green numbers give full details.     |    back to list of philosophers     |     unexpand these ideas

Ideas of Robert C. Stalnaker, by Text

[American, b.1940, Professor at Cornell University, then at the Massachusetts Institute of Technology.]

1968 A Theory of Conditionals
p.34 p.18 In nearby worlds where A is true, 'if A,B' is true or false if B is true or false
     Full Idea: Consider a possible world in which A is true and otherwise differs minimally from the actual world. 'If A, then B' is true (false) just in case B is true (false) in that possible world.
     From: Robert C. Stalnaker (A Theory of Conditionals [1968], p.34), quoted by Dorothy Edgington - Conditionals (Stanf) 4.1
     A reaction: This is the first proposal to give a possible worlds semantics for conditional statements. Edgington observes that worlds which are nearby for me may not be nearby for you.
p.34 p.18 A possible world is the ontological analogue of hypothetical beliefs
     Full Idea: A possible world is the ontological analogue of a stock of hypothetical beliefs.
     From: Robert C. Stalnaker (A Theory of Conditionals [1968], p.34), quoted by Dorothy Edgington - Conditionals (Stanf) 4.1
     A reaction: Sounds neat and persuasive. What is the ontological analogue of a stock of hopes? Heaven!
1970 works
p.82 Conditionals are true if minimal revision of the antecedent verifies the consequent
     Full Idea: Stalnaker proposes that a conditional is true if its consequent is true in the minimal revision in which the antecedent is true, that is, in the most similar possible world in which the antecedent is true.
     From: report of Robert C. Stalnaker (works [1970]) by Stephen Read - Thinking About Logic Ch.3
     A reaction: A similar account of counterfactuals was taken up by Lewis to give a (rather dubious) account of causation.
1976 Possible Worlds
p.303 We can take 'ways things might have been' as irreducible elements in our ontology
     Full Idea: Stalnaker suggests talking 'ways things might have been' as sui generis elements of our ontology - actual abstract entities in their own right, not to be reduced to more familiar items.
     From: report of Robert C. Stalnaker (Possible Worlds [1976]) by William Lycan - The Trouble with Possible Worlds 09
     A reaction: This seems to rest on an ontology of 'states of affairs', favoured by Armstrong, and implied in the Tractatus. How big is a state of affairs? How manys states of affairs can be co-present?
1978 Assertion
p.37 An assertion is an attempt to rule out certain possibilities, narrowing things down for good planning
     Full Idea: Stalnaker's guiding idea is that in making an assertion the speaker is trying to get the audience to rule out certain possibilities. ....If all goes well, further planning will proceed on the basis of a smaller and more accurate range of possibilities.
     From: report of Robert C. Stalnaker (Assertion [1978]) by Laura Schroeter - Two-Dimensional Semantics
     A reaction: This sounds intuitively rather plausible, and is a nice original thought. This is what we pay clever chaps like Stalnaker to come up with. It seems to imply some notion of verisimilitude (qv. under 'truth'), depending on how much narrowing happens.
p.126 An assertion aims to add to the content of a context
     Full Idea: Stalnaker starts with the general thesis that the role of a successful assertion of s is to update the context by adding to it the content of s.
     From: report of Robert C. Stalnaker (Assertion [1978]) by Ofra Magidor - Category Mistakes 5.3.2
     A reaction: This is to be compared with criteria of meaningfulness, such as verificationism, and with Grice's rules of conversational implicature. Presumably if you assert what the context presupposes, you fail to assert, without being meaningless.
1979 Anti-essentialism
p.71 p.71 An essential property is one had in all the possible worlds where a thing exists
     Full Idea: If necessity is explained in terms of possible worlds, ...then an essential property is a property that a thing has in all possible worlds in which it exists.
