The Modal Metaphysics of Lost Highway

The following is an attempt to track some of the salient elements of David Lynch’s films largely through the different but related lenses developed by contemporary philosophers Timothy Williamson and Kit Fine. I do so because I think they both offer models to understanding the films more fruitfully than competing interpretive models. Or maybe it runs the other way: I’m using David Lynch to model the modes of modality of Williamson and Fine. I guess I’m doing both. Kind of. And I end up with different conclusions depending on how far into the thing you get. One minute a film is definitely Williamsonian, the next absolutely Finean. The plot of this is as confusing as Lynch at his best, but unlike Lynch the confusion here is caused as by my own confusion rather than brilliant design. So bite me! 

David Lynch’s films are often described as confusing, dreamlike, or surreal, but those descriptions miss something decisive. The confusion they generate is not simply interpretive. It is structural. Viewers do not merely wonder what a scene means. They wonder what kind of world would allow such a scene to occur at all. This is already a metaphysical question, whether it is recognised as such or not. 

In most narrative cinema, the metaphysical background is stable. Characters persist through time. Events occur in a determinate order. Possibilities narrow as stories progress. Even when narratives deceive or withhold information, they do so against a fixed background of what must remain the case. Lynch disrupts that background. He constructs films in which identity can fracture without explanation, where events repeat without becoming copies, and where worlds seem to fold back on themselves rather than advance.The claim of this book is that these effects cannot be adequately understood using the usual interpretive tools. Psychological explanations treat the films as depictions of interior states. Symbolic readings turn every anomaly into metaphor. Both approaches assume that the world of the film is ontologically ordinary and that all strangeness occurs within it. Lynch’s cinema resists that assumption. The strangeness belongs to the world itself.

To understand that world, we need a theory of modality.

Timothy Williamson’s paper “Accepting a Logic, Accepting a Theory”, when reading through a collection of essays in Yale Weiss and Romina Birman’s Saul Kripke on Modal Logic early last year. It addresses issues in the philosophy of logic, particularly how one should understand the acceptance of a logic standardly conceived in analytic philosophy. Williamson takes up a challenge motivated by Kripke’s critiques of alternative logics and the broader question of whether adopting a non-classical logic is really a “change of subject” or something more like adopting a new scientific theory. Williamson’s core question is deceptively simple: what does it amount to accept a logic? And his answer is deliberately deflationary. Accepting a logic is not a sui generis philosophical act, it is the same kind of thing as accepting any other theory.

This is what later came to be labelled anti-exceptionalism about logic. He starts by targeting a deeply entrenched assumption in analytic philosophy, namely that logic is special in a way that makes it immune to ordinary epistemology. On the “exceptionalist” picture, logic is constitutive of thought itself. To reject classical logic would therefore not be to make a mistake but to change the subject. You see this idea in various forms in Quine, Carnap, and in some readings of Wittgenstein. Williamson thinks this picture is unstable. If logic is constitutive of thought, then disagreement about logic should be impossible in a very strong sense. But disagreement clearly happens, and not just verbally. Logicians argue, persuade, revise, and sometimes abandon systems. That social fact is his starting datum. His first major move is the folk theory vs scientific theory analogy. Williamson asks us to compare logic with physics. Ordinary people have inferential dispositions about motion, causation, and space. Those dispositions form something like a “folk physics”. Scientific physics does not merely describe those dispositions, it often corrects them. Yet we do not think adopting relativity means changing the meaning of “motion” or “time” in a way that makes disagreement with Newton impossible. The analogy is meant to dissolve the idea that logical revision must be conceptually incoherent. If ordinary inferential practice can be improved upon, why not logical practice as well? Crucially, Williamson does not say that logic is empirical in the same way as physics. The analogy is epistemological, not metaphysical. It concerns how we justify acceptance, not what the subject matter is. 

This sets up his second key distinction: logic vs metalogic. Many arguments for the special status of logic rely on results about logical systems, for example soundness, completeness, consistency proofs. But those are metalogical results, typically carried out using background mathematics and often classical reasoning. Williamson insists that using classical metalogic to study non-classical logics is not circular in a vicious sense. It is no worse than using current physics to evaluate rival physical theories. Acceptance happens at the object level, not at the level of the tools used to evaluate candidates. This blocks a common objection that you cannot rationally argue for a logic without presupposing it. Williamson’s reply is that presupposition in practice does not entail incorrigibility in principle. The most subtle part of the paper concerns inferential dispositions. 

Williamson denies that accepting a logic consists in explicitly believing axioms or rules. Instead, acceptance consists in having defeasible dispositions to infer in certain ways. By calling dispositions defeasible philosophers just mean that they are not exceptionless. They can be overridden, trained, refined, and sometimes corrected. This is crucial for explaining how rational disagreement about logic is possible. If logic were a set of explicit meanings fixed by stipulation, disagreement would collapse into verbal dispute. But if logic is embodied in fallible cognitive practices, then disagreement can be substantive. This is where Williamson shifts the debate from semantics to epistemology. The question is no longer “what do logical constants mean?” but “what inferential practices should we endorse, and why?” 

At this point he turns explicitly to the Quinean ‘change of subject’ objection. The objection says: when a classical logician and an intuitionist disagree about excluded middle, they are not really disagreeing, because “not” and “or” mean different things for them. Williamson’s reply borrows heavily from Kripkean externalism. He invokes the general idea, familiar from Saul Kripke and Hilary Putnam, that meaning is not fixed by what is “in the head” of individual speakers. Meaning is anchored in communal practice, deference, and causal history. Just as chemists can disagree about whether water is H₂O without changing the subject, logicians can disagree about valid inference without talking past one another. Disagreement does not require perfect semantic alignment, only enough shared practice to sustain correction and critique. Thus Williamson thinks Quine’s argument overreaches. It assumes a descriptivist picture of meaning that Kripke himself undermined. 

The role of Kripke is important here. Kripke famously argued that some logical laws, especially those tied to necessity, have a kind of objectivity that resists conventionalism. Williamson is sympathetic to this realism but rejects the conclusion that logic is unrevisable. Objectivity does not entail immunity from error. In effect, Williamson is offering a realism without exceptionalism. Logical facts are real, but our access to them is mediated by fallible theory, just like everywhere else. The payoff comes at the end. If accepting a logic is accepting a theory, then debates between classical, intuitionistic, relevant, or paraconsistent logics should be assessed using familiar theoretical virtues: explanatory power, simplicity, coherence with mathematics and science, fruitfulness. There is no sharp line where philosophy stops and logic begins. Logical theory choice is continuous with rational inquiry more generally. This is a radical conclusion. It undercuts both logical dogmatism and the idea that logic is merely conventional. Logic becomes a domain of responsible, corrigible judgement. 

In The Philosophy of Philosophy and later in Modal Logic as Metaphysics, Williamson continues to insist that logic is not methodologically special. There is no autonomous “first philosophy” that settles logical truth prior to theory. Logic is continuous with mathematics, metaphysics, and science, and its acceptance is governed by the same epistemic norms. As I’ve just sketched, in the early paper this continuity is defended mainly at the level of epistemology. Accepting a logic is like accepting a theory because both involve defeasible inferential dispositions, sensitivity to counterexample, and communal standards of correctness. In The Philosophy of Philosophy, this gets generalised. Williamson argues that all philosophical methodology works this way, including thought experiments, modal intuitions, and conceptual analysis. What was a claim about logic becomes a claim about philosophy as such. The same is true of the anti-Quinean move. In the paper, Williamson uses Kripkean externalism to block the “change of subject” objection. In later work, especially in Modal Logic as Metaphysics, this becomes a full-blooded realism about modality. 

Logical disagreement is possible because there are objective modal facts, even if our access to them is theory-mediated and fallible. The early defence of disagreement sets up the later defence of realism. Where things really harden is Williamson’s attitude toward necessity and classical logic. In the early paper, Williamson is careful not to prejudge which logic is correct. His anti-exceptionalism is officially neutral. Classical logic, intuitionistic logic, and others are all in principle revisable. What matters is how well they function as theories. By the time we get to Modal Logic as Metaphysics, that neutrality is gone. Williamson now explicitly endorses classical logic and classical modal logic as the correct framework for metaphysics. 

Necessity is no longer treated as something we infer via fallible dispositions. It becomes metaphysically robust, mind-independent, and governed by S5-style principles. Ok. Its hard to avoid technical terms of art like S5 and the like in this area of philosophy. It can be off putting to non specialists but I think once you keep track of what these technical terms are doing its actually interesting. The trouble comes trying to remember what they mean each time they crop up because often the specialists forget to ensure their readers are keeping up with them and they rush on and you get lost. So throughout I’ll keep reminding readers what these terms of art mean each time so you don’t have to scramble back looking for the explanation. 

So there are these S-systems, and thinking about them helps to see why S5 is powerful but also why some philosophers resist it. All of these systems are ways of regimenting how necessity and possibility behave. They differ in how “rigid” they make modal space. (Modal is just about necessity and possibility). The weakest commonly discussed system is K. K just gives you the bare minimum needed to reason with necessity and possibility at all. It says, roughly, that if a necessary rule holds, then it holds in all cases where it is applied. But K places almost no structural constraints on how possibilities relate to one another. It is very permissive and very thin. Philosophically, it corresponds to a view where modal structure is largely unspecified. 

A step up is T. T adds the idea that whatever is necessary is true. This is often taken to be non negotiable once necessity is understood as “could not have been otherwise”. T already rules out some exotic views, such as necessities that somehow fail to obtain. Then comes S4. S4 adds the principle that if something is necessary, then it is necessarily necessary. This gives necessity a kind of stability. Once a fact counts as necessary, it stays necessary across all possible perspectives. Many philosophers find S4 attractive because it captures the idea that necessity does not flicker or weaken, but it still allows possibility to be more flexible. 

Finally there is S5, which adds the strongest principle: if something is possible, then it is necessarily possible. This collapses many distinctions between different “levels” or “grades” of possibility. All possible worlds become mutually accessible. Modal space becomes fully symmetric and flat. You can also think of these systems in terms of how they treat access. Imagine each possible world can “see” some others. K allows almost any pattern of seeing. T says every world sees itself. S4 says if a world sees another, it also sees everything that other world sees. S5 says every world sees every other world. 

Why does this matter philosophically? Because choosing one of these systems is not just a technical decision. It reflects how you think reality itself is structured. Williamson is comfortable with something very close to S5 because he thinks necessity is grounded in what things are, their identity and existence, not in perspective, language, psychology, or history. Once the world fixes what is possible, that fixity applies everywhere. (Someone like Kit Fine is much more cautious. Even though he is not hostile to modal logic, he resists the idea that one single modal operator captures all the ways in which things can be necessary. He thinks there are different kinds of necessity, grounded in essence, law, norm, practice, affect, and structure. For that reason, flattening all modality into something S5-like risks losing important distinctions. We’ll be coming to Fine later. This is just so you can see where these two diverge a bit. ) 

Back to Williamson for now. The early paper clears the methodological ground by saying logic could be revised. The later work says that, given our best theories, it should not be. Anti-exceptionalism survives, but it no longer motivates pluralism. Instead it legitimises realism and theoretical conservatism. You can see this clearly in how Williamson treats counterexamples (e.g. if the match had been wet then it wouldn’t have ignited). In the early paper, deviant intuitions count as data that might push us away from classical logic. In the later work, many such intuitions are reclassified as cognitive error, performance failure, or confusion about the metaphysics of necessity. The same epistemology is still in play, but the verdicts are firmer. What quietly drops out over time is the emphasis on inferential practice. 

In the paper, Williamson leans heavily on inferential dispositions. Acceptance of logic is grounded in how agents reason, how they are trained, and how they revise their habits. This gives the paper a faintly sociological or pragmatic flavour, even though Williamson is not a pragmatist. In Modal Logic as Metaphysics, this dimension recedes. The focus shifts from practice to structure. Logical systems are evaluated less by how they regulate reasoning and more by how well they capture objective modal reality. The metaphysics does the heavy lifting. Inferential dispositions become evidence rather than constitutive elements. This shift matters philosophically. The early Williamson leaves space for a kind of fallibilist modesty about logic, even if he is not a pluralist. The later Williamson is far more confident that we already have the right tools. Critics often say that anti-exceptionalism ends up functioning as a ladder that is kicked away once classical modal logic is safely installed. 

Seen this way, the paper is best understood as transitional. It dismantles the idea that logic is beyond theory choice, but it does not yet assert a strong metaphysical picture of necessity. Modal Logic as Metaphysics supplies that picture and then uses it to justify a fairly orthodox logical framework. One way of putting it is this: early Williamson is trying to make logical disagreement intelligible; later Williamson is trying to explain why most of it is mistaken. Williamson’s opening move in Modal Logic as Metaphysics is to take something that looks, at first glance, like an abstract piece of logical machinery and put it under metaphysical pressure. 

He begins from an ordinary thought: things could have been otherwise. A coin you are holding could have landed tails rather than heads. A tree that exists might never have grown. And it can seem just as obvious that what things there are could have been otherwise too: perhaps there might have been fewer things, or more things, or even nothing at all. Williamson starts by letting that intuitive picture speak in its most natural voice: the universe could have evolved differently so that there was never any money, and in that case there would not merely have been no coins, there would have been no such particular coin at all, because it simply would not have been. Conversely, there could have been something that is actually nothing, like a coin design never minted. So, he says, there is something that could have been nothing, and something that could have been something. 

That is the intuitive entry point for a dispute about whether “what there is” is contingent. But he immediately brings into view a rival tradition. Some philosophers, he notes, have denied that it is contingent what there is, while allowing that it is contingent how what there is is arranged or behaves. The pieces are fixed, only the pattern changes. He uses Wittgenstein’s Tractatus as a salient representative of the idea that there is an “unalterable form” of objects, with variability located in their configuration rather than in the inventory itself. Williamson is careful not to get bogged down in exegesis, and he signals that the real aim is not about Wittgenstein in particular, but about a view: that it is necessary what there is. 

At this point he introduces two names that become organising labels for the whole book. Call the proposition that it is necessary what there is “necessitism”, and its negation “contingentism”. So either “Ontology is necessary” or “Ontology is contingent”. ('Ontology' just means what exists. If you say that Yuxin exists, you’re doing ontology.) But Williamson is also explicit that these are not yet detailed theories. They are high altitude generalisations that can be combined with many auxiliary assumptions, and neither side by itself settles particular case questions like whether there are animals, or which animals there are, or whether numbers exist. 

The disagreement is about the general claim that necessarily everything is necessarily something. Note that the necessity at issue is metaphysical, not epistemic. Williamson insists that the question is not what can be known, thought, or said, but what could not have been otherwise, in the relevant metaphysical sense. This is important because many people hear “necessary” and slide into “knowable a priori” or “analytic”. Williamson is trying to keep those debates at arm’s length. He is also explicit that the dispute is not immediately about “laws of physics” or about science fiction scenarios. You can be a necessitist while denying that the laws of nature are metaphysically necessary. You can allow that the laws might have been different, while still maintaining that the inventory of “somethings” is fixed. 

Conversely, you can be a contingentist while granting that some items, like numbers on a Platonist view, are non contingent. The core issue is about the general, unrestricted claim, not about whether there are special necessary beings. Williamson refuses to let the debate stay at the level of vibe. If necessitism sounds crazy, that is not yet an argument. If contingentism sounds like common sense, that is not yet decisive either. 

The first chapter is an exercise in making the debate precise enough that later logical machinery can actually bite. One way he does this is by pairing the modal dispute with a temporal analogue. If necessitism says “it is necessary what there is”, a temporal cousin would say “it is always the case what there is”. That temporal thesis is what he calls permanentism, and its negation temporaryism. The point is use a structural analogy: both debates involve quantifying “always” or “necessarily” over “everything”, and both invite similar confusions about what is being held fixed. Williamson suggests that a lot of resistance to necessitism comes from conflating it with claims that would be better classified as permanentist, or with muddled combinations of modal and temporal readings. The pairing also helps him later, because temporal logic and modal logic share formal patterns.   

From there he asks : why take necessitism seriously at all? Surely it is obvious that many things could have failed to exist. Williamson’s distinguishes the attitude “it seems obvious, therefore it must be false” from the attitude “it seems obvious, therefore any argument for the opposite must be fallacious”. He complains that the latter attitude can deadlock inquiry. If we treat every contrarian hypothesis as equally serious, we get swamped by blogger level proliferation. If we refuse to treat any contrarian hypothesis as serious because it offends the obvious, we risk missing cases where the obvious turns out to be wrong. So we need some principled way to decide which hypotheses are worth real theoretical attention. 

That is already a preview of his broader posture: treat logic and metaphysics as parts of theoretical inquiry, to be assessed by something like the standards we use in other theoretical domains. But before he gets there explicitly, he runs a series of “forms of necessitism” thought experiments that show how quickly naive objections misfire. Consider a coin again, now understood as a macroscopic object made of microscopic parts. Williamson sketches two very different atomist manoeuvres a necessitist might try. 

An eliminativist says, strictly speaking there is no coin, only atoms. Coin talk is a convenient shorthand for complicated talk about atoms. If so, then the coin is not a counterexample to necessitism because there is no such object as the coin. A reductionist says, by contrast, that there really is a coin, namely the mereological sum of its atoms. (Mereology is just about the relation of wholes to parts and vice versa). But then, given how the atoms change over time, the reductionist starts to look committed to saying that the coin in your pocket is never strictly the same coin from moment to moment, because the exact sum differs as atoms are gained or lost. 

Williamson thinks neither defence is attractive as a way of making necessitism plausible, and the deeper point is that simply appealing to microphysics does not explain away the appearance that there could have been fewer things. Even if physics told us that the number of atoms is fixed, why could there not have been more atoms, or fewer atoms, under different laws? (Fine thinks anyone saying that coin talk is strictly speaking false because of physics is just wrong; there is no good reason to think people are making a mistake when they do coin talk and that linguistic data is important and decisive.) 

Then Williamson introduces what becomes one of his recurring ontological distinctions: concrete versus non concrete. He uses “concrete” informally, as roughly “located in space and time, causally engaged, physical-ish. The key idea is that a necessitist does not have to say the coin is necessarily a coin, or necessarily concrete. The necessitist can say: necessarily, the coin is something, but it might have been non concrete, in which case it would not have been a coin. That is, contingency migrates from “whether it exists” to “whether it is concrete”. Williamson’s early chapters are an extended attempt to make that migration intelligible rather than merely verbal. 

A natural reaction is to say: if it is non concrete, is it an abstract object, like a number? Williamson blocks that move. “Non concrete” is not automatically “abstract” in a positive sense, because “abstract” is not merely “not concrete”. “Abstract” has its own paradigms, like numbers and directions, tied to theoretical roles. If the coin had not been concrete, it would not thereby have played the role of a number. Williamson uses a temporal analogy to make the point vivid: when an iceberg melts, it does not become abstract, it simply ceases to be concrete. The contrast is meant to create conceptual space for entities that are neither concrete nor abstract in the familiar platonist way, which is exactly the sort of thing necessitism will later need. 

Here he starts to sound like he is inventing a new category of being. He does not deny that. He calls it a cost. He also notes a familiar pattern from mathematics: set theory often posits huge infinities of sets, not because common sense wants them, but because theoretical principles that do serious work seem to require them, and the alternative can be massive loss of simplicity and elegance. This analogy is meant to loosen the grip of “but that sounds ontologically extravagant” as an immediate refutation. The question is whether the theoretical pay off warrants the ontological cost. 

Williamson pivots to the ambiguity of phrases like “possible F”. The phrase “a possible president” can mean at least two things. On one reading, it is predicative: “Trump is a possible president” means Trump is a president and Trump could have existed. That is boring, and it makes “possible” function like “existing in some possible world” tacked onto existence. On another reading, it is attributive: “Trump is a possible president” means Trump could have been a president. That reading does not require Trump to be a president, or even to exist, in the actual circumstances. It is structurally like “alleged diamond”, which does not entail “diamond”. Williamson argues that the attributive reading is the one relevant to necessitist talk of merely possible objects. When you say “we are all possible murderers”, you ordinarily mean each of us could have been a murderer, not that each of us is a murderer who could have existed. 

Contingentists often hear “possible object” in a way that makes it sound like Meinong’s notorious “non existent objects” with impossible properties. Williamson wants to disarm that association. He says: the necessitist’s “merely possible F” is naturally understood on the attributive reading: something that could have been an F but is not an F. On this reading, “merely possible stick” is coherent: it could have been a stick, but is not a stick. The point is not that we can picture such a thing, but that the grammar of modal attribution already supports the idea that possibility can attach to predicates, not merely to existence. In Lynch’s Twin Peaks there’s the Black Lodge where the idea that someone could have been a tree but isn’t a tree is illustrated by depicting the person who could have been a tree as a tree even though he isn’t and couldn’t have been. Well now. 

Williamson then introduces what looks like a metaphysical challenge: “Where is the merely possible coin, actually?” The answer is: nowhere. In Lynch’s Twin Peaks the same answer is: In the Black Lodge. But Williamson argues that this is not decisive, because the analogous temporal question also has a “nowhere” answer. Where is the past coin now? Nowhere. That does not show that past objects are incoherent, or in the Black Lodge. It shows that location talk can mislead when applied outside the concrete present. This is part of a broader strategy in the early chapters: use temporal cases, which many people are already willing to treat seriously, to make modal cases less alien. 

A further challenge says: modal properties must be grounded in non modal properties. A lump of clay is malleable because of its microstructure, and the microstructure is non modal. So, if an object is merely possible and has no concrete microstructure, how can it have modal properties like “could have been a stick”? Williamson pushes back by questioning whether the modal versus non modal distinction is even clear in the way that objection presupposes. Microstructure has consequences for what the object can do, so why is it “non modal”? If “non modal” just means “lacking explicit modal vocabulary”, that is not a serious metaphysical distinction. The upshot is not that grounding is impossible, but that we cannot assume a simple “modal grounded in non modal” picture to rule out necessitism by fiat. 

One of the most important clarifications for the whole book arrives in section 1.4: unrestricted generality. Williamson says that in the core statement of necessitism the quantifiers “everything” and “something” are absolutely universal, with no tacit contextual restriction. People often soften the claim by restricting “everything” to “everything concrete”, or “everything that exists in space and time”, and then they say “of course everything is necessarily something” because they have built existence into the domain. Williamson’s point is that the interesting dispute only begins when both sides are quantifying without restriction. In ordinary conversation we often do restrict quantifiers. “Everything is wet” might mean “everything in the laundry basket is wet”. 

But when we are doing metaphysics of being, Williamson wants the wide open reading. This immediately resonates with Lynch’s universe which so often refuses contextual restrictors on quantifiers. This leads to an objection many logicians will recognise: unrestricted quantification is entangled with set theoretic paradoxes, because if you quantify over absolutely everything you are tempted to treat “everything” as forming a set, and then Russell style contradictions loom. (ie is the set of all sets in itself?) Williamson’s argues against this by saying that unrestricted quantification does not by itself generate contradictions. Contradictions arise only with auxiliary assumptions, like that the things quantified over form a set. We can reject those auxiliary assumptions without rejecting unrestricted quantification. He treats unrestricted quantifiers as intelligible. 

Necessitism starts looking like a consequence of taking certain logics of modality with unrestricted quantification and identity seriously. Contingentism looks like the attempt to keep the logic while blocking the metaphysical consequence, typically by complicating either the logic (special rules, restricted inference) or the semantics (variable domains, free logic). Williamson is preparing the ground so that when the technical results arrive, the reader cannot escape by saying “oh, you just meant ‘exists’ in the restricted sense anyway”. 

Some philosophers, especially David Lewis, treat modal operators like “possibly” and “necessarily” as quantifiers over possible worlds, understood as maximal concrete spatiotemporal systems, big, concrete realities like our own, just causally isolated from us. On this view, when you say “possibly P”, what you really mean is “there is some possible world in which P is true”. So take a simple sentence. “There could have been no donkeys.” On the Lewisian picture, this is true because there is a possible world which is just like ours in many respects but where evolution went differently and no donkeys ever existed. The sentence is true because one of the worlds has no donkeys. So far, this seems intuitive and harmless. Now bring in quantification, words like “everything”, “something”, and “nothing”. Lewis and many other metaphysicians often use unrestricted quantification. That means when they say “everything”, they really mean everything there is, not just everything in some domain like “everything in this room” or “everything in this world”. 

This is where the trouble begins. Consider the necessitist claim Williamson wants to make precise: “Necessarily, everything is necessarily something.” In plain terms, this says: anything that exists exists necessarily. Nothing winks in and out of existence across possible worlds. Now translate that into Lewis’s framework. “Necessarily P” means “in every possible world, P is true”. So the necessitist sentence becomes something like: “In every possible world, everything is necessarily something.” But now ask: what does “everything” range over? If quantification is unrestricted, then “everything” already ranges over all objects across all possible worlds. Not just the objects in one world, but all objects full stop. If that is how the quantifier works, then the sentence starts to collapse. Why? Because if “everything” already includes all objects, then saying “everything is something” is just a logical truism. Of course everything is something. Otherwise it would be nothing, and then it would not be included under “everything”. So when you combine these two moves treating necessity as quantification over worlds and treating object quantifiers as unrestricted you risk turning the deep metaphysical claim of necessitism into something like: “In every world, everything is something.” Which is as empty as saying: “Everything exists.” That is not a substantive thesis about modality or existence. It is just a logical truth. 

To see this more concretely, imagine a toy universe of objects. Suppose your ontology includes Alice, Bob, and a unicorn that exists only in a possible world W2. If your quantifier “everything” ranges over Alice, Bob, and the unicorn regardless of world, then saying: “Everything exists in every world” cannot mean what the necessitist wants. Because the unicorn does not exist in the actual world in the Lewisian sense. So either the statement is false in an uninteresting way, or you quietly shift the meaning of “exists” so that “exists” just means “is in the unrestricted domain”. Either way, the claim loses its intended force. Williamson doesn’t argue that Lewis’s modal realism is incoherent or false. He says that using modal realism as the framework for debating necessitism versus contingentism distorts the issue. If necessity just means “true at all worlds”, and existence is already handled by an unrestricted quantifier, then modal claims about existence stop doing real work. They become artifacts of how we set up the semantics, not discoveries about reality. 

That is why Williamson insists that the debate should not be conducted in a framework that reduces modality to quantification over concrete worlds. He wants a framework where necessity is a genuine metaphysical notion, not a by product of how we count worlds. If you treat modal operators as nothing more than world quantifiers, and you treat object quantifiers as unrestricted, then deep metaphysical disputes about existence risk being flattened into trivialities. To preserve the substance of the necessitism debate, you need a conception of modality that is not simply world counting. Another potential confusion he targets is the “actualism versus possibilism” terminology in the literature. In ordinary language, “actually” often sounds weighty. It seems to point to what really exists, as opposed to what merely could have existed. 

Because of this, philosophers introduced a contrast between actualism and possibilism, thinking they were marking a deep metaphysical divide. The usual gloss goes like this. The actualist says that everything that exists is actual, and there are no merely possible objects. The possibilist says that there are things that are possible but not actual, like possible people or possible worlds. Williamson says that once you look carefully at how the word “actually” behaves in modal logic, that way of framing the dispute collapses. 

In modal logic, “actually” is treated as an indexical operator. It works a bit like “here” or “now”. Saying “actually P” just means “P is true at the actual world”. Crucially, unless it appears inside another modal context, it does not change the truth value of what you say. So take the sentence: “Everything that exists actually exists.” Under the modal logic reading, this becomes trivial. It is like saying: “Everything that exists exists here.” That cannot possibly be false. Whatever exists, exists where we are evaluating existence from. Nothing metaphysical is being said. Here is a simple example. Imagine you say: “There are dinosaurs.” That is false. But if you say: “There are actually dinosaurs,” it is still false. Now imagine you say: “There are mammals.” That is true. And if you say “There are actually mammals,” it is still true. Outside of special modal contexts, “actually” does no work. It does not add metaphysical depth. 

Now consider a sentence where “actually” does matter. “It is possible that there are flying pigs.” That sentence is true in a loose sense. But now say: “It is possible that there are actually flying pigs.” This second sentence is false, because it says there is a possible scenario in which, at the actual world, there are flying pigs. That is not the case. Here “actually” matters because it sits inside a modal operator. Williamson’s point is that if the actualist versus possibilist debate is framed using “actually” in the trivial sense, then the dispute reduces to something empty. Of course everything that exists actually exists. That is not a metaphysical thesis, it is a logical truism. 

So if philosophers really want to argue about what exists, they must mean something else by “actual”. But once they introduce a heavier, metaphysically loaded notion of actuality, the terminology becomes unstable and unclear. Different authors mean different things, and arguments slide past one another. This is why Williamson thinks the language of actualism and possibilism has become hopelessly muddled. It mixes a harmless logical operator with a much stronger metaphysical idea, without clearly separating them. 

The real issue, he says, is not whether everything is actual, but whether everything that exists exists necessarily. That is the dispute between necessitism and contingentism. The necessitist says: everything exists in every possible world, though it may have different properties in different worlds. The contingentist says: some things exist in some worlds but not others. That disagreement is genuinely metaphysical, clearly stated, and not hostage to the ambiguities of the word “actually”. 

So the upshot is that Williamson is not denying that there are deep questions about existence and modality. He is saying that the traditional actualism versus possibilism labels obscure those questions rather than clarifying them. By switching to necessitism versus contingentism, the debate becomes precise enough to do real metaphysical work. Similarly, in the philosophy of time, “presentism versus eternalism” is often glossed as “everything is present” versus “not everything is present”. Williamson runs an analogous complaint: on the straightforward temporal logic reading of “presently”, “whatever is, presently is” is trivial. So again the interesting metaphysical dispute should be reformulated as permanentism versus temporaryism, which targets the claim that “always everything is always something” rather than a linguistically unstable “present” predicate. 

Williamson wants logic and metaphysics to be assessed together. He discusses the necessity of identity and distinctness. In standard modal logic with identity, if x is identical with y, then necessarily x is identical with y. Likewise, if x is distinct from y, then necessarily x is distinct from y, given some plausible background principles. He sketches arguments for these claims and says that, in natural models, they come out valid. You can always tinker with identity interpretation to make them fail, but you can do that for any argument, and it would be bad methodological taste to uglify a strong, elegant logic of identity just to avoid ontological inflation. 

The book’s central method is to start with a powerful logic, then treat metaphysical conclusions that fall out of it as serious candidates, unless there is overwhelming reason to distort the logic. This is a clue to Williamson’s ambition: modal metaphysics is  a structural picture of objects across counterfactual and temporal variation. He then begins the process of showing how the metaphysics is “intimately connected with some technical issues” in quantified modal logic, especially the Barcan formula and its converse. 

Quantified modal logic is just first order logic (with “for all” and “there exists”) plus modal operators (usually “necessarily” and “possibly”). It inherits the familiar truth functional connectives like “and”, “or”, “not”, “if…then”, and adds modal operators that shift evaluation across possibilities. Williamson defines this as a combination of the two logics, plus new problems about their interaction. The Barcan formula, in one common modern notation, says roughly: if it is possible that there exists an x such that A(x), then there exists an x such that it is possible that A(x). In symbols, ◇∃x A → ∃x ◇A. Intuitively, it moves an existence claim out of the scope of “possibly”. The converse Barcan formula goes the other way: if there exists an x such that possibly A(x), then possibly there exists an x such that A(x). 

The philosophical anxiety is that the Barcan formula seems to imply that whenever it is possible that something of a certain kind exists, there is already something, actually, that could have been that kind of thing. That sounds like it commits you to “mere possibilia” as actual entities. On Williamson’s picture, that is not a bug but a feature, because it is closely allied to necessitism. 

Williamson explains that the Barcan formula emerged in a technical, even austere, setting in the mid twentieth century. He traces it to Ruth Barcan Marcus’s 1946 work on a formal system of quantified modal logic built on C I Lewis and Langford’s strict implication framework. In that setting the work is “syntactic”: it lays down axioms and rules, proves theorems, and does not supply a semantics or even an informal English reading. The symbols are intended to function as quantifier and possibility operators, but the metaphysical significance is not argued for. This historical point matters to Williamson because he is trying to dissolve the idea that logic and metaphysics are disjoint, and that therefore theorems of a formal system cannot have metaphysical consequences. If logic is “good” and metaphysics “bad”, then “no logic is metaphysics”, and you can treat the Barcan formula (roughly: whenever it is possible that something of a certain kind exists, there is already something, actually, that could have been that kind of thing) as a harmless formal convenience. 

Williamson thinks that posture became harder to maintain, precisely because once we start using quantified modal logic to represent metaphysical claims about necessity, possibility, and existence, the boundary becomes porous. The early history of the Barcan formula is, for him, a case study in how technical results acquire metaphysical weight once we begin interpreting them in the relevant ways. At first the Barcan principles are adopted in a setting where semantics is suppressed, then later, as semantics and intended interpretations develop, the principles begin to look “metaphysically problematic”, and contingentists try to avoid them. Those attempts, he hints, often “come to grief”, pushing contingentists toward significant restrictions in their logic. He aims to show “how early contingentist attempts to avoid them came to grief” and why later attempts paid a high price. 

The Barcan formula (roughly: whenever it is possible that something of a certain kind exists, there is already something, actually, that could have been that kind of thing)  is not just a random axiom. In many semantic treatments, it corresponds to a structural assumption about domains of quantification across possible worlds. Roughly, if the domain is constant across worlds, then moving ∃x out of ◇ is safe, because the things you quantify over do not vary from world to world. If the domain varies, then the move can fail, because something might exist in one world but not be among the things quantified over in another. This is how he will connect logical principles to metaphysical theses. 

Suppose you think Wittgenstein could have had a child. Contingentism, in its intuitive form, says: there is a possible world where Wittgenstein has a child, and in that world there is an individual who is his child, but in the actual world there is no such individual at all, because Wittgenstein in fact had no children. The Barcan formula, read naively, seems to say: if it is possible that Wittgenstein has a child, then there exists, actually, something that could have been Wittgenstein’s child. That “something” is not any actual person, because there is no such person. So it looks like you have to accept some merely possible individual as an inhabitant of your actual ontology. 

Ted Sider, summarising Williamson’s position, puts it starkly: “if there could have been a child of Wittgenstein, then there in fact exists something that could have been a child of Wittgenstein”, and this thing is “nonconcrete”, a “bare possibilium”. Notice how that connects directly back to Williamson’s insistence that “non concrete” need not mean “abstract like a number”. The child of Wittgenstein that could have been is not, on Williamson’s view, a concrete human being with mass and location. It is something that could have been concrete, but is not. 

If you accept that kind of entity, the Barcan formula starts to look like a natural logical reflection of it, rather than a metaphysical horror. Of course in Lynch it does look like metaphysical horror too. 

At this point the dispute becomes methodological. The contingentist can respond by rejecting the Barcan formula (roughly: whenever it is possible that something of a certain kind exists, there is already something, actually, that could have been that kind of thing), or by changing the logic so that the inference patterns that lead to it are blocked. But Williamson’s project, as both his own framing and sympathetic summaries emphasise, is to argue that the “best” modal logic, judged by a mixture of simplicity, strength, and theoretical utility, includes principles that entail the Barcan schema and thereby support necessitism. 

Sider’s reconstruction of Williamson’s strategy is especially helpful for understanding that Williamson is not trying to derive necessitism from “pure reason” alone. He is arguing that when you adopt the classical rules of inference for modal logic, and you do not impose ad hoc restrictions motivated only by the desire to save contingentism, you get a logic that carries with it necessitist commitments. Contingentists then have to decide where to resist: perhaps by restricting existential generalisation, adopting a free logic, restricting necessitation, or adopting variable domain semantics with special truth conditions. Each resistance tends to come with a cost in complexity, loss of elegance, or loss of inferential power. 

Williamson’s wager is that those costs are not worth paying, so the theoretically preferable package is the strong logic plus necessitist metaphysics. Williamson’s “Possible Worlds Model Theory”, connects the metaphysical issues to possible worlds model theory, understood as a mathematical framework used in formal semantics, and he calls it the “main technical achievement of modal logic”. His deeper ambition is to explain how to read systems of modal logic as metaphysical theories, so that the model theory can be applied to metaphysics rather than treated as a merely linguistic tool. 

That is exactly the bridge Williamson needs if the Barcan formula is to be more than a theorem inside a calculus. If the model theory is read as representing the structure of metaphysical modality, then constraints like constant domains are not arbitrary modelling choices, they are metaphysical commitments. Conversely, if you refuse that reading, then the model theory cannot straightforwardly tell you what metaphysics to accept. 

A basic picture of possible worlds model theory, in the minimal Kripkean sense, goes like this. You have a set of “worlds” (often understood abstractly, not necessarily Lewisian concrete universes), an “accessibility relation” that says which worlds are relevantly possible relative to which others, and a valuation that assigns truth conditions to sentences at each world. For quantified modal logic, you also need domains of individuals at each world. If the domain is constant, you quantify over the same individuals at every world. If it is variable, the domain can expand or contract from world to world. 

In the simplest models of constant domain semantics with identity, the Barcan formula and its converse typically come out valid, and in that setting necessitism looks like the natural metaphysical interpretation: everything is something in every world, though it may be concrete in some worlds and non concrete in others. In variable domain models, you can invalidate the Barcan formula, and that seems to match contingentism’s intuition that some individuals “do not exist” in some worlds. 

So then we can ask: what justifies choosing the more complex variable domain machinery? Is it independently motivated, or is it a patch built solely to block a metaphysical consequence we find uncomfortable? Sider points out that the variable domain approach faces a formal challenge of assigning truth values to statements in worlds where objects denoted in the statement do not exist, and a philosophical challenge of explaining how such models can be “intended” if possibility and necessity are not reducible to facts about worlds and their inhabitants, which many contingentists also agree on. 

Williamson’s point is that, once you stop treating possible worlds as metaphysically basic concrete realms, you lose the easy story that makes variable domains look like “obvious metaphysics”, and the semantics starts to look like a technical workaround whose metaphysical interpretation is unstable. What is distinctive in Williamson’s approach is that he treats these choices as theory choices. He explicitly says, in the context of identity, that we should be guided by a conception of theories in logic and metaphysics as scientific theories, to be assessed by the same overall standards as theories elsewhere, including mathematics. That is a striking claim because it insists on theoretical virtues like simplicity, power, and integration. 

And it is meant to reverse a familiar order of deference. Many philosophers think we should first settle metaphysics by intuition or conceptual analysis and then choose a logic that fits. Williamson wants to treat the logic itself as part of our best theorising about structure, and then adjust metaphysics to fit the best logic, unless there is a compelling reason not to. Keeping that methodological stance in view Williamson can be read as building a sequence of pressures. The first pressure is conceptual: once you insist on unrestricted quantification, the simple statement “necessarily everything is necessarily something” cannot be defused by ordinary contextual restriction tricks. That forces contingentists to say something substantive about what the quantifiers range over and how “is” and “exists” function in their framework. Williamson even recommends avoiding “exists” as a key term because it carries misleading restricted readings, preferring to formulate the theses with “something” and “everything” instead. 

The second pressure is linguistic: the ambiguity of “possible F” and related constructions shows that our ordinary modal talk already supports attributive readings on which something can “be a possible F” without being an F, which opens the door to merely possible objects understood as could have been Fs. This does not prove their existence, but it undermines the complaint that such talk is grammatically incoherent or necessarily Meinongian. 

The third pressure is logical: classical modal logic with identity delivers strong principles like the necessity of identity and distinctness. Williamson argues it would be perverse to distort that logic just to avoid the ontological consequences that arise when you combine it with modal claims about what could have been. Better, he suggests, to accept the ontological inflation and then ask whether it is theoretically harmless, in the way that large infinities can be theoretically harmless or even theoretically required in mathematics. 

The fourth pressure is historical and interpretive: the Barcan formula (roughly: whenever it is possible that something of a certain kind exists, there is already something, actually, that could have been that kind of thing) and its converse emerged in a setting where semantics was suppressed, but as soon as we start using quantified modal logic to express metaphysical claims, we cannot pretend that theorems are metaphysically inert. The history is meant to show that the logic did not begin life as a metaphysical doctrine, but also that it became one once the boundary between logic and metaphysics started to dissolve. 

And the fifth pressure is methodological: if modal logic provides a structural core for metaphysics, then we should prefer the best modal logic by theoretical virtues, even if it forces us toward radical metaphysical conclusions. That’s Williamson’s ambition: metaphysical issues concerning necessity and possibility should be investigated by means of formal logic, and the best modal logic is a strong higher order S5 with classical rules, with necessitism as “part and parcel” of that logic, while alternatives “fare less well”. 

