Philosophy 1870: Theories of Truth
Syllabus

• The first issue Austin discusses is what is sometimes called 'the problem of truth-bearers'. There are many things to which we ascribe truth. It seems reasonable to suppose that one of these is fundamental and that others are derived. Austin suggests that what is fundamentally true is a statement. What does he mean by a 'statement'? Is he entirely consistent in that usage?
• Austin sets out to defend a form of the correspondence theory of truth, according to which a statement is true if it corresponds to the facts. In developing this view, he distinguishes two types of 'conventions' which relate words to the world: descriptive conventions and demonstrative conventions. What are these? Can you give an example to illustrate how this distinction might apply in a particular case? How does Austin use these to define truth?
• Austin, as you will see, is somewhat taken with his own wit. But there are turns of phrase well worth unpacking. Here is one: "...[I]n discussing truth, ...it is precisely our business to prise the words off the world and keep them off it" (p. 118). What might he mean by that?
• Austin emphasizes that "...the relation between [a statement] and the world which [the statement that that statement is true] asserts to obtain is a purely conventional relation" (p. 122). This might seem to imply that the statement "Bill's statement that the cat is on the mat is true" is about the conventions of whatever language Bill speaks. Does it? Should it?
• In §4, Austin discusses the so-called 'redundancy' theory of truth, according to which the use of the word 'true' is, at least in core cases, wholly dispensible. Austin takes there to be a short argument against this view. What is it? How good is it?

• Strawson argues, in effect, that every aspect of Austin's view is confused: His accounts of statements and facts, and of how truth relates them.
• Concerning statements, Austin regards them as 'historical episodes', basically utterances. Strawson argues that truth is not attributed to such episodes, but rather to what is said by the speaker, though he hems and haws about what 'what is said' might be. (Nowadays, such a view would be expressed by saying that the fundamental truth-bearers are propositions.) What are his reasons?
• Regarding facts, Strawson argues that subject-predicate sentences are 'about' only that to which the subject refers, that " there is nothing else in the world for the statement itself to be related to" (p.134). In particular, Strawson denies that 'facts' are 'things in the world' to which statements might relate. What is his best argument for this claim?
• Strawson insists that 'the problem of truth' is to be resolved by giving an account of the use of the word "true". But, on pp.141-3, he also distinguishes a different problem, that of "elucidating the fact-stating type of discourse". What is this other problem? Is it more or less interesting or important that 'the problem of truth'?
• Strawson repeatedly remarks that 'the problem of truth' involves exploring the use of the word "true" within fact-stating discourse. Relatedly, he claims that "the occurrence in ordinary discourse of the words 'true', 'fact', etc., signalizes, without commenting on, the occurrence of" fact-stating discourse. This is a definite error. Why so?
• A related worry, discussed in §5, is that Austin's model of 'fact' does not extend easily to sentences more complicated than "The cat is on the mat", e.g., to negations and quantifications. How might Austin respond to that objection? Does he really need 'negative' facts? or else a different sense of "true" for negations?
• In §3, Strawson addresses Austin's claim that the correspondence relation is "purely conventional". He vehemently denies that claim. The crucial basis for his denial is a distinction drawn on p.143. What is this distinction? How exactly is it to be applied to the case of assertions of the form "What A said is true"?
• In §4, Strawson discusses Austin's view that, when says e.g. "What A said is true", one is making a claim about a statement (i.e., proposition). In particular, he argues that, in at least some cases, saying "What A said is true" amounts simply to re-asserting whatever A said. No doubt Strawson is correct that, in some such cases, we are re-asserting what A said. But does that imply that we are not also making an assertion about a statement?

