THE SEMINARS
INFORM INSPIRE INTEGRATE
The PhiLang Seminars in Philosophy of Language and Argumentation are a series of events for philosophers and linguists. Each seminar features a talk by a distinguished figure in research and a discussion session where participants can raise questions and expand the range of the debate. The events take place online via the Teams platform and are open to everyone, upon registration.
Our goals are to:
Inform the community about cutting-edge research
Inspire new research and new researchers
Integrate people, ideas, and approaches for a better understanding of our work and each other.
The speakers and their topics both draw on the tradition of the PhiLang Conference series and help set the agenda for the 9th edition of the event, to be held in Lodz in May, 2025.
Join the debate by registering here. You need only register once for the entire series.
LINKS
PREVIOUS SPEAKERS
27/11/24 Natalia Karczewska
University of Warsaw
Disagreement and the Common Ground
Common ground is a concept used in explaining conversational dynamics in its many manifestations. Stalnaker (2002) characterizes it as "presumed background information shared by participants in a conversation". There are a few proposals on how to understand this background information. One of them views common ground as a set of possible worlds (Stalnaker 1970). The effect of asserting p on the common ground would be restricting it to the effect that the possible worlds where p isn’t true are eliminated. Another way of thinking about the common ground is to say that it is made up of psychological states, such as beliefs or knowledge states (Bach and Harnish 1979). For a belief B to be common ground, it has to be the case that not only all the participants hold B, but also that all of them believe that all the others hold B (and so on). The third alternative is to think of common ground as a normative condition, e.g. as comprising reasons to believe (rather than beliefs or their contents) (Lewis 1969) or commitments – relations between speakers and propositions.
Many kinds of arguments are raised in the discussion on which of these accounts of common ground should be preferred, e.g. the problem of infinite recursion of mental states, the power to explain speech acts different from assertion, etc. In my talk I will focus on the question of which of these accounts gives the best explanation of disagreement data with a particular focus on evaluative disagreement. I argue that along with a contextual semantics of evaluative terms such as “good” or “beautiful” a problem for the non-normative accounts of common ground emerges. I then follow Geurts (2019, 2020) in exploring the idea of a relational common ground and sketch a way of seeing evaluative disagreement in terms of refusing to accept certain kinds of commitments.
18/09/24 Jakub Pruś
Ignatianum University In Cracow
Fixing the Meaning for the Sake of Argument – Semantic Arguments, the Types and Criteria for Assessing
Suppose one aims to persuade that “Julius Caesar was a criminal.” To support that claim various arguments may be formulated, such as: (a) Joseph Stalin murdered his political opponents who openly opposed him, thus he was a criminal. Julius Caesar did the same. Therefore, if Stalin was a criminal, then Caesar was also a criminal; (b) Most historians studying the history of the Roman Empire consider Julius Caesar to be a criminal, so he was one; (c) Anyone who murders or orders the killing of others is a criminal—regardless of whether it’s in defense of the country or to maintain power. Julius Caesar issued orders resulting in thousands of deaths, so he was a criminal.
Argumentation theory distinguishes and extensively describes different types of arguments, each with its own definition, argumentation scheme, subtypes, and criteria for evaluation. We may, therefore, identify that: example (a) employs the argument from analogy, and (b) is an argument from authority. But what about (c)? Is there a theory for such arguments?
The argument in example (c) establishes the meaning of “criminal” by extending it to those who lead troops into battle or kill in self-defense, and in this way, Caesar can also be included in the extension of a “criminal” (regardless of whether he committed genocide or simply won battles).
It turns out that such a way of arguing, by fixing the meaning by persuasive definitions, persuasive quasi-definitions, verbal classifications or attributions, is not only very common in various types of discourses, but also a very powerful one. Semantic arguments can be found in the debates on "racism", "rape", "peace”/"war", "abortion”, and "violence", but also on more specific topics, such as the definition of "planet” or "wetland.”
This type of argument has been carefully analysed in recent years by myself, along with Andrew Aberdein and Fabrizio Macagno as semantic arguments. The most extensive study of the concept, types and criteria for assessing semantic arguments, which were recently published in the book (in Polish): Argumenty semantyczne (Ignatianum University Press, 2023). In this talk, I discuss this work and its impact.
