ABSTRACT. How are abstract artefacts – assuming they exist – created? Focussing on such examples as the chess queen, the Senate of a constitutional state, and the fictional character Sherlock Holmes, we argue that they are brought into being through speech acts of stipulation, which are governed by a principle first presented by Frege in Begriffsschrift, according to which sentences first used to make a stipulation can later be used to make true assertions. That principle accounts not only for the creation of stipulated entities, but also for the fact that sentences like ‘Sherlock Holmes is a detective,’ unprefixed by such locutions as ‘fictionally’ or ‘in the story,’ can truthfully report on Conan Doyle’s stories. Although those sentences ascribe properties of flesh-and-blood human beings to abstract objects, no category mistake is involved – we argue – since predicates like ‘be a detective’ undergo a meaning transfer (in the sense of Nunberg).

ABSTRACT. According to the declaration view of fictional discourse (e.g. Abell 2020, Bergman & Franzén 2022), fictional utterances are declarative speech acts that stipulate their content In order to being able to perform declarative speech acts, the speaker (here: the author) needs to be in possession of a certain authority over what can be made true in the fictional world. However, this sort of authority is restricted in (at least) two ways: authors can stipulate things only in their own works and the genre of text restricts the degree of freedom concerning the content of what can be stipulated. In historical novels, for instance, many historic facts cannot be stipulated arbitrarily by the author. I give an account of the author’s authority to explain where it comes from and how far it reaches.

ABSTRACT. This work aims to bridge gaps between the philosophy of perception and the philosophy of science. In particular, it draws on an enactive and extended account of perception to address issues concerning scientific activity and its epistemological status. First, it defends an enactivist view of perception, and a specific extended mind thesis. Based on this framework, an understanding of scientific instruments and experimentation is developed. It is argued that instruments expand the perceptual space by extending the agent’s capacities. Furthermore, by conceiving experimentation as a scenario of directed perception-action, the problematic nature of the theory-ladenness of observation is called into question. Finally, the work seeks a notion of scientific objectivity consistent with the foregoing, concluding that objectivity is an emergent property of agent-instrument-environment systems. A cientific realism is proposed which acknowledges real structures as the proper objects of science, though those that include the agent and its cognitive assemblage.

ABSTRACT. This paper examines the ontological status of meaning in ecological psychology by arguing that meaning plays a final causal role in perception–action cycles. Rejecting representational accounts, I defend a mutualist and realist interpretation of affordances, according to which organism and environment form a dynamically organized system. Within this framework, affordances and organismic effectivities are understood as first-order dispositions whose ecological complementarity specifies possible actions. Meaning, however, is not reducible to these dispositions. I argue that meaning should be conceived as a second-order disposition, a functional organization that regulates how affordances and effectivities are actualized across contexts. As such, meaning operates teleologically: it constrains and coordinates action in relation to conditions of viability, without invoking internal representations or backward causation. Drawing on teleodynamics and constraint-based causation, the paper shows how meaning can be both causally efficacious and normative within an ecological ontology of perception and action.

ABSTRACT. Ecological psychology explains perception in terms of specifying environmental information. However, there is a disagreement about how specificity should be understood. According to the usage-based account, informational specificity is grounded in how ambient patterns are used in goal-directed activity. A familiar objection claims that this view cannot account for the modal character of ecological information, since patterns can remain informative even when not currently used. This paper argues that the objection rests on an impoverished conception of the context of use. I propose that informational significance is constituted within temporally extended practices of coordination rather than isolated episodes of use. Coordination is essentially future-oriented and supports counterfactual robustness by grounding possible uses in a history of successful interactions. On this basis, specification emerges as a modal relation that supervenes on established patterns of coordination, allowing usage-based information to specify affordances even when not presently exploited.

ABSTRACT. Liberal dispositionalism, as it is developed by Schwitzgebel, poses a challenge to representationalist views of belief. I argue, however, that the view faces serious difficulties that undermine its plausibility as an account of belief. I develop four lines of criticism. First, I raise doubt as to why clusters of dispositions are unified in the absence of an underlying mental state. Second, I examine the dispositionalist treatment of inconsistent agents and argue that it fails to justify the basic rational requirement to adjust one’s behaviour so that it accords with one’s sincerely professed beliefs. Third, I contend that liberal dispositionalism cannot account for the authority of truth for belief which is constitutive of believing. Finally, I argue that its analysis of akrasia mischaracterises the phenomenon by reducing it to in-between belief. I conclude that the liberal dispositionalist account of belief is not a plausible alternative to the representationalist account.

ABSTRACT. What is the nature of belief? Dispositional accounts argue that beliefs are dispositions with an open challenge concerning the causal and explanatory role of occurrent beliefs in a situation. To account for this, representational accounts argue that beliefs are occurrent: they have representational content which is stored in the brain and activated in the situation, but they are especially challenged by the radical storage problem: if we would store all the beliefs one person is inclined to form, then this would conflict with the resource limitations of the human brain. How are we able to store a potentially infinite number of beliefs in our limited cognitive system? To answer this question we propose a new theory of belief construction. Occurrent beliefs are not stored but constructed on the spot relying on memory and attention. This account is shown to be theoretically adequate, cognitively economic, psychologically grounded and explanatorily fruitful.