It is worth lingering on what “S5” and “higher order” mean here, without drowning in notation. S5 is a standard system of modal logic characterised, in semantic terms, by an accessibility relation that is an equivalence relation (roughly: every world is accessible from every world in the relevant sense). Philosophically, S5 often corresponds to a strong view about modality where if something is possibly necessary then it is necessary, and if something is possibly possible then it is possible. Williamson’s broader project will not be simply to assume S5, but to argue for strong modal principles as part of the best theory. “Higher order” means that the logic quantifies not only over individuals (first order) but also over properties or relations (second order), or even higher. This matters because once you allow quantification over properties, you get powerful comprehension principles, and those interact with modality in ways that generate further necessitist pressures. 

Williamson repeatedly diagnoses confusion as coming from bad formulations and unstable terminology. “Actualism versus possibilism” is muddled because “actual” is ambiguous between a rigid operator in modal logic and a metaphysical notion of being actual that would do harder work. “Presentism versus eternalism” is similarly muddled because “present” slides between a temporal indexical operator and a metaphysical thesis about what exists. Even “exists” is misleading because it invites restricted readings that trivialise the claims. The point is that if you are going to let logic guide metaphysics, you need your metaphysical theses to be expressible in the language the logic actually handles. Otherwise, you cannot tell whether the logic supports or undermines the thesis. 

So by now we have learned that necessitism, formulated with unrestricted quantifiers, is not the childish claim that every ordinary concrete thing must exist no matter what, but a more radical and more subtle claim: the domain of “somethings” is fixed across modal space, while concreteness and other substantial features can vary. The thing that makes that plausible is not an intuition about what exists, but a willingness to accept that the best overall theory might postulate many non concrete entities as the values of unrestricted quantifiers. 

We have learned that much resistance to necessitism comes from hearing it through the wrong lenses: through modal realism, which trivialises the intended claim by turning necessity into world quantification; through “actualism” talk, which often relies on an unanalysed “actual”; through “exists” talk, which slides between restricted and unrestricted readings; or through a Meinongian caricature of possible objects. 

We have learned that the Barcan formula is the technical hinge where metaphysics and logic grip each other. It looks, on its face, like a theorem about the scope interaction of “possibly” and “there exists”. But once you interpret it under unrestricted quantification, it begins to look like a metaphysical statement: possibilities about existence require actual items that could have realised those possibilities. Williamson’s historical story  is designed to show how that hinge was manufactured in technical work before it was understood as metaphysically loaded, and then how it became loaded as the field’s ambitions shifted. And we have learned Williamson’s guiding evaluative norm: do not sabotage a strong, simple, elegant logic to save a metaphysical prejudice. If the best modal logic, judged by theoretical virtues, implies necessitism, then necessitism deserves to be treated as a serious candidate, and the contingentist must show that the costs of necessitism are worse than the costs of complicating the logic to block it. 

Consider Williamson’s knife factory style scenario. Imagine a machine that assembles knives from blades and handles. In the actual run, blade B is paired with handle H to make knife K. But if a belt had been delayed by a second, B would have been paired with a different handle H*, producing a distinct knife that never in fact exists in space and time. Williamson suggests that we can nevertheless uniquely describe “the possible knife that would have been assembled from blade B and handle H* had the belt been delayed by one second”. 

A contingentist wants to say there simply is no such knife, since it was never assembled. A necessitist can say: there is such a thing, but it is non concrete. It is a possible knife on the attributive reading: something that could have been a knife but is not actually a knife. The Barcan style pressure then appears: if it was possible that such a knife existed, then there exists something that could have been that knife. Once you stop restricting quantifiers to the concrete, and once you allow “possible F” to function attributively, the necessitist package starts to look logically natural, even if ontologically extravagant. 

That is why Williamson calls his book Modal Logic as Metaphysics. He is constructing the idea that modal logic is building a framework for stating, comparing, and adjudicating metaphysical theories about the structure of being across possibility and time especially through the detailed development of the Barcan principles, the assessment of competing quantified modal logics, and the higher order arguments that make necessitism harder to avoid without paying heavy theoretical costs. 

Chapter 3: Modal Logic and Lost Highway

What I want to do now is apply all this to the film world of David Lynch. Modal Logic as Metaphysics insists that modal logic is not merely a convenient formal language for talking about metaphysical possibility and necessity, it is one of the main places where metaphysics itself gets done. He means that when we try to say, with any precision, what could have been the case, what must be the case, what could not be the case, and what exists across those modal claims, we very quickly commit ourselves to structural principles. Those principles are substantive metaphysical commitments. 

Lynch’s Lost Highway is unusually apt. Lynch gives you a world in which identity, time, and the boundaries of what is really happening are never simply presented, they are governed by an underlying structure that you infer by watching what can and cannot happen in the narrative. The film is a machine that constrains what interpretations remain live. 

That is close to Williamson’s sense that modal logic is not an optional overlay on metaphysics, it is part of the engine that generates what a metaphysical view can even be. To set up the stakes, it helps to say what “modal logic” is, for a non specialist, without flattening it. The basic modal operators are usually written as a box and a diamond. Read the box as “necessarily” and the diamond as “possibly”. 

So, if P is a statement, then “necessarily P” says that P could not have failed to be true, while “possibly P” says that P could have been true, even if it is not actually true. In ordinary English, you already use these constantly. “It could have rained today” or “You could have died yesterday” are possibility claims. “Two plus two must be four” is a necessity claim. Metaphysics becomes involved when the “must” and “could” are not merely about what someone knows, or about what the laws of physics allow, but about what reality itself permits, even in principle. That is what philosophers tend to mean by metaphysical modality. 

Now add quantifiers, words like “everything” and “something”. In formal logic, these are ∀ and ∃, and they let you say things like “everything is such that …” or “there is something such that …”. When you combine these with necessity and possibility you get quantified modal logic, a framework for saying things like “necessarily, everything is such that …” and “possibly, there exists something such that …”. 

Those combinations are exactly where Williamson thinks metaphysical commitments become unavoidable, because they force you to choose how existence and modality interact. Lost Highway gives you a vivid way to feel that interaction. The film’s central unease is that the same “space of characters” seems to contain Fred Madison and Pete Dayton, and yet the film refuses to let you treat them as two straightforwardly co existing individuals within one stable story world. Fred “becomes” Pete, then later Pete “becomes” Fred, and the woman played by Patricia Arquette appears as Renee and as Alice, with the film simultaneously inviting and resisting the thought that this is one person in two guises. The narrative behaves like a model in which some individuals are present in one “region” but not another, except the regions are not physical locations, they are modal or structural positions in the story. 

When the Mystery Man answers the phone and says he is at Fred’s house, the film is, in its own medium, staging something like a violation of ordinary constraints on where an individual can be at a time, and it does so by making the constraint itself the horror. 

Let me just remind you of what Williamson has been doing. Williamson has put pressure on a venerable suspicion, associated especially with Quine, that quantified modal logic is somehow illegitimate or “unscientific”, “born in sin”, because it involves talk of necessity and possibility applied to quantified claims about objects. One worry is about meaning, the idea that once you talk about what is necessarily true of an object, you need a robust account of what makes it the same object across different possible scenarios. 

Another worry is about existence, the thought that once you quantify inside modal operators you start committing yourself to odd things like “merely possible objects”. Williamson’s first moves are meant to show that these worries do not justify treating modal logic as metaphysically second class. He is clearing the ground for the more aggressive claim that the best metaphysics may actually be guided by the best modal logic. In other words, the logic expands what exists. 

A key background assumption he flags is that metaphysics wants unrestricted quantification. That means, when you say “everything”, you mean everything, not merely everything concrete, or everything in this room, or everything actual in some narrowly chosen sense. Metaphysics, at least in the ambitious tradition, tries to speak about being as such. 

Williamson wants the quantifiers in his modal logic to match that ambition. That stance matters, because many evasions of tricky modal commitments rely on quietly restricting what “there is” ranges over. In Lynch terms, it is like refusing to let the camera pan into certain corridors of the house because you know what is there will break the story. Williamson’s methodological temperament is to insist on going down the corridor. 

Once you accept quantified modal logic as a legitimate arena, a famous cluster of principles shows up almost immediately. These are the Barcan Formula and its converse, usually abbreviated BF and CBF. You do not need the symbolic form to grasp the issue, but the idea is simple to state. Suppose you say, “Necessarily, everyone is such that possibly they could have been a pianist.” That is a claim where a universal quantifier appears inside a necessity claim, and then a possibility claim appears inside that. 

Now compare it to, “Everyone is such that necessarily, possibly they could have been a pianist.” The Barcan type principles concern when you can move quantifiers across modal operators without changing truth. In more intuitive terms, they ask whether the domain of what your quantifiers range over stays fixed when you consider other possible ways reality might have been. 

If the domain stays fixed, then whenever you talk about “everything”, you are ranging over the same stock of objects no matter which possible situation you consider. If the domain shifts, then some objects exist in one possible situation and not in another. 

Lost Highway makes this feel visceral. Is “Pete” someone who exists, full stop, across the whole story world, even in the segments where he is not “there”, or is he the sort of entity that only exists within a certain narrative possibility, something like a locally generated person? The film’s unease partly depends on the fact that either answer seems to cause trouble. 

If Pete exists “all along”, then Fred’s transformation is not a creation ex nihilo but some kind of identity or role shift. If Pete exists only in the “middle section”, then the story has admitted a domain shift, a new individual coming into existence in a way the film never depicts as birth, and vanishing in a way it never depicts as death. 

Williamson’s central metaphysical thesis is necessitism. In one slogan form, necessitism says that necessarily everything is necessarily something. Another way to put it is that no matter how things could have been, the total domain of objects is the same. What varies across possibilities is not what exists, but which of those objects are concrete, or located in spacetime, or causally active, or in any other robust sense “present” in the world. 

Necessitism is a claim about quantificational domains and the metaphysical interpretation of existence claims inside and outside modal operators. Lost Highway is obsessed with the difference between being there and being real, between appearing in the frame and being part of the story’s deep inventory. Renee is “there” in the first part, Alice is “there” in the second, but the film punishes you if you simply treat “there” as the mark of what exists. 

The Mystery Man, in particular, behaves as if he is not bound by the usual conditions of presence, he can be at the house and at the party in a way that breaks ordinary individuation constraints. If you try to make sense of this by saying he is “not really real”, you are reaching for a domain restriction. But the film keeps forcing him back into the story as an operative element. 

Williamson’s necessitism is the claim that what the room contains does not exhaust what there is, and that we should not read ontological conclusions off what is locally manifest. Williamson knows that many philosophers instinctively prefer contingentism, the opposing view on which what exists could have been different, in the strict sense that there might have been fewer objects or more objects. 

The most emotionally compelling argument for contingentism is the common sense thought that you could have failed to exist. You were not inevitable. Surely, the contingentist says, reality could have lacked you. Williamson’s strategy is to separate several things that contingentists slide together. “You could have failed to be concrete” is not the same as “you could have failed to exist”. If we equate existence with concreteness, we are building in what we want to prove. Instead, Williamson suggests that the right metaphysical discipline is to keep the quantifier domain fixed and let properties like concreteness, spatiotemporal location, causal powers, and so on, vary across possibilities. 

The choice of logical framework leads you towards a metaphysical picture. It is not merely reporting a prior metaphysical picture. An example helps. Imagine a lover you never had. Common sense says that lover does not exist. But common sense also says it could have existed. Contingentism seems to model this neatly. There is a possible world with that lover, and the actual world lacks it. Necessitism models it differently. On necessitism, there is an object which is that lover, but in the actual world it is non concrete, not located, not causally engaged, perhaps not even determinately person like in any thick sense, while in some other possible situation it is concrete and has the life you imagine. 

That sounds extravagant, until you notice that contingentism also has to make room for “that lover” in some way in order to say anything informative about it. Even in ordinary discourse, you can refer to and reason about non actual individuals. Williamson’s bet is that once you systematise that practice, a fixed domain is not a gratuitous inflation, it is a way of avoiding worse confusions. 

Lost Highway’s own structure is again suggestive. The film has a circular or spiral narrative structure, sometimes compared to a Möbius strip. The opening message, “Dick Laurent is dead”, appears at the beginning and is spoken again at the end, in a way that makes the story feel as if it folds back on itself. The temptation is to say that certain elements in the film only exist in certain segments. But the story’s very circularity pushes you towards another interpretation, namely that the inventory is stable while the modes of manifestation shift. 

Pete is in the “domain” even when the narrative does not allow him to be concrete. Renee and Alice are in the “domain” even if they are one person under different guises, or two persons connected by some structure of projection or substitution. The film’s horror is not just that identity is unstable, but that the system that generates identity is stable and indifferent to your wishes. 

At this stage you might object that the film is about psychology, not metaphysics. Lynch himself used the phrase “psychogenic fugue”, and commentators often talk about guilt and dissociation. But that distinction is exactly the kind Williamson wants you to handle carefully. The metaphysical question is not “is this psychologically motivated”, but “what structure of possibility and necessity is presupposed by the best explanation”. Williamson’s general orientation is close to a scientific realist temperament: we should treat metaphysical theorising as continuous with other theorising, guided by explanatory power, simplicity, and theoretical integration. A logic is not accepted because it flatters our intuitions. It is accepted because it helps to build the best overall theory. 

That methodological note pushes his modal metaphysics towards something like a disciplined science rather than free form speculation. So Williamson’s struggle is to show you why the machinery of quantified modal logic, once you stop treating it as a purely formal game, points towards necessitism as a structurally natural position. 

The alternative, contingentism, tends to require extra devices to say what you want to say without contradiction. It often involves shifting domains, plus additional principles to regulate how reference to non actual individuals works, plus story telling about how cross-world identity is fixed. Williamson’s view is that a fixed domain plus the right modal logic gives you a cleaner core. However, he spends time on what the alleged costs are. The big cost that alarms people is that necessitism seems to multiply entities, populating reality with merely possible individuals. 

Lost Highway can help you distinguish multiplication from re description. The film does not, in any literal sense, add a second world on top of the first. It reconfigures how you count what is “in” the story by refusing your ordinary criteria for individuation. Once you accept that the film’s ontology is not tracked by the cast list, you can stop being spooked by the idea that there are more entities than you first thought. There were always more constraints than you recognised, and those constraints govern what you can legitimately infer. 

A second cost is epistemological. Even if the domain is fixed, how do we know what exists, especially if some of it is non concrete or otherwise inaccessible? Williamson’s broader work in epistemology emphasises that knowledge is not limited to what is “given” in experience, it is the product of reliable cognitive capacities operating within a world that is not tailored to our accessibility. That outlook supports the thought that we can be entitled to modal knowledge by disciplined theorising, even if we cannot point to the merely possible like an object on a table. In Lynch terms, you cannot “see” the structure that makes the tape appear on the porch, but you can infer that there is a structure from the way events are constrained. 

There is also a technical aspect that Williamson uses carefully, the distinction between semantics and metaphysics. A standard way to explain modal logic is to use possible worlds semantics, where “possibly P” means “P is true at at least one possible world”, and “necessarily P” means “P is true at all possible worlds”. This is a semantic tool, a way to model the logic. 

Williamson’s insistence is that we should not mistake the semantic model for the metaphysical picture. The fact that we can represent modal talk using possible worlds does not force us into a metaphysics where worlds are concrete universes in the way David Lewis famously proposed. Williamson does not set out to refight the Lewisian battle in detail (although I’ve seen him in many videos on line where he does that), and he signals that the focus is on other structural issues. But he is clear that modal logic can be metaphysically significant even if you do not inflate worlds into vast concrete objects. 

Lost Highway again is helpful. You can model the film’s structure in many ways. You could say there are two timelines, or two identities, or two narratives that interpenetrate. Those are semantic or interpretive devices. The metaphysical question, within the fiction, is what the story commits you to as real inside its own universe. 

The Mystery Man might be interpreted as an external demon, a psychological projection, an embodiment of surveillance, or a personification of a camera. Each is a model. The film’s deeper commitment, though, is that whatever he is, he is not optional. The narrative rules treat him as a fixed point. Similarly, Williamson wants modal logic to give you fixed points, constraints that any adequate metaphysics must respect, even when you vary your favourite interpretations. Williamson is building a case that the right “structural core” for modal metaphysics is a quantified modal logic that is strong enough to sustain serious metaphysical generalisations, and that when you let that logic do its work it naturally favours necessitism. 

Ted Sider, in engaging with Williamson, puts the thought sharply by describing quantified modal logic as supplying a central structural core to theories of modal metaphysics, even when disputing what simplicity should mean in that context. The important point for our purposes is that Williamson is trying to get you to stop thinking of these as merely technical choices. They are the metaphysical levers. 

To deepen this, it helps to watch how Williamson handles the sense that contingentism is the “default”, and how he tries to flip that burden. The contingentist begins with an intuition like this: surely it is possible that nothing existed. Or at least, surely it is possible that fewer things existed. It can feel like mere stubbornness to deny that. But Williamson’s method is to push you away from raw intuition and towards disciplined constraint satisfaction. If you say “it is possible that nothing exists”, what exactly do you mean? If by “nothing exists” you mean “there are no concrete things”, then necessitism can accept that as a live possibility, depending on what further metaphysical principles you adopt. If by “nothing exists” you mean “the domain is empty”, then you are adopting an interpretation of the quantifiers that may be unfit for metaphysics, because metaphysics wants the quantifiers to be unrestricted and stable across modal variation. 

You are also threatening the very logic you are using, because many inference rules presuppose non emptiness in subtle ways. Williamson’s instinct is that apparent metaphysical freedom can be purchased only by covertly weakening the conceptual tools in play. He wants to keep the tools sharp. 

Think again of Lost Highway. People often say, after the first viewing, that the middle section “is not real”, or that it “never happened”, as if the film contains a segment that is simply absent from the story’s ontology. That is the analogue of the empty domain move. It feels liberating. It lets you keep your ordinary assumptions intact. But it does not respect the constraints the film itself imposes. The second half is not a detachable dream, because details cross leak, identities overlap, and the loop structure binds the opening and closing around it. The film punishes you for treating a region of its narrative space as ontologically void. 

In Williamson’s terms, shifting domains can be made consistent, but it introduces a question. Why should the logic of modality and quantification have that shape, rather than one in which the domain is fixed? What theoretical work is the shifting doing? Often, it is doing the work of protecting pre theoretical intuitions about existence, especially the intuition that to exist is to be concrete. 

Williamson thinks that is a mistake. Existence, in the quantificational sense, is not the same property as concreteness, and metaphysics should not bake that identification into its foundations. This is where the Barcan related principles return in a more metaphysical register. If you accept a fixed domain, Barcan like principles can become attractive because they express the idea that whether something is in the domain does not vary across modal scenarios. If you reject them, you need a story about why quantification behaves differently depending on which possibilities you are considering. 

That can be done, but Williamson’s claim is that it creates a more complex, less integrated theory without compensating gains. An accessible example is the “unborn child” case. Suppose we name a particular sperm and egg that, in fact, never met. We can stipulate a counterfactual scenario in which they did meet and resulted in a child, call her Clara. Now we want to say something like, “Possibly, Clara exists.” 

The contingentist hears that as literally introducing a new entity into the ontology of that possible scenario. The necessitist hears it as saying that there is an object in the domain which is Clara, and the scenario makes her concrete. The controversial part is not the modal claim, both agree it is possible in an intuitive sense, it is what the quantifier is doing in the background. 

Williamson’s line is that once you allow reference and quantification over non actuals in modal contexts, you are already acting as if the domain is fixed. You are already treating Clara as something you can talk about, even if she is not concrete here. Necessitism makes that implicit practice explicit and systematises it. 

Now bring Lynch back in. Pete Dayton is like Clara. In one register of interpretation, Pete is a “possible person” generated by Fred’s breakdown, not genuinely part of the story’s furniture. In another register, Pete is a stable occupant of the story’s domain whose concreteness in a given segment depends on the narrative modality that is active. 

The second register, strange as it sounds, matches the film’s own repetitious insistence that the identities are not cleanly separable and that the loop structure binds them together. Pete is not merely imagined, because the film treats his arc as causally efficacious inside the overall structure, and causal efficacy inside a fiction is the closest analogue to concreteness. 

Williamson also spends energy early on resisting a particular misframing. Some readers assume that modal metaphysics is mainly about possible worlds, about whether worlds exist, and if so what kind. Williamson says no to that. He does not deny that possible worlds talk can be helpful, but he is much more interested in how quantified modal logic constrains metaphysical theorising about existence, identity, and generality. These metaphysical issues emerged through the development of quantified modal logic, which is a clue to his guiding conviction: the formal development was not merely technical progress, it was metaphysical discovery. 

That has a striking consequence. If modal logic is a site of metaphysical discovery, then metaphysical disputes are not exhausted by trading intuitions. They can be advanced by finding better axioms, better principles, better systematisations, in much the way science advances by finding better theories. This of course has a “scientific” orientation, not in the sense of running experiments, but in the sense of treating metaphysics as theory construction subject to abductive evaluation, where abduction means inference to the best explanation. 

Here the Lost Highway analogy can do a lot of work, because the film trains you into abductive reasoning. Lynch does not give you a key. You construct the best explanation that fits the constraints. Some explanations are too simple, they ignore the tape motifs, the loop, the phone call at the party, and the identity doubling. Some explanations are too baroque, they posit an infinity of hidden events not suggested by anything on screen. The most compelling interpretations are those that achieve a kind of theoretical elegance, covering more of the phenomena with fewer ad hoc moves. That is abductive assessment. It is not proof, but it is rational guidance under constraint. 

Williamson wants metaphysics of modality to be like that. A further subtlety in Williamson’s opening third is that he is not simply advocating necessitism as a free standing metaphysical dogma. He is showing how it fits naturally with a conception of metaphysics that aims at general laws, general truths about being, formulated in a language that includes necessity, possibility, and unrestricted quantification. 

He makes clear that the ambition is to discover which generalisations in such terms are true, and that we do not know ahead of enquiry which quantified modal logic is correct. That remark signals that even the logic is, in some sense, answerable to metaphysical truth, rather than metaphysical truth being whatever fits a stipulated logic. 

That reverses the usual order. Instead of saying, “Here is the correct logic, now let us apply it to metaphysics”, it says, “Here is a disciplined method, let us see which logic metaphysics itself requires.” How does that relate to the early necessitism debate? It means that the defence of necessitism is about presenting necessitism as the best theoretical setting for doing modal metaphysics at all, because it yields a more stable core from which to derive further results. If you like, it is like saying that Lost Highway is not just an odd story about a man who turns into someone else. It is a film that demands a particular kind of interpretive logic from you, and once you accept that logic, the film’s strange events become intelligible within a coherent system. 

The right way to use Lost Highway is to use the film to illustrate why structural constraints matter more than immediate appearances, which in fact would then destabilise any claim that all that’s going on is the logic of dreams, as many claim. Take the videotapes. In the film, a tape arrives showing the outside of the house, then later another arrives showing footage inside the house, including Fred and Renee asleep. The horror is that the boundary of the house, like the boundary of the self, is no longer a reliable delimiter of what is “in” and “out”. 

Williamson’s necessitism targets an analogous philosophical boundary, the boundary between existence and non existence. Many people treat that boundary as absolute, as if “does not exist” means “is not in reality’s inventory in any way”. But modal reasoning, especially with quantifiers, keeps smuggling in cross boundary talk, because we refer to and quantify over what could have existed. The tapes are like those modal claims. They show you that the boundary you trusted is not functioning the way you assumed. 

Now bring in the Mystery Man’s phone call at the party, where he claims to be at Fred’s house and answers when Fred calls home. This is a violation of ordinary constraints on location and identity. In logic, a constraint is something like a rule that must hold in all admissible interpretations. The Mystery Man is not the film’s way of introducing a “non classical” element that forces you to reconsider which constraints are actually governing the story. Williamson shows that the “classical” constraints of quantified modal logic, including fixed domains and their associated principles, are not arbitrary, they are the constraints that make metaphysical theorising coherent. 

Lynch then uses classical logical elements to produce uncanny effects precisely because they are not violations. The uncanny is precisely where home is no longer at home. Alternative logics and contingentism domain switch and so they are no longer talking about home at all. Lynch is talking about home and producing his strangeness within that one totalising domain, as Williamson’s necessitism requires. 

Williamson is unimpressed by the thought that necessitism is “obviously” wrong because it postulates too much. He is asking you to discipline the interface between language and ontology in modal contexts. When you say, “It could have been that there was a Mystery Man”, you are committed to there being a determinate way for your quantifiers and modal operators to interact so that the sentence has a stable truth condition. Necessitism proposes one stable way: the domain is fixed, the variation is in which objects are concrete, and modal operators range over alternative assignments of properties and relations. Contingentism proposes another: the domain itself varies. 

Williamson’s argument is that the first option yields a more unified theory with fewer extra patches. It is useful to isolate three distinct questions that Williamson is disentangling. First, a question of logical form, what principles about quantifiers and modal operators yield a system that behaves well, that supports the inferences we rely on, that avoids paradoxes or unintended trivialities? This allows Lynch to produce within his compelling cinematic stagings the uncanny using well behaved quantifiers and modal operators. 

Second, a question of metaphysical interpretation, given such a system, what does it say about being, about what there is, about whether objects exist necessarily or contingently, about whether existence is best understood as a quantificational matter or as a property like concreteness? Lynch’s cinema presses that question hard. 

Third, a question of methodology, how do we decide between competing packages of logic plus metaphysics, if not by intuition alone? Lynch produces worlds where our intuitions hardly help at all, and this then opens the door to Williamson’s project of using logic to produce metaphysics and ontologies. 

Williamson is heavily concerned with the second and third, but always through the first. That is why the work can feel like it is moving on two levels at once. Lost Highway does that too. On one level you are watching events. On another, you are learning the rules that make those events possible within the film. Lynch never states the rules, but he enforces them. Williamson states the candidate rules, then argues about which package best explains our modal talk and metaphysical ambitions. 

If you want one last Lynch based example that mirrors Williamson’s core move, focus on the line “Dick Laurent is dead”. It is heard at the beginning on the intercom, and it is spoken again by Fred at the end when he buzzes the intercom. The sentence seems to come from nowhere and yet it anchors the loop. It is like a necessity claim inside the fiction, a fixed point that the narrative keeps returning to. 

Williamson is looking for those fixed points in metaphysics, not in sentences, but in structural truths. Necessitism is offered as one such truth, not because it is emotionally obvious, but because it makes the overall system of modal metaphysics loop without tearing. His guiding thought in Modal Logic as Metaphysics is that you can treat the formal machinery of modal logic, the boxes and diamonds, the quantifiers, the model theory, not as a neutral calculus you happen to use for talking about metaphysics, but as something that itself carries metaphysical commitments when you read it in the intended way. 

He is explicit that the central dispute is whether it is necessary or contingent which things there are, and that he moves from a non technical statement of that metaphysical issue, to the Barcan principles, to the possible worlds model theory that lets you see why those principles are not just scholastic curiosities. 

Lynch’s Lost Highway is almost too obliging for this job. Its story behaves like an argument that keeps producing new cases, new pressures, and new apparent counterexamples to any neat picture of existence across possibilities. It does that by making you inhabit a world in which who exists, who is identical to whom, and what counts as the same person from one scene to the next, keeps sliding under your feet. 

The film’s notorious transformation, Fred Madison in prison replaced by Pete Dayton, is the sort of event that forces you to decide whether you are looking at a single thing persisting through different conditions, or at one thing vanishing and another appearing. That is almost exactly the fork Williamson opens with, only without the melodrama and done in clean metaphysical terms. 

A claim is necessary if it could not have been otherwise, it holds no matter how things might have been. In modal logic this is often written with a box, □. A claim is possible if it could have been otherwise, it holds in at least one way things might have been. That is written with a diamond, ◊. 

Then you add quantifiers. “Something exists with property F” is ∃x F(x). “Everything has property F” is ∀x F(x). The pressure point is what happens when you put these together, because then you are no longer just saying “possibly, there is a car crash”, you are saying things like “possibly, there exists someone who is a saxophonist”, and then you must decide whether that “someone” is drawn from a fixed roster of objects that are there in every possibility, or from a roster that varies with the possibility. 

Take the first act of Lost Highway. There is a house, there is a couple, there are videotapes that arrive showing the exterior of the house, then the interior, then the couple asleep, then the aftermath of murder. The tapes behave like a crude possible worlds device. They look like alternate “takes” of the same setting, except the point is not that they are alternate, it is that they are presented as somehow already there, already available, not created by the characters. 

The tapes insist that there is a determinate fact of the matter about what is in the house, even when the characters do not have epistemic access to it. That is the atmosphere Williamson wants for metaphysical modality: not “what I can imagine”, but an objective space of ways reality could have been. Contingentism, in its simplest form, says that which things exist is contingent. There could have been fewer things, or more things, and in some ways things might have been, certain individuals simply would not have existed at all. Necessitism says, very roughly, that existence is not like that. Necessitism says that necessarily, everything is something. Everything necessarily exists, even if it does not necessarily have the same features. 

Williamson’s own preferred version allows that many things are necessarily in the domain, but only contingently concrete, only contingently located in spacetime, only contingently “active” in the world. In the background is the thought that the simplest, strongest modal logics, when read with standard rules, push you toward that picture. You might hear “everything necessarily exists” and think it is obviously false, because surely you could have had a world where you were never born, and then you would not exist. 

But that is where Williamson’s way of sharpening the issue matters. He wants you to separate the everyday, emotionally loaded contrast between being born and not being born, from the formal question of what the quantifiers range over in modal discourse. One way he makes this palatable is by allowing that the objects that “exist necessarily” might be non concrete in many worlds. On such a view, in a world where you were never born, there is still an object that is you, but it is not concrete there, it does not occupy space, it does not have a body, it does not do anything. It is, if you like, a merely possible individual, a “bare possibilia” as Ted Sider says. 

That sounds eerie in the abstract, so let the film do some work which kind of proves that it really is eerie. In the second half of Lost Highway, Alice Wakefield looks like Renee Madison. Same actor, same face, different name, different relations, different social embedding. The film uses that resemblance to make you feel a temptation: perhaps Renee and Alice are “the same woman” under different guises. But it also gives you strong prompts to resist that identification, because the relations do not line up, the photograph changes, and what counts as evidence slides around. 

Now imagine taking the necessitist line on the film’s cast list. The domain of objects, the “there is something”, is fixed across the film’s shifting strands. The object that is Renee is in the domain even in the Pete strand, but it may not be concrete in that strand, or it may be concretely present only under a different cluster of properties that makes the characters call her Alice. 

The contingentist line, by contrast, is that the Pete strand contains an extra individual, Alice, who is not there in the Fred strand, or else that Renee exists in the Fred strand and not in the Pete strand. Lynch makes both pictures feel live, and the discomfort you feel about choosing is the discomfort Williamson is professionalising and systematising. Williamson spends time distinguishing forms of necessitism, and showing how easy it is to talk past each other if one side is using quantifiers in a restricted way, “everything in this world”, while the other is using them unrestrictedly, “everything, full stop”. 

That distinction between restricted and unrestricted quantification is one of the quiet motors of the debate, because some objections to necessitism work only if you silently restrict the domain to what is concrete in the actual world. Williamson wants you to keep an eye on the unrestricted reading. 

If you want a film analogue for restricted quantification, think of the detectives in Lost Highway arriving and treating the house, the marriage, the tapes, as the whole domain that matters. Their professional stance is that the relevant objects are the ones in this evidential frame. But the Mystery Man’s phone call trick, claiming to be at Fred’s house and then answering the call, is a literal intrusion of the thought that the domain is larger than what the investigators have decided to count. 

Williamson’s version of this is methodological: if you want to talk about what could have existed, you cannot simply define the domain so that it includes only what actually exists. Doing that bakes contingentism in by stipulation. You end up with an argument that feels decisive but only because you changed the subject. 

One of Williamson’s early clarifications is about the difference between necessitism and a view often associated with Meinong, sometimes called Meinongianism, on which there are non existent objects that nonetheless have properties. Williamson's necessitism says that everything is something, not that there are non entities with properties. It is trying to defend a robust quantificational realism, not a shadow ontology of “objects that do not exist”. That is part of why he is willing to say that if you quantify over possible Fs, possible individuals, then you are already, in some sense, quantifying over objects. 

Here Lost Highway is again a useful prop, because the film keeps producing things that feel like they are “there” without being straightforwardly there. The videotapes are the simplest example. They function as evidence. They are physical. They can be rewound. And yet they come from a point of view that no ordinary character occupies. They are not merely fantasies. They are not just “non existent objects” in the sense of something imagined. They have causal roles. The film thereby dramatises a difference between “not in my world” and “not real”. 

Williamson wants the metaphysical discussion to respect the analogous difference between “not concrete here” and “not an object at all”. The reason for the historical chapter, and Williamson is explicit about this, is that these principles are the place where the metaphysical dispute and the formal dispute visibly lock together. The intuitive core can be expressed without symbols. The Barcan direction says that possibility does not let you conjure brand new individuals from outside the domain. If “there could have been a saxophonist who never existed” is true, then there is already, among the things you quantify over, some individual such that that individual could have been a saxophonist. That is a necessitist friendly thought, because it fits the fixed roster picture. 

The contingentist, who wants the roster to grow and shrink with the world, is inclined to resist it. The film can make this resistance visceral. Consider the prison cell. Fred is there, convicted of murder. Then, during a check, the guards find he has been replaced by Pete. If you take the film’s surface as a description of a metaphysical possibility, it looks like an individual has been “conjured” into existence, Pete appears where Fred was. 

Now suppose you were forced to translate that cinematic event into a disciplined modal language. You would have to choose between something like: 

1. There is an individual, Fred, who is later identical to Pete, the same object with two clusters of properties, one cluster making him Fred, another making him Pete.
2. Fred ceases to exist and Pete begins to exist, so the domain has changed.

 The first is a fixed domain picture. The second is a variable domain picture. Lynch refuses to adjudicate cleanly, but Williamson’s point is that once you start embedding quantifiers inside modal operators, you cannot avoid adjudication, because your logic will validate or invalidate principles like the Barcan formula depending on how you handle the domain. 

Williamson’s historical tour, through Barcan Marcus, Carnap, Prior, and others, shows that what look like purely technical moves, like accepting the rule of necessitation which says that if a statement is a theorem then its necessity is also a theorem, can become metaphysically explosive once you combine them with identity and quantification. 

One simple derivation Ted Sider for one discusses in this vicinity goes from the trivial identity t = t, to “there exists something identical to t”, to “necessarily, there exists something identical to t”, and that last step, if you take it at face value for arbitrary t, looks like necessary existence. We might feel that something has gone wrong, because it seems like a technical trick is generating an ontological conclusion. Williamson’s counter pressure is: perhaps it is not a trick, perhaps it is exposing what you are already committed to if you treat the logic as capturing our best systematic norms for reasoning about necessity and possibility. If your formal system is strong, simple, and fruitful, you do not get to refuse its metaphysical consequences just because they sound odd at first hearing. 

That is what provoked a lot of discussion of whether Williamson is leaning too heavily on simplicity and strength as theoretical virtues for logic and metaphysics, something that Ted Sider is inclined to think I think. Lost Highway can serve as a pressure gauge. The film is strong and simple in its own way, almost primitive in its motifs. A message, a house, tapes, a double, a highway loop. But that very aesthetic strength produces metaphysical consequences in the viewer’s mind. You find yourself committed to something like a “two story ontology” even if you did not ask for it. Either there is one man with two lives, or there are two men, or there is some third option in which the identities are not classical at all. Lynch is not arguing, but he is forcing commitments by the way he structures what counts as evidence. 

Williamson is doing the same thing in a different register. Williamson also tracks attempts to avoid Barcan style commitments (eg If “there could have been a saxophonist who never existed” is true, then there is already, among the things you quantify over, some individual such that that individual could have been a saxophonist) by weakening the logic, for example by restricting necessitation, or by allowing truth value gaps, where some statements are neither true nor false because their presuppositions fail, such as statements about a person in a world where that person does not exist. 

Those approaches are attractive to contingentists because they let you say “in the world where you were never born, statements about you have no truth value”, thereby blocking certain inferences. But Williamson wants you to notice the costs. You pay with logical complication, with fractured inferential norms, and with less elegant overall theory, as Sider keeps pointing out 

Think of the way Lost Highway sometimes seems to produce “truth value gaps” in its own internal evidence. The photograph that Pete finds, showing Dick, Andy, Alice, and Renee, later seems to fail to contain Alice at all. The film is not merely hiding information, it is making the evidential object itself unstable.  

That is an artistic version of saying: some propositions do not get to be true or false because the objects they purport to be about are not there, or are not well defined. If you build your metaphysics around such gaps, you can often dodge particular inferences, but you are also changing what counts as a stable representational object, whether a proposition has a determinate content across possibilities. 

Williamson’s inclination is to treat that as a red flag unless it buys you something overwhelming elsewhere. So now you can see why Williamson thinks the metaphysical question is entangled with model theory, and why the “possible worlds” talk, so often treated as metaphysical window dressing, is in fact part of the technical mechanism that decides which logics validate which principles. 

Possible worlds model theory can sound intimidating, but the core idea is straightforward. A model theory gives you a mathematical structure in which you can interpret your sentences and evaluate whether they come out true. For modal logic, the classical approach, associated with the late great Saul Kripke, uses “worlds” as points, an accessibility relation between worlds that tells you which worlds are considered possible relative to which, and an assignment of which statements are true at which worlds. For quantified modal logic, you also need a domain of individuals. And then, crucially, you must decide whether the domain is the same at every world or can vary with the world. 

This is where Williamson’s slogan, that modal logic can be read as metaphysics, becomes concrete. A fixed domain Kripke model looks like necessitism in mathematical clothing. A varying domain Kripke model looks like contingentism in mathematical clothing. And principles like the Barcan formula correspond, in the semantics, to constraints on how the domains relate across accessible worlds. 

Go back to Lost Highway and treat each coherent strand, the Fred strand and the Pete strand, as a “world”. The accessibility relation is the film’s own sense of which strand can be reached from which, perhaps via the prison transformation or via the desert cabin where identity flips back. Then the key question becomes: is the cast list the same at both strands. If it is, we are in fixed domain territory. If it is not, we are in varying domain territory. 

The film plays with the idea that some roles, like the Mystery Man, cut across strands with uncanny invariance, while others, like Alice, seem to appear only in one. That is a near perfect illustration of why philosophers are tempted by varying domain semantics: it matches the phenomenology of ‘new characters’ entering the scene when you shift possibilities. 

But Williamson’s point is that you should not let that immediate phenomenology settle the matter. You should look at the systematic costs and benefits. Fixed domains tend to preserve stronger logical principles. Varying domains tend to force you to complicate the logic or to accept the failure of certain inferential moves. As Sider explains, Williamson wants you to see that the most powerful and systematic modal logics, the ones with the greatest unifying explanatory reach, sit naturally with fixed domains, and therefore with necessitism. At the same time, Williamson is not merely a defence of fixed domains but also a lesson in how to avoid a common confusion about possible worlds. 

Many philosophers, and many casual readers, treat possible worlds as if they were giant alternative universes, places like ours but different. That is one metaphysical picture, and David Lewis famously pushed it hard, but Williamson is careful to say that his emphasis is less on the metaphysics of worlds themselves and more on how the semantics works, on the structure of the model theory and what it allows you to represent. That is why Lost Highway is especially apt. The film’s two strands are not presented as two separate universes you could travel between, like science fiction. They are presented as a single, psychologically and narratively folded space, likened by commentators to a Möbius strip, and described by Lynch in terms of a fugue state, an identity break. In other words, the film gives you “worlds” without giving you a metaphysics of worlds. They are representational devices within the work, not objects in the work. 

That is close to Williamson’s attitude to possible worlds semantics. You can use worlds as a semantic framework without taking them as fundamental furniture of reality. The semantics is a tool for tracking necessity and possibility, and the metaphysical conclusions come, for Williamson, not from reifying worlds but from reading off what the best semantic theory, combined with the best logical principles, commits you to about existence and identity. There is another Williamsonian subtlety here that the film can clarify: the idea of metaphysical universality or unrestrictedness, which he explicitly flags in both propositional and first order contexts. 

The thought is that when we are doing metaphysics, we do not want a logic whose quantifiers are surreptitiously restricted to a special kind of entity. We want a framework that can talk about anything there is, in the broadest sense. In the film, the temptation to restrict the domain is constant. You want to say “the real world is the Fred world, the Pete world is a fantasy”, or “the Pete world is the real world, the Fred world is a delusion”. But the film’s own devices, especially the tapes, work against that neat partition. They keep leaking content from one strand into another. They keep implying that the “unreal” strand has causal bite. 

The result is that any attempt to restrict the ontology to one strand feels like an interpretive violence against the work. You can do it, but you pay in explanatory awkwardness, you must add ad hoc rules about what counts as real evidence. That is analogous to Williamson’s complaint that many contingentist manoeuvres achieve their result by restricting quantification or by building in extra semantic machinery that makes the overall theory less systematic. 