12/06/24 Gabriel Dupre
UC Davis
Language as a Theoretical Posit
If linguistic theory is to explain, rather than merely describe, the way that humans speak, it seems that it must identify the psychological structures involved in the acquisition and use of language. This makes language, the target of such theorizing, not an observational given, but a theoretical posit. This then raises the question, familiar from the philosophy of science, of what justifies making such posits.
In this talk, I will identify two major approaches within the philosophy of science to answering this question. One approach centers the predictive power of our scientific theories. Those theoretical posits which maximize our ability to predict (i.e. to predict a greater proportion, and at a higher degree of accuracy) the range of observations associated with a theoretical domain are most justified. This approach is generally captured well by prominent formalist approaches to confirmation, such as Bayesian Confirmation Theory. A rival picture centers around scientific explanations. Theoretical posits function to enable us to provide the deepest explanations possible, for those observations susceptible to such explanation. This may well be a small and atypical subset of the full range of observations.
I will develop a picture of these competing understandings of the role of theoretical posits in the sciences through discussion of some case studies from Jerry Fodor’s work. I will argue that his argument for a modular understanding of the mind exemplifies the latter, explanation-centered approach, while his argument for the reality of the propositional attitudes exemplifies the former, prediction centered, approach.
I will then, finally, turn to some recent debates in the language sciences. In recent work, Stephen Piantadosi has argued that Large Language Models “refute” traditional, generativist approaches to linguistics. I believe that this claim is justifiable only if we also adopt the prediction-centered understanding of the goal of theoretical inquiry. But this has not been the approach adopted by prior work within linguistics, which has rather focused on explanation. So this apparently empirical debate about the proper form a linguistic theory ought to take is transformed into a much more general philosophical debate about the function of scientific inquiry.
15/05/24 Paul Pietroski
Rutgers University
Logically Simple Concepts, Conjunctive meanings
Abstract: Here’s an old idea: human linguistic meanings are instructions for how to build concepts that exhibit intuitively impeccable inferences, including examples like (1) and (2).
(1) A baker buttered a bun with a knife at dawn; so a baker buttered a bun.
(2) No baker buttered a bun; so no baker buttered a bun at dawn.
The old idea raises questions about how the relevant concepts are generated: what kinds of atomic concepts are permitted; how are they combined; if psychological forms of conjunction and negation are invoked, are they computationally simpler than their logical counterparts in (3) and (4)?
(3) ƎeƎxƎy[Rexy & Fx & Gy & He]
(4) ꓯeꓯxꓯy~[Rexy & Fx & Gy & He]
I’ll describe a Language of Thought (LoT) that is very simple computationally, but strong enough to capture important inferential patterns. Then I’ll show how to supplement this LoT, in ways that approximate adequacy for natural language semantics, without modifying the core generative system.
25/10/23 Dan Zeman
University of Warsaw
Retraction: Conceptual and Empirical Challenges
Abstract: Retraction – the second-order act of “taking back” a previous speech act – is both an interesting phenomenon in itself and the basis of an argument that has played an important role in various semantic debates (for example, in that about predicates of taste, epistemic modals and other perspectival expressions). In this talk, I focus on several, more general, questions – of both a conceptual and an empirical nature. Thus, connection to the former, I aim to provide a sketch of an answer to questions like the following: What is retraction exactly? Is it mandatory? If so, under which conditions? In connection to the latter, I discuss the significance of several recent empirical studies testing whether the folk do, in fact, retract previous speech acts, and raise several methodological issues.
14/06/23 Katarzyna Jaszczolt & Chi-Hé Elder
University of Cambrdge Unversity of East Anglia
Dynamic Functional Proposition for Dynamic Discourse Meaning
Abstract: The tenets that the main meaning in discourse can be conveyed in a variety of (direct and indirect) ways and that it is shaped over a number of conversational turns are self-evident and as such not in need of theoretical or empirical support. Nevertheless, theories of utterance meaning in the post-Gricean tradition have typically focussed on the proposition expressed by the speaker in a single conversational turn that is recovered by the addressee with the help of some (conscious or automatic) modification of the uttered sentence. In this tradition, successful communication rests on the assumption that speakers and addressees come to a shared understanding of these propositions as they are produced in conversation.
