal right to trample on the academic freedom of institutions that discriminate on the basis of race. That’s what the Civil Rights Act says.

“Well, they have the legal right, to be sure. But not the moral right.” And here I’m inclined to agree with the objection. But note that this amounts to an objection to the Civil Rights Act. It gives the government a legal right to do that which they have no moral right to do. Indeed, this is the old libertarian objection to the Civil Rights Act. Barry Goldwater is vindicated at last.

Subscribe now

Or is he? I’ve no desire to drive home the point, to twist the knife. I’m a mild, boring sort of libertarian, after all. The world is complicated, and we should never commit ourselves fully to any particular perspective. The Trump administration’s treatment of Harvard is bad. But Jim Crow was worse. If the worst that can be said about the Civil Rights Act is that it gave Donald Fucking Trump the power to run roughshod over liberal institutions, and the best that can be said is that it ended Jim Crow, then the Civil Rights Act comes out way ahead. Jim Crow was that bad.

But the libertarians had a point. I think it’s easier to see that now.

I’ve been paying attention to statements by Trump and his inner circle, and I think there’s a pretty clear picture of what Trump is going for, that everyone is circling around but no one is quite hitting. The closest theory is “Trump just like tariffs,” but it’s a little more complicated than that. What follows is just my own theory, of course. But I’m pretty confident in this one.

Subscribe now

First and foremost, Trump is a mercantilist. That is, he sees trade in zero-sum terms. Trade is not good. Both sides gain and lose in a trade, and those gains and losses offset in the typical case; but when someone gains from trade, the other side is getting ripped off. Mercantilist theories were popular in, like, the 1500s, which not coincidentally corresponded to the rise of European colonialism. Trade isn’t good, because you have to give something up when you trade. Better, then, to just sail gunboats around the world and take what you want. Real wealth comes from extractive colonialism. Trump agrees with this; recall that his biggest complaint about the Iraq war is that we didn’t take the oil. If you’re going to invade a foreign country, you might as well steal all their shit for your trouble.

Trump also thinks of trade surpluses and deficits as a measure of how well you’re faring in the fundamentally zero-sum game of trade. If you have balanced trade, gains and losses are offsetting. If you have a trade deficit, you’re getting ripped off. If you have a trade surplus, you’re ripping other people off. He also believes (and this is his only relevant belief that is only approximately correct) that you can lower trade deficits and get to a trade surplus by raising tariffs.1

Finally, Trump likes ripping people off. This is his entire business career. He’s also got a fundamentally psychopathic and frankly mafioso approach to business dealings. He’ll use whatever leverage he has in virtue of his already great wealth to squeeze as much out of his business partners as possible, even to the point of ruining them in the process. This brings him great pleasure in life.

So then, to put this all together: Trump likes tariffs. Not so much as an end in themselves (although he does like the tariff aesthetic: “the most beautiful word” etc.), but because he sees them as a tool of mercantilist extraction. And he likes mercantilist extraction. Trade is zero-sum, he likes ripping people off, and tariffs are a tool whereby he can rip other people off. Beautiful.

This is why he put tariffs on every country in the world, even countries like Australia and Singapore that we run a trade surplus with and which don’t tariff US goods. He’s not trying to negotiate for free trade agreements. He’s trying to extract wealth from anyone who wants to export goods to the United States. In Trump’s mind, he’s got all the leverage. The US economy is the biggest and best in the world. Everyone wants to sell their goods here. And they can. But for a price. Gotta let the capo get his beak wet, you understand.

And for countries that the US has been running a trade deficit with, the tariffs are higher. This is because those other countries have been ripping us off (or so Trump believes), and that can’t stand. Time for us to rip them off, as is the proper way of the world. Who the fuck does Vietnam think they are, that the US has to run a massive trade deficit with them? They don’t extract wealth from us; we extract wealth from them. Tariff time.

Now when he announces his tariffs, all of the countries of the world come begging to make a deal. He loves this. He’s got all the leverage. So you want to make a deal? Let’s make a deal. Now we’re going to rip you off. That part is obvious, settled. But if you want to negotiate about how you’re getting ripped off, sure, we can have that conversation. You want us to lower our tariffs on you? Fine. Let us discuss other ways that you can pay tribute to the US (or to Trump himself).