     From: Robert C. Stalnaker (Anti-essentialism [1979], p.71)
     A reaction: This seems to me to be a quite shocking confusion of necessary properties with essential properties. The point is that utterly trivial properties can be necessary, but in no way part of the real essence of something.
p.71 p.72 Bare particular anti-essentialism makes no sense within modal logic semantics
     Full Idea: I argue that one cannot make semantical sense out of bare particular anti-essentialism within the framework of standard semantics for modal logic.
     From: Robert C. Stalnaker (Anti-essentialism [1979], p.71)
     A reaction: Stalnaker characterises the bare particular view as ANTI-essentialist, because he has defined essence in terms of necessary properties. The bare particular seems to allow the possibility of Aristotle being a poached egg.
p.73 p.73 Necessarily self-identical, or being what it is, or its world-indexed properties, aren't essential
     Full Idea: We can remain anti-essentialist while allowing some necessary properties: those essential to everything (self-identity), relational properties (being what it is), and world-indexed properties (being snub-nosed-only-in-Kronos).
     From: Robert C. Stalnaker (Anti-essentialism [1979], p.73)
     A reaction: [a summary] He defined essential properties as necessary properties (Idea 12761), and now backpeddles. World-indexed properties are an invention of Plantinga, as essential properties to don't limit individuals. But they are necessary, not essential!
p.76 p.76 For the bare particular view, properties must be features, not just groups of objects
     Full Idea: If we are to make sense of the bare particular theory, a property must be not just a rule for grouping individuals, but a feature of individuals in virtue of which they may be grouped.
     From: Robert C. Stalnaker (Anti-essentialism [1979], p.76)
     A reaction: He is offering an objection to the thoroughly extensional account of properties that is found in standard possible worlds semantics. Quite right too. We can't give up on the common sense notion of a property.
p.79 p.79 Why imagine that Babe Ruth might be a billiard ball; nothing useful could be said about the ball
     Full Idea: I cannot think of any point in making the counterfactual supposition that Babe Ruth is a billiard ball; there is nothing I can say about him in that imagined state that I could not just as well say about billiard balls that are not him.
     From: Robert C. Stalnaker (Anti-essentialism [1979], p.79)
     A reaction: A bizarrely circumspect semanticists way of saying that Ruth couldn't possibly be a billiard ball! Would he say the same about a group of old men in wheelchairs, one of whom IS Babe Ruth?
p.85 p.85 Logical space is abstracted from the actual world
     Full Idea: Logical space is not given independently of the individuals that occupy it, but is abstracted from the world as we find it.
     From: Robert C. Stalnaker (Anti-essentialism [1979], p.85)
     A reaction: I very much like the second half of this idea, and am delighted to find Stalnaker endorsing it. I take the logical connectives to be descriptions of how things behave, at a high level of generality.
1987 Counterparts and Identity
p.159 To say there could have been people who don't exist, but deny those possible things, rejects Barcan
     Full Idea: Stalnaker holds that there could have been people who do not actually exist, but he denies that there are things that could have been those people. That is, he denies the unrestricted validity of the Barcan Formula.
     From: report of Robert C. Stalnaker (Counterparts and Identity [1987]) by Ian Rumfitt - The Boundary Stones of Thought 6.2
     A reaction: And quite right too, I should have thought. As they say, Jack Kennedy and Marilyn Monroe might have had a child, but the idea that we should accept some entity which might have been that child but wasn't sounds like nonsense. Except as fiction…..
1 p.112 Unlike Lewis, I defend an actualist version of counterpart theory
     Full Idea: I defend a version of counterpart theory that is quite different from Lewis's version, as it is tied to actualism (all that exists is part of the actual world) rather than possibilism (possible things may exist without actually existing).
     From: Robert C. Stalnaker (Counterparts and Identity [1987], 1)
     A reaction: This could be the theory I am after. I am sympathetic to both actualism and to counterpart theory. Off to the woodshed….
2 p.113 Extensional semantics has individuals and sets; modal semantics has intensions, functions of world to extension
     Full Idea: Semantic values in extensional semantics are extensions, like individuals for terms, and sets for predicates. In modal semantics we have intensions, functions from worlds to appropriate extensions.
