Ideas of Nathan Salmon, by Theme

[American, b.1951, Professor at the University of California, Santa Barbara.]

green numbers give full details    |    back to list of philosophers    |     unexpand these ideas    |    
2. Reason / D. Definition / 11. Ostensive Definition
Ostensive definitions needn't involve pointing, but must refer to something specific
     Full Idea: So-called ostensive definitions need not literally involve ostension, e.g. pointing, but they must involve genuine reference of some sort (in this case reference to a sample of water).
     From: Nathan Salmon (Reference and Essence (1st edn) [1981], 4.11.2)
4. Formal Logic / D. Modal Logic ML / 2. Tools of Modal Logic / b. Terminology of ML
A world is 'accessible' to another iff the first is possible according to the second
     Full Idea: A world w' is accessible to a consistent world w if and only if w' is possible in w. Being 'inaccessible to' or 'possible relative to' a consistent world is simply being possible according to that world, nothing more and nothing less.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], IV)
     A reaction: More illuminating than just saying that w can 'see' w'. Accessibility is internal to worIds. It gives some connection to why we spend time examining modal logic. There is no more important metaphysical notion than what is possible according to actuality.
4. Formal Logic / D. Modal Logic ML / 3. Modal Logic Systems / d. System T
For metaphysics, T may be the only correct system of modal logic
     Full Idea: Insofar as modal logic is concerned exclusively with the logic of metaphysical modality, ..T may well be the one and only (strongest) correct system of (first-order) propositional logic.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], Intro)
     A reaction: This contrasts sharply with the orthodox view, that S5 (or at the very least S4) is the correct system for metaphysics.
4. Formal Logic / D. Modal Logic ML / 3. Modal Logic Systems / f. System B
System B has not been justified as fallacy-free for reasoning on what might have been
     Full Idea: Even the conventionally accepted system B, which is weaker than S5 and independent of S4, has not been adequately justified as a fallacy-free system of reasoning about what might have been.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], Intro)
In B it seems logically possible to have both p true and p is necessarily possibly false
     Full Idea: The characteristic of B has the form φ⊃□◊φ. ...Even if these axioms are necessarily true, it seems logically possible for p to be true while the proposition that p is necessarily possible is at the same time false.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], Intro)
System B implies that possibly-being-realized is an essential property of the world
     Full Idea: Friends of B modal logic commit themselves to the loaded claim that it is logically true that the property of possibly being realized (or being a way things might have been) is an essential property of the world.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], V)
     A reaction: I think this 'loaded' formulation captures quite nicely the dispositional view I favour, that the possibilities of the actual world are built into the actual world, and define its nature just as much as the 'categorial' facts do.
4. Formal Logic / D. Modal Logic ML / 3. Modal Logic Systems / g. System S4
What is necessary is not always necessarily necessary, so S4 is fallacious
     Full Idea: We can say of a wooden table that it would have been possible for it to have originated from some different matter, even though it is not actually possible. So what is necessary fails to be necessarily necessary, and S4 modal logic is fallacious.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], I)
     A reaction: [compressed]
4. Formal Logic / D. Modal Logic ML / 3. Modal Logic Systems / h. System S5
S4, and therefore S5, are invalid for metaphysical modality
     Full Idea: Salmon argues that S4 and therefore S5 are invalid for metaphysical modality.
     From: report of Nathan Salmon (Reference and Essence (1st edn) [1981], 238-40) by Timothy Williamson - Modal Logic within Counterfactual Logic 4
     A reaction: [He gives references for Salmon, and for his own reply] Salmon's view seems to be opposed my most modern logicians (such as Ian Rumfitt).
S5 modal logic ignores accessibility altogether
     Full Idea: When we ignore accessibility altogether, we have finally zeroed in on S5 modal logic.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], IV)
S5 believers say that-things-might-have-been-that-way is essential to ways things might have been
     Full Idea: Believers in S5 as a correct system of propositional reasoning about what might have been must claim that it is an essential property of any way things might have been that things might have been that way.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], V)
     A reaction: Salmon is working in a view where you are probably safe to substitute 'necessary' for 'essential' without loss of meaning.
