Seminar

This talk will sketch Wittgenstein's view on the relation between philosophy (as he understood it) and empirical science (as he understood it). I will focus on what he takes to be the main difference between the methods of these two disciplines: 'grammatical' investigation, and the construction of explanatory theories, respectively. Special attention will be paid to avoiding the confusion between science and scientism; while Wittgenstein had a rather neutral attitude towards the former, his target is the latter. I shall argue that for him the very point of doing philosophy is precisely to avoid engaging in an activity similar to that of the scientists, namely theory-construction. This is a ‘therapeutic’ approach -- there is no need to worry about providing explanations, since "there is nothing to explain" (PI 126) -- and can be illustrated in various ways; however, in this talk I can only discuss how it works on mathematics. 

Substructural logics are logics weaker than classical logic on account of lacking or restricting inference rules that can be schematically formulated without reference to any logical constants, aka structural rules. In the philosophical literature on substructurality, there is a widely shared consensus that rules of these kind are in some sense prior and more fundamental than the inference rules that govern the behaviour of the logical constants. This consensus was recently challenged by E. Zardini, who argues that the structural properties of a logic are grounded in properties of the logical constants. In this talk I will critically examine and, consequently, debunk Zardini’s arguments. I will wrap up by proposing a different view of structural rules, according to which they are sui generis proof resources, neither grounded in, nor grounding properties of the logical constants. 

In an influential paper, Mandelbaum argues that Gendler's characterisation of alief (Gendler 2008a, 2008b) suggests that alief either has propositional content or associative content, and that this gives rise to a dilemma. On the first horn, if alief has propositional content, then alief is not a robust notion. On the second horn, if alief has associative content, then the account fails to do the explanatory work set out for it. Either way, the account fails. (Mandelbaum 2013) Recently, Danón has suggested that we adjust the alief account (Danón 2021). Her proposal is that we should assign what she calls semi-structured propositional content to alief. According to Danón, this would allow us to evade the dilemma. It would also allow us to explain the peculiar features of aliefs. I argue that attributing semi-structured propositional content has none of the advantages mentioned by Danón. However, since Mandelbaum's argument is unsound, Gendler's account is left untouched. I finish off by clarifying the success conditions of the account put forward by Gendler, and by offering, in light of these conditions, a new challenge for the alief theorist. 

More and more cases crop up where the epistemic achievements we think are primarily due to the realization of a specific epistemic value (truth, coherence, simplicity, empirical adequacy, explanatory or predictive power, and so on – the list is open-ended) might in fact be primarily due to an entirely different value or several such values. Even when conceding that specific epistemic achievements might promote several values at once, claims of value priority remain moot. I explore whether we can warrant epistemic value priority claims or whether, on the contrary, we should settle for skepticism about them. I explore a diachronic variant of this issue by considering cases where historians of science allege different epistemic values were in fact promoted by scientific achievements than what the scientists themselves thought they were promoting at the time. The examples I draw on primarily come from the recent history of cognitive science, specifically the development of neural networks for modeling semantic cognition. 

In Hintikka and Sandu “What is logic?” (2007),  we defended a conception of logic according to which a logical system is adequate only if it can characterize (define) mathematical notions.  According to this view, expressions like “natural number” have a meaning (content, reference) only if we have a formula or a theory in our language which defines natural numbers (categoricity). In other words, only if an expression or theory in our language is able to remove non-standard (non-intended) models, can we provide, through the model theory,  a proper reference for “natural number”. Implicit in this program was the view that reference has priority over inference, which Hintikka typically interpreted as a priority of model-theoretical over proof-theoretical methods (e.g. Hintikka, The Principles of Mathematics Revisited, 1996). In the present talk, I will analyse some of the consequences of this conception and move beyond it. 

Although the prospect of cognitively and physically enhancing soldiers has been subjected to ethical scrutiny, there is no discussion on the ethics of enhancing generals, i.e., the part of military staff in charge of strategy. The aim of this presentation is to argue that within the moral framework of Just War Theory (Walzer 1977, McMahan 2009) the following three theses are plausibly true:

(1) It is morally desirable for any state actor to cognitively enhance generals (i.e., to create super-generals),

(2) If a state actor is morally justified to fight a war, then it is morally desirable to fight a war against state actors that employ cognitively enhanced generals,

(3) It is morally desirable for state actors to share technologies aimed at cognitively enhancing generals.