• Ayer begins by discussing the view that "true" is "superfluous" or, as it is more often put, 'redundant'. Ayer first argues that this view works only in the case in which "true" is applied to a proposition that is "itself mentioned" and not in the case in which a proposition is "only mentioned". A simpler example of the latter would be something like: Fermat's Theorem is true. But the case most often discussed, as we'll see later, is that of generalizations involving "true" such as "Everything Bill says is true". Why can we not proceed as Ramsey does and analyze this as: ∀p(Bill says p → p)? Is this rightly described as a confusion of use and mention?
• Ayer nonetheless agrees with Ramsey that the meaning of the word "true" can be fully explained in terms of the so-called T-scheme:

  It is true that S iff S

  (Not that this is a "scheme", a pattern, to be instantiated by subsituting various sentences for "S"; in particular, "S" is not a variable, so the objection to Ramsey does not apply.) Ayer argues, however, that we need to restrict this scheme---that is, restrict what sentences can be substituted for "S"---on pain of contradiction. What is his argument? What is his proposed solution to this problem?
• Ayer next suggests that, for practical purposes at least, we might better focus on truth as applied to sentences. And he then argues (largely following Tarski, whom we shall read later) that it will be enough to define "true" if we can produce a defintion from which all instances of the 'sentential' T-scheme:

  "S" is true iff S

  can be derived. As Ayer mentions, and as we shall see in detail later, this can be done for certain sorts of formal languages (and natural language semantics is, in many ways, an attempt to extend that treatment to natural languages). But, Ayer argues, the T-scheme cannot itself be regarded as a definition of "true". Why not?
• The next question Ayer raises is whether T-sentences such as:

  "The sky is blue" is true iff the sky is blue.

  are necessary or contingent. Why does this issue matter to Ayer? How does he propose to resovle it? Why does this problem not arise in the case of "It is true that p"?
• On the basis of the arguments given so far, Ayer claims: "To speak of a sentence, or a statement, as true istantamount to asserting it, and to speak of it as false is tantamount to denying it". True or false?
• Ayer has, in effect, claimed that the T-scheme completely characterizes the meaning of the word "true". Nonetheless, he insists that this does not really solve "the philosophical problem of truth". Why?
• What Ayer claims is wanted is "a criterion of validity", by which he means a general account of what it is for a proposition (any proposition) to be true. What obstacle does Ayer think stands in the way of such an account? Why does he think that objection is not wholly fatal to the enterprise?
• Ayer's own positive views about the answer to this question, which are elaborated in the remainder of the paper, are based upon his 'logical positivism' and would not now be widely accepted. But there is a striking affinity between his criticisms of the correspondence theory and Strawson's. How so?

This paper is legendarily hard, so I am going to provide a lot of guidance for your reading of it. Read the paper slowly, and take a lot of deep breaths.

• Dummett begins the paper by expounding Frege's claim that sentences refer to their truth-values. It is easiest to understand this claim when it is put differently: that the "semantic value" of a sentence is its truth-value. And that claim is best understood in terms of the truth-tables: that the central semantic fact about a sentence is that it is true or that it is false. The point of this is really just to introduce the idea of thinking of truth from the perspective of logic.
• Dummett then suggests that, while it's reasonable to think that sentences do have "semantic values", Frege has to earn the right to say that their semantic value is their truth-value. On pp. 142-4, Dummett introduces an analogy between truth and falsity, and winning and losing, to illustrate what Frege would have to do to earn that right. What exactly does Dummett think Frege would have to do?
• Dummett then proceeds to argue, on pp. 145-6, that (the propositional version of) the T-scheme may not even be correct. The argument turns on the idea that there may be sentences that are perfectly meaningful—they express propositions—but are neither true nor false. A putative example would be something like, "The greatest prime number is one less than a perfect square". Frege would have held that this expresses a proposition, but does not have a truth-value, due to the fact that there is no greatest prime. Why, then, does Dummett think that:

  It is true that the greatest prime number is one less than a perfect square iff the greatest prime number is one less than a perfect square.