     From: Robert C. Stalnaker (Counterparts and Identity [1987], 2)
     A reaction: It seems obvious that the meaning of a word like 'giraffe' must include possible giraffes, as well as actual and deceased giraffes.
2 p.114 If possible worlds really differ, I can't be in more than one at a time
     Full Idea: Nothing can be in two places at once. If other possible worlds are really other universes, then clearly, you and I cannot be in them if we are here in this one.
     From: Robert C. Stalnaker (Counterparts and Identity [1987], 2)
     A reaction: This can be sensibly expressed without possible worlds. I can't embody my other possibilities while I am embodying this one (I'm too busy). Insofar as possible worlds are a good framework, they are just a precise map of common sense.
2 p.118 If counterparts exist strictly in one world only, this seems to be extreme invariant essentialism
     Full Idea: Counterparts involve the thesis that domains of different possible worlds are disjoint: possible individuals exist in at most one possible world. This seems to suggest extreme essentialism, where nothing could differ from how it is.
     From: Robert C. Stalnaker (Counterparts and Identity [1987], 2)
     A reaction: He quotes Salmon (1981:236) as saying counterpart theory is particularly inflexible essentialism. This is a long way from my use of 'essentialism'. The problem is just the extent to which my counterpart is 'the same' as me.
1997 Reference and Necessity
Intro p.166 Kripke's possible worlds are methodological, not metaphysical
     Full Idea: The possible worlds framework that Kripke introduces should be understood not as a metaphysical theory, but as a methodological framework.
     From: Robert C. Stalnaker (Reference and Necessity [1997], Intro)
     A reaction: That's certainly how I see possible worlds. I lose no sleep over whether they exist. I just take a set of possible worlds to be like cells in a spreadsheet, or records in a database.
§1 p.166 'Descriptive' semantics gives a system for a language; 'foundational' semantics give underlying facts
     Full Idea: 'Descriptive' semantics gives a semantics for the language without saying how practice explains why the semantics is right; …'foundational' semantics concerns the facts that give expressions their semantic values.
     From: Robert C. Stalnaker (Reference and Necessity [1997], §1)
     A reaction: [compressed] Sounds parallel to the syntax/semantics distinction, or proof-theoretical and semantic validity. Or the sense/reference distinction! Or object language/metalanguage. Shall I go on?
2 p.170 If it might be true, it might be true in particular ways, and possible worlds describe such ways
     Full Idea: A clarifying assumption is that if something might be true, then it might be true in some particular way. …Possible worlds begin from this, and the assumption that what might be true can be described as how a possibility might be realised.
     From: Robert C. Stalnaker (Reference and Necessity [1997], 2)
     A reaction: This is a leading practitioner giving his best shot at explaining the rationale of the possible worlds approach, addressed to many sceptics. Most sceptics, I think, don't understand the qualifications the practitioners apply to their game.
2 p.171 Possible worlds are ontologically neutral, but a commitment to possibilities remains
     Full Idea: I argue for the metaphysical neutrality of the possible worlds framework, but I do not suggest that its use is free of ontological commitment to possibilities (ways things might be, counterfactual situations, possible states of worlds).
     From: Robert C. Stalnaker (Reference and Necessity [1997], 2)
     A reaction: Glad to hear this, as I have always been puzzled at possible aspirations to eliminate modality (such as possibility) by introducing 'possible' worlds. Commitment to possibilities I take to be basic and unavoidable.
2 p.171 Possible worlds allow discussion of modality without controversial modal auxiliaries
     Full Idea: The main benefit of the possible worlds move is to permit one to paraphrase modal claims in an extensional language that has quantifiers, but no modal auxiliaries, so the semantic stucture of modal discourse can be discussed without the controversies.