Social choice is the study of collective choices, preferences, beliefs, attitudes etc. For example,  it studies voting rules like the simple and the absolute majority rules, consensus and unanimity etc. In this talk I shall discuss the logical structure of such rules. I shall present two reconstructions in modal logic and in 3-valued logic and a more general formalization în the frame of the structuralist philosophy of science (van Fraassen, Sneed). Some philosophical implications will be mentioned. 

 Most non-classical logics that have been seriously studied are meta-classical: that is, claims about their consequence relations behave classically. Consider, for example, K3. Though the schema ⇒K3 Av¬A is not valid, the metatheoretic schema Γ ⇒K3 A or Γ ⇏K3 A is valid. In this talk, I make the notion of meta-classicality precise and iteratively extend it to notions of meta-meta-classicality, and, more generally, metan-classicality for any finite n. I then argue, based on considerations from the logic of truth, that the correct universal logic is non-classical all the way up: that is, it is not metan-classical for any n. One upshot of this, which does not hold of logics that are metan-classical for some n, is that our informal background reasoning about logical consequence cannot be unrestrictedly classical. I conclude with a discussion of the significance of this. 

Paradoxes of time, and self-referential paradoxes like 'Liar paradoxes' are very different. In this talk I argue that both can be resolved by appealing to a type of logic that has recently gained traction in the philosophy of mathematics - a two-dimensional temporal logic. Temporal logic has proven fruitful in modelling numbers and sets as indefinitely extendible structures. Two-dimensional logic turns out to allow us to look at stages or times from multiple temporal (or stage-based) perspectives. In this talk I argue that such a logic helps to solve puzzles in time, and self-reference too.

If we say on Sunday: 'There will be a sea battle tomorrow' then non-fatalists seem committed to say that the proposition is indeterminate. If there is indeed a sea battle on Monday, is seems that, from the perspective of Tuesday, Sunday's claim is true. But can propositions change truth value like that? I argue that perspectival, two-dimensional logic can improve upon previous suggested resolutions to such problems. But I argue that it can also do much more than that. The very same logic can also help disambiguate semantic paradoxes such as Liar Paradoxes. If on Sunday John said 'What Mary will say this time tomorrow will be true' then the same time on Monday Mary says 'What John said this time yesterday is false' it seems that each statement can only be true if false and vice versa. This raises major questions about the nature of truth-predication. I argue that two-dimensional temporal logic exposes the temporal structure of such Liar-networks, and that such disambiguation helps to resolve them. Even a single statement such as 'This sentence is a lie' turns out to involve a two-step process that can be disambiguated and solved this way. This offers profound implications for some important areas of philosophy riddled by paradoxes of self-reference. 

 My talk will sympathetically explore a Wittgenstein-inspired view of morality – one that combines recognition of the non-natural character of basic moral concepts (such as OUGHT), the existence of moral facts, and the dim prospects for moral knowledge. This perspective will derive from applying, to ethical/moral discourse, Wittgenstein’s deflationism about TRUTH and FACT, his conception of meaning as ‘use’, anti-theoretical, ‘quietist’ meta-philosophy, his pragmatist view of language as a useful multi-purpose instrument, and his pluralistic appreciation of the variety of functions, types of concept, and norms of belief that this instrument needs to encompass. 

In Writing the Book of the World (2011), Sider argues that joint-carving concepts are intrinsically epistemically better than non-joint-carving concepts. In this talk, I will address the question of whether we can make sense of the intuitive betterness of joint-carving concepts without having to appeal to intrinsic goodness or postulating joint-carving as its own epistemic value. In other words, I will explore whether we can account for the intuitive or apparent betterness of joint-carving concepts in terms of other, more familiar and already widely recognised epistemic values. Three main proposals will be explored: (1) that joint-carving concepts allow for more true generalisations, (2) that joint-carving truths are more valuable than other truths, and (3) that joint-carving representations of explanations generate understanding. I argue that none of these proposals succeed in accounting for the epistemic value of joint-carving concepts. 

The Socratic imperative to examine your own life is linked to the idea of personal autonomy. Therefore, it is quite hard to reject it, but it is equally difficult to apply it, for various reasons. As long as it is philosophically assumed, it has important consequences when it comes to one’s meta-philosophical alternatives. In this talk, I aim to focus on philosophical autonomy and philosophy as a way of life.