  is not itself true?
• Dummett then argues, on pp. 146-9, that, even if its instances are all true, the T-scheme "cannot give the whole meaning of the word 'true'". The argument turns on the assumption that the truth-tables have some explanatory value, in particular, that they embody (at least partial) explanations of the sentential connectives. Another aspect of the argument is that the 'redundancy theory' is incompatible with a truth-conditional conception of meaning. Can you fill in some details?
• On p. 149, Dummett then concludes that a theory of truth must be possible in a certain sense. In particular, he thinks that it must be possible for us to articulate the point of our characterization of assertions as true and as false. There is a sketch of what Dummett has in mind in the paragraph running from p. 149 to p. 150. Try to articulate as best you can what research program he means to be articulating. How might this relate to the what Strawson called the problem of "elucidating the fact-stating type of discourse"?
• On pp. 150-4, Dummett then argues in support of a very general claim that he makes on p. 150: that, given the point of the characterization of assertions as true and as false, there is no need, and no room, for any finer characterization, and so that it is senseless to say that an assertion is neither true nor false. The core of the argument is on p. 153, where Dummett suggests that, although we might call both conditionals with false antecedents and sentences containing non-referring terms "neither true nor false", there is an important asymmetry between the two cases for which this common terminology fails to account. What is that asymmetry?
• Finally, on p. 154, Dummett concludes that "we should abandon the notions of truth and falsity", at least in connection with the explanation of the meanings of statements. In fact, however, that isn't quite what he means. He thinks there is a particular way of using "true" and "false" that is unhelpful and another way of using them that would still be OK. What is the difference?
• Dummett proceeds, on pp. 154-7, to explore whether there might yet be a point in calling certain statements neither true nor false. He argues that there may well be, but that, if there is, it must necessarily concern the way such statements behave when they occur as parts of other statements (e.g., as antecedents of conditionals). How does that relate to Dummett's earlier thesis about the role of the concept of truth in logic?

As mentioned, we probably will not even manage to discuss that much, and we certainly will not manage to discuss more. So you are welcome to stop reading at this point.

• Finally, on pp. 157-62, Dummett introduces a set of considerations that are supposed to show that the notions of truth and falsity that are appropriate to the evalaution of assertions are not the classical notions of truth and falsity. Rather, calling an assertion "true" is like saying it is justified, and calling it "false" is like saying that it is unjustified. This sort of argument is one that became closely associated with Dummett, and he spent much of his career trying to develop it and to fill in the details.

Good luck!

• Field's paper concerns the philosophical significance of Tarski's work on truth. His central claim is that Tarski showed us how to "reduc[e] the notion of truth to certain other semantic notions", but that he did not show that the notion of truth was unproblematic, as Popper and others held.
• Field's paper is organized around the contrast between two theories of truth: T1 and T2.
  • What is the fundamental difference between T1 and T2?
  • What does Field take T1 to accomplish? Why does Field regard that accomplishment as important?
  • Field enumerates a number of advantages he takes T1 to have, when used as a semantic theory for a natural language. In particular, he thinks it helpful cleanly to separate the explanation the theory gives of how the semantic properties of complex expressions depend upon those of their parts. But that is more an issue in philosophy of language, so one that is somewhat outside the scope of our interests here, so do not worry too much about pp. 351-4.
  • What does Field think T2 purports to add to T1? What value does he think other philosophers have thought this addition has?
  • Despite the striking differences between T1 and T2, there are a handful of axioms of the two theories that are the same, namely, the axioms for negation, conjunction, and the universal quantifier. Field does not say anything about why that is, but it is not at all obvious why it is. What would more T1-like clauses for these expressions be like? What would a theory that was formuated in those terms accomplish?
• In §III of the paper, Field discusses one motivation for dissatisfaction with T1: physicalism. How and why does physicalism provide a reason to be dissatisfied with T1? What does Field think answering the physicalist challenge to semantics requires us to do? (To some extent, the answer to that question only emerges in §IV, around pp. 366-7.)
• Field argues in §IV that what T2 adds to T1 is completely trivial and so that T2 is no advance whatsoever over T1. Indeed, Field argues, on p. 370, that we can say, quite precisely and rigorously, exactly what "T2 minus T1" is, that is, what T2 adds to T1. How does that argument go exactly? What does T2 add to T1? Why is that supposed to be "trivial"?
  For what it's worth, I think Field somewhat misformulates this argument. What he calls D2 is itself implied by Tarski's theory T2 given only the claim that there are no other primitive names in the language other than those listed in D2. But that presumably falls out of our syntax for D2. (If that is right, then T2 by itself entails T1, and Field's remark that "T1 is simply a weaker version of Tarski's semantic theory" seems right. Moreover, his claim that T1 ∧ D2 ∧ P2 ∧ A2 is equivalent to T2 then seems correct, as well.)
• Field writes that, for Tarski, "...the notion of an adequate translation is employed in the methodology of giving truth theories, but it is not employed in the truth theories themselves" (p. 355). What does Field mean by this? Might we be able to construe the role that Field thinks (e.g.) a causal theory of reference should play here in the same way?
• In §V, Field discusses the question what significance the notion of truth has for philosophy and why we should not simply abandon it in the face of the physicalist challenge. We shall take up this kind of issue later, when we discuss deflationism. For now, however, we'll set it aside.
• Looking forward to our next reading, here's a question just to think about. Field admits that T1 offends against Tarski's insistence that he will not make use of primitive semantic notions. Why did Tarski impose that constraint? What would he have thought was the cost of flaunting it?

