UNPAYWALLED: being philosophically demure, being conceptually mindful

an introduction to thick concepts

Sorry stalkers, still in exams. That means no fresh bangers for a week. :/

I had one today on the Gospels, arguing, among other things, that Matthew isn’t that antisemitic, and that Jesus wasn’t lying when he said he still wants you to keep the Mosaic Law. Tomorrow is Philosophy of Religion. Wish me luck!

Also, I turned twenty-one on the 3rd, and you probably didn’t even buy me anything, not even a paid subscription, or an expensive flat in Chelsea. Since I am now older than you, I will be the bigger man and let it go. Still, if you want…

In the meantime, here is a freebee, released from its paywall and sent to your inbox. Enjoy!

---

I’ve been reading about thick concepts for a book I’m writing and I thought I’d spill the tea. This should help me absorb the material better — which is inevitably somewhat dry — and help you to be cool and fun, which is the main ‘how-to’ question you lot DM me about.

It begins with Bernard Williams, the insensitive monster who made Derek Parfit cry by not possessing the concept of a normative reason. In Ethics and the Limits of Philosophy, Williams coined the term “thick concept”. Since I can’t find any quotable definitions in Williams’ book, you’ll have to settle for my explanation.

DEMURE, MINDFUL, ELOQUENT, RUGGED, FAIR, FINE. These are all things you’ve called me. But they’re also examples of thick concepts. LEWD, LAZY, BLASPHEMOUS, SLOPPY, BORING, SOPHOMORIC. These are all things you’ve called my Substack secondary, Bentham's Bulldog. They’re thick concepts too.

What in tarnation is a thick concept? Roughly, a concept is thick when it has both a descriptive and an evaluative aspect to it. For example, GENEROSITY picks out a pattern of behaviour. In that sense, it’s descriptive. Yet GENEROSITY — unlike, say, MARSHMALLOW — is also morally loaded: calling someone ‘generous’ is, in most contexts, a positive appraisal.

Why the flip would anyone want to understand thick concepts? Good question, but enough with the profanity. Here are a few ways in which thick concepts are philosophically interesting:

1. Debunking Non-Cognitivism: according to moral non-cognitivism, moral utterances like ‘Torture is wrong!’ or ‘Not subscribing to Going Awol is even worse!’ don’t even try to express something true. Rather, moral utterances are, generally speaking, the expression of an attitude or a command or something otherwise can’t be true or false. For example, non-cognitivists might translate the previous moral utterances as ‘Boo, torture! Thumbs down!’ or ‘Subscribe to Going Awol! Now!’. (Note: this is not meant to be an a priori claim about morality, deducible from in one’s head from the armchair. It’s an empirical claim about how moral language actually operates. In principle, a study could disprove it tomorrow.) According to some philosophers — Philippa Foot, for example — thick concepts show that non-cognitivism is a very failed view. To see Foot’s point, take a sentence like: “Glenn is rude”. Given his pattern of past behaviour — calling me a “very failed Substacker”, despite the fact that many people are calling him “Glenn whatshisface”, “a tad Jelly”, and “likely a Trump toadie or a Russian bot” — I think it’s fair to call him not only rude, but a very failed Substacker indeed. Note, though, that RUDE is a thick concept: it’s a term of opprobrium, yes, but it also has descriptive content packed into it (disregard for etiquette norms, for example, or a tendency to make disruptive fart noises.) When I say “Glenn is rude” in this context, I’m surely making a moral appraisal; yet — as Foot points out — my calling Glenn rude can’t merely be the expression of an attitude like “Glenn? ICK!”. After all, I’m making a set of descriptive, empirical claims about Glenn by calling him ‘rude’: e.g., that he chews with his mouth open, that he wears his hat indoors, that no one likes his company, and so forth. Since thick concepts make up an enormous share of our moral discourse (consider: people often choose to deploy thick terms like ‘nasty’, ‘boorish’, and ‘lewd’ in place of thinner, purely evaluative terms like ‘bad’ or ‘immoral’), non-cognitivism fails as an empirical claim about how moral language really operates, since it fails to account for thick terms in our day-to-day moral discourse.

2. The Normativity of Logic: there’s a debate about whether logic is normative. According to some logicians, logic just describes the entailments of, and inconsistencies between, propositions. According to other logicians, logic is also a normative discipline: it tells us what we ought to conclude, and which propositions we ought not accept in pairs. In “Validity As A Thick Concept”, Sophia Arbeiter argues for the normativity of logic on the grounds that LOGICAL VALIDITY is a thick concept: it describes relationships of entailment between propositions, yes, but ‘validity’ is also a virtue of an argument — if someone declares your argument valid, they’re appraising it positively. Since logic trades on validity, Arbeiter argues, logic is a normative discipline.