Starting with the assumptions that meaning is jointly constructed over a number of turns in discourse and that the main meaning settled on need not be communicated directly, we focus on the fact, supported by wealth of empirical evidence, that speakers and hearers need not always converge on the main expressed proposition in order for communication to proceed unhindered: they may share partial understandings of individual utterances, allowing the overarching discourse meaning to unravel as the interaction progresses. We introduce a novel unit of meaning that accounts for such a dynamic concept that can emerge and develop over several turns at talk. We call it a ‘dynamic functional proposition’. This unit not only includes the linguistic meaning that has been communicated, but also other aspects of the situation, including the social frame, interlocutors’ levels of attention, emotions, and other non-representational aspects. These various aspects will have greater or lesser salience for different speakers, hence offering an explanatory tool for how utterance meanings are negotiated, as well as when and why misunderstandings occur. We also discuss briefly how this unit differs from other related concepts of a proposition utilised in linguistics and philosophy.
We conclude by demonstrating such a unit can be formalised, using a dynamic, contextualist theory of Default Semantics (Jaszczolt 2005, 2021b), pointing out the importance of various social and individual ‘filters’ that can affect the formal representation.
12/05/23 Herman Cappelen
University of Hong Kong
Conceptual Engineering and AI: What kind of agent is ChatGPT?
Abstract: Some (but not all) technological transformations put pressures on our core conceptual apparatus. Current development and use of AI is a paradigm. Conceptual Engineering can help us understand and guide these conceptual transformations. In this talk I use our concept of an agent (someone or something that can act) as an illustration.
14/12/22 Christopher Tindale
University of Windsor, Canada
Patterns of Argument: Aristotelian Topoi and Argumentation Schemes
Abstract: Scheme studies is developing as an important sub-area of argumentation studies. This talk looks to connect the topic to its root in the Aristotelian topoi through a discussion of logos. When Aristotle announces that, along with ethos and pathos, persuasion is produced by logos itself, he invites consideration of how exactly such persuasion arises through the pattern of arguments employed by arguers. Having highlighted important current research I turn to aspects of scheme theory that I judge to illustrate the rhetorical features involved, thereby suggesting an answer to the questions about the persuasive force of logos.
12/10/22 François Recanati
Collège de France, Paris
Reference through memory
Abstract: When you perceive an object and, later, remember it, are you thinking of the object in the same way, or in different ways? Proponents of ‘dynamic senses’ hold that the mode of presentation is the same. I will criticize, and hopefully rebut, what I take to be the main argument in support of that view.
25/05/22 Manuel García-Carpintero
University of Barcelona
Coordination and Presuppositions
Abstract: As several philosophers have pointed out, there is a difference between statements formalized in FOL as a = a and those formalized as a = b: the latter state the obtaining of a relation of de facto or external coreference, while the former presuppose one of de jure or internal coreference, sometimes called 'coordination', expressed in natural languages by relations of anaphoric dependence. Coordination relations are usually picked out by appeals to the understanding faculty (Schroeter 2007, Fine 2007, Pinillos 2011), and they count as analytic under a natural characterization of that notion. In previous work, I have defended a presuppositional account, at odds with formal accounts on which there is “nothing intensional” (Heck 2012, 144) in those relations. In the paper I will state a problem for the account of reference-fixing in general and de jure coreference or coordination that I hold, and I’ll present a solution I am considering.
08/04/22 Ray Jackendoff
Tufts University
The Conceptual Structure of the Word ‘Belief’
Abstract: What concept does the word ‘belief’ express, and how does it fit into the Theory of Mind? In its repertoire of syntactic niches (some not noticed in the literature) and in its range from stereotypical to marginal cases, ‘believe’ is quite a typical word. The analysis offered here reveals striking grammatical and semantic parallels to expressions that denote depictions, such as ‘picture’, ‘map', and ‘performance.’ These parallels provide unexpected solutions to classical philosophical puzzles about ‘belief.’ The analysis is extended to other propositional attitude predicates, to speech-act predicates, to predicates of actional attitude such as ‘intend’ and ‘vow,’ and to commitment to norms of all sorts. These conclusions provide a sense of the richness of folk psychology and of how detailed linguistic analysis can uncover it, presenting challenges for future research.