• NOTE: When Parsons refers to formulae like "(2.5)", he means formula (5) in §II.
• Parsons sets out in this paper to defend Tarski's "hierarchical" treatment of the liar paradox against a series of objections that he discusses in §I. (Parsons's is the most sophisticated and developed form of this view up to that time. Later writers developed related views. See the optional readings mentioned above.)
• Parsons does not present van Fraassen's view in enough detail for it to be intelligible to someone who has not read that paper, so do not worry too much about getting it all right. What's important about that view, for our purposes, is (i) that it attempts to do without the distinction between object-language and meta-language and (ii) that it attempts to do so by allowing that the liar sentence might be neither true nor false. This leads Parsons to object that van Fraassen escapes paradox only at the cost of certain 'expressive limitations', e.g., the absence of a 'strong negation' that would convert any non-true sentence into a true one.
• This is what has come to be called the 'revenge problem': When one tries to formulate a consistent theory in this way, it always seems as if there are things one would like to be able to say about the liar sentence that one cannot, in that theory, say. E.g., we cannot, in van Fraassen's theory, truly say that the liar sentence is not true, even though it isn't. This leads van Fraassen himself to acknowledge that no language can ever be 'expressively complete'. But that, Parsons says, was Tarski's point: It is what leads to the hierarchy. If that is right, then our problem is to understand how it could be that natural language is always expressively incomplete.
• Parsons proposes in §II to make explicit, in a way Tarski does not, the role played in the liar reasoning by the implicit assumption that e.g. the liar sentence expresses a proposition. (It's a natural, indeed common, response to the liar to insist that it is meaningless: Recall Ayer.) The T-scheme can then be rephrased in terms of a notion of propositional truth, as at (5): (In effect, this restricts the sentential version of the T-scheme to sentences that express propositions. A sentence is true, on such an account, if it expresses a true proposition.)
• Parsons then rehearses versions of the liar paradox framed in terms of propositions. Simplifying, we may take the relevant sentences to be:
  (λ) λ expresses a false proposition.
  (κ) κ does not express a true proposition.
  It can then be proven that neither λ nor κ expresses a proposition at all. But this leads to an immediate puzzle: If κ does not express a proposition at all, then presumaby it does not express a true one. But that is what κ says. So it seems that κ actually is true after all, i.e., that it expresses a true proposition. But that will now lead to paradox. This is the so-called strengthened liar. It poses a threat to any view according to which the liar does not express a proposition. (It would be worth writing out the argument, in detail, that κ does not express a proposition. Note particularly where (5) enters.)
• Parsons proposes to avoid this by being more careful about the domain of the quantifiers ranging over propositions, e.g., in κ, which more strictly reads: There is no true proposition that κ expresses. Parsons's idea is that there can be no single all-encompassing domain of propositions: Given any such domain D, (5) implies that there will always be other propositions that are not in that domain, e.g., the one expressed by κ, when the propositional quantifiers range over D. As a result, the sentential truth-predicate will never apply to all sentences in the language, but only to some of them. (Parsons takes this to be analogous to a similar response one might make to the set-theoretic paradoxes.)
• (In §III, which you need not read, Parsons re-formulates all of this in terms of a sentential notion of truth. But, as he remarks later, he thinks that the intensional aspects of the paradox are crucial, so the sentential version is technically useful, perhaps, but not philosophically fundamental.)
• In §V, Parsons addresses the question what any of this really has to do with natural language. He insists that what's really causing the problem is not the notion of truth but the notion of expression, regarded as a relation between sentences and propositions. Truth, Parsons wants to claim, can once and for all be characterized in terms of the propositional version of the T-scheme. It is the relation of expression that varies between object-language and meta-language: Sentences that do not express propositions from the point of view of the object-language do come to express them from the point of view of the meta-language. Parsons suggests that we develop this idea by attributing a certain kind of context-sensitivity either to the quantifiers ranging over propositions or to the words that express (!) the expression relation. (The notion of sentential truth would then also become context-sensitive, since it is defined in terms of these notions.) The shift of contexts is supposed to be requried by "reflection" on what someone has said: that is, by deployment of notions of truth and expression that would apply to what they have said; but that, Parsons wants to say, just corresponds to the move from object-language to meta-language, as Tarski taught.

