The workshop is hosted by the ERC-Starting Grant ‘Truth and Semantics’ (TRUST 803684) at the University of Bristol directed by Johannes Stern. It’s aim is to explore the semantics and pragmatics of the word ‘true’ in natural language. Central topics of discussion will include whether or not ‘true’ is ambiguous, context sensitive, vague, gradable, presuppositional, or lacking content altogether. Other relevant topics will include what sort of property or concept ‘true’ expresses, connections between natural language ‘true’ and the semantic paradoxes, and syntactic properties of ‘true’.
Registration is now open until August 25th. To register, please send an email to Poppy Mankowitz (poppy.mankowitz@bristol.ac.uk) with the following information:
9.00: Conference opening and coffee.
9.30 - 10.45: Dorit Bar-On (University of Connecticut): Truth in Meaning.
10.45 - 11.15: Coffee.
11.15 - 12.30: Bryan Pickel (University of Glasgow): A Redundancy Theory of Truth.
12.30 - 13.30: Lunch.
13.30 - 14.45: Paul Égré (Institut Jean Nicod, Paris): Truth is flat and bumpy – On the gradability of “true”.
14.45 - 15.00: Break.
15.00 - 16.15: Caspar Wilson (King’s College London): Austin’s Alethic Pluralism and a Pragmatic Function of ‘True’.
16.15 - 16.45: Coffee.
16.45 - 18.00: Max Kölbel (University of Vienna): A Spoilsport’s Account of Truth.
With the so-called linguistic turn in philosophy, it has become standard to try to “save the differences” amongst different types of discourses (e.g. ethical, mathematical, epistemological) by positing semantic differences – differences in kinds of meaning, or else in kinds of truth. Such semantic strategies face a dual challenge. If the semantic differences are too deep, it becomes difficult to account for palpable logico-semantic continuities across discourses, as well as the meaning/truth of ‘mixed’ sentences or arguments. If one addresses this challenge by positing a new type of uniform meaning or truth that all sentences possess, then one must face anew the challenge of saving the differences. Using first-person discourse as a case study, I argue that we should resist trying to capture differences across discourses by resorting to pluralism about meaning or truth. We should capture logico-semantic continuities by assigning indicative sentences from all discourses truth-conditions as their meanings. I offer a particular construal of the appropriate notion of truth-conditions and explain how it leaves room for saving the differences not semantically but rather in other (e.g. metaphysical, epistemological) terms.
The classical redundancy theory of truth says that the predicate ’is true’ is descriptively empty. Sentences of the form ’that Annabel is happy is true’ express the same propositions as sentences of the form ’Annabel is happy’. The view is often dismissed as moribund, and is treated as a step on the way to more sophisticated forms of deflationism. In particular, it is thought that the view does not easily extend to generalizations such as ’something you said is true’ or to truth ascriptions involving singular terms such as ’Noether’s first theorem is true’. Sometimes it is suggested that the redundancy theorist must go so far as to reject the view that ‘is true’ is a predicate at all. I argue that these reservations are misplaced. I suggest that we construe the redundancy theory as the thesis that ‘is true’ is semantically vacuous. By following typical approaches to semantic vacuity from formal semantics, we can naturally extend the redundancy theory to generalizations and to truth ascriptions involving singular terms. I show that even the most uncompromising version of the redundancy theory can meet the challenges traditionally posed to it.
While the adjective “true” in English admits modification by “very” and by the comparative and superlative morphemes (“truer”, “most true”), the view of truth as a gradable property is generally met with scepticism (see Frege 1919, and Haack 1980, criticizing fuzzy views by Zadeh 1975). Recently, some novel observations were put forth in Henderson (2021) and in Egré (2021) to reassess the gradability of “true”. In this paper I propose to highlight some of those observations and to discuss further issues. Haack (1980) argued that “true” is neither a relative gradable adjective, nor an absolute gradable adjective. In the first part of the presentation I examine Haack’s arguments. First I will review evidence in favor of “true” (in English) and “vrai” (in French) being absolute adjectives, but moreover I will show that “true” has the features of both partial and total absolute adjectives (in the sense of Rotstein and Winter 2004, viz. “flat” vs “bumpy”). These observations, however, are not enough to conclude that truth is fundamentally gradable. D. Feinmann (private communication) observes that various languages differ from English and French, for instance Spanish, which distinguishes “verdad” (non-gradable) and “cierto” (gradable). This suggests that “true” might be gradable only as a coerced adjective in English and in French, like “pregnant” (see Burnett 2017). Depending on the case, “truer” might mean “more clearly true” and express an epistemic property. I will argue, however, that this interpretation does not cover all cases of “truer”. For generic sentences, or for singular sentences containing gradable predicates, “truer” can mean “more often true” or “closer to fully true”. Both of those uses, I will argue, are not fundamentally epistemic, they tell us that truth comes in different extents.