Now, none of this yet settles the dispute, and Williamson knows it. So he’s trying to show that the apparent metaphysical question, “could there have been different things”, is not to be settled by quick armchair intuition, because those intuitions are too easily entangled with shifting uses of quantifiers, with assumptions about identity across possibilities, and with confusions between semantic frameworks and metaphysical pictures. 

The Barcan principles (If “there could have been a saxophonist who never existed” is true, then there is already, among the things you quantify over, some individual such that that individual could have been a saxophonist. ) and the model theory are introduced as a way of turning the debate from a fog of slogans into a structured comparison of theories. 

Take a mundane modal claim: “It was possible that Fred never met Renee.” If you translate that into a quantified modal framework, you might be tempted to treat it as a story about a world in which there is no Renee. The contingentist can say, in that world, she does not exist. The necessitist can say: there is still that individual, Renee, but she is not concrete, or she does not stand in the meeting relation to Fred. The difference affects what inferences you can make. For instance, if Renee does not exist in that world, then statements like “Renee is identical to Renee” might be treated as lacking a truth value there, depending on your semantics. If Renee exists but is not concrete, then “Renee is identical to Renee” remains straightforwardly true, and you can use it inside modal reasoning. 

Now notice how Lost Highway pushes you toward the second picture. Even when Renee is “gone”, the film keeps her as a structural presence. Fred dreams of her attacked. He “briefly sees” another face on her. Pete finds Alice, a near duplicate. The Mystery Man insists “there is no Alice, only Renee”. The film’s own internal insistence is that the individual is there as an anchor even when her concrete role changes. 

That is a remarkably good intuitive staging of the necessitist thought: the object is there across the modal space, but its mode of being there changes. Another example concerns identity across possibilities. One reason contingentists like varying domains is that they think it respects the intuition “if I had not existed, there would have been no object that is me”. Williamson’s necessitist says, instead, “if you had not existed, there would still have been you, but not concrete”. 

The film gives you a felt version of the dispute in the Fred to Pete transformation. If you insist there is no cross strand identity, you get a clean break but you also make the film’s echoes and loops less intelligible. If you insist there is identity, you preserve the echoes but you must accept a strange kind of persistence that is not the everyday persistence of a body through time. That is very close to Williamson’s strategy. 

He is prepared to accept the strange persistence, because it buys theoretical power and simplicity in the overall logical system. (This is something he does when he thinks about vagueness.) And he thinks the demand that persistence must always look like bodily continuity is not an argument, it is an unexamined prejudice imported from an overly concrete view of what it is to be an object. This is also where the analogy with time, which Williamson flags as a subsidiary thread, becomes illuminating. 

The metaphysics of time has a similar dispute. Some views say that only present things exist, others say that past and future things exist too. You can model those disputes with similar machinery, domains that vary with times or are fixed across times. In Lost Highway, the temporal order is itself strange, the opening message “Dick Laurent is dead” returns at the end, as if the film contains its own future in its past. The narrative loop makes it hard to insist that only the “present scene” exists in the relevant sense. The film again nudges you toward a more expansive ontology, one in which the cast of entities is stable while their temporal or narrative roles, their guises, change. 

If you take seriously the idea that Lost Highway is, as it has often been described, Möbius like, then the film becomes a picture of a space in which “where you are” does not fix “what exists”, it fixes only which aspects are manifest. That is an aesthetic cousin of Williamson’s picture of modal space. Of course, a sceptical reader can say: film is not metaphysics, and Lynch is not offering a theory. True. But Williamson’s ambition is not undermined by that. The point of using the film here is not to treat it as evidence for necessitism. It is to use it as a stable set of imaginative prompts that keeps the distinctions vivid. Williamson’s framework is precisely a way of controlling imagination with formal structure. 

We have three big claims in view  already shaping everything. First, the dispute about what exists across possibilities is a genuine metaphysical dispute which becomes visible only when we insist on unrestricted quantification and on a systematic semantics. 

Second, the Barcan principles are the formal expression of deep choices about whether possibilities can introduce new individuals. Accepting or rejecting them is about what you think existence amounts to when you embed existential claims under modal operators. 

Third, Kripke style model theory is a disciplined environment in which different metaphysical pictures become different semantic constraints, and you can then compare those pictures by theoretical virtues such as simplicity, strength, and explanatory integration. Williamson’s later chapters raise the temperature dramatically, moving to predication, higher order logic, and the interpretation of higher order quantification, but you already see the methodological stance: do metaphysics as if it were a serious theory building enterprise continuous with logic and, in spirit, with science. 

To see how Lost Highway illustrates that last methodological stance, focus on what the film does to you when you try to interpret it. If you settle for a purely local reading, scene by scene, you can always explain each moment in isolation. But the film punishes that approach. Clues recur. Objects reappear. The same line comes back. The Mystery Man’s existence in one strand pressures you about the other. You are pushed toward a systematic theory of the whole. And when you propose a theory, you must evaluate it by something like theoretical virtues. Does it unify the evidence? Does it avoid arbitrary stipulations? Does it preserve the strongest patterns you can find? Does it explain the loop without hand waving? 

That is exactly the temperament Williamson wants for metaphysics. He is making you dissatisfied with ad hoc responses such as “well, in some possible worlds there just is not any fact of the matter”, unless you have a principled semantics that earns that move. It is making you treat the logic as a constraint on what you can coherently say, in the way that the film’s own repeated motifs constrain what you can coherently take the narrative to be doing. 

And the decisive Lynchian echo, the one that most cleanly tracks the dialectical knot Williamson is setting up, is the line Lynch himself is reported to have used about the film, that it is a psychogenic fugue rather than a conventionally logical story. Because Williamson’s book is, in a way, the opposite provocation. He is saying: even if the space of possibility feels fugue like, even if identity seems to slip, even if our ordinary language tempts us toward ambiguity, you can still, and should, bring the strongest formal logic you can to bear, and you should be willing to accept the metaphysical consequences of the best overall system. 

If you want a single closing image for where we have got to, hold on to the intercom loop. At the start, “Dick Laurent is dead” arrives as an ungrounded necessity, a claim that seems to come from nowhere. At the end, Fred speaks the same line into the intercom, as if the message’s source was always internal to the loop. That is a perfect emblem for what Williamson is doing with the Barcan principles and model theory. What at first looks like an ungrounded axiom, an arbitrary formal commitment, is later shown to be generated by the structure of the best overall theory, by the way identity, quantification, and modality interact once you stop treating them as separable. The necessity, in other words, is not an alien intrusion, it is the voice of the system talking to itself. 

So if the opening stretch of Williamson’s book feels like the moment in Lost Highway when the first VHS tape arrives on the porch, the middle stretch feels like what happens after you press play and realise the tape is not merely recording the house, it is somehow inside the house. 

The questions are no longer just, “What is the right way to talk about possibility and necessity?” They become, “What must the world be like, if our best regimented way of talking about possibility and necessity is to be taken seriously?” In Williamson’s hands, the technical apparatus of quantified modal logic is a probe, and in the middle of the book the probe goes in deeper. The pivot is that he stops treating modal logic as a neutral calculus that you can interpret however you like, and starts treating it as a disciplined way of doing metaphysics. 

That sounds ok until you remember what quantified modal logic is. It is modal logic, with necessity and possibility operators, plus quantifiers like “everything” and “something”, plus identity, “is the same as”. Once you have that combination, you can state claims that look metaphysical even before you add any metaphysical vocabulary. You can say, for instance, that necessarily everything is something, or that possibly there is something that is not actually something, depending on how you set things up. Williamson tries to show that if you are serious about metaphysical modality, meaning the kind of possibility and necessity that is not merely physical, or epistemic, or legal, but what could have been the case full stop, then the most natural and powerful logical framework pushes you towards surprisingly robust claims about objects, properties, and what there is. 

Lost Highway is not just confusing, it is confusing in a structured way. The structure is exactly what keeps viewers returning to it, and it is exactly what Williamson thinks logic does for metaphysics. Logic does not tell you which story is true, but it tells you which stories can be told without cheating. A first major theme is model theory, the business of explaining truth in a formal language by giving it a mathematical “model” consisting of things like a domain of objects, an interpretation of predicates, and for modal logic, a set of possible worlds with an accessibility relation connecting them. 

Williamson spends time insisting that this is not an optional extra. If you want to understand what your modal claims amount to, you need to understand what would make them true. In the film, the VHS tapes serve as a crude model theory for Fred’s life. They function like an external semantics. Fred and Renee’s everyday experience is one thing, but then there is the tape, which says, “Here is what is the case”, and it does so with a weird authority because it is not merely someone’s testimony, it is an apparent record. 

When a later tape shows Fred standing over Renee’s dismembered body, the tape becomes a sort of truth condition generator. It is not a proof, but it is treated as evidence of what the world must contain for that content to be true. Williamson’s point is that if you take modal discourse as aiming at truth, then you need some disciplined account of how modal sentences get to be true or false. In quantified modal logic the hard part is always quantification across worlds, because that is where metaphysics sneaks in. 

When you say, “Possibly, there exists an x such that x is a unicorn”, what does your “there exists” range over in the possible world where the unicorn exists? Does it range over a domain that changes from world to world, or a fixed domain that is the same across all worlds? The technical choice corresponds to a metaphysical choice. If the domain varies, you are flirting with contingentism about what there is, the view that there could have been fewer or different objects. If the domain is fixed, you are flirting with necessitism, the view that necessarily, everything is something, meaning that the quantifier domain is necessary, even if what properties objects have varies. That is the familiar battleground, but Williamson tries to show that the battleground is not just a matter of taste. Certain very natural constraints on how the language works, and on how we should reason with it, make fixed domains look like the stable option. 

Lost Highway offers a vivid picture of what a fixed domain feels like. Pete Dayton is introduced as if he is a different person. Different family, different job, different girlfriend, different streets. Yet the film keeps slipping in insistences that the cast has not changed as much as it pretends. Alice resembles Renee, the Mystery Man threads through both halves, the same ominous phone call logic recurs, and eventually Pete collapses back into Fred. The film behaves like a story with a fixed cast list, even when it seems to switch casts. The world shifts, but the inventory of “possible characters” feels strangely constant. That, roughly, is what necessitism wants at the level of ontology. The furniture is all there in the warehouse, even if in a given scene only some items are on stage. 

Williamson’s defence of taking that picture seriously depends on careful work about predication and modality, the question of how we should understand sentences that attribute properties to objects under modal operators. The metaphysical action begins when you say things like, “Socrates could have been a carpenter” or “This very table might not have existed”. Those are de re modal claims, claims about a thing, rather than merely about a description. The distinction is sometimes put as de re versus de dicto, about the thing versus about the saying. Williamson is interested in how the formal machinery can capture the de re side without hand waving. 

The familiar move is to use quantifiers and identity to “reach into” the scope of the modal operator. Instead of “Possibly Socrates is a carpenter”, you treat it as, “There exists an x such that x is Socrates and possibly x is a carpenter.” That shift matters because it encodes a commitment: the same object, Socrates, is being tracked across worlds. In the film, tracking across worlds is exactly what you struggle to do. Is Pete Fred, is he a different person, is he a projection, is he a cover story? Lynch refuses to say, but the film makes the tracking problem feel visceral. 

When Pete sees the photograph and his nose bleeds, then he wanders a hotel corridor that seems to interrupt his own house, you feel the collapse of a stable de re anchor. Williamson’s point is that a lot of contingentist resistance to necessitist frameworks is powered by an intuitive discomfort that looks like this. It feels as if the object you are talking about cannot be present in a world in which it does not exist. It feels like asking to see Pete inside the death row cell. The demand looks category mistaken. Williamson’s reply is to argue that we should not confuse “is in the world” with “is in the domain of quantification for the logic”. The logic’s domain is the full range of things the language can be about. This is where his methodological temperament shows. He trying to show that if you want a systematic logic that supports the inferential practices you already rely on in mathematics, science, and everyday reasoning, then you will end up quantifying over things even when they are not “present” at a world in the way you naively imagine. In other words, he is urging you not to let the phenomenology of presence dictate the semantics of quantification. 

Lost Highway’s Mystery Man is a useful emblem here. He appears at the party, claims he is simultaneously at Fred’s house, proves it by answering the phone, then later turns up filming in the desert, and finally vanishes again. If you treated presence as the criterion for existence, the Mystery Man would be metaphysically impossible. Yet the film insists, within its own rules, that he is the connective tissue that makes the story’s logic run. He is less a character than a device of cross scene identity. 

Williamson does something analogous. He treats cross world identity as a structural requirement for making sense of modal discourse. The demand for a clean semantics is like the demand that the phone call at the party actually connect to something real, rather than being a cinematic trick that you should ignore. The more you insist that it must connect, the more you are pushed towards a world picture with stable cross world objects. This is one reason the Barcan formulas have such gravitational pull in the debate,. Very roughly, the Barcan formula and its converse connect quantification and modality, linking claims about what possibly exists with claims about what exists possibly. 

The informal feel of them is something like this. If it is possible that something satisfies a condition, then there is some thing such that it is possible for it to satisfy that condition, and conversely. A logic that validates both tends to behave as if the domain is fixed across worlds. That makes contingentists uneasy because it seems to entail, in one or another dress, that merely possible objects are “among the things there are”, just not actual. Williamson works hard to dissolve the sense that this is an absurd inflation. He tries to reframe it as a clarification of what our quantifiers were doing all along, once they are treated as unrestricted, which is the metaphysician’s default ambition. 

His next big movement is the step from first order to higher order modal logic. First order logic quantifies over objects, things like people, planets, tables. Higher order logic also allows quantification over properties, relations, or sets, depending on how you construe it. The metaphysical temptation is immediate. If you can quantify over properties, you can state comprehension principles, roughly, for any condition there is a property corresponding to it. But that is dangerous because naive comprehension can generate paradoxes, the way naive set formation leads to Russell’s paradox. Williamson’s strategy here is characteristic. He reacts by trying to say  what principles we should accept, and how they connect to metaphysical claims about the availability of properties and propositions. Return to the film’s tapes. The tapes are not just recordings, they are like higher order entities. They are objects, physical cassettes, but what matters is their content, which is about the objects in the story. The tape can contain Fred’s house, then contain Fred and Renee in bed, then contain a scene of murder. The film thereby encourages you to think in two layers at once, the layer of events and the layer of representations of events. 

Higher order logic encourages a similar doubleness. You can talk about Fred, and you can talk about the property of being filmed, or the proposition that Fred is filmed, and you can ask whether such properties and propositions exist necessarily, or only contingently. A lot of Williamson’s book is an attempt to make this representational layer metaphysically respectable without letting it run wild. One way to see the stakes is to notice that many arguments against necessitism or against robust modal principles rely on what you might call scarcity intuitions. There cannot be that many things. There cannot be an object corresponding to every way the world could have been. There cannot be a property for every condition, because that would be an overpopulated universe, and it invites paradox. 

Williamson’s counter pressure is that once you take modern logic and semantics seriously, scarcity is not your starting point, it is a hypothesis you would need reason to adopt against a background in which abundant mathematical and semantic entities already do a lot of explanatory work. This is where the film’s Möbius strip comparisons become suggestive, even if Lynch himself resists pinning the film down to one explanation. The film’s narrative is often described as looping back on itself, the ending feeding the beginning. In a Möbius strip there is one surface that seems to become two, or two that becomes one, depending on how you trace it. Higher order modal logic has a similar feel because it lets you quantify over entities that are defined in terms of what happens across worlds, and then use those entities to state claims about worlds again. The risk is a kind of self reference. The opportunity is a kind of expressive adequacy, the ability to articulate exactly the metaphysical distinctions you have been gesturing at in ordinary language. 

Williamson’s carefulness comes out in the way he treats comprehension principles intensionally. An intension, in modal semantics, is often a function from worlds to extensions, meaning that instead of thinking of a property as just the set of objects that have it, you think of it as something that tells you, for each world, which objects have it there. This is a standard move in intensional semantics, but Williamson wants to use it metaphysically. 

If properties are intensions, then even if the extension is empty in a world, the property can still exist. The condition “being a unicorn” may have no instances in the actual world, but the property can still be there, ready to be instantiated in other worlds. That again echoes necessitism’s attitude to objects. The domain is stable, even when the instantiation pattern changes. In the film, Alice is the extension, and Renee is the same extension seen under a different intensional profile, at least from Pete or Fred’s point of view. The film’s cruelty is that Pete thinks he is dealing with a new woman, a new extension in his life, but the story keeps insisting that the intensional content is entangled with Renee, that there is no clean break. The Mystery Man even states brutally that “there is no Alice, only Renee.” 

Whether you accept that line as metaphysical truth within the film, it captures the structure Williamson wants you to feel in the logic. You can re describe and re identify, you can shift guises, but the underlying inventory of the things your language can be about does not obediently shrink and expand with your narrative needs. One might worry, at this point, that the analogy is doing too much, that it makes Williamson sound like he is offering a metaphysics of cinematic unreality. But the the most disciplined, way to do metaphysics is to let the logic lead. That is why he pays so much attention to mapping between contingentist and necessitist discourse. 

He has the idea that many contingentist sounding statements can be translated into necessitist friendly language by re locating where the modality sits. Instead of saying “x does not exist in world w”, you say something like “x is not concrete in w”, or “x is not in the spatiotemporal region of w”, or you introduce an actuality like operator, depending on the framework. The point is not that translation always preserves every nuance, but that the alleged deep disagreement may partly be a disagreement about what to take as primitive. 

In Lost Highway, the same thing happens with identity talk. If you insist on the primitive claim “Fred becomes Pete”, you get one kind of metaphysics, perhaps supernatural. If you insist instead on the primitive claim “Fred is in a psychogenic fugue and imagines himself as Pete”, you get a different metaphysics, perhaps psychological, and some mention Lynch’s own phrase “psychogenic fugue” in this spirit. The surface narrative may look like it forces a radical ontological revision, but a re description at the right level can make it look like a story about representation, self deception, and selective memory. The key is that both descriptions are trying to honour the same data, the same scenes, the same tapes, the same phone call. 

Likewise, Williamson thinks rival modal metaphysics must be answerable to the same inferential and semantic data. His sympathy is with the view that makes the least ad hoc sacrifices in our best general purpose theory of modality. This brings out the role of unrestricted quantification. Metaphysicians often want to quantify over absolutely everything. That is what it means to ask what there is, not just what there is in some restricted domain like physics, or biology, or the furniture in your kitchen. But unrestricted quantification is also what makes the logic bite, because if your quantifiers are restricted to what exists in a given world, then many formal principles become slippery. The logic starts to encode metaphysical limitations directly, and then you risk making your metaphysics look true by building it into the semantics. Williamson wants the logic to be a constraint, not a mirror held up to your prior preferences. 

In the film, Fred does the opposite. He seems to want a restricted quantification, restricted to what he can bear to acknowledge. The Pete narrative reads like a restriction of the domain of culpability. It creates a world in which Fred is not the murderer, and in which the violence and jealousy are displaced onto other figures, or onto a gangster plot. Yet the tapes function as an attempt to force unrestricted quantification back in, to insist that all the relevant entities, actions, and responsibilities are in play whether Fred wants them there or not. That is why the tapes are so unsettling. They are like the metaphysician’s insistence that your quantifiers range over everything, including the parts of the story you would like to cut. 

When Williamson turns to higher order resources, he is also responding to a more technical version of the same issue. If you keep the logic first order and relatively weak, you may think you have saved yourself from ontological inflation, but you may also have impoverished your expressive resources so that you cannot state, within the theory, the principles that you implicitly rely on when reasoning about modality, essence, and identity. The move to higher order logic is like admitting that the film is not just a sequence of events but also a sequence of interpretations of events, and that any honest account has to mention both layers. 

Williamson’s project is methodological as much as doctrinal, it is about how modal logic can guide metaphysical theorising. As the formal resources expand the metaphysical reach expands with them. A natural worry is, “How do we keep this from becoming a free for all?” If we can posit necessary existence of objects, and quantification over properties, and comprehension principles, are we not just writing metaphysical fan fiction? Williamson’s answer is that the discipline comes from the same place it comes from in any theoretical domain, the pressure to systematise, to avoid arbitrary exceptions, to match inferential practice, and to preserve powerful explanatory generalisations unless there is a compelling reason not to. Ted Sider’s discussion of the book’s attraction to simplicity captures this aspect, even when he presses questions about it. 

In the film, the discipline comes from the way motifs recur with a kind of necessity. The intercom message “Dick Laurent is dead”, the dark corridor, the phone call, the videotape, the desert cabin, the camera in the Mystery Man’s hands, they keep returning. The film does not let you stabilise your interpretation by discarding the inconvenient scenes. You either find a framework that accommodates them, or you admit that you do not yet understand the film. Williamson thinks modal metaphysics should be treated with the same intellectual honesty. If your preferred metaphysics requires a patchwork semantics that blocks otherwise natural inferences, or refuses otherwise natural generality principles, you have taken on theoretical debt. One of the most delicate issues is the difference between denying necessitism and merely refusing to talk necessitistically. 

A contingentist might say, “There could have been no such object as this table.” Williamson will push you to notice that even stating that claim, in a rigorous semantics, requires you to keep some grip on which object you are talking about across worlds, and therefore to allow that the object is in some sense available to your quantifiers even in worlds where it is not “there” in the ordinary sense. The contingentist may respond by restricting quantifiers to what exists at a world, but then you need a story about the modal profile of the quantifiers themselves, and you risk losing standard logical principles like free use of identity in modal contexts, or you risk treating quantifiers as context sensitive in a way that undermines the metaphysical ambition of saying how reality could have been, not merely how we could have described it. 

This is exactly the sensation of trying to describe the film’s transformations without cheating. You can say “Fred is replaced by Pete” and treat it as brute, but then you have a mystery that is not explained, only named. Or you can say “Fred imagines Pete” and treat Pete as a kind of fictional entity within the fiction, but then you have to explain why Pete’s world has so much stubborn detail, why other characters interact with him independently, and why the film gives him an apparent causal footprint. The film does not let you escape the need for a systematic account. The “replacement” story risks arbitrariness, the “imagination” story risks thinness. 

Williamson’s is a sustained attempt to show that necessitism, combined with a robust modal logic, is the more systematic account. It does not eliminate metaphysical surprise, but it reduces special pleading. Even the film’s ending helps here, because it lands on a gesture of necessity that is not quite repetition, more like inevitability. Fred returns to the house, delivers the same message that began the film, then flees into a chase that dissolves into noise and light on the highway. The story feels as if it has to fold back on itself because its internal constraints demand it. 

Williamson’s view of logic is similar. Once you accept certain natural constraints on modality and quantification, once you accept the inferential roles that modal and quantificational expressions play in our best theorising, you will feel pushed towards certain metaphysical commitments. The commitments are not forced by a single dramatic argument, they are forced by the way the whole system hangs together. 

Williamson is building the machinery that makes it look like the default outcome of a rigorous approach. He moves from the idea of possible worlds as a modelling device to the more provocative idea that the modelling constraints reveal something about the underlying structure of reality, at least at the level of its most general features. He expands the logical resources to include higher order tools because metaphysics needs the expressive power to articulate its own commitments cleanly. And throughout, he presses a methodological moral that the Lost Highway analogy makes unusually vivid. 

If you only accept the scenes you like, you will never understand the film. If you only accept the logical principles that flatter your prior metaphysical intuitions, you will never let modal logic do the work it can do. Williamson starts to make the translation project more explicit, showing how contingentist and necessitist ways of talking can sometimes be mapped onto each other, and he begins drawing out consequences that feel, at first, as unsettling as the Mystery Man’s phone call, but which he wants you to see as the natural price of theoretical coherence. 

In Lost Highway the house is familiar, suburban, dimly lit, and yet it is the site of an intrusion that rewrites what counts as inside and outside, what counts as evidence, what counts as the same person over time. That is the right atmosphere  because Williamson is trying to get you to feel how modal logic, the formal study of necessity and possibility, stops being a decorative overlay on metaphysics and becomes a way of disciplining what you are allowed to say about identity, existence, and properties. 

The hinge is quantification. Quantifiers are words like “everything”, “something”, “nothing”. Quantified modal logic is what you get when you let those words interact with modal operators like “necessarily” and “possibly”. The notorious flashpoints are the Barcan Formula and its converse, principles that connect “possibly” and “everything” in systematic ways. You can read them, roughly, as saying: if it is possible that something is F, then there is something such that it is possible it is F, and conversely. The reason anyone cares is that, once you accept standard ways of interpreting the formal language, these principles pull you towards a picture on which the domain of “things we quantify over” does not vary from world to world. And that, in turn, is one of the main technical pressure points pushing in the direction of necessitism, the view that everything exists necessarily, that is, nothing could have failed to exist. 

Now, in the film, the tapes are the intrusion. A tape arrives, it shows the house from outside, then closer, then inside, and the ordinary boundary between what is there and what is merely imaginable collapses. In Williamson the intrusion is similar in structure. He keeps asking: what is the domain of quantification supposed to be when we are doing serious modal theorising. Is it “all the things that actually exist”, varying as the story varies, or is it a fixed range of objects, some of which may be “non-actual” relative to a given world, but still objects? The tape’s uncanny power is that it seems to show the same house and yet it is not just another viewpoint on it, it is the house as something that can be entered from an angle the occupants do not control. 

Williamson wants the metaphysician to admit that modal discourse does something like that. Once you formalise and try to systematise “could have been” talk, you are forced to decide whether you are controlling the camera by letting the domain fluctuate, or whether you are letting the domain stay fixed and instead allowing the mode of being of things, concrete or not, instantiated or not, to vary. 

This is where his engagement with another contemporary philosopher, Robert Stalnaker matters. Stalnaker is often associated with a kind of sober modal semantics, built around possible worlds thought of as representational devices, not as a plurality of concrete universes in the Lewisian manner. Williamson treats Stalnaker as a subtle test case because Stalnaker tries to respect ordinary modal intuitions while keeping a disciplined semantics for quantified modal logic. The worry that keeps surfacing is that you can make your semantics fit common sense by putting certain “restrictions” in, but then you have to explain what those restrictions amount to, and whether they are principled or ad hoc. 

In Lost Highway you see this as the film’s own self restriction. It keeps presenting you with a coherent noir frame, jealousy, surveillance, a glamorous femme fatale orbit, then it breaks that frame by switching protagonists, as if the film had been trying to honour a familiar narrative semantics but could not do so without turning the “domain” of characters and identities into something unstable. The question becomes whether that instability is a feature of reality or a repair strategy. Williamson’s guiding suspicion is that variable domain strategies, letting the range of “things” change from world to world, tend to be the philosophical equivalent of editing the tapes. They may preserve certain pre theoretical verdicts, for example, “this table might not have existed”, but they do so by building into the semantics a picture of existence as a world relative affair. The trouble is that, once existence is made world relative in that way, you then need extra machinery to recover the logical behaviour of identity and quantification that we rely on elsewhere. 

Williamson is not claiming that variable domain semantics are impossible, he is claiming they come with costs, and that those costs often show up as distortions of what the logical constants are doing. A simple place to see the pressure is identity. Identity is “is the same as”. In standard logic, it is reflexive, everything is identical with itself, and it is governed by substitution principles, if a equals b then whatever is true of a is true of b. When you move to modal contexts, you immediately face the question of “necessary identity”. Is it necessary that a is a, yes, but is it necessary that if a is b then a is b, that is, if they are the same, could they have been different? 

Most philosophers are pulled toward necessary identity for ordinary objects, at least if “a” and “b” are rigid designators, names that pick out the same object in every possible world where that object exists. Williamson’s point is that if you start allowing existence itself to vary by world in a strong way, you start creating scenarios in which the logic of identity is no longer cleanly classical, or else you have to start policing which substitutions are allowed when existence claims are in play. He would rather keep identity logic stable and pay the metaphysical price, than keep the metaphysics intuitive and pay the logical price. 

Lost Highway is, among other things, a sustained staging about identity. Fred is Fred, then Fred is not Fred, then Pete is Pete, yet the film invites you to treat the shift as neither a straightforward disguise nor a simple case of two distinct people. There is a temptation to say the domain has changed, the film has moved to a different world with a different “range” of individuals. Another temptation is to say the domain is fixed and what changes is which individuals are “concrete” in the story’s current segment, which ones are accessible, which ones can be encountered in the narrative space. 

That second temptation is closer to Williamson’s dialectical move. He often distinguishes existence in the unrestricted sense, being among the things there are, from being concrete, being located in space and time, being causally embedded. On a necessitist view you can say, in effect, that the things are all there in the ontology, but not all are concrete in every world. In the film, you can read the Fred, Pete duality as something like that: the ontology of individuals is not being created and destroyed by the narrative, rather, different “modes of manifestation” are being toggled, with the Mystery Man functioning as an agent of that toggling, the one who treats the boundaries as porous because he was never committed to them. 

This leads into Williamson’s ascent from first order to higher order modal logic. First order logic quantifies over individuals, things like people, tables, numbers, whatever your ontology contains. Higher order logic, in addition, quantifies over properties, relations, sets, and so on, things like “being tall”, “being a saxophonist”, “being identical to Fred”, “being the property that…”. The main idea is that higher order quantification lets you say not just “something is F” but “there is a property P such that…”. Williamson’s philosophical point is that if you are trying to state metaphysical principles about what must exist, or about how objects are individuated across possible worlds, you can end up needing higher order resources to do the job without smuggling the conclusion in. 

A crucial notion here is comprehension. Comprehension principles say, roughly, that for any condition you can describe, there is a property, or a set, corresponding to it. In its naive form this is inconsistent, it leads to paradox, so contemporary logic uses carefully constrained versions. Williamson’s interest is that modal and intensional comprehension principles, principles allowing you to form properties sensitive to modal facts, have direct metaphysical consequences. 

If, for example, for each individual there is a corresponding property that picks it out uniquely, a thisness or haecceity, “being Fred and no one else”, then you get powerful tools for tracking individuals across worlds. Williamson explores how such principles interact with necessitism and contingentism.

The Lost Highway analogue is the way the film manufactures and withholds properties of identification. At times the film behaves as if there is a haecceity in play, a property that pins a person down regardless of narrative disguise, the way Renee and Alice seem to share a face but not an identity. At other times the film behaves as if properties are only locally available, as if the world cannot sustain the property “being Fred” across the whole story without contradiction, so it swaps in a different bundle of properties, “being Pete”, and then bleeds the earlier bundle back in through echoes, gestures, objects, and repetitions. 

Watching this, you feel the temptation to say, there just is no single property that uniquely identifies the person across the whole structure, the best we can do is track a role within a segment. Williamson is, in a sense, asking you whether that temptation is a genuine metaphysical insight or a failure of theoretical nerve. Do we really want to say there is no haecceity for an individual, or do we want to say that our representational system is not rich enough unless it can express such thisnesses. When he entertains intensional comprehension, he is probing exactly that issue: are we allowed, on pain of theoretical chaos, to recognise properties whose identity conditions reach across worlds. 

Once you bring this into focus, necessitism starts to shift to being an interpretive key for the film, not only a way of illustrating it. Here is the interpretive turn. The standard contingentist thought, both in metaphysics and in everyday life, is that some individuals are fragile in their existence. If the world had gone slightly differently, Fred might not have existed, Pete might not have existed, Renee might not have existed. The film itself seems to run on that fragile intuition, as if the plot is exploring alternate lives that branch away, as if a different choice, a different jealousy, a different erotic alignment, would have produced a different cast of real people. Under that reading, the transformation is like a counterfactual, a near miss world in which a different person occupies the centre. 

The necessitist reading is stranger and, in the film’s idiom, more faithful to its horror. On necessitism, the individuals are not created by narrative conditions. They are already among “the things there are”, necessarily. What varies is which individuals are concrete in a given world, which are instantiated in the world’s spatiotemporal theatre. That is why necessitism can preserve the intuition that “this table might not have been made”, but reanalyse it as “this table might not have been concrete”, or “might not have been composed in that way”, rather than “might not have been at all”. 

In Lost Highway the Fred and Pete strands need not be two ontologies. They are two concreteness profiles, two patterns of instantiation within a fixed background of individuals and properties. The transformation scene, on this reading, is not the birth of Pete from nothing, but the switch of which already existing individual is now concrete at the narrative centre. The film’s dread comes from the sense that the switch is not under the protagonist’s control because the underlying inventory of individuals is not being managed by him at all. It is managed by whatever the film figures as the metaphysical machinery, which in the story is dramatised as surveillance, videotape, the Mystery Man’s omnipresent phone call, the idea that the inside is always already mapped from outside. 

This interpretive move also helps with the film’s looping structure, the way the opening line returns at the end, as if the story folds back on itself. If you think in terms of variable domains, the loop suggests a reset, the world is rewound and re populated. If you think in terms of a fixed domain with changing concreteness, the loop suggests something closer to a constraint satisfaction problem: the narrative is searching through different allocations of concreteness and properties until it reaches a configuration that can close on itself. The “lost highway” is not just the road in the desert, it is the space of possible allocations, the modal space. The film shows you a few points in that space, not by giving you alternate endings as in a gimmick, but by forcing the same individuals and properties to reconfigure until a stable cycle emerges. 

Where higher order resources become especially potent is in articulating what, exactly, is preserved across the reconfigurations. If we can quantify over properties, we can say: there is a property that Fred has in every segment where Fred appears, and that property is shared, distorted, or mirrored by Pete. We can also say: there is a relation that connects Renee and Alice, not identity but something like counterparthood, the relation of playing analogous roles across worlds. Williamson is wary of counterpart theory, associated with Lewis, as a way of handling trans world identity, because it can make “could have been” claims come out true by changing the subject from the object to its counterpart. 

But the film itself is soaked in counterpart aesthetics, the same face, the same house, the same gestures, the same violent jealousy, re appearing in another register. The necessitist interpretation can treat those counterpart patterns as derivative, as patterns in how concreteness and properties are assigned, rather than as the ground truth about identity. In other words, the film tempts you to think in counterparts, but a Williamsonian lens asks whether the temptation is covering over a more rigid underlying metaphysics. 

Williamson is not simply collecting intuitions, he is weighing theoretical costs, simplicity, uniformity, the power of a framework to integrate disparate phenomena. That is why he is willing to sound, at first, counterintuitive. In the film’s terms, he is willing to accept the uncanny tape rather than insist the house is safe. He is asking whether the metaphysician should accept a stable, uniform logical framework, even if it forces a revisionary account of what it is for things to exist and to be the same, because the alternative is to let existence and identity become local special effects, constantly tweaked to save appearances. That would be a different film – perhaps The Truman Show. 

Lost Highway keeps confronting you with evidence that is at once intimate and external, the tape is of your bedroom, yet it comes from outside you. Quantified modal logic has the same structure. The inferences you make in ordinary modal talk feel intimate, mine, internal to my sense of what could have happened. Williamson’s machinery shows you that those inferences presuppose an external structure, a semantics, a discipline about domains and identity, and once you see that structure, you have to decide whether to obey it or to keep rewriting it to protect your initial self conception. 

The necessitist turn is, on this cinematic reading, the decision to obey the structure, to accept that the inventory of things is not hostage to the narrative segment you are currently living through, even if that means admitting that your sense of fragility, the sense that you or I might never have been, is not a deep metaphysical fact but a surface phenomenology of which individuals happen to be concrete here. And that, I think, is the point at which the film starts to look like a parable about the metaphysical stakes of the choice. 

If the ontology is fixed, then the horror is not annihilation but displacement. It is not that the person ceases to be, it is that he ceases to be here, in this mode, in this concrete manifestation, while still being among the things there are. In Lynch’s world, that is far worse than simple non existence, because it means the universe can keep its inventory intact while swapping the faces at the centre of the room, and the room, like the semantics, does not have to ask permission. 

 Earlier, the dispute with contingentism looked like a dispute about whether everything exists necessarily. Now it becomes a dispute about what you are willing to let your basic theoretical vocabulary do. He argues that necessitism is structurally clarifying, and that the clarifications are not cosmetic. They reorganise what you should expect from explanation in metaphysics. There is a lever Williamson keeps pulling. When he says “necessarily everything is something”, he is pressing on how quantifiers work. Remember, a quantifier is just the “everything” or “something” device in logic. If your quantifiers range over different domains at different possible worlds, then you can say “in that world there are no unicorns” in a very direct way, because the domain at that world simply contains no unicorns. 

Necessitism pushes you toward a different modelling choice: quantifiers range over a fixed domain of objects, and what varies by world is which of those objects are concrete, or instantiated, or spatiotemporally located, or otherwise “present in the world” in the thick sense. Williamson is explicit that necessitism is a claim about metaphysical necessity, not about what we can know or say, and that it does not force you into crude reductions like “only atoms exist” or “a coin has every atom essentially”. Instead, it invites the stranger option that some things can be non concrete at some worlds, while still being things. So unicorns exist but just not concretely. 

That is already a way of reading Lost Highway. Think about what the film does to the status of a person. It does not merely show a man changing, it shows the ontological role of “this man” becoming unstable. Fred is there, then the narrative insists on Pete, then it curls back, and the edges blur in ways that are not well captured by a simple “either Fred exists or he does not”. 

If you bring a contingentist picture to the film, the natural metaphysical gloss is to treat Pete as something like a different inhabitant of a different domain, a different world’s population. Under that gloss, when the story shifts, it is as if the world has swapped domains, and the question “where did Fred go” becomes a question with no answer inside the model. The model has simply stopped ranging over him. Williamson’s necessitist modelling gives you a different and, for his purposes, better behaved gloss. The domain does not swap. The story can shift which objects are concrete, which are visible, which are psychologically integrated, which are socially tracked, but it need not change what there is. The film can then be read as dramatising a metaphysical separation between existence and concreteness. 

Concreteness here is not just being physical, it is being in the world as the sort of thing that can be causally implicated, filmed by the camera, arrested by the police, desired by someone else, located on a map. Williamson’s view is that you can have objects that exist necessarily even if they are non concrete in some worlds, and he spends time warning you not to confuse that with the silly claim that, if a stick had never been concrete, it would still have been a stick. That is where his distinction between two readings of phrases like “possible stick” does real work. On one reading, “possible stick” means “a stick that could have existed”. On another reading, it means “something that could have been a stick”. It is the second reading, the attributive reading, that lines up with necessitism. 

Now notice what that does to Lost Highway. If you say Pete is “a possible Fred” in the first sense, you are forced into a kind of category mistake, as if Pete is already Fred, plus an extra modal property. But if you say Pete is “possible Fred” in the second sense, you can mean: here is something, already in the domain, that could have been Fred, could have been the man tracked as Fred, could have borne those properties, could have occupied that narrative role. The film’s uncanny effect is partly that it keeps you oscillating between these readings without letting you settle. 

At times it pushes you toward the predicative feel, as if Pete simply is Fred under a disguise. At other times it pushes you toward the attributive feel, as if Pete is instead an alternative career of the same underlying object, with the link to Fred being modal rather than straightforwardly factual. Williamson cashes out why he thinks this is not just an optional way of talking. 

He tries to show that the necessitist framework is theoretically attractive in the way a good scientific framework is attractive, because it makes certain principles simpler, more general, less riddled with special cases. The claim is structured around four themes: a necessary framework of objects, pressure on supervenience claims, scepticism about truthmakers, and a reconception of contingency, change, and difference. 

Start with the “necessary framework of objects”. Williamson wants the background ontology to stop fluttering as you move across possible worlds. A simple way to put the motivation is this. Modal discourse, talk about what could have been, is already committed to treating “the same thing” as being available for recombination with different properties. We say the same cat could have been heavier, the same person could have chosen differently. When you formalise that discourse, you either build the cross world availability of objects into your identity conditions, or you hide it in the semantics by letting domains change and then adding a counterpart relation or an existence predicate or some other device to mimic cross world talk. Williamson’s bet is that the cleanest overall theory is the one that lets quantifiers remain unrestricted and lets identity behave in the straightforward classical way, with a stable domain across modal operators. That makes your formal system smoother, at the cost of populating it with “merely possible” objects, objects that are not concrete at the actual world. Critics often treat that cost as immediate absurdity. Williamson treats it as a familiar kind of theoretical trade off, closer to set theory’s willingness to accept huge infinities for the sake of principled mathematics. 

If you pull that thread into the film, you get a way of interpreting the “mystery” not as a supernatural invasion but as pressure from the necessary framework itself. The Mystery Man’s strange authority, his seeming to be simultaneously outside and inside the narrative, can be read as the film’s figure for a fixed domain that does not respect the local boundaries the characters take for granted. On a contingentist picture, the characters could in principle be insulated by domain shifts, the story could move to a place where certain entities are not even available. On a necessitist picture, the insulation is thinner. The background space of objects is already there, and what varies is which of those objects become concrete, which become accessible, which become entangled. The Mystery Man is then the personification of that unsettling thought: the world’s inventory is not just what happens to be on stage. 