• What does Kripke mean by saying on p. 692 that many of our ordinary attributions of truth are "risky"? How does this serve to address (say) Austin's insistence that the liar sentence is somehow defective?
• How is the notion of "groundedness" supposed to contribute to Kripke's analysis of the Liar paradox? How is this notion similar to and different from the one sketched by Ayer? (Hint: How would Kripke and Ayer think about the sentence "Either this sentence is false or snow is white"?)
• Kripke discusses an example early in the paper about Nixon and Dean. What is the central point of that example? Kripke later argues that this example poses a problem for Tarski's hierarchical account. Why so? How does this lead Kripke to search for a way of allowing a language 'to contain its own truth-predicate'? What does he mean when he says that sentences should be allowed to seek their own level?
• Be sure you understand both why and how Kripke wants to allow for 'truth-value gaps', that is, sentences that are neither true nor false. This is explained on pp. 699-700 and involves using three-valued logic, extensions and anti-extensions for predicates, etc. (Why can't we just use classical two-valued semantics here?)
• On p. 701, Kripke describes certain intuitions about truth are that serve to motivate his approach. How are these related to Convention T? How do they motivate the notion of groundedness?

It would be worth re-reading §V of Field's paper "Tarski's Theory of Truth" before reading this one, or at least reviewing your own notes about that paper. In many ways, the view Field expresses here is a result of his souring on the prospects for reducing 'primitive denotation' to anything physical. Since that, as Field continues to see it, is what is required for a proper reduction of the notion of truth, we must abandon that hope, as well, and make do entirely with T1 or something like it: a 'list-like' theory of truth and similarly 'list-like' theories of denotation and satisfaction.

You will see in this paper constructions like "Σp(...p...)" and "Πp(...p...)". Here, Σ and Π are 'substitutional quantifiers'. Roughly, "Σp(...p...)" means: Some instance of "...p..." is true, where the instances are formed by replacing 'p' by sentences. Field would not want us to read it that way but to take substitutional quantifiers as primitive; but that should allow you to understand his uses of this notation.

• Field suggests that the fundamental division among theorists of meaning and content is whether truth-conditions can or should play a central role in such theories. In effect, what Field is doing in section 1 is proposing is that we recast that question as one about how we should understand the truth-predicate. Why is that supposed to be a good idea? How does the recasting go?
• What is a "purely disquotational" truth-predicate? How does Field propose to explain it? He claims, almost in passing, that it is obvious that everyone, including a disquotationalist is entitled to such a predicate. Is it obvious?
• What does Field mean when he says that a purely disquotational truth-predicate can only be applied to sentences that we understand? Why is that a non-optional feature (or bug) of purely disquotational truth-predicates?
• In section 2, Field develops a story about how a disquotationalist might understand meaning and content: in terms of 'conceptual role' and what he calls 'indication relations'. The latter are simply correlations between a given thinker's beliefs and states of the world. Field emphasizes that these need not correspond to truth-conditions in any way and so do not obviously yield a reduction of what he once called 'primitive denotation'. But most of this discussion is by way of trying to indicate a bit more clearly how deflationism differs from 'inflationism'.
• Section 3: How is the inflationary story about the meanings of logical operators supposed to work? Why does Field take it to be problematic? What alternative does he propose? Would an argument similar to the one Field gives also suffice to show that "P because Q" is true iff "P" is true because "Q" is true? If so, why might that be a problem?
• Toward the end of section 4, Field writes:

  All I really hope to motivate here is that we should be "methodological deflationists": that is, we should start out assuming deflationism as a working hypothesis; we should adhere to it unless and until we find ourselves reconstructing what amounts to the inflationist's relation "S has the truth conditions p".