     From: Robert C. Stalnaker (Reference and Necessity [1997], 2)
     A reaction: The strategy introduces the controversy of possible worlds instead, but since they just boil down to collections of objects with properties, classical logic can reign. Possible worlds are one strategy alongside many others.
3 p.172 To understand an utterance, you must understand what the world would be like if it is true
     Full Idea: To understand what is said in an utterance of 'The first dog born at sea was a basset hound', one needs to know what the world would have been like in order for what was said in that utterance to be true.
     From: Robert C. Stalnaker (Reference and Necessity [1997], 3)
     A reaction: Put like that, the idea is undeniable. Understanding involves truth conditions. Does mean involve the understanding of the meaning. What do you understand when you understand a sentence? Just facts about dogs? Or something in the sentence?
4 p.175 In the use of a name, many individuals are causally involved, but they aren't all the referent
     Full Idea: The causal theory of reference is criticised for vagueness. Causal connections are ubiquitous, and there are obviously many individuals that are causally implicated in the speaker's use of a name, but they aren't all plausible candidates for the referent.
     From: Robert C. Stalnaker (Reference and Necessity [1997], 4)
     A reaction: This seems to be a very good objection. Among all the causal links back to some baptised object, we have to pick out the referential link, which needs a criterion.
4 p.176 If you don't know what you say you can't mean it; what people say usually fits what they mean
     Full Idea: If you don't know what you are saying then you don't mean what you say, and also speakers generally mean what they say (in that what they say coincides with what they mean).
     From: Robert C. Stalnaker (Reference and Necessity [1997], 4)
     A reaction: Both these thoughts seem completely acceptable and correct, but rely on something called 'meaning' that is distinct from saying. I would express this in terms of propositions, which I take to be mental events.
4 p.176 To understand a name (unlike a description) picking the thing out is sufficient?
     Full Idea: If we ask 'what must you know to understand a name?', the naïve answer is that one must know who or what it names - nothing more. (But no one would give this answer about what is needed to understand a definite description).
     From: Robert C. Stalnaker (Reference and Necessity [1997], 4)
     A reaction: Presumably this is naive because names can be full of meaning ('the Empress'), or description and reference together ('there's the man who robbed me') and so on. It's a nice starting point though. A number can serve as a name.
5 p.182 Possible worlds allow separating all the properties, without hitting a bare particular
     Full Idea: The possible worlds framework suggests a way to express the idea that a particular is conceptually separable from its properties without relying on the rejected picture of a bare particular.
     From: Robert C. Stalnaker (Reference and Necessity [1997], 5)
     A reaction: As I read him, Stalnaker's proposal just comes down to replacing each property in turn with a different one. 'Strip away' red by making it green. It being green in w1 doesn't throw extra light. Can it be a bare particular in w37?
5 p.185 Rigid designation seems to presuppose that differing worlds contain the same individuals
     Full Idea: A rigid designator is a designator that denotes the same individual in all possible worlds; doesn't this presuppose that the same individuals can be found in differing possible worlds?
     From: Robert C. Stalnaker (Reference and Necessity [1997], 5)
     A reaction: This is part of Stalnaker's claim that Kripke already has a metaphysics in place when he starts on his semantics and his theory of reference. Kripke needs a global domain, not a variable domain. Possibilities suggest variable domains to me.
2003 Conceptual truth and metaphysical necessity
1 p.202 Critics say there are just an a priori necessary part, and an a posteriori contingent part
     Full Idea: Critics say there are no irreducible a posteriori truths. They can be factored into a part that is necessary, but knowable a priori through conceptual analysis, and a part knowable only a posteriori, but contingent. 2-D semantics makes this precise.
     From: Robert C. Stalnaker (Conceptual truth and metaphysical necessity [2003], 1)
     A reaction: [Critics are Sidelle, Jackson and Chalmers] Interesting. If gold is necessarily atomic number 79, or it wouldn't be gold, that sounds like an analytic truth about gold. Discovering the 79 wasn't a discovery of a necessity. Stalnaker rejects this idea.