The supervenience section then sharpens a different worry. Supervenience claims are popular in metaphysics because they promise tidiness. To say A supervenes on B is to say there cannot be a difference in A without some difference in B. People often say mental facts supervene on physical facts, or moral facts supervene on natural facts, or modal facts supervene on non modal facts. 

Williamson is suspicious of sweeping supervenience rhetoric here, not because he wants metaphysics to be messy for its own sake, but because the necessitist framework complicates what it would even mean for some class of facts to be the complete base. Once you allow a realm of non concrete objects, you have to decide whether the base includes facts about them. If you try to insist that the base is just what is concrete, you risk making modal truth float free of the base, because modal claims quantify over the fixed domain, including non concrete things. If you instead include them in the base, you have admitted that the base is not just the concrete, and the neat picture of supervenience as “everything rides on the concrete distribution” starts to look parochial. 

Here the film becomes interpretatively productive. Lost Highway is obsessed with the temptation to treat a certain layer as the base layer. Is the base the videotape evidence, the physical record, the thing that seems to show what happened? Or is the base the interior psychic narrative, the jealousy, the shame, the wish to be someone else? The film refuses to let either level be sovereign, because each seems to generate facts the other cannot comfortably determine. In Williamson’s terms, that is what it feels like when you try to keep modal structure supervenient on a thin concrete base. The story keeps producing “modal residue”, differences that do not track cleanly a chosen base. The switch from Fred to Pete is not presented as an ordinary physical transformation, but it is not presented as pure fantasy either. It is a difference that lives in the space between base layers, and the film’s anxiety is that there may be no base layer that is stable enough to do the supervenience work you want. 

Then comes his discussion of truthmakers. A truthmaker is supposed to be, roughly, a chunk of reality that makes a statement true. Truthmaker theorists often want a metaphysical discipline that forces every truth to be backed by an entity, event, or state of affairs. Williamson’s line is that this discipline is frequently misplaced. The necessitist picture helps him make the point because it heightens a mismatch between modal truths and ordinary truthmaking expectations. If it is true that there could have been a purple cow, what in the actual world makes that true. If you insist on a truthmaker in the actual concrete realm, you will be driven either to baroque entities or to denying perfectly ordinary modal truths. Williamson thinks the right response is to reject the demand, not to distort the ontology to satisfy it. 

Again, the film almost begs to be read through this. The tapes arrive. A crude truthmaker impulse says: the tape is the truthmaker, the tape makes it true that someone was in the house, that something happened. But the film steadily undermines that. The tape does not settle what the tape seems to show, because the showing itself is implicated in the story’s structure. If you look for a single concrete item that makes the relevant claims true, the item slips into the claim. The film trains you out of the thought that every truth has to have a neat object behind it, as if truth were always a matter of pointing. That is close to Williamson’s therapeutic aim with truthmakers in the modal case. Modal reality is not, on his view, a special occult layer, but nor is it something that always reduces to a conveniently localised truthmaking chunk. 

The last sub theme, “Contingency, change, and difference”, is where the film and necessitism start to lock together most tightly, because the film’s surface topic is precisely a kind of change that looks like difference in what there is. Fred seems to vanish, Pete seems to appear. If that is what happens, necessitism is in trouble. Williamson’s response is to insist that we have been too quick to equate contingency with existence variation. Contingency can be carried by changes in concreteness, instantiation, location, or qualitative profile, while existence remains fixed. You can say the coin could have failed to be a coin, or could have failed to be concrete, without saying it could have failed to be something. You can say the oak stick could have failed to be concrete without saying there would have been no object there for modal claims to concern, if you adopt the attributive reading of “possible stick” and the accompanying necessitist ontology. 

Think of the Fred Pete switch as the cinematic presentation of a distinction between an object and its mode of being concrete in a narrative world. On a necessitist reading, there is an underlying object, or perhaps a structured plurality of objects if you want to be cautious, that is capable of being concrete in more than one qualitatively saturated way. One way is Fred, embedded in a marriage, in suspicion, in the claustrophobic domesticity of the early scenes. Another way is Pete, embedded in a different social surface, different desires, a different relation to the femme fatale figure. The film refuses to let you treat these as simply two different people in two different worlds, because the story’s affect depends on their leaking into one another. That leakage is exactly what the fixed domain picture makes intelligible. It is not a leak between worlds with different inventories, it is the same inventory expressing different concreteness patterns. 

Williamson insists that the best support for necessitism is abductive, an inference to the best overall theory, not a single knockdown proof. Critics like Ted Sider present the debate as in part about simplicity and ideological economy, whether the necessitist’s ontological expansion is compensated by simpler principles. Williamson is explicit that multiplying entities can be a necessity for theoretical plausibility, because the alternative can be a massive loss of simplicity and elegance in principles. 

The film gives you a way of feeling what that trade off is like. The “ontological parsimony” version of the film would say: only what is on screen exists, only what is concrete in the current segment is in the domain. But that would force you to add extra machinery to explain the persistent cross segment identity pressures, the way names, guilt, desire, and accusation seem to retain grip even when the concrete cast appears to change. The necessitist version of the film is more lavish in background ontology, because it allows a stable domain of narrative objects that do not always manifest, but it buys you a cleaner account of why the story can coherently continue to be about the same jealousy, the same violence, the same self interpretation, even as the concrete profile shifts. This is where we have “necessitist applications” as interpretation rather than illustration, and the strongest interpretative move is to treat the film’s loop as an S5 shaped intuition pump. (Remember, S5 is a modal logic in which, very roughly, necessity and possibility behave as if the space of worlds is maximally connected. One slogan is that if something is possibly necessary, then it is necessary. The slogan is a structural claim about the accessibility relation between worlds.)

 Williamson is not naively projecting formal systems onto art, but the film’s loop has an S5 feel because it suggests that the modal space is not layered into separate regions. The narrative keeps returning, as if the possibility space is internally tight. The loop is not just repetition, it is a kind of modal closure. When the film circles back to the opening message, it feels less like time travel and more like a necessity emerging from possibility, as if the story could not escape its own structure. Under necessitism, that structural closure is not explained by the world containing only what it locally contains. It is explained by the underlying domain being there regardless, with the story’s local “actuality” being a matter of which parts of the domain become concrete together. The Mystery Man’s phone call, the uncanny simultaneity, becomes a cinematic way of indicating that the same object can be implicated in different concrete patterns without requiring two separate existences. 

The phone call collapses distance the way a modal operator collapses worlds, by shifting what counts as accessible without changing what exists. You can even push Williamson’s “possible F” ambiguity into the film’s central erotic and violent structure. The femme fatale figure, in her different presentations, is persistently treated as both predicatively and attributively “possible”. On the predicative pull, she is the same woman, just appearing under different conditions, and the narrative is about uncovering the same concrete person. 

On the attributive pull, she is whatever in the domain could have been that woman, a role that can be concretely occupied by different qualitative profiles while preserving a modal link to an underlying object. The film’s unease comes from not settling whether the character is a determinate concrete individual across the whole narrative, or an object whose identity is carried by modal role rather than by stable concrete continuity. 

Williamson’s framework legitimises the second kind of identity, not as a cheapening of individuality but as a consequence of separating existence from concreteness and allowing modal recombination to be literal rather than metaphorical. Williamson is not merely proposing an ontology, he is proposing a way of doing metaphysics that treats modal logic as a “structural core”. He presses the idea that you should assess metaphysical frameworks the way you assess scientific frameworks, by theoretical virtues like simplicity, explanatory integration, and resistance to ad hoc patches. 

In the context of Lost Highway, the comparable claim would be that you do not interpret the film by collecting “what literally happened” as if it were a police report, you interpret it by finding the simplest and most integrated structural account of why the film insists on certain transitions and forbids certain stabilisations. The film itself seems to punish ad hoc repairs. Every time you think you have patched the story with a local fix, a later scene makes the fix feel like a refusal to see the structure. 

So Williamson is about persuading you that a metaphysics that keeps changing the domain of quantification across worlds is like an interpretation that keeps changing what characters count as being in the story, and then acts surprised when it cannot state the most basic cross scene claims without extra apparatus. Necessitism, for Williamson, is the refusal of that apparatus when the apparatus is only compensating for an earlier decision to let the domain wobble. 

On a contingentist reading, Lost Highway is a film about ontological rupture, about the replacement of one existent by another. On a Williamsonian necessitist reading, it is a film about the instability of concreteness, about how a fixed inventory of objects can underwrite radical shifts in manifestation without any literal coming into existence or passing out of existence. The horror, and the seduction, is that the inventory is not under the characters’ control. They can try to flee into a different concrete life, but the domain they are fleeing within is still there, and the modal structure of what could have been, what would have been if, keeps its grip. That is why the loop feels like the world’s necessary framework asserting itself, the way Williamson thinks a good theory’s structural principles assert themselves when you try to do without them. 

By the end of Modal Logic as Metaphysics, Williamson is no longer merely saying that necessitism is coherent, or even that it is defensible. He is saying that, given a serious commitment to modal reasoning, identity, and unrestricted quantification, necessitism is the theory that lets you stop improvising repairs. It is the theory that allows you to say, “this is how the structure works”, and then live with the consequences. 

The film can now be read the same way, not as a riddle to be solved scene by scene, but as a world whose structure either stabilises under a necessitist ontology or fragments under a contingentist one. So consider two rival metaphysical readings of Lost Highway, now treated explicitly as theories rather than moods. 

On the contingentist reading, the film’s world is one in which existence itself is fragile and world bound. In one narrative segment, Fred exists. In another, he does not. Pete exists instead. Alice exists where Renee does not. The domain of objects changes as the story shifts. Each segment comes with its own inventory, and the transitions are ontological breaks. Under this reading, when Fred becomes Pete, something literally ceases to exist and something else literally comes into existence. The metaphysics here mirrors a variable domain semantics. Quantifiers range only over what is concrete in the current segment, and talk of “the same person” across segments is either metaphorical or illicit. This reading has an immediate appeal because it aligns with the phenomenology of rupture. It honours the shock of the prison transformation. It preserves the intuition that identity is anchored in concrete embodiment. It allows you to say, straightforwardly, “Fred is gone”, without qualification. 

But the cost emerges the moment you try to state anything general about the story. How do you explain the persistence of guilt, jealousy, and accusation if the bearer of those states literally ceases to exist? Why does Pete inherit Fred’s anxiety? Why do the same objects, the house, the saxophone, the videotape apparatus, keep functioning as if they belong to a single world rather than two unrelated ones? Why does the loop close, with the same sentence reappearing, if the ontology has been reset in between? To preserve those connections, the contingentist reading has to add auxiliary devices. Psychological continuity. Narrative projection. Dream logic. Symbolic substitution. Each device does local work, but the overall picture becomes a patchwork. The ontology fragments, and the explanatory burden shifts to metaphor. 

Now contrast the necessitist reading. On this reading, the film’s world has a fixed inventory of objects, persons, properties, and relations. That inventory does not change as the narrative unfolds. What changes is which of those objects are concrete, which are instantiated, which are embedded in spacetime, which are accessible to perception, which are socially recognised. Fred and Pete are not created or destroyed. They are objects whose concreteness status varies across the film’s modal structure. Renee and Alice are not ontological alternatives, one existing instead of the other. They are distinct objects or distinct modes of one object whose concrete instantiation is governed by the story’s shifting constraints. The Mystery Man is not an exception to ontology. He is an index of it. He behaves as if the domain never shrinks because, on this reading, it does not. 

Once you adopt this framework, several things fall into place at once. First, the loop stops being mysterious. The return of the opening line is not a paradoxical reset. It is the closure of a modal cycle. The world does not have to create new entities to restart the story. It recombines what is already there. The sense of inevitability comes from the fact that the modal space is internally constrained. Certain configurations of concreteness lead, necessarily, to others. The highway is “lost” not because it leads nowhere, but because it is not anchored to any single concrete trajectory. It is a path through a fixed space of possibilities. 

Second, identity ceases to wobble. It becomes rigid in Williamson’s sense, even when manifestation is not. The question “is Pete Fred?” is no longer the right question. The right question is “which object is concrete here, and under which qualitative profile?”. The film’s refusal to give you a definitive answer to identity puzzles starts looking like fidelity to a deeper structure. Identity is not determined by narrative role alone. It is determined by which object the quantifiers pick out, and the quantifiers, on a necessitist view, always range over the same domain. 

Third, the film’s obsession with surveillance makes metaphysical sense. Surveillance is a model of unrestricted quantification. The camera sees regardless of what the characters acknowledge. The tape arrives regardless of consent. The Mystery Man is already at the house because the house is not the limit of what exists. In Williamson’s terms, the error of the characters is to confuse concreteness with existence, to think that what they can see or inhabit exhausts what there is. The horror is the discovery that it does not. 

Williamson’s rejection of truthmaker maximalism, his scepticism about global supervenience on the concrete, and his defence of a necessary framework of objects all converge on this. Reality does not owe us a base layer that coincides with what is manifest. Modal truths do not wait for concrete anchors. They are licensed by the structure of the domain itself. 

In the film, the tape does not need Fred’s acknowledgment to be true. In the metaphysics, the possibility that Fred could have been otherwise does not need a concrete surrogate in the actual world to be meaningful. Seen this way, necessitism is a refusal to let ontology be hostage to perspective. It insists that “there is” does not mean “there is here”, “there is now”, or “there is in this story segment”. It means “there is”, full stop. Everything else, including concreteness, actuality, presence, embodiment, is downstream. 

That insistence is what allows Williamson to dismantle a certain style of metaphysical reassurance. The reassurance says: do not worry, even if modal logic looks strong, even if Barcan principles look tempting, we can always retreat to a picture where only the concrete exists, and the rest is only talk. Williamson’s says once you accept the expressive power of quantified modal logic, once you let higher order resources do their work, once you insist on unrestricted generality, the retreat is a distortion. 

The same is true of Lost Highway. You can retreat to a psychological reading, or a dream reading, or a symbolic reading, and many viewers do. Those readings are retreats. They let you keep your ordinary ontology intact at the cost of treating the film’s structure as a kind of noise. The necessitist reading, by contrast, treats the structure as the signal. It says: the film is about the instability of concreteness within a fixed ontological field. The terror is not that something might cease to exist, but that it might cease to be the thing you thought you were, while still being. Williamson is offering necessitism as the picture that remains when you stop insulating existence from modality. He is explicit that this picture will offend some intuitions. But his wager, stated most clearly at the end of his book, is that metaphysics should not be intuition management. It should be theory construction under constraint. If the best overall theory tells us that the domain is fixed and that objects exist necessarily, then we should accept that and revise our understanding of contingency accordingly. And that revision is precisely what the film demands. 

The ordinary understanding of contingency says: I could have been someone else, or I might not have been at all. The film’s understanding says something colder. You are what you are because this configuration of concreteness has been selected, not because the alternatives were unreal. The alternatives are there, pressing, waiting, capable of being instantiated. Escape is not annihilation, it is recombination. 

In that sense, Lost Highway is one of the clearest aesthetic representations of what a necessitist universe would feel like from the inside. A universe in which the inventory never shrinks. A universe in which guilt does not disappear when you flee. A universe in which the camera keeps filming because the domain keeps ranging. A universe in which the most frightening thought is not that you might never have existed, but that existence is not optional, only its modes are. Lost Highway ends with a scream, a chase, and a loop. But the scream is the same methodological scream as Williamson’s moral. It is the sound of a perspective discovering that it is not sovereign. 

Where contingentism reads the film as a story about things coming into and going out of existence, necessitism reads it as a story about existence being inescapable, and about the terror of living inside a world where only the distribution of concreteness changes. That is exactly the terror Williamson thinks we should be prepared to face, if we are serious about modal metaphysics. And in Lost Highway Lynch seems to be presenting an aesthetic staging of his ideas. 

2. Fine

But lest we think we have landed the Lynchean world a clean necissitist metaphysics, I need to immediately pivot to another approach because it seems to provide equally compelling fruitfulness. This is an approach I attribute to Kit Fine and is what I’ve used in other efforts to grasp Lynch’s work. So I need to compare and contrast how Fine and Williamson make sense of the Lynch films and judge them by their interpretative fruits. To do this I’ll need to quickly summarise what I’ve made of Williamson and a Williamsonian reading of Lost Highway so far, which will mean I’ll be repeating myself somewhat. You can skip the next few paragraphs if you think you’ve heard enough of that Williamsonian necessitist interpretation. 

So, I think to get the comparison to land, it helps to keep three things in view at once, and to let Lost Highway do some of the work. First, Lynch gives you a world in which identity looks both rigid and fluid, rigid because there is a single suffocating knot of guilt and desire that keeps returning, fluid because the film seems to permit a person to “become” someone else without any clean explanatory bridge. Second, Williamson’s modal metaphysics is built to respect a very demanding kind of objectivity about modality, and to do it with the same sorts of theoretical virtues we use elsewhere, simplicity, explanatory power, systematic fit. Third, Fine’s modal metaphysics is built to resist a particular picture of modality, roughly, that modality is best understood by quantifying over a space of complete alternative worlds and then treating the modal facts as what comes out true across that space. 

So here we go. Fine’s alternative centres essence, dependence, and a sharp sense that the actual has a certain primacy over merely possible constructions. Once you hold those together, you can see why they sometimes look like allies and sometimes like they are reading two different films. Williamson’s starting posture is that modality is a respectable subject matter, metaphysical necessity and possibility are among the things we can theorise about, and the right way to proceed is in the same abductive spirit as elsewhere, we try to build the best overall theory of the modal generalisations we care about, with formal tools that let us test consequences sharply. In Modal Logic as Metaphysics he is explicit that the project is metaphysical, with the modal operators read as metaphysical necessity and possibility and the quantifiers read as unrestricted, not tied to what we happen to be able to name or conceive. That “unrestricted quantification” piece is as we've seen one of the hinges on which the necessitism argument turns. Necessitism, in Williamson’s compressed slogan, is that ontology is necessary: necessarily, everything is necessarily something. In longer form, it is necessary that for everything x there is something y such that necessarily y is identical with x. 

The point is not that every ordinary kind is instantiated in every world, Williamson is happy to say it is contingent that there are animals, or saxophonists, or videotapes. The point is that the domain of quantification, what there is in the most general sense, does not fluctuate across possibilities. What fluctuates is which things are animals, which things are saxophonists, which things stand in which relations, which things have which properties. 

Lost Highway tests how you think about “what there is” versus “what is the case”. The film keeps returning to the thought that something has happened, something is fixed, “Dick Laurent is dead”, and yet the narrative seems to slide into a different register where the cast of individuals appears to change. On one common line, Pete Dayton is a fantasy construction, a dissociative escape hatch for Fred Madison, in which the same underlying reality is reconfigured as a different life. Lynch himself talked about it as a “psychogenic fugue” and there are recurring interpretations discussing its Möbius like structure. If you take that seriously, you have a natural Williamsonian temptation: do not treat the shift from Fred to Pete as a shift in what exists, treat it as a shift in which properties and roles are instantiated, and in which descriptions apply, while the underlying ontology is held fixed. That is, in Williamson’s idiom, you should be suspicious of treating apparent existential variation as deep ontological variation. 

The contingentist wants to say: there could have been fewer things, some things might not have existed. Williamson pushes back with a battery of considerations, some semantic, some methodological, some to do with how higher order resources let you state the relevant commitments cleanly, and some to do with explanatory challenges for contingentism. You do not need the technical details here to see the interpretive moral for the film: Williamson wants you to avoid building a metaphysics that makes existence itself too cheap, too easily toggled. He would prefer to locate the drama in the pattern of instantiation and explanation, not in domains popping in and out. 

Now notice what happens if you read Lost Highway with that pressure in mind. The film seems to offer two “worlds”, the Fred world and the Pete world, but the film also insists on a set of trans world like anchors, the Mystery Man appears in both segments, the video evidence crosses the divide, the names and faces rhyme, Renee and Alice have an eerie near identity, and the whole story curls back on itself. A Williamson style metaphysician is naturally drawn to a unifying account: the ontology is fixed, what changes is which guise is actualised, which narrative is selected, which properties are salient, perhaps even which “presentation” of an object is available to cognition. The Mystery Man’s line about being at Fred’s house when he is also at the party is almost an advert for a certain kind of modal and intensional discomfort, the same individual in two locations is impossible in ordinary space time terms, but the film makes it feel metabolically true. 

At this point you can see both the overlap with Fine and the beginning of the divergence. Fine shares with Williamson a refusal to treat modality as mere verbal hygiene. Both think there are real modal facts to be got right, and both are willing to criticise Quinean suspicion about de re modality, modality about things, not just about sentences. Fine is explicitly associated with defending the intelligibility of de re modality and with taking modal notions as in some sense primitive or not straightforwardly reducible. 

He is also a key figure in the movement sometimes described as “new actualism”, which tries to do justice to modality without reifying Lewis style possible worlds as concrete places. Barbara Vetter’s survey of “modality without possible worlds” frames Fine as central to this reorientation. So you might think Williamson and Fine are simply travelling companions: both are serious modal realists in the broad sense, both reject crude deflationism, both engage the formal apparatus with sophistication. 

But Fine’s deepest instinct is to resist the thought that possible worlds semantics, especially in its standard shape, should be doing the metaphysical leadership. He thinks the tail often wags the dog. In particular, Fine’s work on essence argues that what is necessary about a thing is often grounded in what it is, in its essence, rather than in what holds across an antecedently given space of worlds. 

This matters immediately for Lost Highway, because the film keeps teasing you with the question: are Fred and Pete one and the same, or are they distinct, or is one an “unreal” construction that nevertheless has narrative agency? A Williamson leaning reading says: treat the quantifiers as unrestricted, take identity seriously, and be cautious about multiplying entities. If you are tempted to say “Pete does not exist in the first half”, Williamson will want to hear why that is better than saying “there is such a thing as Pete, but in that segment he is not instantiated as the salient guise”, or perhaps “there is such a thing as the individual who is in fact Pete in the second half, but in the first half he is not identical with anyone visible to us”, because on necessitism the domain is fixed even when the story is not. 

You can translate that into a filmic register: the film’s horror is not that new beings blink into existence, but that the same being’s route through guilt and desire can force a reconfiguration of self description so radical that it looks like a birth of a new person. Lynch even gives you birth imagery and headache collapse around the prison transformation, as if to mock the idea that a new entity has arrived while simultaneously inviting it. 

Fine will be more wary of that kind of unifying metaphysics if it bulldozes essence. Here is the basic Fine shaped objection, stated without technical machinery. If Fred is essentially Fred, if being Fred is not just a role he happens to play but part of what he is, then he cannot literally be Pete, any more than a particular number could literally be a particular colour. The film encourages exactly this intuition by making Pete not merely a renamed Fred but a different sexual economy, a different social embedding, a different age coded embodiment. The point is not just psychological, it is ontological in the sense that the film treats the difference as difference in what the subject is like at the level of kind and powers. If you accept an essence centred approach, you will resist the claim that one and the same object is, across possibilities or across segments, both essentially Fred and essentially Pete. 

At the same time, Fine’s way of resisting does not have to be “so Pete is a different possible world person”. Fine has alternatives. Fine is famous for distinguishing necessity from essential necessity, and for arguing that essence is not reducible to modality, it can ground it. That gives you an interpretation that can say: perhaps Pete is not identical with Fred, but Pete is an essentially dependent projection, an “ersatz” persona whose existence and character is grounded in Fred’s fractured agency. In that case the film’s Möbius structure is not best captured by saying “in another world Fred is Pete”, but by saying “within the actual’s structure of dependence, there is an entity or role that is generated by Fred’s attempt to evade an essential fact about himself”. 

This kind of move is very Fine like: you keep the metaphysics close to the internal natures and dependence relations, rather than letting a global space of worlds decide what is possible. Now we can make the comparison more explicit by using Lost Highway as a test case for three classic pressure points in modal metaphysics: the status of possibilia, the logic of quantification in modal contexts, and the explanatory direction between essence and modality. 

On possibilia, Williamson’s necessitism is, in one sense, aggressively generous: everything exists necessarily, so there are no “merely possible” entities in the sense of entities that exist in some worlds but not others. That does not mean Williamson thinks every fictional character is real, or that every description picks out something. But it does mean that if there could have been a child of Fred and Renee, then there is something that is that child in the relevant possibilities, and what varies is whether it is a child, not whether it exists. 

Fine, by contrast, often insists on the primacy of the actual over the possible, and is suspicious of inflating our ontology by taking the space of possibilities as ontologically prior. If you bring that to Lost Highway, Williamson is more comfortable saying “there is such a thing as the Pete individual even ‘before’ he shows up”, because “before” is not the right metaphysical category here. Fine is more comfortable saying “do not treat Pete as an already there entity waiting off screen, treat Pete as a dependent construction, perhaps grounded in Fred’s actual psychological and normative situation”. Both readings give a different metaphysical feel to the uncanny. 

Williamson’s uncanny is about stable being under unstable predication. Fine’s uncanny is about the way actuality can generate layered structures of dependence that mimic alternative lives without literally realising them as independent entities. On modal quantification, Williamson takes seriously the Barcan style issues, roughly, how quantifiers interact with modal operators, and for arguing that once you take the quantifiers as unrestricted, much of the resistance to certain principles is misplaced. (roughly, S5 says that if something is possibly necessary, then it is necessary.) You do not need to know the formulas to feel the connection: Lost Highway constantly forces you to ask whether “there exists someone who could have been Fred instead” makes sense, and whether the existence claim is inside the “could have” or outside it. When Fred says, in effect, “I like to remember things my own way”, the film is staging a dispute about which existential commitments are built into a story and which are introduced by our way of describing it. 

Williamson’s style is to regiment these disputes so that we can see what follows from what, and then to choose the best overall package. Fine’s style is to warn that regimentation can mislead if it forces all modal talk into one homogeneous operator plus quantifier shape. He is attentive to differences among modalities, logical, metaphysical, normative, and to the thought that essence adds a layer that is not captured by plain modal operators. 

In film terms, Williamson is tempted to say, “let us decide whether Pete exists necessarily and whether Fred could be Pete, then the rest will be bookkeeping”. Fine is tempted to say, “you are flattening the difference between what Fred could have been given his nature, what the narrative can depict given cinematic licence, and what is metaphysically coherent given essences”. Williamson often treats necessitism as part of a best overall theory of modality and logic, an abductive winner. He is not trying to deduce it from a single intuition but to show it integrates better with a theoretically virtuous package. Fine treats essence as explanatory bedrock for many necessities. 

This yields the most vivid split in their preferred Lost Highway readings. A Williamson leaning reading of Lost Highway takes the spiral seriously. The film’s loop is not a mere narrative trick, it is a metaphysical hint that the space of ways things could be is structured by deep necessities about identity, agency, and reidentification. The Mystery Man functions like an operator that forces collapse of subjective perspective into objective fact, he is at your house now, you cannot keep the “could be” safely quarantined. 

In this spirit, the Pete episode is not the introduction of a new being but a reallocation of properties across an already fixed domain. Fred’s attempt to run from an actuality, the murder and what it means about him, generates a modal illusion, as if there were a nearby possibility in which he is different. But on necessitism, there is no ontological escape route: the “different life” is not secured by absence of entities, it is at most secured by different predications of the same entities. The horror is precisely that the world is ontologically inescapable. If “Dick Laurent is dead” is a fact, then the domain is already stocked with everything required for that fact to be unfolded, even if Fred refuses to remember it. 

A Fine leaning reading treats the film as an essay in essential constraint and dependent projection. Fred cannot simply be Pete because the shift is too deep to be mere contingent variation in properties. Instead, Pete is best understood as an essentially dependent persona, a role generated by Fred’s attempt to reconstitute himself under a different description, and the film’s Möbius shape is a dependence structure, not a tour through a space of alternative worlds. Fine’s emphasis on essence gives you a clean way to express what the film makes you feel, certain things are not negotiable, not because of causal constraint but because of what the characters are. 

Renee and Alice, for example, are not “the same woman in two worlds” on this view, they are two nodes in a structure where one is a reconfiguration of the other under fantasy, and the dependence runs through Fred’s agency and desire rather than through a global modal space. This is very close to how many interpretations treat the film’s women as obsessions or projections rather than as independent individuals in a realist sense, while still allowing you to treat the projection as a real part of the actual psychological economy. 

Notice the overlap. Both readings are, in their own way, anti Lewisian in spirit. Williamson is not embracing concrete possible worlds, he is using possible worlds semantics as a tool while defending necessitism about the domain. Fine is explicitly central to the project of doing modality without possible worlds as metaphysical primitives. Both also give you a principled way to respect Lynch’s refusal to settle the plot into a single ordinary explanation. Lynch wants the audience to live with an abstract structure about identity, and both Williamson and Fine think modality and identity are the right register for that. 

Where they come apart is what they think the best metaphysical explanation of that structure looks like. Williamson’s package makes it attractive to say that the film is dramatising a kind of modal collapse, a slide from seeming contingency to necessity. Fred behaves as if there is a nearby possibility where he is innocent, virile, untroubled, and the film shows that this “possibility” is not a genuine alternative but a reshuffling inside a fixed domain that cannot remove the underlying necessities about what has happened and who he is. The spiral that comes around “a little bit higher” becomes, on a Williamson reading, a metaphor for abductive pressure, your best overall theory must accommodate the returns, the stable anchors, the inescapable quantificational commitments. 

Fine’s package makes it attractive to say that the film is dramatising essential impossibility rather than mere unactualised possibility. Fred’s fantasy does not show “what could have been”, it shows what cannot be, because it violates the essential structure of the actual agents involved. That is why the fantasy curdles. On this reading, the Mystery Man is not merely the enforcer of fact but the emissary of essence, the part of the structure that makes it impossible for Fred to become the kind of thing he wants to be, while still letting him generate dependent simulacra of that life. 

Fine’s own work on essence and modality is the philosophical analogue of this thought, it is often not enough to say “in some possible world”, you need to ask whether the thing in question could have been otherwise given what it is. Williamson is more willing than Fine to treat the film’s identity shifts as compatible with strict identity across the underlying metaphysical domain, because his necessitism and his methodological preference for overall theoretical virtue encourage a unifying ontology. When he sees the same motifs reappearing across the Fred and Pete segments, he is pushed toward the thought that the best explanation preserves a single stock of entities and treats the variation as variation in properties, roles, and epistemic access. That makes the film a parable about how ontology stays fixed while our ways of classifying and remembering can become radically unstable. 

Fine on the other hand is more willing than Williamson to say that strict identity is the wrong relation to force on the Fred and Pete divide, because essence forbids certain identifications. He will therefore look for other relations, dependence, projection, role embodiment, perhaps something like an arbitrary object, a schematic placeholder that can be filled in different ways by the actual. What matters is the direction of explanation, he will explain modal and identity-like appearances by the essences and dependence relations of the actual, rather than by treating the essences as derivative from what is true in the best modal logic. 

Think about the highway itself in the film. Williamson reads it like a single domain with different routes, the cars can take different turns, but the map is there regardless, and the most general inventory does not change. Fine reads it like a structure of lanes that some vehicles cannot enter because of what they are, and when they seem to enter, it is because a dependent image has been generated, a cinematic lane that looks drivable until it collapses back into the actual. 

Both readings are powerful, but they put the metaphysical weight in different places. Williamson makes Lost Highway feel like a study in the inescapability of being, and the way modal reasoning, properly regimented, tells against fantasies of ontological disappearance. Fine makes it feel like a study in the inescapability of what something is, and the way the actual can generate shadow lives that are not genuine alternatives but dependent distortions. For a Williamson style reading , when Fred “becomes” Pete, the film’s metaphysical suggestion is not that a new entity has been created out of nothing, nor that an entity has ceased to be. Rather, it is that the film is showing you a world, or a scenario, in which different parts of a fixed domain of entities are “lit up” as concrete occupants of the story’s foreground. The film is an experiment in what it is like for the actual to be a moving spotlight over a necessarily existing cast. 

The Mystery Man’s unnerving omnipresence, his capacity to be “at your house” and “here at the party” in the same breath, even starts to look like a grotesque dramatisation of what necessitists often say , that domain membership is not itself a local, time bound fact, and that the metaphysical framework is not rearranged by ordinary events. 

Kit Fine, by contrast, is one of the most influential sources of resistance to the idea that modality should be read off from possible worlds machinery, or from a smooth, extensional semantics in which all that matters are truth values across worlds. Fine’s approach is sometimes described as hyperintensional and essence first. 

“Hyperintensional” here means that two claims can be necessarily equivalent, true in exactly the same possible worlds, and yet still differ in a way that matters for explanation, for grounding, for what is essential to what (Louis Lane knows Superman can fly but doesn't believe she can escape through the top floor window with Clark Kent and survive, because she doesn't realise Clark is Superman. You couldn't explain the situation if all you took into consideration is the identity relation between Clark and Superman). Fine’s argument is to show that standard modal tools cannot capture the fine structure of metaphysical dependence, because necessity is too coarse grained. His essay “Essence and Modality” is the canonical locus, where he argues that essence cannot be reduced to modality, and that the direction of explanation often goes the other way, the modal facts are to be grounded in essential facts 

This creates an immediate contrast with Williamson’s kind of system building. Williamson is willing to let the best overall modal logic, chosen partly for theoretical virtues, exert pressure on ontology. Fine is much more likely to say that even a perfect logic of necessity and possibility leaves out the structure that metaphysics most cares about, namely what something is, in virtue of what, and how that internal character constrains and explains its modal profile. 

That does not mean Fine is a simple contingentist in the ordinary sense, because Fine’s view is not just “some things might not have existed”. It is rather that existence, identity, and possibility must be articulated through a more discriminating metaphysical vocabulary than bare quantification across worlds, and that standard possible worlds talk risks confusing a representational device with the source of modality itself. 

A Williamson inflected interpretation, as I just sketched, will tend to treat the film’s transformations as switches of concreteness or actuality within a fixed ontological domain. A Fine inflected interpretation will tend to treat them as shifts in essence attribution, shifts in what is taken to be constitutive of the central subject, and therefore shifts in which modal claims about that subject are even well formed. Put more simply, Williamson encourages the thought that the film is showing you how the same world can have different “profiles” while keeping the same underlying furniture. Fine encourages the thought that the film is showing you how different profiles correspond to different candidates for what the subject even is, because the film is playing directly with criteria of identity, not merely with occupancy of an already fixed domain. 

Here the Fred, Pete, Renee, Alice quadrilateral is decisive. The film invites you to treat Renee and Alice as the same woman under different guises, and also invites you to treat them as importantly different, because each comes with a different erotic and moral economy, and a different set of narrative permissions. On a worlds based modal reading, you might be tempted to say, there are worlds in which she is Renee and worlds in which she is Alice. Fine will press you to ask, are we holding fixed what matters to identity when we do that, or are we illicitly sliding between two different essence claims, one about a wife in a claustrophobic domestic scene, another about a femme fatale who functions as a lure into a different criminal reality? If the essence of the character is tied to her role in an economy of guilt and desire, then “Renee” and “Alice” may not be two world indexed presentations of the same object, but two different objects, or better, two different identity candidates, generated by the film’s own internal rules of constitution. Fine’s central complaint against some modal metaphysics is that it pretends you can do identity first and essence later, and he thinks the order should be reversed. 

This is also where Williamson and Fine overlap more than it might first seem. Both resist a crude linguistic quietism that says modal disputes are merely verbal. Both treat modality as substantive, and as something that should not be reduced to talk about “ways the world could have been” without metaphysical cost. Fine, in particular, defends the intelligibility of de re modality, claims about what a thing could or must be, not merely about de dicto modality, what a sentence could or must be. Williamson likewise treats quantified modal logic as a serious descriptive and normative enterprise, and he is willing to use technical results as constraints on metaphysical options. 

In terms of Lost Highway, both will reject the lazy response that the film is “just” ambiguous, therefore any modal reading is arbitrary. Both will instead see the film’s systematic ambiguities as evidence that it is staging a real metaphysical problem, what identity through possibility could even mean when the criteria of identity are under strain. But the places they come apart map very neatly on to two rival readings of the film. 

One reading, call it the fixed domain reading, says that the cast of entities is stable, and the film’s horror is the mobility of perspective, access, and actuality. The message “Dick Laurent is dead” arrives like a rigid fact that seems to loop, as if necessity itself has become a trap, the end is built into the beginning. The Mystery Man’s camera produces a kind of external, object level record that Fred cannot domesticate into memory, the tapes show the house, then the bedroom, then the murder scene, as if the world’s facts outrun the subject’s self narration. The key metaphysical effect is that there is a determinate reality, with a determinate inventory, and the subject’s transformations are not ontological miracles but re assignments of the role of “the actual” within a stable modal space. That is recognisably Williamsonian in spirit, even if the film does it with dread and neon rather than with axioms. 

The other reading, call it the essence shift reading, says that the cast of entities is not stable in the right way, because the film is not merely switching what is foregrounded, it is switching what counts as constitutive. Fred’s problem is not just that he is in an odd world, it is that he cannot sustain an essence for himself that can bear the facts, so the film generates a different identity candidate, Pete, with different constitutive traits, different social placement, different erotic competence, different narrative affordances. The shift is then not a journey to another possible world containing the same objects, but a reconstitution of the subject, and with that, a reconstitution of the space of possibilities that are live. 

In Fine’s terms, you should not expect standard necessity, truth in all possible worlds, to capture this, because the film is precisely about how the “same” necessities can be carried by different constitutive stories, and how those stories determine what counts as possible for the subject. That is the Fine style insistence that modal profile is grounded in what things are, not merely in an extensional distribution across worlds. 

You can see the same split in how each would treat the film’s looping structure. Williamson’s necessitism is often paired with a taste for S5 style modal principles, where the accessibility relation between worlds is, in effect, very generous, and necessity behaves with a kind of stability across possibilities. One intuitive upshot is that the modal space has a highly regular geometry. This sits well with Lost Highway’s sense that the narrative is on rails, the beginning line returns at the end, and the end seems already implicit at the start. If you let that geometry take interpretive control, the loop is not merely psychological repetition, it is a metaphysical structure, the space of possibilities for Fred is fixed in advance, and what the film shows is not that the ontology changes but that the subject is driven along a path that reveals what was always in the modal space. 

Fine will be much more suspicious of moving too quickly from a narrative loop to a metaphysical loop. For him the interesting question is what grounds what. If the loop is real in the story, what in the story makes it real, and is that “maker” itself a metaphysically basic fact, an essential feature of the subject, or a constructed artefact of some representational system. Fine is drawn to grounding talk and to distinctions between different kinds of necessity, for example logical necessity, metaphysical necessity, and necessities grounded in essence. Those distinctions are not always front and centre in the popular reception of his work, but they are part of the general Fine orientation, that modal vocabulary is too undifferentiated to carry all the metaphysical weight we want it to carry. 

For Lost Highway, that means he will read the loop less as an S5 like global regularity, and more as a symptom of an essential constraint on the subject’s self constitution, a constraint that forces certain recombinations and displacements of identity whenever guilt becomes explicit. This is also where their readings of the Mystery Man would tend to diverge sharply. A Williamsonish reading can treat him as an invariant across the film’s shifts, something like an operator that is always available, a function that maps each “world” segment to its record, its tape, its external witness. The Mystery Man then becomes the film’s dramatisation of a necessity operator, the thing that does not vary as the subject varies, the condition of possibility of there being a determinate fact of the matter about what happens. 

A Fineish reading is more likely to treat the Mystery Man as constitutive rather than invariant, not a mere cross world constant but an internal ingredient in the essence of the subject as split, a figure whose role is to enforce the film’s rules of re identification, and therefore to police what counts as the same person across the transformation. Those are different metaphysical roles, and you feel the difference interpretively. On the first, the Mystery Man is an ontological anchor. On the second, he is a constitutional mechanism. 

It is worth making one overlap between the philosophical approaches explicit that is easy to miss. Both Williamson and Fine are, in different ways, anti deflationary about modality. I've said that before but it's worth saying it again. Williamson thinks modal claims can be known and theorised in disciplined ways, and he is open to robustly metaphysical conclusions like necessitism, partly because he thinks theoretical choice in logic has teeth. Fine also thinks modal facts are not mere shadows of language, and he is willing to posit a deep metaphysical structure of essence and dependence to explain them. So both can treat Lost Highway as a film whose basic imaginative labour is metaphysical, it is trying to show you what it is like for the space of possibilities to be organised, constrained, and policed, not merely what it is like to be confused. 

Where they really come apart is that Williamson is willing to let a unified modal logic, plus theoretical virtues, dictate the shape of metaphysical space, whereas Fine is more willing to fracture the space into different explanatory layers, where necessity and possibility are downstream of essence, and where sameness across possibilities is not a primitive given but a metaphysical achievement that depends on what a thing is. 