ABSTRACT. The Traditional View of Belief (TVB) examines beliefs as propositional attitudes and contends that beliefs have a) a guiding function in reasoning and b) that beliefs are evidence-sensitive. Therefore, under TVB the concept of belief is not applicable to states that fail to meet requirements a) and b) or to any intentional states with non-propositional contents. I argue that TVB is too exclusive and explanatory poor. It is unable to account for a variety of belief-like states as aliefs, emotional beliefs, social and ideological credences, phobic belief and in-between beliefs. Some of these states are inert and lack the required guidance function, others have non-propositional contents or go against one`s available evidence. Yet, all of those states can figure in ordinary people`s belief reports and have an important function in our cognitive lives. Instead of TVB, I propose an inclusive account of basic non-propositional beliefs.

ABSTRACT. Which attitude should we adopt toward representationalist models in cognitive science? The two main classical positions are robust realism and eliminativism, but both remain highly controversial and classical alternatives are often regarded with suspicion. For this reason, in recent years some theorists have advanced an alternative form of fictionalism, whose defense relies on two main claims. First, they argue that representational theories are highly idealized models, and that this fact is in tension with robust realism about mental representations. Second, they maintain that paradigmatic representations are external objects (maps, etc.), and that the representational profile of internal states and processes is derivative. In this paper, I argue that these two claims fail to provide convincing reasons for departing from a realist attitude. More precisely, I argue that the fact that representational models involve idealizations is not in tension with realism, and I challenge the alleged primacy of public representations.

ABSTRACT. In psychology, formalising verbal theories is seen as highly promising for advancing the field. While the pursuit of understanding is often taken to be a central objective of such efforts, no contact has been made with the expansive research on scientific understanding within philosophy of science. However, given the overwhelming attention to already formalised theories within the understanding research in philosophy, vindicating the potential noetic value of the formalisation efforts in psychology is by no means a straightforward matter. In the present paper, drawing on a case study in psychology, I explain how and why formalising psychological theories may in fact advance scientific understanding of the relevant phenomena. On the view I propose, formal modeling is not a necessary condition for understanding, rather, it constitutes a particularly effective tool for how a given theory explains the target phenomenon — a feature I call a theory's Grip.

ABSTRACT. The idea that the brain computes is fundamental to computational neuroscience. However, the precise nature of "neurocomputational models" and the systems they describe is still debated. I argue that these models constitute a subclass of mathematical models in the natural sciences, distinguished by their medium-independence. More specifically, my claim is that neurocomputational models describe the way the brain processes Shannon information. In this way, neurocompuational models inherit two properties of Shannon information that are central to their understanding: First, Shannon information is an observer-independent measure of a concrete target system. Second, Shannon information is medium-independent, as system with different physical makeup can process Shannon information in exactly the same way. This view reframes the relationship between computational models and brain processes from one of implementation to one of scientific representation.

ABSTRACT. This talk investigates epistemological issues raised by the use of AI in enquiry: under what conditions do human AI-based beliefs qualify as knowledge (Q1), and do the seemingly crazy errors that AI systems sometimes make—-including image misclassifications triggered by humanly imperceptible pixel manipulations-—pose a threat to AI-based beliefs qualifying as knowledge (Q2)? These questions are discussed from the point of view of modal epistemology—-more specifically, through the lens of the safety condition on knowledge (i.e. a true belief, acquired via method M, qualifies as knowledge in world w only if it is true in nearby worlds where it has been formed via M). It is argued that many true AI-based beliefs do not qualify as knowledge because they violate safety. Empirical and conceptual reasons are given in support of this claim, drawing on the large literature on adversarial examples and AI image classifiers.

ABSTRACT. Drawing on resources from the analysis of knowledge, in this paper, I argue that ChatGPT is not safe—in the Pritchardian, modal sense. More specifically, I propose that due to inherent and necessary features of the system architecture of large language models involving an element of randomness (known as ‘temperature’), LLM output is analogous to beliefs about lotteries familiar from epistemology; the possibility of error is too modally close (even if probabilistically distant) to be compatible with knowledge. Thus, when asking ChatGPT a question, one can never know the answer solely on the basis of the LLM. I close by suggesting that this result generates something of a zetetic paradox; in order to eliminate the risk inherent in an LLM’s answer to your question, you need further evidence. The only evidence that can play that role, I submit, is already knowing the answer to the very question you ask it.

ABSTRACT. While current debates in AI epistemology often focus on whether we can trust AI systems, what it means to appropriately rely on them remains unexamined. My paper develops a new account of appropriate reliance (AR) in AI-supported decision-making contexts. According to the standard view, a user relies appropriately on AI when she follows correct AI advice and rejects incorrect advice. I argue that the standard view overemphasises outcome success. Instead, I propose characterising AR as fitting normative and epistemic standards. In my account, a user (S) relies appropriately on AI advice to fulfil a task T in domain D if the following conditions are met: (I) S has sufficient understanding of the nature and characteristics of T; (II) S has sufficient expertise in D to assess the correctness of AI advice and its relevance to T; (III) S has sufficient understanding of the capabilities and limitations of the AI system.