1 p.203 The necessity of a proposition concerns reality, not our words or concepts
     Full Idea: The necessity or contingency of a proposition has nothing to do with our concepts or the meanings of our words. The possibilities would have been the same even if we had never conceived of them.
     From: Robert C. Stalnaker (Conceptual truth and metaphysical necessity [2003], 1)
     A reaction: This sounds in need of qualification, since some of the propositions will be explicitly about words and concepts. Still, I like this idea.
1 p.203 Conceptual possibilities are metaphysical possibilities we can conceive of
     Full Idea: Conceptual possibilities are just (metaphysical) possibilities that we can conceive of.
     From: Robert C. Stalnaker (Conceptual truth and metaphysical necessity [2003], 1)
2 p.204 Meanings aren't in the head, but that is because they are abstract
     Full Idea: Meanings ain't in the head. Putnam's famous slogan actually fits Frege's anti-psychologism better than it fits Purnam's and Burge's anti-individualism. The point is that intensions of any kind are abstract objects.
     From: Robert C. Stalnaker (Conceptual truth and metaphysical necessity [2003], 2)
     A reaction: If intensions are abstract, that leaves (for me) the question of what they are abstracted from. I take it that there are specific brain events that are being abstractly characterised. What do we call those?
2 p.205 Two-D says that a posteriori is primary and contingent, and the necessity is the secondary intension
     Full Idea: Two-dimensionalism says the necessity of a statement is constituted by the fact that the secondary intensions is a necessary proposition, and their a posteriori character is constituted by the fact that the associated primary intension is contingent.
     From: Robert C. Stalnaker (Conceptual truth and metaphysical necessity [2003], 2)
     A reaction: This view is found in Sidelle 1989, and then formalised by Jackson and Chalmers. I like metaphysical necessity, but I have some sympathy with the approach. The question must always be 'where does this necessity derive from'?
2 p.205 A 'centred' world is an ordered triple of world, individual and time
     Full Idea: A 'centred' possible world is an ordered triple consisting of a possible world, an individual in the domain of that world, and a time.
     From: Robert C. Stalnaker (Conceptual truth and metaphysical necessity [2003], 2)
4 p.209 In one view, the secondary intension is metasemantic, about how the thinker relates to the content
     Full Idea: On the metasemantic interpretation of the two-dimensional framework, the second dimension is used to represent the metasemantic facts about the relation between a thinker or speaker and the contents of her thoughts or utterances.
     From: Robert C. Stalnaker (Conceptual truth and metaphysical necessity [2003], 4)
     A reaction: I'm struggling to think what facts there might be about the relation between myself and the contents of my thoughts. I'm more or less constituted by my thoughts.
5 p.211 One view says the causal story is built into the description that is the name's content
     Full Idea: In 'causal descriptivism' the causal story is built into the description that is the content of the name (and also incorporates a rigidifying operator to ensure that the descriptions that names abbreviate have wide scope).
     From: Robert C. Stalnaker (Conceptual truth and metaphysical necessity [2003], 5)
     A reaction: Not very controversial, I would say, since virtually every fact about the world has a 'causal story' built into it. Must we insist on rigidity in order to have wide scope?
2010 Merely Possible Propositions
p.22 p.22 A 'Russellian proposition' is an ordered sequence of individual, properties and relations
     Full Idea: A 'Russellian proposition' is an ordered sequence containing the individual, along with properties and relations.
     From: Robert C. Stalnaker (Merely Possible Propositions [2010], p.22)
     A reaction: Since Russell took properties and relations to be features of reality, this made the whole proposition a feature of reality. This is utterly different from what I understand by the word 'proposition', which is a feature of thought, not of the world.
p.28 p.28 Predicates can't apply to what doesn't exist
     Full Idea: Nothing can be predicated of something which does not exist.