That difference yields two distinct readings of the Fred, Pete relation. On the Williamson side, you can push a strikingly strong interpretive thesis, Fred and Pete are distinct necessarily existing individuals, and what the film shows is a shift in which individual is concrete in the relevant segment, perhaps induced by some lawlike constraint of the narrative world, or by a psychological mechanism that is itself a concrete realiser of a deeper metaphysical constraint. The film’s trick is then to tempt you into thinking that something has come into existence, or gone out of existence, when really you are watching a fixed domain under shifting concreteness. That is exactly the kind of illusion necessitists say contingentists are prone to, they reify the difference between being actual and being in the domain of quantification. 

On the Fine side, you can push an equally strong but different thesis, the film is refusing to settle the essential facts that would make “Fred is Pete” either true or false under a single criterion of identity, and the refusal is the point. The film forces you to experience, not just a puzzle about which possible world you are in, but a breakdown in what would count as re identification across worlds. When Pete finds the photograph, when the picture fails to cohere, when Alice vanishes from it, the film is not merely changing an empirical detail, it is changing what would ground the identity claims you are trying to make. That is a Fine style complaint about possible worlds talk, it can leave you thinking the hard work is done once you say “in some world”, but the real work is specifying the constitutive facts that make it the same object in those different cases. 

Williamson’s moral is that the film’s horror is the horror of a fixed ontology that you cannot escape, combined with a volatile condition of concreteness that you cannot control. The modal space is there, regardless of what the subject can bear, and the subject’s attempted escape, the Pete segment, is not an ontological creation but a reallocation within that space. Even the looping line “Dick Laurent is dead” can be read as the film’s way of telling you that the world’s inventory and its fundamental constraints do not come and go with your stories about them. 

Fine’s moral is that the film’s horror is the horror of essence under pressure, the subject cannot keep a stable constitutive profile, so the space of possibilities splinters, not because reality is many worlds, but because the grounds of modal truth are being reconfigured. The film does not just show that there are other ways things could have been. It shows that what counts as “the same thing” across those ways is itself a metaphysical problem, and one that no amount of extensional possible worlds bookkeeping can resolve without extra, essence laden structure. 

If you take Williamson’s necessitism seriously as an interpretive constraint, you get a distinctive reading of Lynch’s use of disappearance. Alice’s line, “You’ll never have me”, followed by disappearance, is typically read psychologically, as the impossibility of possession, or the collapse of fantasy into guilt. On necessitism, you can add something colder and stranger, disappearance is not non existence, it is a transition from concreteness to non concreteness relative to the story’s current modality. The film’s erotics then become metaphysical, desire reaches for the concrete, but the domain contains more than the concrete, and the subject’s punishment is to be forced to confront that the object of desire is not annihilated, it is simply withdrawn from the concrete register in which possession could make sense. 

Fine can also do something with this scene, but his version would treat the disappearance as a signal that the film refuses to supply the constitutive grounds needed to identify Alice with Renee under a single essence, so the desire claim fails because its target is not metaphysically well stabilised. Same scene, radically different metaphysical diagnosis. 

I think we need to look more closely at what Fine's metaphysics is up to. Fine’s collected volume Modality and Tense is a set of essays rather than a single monograph, but it is unified by two large commitments. Fine is against empiricist reductions of modality, attempts to replace “could” and “must” with talk of regularities or with Lewis style quantification over a pluriverse. Fine is also against modal monism, the Kripke inspired tendency to treat metaphysical possibility as one primitive notion to which all other modalities reduce. Fine wants a variety of modal notions, and he wants modality to be irreducible, but also more articulated than the standard possible worlds package tends to suggest. If you translate that into the cinema register, Fine is the sort of theorist who will tell you that Lost Highway is not just trading in one modality, metaphysical possibility, but in several distinct kinds of “could have been”, and that collapsing them will hide what the film is doing. There is something like epistemic modality in the first movement, the uncertainty Fred has about what is happening. There is something like psychological modality, the fantasy space in which Pete lives. There is something like normative modality, the sense of guilt and punishment staged by the Mystery Man. And there is something like metaphysical modality, the way the film itself seems to allow identity and time to fold. 

But Fine is most famous, in this neighbourhood, for insisting that essence is not to be reduced to modality. The common Kripkean thought is that to be essential to an object is, roughly, to be necessary for it, it has the property in every possible world in which it exists. Fine argues that this assimilation is misguided. The classic shape of his argument is that there can be two necessarily equivalent truths, both true in all possible worlds, such that one is essential to an object and the other is not. Williamson’s later paper summarises the canonical example and then targets it, but the summary itself nicely captures what made Fine’s challenge so influential, essence seems to cut finer than necessity. 

Why does this matter for our comparison, and for rival readings of Lost Highway? Because Williamson’s modal metaphysics wants a close fit between the modal logic and the metaphysics, while Fine wants to say that even if modal logic is in good shape, there are metaphysical joints it cannot see, because it is not hyperintensional enough.  

“Hyperintensional” is a technical term worth glossing carefully. An intensional context is one where you cannot always replace one expression with another that has the same extension, the same reference, without changing truth. For example, “Lois believes that Superman can fly” may be true while “Lois believes that Clark Kent can fly” is false, even though Superman is Clark Kent. A hyperintensional context is even finer grained. It can distinguish between sentences that are not merely co referential, but necessarily equivalent. If “A” and “B” are necessary equivalents, true in the same possible worlds, standard possible worlds semantics treats them as having the same modal content. Fine thinks metaphysics sometimes needs distinctions finer than that, distinctions grounded in how truths hold in virtue of the natures of things. 

In Lost Highway, the hyperintensional temptation is everywhere. “Renee” and “Alice” might be, in some sense, necessarily the same woman in the film’s world, because the casting and the narrative insist on their identity, yet the film also insists that being Renee is not the same as being Alice. Not in the sense of mere costume, but in the sense of what each figure is for, what relations define them, what role they play in a structure of desire and guilt. Likewise, “Fred is Fred” and “Fred is the one who appears as Pete” might be necessarily equivalent given the film’s arc, yet the film wants to make them feel metaphysically different. The question becomes: are those differences merely differences in our representation, or are they differences in reality, in essence? 

Here the two philosophers begin to separate in a way that can be stated cleanly, even if the downstream consequences are subtle. Williamson’s preferred strategy is to treat many fine grained distinctions as epistemic or representational rather than metaphysical, 'dodgy data' built out of pragmatic heuristics. If two propositions are necessary equivalents, he is suspicious of building a metaphysical theory that insists they come apart in the furniture of the world, because he expects our sense that they come apart may be produced by heuristics about how we grasp identity, reference, and composition. 

That is exactly the line he takes against Fine style arguments for hyperintensional essence, warning that we may be mistaking a cognitive shortcut for a metaphysical datum. Fine’s preferred strategy, by contrast, is to treat many of those distinctions as real distinctions, because he thinks talk of what holds “in virtue of the nature” of an object is not captured by possible worlds necessity. Fine is pushing for strict separations between issues, and for a repertoire of modal notions rather than one. In that spirit, Fine will be readier to say that the film’s identity play is not merely a trick of shifting concreteness assignments within a fixed domain, but an exploration of different ways an individual’s essence can be expressed, concealed, or reconfigured, and that some of those “could” claims concern natures rather than worlds. 

Now, one might think this sets them up as opponents about necessitism itself, but the situation is more interesting. In Williamson’s published reply to Fine, he explicitly reports that Fine thinks necessitism is obviously true, even saying that “anyone in their right mind should accept necessitism”, though Fine frames his broader modal metaphysics through the actualism versus possibilism dispute and Williamson thinks that dispute is less clear than necessitism versus contingentism as we’ve already noted right at the start. 

So, in one important respect, their ontological baseline might overlap. Both can accept that there are “merely possible” individuals, like Wittgenstein’s possible children, without treating them as fictional. The fight is about how to understand the status of those individuals, and about which conceptual machinery is doing the real explanatory work. This creates a striking situation for interpreting Lost Highway. On a shared necessitist baseline, you might say that “Pete” is not nothing, not an illusion, not a mere label, he is something, in the widest sense, even if his concreteness is conditional. But then the divergence begins. 

Williamson’s Lost Highway would treat the film’s ontology as fixed, and interpret the shifts as shifts in concreteness, in which properties are instantiated, in which relations become salient, and in which epistemic access routes are available. The Mystery Man’s eerie persistence becomes an emblem of the fixed domain. He is not created by guilt, he is what guilt latches onto. The videotapes become the “wide domain” pressing through the narrative’s attempt at concealment. The looping structure becomes a modal structure, the same objects under different world like circumstances, but without literal world creation. 

Fine’s Lost Highway would be more willing to say that the film’s real subject is essence. Not essence as a list of necessary properties, but essence as the ground of modal truths, what makes certain possibilities and necessities hold. A Finean would be attracted to the thought that Fred’s transformation is not just a redistribution of properties across a fixed individual, but a passage between different essential profiles, different “ways of being” that cannot be captured by the thin notion of which worlds the individual inhabits. “Alice” and “Renee” can then be treated as distinct essential roles or natures that the same underlying individual can occupy, but which are not reducible to mere necessary truth conditions. This is why Fine’s insistence on multiple modal notions matters. 

The necessity that “Renee is Renee” is not the necessity that “Renee is the woman in the porn film”, and even if both are necessary in the same worlds, only one might be true in virtue of her nature, and the other might be true in virtue of contingent entanglements with Eddy’s world. 

At this point, it helps to get very explicit about what “in virtue of” does, because it is the hinge. Williamson wants modal metaphysics to be answerable to the combined pressure of logic and theoretical virtues, simplicity, unity, explanatory power. Fine, by contrast, often insists that explanatory order runs from natures to necessities. A truth is metaphysically necessary because of what things are, not conversely. In more recent discussion of Fine you also see the same idea presented in an Aristotelian idiom, that it is things, not worlds, that “give rise” to modality. 

So imagine the desert cabin scene again. Alice says “You’ll never have me” and disappears. Immediately Pete becomes Fred. A Williamsonian treatment will say: nothing popped out of being. The same domain persists. What happened is that the property profile that was being instantiated, the Pete profile, ceased to be instantiated by a concrete individual in that situation, and another profile, the Fred profile, became instantiated. The change is dramatic but it is not ontologically creative. The metaphysical work is being done by modal logic plus an account of concreteness as world relative. 

A Finean treatment will say: the scene is about the limits of one essential role. Alice, as Alice, is essentially unavailable to possession by Pete, as Pete. “You’ll never have me” is not just psychological taunting, it is a statement of an essential constraint within that modality of desire and identity. Her disappearance is the manifestation of that constraint. The transformation is not just the turning on of a different set of predicates, it is the collapse of one essential pattern and the re emergence of another. The metaphysical work is being done by grounding, by essences, by hyperintensional distinctions that modal logic alone cannot see. 

These are not mutually exclusive in the sense that one is true and the other false, because films are not metaphysical treatises. But they come apart as interpretive visions, and the disagreement can be sharpened by asking a single question. Does the film’s horror come from the fact that the domain cannot be shrunk, that the unwanted objects remain there, or does it come from the fact that there are truths about what the characters are that cannot be rewritten by shifting worlds or shifting narratives?

Williamson’s answer leans to the first. Fine’s answer leans to the second. Now, there is a complication. Williamson is not blind to essence talk, and Fine is not blind to formal discipline. Williamson’s more recent critique of Fine, while hostile to certain hyperintensional moves, explicitly allows that hyperintensional essentialism may be intelligible, he is contesting its evidential support and its tendency to overfit. Fine, meanwhile, is perfectly comfortable with rigorous logic, but he does not want the logic to dictate the metaphysics by impoverishing the metaphysical distinctions in advance. 

That dialectic plays out in the way each philosopher would handle the film’s most metaphysically loaded object, the Mystery Man. In many popular discussions the Mystery Man is treated as a projection or an internal faculty, something like conscience, guilt, superego, or a policing function. Those readings are psychologically plausible, but notice that the film also treats him as ontologically strange. He is at Fred’s house and at the party. He seems to operate with privileged knowledge. He carries a camera as if recording were an ontological operation. 

A Williamsonian reading will be tempted to say: the Mystery Man is the film’s personification of the fixed domain and the rigid constraints of identity. He is what remains when you try to run away from what there is. He is, as it were, the Barcan principle walking around at night, enforcing that you cannot make possibilities safe by treating their objects as non entities. He is also what pushes the narrative back from the fantasy’s de dicto comfort, the comfort of descriptions, to the de re insistence on the thing itself, “what the fuck is your name”, as the quoted analysis puts it. 

A Finean reading will instead see the Mystery Man as the guardian of essence, the agent that forces the story to respect what the characters are in virtue of their natures, not just what they happen to be described as. His camera is then not just evidence gathering but essence revealing, a device that shows what is there in the way it is there, regardless of the narrative’s substitutions. He is the one who says, in effect, you can change the description, but you cannot change the nature that grounds the modal profile. 

On this difference, you can now see where Williamson and Fine overlap and where they come apart. They overlap in their impatience with certain Quinean ways of deflating modal disagreement. Both treat de re modality as intelligible, and both are willing, in different ways, to take modal discourse at face value rather than translating it into something else. Fine’s project can be understood partly as a defence of irreducible modal notions against empiricist reductions, and Williamson’s whole book is a repudiation of the thought that modal logic is merely a calculus without metaphysical bite. They also overlap, at least in the exchange Williamson reports, in being broadly comfortable with necessitist counting of merely possible individuals. Williamson explicitly notes Fine’s impatience with contingentism and his willingness to treat “possible knives” style cases as straightforward evidence that there are such things. That makes both of them unusually well suited, compared with many mainstream metaphysicians, to treat Lost Highway as a film in which “mere possibilities” are not just thoughts but part of the world’s stock. 

But they come apart over explanatory priority and over the granularity of metaphysical distinctions. Williamson’s priority is modal logic plus theoretical virtues, with an accompanying suspicion of hyperintensional metaphysical categories that outrun what modal discipline can justify. Fine’s priority is essence and grounding, with modal truths as downstream, and with a willingness to recognise distinctions that do not correspond to any separation in possible worlds. 

So, in their preferred readings of Lost Highway, the overlap shows up as agreement on a basic ontological generosity. The film’s doubles, its possible lives, its alternative guises, are not to be treated as nothing. There is something there, even when it is not concrete in the everyday sense, or when it is only “there” in the film’s modal space. The disagreement shows up in what they think the film is teaching you about why the horror is inescapable. 

Williamson’s reading makes the inescapability structural. The domain does not change, the objects are there, and the narrative’s attempt to escape is an attempt to cheat on quantification. The road loops because the ontology cannot be edited away. The intercom message repeats because the objects implicated by it do not blink out. The videotapes arrive because the world’s inventory is indifferent to the agent’s self narration. 

Fine’s reading makes the inescapability essential. The film’s loop is not just a fixed domain but a fixed nature. Fred cannot simply become Pete in the sense of shedding his nature, because the nature grounds what is possible for him, including which fantasies can be sustained and which collapse. Alice’s vanishing is not just a change in what is concrete, it is an expression of an essential constraint on possession, knowledge, and desire. The Mystery Man is terrifying because he is not merely a reminder that “the object is there”, but because he is a reminder that “the object is what it is”, and that modal space is carved by that fact. 

You can even phrase the contrast as two metaphysical aesthetics of Lynch. Williamson gives you an austerely logical Lynch, in which the weirdness is what happens when a world refuses to let you get away with sloppy quantifiers and shifting domains. The film becomes an allegory of why the simplest adequate modal logic is metaphysically constraining. 

Fine gives you an Aristotelian Lynch, in which the weirdness is what happens when essence, the natures of things, is not transparent to consciousness, and when a story tries to replace nature with description. The film becomes an allegory of why modal space is grounded, not merely charted, by worlds. 

The one concrete place where their readings would most sharply diverge is the status of identity statements in the film. Williamson will tend to treat “Fred is Pete” as an identity claim whose intelligibility depends on how we are modelling individuals across modal variation, and he will push the idea that our resistance to such claims is often psychological, a heuristic response, rather than decisive metaphysical evidence. Fine will tend to treat identity as only one layer, and will ask whether “Fred as Fred” and “Fred as Pete” pick out the same essence, the same ground of modal truths, or whether the film is precisely staging that essence is more structured than identity plus world membership can capture. Both readings can be made to fit, but they make different features of the film central. 

Williamson’s makes the film’s bureaucracy, the police, the tapes, the intercom, the repeated message, feel like metaphysical operators. Fine’s makes the film’s erotic metamorphosis, the shifting names, the vanishing woman, feel like essence in action. And once you see that, you can also see why the comparison is not an external academic exercise. It is a way of noticing that Lynch’s film is itself a laboratory for rival pictures of modality. Is the nightmare that you cannot make the domain smaller, or is it that you cannot make the natures looser? Williamson: the nightmare is that the quantifiers do not care what you want. Fine: the nightmare is that the natures do not care what you narrate. 

Let’s keep pushing on this because none of this is easy and I need to be able to have concrete examples rather than mere schemas. So, the party phone call is where the film announces its metaphysical ambitions really clearly. Fred meets the Mystery Man at a party. The Mystery Man tells Fred that he is at Fred’s house right now. He hands Fred a phone. On the other end, the Mystery Man answers. 

On a Williamsonian reading, this scene is doing something very precise. It is forcing a collapse between what Fred takes to be mutually exclusive locations, here and there, now and now. The discomfort is that the scene refuses to respect a tacit domain restriction. Fred’s ordinary reasoning presupposes that the relevant objects are localised, that if someone is here they are not there. But the scene operates at a different level. The Mystery Man is not being treated as a localised concrete individual in the ordinary sense, he is being treated as something whose identity and availability are not governed by Fred’s perspective. 

In modal–logical terms, the scene is an attack on de dicto safety. Fred wants to reason about “the man here” and “the man there” as if those descriptions carve reality. The scene insists on a de re grip: this very individual does not respect your descriptive partition. The horror is the same horror Williamson attributes to unrestricted quantification. The domain does not shrink to fit your epistemic comfort. When the Mystery Man says “I’m at your house”, what is unbearable is not the content but the implication that the world’s inventory is not under Fred’s control. 

On a Finean reading, the pressure point is different. The phone call is not primarily about the size of the domain, but about the violation of essential constraints. The Mystery Man is not frightening because he ranges over all objects, but because he occupies an impossible role relative to ordinary human essence. A human being, as such, cannot be both here and there. The scene works because the Mystery Man is not treated as merely violating a physical law, but as violating what it is to be a person at all. 

The call is therefore a confrontation with essence, not with quantifiers. The film is telling you that this figure does not fall under the same kind as Fred, even if he wears a human face. The metaphysical disturbance is hyperintensional. Two propositions may be necessarily equivalent, “the Mystery Man is human” and “the Mystery Man looks human”, but only one tracks what he is. The scene unsettles because it exposes the insufficiency of modal equivalence for capturing nature. 

Already you can see the split. Williamson hears “you cannot hide behind descriptions”. Fine hears “you cannot pretend this thing has the same nature”. The prison transformation is the most violent ontological moment in the film. Fred is on death row. He suffers headaches. The guards open the cell. Pete Dayton is there. Williamson’s preferred description is that nothing came into existence. Nothing went out of existence. The domain is fixed. What happened is that the object that was concrete under the Fred profile ceased to be concrete in that way, and another object, already in the domain, became concrete. The shock comes from confusing concreteness with existence. 

Williamson would insist that our intuition that “Fred no longer exists” is a heuristic misfire produced by our ordinary way of tracking persons, which is tied to bodily continuity, memory, and social recognition. But modal metaphysics, regimented properly, has no need of that heuristic. Pete did not arrive from nowhere. Fred did not vanish into nothing. The film’s showing how completely the conditions of concreteness can change while the ontology remains indifferent. 

Fine’s reading resists that flattening. The prison scene is not merely a switch in which individual is concrete. It is a collapse of one essential profile and the emergence of another. Fred cannot continue as Fred, not because of contingent circumstances, but because what Fred is cannot bear what has become true. Pete is not simply another object stepping into the spotlight. He is a dependent reconstitution of agency, an attempt to generate a subject with a different essential structure, younger, less burdened, more erotically competent, less tied to the act of murder. On this view, the transformation is not explained by modal recombination but by essential failure. The film is not saying “the world allows this object to be concrete instead”. It is saying “this essence cannot survive this truth, so a different one is instantiated”. The metaphysical violence is not about existence but about what can count as the same person. 

The photograph scene deepens the divide. Pete finds a photograph of himself with Alice. The photograph changes. Alice disappears from it. For Williamson, this is a lesson in the instability of representational anchors. The photograph is treated by the characters as a truthmaker, a local object that secures identity. But the film refuses that comfort. Identity is not fixed by local artefacts. The disappearance of Alice from the photograph does not mean she no longer exists. It means that the configuration of properties and relations that made her concrete in that narrative strand has ceased to obtain. The photograph changes because it is not metaphysically fundamental. It tracks concreteness, not being. This is a visualisation of Williamson’s scepticism about truthmaker maximalism. Not every truth has to be backed by a concrete thing in the way we expect. Modal structure outruns the photograph. 

For Fine, the photograph scene is textbook hyperintensionality. The proposition “Alice was there” and the proposition “the woman in the photograph was there” may be necessarily equivalent at some coarse level, but the film insists that they come apart in grounding. The disappearance is not merely a shift in what is concrete. It is a signal that the identity conditions have failed. Alice, as Alice, does not have the right essential relation to Pete to remain fixed across representational media. The photograph cannot stabilise her because the metaphysical dependence runs the other way. She is not grounded by the photo, nor by Pete’s desire. She is grounded by an essential role that is no longer operative. The image collapses because the essence has withdrawn. 

The desert cabin scene is where the two readings become maximally distinct. Pete and Alice reach the cabin. Alice says, “You’ll never have me.” She walks away. She disappears. Pete becomes Fred. Williamson hears this as the cleanest case of modal disappointment. The fantasy reaches its limit. The object of desire simply ceased to be concrete in that configuration. Desire aimed at concreteness, at possession, but the domain was always larger than that. The transformation back to Fred is not punishment imposed from outside. It is the reassertion of the fixed ontology. The attempt to escape by changing roles fails because the roles were always parasitic on a stable domain. The desert, empty and overexposed, is the film’s image of unrestricted quantification. There is nowhere to hide because nothing has been removed. 

Fine reads the same moment as the triumph of essence. “You’ll never have me” is an essential impossibility. Alice’s disappearance is the revelation that the conditions for identity were never met. Pete cannot possess Alice because Pete, as Pete, is not the kind of thing that could stand in that relation to Alice, as Alice. When that essential mismatch becomes explicit, the constructed identity collapses. Fred returns because Fred is the only essence available that can bear the underlying facts. The desert is not a domain but a testing ground, a place where essential constraints are stripped of disguise. 

Finally, the closing chase and the looping message, “Dick Laurent is dead”. Williamson treats the loop as modal closure. The message repeats because the ontology does not reset. The same objects are implicated again. The chase is toward recognition that the modal space is already structured. The beginning was always part of the end. This is the cinematic analogue of S5-style stability. Possibility does not open infinite exits. It rearranges a fixed set of necessities. The scream at the end is the scream of an agent confronting the fact that the quantifiers were never local. 

Fine treats the loop as essential return. The message repeats because the essential structure has not been altered. Fred cannot narrate himself into a different nature. The loop is not modal exhaustion but constitutive inevitability. The end is the beginning because the essential facts that ground the story never changed. The chase is not through a space of worlds but through failed reconstitutions of selfhood. 

So when you put the two philosophers fully into the film, the contrast becomes very sharp. Williamson’s Lost Highway is about the cruelty of a fixed ontology combined with volatile concreteness. The horror is that existence does not bend to narrative, memory, or desire. Modal logic, properly regimented, shows why the escape routes are illusory. Fine’s Lost Highway is about the cruelty of essence. The horror is that what something is constrains what it can be in ways no amount of modal reshuffling can dissolve. The escape routes fail not because the domain is fixed, but because the natures are. They overlap in refusing to let the film dissolve into “it’s all in his head”. Both insist that the film is metaphysically serious. They diverge in what they think the seriousness consists in. Williamson says the seriousness is logical and ontological. Fine says the seriousness is constitutive and essential. And Lynch, characteristically, gives you just enough structure that both readings remain live, while making each one feel, in isolation, terrifyingly incomplete. 

If we ask the harder question seriously, not “can Lost Highway be read in a Williamsonian or a Finean way”, but “which metaphysical picture does the film itself exert pressure towards, once you stop rescuing it with external theory”, then maybe the answer is that the film’s internal discipline quietly favours one over the other. The temptation, especially for philosophers, is to say that Lynch is staging an undecidable metaphysical experiment, that the film is deliberately underdetermined, and that any attempt to settle it is a projection. That temptation mirrors a familiar quietism in philosophy of modality, the thought that once we have multiple coherent frameworks, there is no fact of the matter which one reality prefers. 

But Lost Highway is punitive. It punishes certain interpretive moves and rewards others. If you follow the film’s own economy of explanation rather than imposing your own, it begins to lean. The first thing to notice is what the film refuses to let stabilise. It refuses to stabilise identity criteria at the level of essence. Every time the viewer tries to say “this is what Fred really is”, the film undercuts the claim by showing how easily that alleged essence can be reconfigured, re clothed, or displaced. Fred as husband, Fred as jealous man, Fred as murderer, Fred as saxophonist, none of these roles is treated as metaphysically basic. They are all treated as fragile, contingent surfaces that can be shed. 

Pete’s youth, sexuality, and social embedding are not presented as violations of an underlying essence, but as available recombinations. The film never signals that a constitutive boundary has been crossed in the way Fine’s essence first metaphysics would require. There is no moment where the narrative says, in effect, “this cannot be the same thing, because the nature has been violated”. Instead, the film behaves as if nature itself is plastic, or at least not the explanatory bedrock. What the film does insist on, relentlessly, is the persistence of something that is not reducible to character traits, social roles, or self understanding. Guilt persists. Violence persists. The murder persists. The message “Dick Laurent is dead” persists. These are not treated as properties of a particular essence, but as facts that remain in play regardless of who is concrete, who is visible, or who is narrating. 

The film’s invariants are not essences, they are facts and relations that cut across identity shifts. The Mystery Man does not appeal to what Fred essentially is. He appeals to what has happened. He does not say “you are the kind of thing that must be punished”. He says, effectively, “this occurred, and it is still there”. That is a deeply Williamsonian pressure point. Modal structure is anchored in what there is and what is the case, not in an essence laden story about what things are in themselves. Identity can wobble, descriptions can fail, self conceptions can fracture, but the domain and the facts do not go away. The film’s horror repeatedly comes from the same source, the world does not forget just because the subject does. 

An essence first metaphysics of the kind associated with Kit Fine would predict a different cinematic grammar. If the film were fundamentally about essential impossibility, you would expect moments where the narrative signals metaphysical impossibility rather than mere failure. You would expect certain transitions to be marked as violations of nature rather than collapses of fantasy. But Lost Highway does not treat the Fred to Pete transformation as metaphysically forbidden or incoherent. It treats it as horrifyingly easy. That ease is not compatible with a view on which essence tightly constrains what an object can be. It is compatible with a view on which what is constrained is not what things are, but which configurations of properties and relations can remain concrete together. 

The film also consistently privileges external, impersonal registers of reality over internal, essentialist ones. Cameras, tapes, telephones, police files, photographs, these are the film’s truth bearing devices. They are crude, mechanical, indifferent to self interpretation. They track the world in a way that bypasses essence. A Finean reading might try to say that these devices reveal essence, but the film never uses them that way. They do not tell you what someone is. They tell you that something happened, that something is still there, that the world has not been edited. 

This aligns almost perfectly with Williamson’s scepticism about truthmaker fetishism combined with his insistence that modal facts outrun our representational preferences. The tape does not ground the truth, but it enforces it. Even the erotic structure of the film supports this. Alice’s line, “You’ll never have me”, is often read as an essentialist refusal, as if she is saying “I am not the kind of thing you can possess”. But the film immediately undermines that reading by making her disappearance coincide with a shift in concreteness rather than a revelation of nature. She does not remain as an inaccessible essence. She vanishes as a concrete object in that configuration. The desire fails not because it targets the wrong essence, but because the conditions under which that object was concrete have collapsed. This is a world in which possession fails because actuality is unstable, not because nature draws an absolute boundary. 

The looping structure at the end clinches the case. The return of the opening line is not framed as a return to an essential truth about Fred. It is framed as the return of a fact. The same message, the same content, the same worldly commitment. The film does not say “this is who you are”. It says “this is still there”. That is a metaphysics of persistence without essence, of necessity without nature. It is precisely the kind of metaphysical environment in which necessitism feels at home. The ontology does not shrink. The facts do not dissolve. Only the modes of inhabiting them change. 

So Lost Highway is more naturally read as siding with Williamson’s modal metaphysics than with Fine’s. Not because it illustrates necessitism in any didactic way, but because its internal economy punishes essence based stabilisations and rewards ontology first ones. It is a film in which identity is cheap, but being is expensive. In which what you are can change, but what there is does not. In which escape through redefinition fails, but escape through annihilation is never even on offer. If Fine’s metaphysics is haunted by the thought that modality must ultimately be grounded in what things are, Lynch’s film is haunted by the opposite thought, that what things are is too fragile to ground anything, and that the true source of horror lies in the indifference of being to self description. 

That is why the film never lets you say, with confidence, “this is the real Fred”, but it never lets you say, with confidence, “this is no longer the same world”. In the end, Lost Highway does not ask whether we could have been otherwise. It asks whether anything could ever stop being there. And its answer, like Williamson’s, is quietly and relentlessly, no. 

So after that, can we assume that all Lynch is best fitted to a Williamsonian modal metaphysics? To test that idea, let’s turn to Inland Empire once more and see how the fit might go. The most useful way to approach Inland Empire through Timothy Williamson is to begin from the thought that this is Lynch’s most ontologically explicit film. Not explicit in the sense of being clear, but explicit in the sense that it strips away the last narrative protections that allowed earlier films to be rescued by psychology, symbolism, or dream talk. 

Where Lost Highway and Mulholland Drive still allowed the fantasy that there is a stable level at which “what really happened” could in principle be reconstructed, Inland Empire is organised so that any attempt to locate such a level collapses. Scenes do not merely contradict each other, they coexist without hierarchy. Characters are not merely doubled, they are multiply instantiated without any clear priority relation. Time does not merely loop, it fragments into overlapping presents. The film does not invite you to ask which layer is real, it punishes you for asking. This looks like a film where a Williamsonian modal metaphysics becomes almost unavoidable. Williamson’s central insistence is that we must sharply distinguish between what there is and how it is, between the domain of objects and the patterns of concreteness, instantiation, and representation that vary across possibilities. The temptation to explain modal phenomena by shrinking or expanding the domain is, on his view, a deep source of confusion. 

Inland Empire can be read as a sustained exercise in breaking that temptation. From the opening scene onwards, the film establishes that the domain will not be curated for the viewer. The Polish conversation, the crying woman, the ominous declaration that “it happened yesterday but it’s happening again tomorrow”, already signals that the film will not respect the ordinary alignment of time, event, and narration. But crucially, these are not presented as alternative realities replacing one another. They stack. They remain. They return. Nothing is ever cleanly discarded. This is the first major point of contact with necessitism. The film behaves as if once something is introduced into the ontological field, it cannot be removed, only displaced, recontextualised, or rendered inaccessible from certain vantage points. 

Nikki Grace, Laura Dern’s central figure, is not merely a character who becomes confused about her role. She is a site at which multiple roles, identities, and narrative positions become concrete in overlapping ways. At various points she is Nikki, an actress. At others she is Susan Blue, the character she is playing. At others she is a woman in a Polish story, or a prostitute in Los Angeles, or a figure watching herself on a television screen. The key Williamsonian move is to refuse the idea that these are separate ontologies or mutually exclusive worlds. The film never gives you a principled reason to say that one of these is merely imagined while another is real. Instead, it presents a fixed, overcrowded domain in which different identities become concrete under different conditions. In Williamson’s terms, what is shifting here is not existence but concreteness. The entities are all there. Nikki, Susan, the Polish woman, the rabbits, the Lost Girl, the phantom figures, the domestic interiors, the sets, the television screens, all of them persist as items in the film’s world. 

What changes is which of these items are concrete for whom, at what moment, and under what narrative or perceptual constraints. This is exactly the distinction Williamson insists upon when he argues that many contingentist intuitions are really intuitions about concreteness masquerading as intuitions about existence. When viewers say “this character ceases to exist” or “this was only a dream”, they are often responding to a change in accessibility or instantiation, not to a change in what there is.  Inland Empire systematically exploits this confusion. Scenes bleed into one another without transition. A door opens and we are elsewhere, not because we have changed worlds, but because the film has shifted which region of the fixed domain is now concrete for the camera. 

The famous corridor sequences, the repeated movement through domestic spaces that never quite stabilise, function as cinematic analogues of unrestricted quantification. There is always more there than is currently accessible. The camera’s movement does not create new objects, it reveals or conceals objects that were already part of the domain. This is why attempts to read the film as a psychological breakdown, while not wrong, are metaphysically insufficient. A psychological reading implicitly treats certain scenes as internal and therefore ontologically downgraded. But the film refuses to mark any scene as merely internal. Even hallucination like sequences are treated with the same ontological weight as “external” events. The prostitutes’ dance to Nina Simone, the violent stabbing, the moments of confession, the scenes with the rabbits, all occupy the same cinematic register. 

The film does not cue the viewer to perform an ontological filtering operation. This refusal aligns very closely with Williamson’s hostility to domain restriction as a metaphysical crutch. The rabbit sitcom sequences are especially revealing here. They are often dismissed as symbolic, absurdist, or Brechtian interruptions. But from a Williamsonian perspective they function as ontological anchors. The rabbits are not metaphors standing in for something else. They are entities in the domain whose mode of concreteness is peculiar. They exist behind a laugh track, within a set that looks like a television show, yet they bleed into the rest of the film through sound, dialogue, and repetition. Their presence makes it impossible to say that the film’s world is exhausted by human characters, realistic spaces, or linear narrative. The domain is wider than the story we are trying to tell, and the rabbits are a reminder of that excess. This excess is not chaotic. It is structured by repetition, echo, and persistence. Lines recur. Gestures recur. Faces recur. The phrase “Axxon N” recurs. The same scenes appear with slight variation. 

This is where Williamson’s modal logic becomes interpretively powerful. The film behaves like a space of possibilities in which the same objects are recombined under different patterns of instantiation. What we see are not different worlds in the Lewisian sense, but different configurations within a single modal space. The film’s refusal to privilege any configuration as “actual” mirrors Williamson’s refusal to let actuality do heavy metaphysical lifting. Actuality, for him, is an indexical notion (like’I’, ‘here’, ‘there’ etc), not an ontological one. Inland Empire treats actuality the same way. What is actual is whatever is currently concrete for the camera, but that concreteness is always provisional. Time in the film works the same way. The repeated insistence that events have already happened and are happening again does not suggest time travel or alternate timelines. It suggests that temporal location, like personal identity, is a mode of concreteness rather than a marker of existence. Events do not come into being and pass out of being. They are instantiated, de instantiated, and reinstantiated under different narrative conditions. 

This is why the film can present the same death, the same confession, the same moment of recognition, without implying that one is the real one and the others are copies. The film’s temporality is modal rather than linear. It is organised around what could be the case for which objects, rather than around a single privileged history. Violence in the film further supports this reading. When Nikki or Susan is stabbed, the act is not framed as an event that cleanly terminates a character. Death does not remove entities from the domain. Instead, it alters their mode of presence. The stabbed woman continues to appear, continues to speak, continues to act. Death is not annihilation. It is a change in concreteness. 

This aligns almost perfectly with Williamson’s insistence that metaphysics should not build annihilation into its basic vocabulary. Objects do not wink out of existence when they cease to be concrete in a given world or narrative. They persist as members of the domain, available for re instantiation. This is also why the film’s use of screens within screens is so important. Characters watch themselves on televisions. Scenes appear to be filmed within the film. These devices are often read as postmodern commentary, but they can also be read as metaphysical devices that foreground the distinction between existence and access. The fact that a character appears on a screen does not mean they are less real. It means their mode of presence is mediated. 

Williamson’s metaphysics is deeply sympathetic to this idea. He repeatedly emphasises that our epistemic access to objects should not be conflated with their ontological status. Inland Empire makes this separation experiential. Perhaps the clearest Williamsonian moment is the film’s ending. The Lost Girl is released. The rabbits sit peacefully. Nikki smiles. There is a sense of closure, but not of resolution in the ordinary sense. Nothing has been eliminated. No ontology has been simplified. The film ends not by explaining away its excess, but by stabilising it. The domain remains vast. What has changed is the pattern of concreteness. Suffering has ceased for certain figures. Access has been granted. But the world has not been reduced. This is exactly the kind of ending one would expect from a necessitist metaphysics. Salvation, if it exists, is not achieved by making the world smaller. It is achieved by reconfiguring how the existing world is inhabited. 

Seen this way, Inland Empire is Lynch’s most uncompromisingly Williamsonian film. It abandons the last vestiges of essence based identity, psychological grounding, or world switching explanations, and instead confronts the viewer with a single, overcrowded ontological field in which everything persists and nothing can be cleanly excluded. The terror of the film is not that reality fractures, but that it does not. There is no safe outside. No layer can be declared merely representational. No object can be declared merely possible and therefore dismissible. The domain is fixed, and the only thing that changes is which parts of it we are forced to confront. 

Having said all that, I now want to see what Kit Fine and his essences might make of the film and as might be expected, as soon as we shift registers and let Fine take interpretive control, Inland Empire begins to look very different, not because the film changes, but because the metaphysical questions we allow ourselves to ask change. Where a Williamsonian reading is patient with ontological excess and tries to discipline it through a distinction between existence and concreteness, a Finean reading is much less tolerant of treating identity shifts as merely redistributions within a fixed domain. Fine’s metaphysics is driven by the thought that modality, identity, and necessity are grounded in what things are, in their essences, and that possible worlds style recombination, even when handled with care, risks flattening the very distinctions that matter most. 

The first thing a Finean reading notices about Inland Empire is that the film is saturated with dependence relations. Nothing stands alone. Characters depend on roles, roles depend on scripts, scripts depend on earlier scripts, films depend on cursed films, identities depend on performances, performances depend on cameras, cameras depend on watchers. The film’s ontology is not merely large, it is layered, and the layers are asymmetric. Some things exist only in virtue of others. 

The Polish story exists in virtue of the American film. Susan exists in virtue of Nikki. Nikki exists in virtue of being cast. The Lost Girl exists in virtue of being abandoned in a story that has already happened. Fine sees that these are not just narrative facts but are metaphysical facts about grounding. Fine’s work consistently insists that grounding, the “in virtue of” relation, is prior to modality. What is possible or necessary is explained by how things are constituted, not the other way around. From this perspective, Inland Empire is not best understood as a space of recombinations of already existing objects, but as a dynamic structure in which new entities, identities, and even times come into being through acts of dependence. 

When Nikki becomes Susan, this is not simply a matter of a different object becoming concrete. It is a matter of a new dependent entity coming into existence, grounded in Nikki’s performance, the script, the film production, and the cursed narrative inherited from the Polish original. This immediately changes how we should understand the film’s apparent refusal to privilege any level as “real”. On a Williamsonian reading, this refusal supports ontological egalitarianism. On a Finean reading, it supports ontological stratification. The film does not flatten reality. It thickens it. It multiplies layers of being, each grounded in others, without ever collapsing them into a single domain. The fact that the film does not tell us which layer is fundamental does not mean that none is. It means that the film is dramatising the difficulty of tracing grounding relations when they become recursive, circular, or opaque. 

The recursive curse narrative is crucial here. The film within the film is itself a remake of an unfinished Polish film whose actors were murdered. This is not just a plot device. It is an ontological engine. The new film depends on the old one. The old one depends on a prior story of infidelity and violence. That story, in turn, seems to bleed into the lives of the actors who attempt to tell it. From a Finean perspective, this is not modal recombination but essential inheritance. Certain structures cannot be instantiated without importing their essential constraints. When Nikki steps into the role of Susan, she does not merely take on contingent properties. She takes on an essence laden role whose internal relations, jealousy, violence, abandonment, are constitutive. This is why the film repeatedly suggests that once the role is entered, certain outcomes are unavoidable. 

That inevitability is captured by essential necessity grounded in the nature of the role itself. The rabbits take on a different metaphysical significance under this lens. Instead of functioning as reminders of ontological excess, they function as paradigms of dependent being. They exist as a sitcom. Their dialogue is laugh tracked. Their space is a set. Their pauses are scripted. They are not simply strange entities in a wide domain. They are entities whose essence is to be part of a televisual structure. They cannot be abstracted from that structure without ceasing to be what they are. 

