- Below are very substantial extracts from the first two sections, and the concluding section.
- I’ve removed the logical schemata – mainly because of the bother of finding the HTML codes for the symbols – and the diagram(s).
- I’ve also removed some detail, indicated by lacunae (“…”).
- I’ve also removed the references – copious and useful in parts.
- The extracts below discuss the logic of possibility, and not zombies or ghosts – and their metaphysical import – as such.
5.1 The Basic Arguments
- Zombies are special twice over. They're physically like us but lack Subjectivity. There's nothing it's like to be them. Though like you in all physical respects, zombie-you isn't like you 'upstairs'. It has no phenomenal consciousness. Its lights are out on the inside. Zombies are the undead of philosophical thought experiment1. They motivate an influential argument for dualism:
P → It's conceivable that zombies exist. (The argument) is formally valid. But conceivability and possibility must each be clarified before soundness can be judged. The key is whether a sensible take can be found to make (the three premises) jointly true.
P → If it's conceivable that zombies exist, zombies can exist.
P → If zombies can exist, dualism is true.
C → Zombies can exist.
C → Dualism is true.
- Ghosts are special twice over. They're phenomenally like us but lack bodies. Though like you in all phenomenal respects, ghost-you isn't like you 'downstairs'. It has no body. Its lights are on with no inside. Ghosts are the disembodied2 of philosophical thought experiment3, They motivate an influential argument for dualism:-
P → It's conceivable that ghosts exist. (The argument) is formally valid. But once more conceivability and possibility must be clarified before soundness can be judged. The key is whether a sensible take can be found to make (the three premises) jointly true.
P → If it's conceivable that ghosts exist, ghosts can exist.
P → If ghosts can exist, dualism is true.
C → Ghosts can exist.
C → Dualism is true.
- We investigate in reverse order.
→ §5.2 covers possibility.
→ §5.3 covers conceivability.
→ §§5.4-5 deal with arguments which employ them.
5.2 Genuine Possibility
- Suppose φ can be true. This is a genuine possibility if it's a mind- and language-independent fact. φ is genuinely possible if its possibility does not spring from how we think or talk (even in the rational ideal). Genuine possibility is like genuine actuality. It does not depend on us for its existence. It does not depend on us for its nature. Genuine possibility is a realistic domain of fact.
- Ordinary practice harbours many senses of 'can'. Not all mark genuine possibility. Our practice is modally4 muddled. It all but obliges equivocation. We must respect it but avoid its pitfalls. Above all, we must not mistake ersatz possibility for the genuine item. In what follows we should highlight the type of 'can' being dealt with. We should parade that type5 on the surface of discussion.
- To that end, consider four claims:
- Water is made of oxygen and not made of oxygen.
- Lincoln survived his assassination.
- Gold is uncomposed6.
- David Lewis jumps Mount Everest in a single bound.
- It's standard to assess their possibility via four criteria;
- (F) The Formal criterion marks possibility by logical form. It says a claim can be true when it's not of the form φ&¬φ. Explicit contradictions fail the test. Everything else passes. From the above list, then, (1) fails while (2) thru (4) pass. The latter three are tagged possible on the basis of their logical form, (1) is not. Let's symbolize this … In the vernacular; (1) is not formally possible; (2)-(4) are formally possible.
- (C) The Conceptual criterion marks possibility by conceptual content. It says a claim can be true when its falsity is not ensured by the concepts from which it's built. In other words, it says a claim can be true when full grasp of its content reveals nothing to preclude truth. Explicit contradictions fail the test. So do conceptual absurdities. Everything else passes. From the above list, then, (3) and (4) make the grade. They're tagged possible on the basis of content. A full grasp of that content reveals nothing to preclude truth. This distinguishes them from (1) and (2). A full grasp of these latter claims reveals formal/conceptual barriers to truth. Let's symbolize this … In the vernacular: (1) and (2) are not conceptually possible; (3) and (4) are conceptually possible.
- (M) The Metaphysical criterion marks possibility by essence. It says a claim can be true when its falsity is not ensured by the essence of its truth-makers. It says a claim can be true when the entities and features of which it speaks do not, by their nature, make the claim false. Explicit contradictions fail the test. So do conceptual absurdities. And since gold is essentially composed – let's say7 - (3) fails as well. From the above list, then, only (4) makes the grade. It alone gets tagged possible on the basis of the metaphysics of its subject matter. Let's symbolize this … In the vernacular; (1)-(3) are not metaphysically possible; (4) is metaphysically possible.
- (N) The Nomic criterion marks possibility by natural law. It says a claim can be true when its truth is not prevented by such law. Explicit contradictions, conceptual absurdities, and metaphysical slip-ups fail the test. Claims like (4) do as well. Natural law precludes Lewis jumping Mount Everest. Let's symbolize this … In the vernacular: (1)-(4) are nomically impossible.
- The formal mark of possibility applies to (2)-(4). The conceptual mark applies to (3) and (4). The metaphysical mark applies to (4) alone. And the nomic mark applies to none of the claims.
- This illustrates something important. The pat criteria used to mark possibility nest. Satisfaction of (N) implies that of (M). Satisfaction of (M) implies that of (C). Satisfaction of (C) implies that of (F). But not the other way round. …
- Here's a heuristic. Start with claims in a bag. Think of modal8 criteria as filters. Dump the bag through the filters. The formal one lets most everything through. It's extremely coarse-grained. Conceptual, metaphysical and nomic filters winnow out more and more claims. … Our job is to see which criteria mark genuine possibility.
