Does Williamson Trump Fine?: A Note on Metaphysics, Logic and Epistemology in Philosophy of Education (14)

Williamson threatens to turn many Finean distinctions into products of a superficial relevance heuristic.  Fine thinks that this threat has been overstated, that Williamson’s own semantic assumptions are far from decisive, and that even if some examples are shakier than once thought, the larger essentialist project remains methodologically and philosophically productive.

The first thing Fine does is clarify what sort of thesis Essence and Modality was meant to defend. Williamson had suggested that Fine’s paper can be read in a modest way, as defending only the intelligibility of a hyperintensional essentialist position, not its truth. Fine says that this is only partly right. There are really three levels in play. There is a very weak thesis, that such claims are intelligible. There is an intermediate thesis, that they are not only intelligible but, given the present state of metaphysical inquiry, reasonably assertible, that there is a live and substantive issue here. Then there is the stronger thesis, that the claims themselves are true. Fine admits that he sometimes slides between these levels, but he insists that his intended fallback was not the very weak claim of mere intelligibility. It was the intermediate claim, that even if one does not yet accept the stronger thesis outright, one should still treat these essentialist claims as real candidates for truth, as matters of substance rather than semantic confusion.

Williamson’s argument, if successful, would certainly threaten the stronger thesis in some cases. But Fine can still say that the intermediate thesis survives, that the cases opened up a genuine metaphysical space and showed that essence could not simply be treated as a transparent restatement of necessity. In other words, even if some intuitions are unsafe, the conceptual and methodological role of essence as a distinct metaphysical notion may still be vindicated. This is already important for our educational discussion, because it means that even if some of our sharper Finean distinctions about educational concepts turn out to be less secure than we thought, that does not force us back into crude extensionalism or policy pragmatism. The intermediate thesis still matters. It says that there are genuine questions here about the nature of educational things, not just about how we happen to talk.

Fine then turns to Williamson’s number cases. Williamson had argued that our intuition that “It is essential to 8 that it is the successor of 7” might differ in truth-value from “It is essential to 8 that it is the predecessor of 9” is demonstrably mistaken, because the functional terms involved are directly referential and thus interchangeable salva veritate. Fine’s first move is surprisingly cautious. He says that even he is not fully sure the “successor of 7” formulation should be treated as essential in the exact way Williamson assumes. Perhaps what is essential to 8 is not its identity with s(7) but its standing in the immediate successor relation to 7. That is already significant. It shows Fine trying to relocate the essentialist insight away from the specific linguistic packaging Williamson targets. The metaphysical point may concern structural relation rather than identity with a functional term. But Fine does not stop there. He argues that Williamson’s semantics for functional expressions is much less obviously compulsory than Williamson makes it sound. Williamson assumes that expressions like p(9) and s(7) are directly referential in the relevant sense, that they contribute only their referent to the content of the larger sentence. Fine questions that. He points out that even if one can provide such a semantics for extensional languages, it does not follow that it is the right semantics in non-extensional environments such as essentialist contexts. He also argues that the fact that functional expressions are not definite descriptions does not show that they are “scopeless” in the relevant way. There may still be distinctions in content corresponding to different ways of understanding the role of the functional expression within the larger essentialist claim.

Ok, this part of the reply is difficult to get, but I think the core point is not so hard to state. Williamson had tried to debunk Finean intuition by saying, in effect, “You are misled because two expressions that feel different really contribute only the same referent.” Fine’s answer is, “That conclusion depends on a controversial semantic picture. You have not shown that this is the right way to understand such expressions in these contexts.” So Williamson has not yet proved what he claims to have proved. At the very least, the case is not closed. Fine reinforces this by introducing the idea of “objectual content.” Roughly, a sentence may differ not just in truth-value or ordinary semantic value, but in which objects it is directly about. On one natural view, “8 = s(7)” is objectually about 8 and 7, whereas “8 = p(9)” is objectually about 8 and 9. If essential truth is sensitive to what a proposition is directly about, then the two sentences may differ in the relevant way after all. Again, Fine is not claiming to have settled the semantics decisively. He is saying that Williamson’s route from direct reference to debunking is not a shoe in. There are respectable semantic options under which the Finean discrimination survives.

He then adds a second line of resistance. Williamson’s argument relied on very special functional expressions that happen to be rigid, always denoting the same object. But many natural functional expressions are not like that. “The number of planets,” “the velocity of body x at time t,” “the acceleration of y,” these are naturally read as world-sensitive or circumstance-sensitive. Once one allows such expressions, it becomes harder to maintain the neat divide Williamson needs between essentialist and modal scepticism. Fine even suggests that Williamson’s own reasoning might then generate analogous doubts in straightforward modal cases, which would be a cost Williamson presumably wants to avoid. The broader point is that Williamson has chosen an unusually favourable class of examples, and their general significance is not obvious.

Fine then turns to the issue of whether Williamson’s diagnosed rot spreads. Suppose, for the sake of argument, that in the functional-expression cases we really are being misled by some mistake. Why think the same mistake explains the original Socrates/set examples. Williamson had proposed a crude relevance filter operating at the linguistic surface. Fine offers a different diagnosis. Perhaps we are misled in the number case because we assimilate functional expressions like p(9) to paraphrases such as “the predecessor of 9,” thereby reading them as if they were descriptions rather than directly referential terms. If so, the source of error is quite local and does not automatically transfer to the kinds of case in Essence and Modality. Alternatively, even if one follows Williamson in talking of irrelevance, Fine insists that there is an important difference between a sound ontological principle and a dubious linguistic surrogate for it.