This is a textbook Finean point. Some entities are essentially dependent. They do not merely happen to rely on other things. Their nature includes that reliance. The rabbits are not rabbits that happen to be in a sitcom. They are sitcom rabbits. This also reframes the film’s treatment of time. On a Williamsonian reading, time is another dimension of concreteness. On a Finean reading, time in Inland Empire is grounded in narrative structure. Events do not merely recur. They recur because the story requires them to. The line “it happened yesterday but it’s happening again tomorrow” is not a statement about temporal loops in a modal space. It is a statement about the essential incompleteness of a narrative whose grounding conditions have not been discharged. The story cannot finish because its constitutive wrong has not been addressed. Time persists because the grounding failure persists. 

Violence, on this view, is not a mere invariant fact that survives identity shifts. It is an essential feature of certain roles and stories. The woman is stabbed because that is what happens in this story. When Nikki or Susan is stabbed, the act does not merely change her mode of presence. It completes, or partially completes, an essential narrative arc. Death in Inland Empire is not ontological thinning. It is narrative closure or failure thereof. This is why characters continue after death. The role has not been discharged. The essence of the character includes unfinished business. The Lost Girl is the clearest case. She is trapped in a room watching television. She cries. She waits. She exists in a state of suspended grounding. Her existence depends on a story that has gone wrong. She cannot be released by changing which objects are concrete. She can only be released by rectifying the grounding relations that sustain her. When Nikki finally confronts the woman with the screwdriver and then embraces the Lost Girl, this is not a redistribution of concreteness. It is a grounding repair. The essential dependency that kept the Lost Girl trapped is dissolved. This is why the ending feels like salvation rather than explanation. Something has been put right, not something has been moved. 

This also changes how we should understand Nikki’s multiplication. When she watches herself on screen, this is not simply epistemic distance or modal self reference. It is the emergence of higher order entities. The watching Nikki and the watched Nikki are not numerically identical entities under different guises. They are distinct dependent entities, one grounded in the other. Fine’s metaphysics is comfortable with this proliferation because identity is constrained by essence. If two entities have different grounding conditions, they are not the same entity, even if they are necessarily correlated. This is precisely the sort of hyperintensional distinction that Fine insists modal logic alone cannot capture. 

The Finean reading also makes better sense of the film’s emotional texture. Inland Empire is not primarily frightening because it overwhelms us with too much being. It is frightening because it shows how easily one can be constituted into something one cannot control. Nikki does not lose herself because the domain is too large. She loses herself because she becomes essentially entangled in structures that define her possibilities from the inside. This is an existential horror, not an ontological one. It is closer to tragedy than to cosmic dread. 

Seen this way, Inland Empire is less about the indifference of being and more about the tyranny of structure. Roles trap. Stories bind. Essences constrain. The film’s refusal to sort its layers is not a statement that all layers are equal, but that grounding relations can become so tangled that no single perspective can survey them. This is very much in the spirit of Fine’s scepticism about treating modality as a single unified space of worlds. The film’s modal space is fractured because its grounding space is fractured. 

At the same time, the Finean reading is under strain in places. The film’s apparent ontological generosity, the way it lets entities persist even after their grounding conditions seem to have dissolved, pushes against a strict essence first account. The rabbits linger even when their narrative function is unclear. The Mystery figures recur without clear grounding. Some scenes feel less like dependent constructions and more like ontological residue. These moments threaten to pull the Finean reading back toward a Williamsonian one, because they suggest that the film sometimes treats being as more basic than grounding. Still, as a whole, the Finean reading captures something the Williamsonian reading risks missing. Inland Empire is not only about how many things there are, or how they are instantiated. It is about how stories, performances, and identities generate new kinds of beings with their own internal necessities. It is about how you can become something you did not choose, not by metaphysical accident, but by stepping into a structure whose essence already contains your fate. 

To weigh the two readings properly, we have to resist a tempting but shallow move, namely to ask which philosopher gives us the cleverer interpretation, or which framework can be made to “fit” more scenes with fewer loose ends. Lynch’s films are unusually resistant to that kind of scorekeeping. What matters is not coverage but pressure: which metaphysical picture is the film itself exerting pressure towards, once we take seriously what it stabilises, what it lets drift, and what it actively punishes when the viewer tries to domesticate it. The question, then, is not whether Inland Empire can be read in a Williamsonian or a Finean way. 

It clearly can be read both ways, as we have just seen. The harder question is which reading better respects the film’s own internal asymmetries, its distribution of explanatory weight, and its sense of what counts as resolution versus evasion. When we ask that harder question, the balance tips, but not immediately, and not cleanly. Start with what both readings get right. Both Timothy Williamson and Kit Fine offer resources that are unusually well suited to Lynch. Both reject the idea that modality is a thin linguistic veneer. Both insist that questions about what could be, what must be, and what is, are substantive questions about the structure of reality. Both are impatient with the kind of interpretive quietism that says ambiguity dissolves metaphysical commitment. In that respect, both philosophers are better guides to Inland Empire than the more common psychologising or purely symbolic readings, which tend to flatten the film into an allegory of trauma, guilt, or artistic anxiety. Where they diverge is not on whether the film is metaphysically serious, but on where that seriousness lives. 

The Williamsonian reading locates seriousness at the level of ontology and quantification. The world of Inland Empire is terrifying because it is ontologically promiscuous and ontologically stubborn at the same time. Promiscuous, because it allows a vast array of entities, scenes, characters, times, and places to coexist without hierarchy. Stubborn, because none of them can be cleanly erased. The film behaves as if once something is there, it remains there, even if it becomes inaccessible, occluded, displaced, or transformed. Identity shifts, narrative breaks, and temporal dislocations do not thin the ontology. They thicken it. 

The Finean reading, by contrast, locates seriousness at the level of essence and grounding. The world of Inland Empire is terrifying because of the way roles, stories, and performances generate entities that trap those who enter them. The horror is not that the world contains too much, but that certain structures, once activated, impose internal necessities that cannot be escaped by changing perspective. Identity shifts are failures or transformations of essence. Narrative recursion is grounding collapse. Both readings can account for much of what happens in the film. The task is to see which one better explains the film’s asymmetries, especially the asymmetry between what the film treats as negotiable and what it treats as non negotiable. 

One way into this is to look closely at how Inland Empire handles identity. Identity in the film is radically unstable at the surface level. Nikki Grace, Susan Blue, the Polish woman, the prostitute, the Lost Girl, are not neatly sorted into one person versus many. The film encourages, then frustrates, attempts to line them up as either the same individual or distinct individuals. This instability initially looks like a point in favour of the Finean reading, because Fine insists that identity cannot be treated as a simple matter of world membership or modal profile. He emphasises that essence and grounding determine identity conditions, and Inland Empire seems to be obsessed with precisely that question, what counts as the same person when roles, names, histories, and bodies no longer line up. 

But notice what the film does when viewers try to stabilise identity through essence. Every time we try to say “this is what Nikki really is”, the film undermines the claim. Nikki as actress, Nikki as victim, Nikki as murderer, Nikki as saviour, all of these roles are treated as contingent and revisable. There is no moment where the film signals that an essential boundary has been crossed. Even the most extreme transitions, from actress to prostitute, from living to dead, from observer to observed, are handled without metaphysical alarm bells. The film does not present these shifts as violations of nature. It presents them as horrifyingly easy. That ease matters. 

A Finean metaphysics predicts that essence should bite. There should be points where the film insists “this cannot be the same thing”, not merely because of confusion or fantasy, but because the constitutive conditions have failed. Yet Inland Empire consistently refuses to mark such moments. Instead, it treats identity as cheap and malleable. What is not cheap is the persistence of certain facts, certain relations, certain sufferings. This is already a sign that the film’s metaphysical weight is not being carried by essence. The handling of death reinforces this. In a Finean framework, death often marks an essential boundary. What it is to be a living person is essentially different from what it is to be a dead one, even if the narrative continues to track a ghost, memory, or dependent entity. 

But in Inland Empire, death does not function that way. When Nikki or Susan is stabbed, the event does not clearly terminate an entity or close an essence. The stabbed woman continues to speak, move, appear, and act. Death is not a metaphysical cutoff. It is a change in mode of presence. This is much more naturally captured by Williamson’s distinction between existence and concreteness than by Fine’s emphasis on essential natures. 

The same pattern appears in the film’s treatment of time. A Finean reading treats time as grounded in narrative structure. Events recur because the story has not discharged its essential wrong. That reading works up to a point, but it struggles with the film’s indifference to narrative completion. The film does not resolve by finishing a story. It resolves by reconfiguring access. The Lost Girl is released not because a story has been properly completed in a classical sense, but because Nikki performs an act that changes how the domain is inhabited. The film’s ending does not feel like the resolution of a tragic structure. It feels like a redistribution of presence. 

This is where the Williamsonian reading gains traction. Williamson insists that many things we describe as coming into or going out of existence are really changes in which objects are concrete, accessible, or instantiated. Inland Empire behaves exactly as if that were true. Characters do not leave the world when their narrative role ends. They drift to the margins, reappear under other guises, or persist as images, sounds, or memories. The world accumulates rather than resolves. The rabbits are an especially telling test case. Under a Finean reading, they are essentially dependent entities, sitcom beings whose nature ties them to a televisual structure. That is true as far as it goes. But the film does not treat their dependence as limiting their ontological reach. They bleed into the rest of the film in ways that violate their supposed grounding. Their dialogue echoes elsewhere. Their presence affects the emotional tone of scenes that have no narrative connection to them. They are not contained by their essence. They exceed it. 

This excess is exactly what a Williamsonian ontology predicts. Once something is in the domain, its mode of instantiation can shift in ways that outrun the grounding relations that originally brought it into view. Fine’s framework is less comfortable with this kind of ontological afterlife. It wants dependence relations to explain why entities exist and what they can do. Inland Empire repeatedly shows entities doing things that seem to outstrip any clear dependence story. The film’s technological motifs further tip the balance. Cameras, screens, recordings, and broadcasts are not treated as mere representations. They are treated as ways in which reality insists on itself. When a character appears on a screen, the film does not cue us to downgrade their ontological status. If anything, screen mediated presence often feels more authoritative than embodied presence. 

This fits neatly with Williamson’s insistence that epistemic access and ontological status must be kept apart. Fine’s framework can accommodate mediated presence, but it tends to read it as derivative, grounded in something more fundamental. Inland Empire does not give us that more fundamental layer. Perhaps the strongest evidence comes from the film’s attitude toward explanation itself. Fine’s metaphysics is explanatory in a very specific sense. It aims to tell us what grounds what, what holds in virtue of what, and why certain modal facts obtain. It is a metaphysics that promises depth. Williamson’s metaphysics, by contrast, is often accused of being shallow, of refusing to go “beneath” modal facts to something more fundamental. But that shallowness is deliberate. Williamson thinks many metaphysical demands for deeper explanation are misguided. Sometimes the right answer is that this is just how the structure is.  

Inland Empire consistently frustrates explanatory depth in exactly this way. Every attempt to say “this is happening because…” is met with further proliferation rather than grounding. The Polish curse does not explain the American film. It generates more scenes. The film within the film does not ground Nikki’s suffering. It multiplies her identities. Even the apparent act of redemption at the end does not explain away what came before. It coexists with it. The film’s metaphysical temperament is one of accumulation, not reduction. That temperament aligns much more closely with Williamson’s anti exceptionalism about logic and metaphysics than with Fine’s essence first ambitions. Williamson is comfortable with a metaphysics in which the world is structured but not narratively neat, in which modality constrains but does not redeem, and in which ontology is larger than any story we tell about it. Inland Empire feels at home in that picture. 

This is not to say that the Finean reading fails. On the contrary, it captures something vital about the film’s sense of entrapment, especially the way roles and stories exert internal pressure on those who inhabit them. It is particularly good at explaining the film’s emotional logic, the way certain patterns feel inescapable even when identities change. But that strength is also its weakness. The film does not ultimately present those patterns as essential in Fine’s sense. They are breakable. They are disrupted. They are undone not by discovering a deeper nature, but by reconfiguring presence. When Nikki embraces the Lost Girl, the moment does not feel like an essence being fulfilled. It feels like an access barrier being lifted. The Lost Girl does not reveal who she essentially is. She stops being isolated. That is a change in concreteness, not in nature. The world has not been re grounded. It has been re inhabited. 

This final point shows where the film locates hope, such as it is. In a Finean universe, hope would lie in aligning oneself with one’s true nature or in repairing grounding relations. In Inland Empire, hope lies in surviving the fact that the world is too big, too crowded, too persistent to be mastered. The ending does not restore order. It restores coexistence. So, when we weigh the two readings carefully, the conclusion is not that the Finean reading is wrong, but that it is secondary. It captures local structures within the film, especially the way narratives generate internal necessities, but it does not capture the film’s global metaphysical mood. That mood is Williamsonian. The deepest horror of Inland Empire is not that essences trap us. It is that being does not care enough about us to let us escape by redefining ourselves. Lynch’s most radical move in Inland Empire is to deny us the comfort of a ground floor. There is no final layer where things make sense. There is only an ever expanding domain in which different patterns of suffering, identity, and release become concrete, then fade, then return. 

That is not a world governed by essence first metaphysics. It is a world governed by a relentless, indifferent modal structure. In that sense, Inland Empire completes a trajectory. Mulholland Drive leans Finean, because it still clings to the idea that identity and desire are structured by essential misrecognition. Lost Highway leans Williamsonian, because it stages the indifference of being to narrative repair. Inland Empire goes further. It strips away the last illusions of grounding and leaves us with a world that is not haunted by what things are, but by the fact that whatever is there, stays there. That, ultimately, is a necessitist nightmare. And it is why Inland Empire feels so much more exhausting, and so much more honest, than anything Lynch made before. 

But maybe that is too quick! Fine would say that I have been too fast to treat the film’s identity plasticity as hostility to essence. On the contrary, the woman across Inland Empire can be read as a single essence that persists throughout, while what proliferates are dependent manifestations, thinnings and thickenings of that essence. These manifestations are not numerically identical individuals in competition. They are ontologically subordinate entities, grounded in the same fundamental essence but differentiated by their relations to time, narrative role, social normativity, and epistemic position. 

This is exactly the way Twin Peaks: The Return handles doubles, tulpas, and doppelgängers. Cooper, Mr C, Dougie, Richard, these are not new essences popping into being. They are dependent offshoots of the same underlying personhood, generated by different grounding conditions. The essence remains, but it is variously realised, distorted, or partially instantiated. On this view, Lynch is not rejecting essence. He is dramatising its fragmentation under abnormal grounding relations. 

Fine’s metaphysics is unusually well equipped to make sense of this, because he insists that essence is prior to modality and that modal facts can be grounded in essential facts in many different ways. There is not just one kind of necessity. There are logical necessities, metaphysical necessities, essential necessities, and necessities grounded in social or narrative roles. Inland Empire could then be read as a film that deliberately collapses these kinds of necessity into one another, forcing the viewer to experience what happens when essential structure is overdetermined by incompatible grounding relations. 

On this picture, Nikki, Susan, the Polish woman, the prostitute, the Lost Girl are not mere recombinations within a flat domain. They are dependent entities grounded in a single essence whose full instantiation has been damaged. The horror of the film is not that essence is absent, but that it is too fundamental to be cleanly realised. The woman’s essence persists, but it can only appear in fractured, temporally misaligned, normatively broken forms. That is why none of the manifestations feels fully real, and why the film circles endlessly around the question of “who is the woman”. This also lets Fine explain something important that the Williamsonian reading risks smoothing over, namely why the film’s suffering feels personal rather than cosmic. 

The pain in Inland Empire does not feel like the impersonal indifference of being. It feels like the suffering of someone who is being misinstantiated, misgrounded, or trapped in derivative roles that fail to do justice to what she fundamentally is. That is a deeply Finean intuition. It treats the film’s multiplicity not as ontological excess, but as essential violation. This is a very strong position. It shows that Inland Empire can indeed be read as supporting Fine’s pluralism about necessity and his essence first metaphysics. It also shows that my earlier contrast was too quick if it suggested that Fine must treat essence as rigid or monolithic. Fine’s essence is flexible in exactly the right way to allow thinning, thickening, and partial instantiation. 

However, even granting all of that, the decisive issue is not whether Fine can accommodate the film, but whether the film itself behaves as if essence is doing the explanatory work. This is where the Williamsonian pressure reasserts itself. Ask a very precise question: when something goes wrong in Inland Empire, what kind of repair fixes it? If the Finean story were primary, then repair should involve re grounding the essence correctly. The film should move toward a moment where the woman’s true nature is restored, recognised, or properly instantiated, even if only partially. Essence should reassert itself as explanatory bedrock. But that is not what happens. The ending does not restore an essence. It restores access. The Lost Girl is released not because her essential nature is finally honoured, named, or actualised. She is released because a blockage is removed. A door opens. A barrier dissolves. The change is relational, not constitutive. Nothing about the woman’s nature is clarified at the end. We do not learn who she really is. We do not see her finally instantiated as what she essentially is. We simply see her no longer trapped. Fine’s framework predicts that essence, even when fractured, should be explanatorily decisive. But Inland Empire consistently treats explanation as secondary to traversal. You do not fix the world by understanding what something is. You fix it by moving through the world differently. 

This is where the Williamsonian reading has the edge. Williamson is explicitly sceptical of metaphysical demands for deeper grounding when the phenomena can be explained by shifts in accessibility, instantiation, or perspective within a fixed ontology. He thinks philosophers often mistake a craving for explanation for evidence that there must be a deeper level. Inland Empire behaves as if that scepticism is warranted. 

Consider again the idea of tulpas or doppelgängers. In Twin Peaks: The Return, the tulpa logic is indeed Finean friendly. There are clear signs that something is a derivative being, less complete, less grounded, ontologically thinner. Inland Empire never gives us those markers. Nikki’s manifestations are not treated as ontologically weaker than one another. The prostitute is not less real than the actress. The Lost Girl is not less real than Susan. They suffer equally. They act equally. They persist equally. This equality is crucial. Fine’s metaphysics allows for dependent entities, but it does not treat all entities as equally fundamental. Inland Empire does. Or rather, it refuses to tell us which entities are less fundamental in a way that ever matters for how the world behaves. Even the idea that there is one underlying woman is never secured by the film itself. It is a powerful interpretive impulse, but the film never confirms it. There is no moment of essential unification. No “this is who you are” revelation. The film ends without telling us whether there was ever a single woman beneath the multiplicity. That silence is hostile to essence first metaphysics. 

Fine would want to say: the essence is there even if the film does not reveal it. But Inland Empire repeatedly punishes precisely that move. Every time the viewer tries to posit a hidden ground, the film responds by proliferating further scenes, further women, further times. The film does not behave as if there is a concealed essential truth waiting to be honoured. It behaves as if the demand for such a truth is itself part of the trap. This is the key difference from Mulholland Drive, which really does feel Finean, because it gives you a powerful sense that there was a misrecognised essence of desire and identity, and that the tragedy consists in its distortion. Inland Empire withholds that consolation. It does not say “you misunderstood who you were”. It says “there is no level at which that question stabilises”. 

So yes, Fine can give a sophisticated, compelling reading of Inland Empire in terms of a single essence fragmented into dependent manifestations, supported by a plurality of necessity relations. That reading captures the film’s affective intensity and its sense of personal suffering better than a crude Williamsonian flatness would. But when we ask what the film itself treats as explanatorily decisive, the answer is not essence, not grounding, not restoration of nature. It is movement, access, persistence, and traversal within an irreducibly overcrowded reality. That is why, even after taking the Finean reply seriously, Inland Empire still tips Williamsonian. Not because essence is denied, but because essence never does the work that Fine’s metaphysics needs it to do. In Inland Empire, the nightmare is not that the woman’s essence is fractured.The nightmare is that even if there were such an essence, it would not be enough to save her. 

But again, I can see that Fine wouldn’t let things rest there. Let’s look at a powerful second-order Finean move, one that does not merely reinterpret Inland Empire, but challenges the criterion I have used so far to judge between the readings. Let me take this in three steps. First, I’ll restate the Finean counter in its strongest possible form. Second, I’ll show how this shifts what counts as “repair” or “resolution” in the film. Third, I’ll say why, even after accepting this, the disagreement with Williamson is not dissolved but relocated, and why the choice between them ultimately turns on a deep difference about what it means for metaphysics to matter. 

The new Finean move says this: the fact that nothing is “fixed” in Inland Empire is not evidence against essence-first metaphysics, it is evidence for it. If essence precedes necessity, then once essential relations are damaged or reconfigured, there is no route back to an original state via modal rearrangement. You cannot undo trauma by moving to a different possible configuration, because the very space of possibilities has been altered. Release, in this sense, is not restoration but ontological thinning, a permanent loss of structure. 

This is exactly what happens in Twin Peaks: The Return. Cooper’s final attempt to “fix” Laura Palmer does not return the world to an earlier, healthier state. It destroys the coordinates that made that state intelligible. The scream at the end is not psychological failure, but metaphysical one. The essential relations that once held, Laura as victim, Cooper as saviour, Twin Peaks as a place with a history, no longer ground the same necessities. The world has been re-essentialised, and not for the better. On this view, Inland Empire’s ending should not be read as hopeful in any ordinary sense. The Lost Girl’s release is not a healing. It is a collapse of a particular essential configuration. What comes after is not a redeemed world but a thinner one, one in which fewer necessities bind, fewer identities cohere, fewer stories can recur. The rabbits’ calm at the end is ontological exhaustion. The sitcom form has lost its tension because the essential relations that generated it have burned out. 

This reframes my earlier claim that the film restores access rather than essence. Fine would reply that this distinction is misleading. Access is not metaphysically innocent. Access relations themselves are grounded in essence. If the world becomes more traversable at the end, that is because the essential constraints that once structured traversal have been destroyed. Nothing is “repaired” because repair would require reinstating those constraints, and that is precisely what is no longer possible. So the Finean verdict is bleak but coherent: Inland Empire is a film in which metaphysical damage is irreversible. The woman is released not into wholeness but into ontological aftermath. The world continues, but it continues with altered essential joints. That is why nothing feels fully resolved, even when suffering seems to stop. The film shows what happens when essence has been permanently reconfigured. 

Up to this point, this is an extremely strong reply, and it directly answers my earlier objection. It shows that Fine does not need the ending to restore a stable essence in order to claim explanatory priority. Essence can explain why restoration is impossible. At this point, however, the disagreement with Timothy Williamson reappears at a deeper level. The real question is no longer “does the film end in repair or not”. It is “what kind of impossibility is the film ultimately committed to”. Fine’s impossibility is essential. The world cannot be returned to an earlier state because the essential relations that made that state possible no longer obtain. The space of necessity itself has changed. This is a metaphysics of irreversible deformation. Williamson’s impossibility is ontological but non-essential. The world cannot be returned to an earlier state because nothing ever left the ontology in the first place. What looks like loss or thinning is really a redistribution of concreteness and salience. The domain remains maximal. What changes is which structures remain live for agents within it. Notice how subtle the contrast is here. Both views agree that nothing can simply be “fixed”. Both reject naive modal reset fantasies. Both are deeply anti-restorative. But they disagree about why restoration is impossible. 

Fine says it is impossible because the essence has changed, and essence is prior to modality. Williamson says it is impossible because the ontology never changed, and modality never had the power to erase what was there. This difference shows up if we ask a very precise question: does the film treat loss as subtraction or as reconfiguration? In Twin Peaks: The Return, loss really does feel subtractive. Places disappear. Names fail. Laura vanishes into a scream that seems to take the world with it. That is why the Finean reading fits that series almost perfectly. The essential map has been damaged beyond repair. 

In Inland Empire, by contrast, loss almost never functions subtractively. Scenes do not vanish so much as drift. Characters do not disappear so much as become inaccessible. Even when something seems to end, it tends to persist elsewhere, in another register, another screen, another woman, another voice. The world does not feel thinned in the way The Return does. It feels congested, saturated, overdetermined. This is the crucial experiential difference. 

Inland Empire is exhausting because it leaves us with too much. Even when the woman is released, the film does not present a world with fewer beings or fewer structures. It presents a world in which the pressure has shifted, but the density remains. Fine can say that this density reflects lingering dependent entities whose grounding relations have collapsed. But the film never marks them as ontologically residual or second-order. They retain full presence. They are not ghosts of an essence that has failed. They are just there. This is where Williamson’s framework continues to have the upper hand. Williamson does not need to posit irreversible essential deformation to explain why nothing can be fixed. He only needs the claim that the ontology is never pruned. Trauma does not damage essence. It reorganises which parts of reality agents can inhabit without suffering. So the adjudication now looks like this. Fine gives the best account of why attempts at restoration fail in a metaphysically tragic sense. He captures the sense that there is no return, no innocent present, no untouched structure waiting beneath the damage. That is a genuine strength, and it explains why Inland Empire feels kin to Twin Peaks: The Return at the level of affect. 

Williamson gives the best account of what remains once restoration fails. He captures the film’s refusal to let anything be over, erased, or metaphysically downgraded. He explains why the film’s world feels relentlessly populated even after its crises, and why the ending feels less like devastation than like uneasy coexistence. The final choice between them depends on what you take to be the film’s deepest claim. If you think Inland Empire is fundamentally about irreversible metaphysical damage, then Fine is right, and the film belongs with The Return as an exploration of what happens when essence itself is broken. If you think Inland Empire is fundamentally about the impossibility of erasure, about the way reality accumulates rather than heals, then Williamson is right, and the film is about surviving in a world where nothing ever fully goes away. 

My own judgement at this stage, stated as carefully as possible, is this: Inland Empire borrows Fine’s tragic insight but stages it in a Williamsonian universe. It shows us what it is like to experience something that feels like essential devastation, while inhabiting a reality that refuses to thin accordingly. That mismatch, the persistence of being in the face of shattered meaning, is what makes the film uniquely unbearable. Fine explains the shattering. Williamson explains the persistence. And Lynch, as ever, refuses to give us the comfort of choosing only one. 

But I think Fine might push harder. I can think of two further different but converging Finean criticisms. First, the charge that Williamson illegitimately ignores affective necessity. Second, the charge that necessitism secretly smuggles in an essence of the world, thereby collapsing into a Finean position it claims to reject. Taken together, these are the hardest things that can be said against the Williamsonian reading, and they deserve to be met head on. 

Fine’s pluralism about necessity is a core methodological claim. He thinks philosophers go wrong when they treat metaphysical necessity as the one true modality and demote all others, epistemic, normative, practical, affective, to second class status. On Fine’s view, different kinds of necessity answer to different kinds of grounding. What is necessary in virtue of an essence is not the same as what is necessary in virtue of a social role, a norm, a narrative structure, or an affective economy. When you apply that to Inland Empire, the point bites immediately. 

The film is saturated with dread, terror, grief, exhaustion, and ontological panic. These are not incidental. They are not psychological colour layered on top of a metaphysical skeleton. They are the skeleton. The film’s structure is felt before it is understood. The dread is not a reaction to metaphysical facts. It is part of the metaphysical situation itself. Fine would say that this dread is evidence. It is data. It tells us something about what is necessary in the world of the film. Specifically, it tells us that certain configurations cannot be endured, not merely that they cannot be escaped. The terror in Inland Empire has a modal character. It has the feel of “this cannot go on”, “this must end”, “this cannot be otherwise”. That is affective necessity. It is not captured by quantification over objects or by distinctions between existence and concreteness. It is captured by the way the world presses on the subject as inescapable and intolerable. 

Here Fine’s criticism of Williamson is sharp. Williamson explicitly warns against treating intuitions, including modal intuitions, as anything more than defeasible data generated by cognitive heuristics. He thinks that strong feelings of necessity often track psychological shortcuts rather than metaphysical structure. Fine thinks this is exactly backwards in cases like this. In Inland Empire, the affective pressure is not a distortion of the metaphysics. It is the manifestation of it. From that perspective, the Williamsonian reading really does flatten the film by abstracting away from what is doing the modal work. 

When Williamson says that the world’s horror lies in the persistence of a fixed ontology, Fine replies that this misses the point. The horror is not that everything exists. The horror is that certain relations must hold, certain roles must be played, certain suffering must occur, until the essential configuration breaks. Those “musts” are not captured by quantifiers. They are captured by affective necessity grounded in essence. If you ignore that, Fine would say, you are not being more metaphysically rigorous. You are discarding data in order to preserve a preferred formal fit. And for Williamson, who insists that logic must answer to the world rather than the world being trimmed to fit logic, this is a devastating charge. 

Now turn to the second point, which is even more destabilising. I suggest that Fine might suggest that once Williamson says “everything necessarily exists”, he has effectively posited an essence, namely the essence of the world as totality. The world’s essence becomes: containing everything. Once that move is visible, Fine’s framework seems to reappear inside Williamson’s. This is not a cheap gotcha. It goes straight to the heart of what necessitism is doing. 

Williamson insists that necessitism is not essentialism. He wants to say that necessary existence is not a property grounded in what things are, but a structural feature of quantification and modality. Things necessarily exist not because it is in their nature to do so, but because the best modal logic quantifies over a fixed domain. Fine would reply that this distinction cannot be sustained. If something exists necessarily, then there is something about it in virtue of which it exists necessarily. You can call that “structure” rather than “nature”, but the metaphysical role is the same. You have posited a feature that holds in all possibilities and constrains what can be otherwise. That is exactly what essence is supposed to do. 

Once you apply this to the world as a whole, the problem sharpens. Williamson’s world has an invariant domain. That invariance functions like an essence of reality. It tells you what cannot be lost, what cannot be erased, what cannot fail to be. Fine would say: you have simply relocated essence from objects to the totality. You have not eliminated essentialism. You have globalised it. And once that is seen, Fine’s metaphysics rushes back in through the front door. If the world has an essence, then questions about grounding, dependence, and kinds of necessity become unavoidable again. You cannot say “everything exists necessarily” without asking in virtue of what this is so. And once you ask that question, modal logic alone is no longer enough. You need a story about what the world is. 

This is where Inland Empire becomes decisive again. If Williamson’s necessitism were really free of essence, the film’s world would feel indifferent but neutral. Everything would be there, but nothing would demand anything. Yet the film is anything but neutral. Certain configurations scream to end. Certain scenes feel metaphysically wrong, not merely unpleasant. Certain recurrences feel compulsory, not merely persistent. Fine would say: that is because the world’s essence is not “everything exists”, but something like “this story must play itself out until its essential relations collapse”. The necessity in the film is not ontological maximality. It is tragic inevitability. That inevitability is not global. It is local, structured, layered. It comes and goes. It breaks. This is exactly why Fine insists on plural necessity. The affective necessity that drives the film toward release is not the same as the ontological necessity that keeps entities from vanishing. 

The Williamsonian reading collapses these into one by treating ontological persistence as the primary metaphysical fact. Fine refuses that collapse. At this point, the disagreement cannot be resolved by appealing to scenes, motifs, or narrative structure alone. It turns on a prior commitment about what metaphysics is allowed to explain. Williamson wants metaphysics to explain structure while remaining austere about grounding, affect, and essence. He is willing to say that some features of our experience, even very powerful ones, are not metaphysically fundamental. Fine wants metaphysics to explain why those experiences have the modal force they do. He is willing to complicate necessity, essence, and grounding to avoid flattening the phenomena. 

In the context of Inland Empire, this leads to a final, uncomfortable conclusion. If you think that the dread, devastation, and affective overload of the film are merely reactions to an underlying ontological situation, then the Williamsonian reading remains viable, even elegant. If you think that those affects are themselves modal data, telling us what must hold and what must break, then the Williamsonian reading does exactly what Fine warns against. It preserves formal clarity by discarding the very necessities the film is staging. And once you see that Williamson’s “everything exists necessarily” functions like a world essence, the claim that his reading avoids Fine’s metaphysics becomes much harder to sustain. 

So the deepest truth may be this. Inland Empire exposes a limit case for Williamson. It shows a world in which ontological persistence is not the dominant modal fact. The dominant modal fact is that certain structures cannot be lived with. That is not a truth about what exists. It is a truth about what must end. Fine has the better tools for that. Which means that if Lost Highway was Lynch’s most Williamsonian film, Inland Empire may be the film where Williamson’s metaphysics finally breaks, not because it is incoherent, but because it is too thin to carry the affective necessity the film insists upon. 

But is it right that Williamson’s necessitism plus his preferred modal logic quietly forces a kind of essentialism at a higher level, an essentialism not of particular things so much as of the domain and its modal profile? And if he does is Fine entitled to say, “once you have done that, you have reintroduced my explanatory structure, you have just hidden it in the background”. 

If that charge lands, it lands like this. Williamson’s flagship thesis, necessitism, is often given the vivid gloss “everything necessarily exists”, though he is careful about the verb “exists” and prefers a formulation in terms of unrestricted quantification and the claim that necessarily everything is something. The point is constant domain quantification across worlds, in the intended metaphysical reading.   In that setting, contingency is pushed away from being to being thus, from what there is to how things are. That already looks structurally like an essentialist manoeuvre, because you are drawing a firm line between a level that is modally fixed and a level that varies. The natural next thought is, “is that line just the line between essence and accident, except applied to the world’s inventory rather than to Socrates”. 

Fine’s way into the debate is to insist that essence is not just another kind of necessity, or necessity in disguise. Essence is a more discriminating, more hyperintensional kind of explanation than any merely modal claim can deliver. That is the centre of his canonical complaint against the “assimilation of essence to modality”. Modal facts, for Fine, are downstream of essential facts, not upstream. So when he hears a thesis like necessitism, he is likely to hear, not an absence of essence talk, but an essence claim stated in modal garb, something like: it is essential to reality, or to the domain, that it contains every object that it contains, regardless of which world is actual. 

That is my “Williamson is an essentialist at a different level” worry. In a Lynch register, it is the feeling that the film is insisting on a fixed cast list even while it lets the roles, scenes, and identities slide. Even when the story fractures, the production still had to have the actors contracted. The continuity is not in the narrative but in the inventory. That is what necessitism can look like from a Finean angle. 

Now, a Williamsonian reply begins by refusing the explanatory direction. The slogan “essence precedes necessity” is, for him, not a discovery but a methodological temptation, a way of mistaking a tool of description for a source of metaphysical ground. In Modal Logic as Metaphysics, the stance is that quantified modal logic gives you a disciplined framework in which to articulate and test metaphysical generalisations, rather than first installing a primitive essence operator and using it to legislate modal space. Williamson does not have to deny that we can talk in essentialist ways. He can say that essentialist idioms are at best derivative glosses on modal and non modal facts, or heuristic summaries of patterns we find once we do the systematic work. 

This connects to a second, subtler Williamsonian move. He will say that the necessitist constant domain is not an essence thesis but a semantic and logical constraint motivated by the overall theoretical virtues of the resulting system, including how it handles quantification into modal contexts, Barcan style principles, and the costs of the alternatives. Put differently, Fine wants to hear “ground”, Williamson wants you to hear “best theory”. If you accuse him of smuggling in an essence of reality, he answers that you are projecting a Finean metaphysical explanatory appetite onto what is, in his eyes, a theory choice constrained by logic, language, and metaphysical simplicity, a package deal rather than a hidden primitive. 

But can Williamson really keep the essence like work from reappearing, because the package he endorses seems to need something that plays the essence role, even if it is not named. Here is the pressure point. Fine’s essence claims do two things at once. First, they are selective. They sort necessities into kinds. Fine is explicitly a modal pluralist about necessity, arguing that there are different irreducible necessities, for example metaphysical, natural, normative, with different sources. That pluralism gives you conceptual space for what I’ve called “affective necessity”, the thought that some constraints in a work like Inland Empire feel not optional, not because of physics or metaphysics, but because of the film’s own normative and affective structure, its demands on attention, dread, compulsion, shame, fascination. 

A Finean can say, if the data include the film’s affective architecture, you are not allowed to flatten that architecture into a single modal operator without losing the phenomenon. The film is not merely a set of counterfactual plot rearrangements, it is a system of felt constraints. 

Second, Fine’s essence claims are explanatory. They explain why something is necessary by tracing it to what a thing is, not merely to what holds across possible scenarios. That is why he thinks modal equivalence is too coarse grained. The classic form is that two propositions can be necessarily equivalent (eg ‘Socrates is a man’, and ‘Socrates is a man and 2+2=4’), yet one is essential to an object and the other is not, and that shows essence is hyperintensional, not reducible to necessity. Once you see those two functions, you can restate your charge in its strongest form. 

Williamson’s necessitism is not merely a thesis about what exists. It is a thesis about what cannot fail to be in the domain. That looks like an explanatory stopping point. If someone asks, “why are those objects always in the domain”, the necessitist can say “because necessarily they are something”. But that is a modal answer, and Fine will say it is not explanatory, it is a re description at the same level. The Finean will then propose: you are treating the domain itself as having an essence, even if you do not call it that, and you are allowing that essence to do explanatory work by refusing further “why” questions. The essence is “everything”, the domain’s whatness is maximal, so of course everything is in it. Your metaphysics has become essentialism where the essence of the world is totality. That is the hard version of the objection. It is that Williamson is committed to a monolithic world essence that trivialises the Finean project by absorbing it, but in doing so he loses the very discriminations that essence talk was introduced to capture. 

How would Williamson respond carefully, rather than with a dismissive “I just do not do grounding”? He has a few options, and the most interesting is to meet Fine on the hyperintensional battlefield without giving away the anti exceptionalist, logic first posture. One response is to concede that there are hyperintensional distinctions, but deny that they require primitive essences. For example, one can accept that not all necessary equivalences are cognitively, explanatorily, or representationally alike, while insisting that those differences are not metaphysically fundamental. They belong to language, concepts, or theorising practice, not to the modal structure of reality. This is the kind of manoeuvre that also appears in Williamson’s critical engagement with Fine on essence, where he targets Fine’s style of argument for hyperintensionality and tries to diagnose it as driven by unreliable heuristics. If that diagnosis works, then the Finean inference from “there is a distinction” to “there is a primitive essence operator in reality” is blocked. 

A second response is to say that even if Fine is right that essence precedes necessity, that does not refute necessitism, because necessitism is not trying to explain modal facts in Fine’s way. It is a claim about the correct overall semantics and logic for metaphysical modality when paired with unrestricted quantification, and it is defended by theoretical virtues, by how the system behaves as a whole. The fact that Fine can redescribe necessitism in essentialist language does not show that Williamson has smuggled in Finean primitives. It shows only that there are multiple vocabularies for the same commitments. 

But Fine can tighten the screw. He can say, “if you admit that redescriptions are possible, then you have conceded that your framework has the resources to state essentialist commitments, and the question becomes whether your framework can match the explanatory and discriminating power of mine”. Here my “affective necessity” complaint becomes a methodological accusation that Williamson’s monistic modal operator, metaphysical necessity, plus a constant domain, makes the theory too coarse grained to respect the phenomena, whether the phenomena are about essence, about explanatory asymmetry, or about art’s internal norms. 

This is exactly where Williamson’s own methodological line, “fit the metaphysics to the data, not the data to the metaphysics”, can be turned back on him. Modal Logic as Metaphysics presents itself as inquiry driven, not stipulative. So the Finean can say, if your interpretation strategy for Inland Empire forces you to ignore, or redescribe away, whole categories of constraint that the work itself makes salient, then you are doing the wrong kind of fitting. You are compressing the domain to make your operator look good. Williamson will not be happy with talk of “affective necessity” as necessity. He will likely treat it as either metaphorical, or as belonging to the normative and psychological registers rather than the metaphysical one. But Fine’s pluralism is precisely designed to resist that kind of gatekeeping. If there are many necessities, there is no reason to assume in advance that only one of them is philosophically serious. 

A Lynch film then becomes a neat test case, because it foregrounds that there are constraints that bite with the force of necessity for the viewer, even though they are not the constraints of physics or logic. 

So who wins, on the narrow charge that Williamson is just an essentialist at a different level? My sense is that Fine can make the charge stick only if he insists on a strong principle: any respectable metaphysics must supply Fine style explanatory grounds, not merely state modal truths in a regimented logic. If you grant that principle, then necessitism will look like an explanatory stopping point that behaves like an essence claim about the domain, namely that the domain is essentially maximal. From that standpoint, Williamson has not escaped essence, he has inflated it, and the Finean world reappears as soon as you ask the “why” questions that necessitism answers only modally. 

But Williamson can resist the principle itself. He can say: your demand for a deeper ground is a philosophical itch, not a constraint on theory. The right comparison is not “who has the more satisfying ground”, but “which overall package yields the most systematic, simple, and powerful theory of modality and quantification, consistent with our best logical practice”. On that score he can defend necessitism without thereby becoming a Finean essentialist, because he is not in the business of positing primitive essences to underwrite modality in the first place. 