The unsatisfactory counterpart-theory allows the retention of S5
     Full Idea: Counterpart-theoretic modal semantics allows for the retention of S5 modal propositional logic, at a considerable cost.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], V n18)
     A reaction: See the other ideas in this paper by Salmon for his general attack on S5 as the appropriate system for metaphysical necessity. He favours the very modest System T.
4. Formal Logic / D. Modal Logic ML / 4. Alethic Modal Logic
Metaphysical (alethic) modal logic concerns simple necessity and possibility (not physical, epistemic..)
     Full Idea: Metaphysical modal logic concerns metaphysical (or alethic) necessity and metaphysical (alethic) possibility, or necessity and possibility tout court - as opposed to such other types of modality as physical necessity, epistemic necessity etc.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], Intro n2)
7. Existence / D. Theories of Reality / 10. Vagueness / g. Degrees of vagueness
It can't be indeterminate whether x and y are identical; if x,y is indeterminate, then it isn't x,x
     Full Idea: Insofar as identity seems vague, it is provably mistaken. If it is vague whether x and y are identical (as in the Ship of Theseus), then x,y is definitely not the same as x,x, since the first pair is indeterminate and the second pair isn't.
     From: Nathan Salmon (Reference and Essence: seven appendices [2005], App I)
     A reaction: [compressed; Gareth Evans 1978 made a similar point] This strikes me as begging the question in the Ship case, since we are shoehorning the new ship into either the slot for x or the slot for y, but that was what we couldn’t decide. No rough identity?
9. Objects / D. Essence of Objects / 7. Essence and Necessity / a. Essence as necessary properties
Essentialism says some properties must be possessed, if a thing is to exist
     Full Idea: The metaphysical doctrine of essentialism says that certain properties of things are properties that those things could not fail to have, except by not existing.
     From: Nathan Salmon (Reference and Essence (1st edn) [1981], 3.8.2)
     A reaction: A bad account of essentialism, and a long way from Aristotle. It arises from the logicians' tendency to fix objects entirely in terms of a 'flat' list of predicates (called 'properties'!), which ignore structure, constitution, history etc.
9. Objects / D. Essence of Objects / 15. Against Essentialism
Any property is attached to anything in some possible world, so I am a radical anti-essentialist
     Full Idea: By admitting possible worlds of unlimited variation and recombination, I simply abandon true metaphysical essentialism. By my lights, any property is attached to anything in some possible world or other. I am a closet radical anti-essentialist.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], II)
     A reaction: Salmon includes impossible worlds within his scheme of understanding. It strikes me that this is metaphysical system which tells us nothing about how things are: it is sort of 'logical idealist'. Later he talks of 'we essentialists'.
10. Modality / A. Necessity / 3. Types of Necessity
Logical possibility contains metaphysical possibility, which contains nomological possibility
     Full Idea: Just as nomological possibility is a special kind of metaphysical possibility, so metaphysical possibility is a special kind of logical possibility.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], III)
     A reaction: This is the standard view of how the three types of necessity are nested. He gives a possible counterexample in footnote 7.
10. Modality / A. Necessity / 5. Metaphysical Necessity
Metaphysical necessity is NOT truth in all (unrestricted) worlds; necessity comes first, and is restricted
     Full Idea: A mythology gave us the idea that metaphysical necessity is truth in every world whatsoever, without restriction. But the notion of metaphysical modality comes first, and, like every notion of modality, it is restricted.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], IV)
Metaphysical necessity is said to be unrestricted necessity, true in every world whatsoever
     Full Idea: It is held that it is the hallmark of metaphysical necessity is that it is completely unrestricted, the limiting case of restricted necessity, with no restrictions whatever. A proposition is necessary only if it is true in absolutely every world whatever.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], II)
     A reaction: This is the standard picture which leads to the claim that S5 modal logic is appropriate for metaphysical necessity, because there are no restrictions on accessibility. Salmon raises objections to this conventional view.