This is why Trump is happy to go back and forth between imposing tariffs and suggesting he’ll lift them to make a deal. His ultimate goal is, again, to use the size of the US economy as leverage to extract as much as wealth as possible from other countries. Tariffs are the preferred, default method of doing so. But that’s negotiable, if you’d prefer we extracted wealth from you in some other way.

I write this in the interest of explaining, not justifying. This reasoning is built on laughably false empirical beliefs about how international trade works, together with deplorable attitudes towards how the US should deal with its trading partners. But Trump is a man with many false empirical beliefs and deplorable attitudes. I’m not inside his head, of course. But this is a relatively parsimonious theory that fits the data. I suspect it will hold up over the next few weeks. I guess we’ll find out.

Subscribe now

Some Twitter sleuths figured out that it looks like the tariff rates are, for every country, set by the simple formula of min(10%,(Trade deficit/total imports)/2). An indignant White House Deputy Press Secretary, one Kush Desai, responded that, in fact, tariffs were set by the following formula (which was then divided by two out of a sense of mercy):

But do the basic work of unpacking what those variables say, then look at the supplementary material below in that post, and you’ll see that ɛ̝ is 4 and φ is .25, so the change in the tariff rate (∆τ) is just (x-m)/m, which is the trade deficit (exports - imports) divided by total imports. Divide by two to get the final tariff rate. So the Twitter sleuths were right.

What is going on here? I’ve seen a few attempts to explain the math, but none really hits the nail on the head. I’ve read the supplementary materials a couple times (it’s short), and it’s not too hard to understand (modulo one point I’ll come back to shortly). So let me explain it to you. PLEASE NOTE: This is an explanation, not a justification. All of this is dumb.

Subscribe now

So Trump’s basic starting point is that he thinks trade deficits are bad. They’re not, but he thinks they are. He has this very dumb idea that a trade deficit is a deficit of trade, and the dollar amount of the trade deficit is the amount that other countries are cheating us out of through nefarious… I dunno, trading?

So his solution to this injustice is to try to inflict an equal and opposite amount of damage to the other country’s economy. If they’re taking X dollars from us in trade deficit, we’ll take X dollars from them in tariffs. And that’s precisely what the formula is supposed to do. The other country is stealing X dollars from us (per the trade deficit), so we’ll take X dollars right back (with a tariff).

So, a toy example: Suppose that there is another country that we import $100 of goods from and export $21 of goods to. That’s a trade deficit of $79. So the goal is to inflict $79 of damage to them by using tariffs. If we’re trying to inflict $79 worth of damage by putting tariffs on $100 of imports, that would mean a 79% tariff rate on those $100 of imports. We get that 79% number by dividing the trade deficit ($79 in our example) by the total number of imports ($100 in our example). And hey! That’s the formula that Desai shared. So that’s pretty much the whole story.

What about that ɛ̝ and φ though? Well, in one sense, they don’t matter. The White House team assigned values to those variables that directly offset so they don’t have an impact on the final calculation. But what are they doing in the equation to begin with? They’re in there because every $1 of tariffs doesn’t inflict $1 of economic harm. First there’s a question of passthrough rate (basically, how much of the tariff gets through to impact the economy). And second there’s a question of elasticity (basically, is there a multiplier of some sort which magnifies or shrinks the total effect on the economy). The White House says that only 25% of the tariffs will get through, but the effect will have a 4x multiplier effect, and so those cancel out to a 1:1 net effect for every dollar of tariff raised.

Now I’m pretty sure that they’re making a mistake with how those numbers are applied. The passthrough effect and elasticity are, I’m pretty sure, measures of the effect of tariffs on the domestic market, not on the foreign market. So the logic here seems to be something like “If Uruguay (or whoever) is stealing $1B of money from us through a $1B trade deficit, we want to put tariffs on imports from Uruguay that will hurt US consumers by precisely $1B.” So as far as I can tell, the approach here is to shoot the US economy in the foot by the EXACT AMOUNT of the trade deficit. (That’s not how they think of things, of course, but I’m pretty sure that’s what their equation will yield.)

But even if I’m getting that wrong and the White House has indeed correctly come up with a formula that will inflict pain on other countries in the exact amount of the trade deficit, this is still a very stupid thing to do.