Where that leaves the films, and this is where my “tulpas, thinning and thickening, altered normativity” intuition becomes philosophically sharp, is that Fine and Williamson are not merely offering two readings of the same cinematic data. They are offering two standards for what counts as respecting the data. A Williamsonian reading wants a clean separation between existential inventory and contingent narrative configuration. That suits Lost Highway: the cast is fixed, the narrative roles permute, identity is tracked by something like cross world counterpart structure under a rigid domain, and the horror is how much can vary while being itself is held steady. A Finean reading insists that the very identity conditions and explanatory structure are part of the metaphysical story, and that the film is showing essence like constraints being deformed, not merely accidental properties being rearranged. Those deformations can be normative and affective as well as metaphysical, because necessity comes in kinds. 

So the hardest question is this: is Williamson’s framework sufficiently fine grained to represent and explain the distinctions that Fine calls “essential”, without quietly relying on an essence of the domain, and without flattening the film’s internal necessities into mere psychology or rhetoric? Williamson’s careful answer is: I can respect the distinctions as distinctions in representation and theory choice, while keeping the metaphysics lean, and I do not need primitive essences to do that, and I do not need to count every felt constraint as a genuine necessity. Fine’s careful answer is: if you refuse to treat at least some of those distinctions as metaphysically explanatory, you have not merely chosen a leaner theory, you have changed the subject, because the subject was precisely what makes necessities necessary, and you have replaced it with a single operator and a constant domain whose very maximality functions as a hidden essence. 

That is the point where, in Inland Empire, Fine would say the attempt to “return” to a pre traumatised present is not just contingently thwarted. It is blocked by altered essential relations, by the fact that the space of necessities itself has been re carved, and therefore nothing can be fixed in the old way, even if you try to fix it by saying everything necessarily exists. And Williamson would answer that this is a powerful interpretive metaphor, but it is not an argument against necessitism unless you can show that the best overall modal theory must make room, at the level of metaphysical modality, for those hyperintensional and plural constraints. 

Another possible Finean move would be to pivot from Williamson’s own methodological self-description rather than attacking him from the outside. Let me lay it out carefully, because this is where the disagreement stops being about modality alone and becomes about what philosophy is allowed to be about at all. 

Williamson repeatedly insists that philosophy is continuous with science, but not in the crude sense that it is a junior partner of physics. Philosophy, for him, is a theoretical discipline governed by broadly scientific norms, clarity, systematicity, explanatory power, openness to revision, even when its subject matter is abstract. Logic is not exceptional, metaphysics is not exceptional, epistemology is not exceptional. That is the anti-exceptionalist credo. But crucially, he is explicit that philosophy is not a natural science. It does not discover laws of nature, it does not rely on controlled experiments, and it does not reduce to empirical generalisation. Its subject matter includes abstract structure, necessity, modality, meaning, knowledge. 

Now Fine’s move is to say: good. If that is the methodological posture, then Williamson cannot help himself to a natural-science style austerity about ontology when the phenomena under investigation are not natural-scientific phenomena. English literature and history are excellent examples here, and Fine would happily accept them as philosophy-adjacent disciplines in Williamson’s sense. They are theory-laden, interpretive, evidence-sensitive, revisable, and non-naturalistic. But they are not ontologically thin. They traffic in norms, meanings, structures, roles, obligations, styles, genres, forms of life. Crucially, they take these things seriously as part of what there is to be explained, not as mere projections to be eliminated once the “real” ontology is fixed. 

So the Finean pressure point is this: if philosophy is like science but not like physics, and if literature, history, and related disciplines count as legitimate theoretical inquiries, then why is Williamson allowed to dismiss plural necessities, grounding relations, and essences as metaphysically second-rate without argument, rather than as part of the ontology those inquiries are already committed to? Fine would say that Williamson’s attitude to necessity looks suspiciously like a natural-science hangover in a domain where natural-scientific minimalism is inappropriate. In physics, we have good reasons to compress ontology, to eliminate explanatory redundancies, to seek unifying laws. In literary criticism, history, ethics, aesthetics, and large parts of metaphysics, we do not. There, explanatory adequacy often requires multiplying kinds of necessity, kinds of dependence, kinds of constraint, because the phenomena themselves are structured that way. When Williamson responds to Fine’s hyperintensional distinctions by saying that they are “just” differences in representation, or that they are generated by unreliable heuristics, Fine can reply: that response makes sense if your paradigm of evidence is physical science, where feelings, norms, and roles are indeed secondary. But it makes much less sense if your paradigm includes history and literature, where the “felt necessity” of events, the normative inevitability of outcomes, and the internal logic of narratives are precisely what theory is trying to capture. 

This is where Inland Empire becomes philosophically diagnostic. The film does not present its necessities as optional glosses layered on top of a neutral ontology. The dread, compulsion, exhaustion, and inevitability are not reactions to metaphysical facts. They are the facts. The film’s internal economy treats certain developments as unavoidable in a way that is not reducible to metaphysical necessity in Williamson’s sense. They are unavoidable because of how roles, narratives, trauma, and repetition function. Those are not psychological accidents. They are structural features of the world the film constructs. Fine would say: if you take philosophy to be continuous with interpretive disciplines like literature and history, then you are not allowed to say, “those necessities are merely affective, therefore not metaphysical”. 

That is exactly the kind of dismissal Williamson warns against when he criticises exceptionalism. You are declaring in advance that only one kind of necessity counts, and then accusing other kinds of being illusory. Fine’s critique is not saying Williamson is wrong to value simplicity or formal clarity. He is saying Williamson has not earned the right to treat his modal operator as the only metaphysically serious one. Williamson might reply that philosophy, even when it is not a natural science, should still aim for the leanest theory that explains the data, and that plural necessities overfit. But Fine’s counter is that “the data” here include normative and affective constraints that do not behave like noise. 

In Inland Empire, for example, the sense that “this must end”, “this cannot continue”, “this repetition is intolerable”, is not a subjective add-on. It governs the film’s structure. A theory that cannot represent that necessity without translating it into mere psychology is, Fine would say, explanatorily inadequate. And this loops back to the earlier charge that Williamson’s necessitism may be essentialism in disguise. If Williamson insists that all genuine necessity is metaphysical necessity, and that metaphysical necessity is captured by quantification over a fixed domain, then the world’s modal profile becomes monolithic. Everything that cannot be otherwise must be so in the same way. 

Fine’s pluralism rejects that. It says: the way in which a murder must be punished, the way in which a narrative must repeat, the way in which a trauma must reassert itself, and the way in which an object must exist are not the same kind of “must”. Once you accept that, Williamson’s world-level necessitism starts to look like an ontological flattening, not just a formal simplification. It treats all necessities as if they were of one kind, and it treats the domain as if its maximality were explanatorily innocent. Fine’s accusation is that this is not neutrality. It is a substantive metaphysical bet that sidelines whole classes of necessity by refusing to recognise them as such. 

Williamson would insist that recognising a phenomenon as real or important does not require treating it as metaphysically fundamental. One can take literary necessity, historical inevitability, and affective compulsion seriously as objects of study without building them into the metaphysics of modality. They can be explained in terms of human practices, psychological architecture, social norms, narrative conventions, without positing new kinds of necessity at the metaphysical level. 

Second, he would argue that Fine’s pluralism risks collapsing metaphysics into a catalogue of explanatory idioms, each with its own necessity operator, without a clear account of how these relate or why they deserve equal ontological status. Williamson’s worry is not that pluralism is false, but that it blurs the line between metaphysics and description. If every felt constraint is elevated to a necessity, metaphysics loses its discipline. 

Third, and this is the most important move, Williamson would say that philosophy’s continuity with science does not mean continuity with every interpretive practice. Philosophy is theory-building, not criticism. It can draw on literature as data, but it is not obliged to reproduce literature’s internal norms as metaphysical primitives. The fact that Inland Empire feels necessary in certain ways does not show that those necessities are metaphysically on a par with existence or identity. It shows that human cognition responds powerfully to certain structures. 

Fine would say that Williamson is now quietly reinstating a hierarchy of domains, where natural-science-adjacent necessities are metaphysically respectable and others are downgraded to “merely” psychological or social. That hierarchy is exactly what Fine’s work is meant to dismantle. If philosophy is not physics, then why privilege physics-friendly necessity? Williamson thinks metaphysics should aim for a unified, austere account of modality that interfaces cleanly with logic and quantification, even if that means treating many compelling experiences of necessity as non-fundamental. Fine thinks metaphysics should aim to articulate the different ways the world constrains itself, even if that means giving up unification and accepting plural, irreducible necessities grounded in different kinds of essence.   

Clearly Inland Empire does not settle that dispute. But it does something important. It shows that if you ignore affective and narrative necessity, you miss what the work is about. And it shows that if you insist on a single modal register, you will be tempted to redescribe away exactly the constraints that make the world of the film intelligible. 

So the Finean move is a challenge to Williamson’s self-image as an anti-exceptionalist. If philosophy is continuous with interpretive disciplines, then metaphysics must be open to the kinds of necessity those disciplines reveal. If Williamson refuses that openness, Fine can say that he is not being more scientific, he is being selectively austere, and that selectivity itself is a substantive metaphysical commitment that stands in need of justification. 

That, ultimately, is why Inland Empire suddenly looks like a dangerous film for Williamson. It does not refute necessitism. It asks whether necessitism has the expressive resources to describe a world in which necessity is felt, layered, and destructive, rather than merely structural. And Fine’s answer is: not without becoming Finean. 

Fine’s further move would be this. Williamson’s necessitarianism says, at its core, that everything necessarily exists, or more carefully, that the domain of quantification is fixed across modal space. Williamson wants this to be ontologically austere rather than bloated, because he thinks it avoids mysterious “coming into existence” and “going out of existence”, although it is ontologically expansive as well, given that it has all possibilia existing too. But Fine can now point out that this very austerity creates a problem once Williamson widens philosophy beyond the natural sciences. If everything necessarily exists, then psychological facts, social roles, historical structures, linguistic conventions, aesthetic norms, moral obligations, genres, narratives, institutions, practices, and affects all necessarily exist too. They are not optional extras. They are not metaphysical afterthoughts. They are in the domain. 

At that point, Fine presses: what principled right does Williamson have to say that some of these entities are metaphysically salient and others are not? Williamson cannot say “they don’t really exist”, because necessitism forbids that. He also cannot say “they exist but are metaphysically irrelevant” without explaining what metaphysically relevant means, because relevance itself now cannot be defined in terms of existence or non-existence. 

This is the key twist. Necessitism removes Williamson’s most obvious exit. Under a contingentist ontology, one can say: some things exist in some worlds but not others, and metaphysics concerns itself primarily with those existence conditions. Under necessitism, existence is no longer a discriminator. Everything is always there. So metaphysical salience must be drawn somewhere else. Fine’s accusation is that Williamson has not given a non-question-begging account of where that line is. 

This connects directly to Williamson’s anti-exceptionalism. Williamson explicitly says philosophy is not a natural science like physics, but is continuous with inquiry more generally. He regularly treats philosophy, mathematics, logic, linguistics, history and cognitive science as legitimate theoretical disciplines. He also rejects the idea that metaphysics is a special, insulated armchair enterprise. All of that pushes in the direction of inclusivity, not restriction. 

So Fine can say: if philosophy is like science but not physics, and if you accept that history, literature, and art are disciplined inquiries into reality rather than mere projections, then you owe us an explanation of why the necessities internal to those domains are not metaphysically serious. Williamson wants to say that metaphysical necessity is one thing and psychological, social, narrative, affective “necessities” are something else and further that only the first belongs to metaphysics proper. But Fine thinks that distinction cannot be drawn by appeal to existence, because necessitism equalises existence. It also cannot be drawn by appeal to mind-independence, because social institutions, languages, and histories are mind-dependent yet objectively real. It cannot be drawn by appeal to normativity either, because moral and logical norms already play a central role in Williamson’s epistemology. So where exactly is the boundary? 

If Williamson says “metaphysical necessity is necessity that concerns what exists and what properties things have”, Fine answers: social roles, historical structures, and psychological states are properties of things that exist necessarily. If Williamson says “metaphysical necessity is necessity captured by modal logic”, Fine answers: that is stipulative unless you explain why modal logic has priority over other necessity-tracking practices. 

Fine can say that Williamson’s necessitism forces a kind of metaphysical flattening, because it removes existence as a discriminating variable while refusing to admit alternative forms of necessity as metaphysically explanatory. The result is a world in which everything is equally real but only some things are allowed to matter metaphysically, and the criterion for mattering is left implicit. And once you try to make that criterion explicit, you end up doing essence-talk again. You end up saying things like: “It is in the nature of metaphysical reality that only structural features matter,” or “It is in the nature of modality that only world-invariant features count” or “It is in the nature of explanation that some dependencies are deeper than others.” 

For Fine, those are essence claims. They may not be object-level essences like “Socrates is essentially human”, but they are meta-essentialist commitments about what the world fundamentally is like. In particular, they amount to something like: it is essential to reality that it be exhaustively describable by a single modal framework. That is precisely the sort of claim Fine thinks metaphysics should argue for, not assume. 

So Fine argues that Williamson says three things: Everything necessarily exists. Philosophy is a science-like inquiry broader than physics. Metaphysics should be austere and unified. And Fine responds by saying: If everything necessarily exists, then all domains of inquiry concern real entities. If philosophy is broader than physics, then affective, narrative, social, and historical necessities are data, not noise. If you refuse to treat those necessities as metaphysically salient, you are drawing a metaphysical boundary you have not justified. And if you try to justify it, you will inevitably appeal to essential features of reality, collapsing your view back into a form of essentialism. 

Williamsonian won’t deny that these domains exist, but he’ll deny that metaphysics must mirror every legitimate explanatory practice. He can say that metaphysics aims to describe the most general structure of reality and that other disciplines describe local, domain-specific structures. He can go on to say that the fact that a structure is real does not mean it belongs to the most general level. In other words, metaphysical salience is not about existence, but about generality. He’d say something like: Psychological, social, historical necessities are real but parochial. They depend on contingent features of agents, cultures, and practices. Metaphysical necessity, by contrast, concerns what holds regardless of such features. 

Fine’s reply to that would be to argue that Williamson just reintroduces necessity as the criterion of metaphysicality, which is circular unless you already know which necessities count. Moreover, history and language also exhibit necessities that hold regardless of individual psychology. Narrative inevitability, institutional constraint, and trauma repetition are not local quirks. They structure worlds. This seems to be a deep fault line between the two viewpoints. Williamson believes metaphysics should be maximally general and minimally expressive.

Fine believes metaphysics should be maximally explanatory even at the cost of plurality. Necessitism magnifies this conflict because it removes the usual ontological escape hatches. Once everything is necessarily there, the only way to avoid Fine’s pluralism is to declare large regions of reality metaphysically second-rate by fiat. 

That is why Inland Empire is such a perfect stress test. The film constructs a world in which psychological, narrative, affective, and historical constraints are not optional overlays but the very engines of inevitability. A theory that says “those constraints are real but not metaphysical” starts to sound like it is changing the subject. The result is not that Williamson is refuted, but that he is forced into a choice of either articulating a principled, non-circular account of metaphysical salience that does not rely on hidden essentialist commitments, or accept that his necessitism, once broadened beyond natural science, naturally expands into a Fine-style pluralist metaphysics of grounding and necessity. 

Now the dispute no longer looks like a disagreement between two modal theories. It looks like a disagreement about whether metaphysics should discipline reality or listen to it. If we step back and situate the Fine–Williamson dispute against the longer philosophical landscape from Immanuel Kant onwards, what becomes visible is that this is not a local disagreement about modal logic or necessitism at all. It is a re-emergence, in technically sophisticated form, of a fault line that runs through post-Kantian philosophy: the struggle to decide whether necessity belongs primarily to structure or to constitution, to form or to ground, to conditions of intelligibility or to the natures of things. Seen this way, Fine and Williamson are not opposites in any crude sense. They are heirs to different post-Kantian lineages that overlap, cross, and repeatedly attempt to absorb one another. 

Start with Kant, because he fixes the terms of the debate even where neither Fine nor Williamson would endorse his framework wholesale. Kant’s decisive move is to deny that metaphysical necessity can be read directly off things as they are in themselves. Necessity, for him, is bound up with the conditions under which objects can be experienced, judged, and thought. Modal notions are therefore tied to forms of intuition, categories of the understanding, and the unity of apperception. The necessity of causation, substance, or identity is not discovered in the world but imposed by the structure of cognition. That already splits necessity from empirical contingency while refusing to locate it in the essences of noumenal things. 

Williamson inherits something important from this Kantian move, even though he totally rejects Kant’s transcendental idealism. The inheritance is methodological rather than doctrinal. Williamson treats modal structure as something we get at through disciplined theorising about representation, inference, and quantification, not through direct metaphysical inspection of essences. His anti-exceptionalism echoes Kant’s insistence that logic and necessity are not magical access points to reality but are governed by the same standards of theory choice as other inquiries. Modal logic, for Williamson, is not a window into noumena, but a formal articulation of the best way to regiment necessity talk given our inferential practices. 

However, Williamson rejects Kant’s restriction of necessity to the conditions of possible experience. He relocates necessity from transcendental subjectivity to objective modal reality. In this sense, Williamson is post-Kantian but anti-Kantian. He keeps the emphasis on structure and form while discarding the transcendental subject. 

Fine, by contrast, pushes back toward something Kant explicitly tried to block: the idea that necessity is grounded in what things are. But Fine does not simply revive pre-critical essentialism. His move is subtler and much more post-Kantian than it might first appear. Fine agrees with Kant that modal notions cannot be reduced to empirical generalisation. But he rejects Kant’s idea that necessity is exhausted by the conditions of representation. Instead, Fine introduces a layered metaphysics of grounding that sits between Kantian form and Aristotelian substance. 

Essence, for Fine, is not a mysterious inner core accessible by metaphysical intuition. It is a way of articulating what explains why certain necessities hold. This is why Fine insists that essence is hyperintensional: it tracks explanatory priority, not merely modal invariance. Historically, this places Fine closer to the German Idealist and neo-Aristotelian traditions than to Kantian formalism. There are clear affinities with Hegel, though Fine would reject Hegel’s dialectical metaphysics . The affinity lies in the idea that necessity is not merely a matter of logical form but of internal relations. For Hegel, what is necessary follows from what a thing is within a system of relations. For Fine, what is necessary follows from what a thing is in virtue of its essence and grounding relations. The scale is different, but the explanatory impulse is recognisably similar. 

The divergence between Fine and Williamson also mirrors a split that opens up in nineteenth-century philosophy between formalist and constitutive approaches to reason. On one side you have figures like Gottlob Frege, for whom logic provides an objective structure that constrains thought independently of psychology, history, or culture. On the other side you have traditions that insist that meaning, normativity, and necessity are inseparable from practices, roles, and forms of life. 

Williamson is very much a Fregean heir in this respect. His commitment to classical logic, to bivalence, to quantification over a fixed domain, and to the idea that modal facts can be regimented without appeal to grounding relations, places him firmly in the tradition that treats formal structure as metaphysically revealing. His necessitism is, in this sense, a late Fregean move: once the logical form is fixed, the ontology follows. 

Fine, although perfectly comfortable with logic, is closer to the post-Fregean reaction that worries about what formal equivalence leaves out. Here the comparison with Edmund Husserl is instructive. Husserl argued that logical laws alone cannot capture the intentional and essential structures of meaning. Fine’s complaint that modal equivalence is too coarse grained is a distant cousin of Husserl’s complaint that logical form abstracts away from what makes meanings what they are. 

Move forward into twentieth-century analytic philosophy and the contrast sharpens further. Williamson is a direct descendant of the Kripkean revolution, but a specific strand of it. Saul Kripke reintroduced necessity and essence into analytic metaphysics by arguing that some truths are necessary a posteriori (e.g. water is H20) and that objects have essential properties, such as origin. Fine famously challenges exactly this assimilation of essence to necessity. Where Kripke says essence just is necessity across possible worlds, Fine says this misses explanatory asymmetry. 

Williamson sides with Kripke against Fine on this point, but with an important twist. He radicalises the Kripkean framework by making the domain constant and embracing necessitism. In doing so, he strips essence of its traditional role and tries to let modal logic carry the metaphysical weight. That is why Fine’s charge that Williamson has merely relocated essence to the level of the domain is so pointed. It echoes older criticisms of structuralist metaphysics, namely that structure without constitutive explanation becomes dogmatic. Fine’s position, by contrast, aligns him with the neo-Aristotelian revival in contemporary metaphysics, alongside figures like Aristotle (anachronistically), but also modern thinkers who emphasise grounding, dependence, and explanation. Fine’s essence is not Aristotle’s form, but it plays an analogous role. It explains why modal facts obtain rather than merely recording that they do. 

If we widen the lens further, the Fine–Williamson dispute also maps onto a long-standing divide between Enlightenment and Romantic sensibilities, though again in a highly refined form. Williamson’s metaphysics embodies the Enlightenment aspiration to a unified, austere account of reality governed by general principles. Even when he insists that philosophy is not physics, the model of explanation he favours is still one of global coherence and formal discipline. 

Fine’s metaphysics, especially when extended to plural necessities, resonates with the Romantic and post-Romantic insistence that reality is structured by heterogeneous forms of necessity: ethical, aesthetic, historical, tragic. Think here of Friedrich Schelling, for whom freedom, necessity, and tragedy are irreducible dimensions of reality, not to be flattened into a single explanatory register. Fine’s pluralism is a domesticated, analytic descendant of that impulse. 

This is why, when you bring literature, history, and art into the picture, Fine suddenly looks like the more genuinely post-Kantian thinker. Kant opened the space by denying that metaphysics could be a science of things in themselves. Fine fills that space by allowing multiple kinds of grounding and necessity to coexist without forcing them into a single formal mould. Williamson, by contrast, closes the space again by insisting that one modal framework is enough, provided it is regimented correctly. 

So where do they agree? They agree, crucially, against positivism and deflationism. Both reject the idea that modality is merely linguistic, conventional, or expressive. Both think metaphysical necessity is real. Both think philosophy has genuine subject matter beyond empirical science. In this respect, they stand together against much twentieth-century empiricism. 

They part company over whether metaphysics should aim at unity or articulation. Williamson wants a single, powerful framework that captures the general structure of reality, even if it treats many distinctions as derivative. Fine wants a layered framework that captures explanatory asymmetries, even if it sacrifices unity. 

From Kant onwards, these have always been the two options. Kant himself vacillates between them. The post-Kantian tradition splits accordingly. Hegel, Schelling, and later phenomenology push toward articulation. Frege, Russell, and later analytic logic push toward unity. Fine and Williamson are the contemporary inheritors of that split, operating with extraordinary technical sophistication but reenacting a very old disagreement. Seen in that light, their dispute is a live expression of one of philosophy’s deepest tensions: whether necessity is something we discover by formalising or something we understand by explaining. Kant is not just a historical preface here, he is the reason the Fine Williamson dispute has the peculiar feel it has, as if they are arguing about what counts as a metaphysical question before they are arguing about the answer. 

Prior to Kant, it is easy to picture modality, possibility and necessity, as a feature of being itself. Something is necessary if it cannot fail to be, possible if it can be, impossible if it cannot. The form of the question is ontological in the most direct sense. Then Kant arrives and forces a new option into the space. Perhaps necessity is not first of all an ingredient in the furniture of the world, perhaps it is a constraint on how a finite rational subject must judge if it is to judge at all. The centre of gravity shifts. Modal talk becomes entangled with the conditions for experience and thought. Even if you ultimately reject Kant, you inherit the problem he creates, you now have to say whether modality is primarily discovered in the world or installed by the structure of judgement, or some complicated hybrid of the two. 

One way to see this is to begin with what Kant does to the idea of metaphysics as a free floating speculative science of “what there is”. He does not deny that we make metaphysical claims, he says we cannot make them responsibly without first asking what makes knowledge possible. That is what “transcendental” means in his sense, it is about the enabling conditions of representation, judgement, and experience. A modal claim, on this approach, is a kind of commitment about what must hold for any object of experience, or for any coherent judging activity. That is why post Kant philosophy is repeatedly pulled toward treating logic, and especially modal logic, as a window onto the structure of thought and objecthood. 

Williamson is explicit that modal logic is something like a structural core of metaphysical enquiry, where you fix an interpretation of modal operators as metaphysical necessity and possibility, treat the quantifiers as unrestricted, and then ask which generalisations in that language are true, as part of the overall project of metaphysics. That looks, at first glance, like a reversal of the Kantian turn. It looks like a return to pre Kant metaphysics, in which logic is a tool for describing being. But it is not so simple, because Williamson is also deeply marked by the post Kant demand for methodological seriousness. He does not say, metaphysics is a realm of pure a priori insight immune to theoretical constraint, he says, treat it as inquiry, subject to pressure from theory choice, from simplicity, from explanatory power, from how our best overall picture hangs together. That is a recognisably post Kant move even if its target is realist rather than transcendental. It is Kantian in its insistence that metaphysics is answerable to norms of justification, even if it is anti Kantian in the sense that it aims at the world as it is, not merely at the conditions of appearance. 

You can see that methodological orientation in Williamson’s own meta philosophical remarks about philosophy as continuous with serious inquiry rather than a special insulated domain. Fine arrives at the same post Kant crossroads and takes a different exit. He does not deny the seriousness of logic, but he resists the temptation to treat modal logic, or possible worlds talk, as the primary engine of modal metaphysics. The key Fine move, the one that has become a pivot point in contemporary metaphysics, is that the contemporary assimilation of essence to modality is misguided, and that metaphysical necessity should be understood as grounded in essence rather than essence being reduced to what is necessary. 

Here “essence” means, roughly, what a thing is, in a way that supports explanation. Take a triangle. Its essence involves being a three sided polygon, not because we have consulted all possible worlds and found that wherever it exists it has three sides, but because being three sided is part of what it is to be a triangle. The modal fact, that necessarily triangles have three sides, is then downstream. It is true in virtue of what triangles are. Fine’s complaint is that possible worlds style modal metaphysics flips that order. It begins with necessity and then tries to define essence as a special subset of necessary truths, truths that are necessary “in the right way”. Fine thinks that is backwards and loses the explanatory direction. 

Kant’s shadow falls in two places. First, in the idea that there is an order of explanation that is not merely empirical but also not merely arbitrary. Kant says, you must explain how objectivity is possible. Fine says, you must explain why modal truths hold, and the engine of that explanation is essence, not a brute modal operator. Second, Kant’s shadow falls in the distrust of “metaphysical monsters”, claims that purport to describe the world while covertly relying on features of our conceptual scheme. Fine’s suspicion of modal reductionism can be heard as a suspicion of that kind. If you define essence in modal terms, Fine thinks you may be smuggling in a conceptual apparatus that does not track the grounding facts. It is not that the modal truths are false, it is that the account of why they are true is upside down. 

This is the first big map point. Williamson and Fine are both post Kant in the sense that they take the norms of enquiry, and the order of explanation, very seriously. They are not playing the older game of metaphysical system building by pure stipulation. But they disagree about where the explanatory constraints live. Williamson thinks modal logic, with quantifiers and necessity operators, provides a disciplined arena in which metaphysical theses can be stated precisely and evaluated, and he develops that into an argument for necessitism, the view often glossed as “everything necessarily exists”, or more carefully, that necessarily everything is necessarily something, formalised in a quantified modal schema. 

Fine thinks that if you start there, you are already committed to a monolithic “metaphysical necessity” that may be too coarse grained, and you are likely to lose the explanatory primacy of essence. To make this intelligible without technical machinery, you can use a cinematic analogy. Think of a film’s internal logic. 

Take Lost Highway. There is the level at which we can describe regularities, every time the Mystery Man appears the film’s causal texture warps, every time the phone rings identity starts sliding. That is like working at the level of modal generalisation, you are describing patterns that hold across the film’s space of scenes. Then there is the deeper level at which you try to say what the characters are, what kind of being Fred or Pete has in the film’s ontology, what sort of entity the Mystery Man is, and what the film is doing when it splits and recombines a life. Fine’s claim, translated, would be that if you want to explain the regularities properly you have to start from that deeper “what it is” level, the essence level, rather than trying to reconstruct the “what it is” level from the observed regularities across scenes. 

Williamson’s instinct is that the disciplined description of modal structure is not merely an after the fact summary, it is a route into what the film’s ontology must be like if the structure is to be true. That is why their disagreement can look like a disagreement about which direction the camera should point, from essences to necessities, or from modal structure to ontological commitment. 

Kant enters again when we ask whether modality is a feature of things or of our representational practices. Williamson is a realist about modality in the robust sense. When he treats modal logic as metaphysics, he is treating necessity and possibility as worldly, not merely epistemic, not merely about what we can know. His necessity is metaphysical necessity, necessity in the way the world could not have failed to be, not necessity in the sense of being knowable a priori or being analytic. That is part of his broader project to resist a sharp boundary between logic and metaphysics, the logic is about the world’s structure, and therefore can be part of metaphysical discovery. 

Fine is also a realist about modal facts, but he is realist in a different explanatory key. The modal is real because essence is real, and essence is not, for him, a mere projection of our classificatory habits. It is anchored in the natures of things. The difference is not whether modality is real, the difference is what makes it real and how many different modalities there are. This is where the Kant lineage becomes more fine grained. 

After Kant you get several recurring strategies. One strategy, you could associate it with a certain logical positivist and Quinean temper even if the details vary, is to demote metaphysical modality and treat necessity as linguistic or conceptual, a matter of how we use words or fix meanings. On that picture, modality is not a joint in the world, it is a shadow of our conceptual scheme. That strategy is one of the main background enemies for both Fine and Williamson, though they fight it differently. 

Williamson fights it by making modal logic do metaphysical work, as if to say, you cannot cordon modality off into language, it is implicated in our best total theory of what there is. Fine fights it by saying, if you want grounding and explanation, essences give you more than language gives you, because essences are tied to what things are, not merely to our descriptions. 

Another post Kant strategy, you see it in neo Aristotelian revivals, is to treat modality as derivative from powers, natures, and essences, in something like Fine’s sense. The world has structured kinds of things, and those natures generate constraints, which show up as necessities. That strategy is Fine adjacent even when Fine is not explicitly invoked. It is also why Fine’s work is often described as reviving an Aristotelian approach to modality, where things, not worlds, are primary. 

A third strategy is structuralist. It says the world’s modal profile is best captured by a formal structure, and we should be willing to follow the formal structure wherever it leads, even if it delivers counterintuitive ontological commitments, provided the overall theory is simpler, more systematic, and more explanatory. That is Williamson’s zone. His defence of necessitism is not presented as a whim, it is presented as the upshot of taking a certain disciplined picture of quantification and modality seriously, and of comparing the costs of necessitism with the costs of contingentism, the view that not everything exists necessarily. 

So if Williamson claims to be methodologically sensitive, and claims philosophy is a kind of serious inquiry, and does not want to flatten the humanities, then why does he not accept Fine’s pluralism about necessity, including the affective and normative dimensions you pointed to? That is a question about what counts as metaphysical evidence, which is precisely the kind of “Kantian” question that survives even in realist metaphysics. What are the conditions for a claim to count as part of metaphysical enquiry, and what sorts of constraint should metaphysics respect. 

To see how Kant frames that, consider the Kantian worry about category mistakes. If you treat something that belongs to the conditions of representation as if it were a property of the represented object, you generate illusory metaphysics. In a Lynch film, it is easy to make exactly that kind of mistake. You can treat the film’s editing, its sound design, its shifts in grain and texture, as if they were events inside the story world, when they might be the film’s way of presenting the world, not an additional thing happening in it. Philosophically, the analogue is mistaking representational necessity for worldly necessity. 

Fine’s pluralism makes this sort of diagnostic manoeuvre easier. He can say, there is a necessity internal to a normative order, or to an aesthetic order, which need not be reduced to, or identified with, metaphysical necessity.   Williamson, by contrast, wants the main modal operator to be metaphysical, with other “necessities” either analysed as something else or treated as derivative. That is not because he is blind to aesthetics or normativity, but because he is trying to keep the core metaphysical operator clean enough to support general theory. 

The Kantian framing therefore yields a precise version of the question I’m pressing. Is Fine’s talk of different “orders”, natural, normative, metaphysical, and my suggestion of affective necessity, best understood as tracking different kinds of necessity in the world, or as tracking different kinds of constraint in our practices of explanation, evaluation, and interpretation. Fine wants them to be real in the sense that they are not just linguistic conveniences, they correspond to different sources of necessity, coming from the identities of things, the natural order, the normative order, and so on. 

But a Williamsonian can respond in a Kant inflected way, saying, be careful not to inflate every legitimate constraint on judgement into metaphysics. Some necessities might be necessities of representation, of rational agency, of evaluation, without being necessities in being. That is the sort of line drawing pressure Kant makes unavoidable. Once you recognise that there are different kinds of constraint, you have to decide which constraints are metaphysical. 

So even before we get to the explicit Fine versus Williamson clash, Kant already tells us it will not just be about the truth of necessitism or the priority of essence, it will be about the boundary between metaphysics and other forms of intelligibility. What is striking is that both Fine and Williamson do not want to say, metaphysics is merely about our language, nor do they want to say, metaphysics is a free for all immune to rational constraint. The only remaining option is a hard, technical, and ultimately evaluative dispute about explanatory order, theoretical virtue, and ontological commitment. That is why Williamson can write a reply to Fine that reframes the debate, complaining that Fine’s original actualism versus possibilism dispute is confused and that the clearer dispute is contingentism versus necessitism. 

Reframing the dispute is a way of saying, here is the proper object of metaphysical enquiry, in a way that reflects a methodological stance about what counts as clarity, what counts as a genuine thesis, what counts as an ill posed issue. That is again a post Kant move, because it treats the space of metaphysical questions as itself something that must be disciplined, not given. 

Bring this back to Lynch for a moment, because it will matter later when I ask whether the film data, dread, terror, devastation, are “metaphysically salient”. The Kantian lesson is that there is always a question of level. The dread might be an epistemic or affective access route to a metaphysical structure, or it might be part of the representational condition the film imposes, a necessity of how the world is given rather than a necessity of the world given. Fine is naturally sympathetic to letting such things count as data for modality, because he already thinks there are different sources and different registers of necessity. Williamson is naturally suspicious, not of the dread, but of the inference from dread to metaphysics, because he wants to keep metaphysical necessity tied to what is the case across counterfactual variation in the world, not to what is demanded by an evaluative or affective stance. 

Fine looks like an heir to the Aristotelian strand that Kant partially displaces but does not eliminate, the strand that says, the world has natures, and the natures explain the modal profile. His innovation is to make that strand logically sharp without reducing it to possible worlds. When he says essence precedes necessity, he is also saying that explanation precedes mere modal classification. 

Williamson looks like an heir to the rationalist and structuralist strand that runs through Kant but becomes realist again after Kant, the strand that says, disciplined formal articulation can reveal deep structure of reality, and metaphysics should be continuous with rigorous inquiry rather than insulated. His innovation is to treat quantified modal logic as a live instrument of metaphysical discovery, and to accept the ontological costs, like necessitism, if they buy overall theoretical virtue. Both inherit from Kant the demand to justify the form of metaphysical questions, not just their answers. Fine does it by grounding modality in essence and then allowing plurality of modal notions. Williamson does it by insisting on clarity of theses and by letting the formal framework regulate what counts as a coherent position. 

What Kant does, as we saw, is fracture the naive picture in which necessity simply belongs to things in the same way colour or mass might. After Kant, no one can responsibly talk about necessity without saying something about explanation, justification, or conditions of intelligibility. But Kant leaves open two radically different ways of taking that lesson, and the modern landscape splits accordingly. One way is to say: necessity is best understood through form. Through the structures that govern judgement, inference, representation, or theoretical articulation. The other way is to say: necessity is best understood through constitution. Through what things are, how they are put together, what grounds them, what makes them the kinds of things they are. The former trajectory leads toward modality-first views. The latter toward essence-first views. Williamson and Fine sit squarely on opposite sides of this divide, but with a twist that makes their disagreement unusually subtle. 

Neither is naive. Neither ignores Kant. Both think metaphysics must be disciplined. What they disagree about is what does the disciplining. I’ll start with Fine, because his position can easily be caricatured as a return to pre-Kantian essentialism unless we are careful. Fine’s core claim is that explanatory priority cannot be captured by modal profile alone. Two facts can be necessarily equivalent and yet differ in what explains what. That is the hyperintensionality point, and it is the engine of everything else he does. Take the simplest example. It is necessary that a triangle has three sides. It is also necessary that a triangle has angles summing to 180 degrees. But one of these facts explains the other, at least on the standard Euclidean picture. Being three-sided is part of what a triangle is. Having interior angles summing to 180 degrees follows from that. Modal equivalence, necessity in all possible worlds, does not track that explanatory asymmetry. Essence does. This is why Fine thinks the Kripkean move, defining essence as necessary truth about an object, gets things backwards. It tells you that something must be so, but not why. Essence is meant to answer the “in virtue of what” question. It is a grounding notion. 

Now notice something crucial. This is not a rejection of modality. Fine is not saying modal facts are unreal or unimportant. He is saying they are downstream. They are the output of deeper constitutive facts. Modal space is shaped by essence, not the other way around. This places Fine squarely in a post-Kant but non-Kantian lineage. Kant says necessity comes from the conditions of representation. Fine says necessity comes from the natures of things, but those natures are accessed through explanation, not through brute intuition. It is a modernised Aristotelianism, but one that is fully aware of Kant’s critique and therefore insists on clarity, logical precision, and anti-mysticism. 

Once you accept that, Fine’s pluralism about necessity is almost inevitable. If different kinds of things have different kinds of natures, then different kinds of necessity will arise. Logical necessity tracks logical form. Metaphysical necessity tracks constitutive structure. Normative necessity tracks the essence of practices or roles. Aesthetic necessity tracks the internal demands of forms. There is no reason to expect a single necessity operator to do justice to all of these, because there is no reason to expect a single kind of grounding relation. This is why Fine is so resistant to the idea that modal logic, even very sophisticated modal logic, can be metaphysically foundational. Modal logic tracks invariance across scenarios. Essence tracks explanation. Those are different tasks. 

Now turn to Williamson. Williamson’s starting point is not “what explains necessity” but “what is the best overall theory of modal truth”. His methodological instincts are recognisably post-Kantian in a different way. He does not trust appeals to metaphysical intuition, but he also does not think metaphysics collapses into linguistics or psychology. Instead, he thinks metaphysics should be pursued the way we pursue other theoretical disciplines: by proposing precise frameworks, assessing their costs and benefits, and seeing which yields the best total package. Quantified modal logic, for Williamson, is not just a representational convenience. It is a way of stating metaphysical hypotheses clearly enough that they can be evaluated. Once you do that, certain theses, such as necessitism, emerge as theoretically attractive. 

Necessitism, in its cleanest form, says that the domain of quantification is fixed across possible worlds. Everything that is something is something necessarily. Contingency attaches to properties, relations, and states of affairs, not to existence itself. Williamson’s defence of this is that alternatives are theoretically worse. Allowing the domain to vary introduces complications, puzzles about quantifying into modal contexts, awkward restrictions on inference, and ad hoc machinery to block unwanted results. A constant domain yields a simpler, more systematic theory. 

Notice the structural difference already. Fine asks: what explains why something must be the case. Williamson asks: what theoretical framework best captures modal facts. Fine’s metaphysics is explanatory-first. Williamson’s is theory-first. This difference matters because it explains why Williamson is unmoved by the hyperintensionality argument in the way Fine wants him to be. Williamson can happily concede that modal equivalence does not capture all explanatory distinctions, while denying that those distinctions belong to metaphysics proper. They may belong to semantics, to conceptual analysis, to theory choice, or to pragmatic explanation. Metaphysics, on his view, is not obliged to mirror every explanatory asymmetry we care about. 

Here is where the post-Kant fault line really opens. Fine thinks that if a distinction plays an explanatory role in why things must be the way they are, then it belongs to metaphysics. Williamson thinks that metaphysics has a narrower remit: it concerns the most general structure of reality, not every explanatory distinction that arises in inquiry. This is why Fine accuses modality-first approaches of being too coarse grained. And it is why Williamson accuses essence-first approaches of overpopulating metaphysics with primitives. 

This connects to my earlier pressure about psychological, social, historical, aesthetic, and affective necessity. From a Finean perspective, if these domains exhibit genuine “musts” grounded in what those things are, in what practices, narratives, or forms essentially involve, then those musts are metaphysically serious. They are not metaphysically on a par with logical necessity, but they are real necessities grounded in real essences. From a Williamsonian perspective, these “musts” are perfectly real phenomena, but they are not metaphysical necessities. They are necessities relative to certain structures, practices, or responses. They can be explained without expanding the metaphysical modal repertoire. Metaphysics proper remains focused on what holds across counterfactual variation in the world’s most general structure. 