     From: Robert C. Stalnaker (Merely Possible Propositions [2010], p.28)
     A reaction: [He says he is 'agreeing with Plantinga' on this] This seems very puzzling, as you can obviously say that dragons do not exist, but they breathe fire. Why can't you attach predicates to hypothetical objects?
2012 Mere Possibilities
Pref p.-6 Given actualism, how can there be possible individuals, other than the actual ones?
     Full Idea: My main focus is on how, on an actualist interpretation of possible worlds as ways a world might be, one is to account for the possibility that there be individuals other than those that actually exist.
     From: Robert C. Stalnaker (Mere Possibilities [2012], Pref)
     A reaction: The obvious thought would be that they are constructions from components of actual individuals, such as the chimaera, or fictional characters. We need some psychology here, which is not Stalnaker's style.
1 p.1 Some say what exists must do so, and nothing else could possible exist
     Full Idea: Some philosophers deny there could have been anything other than what in fact exists, or that anything that exists could have failed to exist. This is developed in very different ways by Wittgenstein (in 'Tractatus'), Lewis and Williamson.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 1)
     A reaction: This could come in various strengths. A weak version would say that, empirically, that all talk of what doesn't exist is vacuous. A strong necessity (Williamson?) that totally rules out other possible existence is a very odd view.
1 p.2 Modal concepts are central to the actual world, and shouldn't need extravagant metaphysics
     Full Idea: Modal concepts are central to our understanding of the world - the actual world - and understanding them should not require extravagant metaphysical commitments.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 1)
     A reaction: I agree. Personally I think powers and dispositions do the job nicely. You just have to embrace Leibniz's emphasis on the active nature of reality, and the implausible metaphysics starts to recede.
1 p.3 Possible worlds are properties
     Full Idea: Possible worlds are (to a first approximation) properties. [p.12] They are properties of the total universe.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 1)
1.1 p.4 Necessity and possibility are fundamental, and there can be no reductive analysis of them
     Full Idea: My view is that if there were a nonmodal analysis of the modal concepts, that would be a sure sign that we were on the wrong track. Necessity and possibility are fundamental concepts, like truth and existence.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 1.1)
     A reaction: The mystery of modality is tied up with the mystery of time (which is a very big mystery indeed). You get a nice clear grip on the here and now, but time and motion whisk you away to something else. Modality concerns the something else.
1.1 p.4 A nominalist view says existence is having spatio-temporal location
     Full Idea: A nominalist definition of existence is 'having spatio-temporal location'.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 1.1)
     A reaction: This would evidently be physicalist as well as nominalist. Presumably it fits the 'mosaic' of reality Lewis refers to. I find this view sympathetic. A process of abstraction is required to get the rest of the stuff we talk about.
1.1 p.5 I don't think Lewis's cost-benefit reflective equilibrium approach offers enough guidance
     Full Idea: Lewis articulated and made fashionable the cost-benefit reflective equilibrium methodology, but I have my reservations as it does not offer much guidance.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 1.1)
     A reaction: Stalnaker suggests that this approach has 'run amok' in Lewis's case, giving reality to possible worlds. He spends much effort on showing the 'benefits' of a profoundly implausible view. The same can be said of 4D Perdurantism.
1.2 p.11 Properties are modal, involving possible situations where they are exemplified
     Full Idea: I take properties and relations to be modal notions. Properties are to be understood in terms of what it would be for them to be exemplified, which means understanding them in terms of a range of possible situations.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 1.2)
     A reaction: I can't make head or tail of a property as anything other than a feature of some entity. Treating properties as a 'range of situations' is just as baffling to me as treating them as sets of objects.
1.2 p.11 Possible worlds don't reduce modality, they regiment it to reveal its structure
     Full Idea: It is not reduction (of modality) but regimentation that the possible-worlds framework provides - a procedure for representing modal discourse, using primitive modal notions, in a way that helps reveal its structure.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 1.2)
     A reaction: I think this is exactly my view. All discussion of the ontology of possible worlds is irrelevant. They no more exist than variables in logic exist. They're good when they clarify, but dubious when they over-simplify.