  Are we required to play by these rules?
• Section 5: Field argues that a purely disquotational truth-predicate is essential to the formulation of certain sorts of claims about the physical world. Which? Why? (And how is this related to earlier authors we've read?) Field argues that, in order for "true" to play this role, it needs to be 'purely disquotational' and, in particular, to be "use-independent" (p. 266). What does that mean? Why is it supposed to be essential?
• The issue here is dialectically important. In effect, what Field is arguing is that, when "true" is used in the sorts of generalizations that make "true" seem indispensible, it must be used purely disquotationally. If so, then even inflationists seem to need a purely disquotational truth-predicate and so are forced to argue that we also need one that is not purely disquotational. I.e., it looks as if inflationists are committed either to thinking that "true" is ambiguous between these or else that the inflationist notion of truth is a 'theoretical' one that is not expressed by the ordinary word. Both of these views look uncomfortable. Is the inflationist really forced into such a position? See the second to last paragraph of section 6 for one possible line of response.
• Section 6: What does Field take to be the fundamental contrast between a disquotational truth-predicate and a Tarskian one? Why must a deflationist forego any serious interest in compositionality? (Special question for those with some prior exposure to these issues: Can Field consistently affirm that we understand all the sentences over which we are quantifying when we say that all the axioms of PA are true?)
• In section 8, Field responds to the objection that we seemingly can apply the ordinary English truth-predicate to sentences we do not understand. What is the response? What are the limitiations of Field's strategy? Note that the sentences in question need not be non-English sentences but might just contain words (e.g., technical language) with which one is not familiar. (The case of applying "true" to sentences of other languages that I do understand is far less important.)
• In section 9, Field discusses a modal worry we have encountered previously. It seems as if the use-independence of the purely disquotational truth-predicate implies that, if "true" is purely disquotational, then

  (*) Even if 'snow is white' had meant that pigs fly, it would still have been true.

  is true. Why? How does Field respond (on pp. 277-8) to the objection that (*) is false? How good is that response? What does that response say about the relation between his theory and 'ordinary' uses of "true"?
• The purely disquotational truth-predicate is originally explained for sentences. In section 10, Field considers the question how it might be extended to utterances, so as to account for indexicality. (Field also discusses the case of ambiguity, but this is again less central.) How does he propose to account for the application of "true" to, say, "She is a philosopher"? The crucial issue arises at the bottom of p. 280, when Field is forced to concede that, in some sense, there must be "a correct answer to the question of who another person was referring to with a particular application of 'she'...". How does Field propose to handle that fact? How good is his account?