3. Are Catholics Cannibals?: Today I spooled a Twitter thread on why thinking about thick concepts can help us see why Catholics aren’t cannibals, even if they do eat Jesus’ flesh. (Oy! Follow my Twitter!)

4. Diagnosing The Gettier Problem: I’ve substacked about the Gettier problem before, but here’s a cheeky recap. In 1973, Edmund Gettier cooked up two devastating counterexamples to the ‘justified true belief’ definition of knowledge. Since then, epistemologists have been scrambling to find an adequate definition of knowledge, but every definition they come up with — it seems — faces a new and equally devastating counterexample. Increasingly, epistemologists are starting to think that the task of defining knowledge is not only hard, but impossible, and are spending their time trying to explain why no airtight definition of knowledge is forthcoming. In “Knowledge as a thick concept: explaining why the Gettier problem arises”, Brent Kyle argues that KNOWLEDGE is a thick concept, and that its thickness explains why defining it will forever be impossible. The argument is basically this: first, KNOWLEDGE is thicker than a snicker. Recall that for a concept to be thick, it has to contain both descriptive and evaluative content. KNOWLEDGE, it seems, contains both. On the descriptive side, KNOWLEDGE entails true belief — and ‘truth’ and ‘belief’ are both descriptive properties. On the evaluative side, KNOWLEDGE entails epistemic justification — or so Kyle thinks. While there’s significant controversy about what epistemic justification amounts to, saying ‘S is justified in believing p’ seems roughly equivalent to saying ‘S has good reason to believe p’, ‘S’s belief in p is permissible’, or ‘S’s grounds for believing p are adequate’. Since GOOD REASON, PERMISSIBLE, and ADEQUATE are evaluative, EPISTEMIC JUSTIFICATION must be too. And since KNOWLEDGE entails both TRUE BELIEF and EPISTEMIC JUSTIFICATION, KNOWLEDGE meets the definition of a thick concept. Second, Kyle argues, the fact that KNOWLEDGE is thick with twelve Cs explains why epistemologists have failed to analyse it. Why? Well, for reasons that are too fiddly to sum up here (sorry), Kyle thinks thick concepts, in particular, might be unanalysable in virtue of their thickness. This view charts a middle course between diagnoses of the Gettier problem that say the reason we haven’t analysed knowledge yet either has something to do with KNOWLEDGE, in particular, or something to do with human concepts generally. If Kyle is right, the Gettier problem arises for thick concepts generally — not just KNOWLEDGE — but not necessarily for all concepts.

Those were just a few examples. When you go digging, thick concepts turn up all over the place. As philosophers in training, we’d do well to understand the debates around thick concepts. Unfortunately (for you. I’m fine with it), I won’t unpack every disagreement philosophers have about thick concepts: I haven’t the intellectual facility, and we’d probably be here till Christmas.

Instead, I’ll just outline the two sides of the most important debate about thick concepts — separationism vs. non-separationism — and try to keep it peppy. If you want to read the only ‘splainer on the separationism/non-separationism debate available on the World Wide Web that won’t make you want to gouge those pretty little eyes out (*bites lip*), you’ll have to pay me money.¹ The more I’m able to wring from Substack’s coffers, the more time I’ll have to write things like this, for you! Besides, chipping in is very demure, very mindful, whereas not chipping in is Scrooge-like and miserly.

Ciao, beautiful paymasters!

Here is the run-down you were promised.

Basically, the separationism/non-separationism debate boils down to whether it’s possible — in the abstract, on reflection — to separate the evaluative aspect of a thick concept from its evaluative aspect.

Who the flip cares? Good question, but please stop cussing, thank you. The reason the separationism/non-separationism debate matters is that it bears on the fact/value distinction. (The debate bears on other things too, but the fact/value distinction is the Big Mamma-Jamma.)

According to philosophical orthodoxy, there’s a distinction between descriptive facts on the one hand, and values on the other. In a recent publication — open access on the online journal Truth Social — political thinker D. J. Trump gives the following, helpful explanation:

Biden came to me, tears in his eyes, and said: “Donald, Donald, I don’t have facts, I don’t have values! Where the hell am I? I don’t know what I’m saying!” I said: “It’s true.” Call ‘em facts, we call them values. Facts and values, I was the first to call them that. Now everybody’s saying it, even Democrats, it’s true. Facts and values, that’s what I like to call them.

In case that wasn’t clarifying, it seems like there’s an intuitive distinction between value statements — e.g., “honesty is good” — and statements of descriptive fact — e.g., “the Sun is bigger than the Earth”. It’s controversial how, exactly, we should cash out this supposed distinction, but the claim that there is some distinction looks tempting.