1.2 p.11 I take propositions to be truth conditions
     Full Idea: I will defend the view that propositions are truth conditions.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 1.2)
     A reaction: This sounds close to the Russellian view, which I take to equate propositions (roughly) with facts or states of affairs. But are 'truth conditions' in the world or in the head?
1.2 p.13 I think of worlds as cells (rather than points) in logical space
     Full Idea: I prefer to think of the possible worlds not as points in logical space but as cells of a relatively fine-grained partition of logical space - a partition that makes all the distinctions we need.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 1.2)
     A reaction: Since he regards possible worlds as simply a means of regimenting our understanding of modality, he can think of possible worlds in any way that suits him. I find it hard work tuning in to his vision.
2 p.22 Propositions presumably don't exist if the things they refer to don't exist
     Full Idea: It seems plausible that singular propositions are object-dependent in the sense that the proposition would not exist if the individual did not. It is also plausible that some objects exist contingently, and there are singular propositions about them.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 2)
     A reaction: This replies to the view that possible worlds are maximal sets of propositions, and so must exist for the worlds to exist; e.g. Lowe 1999:248. That is yet another commonplace of contemporary philosophy which I find utterly bewildering.
2 p.24 A theory of propositions at least needs primitive properties of consistency and of truth
     Full Idea: A minimal theory of propositions can make do with just two primitive properties: a property of consistency applied to sets of propositions, and a property of truth applied to propositions.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 2)
     A reaction: I would have thought a minimal theory would need some account of what a proposition is supposed to be (since there seems to be very little agreement about that). Stalnaker goes on to sketch a theory.
2.2 p.30 Possible world semantics may not reduce modality, but it can explain it
     Full Idea: Most theorists agree that possible worlds semantics cannot provide an analysis of modal concepts which is an eliminative reduction, but it can still provide an explanation of the meanings of modal expressions.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 2.2)
     A reaction: Stalnaker cites Kit Fine for the view that there is no reduction of modality, which Fine takes to be primitive. Stalnaker defends the semantics, while denying the reduction which Lewis thought possible.
2.4 p.38 In modal set theory, sets only exist in a possible world if that world contains all of its members
     Full Idea: One principle of modal set theory should be uncontroversial: a set exists in a given possible world if and only if all of its members exist at that world.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 2.4)
     A reaction: Does this mean there can be no set containing all of my ancestors and future descendants? In no world can we coexist.
3 p.52 Anti-haecceitism says there is no more to an individual than meeting some qualitative conditions
     Full Idea: The anti-haecceitist strategy holds that a purely qualitative characterisation of a possible world would be a complete characterisation; there is, on this view, nothing to being a particular individual other than meeting certain qualitative conditions.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 3)
     A reaction: Not quite the same as the bundle theory of objects, which says the objects are the qualities. This is about individuation, not about ontology (I think). I don't like anti-haecceitism, but I also don't like haecceitism. Hmm.
3.4 p.60 Dispositions have modal properties, of which properties things would have counterfactually
     Full Idea: Dispositional properties deserve special mention since they seem to be properties that have modal consequences - consequences for what properties the individuals that instantiate them would have in counterfactual circumstances.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 3.4)
     A reaction: I take this to be the key idea in trying to understand modality, but Stalnaker makes this point and then moves swiftly on, because it is so far away from his possible worlds models, in which he has invested a lifetime.
3.6 p.71 The bundle theory makes the identity of indiscernibles a necessity, since the thing is the properties
     Full Idea: On the bundle theory, the identity of indiscernibles (for 'individuals') is a necessary truth, since an individual is just the co-instantiation of all the properties represented by a point in the space of properties.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 3.6)
     A reaction: So much the worse for the bundle theory, I presume. Leibniz did not, I think, hold a bundle theory, but his belief in the identity of indiscernibles seems to have had a theologicial underpinning.
