Summer Reading Group II

Intro

This is to be a space for the summer two-dimensional semantics reading group. Most of the readings will come from an anthology edited by Manuel Garcia-Carpintero and Josep Macià, Two-Dimensional Semantics. If you are interested in joining, please email dsidi [*at*] u [*dot*] arizona [*dot*] edu.

To add the first time: (1) register with the wiki, (2) click the "edit" link above (3) click and read 'show formatting help', then click and read 'more formatting help' (4) choose a legible color for your posts (the last person who adds should append a signature with his/her name and the date). Please type '%ENDCOLOR% %[whatever color]%' before your addition, and '%ENDCOLOR%' immediately following it.

Here is a list of TWiki TextFormattingRules. If you have trouble doing something, you can also look at the preexisting text in the "edit" box as an example. Its pretty easy, just play around a bit.-- DavidSidi - 17 May 2006

David Sidi's remarks are in indigo.

Links

• Chalmers's weblog post for May 02 is on 2DS, and includes a few links.
• Chalmers's syllabus for readings on 2DS

Discussion

• Motivations for two-dimensional semantics (to be updated):
  • accounting for modal illusions (some necessary truths appear contingent, and vice versa) in a way that respects our epistemic access to modality (things about counterfactual situations are stipulated, not found out).
    • modal illusions occur for indexicals, demonstratives, natural kind terms, proper names, definite descriptions
  • to describe the structure of discourse generally (without the details of how discourse happens in a particular language).
  • to explain a priority and necessity in terms of epistemic dependence
  • to explain the relationship between concepts and their contents
    • e.g.: indexical concepts (with contextually determined content), phenomenal concepts

Introduction - Macià and Garcia-Carpintero

• Kripke gives an error theory for modal illusions: identities may be contingent between nonrigid designators which fix the reference of rigid designators (or between a rigid designator and a nonrigid designator). Still, identities between rigid designators are necessary.
  • one way that the two dimensions of 2DS can be understood is as variations in the referents, and variations in the reference-fixers.
• "[Stalnaker] assumes that propositions, the contents of assertions and beliefs, can be modelled by means of classes of possible worlds"
  • So possible worlds are to be linguistically innocent. Is the discourse used to produce a theory of possible worlds supposed to be linguistically innocent too? Stalnaker says things about metasemantic interpretations for diagonal propositions -- but all that is explained (at least in the introduction) is how the rows are interpreted metasemantically. What about the other half, the columns?
    • Isn't the relevant notion of possibility going to require a language?
    • If the answer is 'yes,' this makes me think we can iterate the 2D matrix, like so:

	g			h
g		i	j		i	j
	i	T	F	i	?	?
	j	F	T	j	?	?
h		i	j		i	j
	i	?	?	i	?	?
	j	?	?	j	?	?

• • • The idea is to take the composition of the two-dimensional intension (which is a function from pairs of possible worlds to truth values) with a higher-order intension.
      • in the case above: where k is a function from pairs of worlds and centered worlds to truth values, and f is a function from truth values to theories.-- DavidSidi - 07 Jun 2006
        • this might be a way to arbitrate between empirically equivalent but theoretically distinct theories.
    • Given the possibility of such metaphilosophical interpretations of 2DS, what is the general importance of the two dimensions present in all the matrices? I suspect they tell us something about how we use language in theorizing about it, but I haven't thought enough about it yet.-- DavidSidi - 20 May 2006

• • • For example, in Chalmers' article "The Foundations of Two-Dimensional Semantics," he discusses linguistic contextual intensions (LCI). LCI are functions from tokens of the same linguistic type (which express the same thing in a language) to their extensions. He goes on to accept that if there is a difference in reference between expression tokens A and B, they are not of the same linguistic type. This seems to be to be accepting only one dimension in theorizing about linguistic types.
      • E.g., names need not "have their referents essentially"; we can also think of expression tokens as being of the same linguistic type if what they express in a language is determined in the same way. This addresses the substantive question Chalmers brings up earlier: "on the pluralist approach ... can expressions (whether types or tokens) be associated with values that have such and such properties?"

The Foundations of Two-Dimensional Semantics - Chalmers

• the big questions:
  • "What are the fundamental concepts underlying different interpretations of the framework? How are these related? How do the differences between these interpretations explain the difference in the scope and strength of the claims that are made for them? Which interpretations of the framework yield the strongest connections between the first dimension and the rational domain?" p. 63
• Core thesis: For any sentence S, S is a priori iff S has a necessary 1-intension.
  • this is an update of the Kantian part of the "golden triangle" between reason, meaning, and modality.
  • If this is accepted, then we have what we need to support a dimension answering Frege's puzzle, since cognitive significance is tied to a 1-intension, which is tied to the a priori. So the modal account of sense given by Carnap is now explained in purely epistemic terms.
• Does semantic pluralism deny the project of reducing some semantic values to others? Chalmers describes semantic pluralism as allowing expressions to be associated with semantic values (extensions, various intensions, structured propositions, conventionally implied contents) in many different ways, all equally legitimate.
  • Centered worlds are fairly extravagant as an explanation of variation in the use of an utterance.
    • Further, in the examples given to motivate taking 1-intensions as centered worlds, the use of a word like 'Phosphorus' would not be only a matter of the use of the person at the center of the world. As in the actual world, it seems emphasis should be placed on the normative constraints of a language community. Chalmers doesn't give an account of how to get from individual to community.
    • Does Chalmers' position reflect a problem with his ambitious aims for 2D to yield an account of narrow content, and a grounding of the connection b/w conceivability and possibility?
• Cool argument against contextual interpretations of 2DS on p. 70: any sentence token that means the same as "A sentence token exists" is true, and further, has a necessary linguistic contextual intension and semantic contextual intension. But it is also a posteriori, and so it is incompatible with the core thesis. Nice!
• I can't help but think that "deep epistemic necessity" is just a slightly spooky version of analyticity. The basic notion is a priori thought, not sentences. To deny an epistemic necessity is a contradiction, since the a priori rational reflection is ideal.
  • he claims ideal rational reflection as primitivs, but why would we accept that?
• in the construction of scenario in purely epistemic terms, Chalmers makes use of a language L which can express every possible concept, "or at least any concept of a sort" (whatever that means -- it is left unexplained) (p. 84).
  • For Quinean reasons, could we say that this will not be the only such language, nor will it be compatible with some of the others, and so different scenarios will have different epistemic possibilities?
• I do not understand "semantically neutral expressions"

Assertion Revisited - Stalnaker

• The problem motivating 2D for Stalnaker: "with some statements, there is a tension between global intuitions about the information that the statement conveys, as represented by the possibilities that the statement seems to exclude, and what semantic theories that are otherwise well motivated say about the truth conditions of the statement.... The clearest cases are the necessary a posteriori statements that Saul Kripke brought to our attention in _Naming and Necessity_" p. 294
  • propositional concept accounts for
    • facts determining reference, and
    • facts determining semantic value.
• So the diagonal proposition links the semantic facts as someone thinks they are to the semantic facts that do obtain, and which determine reference.
• Whichever semantic facts obtain determine where you start the diagonal.