Some philosophers, drunk on nuance, want to complicate the fact/value distinction. In their view, there’s no sharp line to be drawn. And many arguments for blurring the fact/value distinction rely, crucially, on non-separationism. Refute non-separationism, and these arguments are dead in the water; fail to refute non-separationism, and these arguments may just succeed.

Let’s roll through a few examples.

One way of cashing out the fact/value distinction is to say that facts and values are mutually exclusive categories. If something is a value, it can’t be fact (and vice versa). This version of the fact/value distinction is very strong, since it implies that there can be no evaluative facts (i.e., that there can be no such fact as ‘torture is objectively wrong’, or whatever.)

Seemingly, thick concepts make trouble for this distinction. If thick concepts are both evaluative and factual at once, then, you might think, facts can be evaluative after all.

But as Pekka Väyrynen points out in The Lewd, The Rude, and The Nasty: A Study of Thick Concepts in Ethics, “[t]hick concepts would challenge this distinction if, and only if, [non-separationism] were true. For only in that case would thick terms and concepts seem to be irreducibly factual and evaluative at the same time.”²

Another way of cashing out the fact/value distinction says that even if there are evaluative facts — facts like “Bush’s HIV program is good” and “recent cuts to it are unspeakably evil” — they’re somehow qualitatively different from descriptive facts like “Bush’s HIV program has saved over 25 million lives”.

Writes Väyrynen: “[t]hick terms and concepts might be used to undermine this sort of fact-value distinction if [non-separationism] were true, since in that case thick terms and concepts would combine evaluation and non-evaluative description in some qualitatively continuous fashion…”³

I won’t bore you by droning on about this, but the takeaway is just that thick concepts lie at the heart of many challenges to the fact/value distinction, and figuring out whether non-separationism is true matters for figuring out whether these challenges have merit.

I. Separationism

Let’s have a look at separationism, the view that, in the abstract, the evaluative and descriptive aspects of thick concepts can come apart.

It is important to get clear on what separationists are not saying. For be warned: if you ever misrepresent Separationism, its partisans will march over to your house and — with a lethal, sparkling axe — chop you some extra firewood so that you can be warm this Christmas. (Separationists are nice people, so be sure to get their view right!)

With that in mind, here are three things separationists aren’t saying⁴:

1. “Phenomenologically, thick concepts like DULLARD feel like they come in two separate parts.” On the contrary, you dullard, separationists only say that the evaluative aspects of DULLARD can come apart conceptually, on reflection, not that DULLARD feels — prior to reflection — like it’s composed of two loose parts.

2. “Since the evaluative and descriptive aspects of DOLT can come apart, it’s easy to introspect and tell exactly where the descriptive aspect ends and the evaluative aspect begins.” On the contrary, you dolt, separationists don’t say this. Again, all they say is that the descriptive and evaluative aspect of a concept like DOLT can be pulled apart in the abstract — they needn’t say all of us can identify the bright red line between the descriptive and the evaluative, or even that anyone can.

3. “The relationship between the descriptive and evaluative content is a thick concept like DIM-WITTED is exactly the same as the relationship between the descriptive and evaluative content in a thick concept like DUNDERHEAD — indeed, on separationism, the relationship between descriptive and evaluative content is the same in all thick concepts!” On the contrary, you dim-witted dunderhead, separationists often say that the relationship between descriptive and evaluative content varies from concept to concept. (E.g., it’s plausible that the evaluative content comes through stronger in MURDER than it does in LIE. Both are negatively morally laden, but the immorality implied by MURDER seems more tightly fixed to the concept than is the case with LIE. It’s easier to dream up an intuitively permissible lie than it is to dream up and intuitively permissible murder. Separationists and non-separationists will haggle over who can best explain the differences between thick concepts, but separationists are happy to acknowledge the differences.)

Now, there are different forms of separationism on the market, each with arguments to recommend them. Since Kier Starmer just blew up my DMs begging me to take the reins for a few hours (“Amos. Please. Britain needs leadership that listens, that learns, that delivers. Britain needs a Substacker. Britain needs you.”), I won’t get into the minutia of different separationist views.