For one thing, again: trade deficits are not bad, we are not getting ripped off in any sense, and retaliating for the exact amount of the trade deficit is approximately the stupidest thing you can do.

For another thing, this is being pitched as a way to balance out trade, to take back the exact amount that was stolen from us (again, it wasn’t stolen). But the elasticity multiplier is about net impact on the economy, not a multiplier to gains from trade. So if a $1B tariff will only pass through 25% ($250M), and that has a multiplier of 4 on top, we’d be getting $250M from the tariff but inflicting $1B in damage to the other economy. Good for vengeance, bad for balancing trade. (I’m not sure about this point, though; the reasoning around elasticities and pass-through is very under-explained.)

But finally, and most importantly, the tariff calculation is made “Assuming that offsetting exchange rate and general equilibrium effects are small enough to be ignored.” And that is absurd. These tariff rates are the highest since the 1910s, higher even than the Smoot-Hawley tariffs that caused the great depression. The exchange rate and general equilibrium effects will not be small enough to be ignored. They will be absolutely humongous. The exchange rate effects are complicated and might, on net, be slightly positive. But the general equilbrium effects are where the horror lies. “General equlibrium” is econ-speak for talking about the total state of all the important variables in the economy (employment, output, interest rates, inflation, etc etc etc). So the White House’s calculations assume that the total effect on the overall economy from these massive tariffs will be “small enough to be ignored.” Reader, they will not be.

And exchange rate and general equlibrium effects aside, this leaves out the political dimension of the analysis. When countries get tariffs levied on them, they generally respond by retaliating and raising tariffs of their own. This is precisely what we should expect, and the effects of that on both the US economy and global trade will be profoundly terrible.

Subscribe now

So there you have it! That’s the logic behind the size of Trump’s tariffs. It is not completely logically incoherent. But it is still very dumb, and bad things will result.

Happy Liberation Day!

One of the better responses I got to that was from Brad Skow, in the comments, who wondered how the reforms I had in mind could be accomplished in a way consistent with academic freedom. I replied that this would all be consistent with academic freedom if it was a faculty-led initiative. And, since I’m a staunch defender of academic freedom, I said that this must be a faculty-led initiative. Another good response was from my dad, who doesn’t like to comment on my posts because he thinks I’ll be embarrassed, but who texted me a day later to share his thoughts. He said that it didn’t matter how unpalatable the professoriate found any of the reforms I was proposing. University presidents and trustees exist to keep the money spigot flowing, and they’d fold like a cheap suit under pressure from the Trump administration. I didn’t have a snappy comeback to that one. “We’ll see,” I replied.

Unfortunately, it looks like my dad, despite not having a PhD in philosophy, is smarter than me.

Subscribe now

A month ago, the Trump administration had threatened to withhold about $400 million in federal funding from Columbia University, in response to perceived inadequacies in its response to student protests that made national headlines late last spring. This Friday, Katrina Armstrong, the interim president of Columbia, released a letter capitulating entirely to the Trump administration’s demands. These included reforming disciplinary procedures at Columbia for student protest and, most alarmingly, putting an international “studies” departments into receivership.

Despite agreeing with some of these reforms on the merits, I’m not happy.

For one thing, while I agree with some of the aspects of these reforms, I definitely don’t agree with the whole package. I think that a lot of shoddy scholarship goes on in “studies” departments, but this isn’t a blanket re-evaluation of that kind of scholarship. It is, instead, a selective targeting of the Department of Middle East, South Asian, and African Studies, which studies areas of the world with significant Muslim populations. Scholars in these areas are either Muslim themselves or sympathetic to perspectives of Muslims in the global south, and they are accused of fostering anti-semitism in their teaching and research regarding Israel and Palestine. There is doubtlessly some truth to the allegations of anti-semitism. But there are legitimate moral, political, and historical reasons to criticize the state of Israel, and these reforms seem targeted at preventing any such criticism. Even when those criticisms do shade into anti-semitism, hate speech is free speech. This censorship is inconsistent with freedom of speech and academic freedom. Similarly, while I have argued before that student protesters who break the law shouldn’t be treated with kid gloves just because they’re students, the particular nature of the demanded crackdown suggests not so much a concern with upholding campus security in general, but a desire to punish anyone on the “wrong” side of the heated debates over the Gaza war.