3.6 p.71 Modal properties depend on the choice of a counterpart, which is unconstrained by metaphysics
     Full Idea: Things have modal properties only relative to the choice of a counterpart relation, and the choice between alternative counterpart relations is not constrained by the metaphysics.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 3.6)
     A reaction: Stalnaker is sympathetic to counterparts, but this strikes me as a powerful objection to the theory. I take the modal properties of something to be fixed by its actuality.
4 p.91 We still lack an agreed semantics for quantifiers in natural language
     Full Idea: We still do not know how to give a direct semantics for the quantifiers of a natural language; that is something that we still do not know how to do (or at least how it is done remains controversial).
     From: Robert C. Stalnaker (Mere Possibilities [2012], 4)
     A reaction: I am struck by how rapidly the domain of quantification changes, even in mid-sentence, in the course of an ordinary conversation. This is decided almost entirely by context, not by pure ('direct'?) semantics.
4.2 p.98 We regiment to get semantic structure, for evaluating arguments, and understanding complexities
     Full Idea: The point of regimentation is to give a perspicuous representation of the semantic structure of an expression, making it easier to evaluate the validity of arguments and to interpret complex statements.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 4.2)
     A reaction: This is an authoritative summary from an expert of why all philosophers must take an interest in logical form.
4.2 p.100 In 'S was F or some other than S was F', the disjuncts need S, but the whole disjunction doesn't
     Full Idea: In 'either Socrates was a philosopher or someone other than Socrates was a philosopher', both propositions expressed by the disjuncts depend for their existence on the existence of Socrates, but the whole disjunction does not.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 4.2)
     A reaction: Nice example, just the sort of thing we pay philosophers to come up with. He is claiming that propositions can exist in possible worlds in which the individuals mentioned do not exist.
4.3 p.103 Strong necessity is always true; weak necessity is cannot be false
     Full Idea: Prior had a strong and a weak reading of necessity, where strong necessity is truth in all possible worlds, while weak necessity is falsity in no possible world.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 4.3)
     A reaction: [K.Fine 2005:Ch.9 is also cited] The point of the weak one is that in some worlds there might not exist the proposition which is the candidate for truth or falsehood.
4.3 p.111 'Socrates is essentially human' seems to say nothing could be Socrates if it was not human
     Full Idea: It seems natural to paraphrase the claim that Socrates is essentially human as the claim that nothing could be Socrates if it was not human.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 4.3)
     A reaction: In ordinary speech it would be emphasising how very human Socrates was (in comparison with Frege, for example). By this token Socrates essentially breathes oxygen, but that is hardly part of his essence.
4.3 n17 p.112 Non-S5 can talk of contingent or necessary necessities
     Full Idea: One can make sense of necessary versus contingent necessities in a non-S5 modal semantics.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 4.3 n17)
     A reaction: In S5 □φ → □□φ, so all necessities are necessary. Does it make any sense to say 'I suppose this might have been necessarily true'?
4.4 p.118 I accept a hierarchy of properties of properties of properties
     Full Idea: I myself am prepared to accept higher-order properties and relations. There is the property of being Socrates, …and the property of being the property of being Socrates, ..and so on.
     From: Robert C. Stalnaker (Mere Possibilities [2012], 4.4)
     A reaction: Elsewhere I have quoted such a hierarchy of vacuous properties as an absurdity that arises if all predicates are treated as properties. Logicians can live with such stuff, given their set hierarchy and so on, but in science and life this is a nonsense.
5 p.133 How can we know what we are thinking, if content depends on something we don't know?
     Full Idea: How can we know what we ourselves are thinking if the very existence of the content of our thought may depend on facts of which we are ignorant?
     From: Robert C. Stalnaker (Mere Possibilities [2012], 5)
     A reaction: This has always been my main doubt about externalism. I may defer to experts about what I intend by an 'elm' (Putnam's example), but what I mean by elm is thereby a fuzzy tall tree with indeterminate leaves. I don't know the meaning of 'elm'!