That said, we can pool together some broad attractions of separationism, understood as the generic thesis that the evaluative and descriptive aspects of thick concepts can come apart conceptually:

1. Separationism helps preserve a robust fact/value distinction, which you might just take to be intuitive, simple, and clear. As Simon Kirchin puts the thought, “[N]ot only is the separation of evaluation and description supposedly reflected in much of modern thinking, it is simple and clear. These things are values, those things are (nonevaluative) facts, and never the twain shall meet.”⁵

2. Separationism (supposedly) has a nice story to tell about ‘evaluative flexibility’ — the fact that the evaluation conferred by thick concepts is flexible, depending on the context in which they’re deployed. For instance: the concept ELEGANT usually communicates something positive; but if you call the wrong sort of poem ‘elegant’ — a poem that was supposed to be rough and jagged — the positive evaluation conferred by ELEGANT seems to melt away. The separationist, seemingly, can explain this phenomenon with too much trouble: the reason ELEGANT is evaluatively flexible is that the evaluative component of ELEGANT can be disentangled from its descriptive component. (To be clear, non-separationists have things to say about evaluative flexibility too. You might just find what they say less plausible. For instance, the non-separationist can be like “there are really two concepts at play here — ELEGANT-pro and ELEGANT-con — that are picked out by the same English word in different contexts” or “ELEGANT is essentially evaluative, but whether it’s evaluative in a good way or in a bad way depends on the context in which it’s deployed”, or something else besides.)

3. Non-separationism is objectionable for X, Y, and Z reasons, so separationism is like totally the way to go.

There are other motivations, of course, but they strike me as pretty flimsy, so I won’t dump them onto the internet. (Probably, the reason I think the other motivations are flimsy is that I haven’t understood them properly, which is even more reason not to go off about them on Substack.)

Non-Separationism

Non-separationists are nice people too. Here is what they believe.

According to non-separationists, paradigmatic thick concepts like LEWD, RUDE, and NASTY are not composed out of evaluative and descriptive concepts. LEWD, RUDE, and NASTY are unitary, evaluative concepts, not composed of more fundamental conceptual elements. As Brent Kyle explains:

“[Non-separationists] hold that the meanings of thick terms are both descriptive and evaluative, although these features are not due to constituent contents within the meanings of thick terms. In slogan form, thick concepts are irreducibly thick. For example, the thick term ‘brutal’ expresses a sui generis evaluative concept, which is not a combination of bad or wrong along with some descriptive content.”

Here are two things that non-separationists aren’t saying:

1. They’re not saying you can’t think about the fact that a thick concept is evaluative, and then separately think about the fact that it’s descriptive. That would be cray-cray. All they’re saying is that, ultimately, thick concepts aren’t built out of more fundamental, evaluative and non-evaluative concepts.

2. They’re not saying, necessarily, that thick concepts like ELEGANT aren’t evaluatively flexible — i.e., they’re not all saying a thick concept can’t evaluate differently in different contexts.

And now, the thing you’ve allllllllllllllllllllllllllllllllllllllllll been waiting for! *drum roll intensifies* The one, the only, MEASLY AND INCOMPLETE LIST OF TWO REASONS WHY SOME PHILOSOPHERS FIND NON-SEPARATIONISM ABOUT THICK CONCEPTS ATTRACTIVE! *crowd goes wild*:

1. You might think that the evaluative aspects of thick concepts like GOOD and RIGHT are too thin, too wafery, to do justice to the evaluative character of a thick concept like, say, JUSTICE. Writes Kirchin: “I regard evaluation as something that is both conceptually basic and also a complex aspect of our lives that, in turn, is given life only by the concepts which are considered by mature users as evaluative. Although appealing, thinking of evaluation as something exhausted by PRO and CON and hived off from the concepts, even if such notions live in a few of them inherently, seems to me to be too narrow a view of evaluation and our evaluative lives.”⁶ (This was a fairly vibesy way to dress up the argument. For a rigorous exposition, see chapter 6 of Kirchin’s Thick Evaluation.)

2. Separationism is objectionable for X, Y, and Z reasons, so non-separationism is the way to go. After all, separationism is a reductive view. That is, it tries to reduce or explain thick concepts in terms of more fundamental concepts. It’s always a kosher move — in the context of any reductionist scientific or philosophical programme — to look at the proposed reductions, declare them implausible, and lose faith in the reductive programme.

Outro

For rigorous, vibes-based reasons, I lean toward non-separationism. I’m still sorting through the arguments, though, and probably will be for centuries, perhaps millenia.

Anyway, this marks the end of today’s blog post. Go back to fighting crime, saving shrimp, or whatever it is my paid subscribers do. And if you liked what you read, don’t forget to like and share! That would be mindful and cutesie. xx

1. I exaggerate: the SEP and IEP entries on thick concepts are both pretty good. As usual, the SEP does better on depth and the IEP entry does better on ease of reading. Of course, you will only see this footnote if you made it past the paywall so… lol.

2. Väyrynen, Pekka. 2013. The Lewd, The Rude, and The Nasty: A Study of Thick Concepts in Ethics. Oxford: Oxford University Press: p. 15.

3. Ibid., p. 16.

4. Cf. Kirchin, Simon. 2017. Thick Evaluation. Oxford: Oxford University Press: pp. 21-22.

5. Ibid., p. 40.

6. Ibid., p. 164.