• Gupta's overall strategy is to distinguish what he takes to be the core claims of deflationary theories of truth from the larger philosophical consequences that those claims are alleged to have. In particular, Gupta argues that these consequences follow only on an implausibly strong reading of the core claims, the most important of which are the Disquotation Thesis and the Infinite Conjunction Thesis. What are the weaker and stronger readings of the Disquotation Thesis and the Infinite Conjunction Thesis?
• Section III of Gupta's paper is concerned with what is sometimes known as the "success argument" against deflationism, which claims that the notion of truth plays an important role in explaining behavioral success, e.g.: People with true beliefs are more likely to get what they want.
  • How does Horwich (as Gupta reads him) respond to this argument?
  • Why does Gupta think that Horwich's response depends upon the strong reading of the Infinite Conjunctions Thesis?
  • Why does Gupta think, as he says on p. 67, that this thesis is "plainly false" so read?
  Gupta's central point here is that there is a difference between explaining all instances of a generalization and explaining the generalization itself. Can you come up with your own example to illustrate this difference?
• Section IV is concerned with the deflationist thesis that truth can play no role in semantics, in particular, that the meaning of a sentence cannot be taken to be its truth-condition.
  • Why does Gupta think this argument depends upon the strong reading of the Disquotation Thesis?
  • What is Gupta's argument (on pp. 70-2) against the strong reading of the Disquotation Thesis? To what extent do these considerations rest upon the issue Gupta has deferred to section V?
  • One point worth careful analysis: Gupta suggests that the claim that someone who asserts "'Snow is white' is true" typically asserts 'no more and no less' than that snow is white need not be explained in terms of the synonymy (or something like it) of the two sentences. What is his alternative explanation?
  (My own view, by the way, is that Dummett did not mean to endorse the deflationary argument. He certainly did not do so in his later work. Rather, he took the argument to be a reductio of deflationism.)
• In Section V, Gupta considers the possibility of replacing the Disquotation Thesis with some other principle that might imply it: one that conceives of there as being some simple rule, or general principle, that governs the use of "true" and that thereby explains its meaning. The most interesting of these is the inferential account that he considers on pp. 74-5. What are his objections to this account? A different worry, expressed on p. 75, is that the inferential account does not yield any claim strong enough to do the work the Disquotation Thesis was supposed to do. How so?
• At the end of the paper, Gupta suggests that "true" plays an important role in our thinking precisely because the Disqutation Thesis and Infinite Conjunction Thesis are false. What are his reasons for this claim?
• Central to many of Gupta's objections is a worry about the 'ideology' that is required by the deflationary account of the meaning of "true". What is this worry? How serious is it?

BH&L, as we shall call them, argue here that the meaning of a sentence "must at least in part be a truth-condition". From that, it will follow either that, unless deflationism can be shown to be compatible with truth-conditional semantics (which most believe it is not), then deflationism must be false. Note that what's really at issue here is deflationism about sentential truth, i.e., what Field calls 'disquotationalism'.

• In section I, BH&L lay out what they call the Determination Argument for the claim that meaning must at least determine truth-conditions (though it may include more). Make sure you understand that argument. The rest of the paper considers a series of attempts a deflationist might make to resist the argument.
• As BH&L argue in section II, it is difficult to see how to resist the main premise of the Determination Argument. What it says, very roughly, is no more than this: Once the meaning of a sentence has been fixed, then whether it is true or false in given circumstances has also been fixed. So denying the premise seems to mean accepting that there could be two sentences, A and B, with the same meaning, but A would truly describe certain circumstances whereas B would not. That does seem incoherent.
• The most interesting (and plausible) objection that BH&L consider this section II is the third. The objection doubles down on disquotationalism: The question whether "snow is white" is true in certain circumstances simply is the question whether snow is white in those circumstances; so circumstances 'plus meaning' determine truth-value, but meaning here does no work; it's the circumstances alone that determine truth-value. BH&L's response seems, in many ways, a version of the modal problem we have previouly considered: They insist that meaning does play a role, since "Snow is white" would not be true in a world in which snow was the same color it is now but "white" meant black. Should we conclude that the problem here is therefore one disquotationalism had anyway?
• BH&L press the point by arguing that, if instances of the T-scheme are to be guaranteed to be correct, then the sentence that is used on the right-hand side must somehow be guaranteed to have the same meaning as the sentence that is mentioned on the left-hand side. (One way to see this is to consider ambiguous sentences, like "Alex went to the bank".) As they discuss, there are ways to try to secure this condition without mentioning meaning, but these restrictions, they argue, artificially restrict the scope of the things to which "true" can be applied. Here again, however, Field seems to have been aware of this point: the pure disquotational notion of truth applies only to 'sentences I understand'. Do BH&L succeed in making that aspect of Field's view less comfortable?
• (By the way, it seems to me that BH&L's arguments would benefit from greater clarity about the role being played by modal considerations. At times, they seem to be discussing what is sometimes called truth with respect to a given world; in that sense, what a sentence means in that world is irrelevant (and there may not even be any speakers in that world). At other times, they are clearly discussing a different notion we might call truth in a world, and in that case what the sentence means in that world is crucial. This is closely related to what David Chalmers calls 'primary' and 'secondary' intensions, where secondary intensions correspond to truth with respect to worlds and primary intensions to truth in worlds. Chalmers argues at some length that truth-conditions should be understood as primary intensions.)
• In section III, BH&L consider whether a deflationist can simply accept the Determination Argument by insisting that the conclusion it establishes is too weak to underwrite a properly truth-conditional account of meaning. The argument here is extremely complex—which is part of why you may skim or even skip it—but I take their main point to be that any view on which truth-conditions are not straightforwardly part of content owes an explanation of how content, so understood, determines truth-conditions. It is, for example, often considered an objection to (narrow) conceptual role theories that conceptual role does not determine truth-conditions. I take it that Field means to avoid this kind of objection, in part, by insisting, once again, that truth applies only to 'sentences I understand': If a sentence means something to me, then I can state its truth-condition simply by uttering that sentence. But BH&L have already offered reasons to reject such a restricted notion of truth—or, at least, not to rest with it.
• In section IV, BH&L consider a variant of the Determination Argument that is stated in epistemic terms. There are some complications here. To what extent do they undermine the appeal of this version of the argument? Does it seem as strong as the earlier one?