- I work with the standard view. It says
- The Formal criterion does not mark genuine possibility;
- The Conceptual criterion does not mark genuine possibility;
- The Metaphysical criterion does mark genuine possibility; and
- The Nomic criterion also marks genuine possibility.
- Let's take each in turn. …
- 5.3 Conceivability
- 5.4 Rejecting the Arguments
- 5.5 Common-sense Ghosts
(Author’s) Discussion Points
- This chapter traffics in three main areas: conceivability, concepts and modality10. None are well understood. I expect debate in these areas to flourish in the immediate future, and results therefrom to augment our take on Zombies and Ghosts. This is exciting when one thinks where debate will take place. The theory of concepts, after all, is part of psychology; that of modality11 is part of logic / metaphysics. Meshing the areas should prove most interesting.
- Here I shoot for non-trivial progress with conceivability. And I commit to realism about genuine possibility. Other than that I do one of two things: remain neutral or work with received opinion. For instance, I go with the assumption that modalities12 nest …. On this common-but-normally-inarticulate view: the space of formal possibility contains that of conceptual possibility; the space of conceptual possibility contains that of metaphysical possibility; the space of metaphysical possibility contains that of nomic possibility; but just the latter two are genuine; only they are mind- and language-independent modality13.
- I don't much like the perspective. For one thing, metaphysical modality14 is got through intuition bashing. One confesses intuition about essence or whatever and then form-fits metaphysical modality15 accordingly. By my lights, though, there's little reason to think such intuition is reliable or stable. And for this reason, metaphysical modality16 strikes me as a creature of darkness (to borrow a phrase from a willing lender). But nothing in this chapter hangs on that impression.
- Further: Chapter 4 exposed a procedural sense of necessity which might be conceptual necessity. If they turn out to be one, however, the space of conceptual possibility will probably fail to contain that of genuine possibility (metaphysical or nomic). One should thus ask: is procedural modality17 conceptual modality18?
- The answer turns on how concepts are individuated. Epistemic views portend a yes … atomistic views do the reverse ... I remain basically neutral. It seems to me we have little to work with but our pre-theoretic concept of concept. The philosophical literature is in post-Twin-Earth disarray. The psychological literature is in its infancy. Full-dress theory is yet to come. We're left with our pre-theoretic take. And it falls between a purely epistemic view and a purely atomistic one. ….
- This chapter depends on the view that neither conceptual possibility nor experiential imagination secure genuine possibility. It rests on the claim, for short, that conceivability does not entail possibility.
- One might defend the entailment by appeal to so-called two-dimensional semantics. On this approach, concepts have two 'readings' (or functions-to-extension - which may or may not differ). One goes with apriori matters such as conceivability. The other goes with aposteriori matters such as essence. Setting logical concepts aside, then, 'X is Y' can be heard four ways:
Subscripted '1's and '2's mark when a concept is heard in the apriori / aposteriori way respectively. The machinery permits an argument:
- X1 is Y1
- X1 is Y2
- X2 is Y1
- X2 is Y2
Conceivability and possibility do not pull apart. Appearances to the contrary deceive. Specifically, they result from projecting the modal19 status of readings like (ii) - (iv) onto readings like (i). If a claim is conceivable when read in the purely apriori way, however, it is possible. Or again: if a claim is conceivable when all its constituent concepts are read in the apriori way, it is possible. Conceivability secures possibility after all. The idea here is simple: when conceivability and possibility look to cohabit, that's because we 'cross read' claims. We read some constituent concepts in the apriori way and others in the aposteriori way. The result is illusion of conceivability without genuine possibility.
- Now, one can motivate semantic dimensions in various ways. Three are currently popular: appeal to the theory of reference, appeal to linguistic intuition, and reflection on supposition. Each can be used to build a two-dimensional semantics. And in turn each two-dimensional semantics can be used to analyse claims like
and their negations. Moreover, the analysis can be used to argue such claims do not break the link from conceivability to possibility. But I say the project is doomed. It's both ill-motivated and badly founded.
- David Lewis is Bruce LeCatt,
- Water is H2O
- It’s ill-motivated for this reason: we should expect conceivability not to secure genuine possibility. After all, such possibility is mind- and language independent. Our methods for interrogating it should turn out fallible. We should expect there to be no method of modal20 interrogation guaranteed to yield truth. It's one thing for thought about an objective domain to contain methods guaranteed to preserve truth. It's another for them to contain methods guaranteed to yield truth. If conceivability entails genuine possibility, however, that's exactly what conceivability is. It's a method of modal21 interrogation guaranteed to yield truth about mind- and language-independent fact. By my lights we should expect not to have such a method. The two-dimensional attempt to create one is ill-motivated.
- And it's unfounded for this reason: no one's made clear where the dimensions come from. No one's made clear how the dimensions in question reflect competent grasp of concepts which bear them. Further, no one's explained why they should be thought capable of playing their designated role in the approach. I do not say this is impossible, of course. But I do say no one has done it. The literature is frustratingly bald hereabouts. It lacks a convincing story about why matters of aprioriticy and modality22 should turn on dimensions built, say, from the theory of reference.
- The two-dimensional approach is pursued vigorously by David Chalmers. See work on his website, ….
Footnote 5: Sturgeon does this by modifying the “possibility” marker (◊) from modal logic, adding underlines, *’s or dots depending on the type of possibility.
Footnote 6: What does this mean?
Footnote 7: Well, yes … but what does this mean? A better example could surely be found.
Footnote 9: I’m not sure why this referencing is used.
Text Colour Conventions (see disclaimer)
- Blue: Text by me; © Theo Todman, 2020
- Mauve: Text by correspondent(s) or other author(s); © the author(s)