In this paper, I compare J.L. Austin sustained discussion of the “highly exceptional” word ‘real’ in Sense and Sensibilia with various things he says about ‘truth’: “that ‘true’ is just such another extraordinary word is obvious.” Although there is no reason to think extraordinary words must all be so for similar reasons, from this comparison emerges a reading of Austin as a peculiar sort of alethic pluralist, one where, in calling something true, one can say any of indefinitely many things depending on the appropriate standards of the occasion. Whether or not Austin’s brand of alethic pluralism has the virtues that motivates other versions, it avoids Tappolet’s mixed inference challenge, and hence plausibly lacks one of their vices. Moreover, this reading identifies an additional function of ‘true’. Miscommunication can occur when speakers disagree over the appropriate standards an occasion calls for. A use of ‘true’ can either clarify whether that is the sort of fit between words and the world at play, or to attempt to impose more stringent standards.
In this paper, I distinguish a number of different questions to which theories of truth can be an answer. I shall distinguish various empirical and a priori questions about meanings and concepts as well as normative questions about which meanings and concepts are worth having. I shall then develop and justify an overall conception of truth that attempts to give a coherent combination of answers to many of the questions previously distinguished. Thus, I will make proposals about to the meaning of “true” in ordinary English, about the concepts typically expressed by it and employed in our thought, about concepts of truth employed in semantics, and about concepts of truth worth having. I will also make some remarks about what one might say about the property or properties of truth. I call it a “spoilsport’s account” because by distinguishing carefully between the different questions surrounding truth, I am treating some apparent disagreements in debates surrounding truth as merely apparent.
I argue for two claims about the predicate ‘true’: first, that ‘true’ is a context-sensitive predicate, and second, that ‘true’ applies almost everywhere except for certain singular points where its application breaks down. The claim about context-sensitivity is supported by an examination of our reasoning with liar sentences, sentences which say of themselves that they are not true. The ‘singularity’ claim accommodates the expressive power of natural language, and is inspired by a remark of Gödel’s, that the paradoxes are analogous to dividing by zero. These two claims are extended to the semantic terms ‘denotes’, ‘extension’, and ‘valid’, each of which, like ‘true’, is associated with paradox. My overall aim is an account of semantic paradox that endorses the substantiveness of our semantic notions, and respects Tarski’s intuition that natural languages are universal.
The adjective “true” can felicitously modify some social kinds but not others. For example, “true artist” and “true scientist” seems to pick out particularly devoted artists and scientists, but experimental evidence has shown people generally judge “true optician” or “true cashier” to be infelicitous. It has been hypothesized that categories like artist and scientist differ from categories like cashier or optician in that while both sets of categories have descriptive content, so-called Dual Character Concepts like artist and cashier have independent normative dimensions associated with idealized values. In other words, you can be an artist by merely having certain commitments to art or beauty, but you cannot be a cashier by merely having certain commitments to the exchange of money or customer service. Despite the importance of the true-modifier in philosophical and psychological discussion of Dual Character concepts, the mechanism by which “true” modifies Dual Character Concepts has not been explored. Through a series of experiments, we first verify that “true” does indeed modulate certain social kind terms to normative, or value-laden, readings. We then demonstrate that while people’s default reading of Dual Character Concepts like “scientist” or “artist” is specifically descriptive and not normative, it is very easy to get people to read Dual Character Concept terms normatively. We hypothesize that “true” as well as the other mechanisms work by raising the salience of certain norms about the values associated with social kinds.
Philosophers that consider truth to be normative have generally taken truth to be the norm guiding acts of adopting a belief or making an assertion. The point of departure of this talk is to pay close attention to the way normative notion relate to truth in natural language. One observation is that the predicate ‘correct’ conveys truth (and just truth) not when it applies to actions (of adopting a belief or making an assertion), but when it applies to entities like beliefs and assertions, attitudinal objects as I call them. This indicates that the normativity of truth has to do with an inherent norm of certain types of content bearers, rather than being action-guiding. I will also discuss whether the correctness (truth) of beliefs and assertions could be on a par with conditions of correctness or appropriateness of emotions, as discussed in recent philosophical work on emotions.
For all further questions about the conference please contact Poppy Mankowitz.