• B&S, as we'll call them, distinguish three sorts of deflationism: "metaphysical" deflationism, "linguistic" deflationism, and "conceptual" deflationism. Make sure you understand what these various positions are.
• Section I of the paper elaborates the importance of "conceptual" deflationism. In effect, what B&S argues is that 'real' deflationism is conceptual deflationism. And they illustrate that the commitments of conceptual deflationism are by considering how a deflationist must regard such a claim as "True beliefs engender successful action": The deflationist must show that the role of the notion of truth is here just 'expressive' and 'logical'. The crucial point is that such a claim cannot be regarded as telling us anything about truth, though it may tell us something about belief and action. The claim is not really 'about' truth, in any significant way.
• There are stronger and weaker ways that B&S describe their goal here. At one point, they describe themselves as arguing that "understanding assertion requires more than the 'thin' concept of truth afforded by deflationary accounts" (p. 68), which would imply that conceptual deflationism is false. Shortly thereafter, however, they say that what they "hope to show" is just "that metaphysical and linguistic deflationism, taken either separately or together, do not entail conceptual deflationism" (p. 69). Which is it?
• Section II discusses Frege's view that asserting a thought (proposition) is presenting it as true (or that judging a thought is acknowledging it to be true). They then raise the question how a deflationist should attempt to 'deflate' that formula: that asserting that p is presenting p as true. They first argue (pp. 71-2) that the strategy of 'denominalizing' will not work here. What is the argument? How good is it?
• B&S consider a second sort of reply a deflationist might make: that we should regard this connection between assertion and truth as part of the explanation of truth itself. Recall here Ayer's remark that "To speak of a sentence, or a statement, as true is tantamount to asserting it". This leads to what B&S call "illocutionary deflationism". (An explicit statement of this sort of view is found in Strawson's first paper "Truth", Analysis 9 (1949), pp. 83-97, JSTOR.) What objections to B&S bring against illocutionary deflationism? (For those who know what it is: How is this related to what Peter Geach famously called 'the Frege point'?)
• Recall that B&S describe themselves as arguing that linguistic deflationism does not imply conceptual deflationism. But one might object that what they really show is just that Frege was a linguistic deflationist but not a conceptual deflationist. Surely no-one would deny that it is possible to hold such a combination of views. The question is whether it possible to do so consistently; maybe Frege's views were just inconsistent here. (Worse, one might think Frege's view that asserting is presenting as true is so unclear that one cannot even evaluate it.) The discussion on pp. 76-8 might be thought of as an attempt to respond to such worries. The central point is to distinguish (on pp. 75-6) between questions about "true" and questions about truth, or between questions about first-order and second-order uses of "true". What are these distinctions? How and why are they supposed to help us see that linguistic deflationism does not imply conceptual deflationism?
• Bar-On and Simmons several times mention an analogy between certain forms of ethical non-cognitivism and certain ways of thinking about the nature of truth. What is the point of this analogy?