This is why Fine’s charge that Williamson risks dismissing whole domains as non-metaphysical bites so hard. It targets exactly this narrowing of metaphysics’ scope. But now we can see the deeper issue. The disagreement is not about whether those domains are real. It is about whether metaphysics must explain them at the same level at which it explains existence, identity, and possibility. Fine says yes, because metaphysics is about explanation and grounding wherever necessity arises. Williamson says no, because metaphysics is about a particular kind of generality, and other explanatory projects, though legitimate, operate at different levels. 

Historically, this maps neatly onto a post-Kant split. On one side you have a tradition that runs from Aristotle through medieval essentialism, is interrupted by Kant, and then re-emerges in modern form through grounding and essence talk. Fine is the most rigorous contemporary representative of this line. On the other side you have a tradition that runs from Kant’s emphasis on form, through Frege and modern logic, into contemporary analytic metaphysics that treats formal structure as revelatory of reality’s deep organisation. Williamson is a leading figure here. 

Both traditions agree that naive metaphysics is impossible after Kant. They disagree about whether the corrective should be explanatory depth or formal discipline. This also explains why the debate becomes so sharp around necessitism. Necessitism is not just a claim about existence. It is a claim about what kind of thing metaphysics is allowed to be. If existence is necessary, then metaphysics cannot explain contingency by appeal to coming-into-being or passing-away. It must explain contingency structurally. Fine hears that and says: structure without grounding is empty. Williamson hears Fine and says: grounding without discipline is obscure. This is a clash of philosophical temperaments rooted in different readings of the Kantian legacy. 

A third relevant subject in trying to place the dispute in a post-Kantian frame is the unity versus plurality of necessity. This is not a detachable side issue but the point at which the Fine–Williamson dispute becomes unavoidable once one takes post-Kantian metaphysics seriously. The first thing to see is that pluralism about necessity is forced on Fine by the explanatory role he assigns to essence. Once essence is doing explanatory work, once necessity is grounded in what things are rather than in a single modal operator, it becomes almost impossible to maintain that all necessities are of the same kind. Different kinds of things have different kinds of constitutive structure, and those structures generate different kinds of “must”. This is easiest to see with examples that are not metaphysically controversial. Logical necessity is not the same as mathematical necessity. Mathematical necessity is not the same as natural necessity. Natural necessity is not the same as moral or normative necessity. A promise must be kept in a way that a stone must fall does not. A tragic narrative must unfold in a way that a chemical reaction must not. These are not just stylistic differences. They involve different sources of constraint, different explanatory grounds. Fine’s claim is that metaphysics has gone wrong by treating these differences as superficial. When philosophers say “that’s not metaphysical necessity, it’s just moral necessity” or “just narrative necessity”, they are, on Fine’s view, refusing to do metaphysical work rather than showing that no work is needed. They are treating metaphysical necessity as a privileged register without arguing for that privilege. 

This is where the post-Kantian background matters again. Kant already showed that necessity does not come in only one flavour. There are necessities of judgement, necessities of inference, necessities of experience. Kant’s mistake, in Fine’s eyes, was to collapse all of these into the conditions of representation, rather than allowing them to be grounded in different kinds of being. Fine’s pluralism can be read as an attempt to correct that over-centralisation while keeping Kant’s insight that necessity is not just brute fact. Once you take that step, the idea that metaphysics should aim for a single unified necessity operator starts to look less like rigour and more like impoverishment. Why should reality cooperate with our desire for unification? Why should the “must” governing triangles, the “must” governing obligations, and the “must” governing trauma all reduce to the same thing? 

This is where Timothy Williamson enters with a very different instinct. Williamson does not deny that there are many ways we talk about necessity. He does not deny that people experience compulsion, inevitability, obligation, dread. What he denies is that metaphysics must treat all of these as revealing distinct kinds of necessity at the same level. For Williamson, metaphysics earns its keep by identifying the most general structural features of reality. Metaphysical necessity, on his view, is the necessity that holds across all counterfactual variation consistent with the world’s structure. Other necessities are either reducible to that or belong to other explanatory projects. Fine thinks plural necessity is forced on us by the world’s heterogeneity. Williamson thinks plural necessity risks turning metaphysics into an unprincipled catalogue of constraints. 

Why does Williamson think this? Because from his point of view, the unity of necessity is what allows metaphysics to remain systematic. If every domain gets its own necessity operator, metaphysics loses its grip. There is no longer a single framework within which claims can be compared, tested, or integrated. Instead, you get a patchwork of local explanations with no clear criteria for what belongs to metaphysics proper. This worry is not trivial. It is tied to Williamson’s broader anti-exceptionalism. He wants metaphysics to look like serious inquiry, not like literary criticism or phenomenology. Serious inquiry, on his view, aims at general principles, not at endlessly multiplying primitives to match every nuance of experience. 

But Fine would say that affective necessity, the sense that something must happen or must end, is not a subjective gloss. It is a real constraint grounded in the structures at play, narrative, psychological, historical. To ignore it is to ignore data. Williamson would reply that treating such felt constraints as metaphysical necessities conflates metaphysics with phenomenology. The fact that something feels unavoidable does not show that it is unavoidable in the metaphysical sense. 

But Fine’s reply to that is sharp. He can say: you are assuming, without argument, that metaphysical necessity is the only necessity that counts metaphysically. That is exactly what is at issue. If philosophy is not physics, if it includes history, ethics, art, and interpretation, then metaphysics cannot be limited to the necessity of fundamental particles and abstract objects. It must also account for the necessities internal to those other domains. This is where pluralism becomes not just plausible but pressing. Once Williamson accepts that social institutions, languages, practices, and artworks necessarily exist, given necessitism, he can no longer use existence as a filter. He must explain why the necessities internal to those things are not metaphysical necessities. And that explanation cannot simply be “because I say so” or “because modal logic does not track them”. 

Fine’s pluralism offers a principled answer. Different necessities correspond to different kinds of essence. The essence of a promise grounds a moral necessity. The essence of a narrative grounds a narrative necessity. The essence of a person grounds psychological necessity. These necessities are not all on a par, but they are all real, and they are all metaphysically explanatory in their own domains. Williamson’s resistance to this is comes from a deep worry about losing the distinction between metaphysics and everything else. If metaphysics expands to include every explanatory practice, it ceases to be a distinct discipline. 

But Fine can turn this worry around. He can say: metaphysics was never meant to be thin. Aristotle’s metaphysics included substance, form, causation, teleology. Kant’s metaphysics included the conditions of experience. Hegel’s metaphysics included history. The idea that metaphysics must be austere and unified is a modern prejudice. At this point, the debate takes on a recognisably post-Kantian shape. It is no longer about whether plural necessities exist, but about whether metaphysics should reflect the structure of reality as it is lived and articulated, or whether it should abstract away from that structure in pursuit of formal unity. This is why the unity versus plurality dispute cannot be settled by technical arguments about modal logic alone. It is a dispute about philosophical temperament. 

Williamson’s temperament is synthetic and unificatory. He wants a framework that can, in principle, accommodate everything without multiplying fundamental kinds. Fine’s temperament is analytic and discriminating. He wants a framework that can explain why different things are constrained in different ways, even if that means accepting multiple irreducible sources of necessity. 

Once you see this, it becomes clear why Inland Empire and similar works exert such pressure. They do not merely present complex narratives. They present worlds in which different kinds of necessity collide: psychological compulsion, narrative repetition, historical trauma, ontological persistence. To treat all of that under a single necessity operator is to risk flattening the very thing the work is doing. From a Finean point of view, this flattening is a metaphysical error. From a Williamsonian point of view, it is a necessary abstraction. The debate is not about whether plural necessities exist in some loose sense. Williamson would happily agree that there are many ways of saying “must”. The debate is about whether metaphysics should track those differences as differences in reality, or whether it should explain them away as derivative. Fine says: track them. Williamson says: explain them away. 

Fine’s intuition is that Williamson’s position risks flattening them. It flattens not by denying existence, but by denying metaphysical salience. And because necessitism guarantees that everything exists, the flattening cannot be done ontologically. It must be done explanatorily. The result is a metaphysics that is ontologically maximal but explanatorily selective.

Fine’s pluralism resists that selectivity. It insists that if something constrains what can and cannot be, and does so in virtue of what it is, then it belongs in metaphysics. The next step is to examine the hardest consequence of this pluralism: the charge that Williamson’s necessitism becomes essentialism at the level of the world itself, an essentialism that Fine can expose precisely because Williamson refuses to acknowledge essences elsewhere. This is the point in the dialectic where everything tightens. Up to now, one might think the dispute is merely about explanatory taste, unity versus plurality, or different modelling strategies. But the charge here is deeper. It says that Williamson’s attempt to avoid essence talk does not succeed. Instead, it displaces essence upward, from objects to the world as such, and then treats that displacement as philosophically innocent. Fine’s claim is that it is not innocent at all. To see why, we need to be very precise about what Fine means by “essence” because the accusation is not that Williamson secretly believes in Aristotelian forms. 

Fine’s notion of essence is not a list of necessary properties. It is a relation of metaphysical priority. To say that it is essential to x that p is to say that p holds in virtue of what x is. Essence answers the “because” question for necessity. When Fine insists that essence precedes necessity, he is insisting that metaphysics must distinguish between facts that are necessary because of deeper constitutive features and facts that are necessary merely as a consequence of how modal space is structured. Now consider necessitism. Necessitism says that necessarily everything exists, or more carefully, that the domain of quantification does not vary across possible worlds. Whatever is something is something in every possible world. Contingency is shifted away from existence and onto properties and relations. 

Fine’s complaint is not that this thesis is incoherent. It is that once you adopt it, you have effectively made existence itself an essential feature of everything. But because Williamson does not want to say that Socrates exists essentially, or that tables exist essentially, he shifts the essentiality to the domain. It is essential to reality, or essential to the world, that it contains all the things it contains. This is not a caricature. It is a direct consequence of treating the domain as modally invariant. Fine can now ask in virtue of what does the domain remain fixed? If the answer is “in virtue of nothing, it is just how modal logic works”, Fine will say that this is not an explanation but a refusal to explain. If the answer is “because that yields the best overall theory”, Fine will say that theoretical convenience is not metaphysical grounding. If the answer is “because everything necessarily exists”, Fine will say that this is circular. 

At this point, the Finean diagnosis is that necessitism functions exactly like an essence claim, but at a higher level. It is an essence of the world or of reality as such. The world’s essence is maximality. The world is such that nothing can fail to be in it. Once you see this, the earlier disputes snap into focus. 

Williamson accuses Fine of multiplying essences unnecessarily. Fine replies that Williamson has collapsed all essences into one massive, unarticulated essence, the essence of the domain, and then declared that no further essence talk is needed. This is why Fine thinks necessitism is not metaphysically austere. It is metaphysically blunt. Now, to be fair, Williamson does not think of himself as positing a world-essence. He would reject that description. His response has several layers. 

First, Williamson would deny that necessitism is an explanatory thesis at all. He would say it is a structural constraint on how modal quantification works. It is not meant to answer “why” questions but to provide a framework in which metaphysical claims can be formulated clearly. Asking “in virtue of what does the domain remain fixed” is, on this view, a category mistake. It is like asking “in virtue of what does modus ponens hold”. It holds because that is part of the logic we are using, and the logic is justified by its overall theoretical role, not by deeper metaphysical facts. This reply is powerful, because it tries to block Fine’s move at the outset. If necessitism is not explanatory, then accusing it of hiding an essence misses the point. 

But Fine can respond that this move is unstable. Modus ponens is a rule of inference. Necessitism is a claim about reality. Williamson insists that modal logic is metaphysics, not just syntax. So the analogy breaks down. If necessitism is a metaphysical claim, then it must answer to metaphysical standards of explanation, not merely to formal convenience. 

But Williamson isn’t done. He can say that Fine’s demand for grounding is itself optional. Metaphysics does not have to explain everything. Some facts are brute. Some structures are fundamental. The fact that the domain is fixed might simply be a fundamental feature of reality. There is no further “because”. And this is where the dispute becomes irreducibly philosophical. Fine does not deny that there can be brute facts. But he thinks that declaring something brute should be a last resort, not a default response to explanatory pressure. In Fine’s eyes, Williamson reaches for brute structure too quickly, precisely at the point where explanatory discrimination is most needed. Fine would say: you are willing to accept brute facts about the domain, but you are unwilling to accept structured essences about objects, practices, narratives, or norms. That is not neutrality. It is a preference for a certain kind of brute fact. 

So the issue becomes, if Williamson is willing to accept brute facts about the world’s modal structure, why is he unwilling to accept brute facts about the essential structure of social practices, moral roles, or narrative forms? Why is it acceptable for “everything exists necessarily” to be fundamental, but not acceptable for “promises essentially bind” or “trauma essentially repeats”? The answer cannot be “because the former is metaphysical and the latter is not”, because that is exactly what is at issue. The answer cannot be “because the former is captured by modal logic”, because that assumes the priority of modal logic. And it cannot be “because the former is simpler”, because simplicity is not the same as truth. 

Fine can argue that Williamson’s necessitism smuggles in a metaphysical hierarchy while pretending not to. At the top sits the modal structure of the domain, treated as fundamental and brute. Beneath it sit all other necessities, treated as derivative or merely apparent. Fine’s alternative is to invert this picture. Instead of a single brute modal structure at the top, we have many grounded necessities at different levels, each explained by the essences of the things involved. There may still be brute facts, but they are local, not global. 

In a Lynch film, identities split, repeat, thicken, thin. Certain patterns feel inescapable. Certain repetitions feel necessary. A Williamsonian reading can say: all of these entities exist necessarily; what varies is how they are instantiated or accessed. A Finean reading says: these patterns arise because of damaged or altered essences, of roles and narratives that generate necessity from within. If the Williamsonian insists that the first explanation is metaphysically sufficient, Fine will say: you have explained persistence but not inevitability. You have explained why nothing disappears but not why something must happen. 

This returns us to the essentialism charge in its most refined form. Fine is accusing Williamson of centralising essentiality in a way that blocks further explanation. The world, on a necessitist picture, has a fixed nature. That nature determines what counts as possible. Everything else is contingent variation within that nature. That is an essentialist picture, even if it does not traffic in essences of chairs and people. 

Williamson’s best reply is that metaphysics should identify the deepest level of structure and stop. It should not attempt to explain everything in terms of something deeper, because that leads to regress or to bloated ontology. Fine’s best reply, in turn, is that stopping at the domain level is arbitrary. There is no reason to think that maximality is the deepest structure. It is simply the structure that becomes visible when you privilege modal logic over grounding. 

So the dispute is about what counts as a satisfactory metaphysical stopping point. Williamson is satisfied when he reaches a framework that is simple, systematic, and expressive enough to state general truths about reality. Fine is satisfied only when those truths are explained by what things are. 

Historically, this maps onto an ancient divide. Plato stops at forms. Aristotle stops at substances. Kant stops at conditions of experience. Hegel stops at absolute spirit. Williamson stops at modal structure. Fine refuses to stop there and asks what structures that structure. 

This matters for discussions about affective necessity, about literature, about history, about metaphysical salience and all the rest. If Williamson’s stopping point is the world’s modal profile, then anything not captured by that profile must be treated as secondary. If Fine’s stopping point is explanatory grounding, then any pattern that demands explanation has a claim on metaphysics. This is why Fine’s essentialism charge is a way of forcing Williamson to say where and why he stops. 

We move, then, to the next subject raised in this debate: methodology, metaphysical salience, and the problem of boundary-drawing once everything necessarily exists. This is the point at which the Fine–Williamson dispute starts to look like a disagreement about what metaphysics is allowed to count as its data. The pressure here comes almost entirely from Williamson’s own self-description. He insists, repeatedly and explicitly, that philosophy is a kind of serious inquiry continuous with science, but not a natural science like physics. That qualification matters. He does not want metaphysics to be reduced to particle physics, nor to inherit physics’ ontological austerity by default. Philosophy, on his view, ranges over abstract objects, meanings, norms, epistemic states, logical relations, and modal structure. It is not ontologically shy. What it is, however, is methodologically disciplined. This combination, ontological generosity plus methodological discipline, is exactly what creates the problem I’ve have been pressing via Fine. 

Necessitism guarantees ontological generosity at the highest possible level. Everything exists necessarily. There is no ontological filter left. Social institutions, psychological states, narrative structures, aesthetic forms, historical events, linguistic conventions, moral obligations, affects, genres, traumas, styles, practices, all of these are in the domain. None can be dismissed as unreal, derivative in the sense of not really there, or merely fictional in the metaphysical sense. Once you accept necessitism, existence itself does no discriminating work. So where, then, does discrimination happen? Williamson needs a criterion of metaphysical salience that does not appeal to existence. But the moment you ask for such a criterion, the ground becomes unstable. 

One obvious candidate would be fundamentality. Metaphysics, one might say, is concerned with what is fundamental rather than with what merely exists. But Williamson is famously sceptical of grounding talk and of metaphysical hierarchies that are not regimented by logic. He does not deny that there are more and less basic facts, but he resists treating grounding as a primitive metaphysical relation. So fundamentality is not something he can easily appeal to without conceding ground to Fine. 

Another candidate would be generality. Metaphysics, Williamson might say, concerns the most general features of reality, those that hold regardless of contingent social or psychological arrangements. This is a more promising route, and it is the one he most often gestures toward. Metaphysical necessity, on this view, is necessity that holds across all counterfactual variation compatible with the world’s structure. Normative or affective necessities hold only given certain agents, practices, or forms of life. 

But Fine’s reply is that this criterion presupposes what it is meant to justify. It assumes that “the world’s structure” is something that can be specified independently of social, historical, narrative, and affective organisation. That is exactly what is in question. Once you admit that institutions, practices, and histories necessarily exist, it is no longer obvious that they are less general than physical laws or modal axioms. The fact that they are historically located does not make them metaphysically local in the relevant sense. They structure entire forms of life. Moreover, this appeal to generality risks collapsing into a Kantian move that Williamson officially rejects. It starts to look as if metaphysical necessity is being defined as necessity relative to a certain abstract perspective on the world, a perspective that strips away anything tied to particular forms of experience or agency. Fine can say: that is not a neutral description of reality, it is a methodological choice masquerading as metaphysical insight. 

A third candidate would be explanatory indispensability. Perhaps metaphysics should concern itself with whatever must be mentioned in our best explanations of reality at the most abstract level. Williamson often appeals to theoretical virtues, simplicity, unification, explanatory power. He might say that modal logic earns its place because it is indispensable to a unified account of counterfactual reasoning, identity, and quantification. But again, Fine can press. Explanatory indispensability depends on what you are trying to explain. If you are trying to explain physical phenomena, you need physics. If you are trying to explain social order, you need institutions. If you are trying to explain narrative inevitability or trauma repetition, you need concepts that track those structures. The claim that modal logic is indispensable to metaphysics does not show that other forms of explanation are dispensable. It shows only that you have chosen a certain explanatory target. 

This is where Williamson’s inclusive conception of philosophy turns into a liability. If philosophy genuinely includes inquiry into language, history, art, ethics, and social life, then metaphysics cannot simply inherit physics’ habit of ignoring everything that does not enter into fundamental laws. Philosophy has no such excuse. It must say why certain realities, though real, are metaphysically secondary. Fine would say that Williamson’s framework has no internal resources to draw this line without begging the question. Because necessitism equalises existence, and because Williamson resists grounding and essence, the only remaining way to mark metaphysical salience is to appeal to the very modal framework whose exclusivity is under dispute. 

This is why Fine insists on grounding and essence. They are tools for drawing principled boundaries. Essence tells you which features of a thing explain others. Grounding tells you which facts depend on which. Without these tools, Fine argues, metaphysics collapses into a list of truths with no internal articulation. Williamson’s response, again, is to say that metaphysics does not need to draw boundaries at the level Fine demands. It can operate with a thinner notion of explanation, one that does not require grounding relations to be metaphysically primitive. Explanatory asymmetries can be handled at the level of theory choice, not ontology. We can say that some explanations are more general, more powerful, more systematic, without saying that the entities they mention are more real or more essential. 

But Fine’s counter is that this turns metaphysics into a meta-theory of theories rather than a theory of reality. It tells us which descriptions are useful, not which structures are explanatory. And once you allow that slide, metaphysics risks becoming merely a methodological supervisor rather than an account of what makes the world the way it is. In literature, history, and art, certain necessities are not optional. A tragedy demands its ending. A genre constrains what can happen. A trauma repeats itself. These are not metaphors. They are structural facts about those domains. To treat them as merely psychological or conventional is to miss what they are doing. Fine’s challenge to Williamson is therefore not that he ignores these domains, but that his metaphysics lacks the expressive resources to take them seriously as metaphysical. He can acknowledge them as phenomena. He can study them as objects. But he cannot let them shape his account of necessity without abandoning the unity he values. 

A film like Inland Empire presents a world in which narrative, affective, historical, and ontological constraints are inseparable. The dread is not a response to the world. It is part of the world’s structure. A metaphysics that treats that dread as merely epistemic or psychological has changed the subject. It has explained persistence and existence while leaving inevitability untouched. Williamson can say that this is a category mistake. Films construct worlds. Metaphysics concerns the real world. But Fine can reply that if metaphysics cannot make sense of constructed worlds without flattening them, then its claim to generality is hollow. After all, constructed worlds are part of reality. They necessarily exist too. 

So we arrive at the deepest methodological tension. Williamson wants metaphysics to be general, unified, and disciplined, even if that means abstracting away from many kinds of necessity that matter to us. Fine wants metaphysics to be explanatory, plural, and responsive to the different ways reality constrains itself, even if that means giving up unification. Necessitism magnifies this tension because it removes the simplest way of avoiding it. Once everything necessarily exists, the only way to say “this matters more than that” is to appeal to explanation, grounding, or essence. And those are precisely the notions Williamson is most wary of. 

This is why Fine’s critique keeps returning, and why it cannot be dismissed as a mere preference. It is rooted in the internal demands of Williamson’s own position. For both Williamson and Fine, modal metaphysics is not a narrowly technical debate about possible worlds semantics or truth conditions alone. It is a metaphysical and epistemological dispute about how necessity is grounded, about essences and their metaphysical salience, and about how human practices (reasoning, language, social norms) interact with modal commitments. Williamson treats logic and modality as subject to rational revision grounded in our best epistemic practices across domains, drawing an analogy between logic and scientific theories. Fine treats modalities as plural, foundational, and irreducible to any single uniform account, necessary truths are grounded in essences, not merely in systemic epistemic practices. 

Oddly, all this seems to pick up on issues in German Idealism concerning the relationship between subject and world, the foundations of thought, the role of history and the human sciences, and the nature of universals and particulars. Frederick Beiser’s commentaries on Idealism is helpful because he repeatedly insists that idealism has been misunderstood when reduced to egocentric solipsism or abstract metaphysical absolutism, and he argues that the real struggle in German Idealism (from Kant through Fichte, Schelling, Hegel and the Romantics) was to chart a middle way between subjective perspectivalism and dogmatic metaphysical monism. 

This struggle resonates deeply with the Williamson–Fine debate. If we step back from the immediate Fine–Williamson dispute and situate it within the philosophical landscape that opens after Immanuel Kant, it becomes clear that what is at stake is not merely a technical disagreement about modal logic, possible worlds, or necessitism. It is a renewed version of a much older question, namely whether necessity belongs primarily to structure or to constitution, to formal conditions of intelligibility or to what things are. 

Beiser’s work on Kant and German Idelism is especially helpful here because he shows how deeply this question has shaped post Kantian philosophy, and how misleading it is to treat modern debates as if they were detached from those historical struggles. Beiser’s central historical intervention is to resist the caricature of German Idealism as a reckless metaphysics of subjectivity or a monolithic doctrine of the Absolute. Again and again, his work emphasises that thinkers from Kant through Fichte, Schelling, Hegel and the early Romantics were responding to a crisis. The crisis was generated by the success of the natural sciences on the one hand, and the collapse of traditional metaphysical certainties on the other. Philosophy could no longer claim access to reality by dogmatic speculation, but it could also not accept that reality was exhausted by empirical description. The question became how to understand necessity, structure, and normativity without collapsing into either scepticism or scientism. 

That question is recognisably the one Williamson and Fine are still asking, even though they ask it with the tools of contemporary analytic philosophy rather than with outmoded and out of date transcendental arguments or dialectical systems. Williamson’s insistence that philosophy is a kind of science but not a natural science is deeply post Kantian in spirit. He rejects the idea that metaphysics is a special domain insulated from standards of rigour, yet he equally rejects the idea that metaphysics must answer to physics as its court of appeal. 

Logic, modality, epistemology and ontology are all, for him, parts of a single enterprise of rational inquiry, subject to revision, comparison, and theoretical assessment. This orientation echoes a Kantian insight even as it rejects Kant’s conclusions. Kant argued that necessity cannot be read directly off the world as it appears, but must be understood in relation to the conditions under which objects can be thought and experienced. Williamson rejects Kant’s restriction of necessity to the conditions of possible experience, but he preserves the idea that modal structure is something we uncover through disciplined theorising rather than through metaphysical intuition about essences. When Williamson treats quantified modal logic as a genuine tool of metaphysical discovery, he is extending the Kantian demand for methodological seriousness into a realist register. Modal claims are about the world, but they are justified by their role in our best overall theoretical picture. 

Fine’s position also emerges naturally from the same post Kantian inheritance, though it moves in a different direction. Fine agrees that naive essentialism is untenable, and that metaphysics cannot simply announce what things are without explaining how such claims are justified. But he argues that the dominant modal framework inherited from Kripke and possible worlds semantics gets the order of explanation wrong. When we say that something is necessary, we are not merely reporting that it holds in all possible worlds. We are implicitly appealing to what that thing is, to its nature, to its constitution. Essence, on Fine’s account, is not a metaphysical residue left over once modal facts are stated. It is the explanatory ground of those facts. Seen in historical perspective, this places Fine closer to the strand of post Kantian philosophy that resisted the reduction of necessity to formal conditions alone. 

Hegel’s insistence that necessity arises from internal relations rather than external imposition, and Schelling’s attempt to articulate different orders of necessity, natural, historical, normative, can all be heard as distant ancestors of Fine’s modal pluralism. Beiser’s great merit is to show that these figures were not rejecting rigour, but trying to articulate kinds of explanation that formal abstraction alone could not capture. This is why Fine’s pluralism about necessity should not be understood as a refusal of unity for its own sake. It is a response to the fact that different domains exhibit different kinds of constraint. Logical necessity, moral necessity, narrative inevitability, historical compulsion, and metaphysical necessity do not all function in the same way. They arise from different kinds of structure, and they demand different kinds of explanation. To collapse them into a single modal operator is not, on Fine’s view, to achieve clarity, but to lose explanatory resolution. Williamson’s resistance to this pluralism is equally rooted in a historical concern. He worries that once metaphysics allows itself multiple irreducible kinds of necessity, it risks dissolving into a patchwork of local explanations without a unifying framework. 

The Enlightenment aspiration to systematicity, which Kant inherits and transforms, is something Williamson wants to preserve. Metaphysics, on this picture, earns its legitimacy by identifying the most general features of reality, not by mirroring every distinction that arises in human practices. What Beiser’s historical work helps us to see is that this tension between unity and articulation is not accidental. It runs through post Kantian philosophy as a structural fault line. German Idealism never fully resolved it. Hegel tried to reconcile unity and plurality through dialectical development, Schelling oscillated between system and fragmentation, and the Romantics openly embraced irreducible multiplicity. 

Fine and Williamson are replaying this drama in a different idiom. Once this background is in view, it becomes much harder to dismiss Fine’s concerns about affective, narrative, and normative necessity as merely psychological or metaphorical. The German Idealists took tragedy, history, and art seriously not as decorative domains but as sites where necessity is experienced and articulated in ways that resist reduction. Beiser’s essays repeatedly stress that philosophy after Kant cannot afford to ignore these dimensions without impoverishing itself. At the same time, Williamson’s insistence on methodological discipline echoes the critical impulse that runs from Kant through analytic philosophy. Without constraints, metaphysics risks becoming unanswerable to reason. 

The question is not whether metaphysics should be rigorous, but what kind of rigour it should aspire to. Should it privilege formal unity even at the cost of flattening distinctions, or should it privilege explanatory depth even at the cost of plurality. Seen in this light, the Fine–Williamson debate is one of the clearest modern expressions of a post Kantian dilemma that has never been resolved. Whether necessity is ultimately a matter of structural invariance or of grounded essence, whether metaphysics should unify or articulate, whether philosophy should abstract from lived constraint or take it as data, these questions have been with us since Kant, and Beiser’s work shows just how much is at stake in how we answer them. 

So when we bring Williamson and Fine into conversation with the history of post-Kantian thought that Beiser reconstructs, what at first might look like an abstract technical dispute about modalities becomes deeply anchored in a longer philosophical struggle about the relationship between mind and world, about how thought is tied to reality, and about what it means to articulate necessity without reducing it either to mere epistemic convention or to a metaphysical fiction. 

Beiser’s account of German Idealism resists the familiar myth that Kant, Fichte, Schelling and Hegel simply inflated the subject until it swallowed the world. He shows instead that what motivates much of their work is not a triumphal subjectivity but an effort to escape the scepticism that arises when knowledge is construed as merely internal representations divorced from an external reality. That sceptical impulse hovers over early modern philosophy from Descartes to Hume, and the German thinkers of the late eighteenth century are more preoccupied with dismantling that legacy than with a crude celebration of subjectivity. 

Kant’s transcendental idealism itself is best understood, on Beiser’s reading, not as an invitation to solipsism but as a strategic effort to secure empirical realism while avoiding naive realism, to show that we can know objects in experience with confidence but that their conditions of possibility are not themselves empirically given. Beiser emphasises that the post-Kantian period is not homogenous; there is a complex taxonomy of positions where subjective or formal idealism and objective or absolute idealism compete, overlap and conflict. 

The immediate successors to Kant see different ways of deepening his project. Fichte turns the transcendental subject into an active principle, a “self-positing I”, but this is in service of combating scepticism and materialism rather than a celebration of inner mental content for its own sake. Schelling pushes back against formulations that collapse the world into consciousness by positing a reality of nature that is not simply the content of thought but a manifestation of deeper powers that we grasp only through our own rational capacities. Hegel, who later systematises these strands, tries to articulate a structure that reconciles self-consciousness with an independent world in a dialectical unity. 

This historical background matters for understanding Williamson’s and Fine’s approaches to modality because it reminds us that their disagreement is not just about how to treat quantifiers or possible worlds; it is about how to ground necessity in a way that avoids collapsing into scepticism on the one hand or empty abstraction on the other. Williamson’s anti-exceptionalist stance, which treats the logic we accept as comparable to accepting a scientific theory, resembles the post-Kantian critical spirit in that it stresses method and practice as normative anchors. For Williamson logic is an enterprise continuous with other rational investigations, subject to revision, adjustment and critique, not immune from the pressures of evidence and coherence that govern good reasoning across domains. In this view modality is embedded in our inferential commitments and holistic epistemic structures rather than anchored in a metaphysical substratum. 

This resonates with the neo-Kantian psychological line that Beiser brings out in the later nineteenth century, where thinkers like Fries and Herbart resist grand metaphysical systems while insisting that philosophy clarifies the structure of knowledge without overreaching into speculative metaphysics. Fine’s metaphysics of essence, in contrast, echoes in some ways the more speculative strains of post-Kantian thought. Where Williamson locates modality in rational practice, Fine treats essences as metaphysically fundamental and necessity as a family of irreducible grounded facts. For Fine, necessity is not merely a product of our best systematised reasoning; it is rooted in what things are, in their essential structures. 

This echoes the objective idealist concern to recover a robust reality that is not reduced to mere representations, the concern that drives Schelling’s Naturphilosophie and the early Romantics’ attempts to articulate a world that is both intelligible and independent yet knowable through reason. For these German thinkers, the world is not a passive object to be mirrored; it has its own dynamic structure that reason participates in rather than imposes. 

When Beiser describes German idealism as a messy struggle against subjectivism rather than a straightforward expansion of an ego into an absolute mind, he is urging us to see the modality debate as part of a longer effort to balance the normative structures of thought with the ontological commitments to an external world. Williamson’s emphasising of epistemic practice pushes toward a kind of Kantian-critical orientation, but if we stop at critique alone we risk the sceptical abyss that Kant’s successors were so alert to avoid. 

Fine’s emphasis on essence recognises that there must be something in the world that our modal claims are about, something more than just the scaffolding of inferential norms. Yet if we treat essences in isolation from the communal, historical, rational practices that give them context, we risk a metaphysics untethered from human reason that slides into unaccountable abstractions. 

The German tradition as discussed by Beiser is centrally concerned with this tension, showing that the idealists never took either extreme as final but sought a mediated account in which reason discloses the structure of reality without reducing reality to mere thought nor leaving thought adrift from what it purports to know. Beiser’s reconstruction also highlights how subsequent movements in philosophy, such as the neo-Kantian revival later in the nineteenth century, responded to the failures of both speculative idealism and naïve realism. 

Neo-Kantians like Fries and Beneke drew on Kant’s critical framework but combined it with empirical psychology, resisting both grand metaphysical systems and the reduction of logic to mere psychology. They insisted on clarifying the logic and methods of the sciences, including human sciences, without assuming a single foundational framework for all reality. This historical mediation anticipates the kind of pluralist but disciplined approach that Fine champions in his plurality of modalities: different necessities rooted in different kinds of structures, each demanding its own account rather than all being absorbed into a single system. 

At the same time, the neo-Kantian insistence that philosophy serve to clarify rather than to dominate reflects Williamson’s anti-exceptionalist orientation toward logic as continuous with other domains of inquiry. Seen against this broader tradition, the Williamson–Fine dispute can be understood as a modern articulation of the deeper struggle post-Kantians faced: how to anchor necessity so that it is neither arbitrary nor detached from our rational engagement with the world. 

Williamson’s model emphasises the normative and scientific-like character of rational inquiry, resisting the temptation to grant logic a special, untouchable status. Fine’s model emphasises the grounding of modal truths in essences, resisting the temptation to reduce modal claims to mere features of inferential practices. Both can be traced back to post-Kantian efforts to secure knowledge against scepticism and to respect the reality of objects without collapsing them into mere constructs of reason. 

Beiser’s work remind us that these concerns are not new; they articulate the same enduring tension between the conditions of thought and the conditions of being, a tension that has animated the tradition from Kant through the neo-Kantians and beyond. If we now return fully to the films, and let them do more than illustrate pre-formed positions, we can see much more sharply how the Fine–Williamson dispute actually changes what we think is happening on screen. 

The films become diagnostic instruments. They force modal metaphysics to show its hand. Take Lost Highway first. On a Williamsonian reading, the central shock of the film is not ontological rupture but ontological excess. Fred Madison does not cease to exist when Pete Dayton appears. Nothing disappears. No entity is annihilated. The domain is stable. What changes are properties, access relations, and epistemic standpoints. Identity fractures because modal space allows distinct property bundles to be instantiated by the same necessarily existing individual across different counterfactual or epistemically inaccessible configurations. The film’s looping structure, its return to its own opening message, functions like a modal operator applied to a fixed domain. Everything that can be quantified over is already there. The horror comes not from loss but from inescapability. You cannot exit the domain. You can only move within it. 

This reading fits Williamson’s necessitism uncannily well. The terror of Lost Highway is the terror of no escape through non-being. Fred cannot annihilate himself, cannot step outside the space of possibilities. Even murder fails to end anything. The Mystery Man does not break the metaphysical rules. He enforces them. He behaves like a witness to modal invariance. “I’m at your house right now” is a claim about unrestricted quantification across epistemic positions. The voice on the phone is terrifying because it exposes the inadequacy of first-person access, not because it introduces an entity that should not exist. From this perspective, the film’s affective force aligns with Williamson’s metaphysics. Dread arises precisely because existence is not fragile. The world is oppressively thick. There is no metaphysical mercy. If everything necessarily exists, then trauma cannot be undone by erasure. It can only be displaced, re-indexed, re-experienced under altered descriptions. That is why the film feels like a closed system rather than a broken one. 

Now contrast this with Mulholland Drive, where the same Williamsonian strategy begins to strain. Here the identities do not merely shift properties. They are narratively generated, dissolved, and reconstituted in ways that feel less like modal recombination and more like ontological dependence on fantasy, desire, and failure. Diane Selwyn’s world does not feel like a fixed domain explored under different access relations. It feels like a world that thickens and thins depending on psychic necessity. Betty does not look like Diane under a different property profile. She looks like an attempt to create a different essence altogether, one that fails. Fine would say that the necessity operating in Mulholland Drive is not metaphysical necessity in Williamson’s sense at all. It is affective and narrative necessity grounded in the essence of the character’s psychic economy. Diane must replay her failure because of what she is, not because the modal structure of the world demands it. The film’s logic is not that of invariant existence but of grounded repetition. The blue box does not merely reveal another region of modal space. It collapses a fragile construction whose necessity was local, internal, and essentially unstable. 

This shows why Williamson’s metaphysics fits some Lynch films better than others. Where identity fracture is governed by invariance and looping, necessitism feels explanatory. Where identity fracture is governed by fantasy generation and collapse, essence and grounding begin to do real work. When we reach Inland Empire, the film actively resists any attempt to stabilise the domain. Characters proliferate, not as alternative instantiations of a fixed individual, but as partial realisations of roles, traumas, and narrative functions that bleed across temporal and spatial boundaries. The woman, the actress, the prostitute, the victim, the watcher are not variants of a single rigid designator. They are manifestations of an essence that is itself fractured, relational, and historically contaminated. 

A Williamsonian approach can still work. One might say that all these figures necessarily exist, and that the film merely explores different ways they are related, misidentified, or epistemically obscured. The rabbits exist. The rooms exist. The loops exist. Everything exists. The film is a vast modal exploration of a fixed ontology. But this is exactly where Fine’s critique bites hardest. Because what Inland Empire insists on, both formally and affectively, is that necessity itself is not uniform. The dread in the film does not arise from invariant existence. It arises from grounded inevitability. Certain scenes must happen because of historical wrongs, exploitation, gendered violence, and repetition across generations. 

These are not necessities that float free of context. They are necessities anchored in what these practices and roles are. A woman caught in a cycle of exploitation does not repeat because modal logic requires it. She repeats because the structure she inhabits makes repetition constitutive. 

Here a Beiser-inflected post-Kantian background can help. Schelling’s insistence that history and freedom generate their own necessities, Hegel’s claim that tragedy arises from conflicting ethical structures rather than abstract laws, and the Romantic emphasis on irreducible affect all point toward a metaphysics in which necessity is layered and internally differentiated. 

Inland Empire behaves like a cinematic instantiation of that insight. The film does not present a world whose horror comes from too much being. It presents a world whose horror comes from damaged grounds. Essences have been altered. Relations that once held no longer do. Release does not restore an original state. It inaugurates a new deformation. This is exactly the Finean move wjereby essence precedes necessity, and when essences change, the modal landscape changes with them. There is no return to a pre-traumatised present because there never was a pure essence to return to. The attempt to “fix” things in Inland Empire fails not because the modal structure forbids it, but because the grounding relations have already been transformed. The world has been thinned and thickened in irreversible ways. 

Williamson might respond that this is all true at the level of narrative explanation but irrelevant to metaphysics. Metaphysics, he could say, does not aim to capture every internal necessity operative in artworks or social structures. It aims to capture the most general structure of what there is. But Inland Empire forces a further question. If the most general structure of what there is necessarily includes social, psychological, historical, and aesthetic entities, then why are the necessities internal to those entities not metaphysically salient? Once existence is guaranteed, salience cannot be filtered ontologically. It must be filtered explanatorily. This is where the films expose the cost of each framework. 

Williamson’s metaphysics gains clarity and discipline but risks flattening necessity into a single register that cannot explain why some repetitions devastate while others merely recur. Fine’s metaphysics gains explanatory depth and sensitivity to difference but risks fragmenting metaphysics into a mosaic of local grounds with no unifying closure. 

Lynch’s films do not settle the dispute, but they reveal its stakes. Lost Highway shows the power of necessitism to capture inescapability. Mulholland Drive shows where that power begins to falter. Inland Empire shows what happens when necessity itself fractures, when the “must” governing events is no longer uniform but stratified, historical, and affective. 

Seen this way, the post-Kantian tradition Beiser reconstructs is the deep grammar of the problem. Kant taught philosophy to ask under what conditions necessity is possible. The Idealists argued about whether those conditions were formal, historical, or constitutive. Fine and Williamson are continuing that argument with new tools. Lynch stages it as horror.