Bizarre identities are logically but not metaphysically possible, so metaphysical modality is restricted
     Full Idea: Though there is a way things logically could be according to which I am a credit card account, there is no way things metaphysically might be according to which I am a credit card account. This illustrates the restricted nature of metaphysical modality.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], III)
     A reaction: His drift is that metaphyical modality is restricted, but expressing it in S5 modal logic (where all worlds see one another) makes it unrestricted, so S5 logic is wrong for metaphysics. I'm impressed by his arguments.
Without impossible worlds, the unrestricted modality that is metaphysical has S5 logic
     Full Idea: If one confines one's sights to genuinely possible worlds, disavowing the impossible worlds, then metaphysical modality emerges as the limiting case - the 'unrestricted' modality that takes account of 'every' world - and S5 emerges as its proper logic.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], IV)
     A reaction: He observes that this makes metaphysical modality 'restricted' simply because you have restricted what 'all worlds' means. Could there be non-maximal worlds? Are logical and metaphysical modality coextensive? I think I like the S5 view.
In the S5 account, nested modalities may be unseen, but they are still there
     Full Idea: The S5 theorist's miscontrual of English (in the meaning of 'possibly possible') makes nested modality unseen, but it does not make nested modality vanish. Inaccessible worlds are still worlds.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], IV)
10. Modality / A. Necessity / 6. Logical Necessity
Logical necessity is free of constraints, and may accommodate all of S5 logic
     Full Idea: With its freedom from the constraint of metaphysical possibility, logical necessity may be construed as accommodating all the axioms and rules of S5.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], III)
     A reaction: He goes on to raise problems for this simple thought. The big question: what are the limits of what is actually possible? Compare: what are the limits of what is imaginable? what are the limits of what is meaningfully sayable?
10. Modality / A. Necessity / 7. Natural Necessity
Nomological necessity is expressed with intransitive relations in modal semantics
     Full Idea: Intransitive relations are introduced into modal semantics for the purposes of interpreting various 'real' or restricted types of modalities, such as nomological necessity.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], II)
     A reaction: The point here is that the (so-called) 'laws of nature' are held to change from world to world, so necessity in one could peter out in some more remote world, rather than being carried over everywhere. A very Humean view of such things.
10. Modality / C. Sources of Modality / 5. Modality from Actuality
Necessity and possibility are not just necessity and possibility according to the actual world
     Full Idea: The real meanings of the simple modal terms 'necessary' and 'possible' are not the same as the concepts of actual necessity and actual possibility, necessity and possibility according to the actual world.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], IV)
     A reaction: If you were an 'actualist' (who denies everything except the actual world) then you are unlikely to agree with this. In unrestricted possible worlds, being true in one world makes it possible in all worlds. So actual necessity is possible everywhere.
10. Modality / E. Possible worlds / 1. Possible Worlds / b. Impossible worlds
Impossible worlds are also ways for things to be
     Full Idea: Total ways things cannot be are also 'worlds', or maximal ways for things to be. They are impossible worlds.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], I)
     A reaction: This unorthodox view doesn't sound too plausible to me. To think of a circular square as a 'way things could be' sounds pretty empty, and mere playing with words. The number 7 could be the Emperor of China?
Denial of impossible worlds involves two different confusions
     Full Idea: Every argument I am aware of against impossible worlds confuses ways for things to be with ways things might have been, or worse, confuses ways things cannot be with ways for things to be that cannot exist - or worse yet, commits both errors.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], III)
     A reaction: He is claiming that 'ways for things to be' allows impossible worlds, whereas 'ways things might have been' appears not to. (I think! Read the paragraph yourself!)
Without impossible worlds, how things might have been is the only way for things to be
     Full Idea: If one ignores impossible worlds, then ways things might have been are the only ways for things to be that are left.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], IV)
     A reaction: Impossible worlds are included in 'ways for things to be', but excluded from 'ways things might have been'. I struggle with a circle being square as a 'way for circles to be'. I suppose being the greatest philosopher is a way for me to be.