This is one of the most philosophically important parts of the reply. Fine introduces what he calls the Ontological Link: if it is essential to y that A, and the proposition expressed by A is directly about x, then y ontologically depends on x. This, he says, is a good principle. If something is really essential to y and directly about x, then x had better not be utterly irrelevant to what y is. But from this one might slide to a Linguistic Link: if A merely contains a term referring to x, then the proposition is directly about x. That stronger linguistic principle is what the functional-expression cases might undermine. So even if Williamson is right that functional expressions can smuggle in apparent irrelevance at the linguistic level, that does not damage the deeper ontological principle Fine actually cares about. In the original E&M cases, Fine says, the relevant judgements are driven by the ontological principle, not by the crude linguistic one.

This is a very Finean move. He is relocating the discussion from language back to metaphysics. Williamson wants to explain our judgement by a superficial relevance filter sensitive to linguistic constituents. Fine answers that what matters in essence is not merely what terms occur but what the proposition is about in the ontological sense, what the object depends on. If that distinction holds, then Williamson’s debunking mechanism may be too shallow to reach the cases that matter most. Fine strengthens the point by offering a kind of controlled reformulation. Instead of saying “It is essential to Socrates that he belongs to singleton Socrates,” he says, in effect, take x to be Socrates and y to be singleton Socrates, then it is essential to y to have x as a member, but not essential to x to belong to y. This reformulation is meant to neutralise the purely linguistic surface effects Williamson appealed to. Fine’s claim is that our judgement remains firm even once the allegedly misleading expressions are removed. So if there is an error, it is not obviously the kind of error Williamson has diagnosed.

Finally, Fine addresses Williamson’s methodological complaint that metaphysicians too quickly abandoned intensionalism on the basis of a few dubious examples. Fine thinks this is unfair. He says that the uptake of Essence and Modality was not actually that fast, and, more importantly, its influence was not due merely to a handful of striking intuitions. It also came from the development of a clear notation, a logic of essence, a semantics, and the productive use of essentialist notions in formulating doctrines that otherwise remained obscure or badly framed. The value of essence, he suggests, lies not only in the examples but in what the concept lets us do. Williamson himself, Fine notes, describes natural numbers in an iterative way. But if one takes that seriously, one is already halfway to an essentialist picture in which the successor structure belongs to the nature of the numbers. Essentialist language is not an ornamental addition. It can articulate structure in a way modality alone may not.

That last point matters a great deal for our discussion. Fine is effectively saying that usefulness is not a mere afterthought. A concept may earn its keep not only by fitting selected examples but by organising inquiry, revealing structure, and enabling doctrines otherwise unavailable. This does not refute Williamson’s warning about heuristics. But it does mean that the verdict on essence cannot be decided by a few contested examples alone. The cost-benefit picture is broader. And that has direct implications for philosophy of education.

Where, then, does this leave the discussion so far? It leaves it in a more complex and, in some ways, more interesting place than either a straightforward Finean confidence or a straightforward Williamsonian scepticism would allow. First, Williamson’s pressure still stands. We cannot simply trust our felt educational judgements about what is central, relevant, constitutive, or essential. The possibility of heuristic distortion is real. This remains a powerful challenge to inflated educational buzzwords, to lazy invocations of “Western philosophy” or “non-Western philosophy,” to under-analysed terms like inclusion, hope, voice, or agency. In all those areas, Williamson helps us see how easy it is for educational discourse to overfit morally loaded data and mistake rhetorical salience for ontology. But, second, Fine’s reply shows that Williamson’s scepticism cannot simply be generalised into a blanket suspicion of all finer-grained structure. The fact that some examples may be unsafe does not mean that essence talk is merely heuristic projection. It may still be that some distinctions of dependence, structure, and direct aboutness are real and philosophically indispensable. In educational terms, that means we should not abandon the attempt to distinguish educational goods from generated display-forms, learner capacities from assessment formats, belonging from mere placement, access from mere co-presence, or discussion as an educational form from mere speaking in class. What Fine’s reply encourages is not abandonment but a more careful grounding of such distinctions.

Third, the educational use of Finean metaphysics now looks less like a hunt for immediate essences and more like a layered practice of discrimination. We must ask, in each case, whether we are dealing with an ordinary modal truth, a generated educational form, a grounding relation, an ontological dependence structure, or merely a relevance-driven way of talking. The machinery remains fruitful, but it becomes more methodologically self-aware.

Fourth, Fine’s distinction between the weak, intermediate, and stronger theses is very helpful for philosophy of education. Much of the time we may not be in a position to defend the strongest thesis, that a given educational concept really has the essence we take it to have. But the intermediate thesis may still be valuable and justified. We may reasonably treat certain educational distinctions as live and substantive, as revealing genuine issues not capturable by cruder vocabularies, even if our final metaphysical verdict remains unsettled. This is especially important in a field where concepts are often either reified too quickly or dismissed too quickly. The intermediate thesis gives us a disciplined middle ground.

So my combined lesson now becomes something like this. Fine gives philosophy of education a way to articulate structural questions about what educational things are, what they depend on, how they are generated, and what follows from their nature. Williamson warns that our confidence in such articulations may be contaminated by heuristic overreach. Fine’s reply shows that this warning, though serious, is not decisive against the larger project. It tells us to distinguish more carefully between superficial linguistic relevance and deeper ontological dependence, between a bad route to an intuition and the intuition’s possible truth, between a few shaky examples and the broader usefulness of a conceptual framework.

That leaves us with a more demanding practice of educational metaphysics. We can still ask whether concepts like inclusion, agency, belonging, decolonisation, demasculinisation, anti racism, class, or hope name genuine structures, generated forms, or only heuristic bundles. But we now have to ask with more than confidence. We need argument, discrimination, mixed methods, and a readiness both to use Finean distinctions and to test them under Williamsonian pressure. I guess that is a harder place to stand, but it seems the right kind of seriousness given how important I think education is.