But more importantly, I see no good end to any of this. For we’re not at the end, of course. Columbia faculty were shocked by capitulation. They’re understandably furious. I’m sure some sort of response targeting the administration will go forward. We’ll see how that shakes out. But no matter who wins, we’re in a bad spot.

If the faculty win a showdown with the administration, we’re back in the scenario I predicted in my previous piece. The university faculty will stand on principle, refusing to bow to the pressure of the Trump administration. This is noble. But the beatings will continue. Columbia will lose $400 million in funding in the near future, and likely more over time as Trump and his cronies find every possible way to cut off every dime of federal funding that goes to Columbia. The result will be massive damage to an institution that, despite its obvious dysfunctions, is a major net positive for the country as a whole and for the overall human state of discovering and preserving knowledge.

If the administration wins, then academic freedom is dead. The Trump administration has learned that they can force complete capitulation with any slate of reforms so long as they threaten to cut off all funding. Columbia’s capitulation this last week sets a horrible precedent. When the Trump administration says “Jump,” while holding in hand an order to sever hundreds of millions of dollars of federal funding, the administration will say “How high?” Knowing that, you can expect them to say “Jump” a lot. Indeed, the damage may already be done on this front. Even if the faculty at Columbia fights back, this tactic will be used again and again against every other university in the country.

Expect a version of this dilemma to play out in every university across the country over the coming months, with dire results either way.

I had suggested in my last piece that university professors had to begin reforming the university, now, in order to get out from under the gun. Sadly, with the news out of Columbia, I’m afraid that advice came far too late. Even if every moderate center-left professor in the country steps up tomorrow to say “It’s time to re-center truth and academic excellence and throw out this half-baked activism,” it will be seen by everyone, from their farther-left colleagues they’d be throwing under the bus to the education commissars in the Republican party as a kind of “anticipatory compliance,” a pre-emptive bending of the knee to political pressure. And they wouldn’t really be wrong, would they?

Moderating reform is good on the merits. But if moderating reform is being advanced under the threat of destruction from hostile political actors, it can’t be done strictly on the merits. Imagine a world where the ideal set of moderating reforms goes through in the next few months. It wouldn’t end there. As much as I, and other likeminded “heterodox” academics would celebrate, the Trump administration would just grin. “A nice start, but we have some more reforms in mind. You don’t want to lose funding, do you?”

Subscribe now

For the benefit of my conservative readers, I’m trying to imagine how they’d respond to my musings. And with a sinking feeling, I hear their response in my head: “What else is new?” The Department of Education has been interfering with university politics for decades. For example, recall that in 2011, the Obama administration issued a “Dear Colleague” letter which called on all schools to use a “single investigator” model for adjudicating campus sexual assault claims. Failure to comply could result in loss of federal funds of various kinds. Schools complied. This was a contentious political issue - conservatives and civil libertarians complained that the “single investigator” model and other reforms mentioned in the “Dear Colleague” letter trampled on the rights of accused students. The Obama administration got their way. So we’ve seen this before.

I don’t agree with that line of response. I can’t think of a time that the federal government actually pressured universities to shutter academic departments, much less a time when a university actually took a significant step in that direction in response to such pressure. This is a difference in degree substantial enough to count as a difference in kind. But still, it is a continuation of a trend.

I’m an optimist by nature, but I find it hard to summon an optimistic spirit here. Universities were never immune to political influence, not now and not ever. But there was a period of relative freedom from politics that I feel is coming to an end. Partisanship and animosity degrade everything they touch, and the degradation seems to be accelerating as partisanship has worsened. Will universities sever themselves from all funding, shrinking dramatically in the process until they are finishing schools for the super-elite, as they were a century ago? Or will federal funding remain a chain around the necks of universities, that will be yanked one way or the other with increasing viciousness every time some new faction comes to power in Washington?

I’d like to believe there’s a third path, where universities take on a principled commitment to being apolitical, suffer some short-term damage as a result of the backlash, and then emerge stronger in the end. A week ago, that’s what I was arguing for. I still hold out some hope that we can reach that end.

But with Columbia’s capitulation, I smell blood in the water, and I’m not the only one. Sharks are circling. I cannot see the land.