10. Modality / E. Possible worlds / 1. Possible Worlds / e. Against possible worlds
Possible worlds rely on what might have been, so they can' be used to define or analyse modality
     Full Idea: On my conception, the notions of metaphysical necessity and possibility are not defined or analyzed in terms of the apparatus of possible worlds. The order of analysis is just the reverse: possible worlds rely on the notion of what might have been.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], IV)
     A reaction: This view seems to be becoming the new orthodoxy, and I certainly agree with it. I have no idea how you can begin to talk about possible worlds if you don't already have some idea of what 'possible' means.
10. Modality / E. Possible worlds / 2. Nature of Possible Worlds / a. Nature of possible worlds
Possible worlds are maximal abstract ways that things might have been
     Full Idea: I conceive of possible worlds as certain sorts of maximal abstract entities according to which certain things (facts, states of affairs) obtain and certain other things do not obtain. They are total ways things might have been.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], I)
Possible worlds just have to be 'maximal', but they don't have to be consistent
     Full Idea: As far as I can tell, worlds need not be logically consistent. The only restriction on worlds is that they must be (in some sense) 'maximal' ways for things to be.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], I)
     A reaction: The normal idea of a maximal model is that it must contain either p or ¬p, and not both, so I don't think I understand this thought, but I pass it on.
10. Modality / E. Possible worlds / 2. Nature of Possible Worlds / c. Worlds as propositions
You can't define worlds as sets of propositions, and then define propositions using worlds
     Full Idea: It is not a good idea to think of possible worlds as sets of propositions, and at the same time to think of propositions as sets of possible worlds.
     From: Nathan Salmon (The Logic of What Might Have Been [1989], I n3)
     A reaction: Salmon favours thinking of worlds as sets of propositions, and hence rejects the account of propositions as sets of worlds. He favours the 'Russellian' view of propositions, which seem to me to be the same as 'facts'.
19. Language / B. Reference / 1. Reference theories
Frege's 'sense' solves four tricky puzzles
     Full Idea: Reference via sense solves Frege's four puzzles, of the informativeness of identity statements, the failure of substitutivity in attitude contexts, of negative existentials, and the truth-value of statements using nondenoting singular terms.
     From: Nathan Salmon (Reference and Essence (1st edn) [1981], 1.1.1)
     A reaction: These must then be compared with Kripke's three puzzles about referring via sense, and the whole debate is then spread before us.
19. Language / B. Reference / 3. Direct Reference / a. Direct reference
The perfect case of direct reference is a variable which has been assigned a value
     Full Idea: The paradigm of a nondescriptional, directly referential, singular term is an individual variable. …The denotation of a variable… is semantically determined directly by the assignment of values.
     From: Nathan Salmon (Reference and Essence (1st edn) [1981], 1.1.2)
     A reaction: This cuts both ways. Maybe we are muddling ordinary reference with the simplicities of logical assignments, or maybe we make logical assignments because that is the natural way our linguistic thinking works.
Kripke and Putnam made false claims that direct reference implies essentialism
     Full Idea: Kripke and Putnam made unsubstantiated claims, indeed false claims, to the effect that the theory of direct reference has nontrivial essentialist import.
     From: Nathan Salmon (Reference and Essence: seven appendices [2005], Pref to Exp Ed)
     A reaction: Kripke made very few claims, and is probably innocent of the charge. Most people agree with Salmon that you can't derive metaphysics from a theory of reference.
26. Natural Theory / B. Natural Kinds / 5. Reference to Natural Kinds
Nothing in the direct theory of reference blocks anti-essentialism; water structure might have been different
     Full Idea: There seems to be nothing in the theory of direct reference to block the anti-essentialist assertion that the substance water might have been the very same entity and yet have had a different chemical structure.
     From: Nathan Salmon (Reference and Essence (1st edn) [1981], 6.23.1)
     A reaction: Indeed, water could be continuously changing its inner structure, while retaining the surface appearance that gets baptised as 'water'. We make the reasonable empirical assumption, though, that structure-change implies surface-change.