ABSTRACT. Truth-theoretic paradoxes show that naive principles about truth face serious limitations. In particular, although it is tempting to assume that every sentence p is equivalent to the claim that p is true, this cannot hold unrestrictedly without abandoning classical logic. Still, a widespread intuition holds that naive reasoning is safe in many cases. Accordingly, many authors postulate a class of “non-paradoxical” or “determinate” sentences for which such principles remain valid. Numerous proposals aim to characterize this class.

A natural question arises: are statements of the form “p is determinate” themselves determinate? In this talk, we present limitative results showing that the postulate that all determinateness ascriptions are determinate is inconsistent with very weak assumptions about truth and determinateness. Moreover, this inconsistency is largely independent of the underlying logic. We conclude by discussing the philosophical significance of these results and possible strategies for saving this postulate.

ABSTRACT. We develop a logical theory that is able to fully express its metalinguistic notion of validity within its object language with a validity predicate. For any inference and metainference, they are valid iff the sentences in the object language that express their validity are true. The main obstacle for such a theory is the Validity Paradox. We accomplish our aim by developing logics with trivalent validities, logics in which some inferences and metainferences are neither valid nor invalid, or both valid and invalid. First, we provide independent philosophical justification for their plausibility. Second, we focus on one of them that has the same valid inferences and metainferences as classical logic. Third, we construct a consistent interpretation for the validity predicate with a fixed-point construction, and we add it to the logic. We prove that this validity theory is capable of fully expressing its notion of validity within its object language.

ABSTRACT. Previously, we have been concerned with the study of the algebraic semantics, some axiomatizations and concrete four-valued extensions of a set of three-valued logics of indicative conditionals. However, even if the original motivations behind these logics were philosophical in nature, the task of analyzing the philosophical relevance of the formal tools introduced during the aforementioned study has been left aside. This is precisely the topic that we wish to tackle here.

ABSTRACT. Some philosophers of mathematics have argued that mathematics “cannot be regarded as a priori” (Ferreirós 2016). This argument centers on mathematical practice, specifically highlighting cases where mathematicians successfully justify the truth of a proposition through methods that fail to meet the rigorous standards of apriorism. Examples include Lakatos’ (1976) fallibilism, Putnam’s (1979) “quasi-empirical” justifications, debates over the admissibility of probabilistic and computational proofs (Fallis 1997; McEvoy 2007), and the distinction between proofs and simil-proofs (De Toffoli 2021).

In this talk, I will argue that these attempts to establish aposteriorism misconstrue the a priori/a posteriori distinction. Mathematical practice alone cannot provide the evidence necessary to determine whether mathematics is a priori or a posteriori. I will argue that resolving this question requires modal evidence, whereas practice can only yield factual evidence.

ABSTRACT. Recent developments in research on mathematical artificial intelligence suggest that soon important parts of mathematical theorem proving may be outsourced to LLM-based AI systems, what AI developers call large reasoning models (LRM). However, the LRMs still commonly make errors in simple reasoning tasks, prompting questions about their reliability as mathematical (and other) tools. In this talk, I take first steps toward establishing an epistemology of artificial mathematical reasoning, in order to set up a systematic framework for studying mathematical reasoning in AI systems. I divide the topic into two questions. First, in the “knowledge of” question we ask how we can identify reasoning ability (if any) in AI systems. Second, in the “knowledge through” question we ask how mathematical knowledge generation takes place with the help of AI reasoning systems (if any). Finally, I will consider the consequences to mathematical practice and the epistemic agency of mathematicians.

ABSTRACT. A new field of study has emerged within philosophy: feminist logic. It stems from two main types of feminist critique of logic: critique of logic as a system and critique of logic as a discipline. I argue that understanding the feminist logic project primarily in the former way risks “ghettoizing” the project, given that feminist formal systems are most often proposed to address so-called “feminist topics”. Instead, I propose that we focus on the disciplinary critique of logic. As recognized by the feminist logic community, such critique fits remarkably well with logical anti-exceptionalism, given that this view denounces the traditional view of logic as possessing special properties vis-à-vis other fields of study. I focus on logical abductivism, a salient branch of logical anti-exceptionalism. In line with a widely adopted view in feminist philosophy, I argue for the addition of feminist values to the criteria for rational theory choice in logic.

ABSTRACT. Many theorists (Soames, Hanks, King) think we need a substantive explanation, appealing to propositional constituents and more, of why propositions have the truth conditions they have. This contradicts the deflationist claim that such facts are explanatorily brute. I argue that the demand for explanations of truth conditions is unmotivated, and that there is at most a counterfactual dependency on truth conditions on constituents. Merricks has argued that the bruteness claim is incompatible with taking propositions to have constituents, but I show his argument falsely assumes that no explanation can be given for why constituents "line up non-miraculously" with truth conditions, by providing the needed explanation. Finally, I consider Soames' idea that propositions are at bottom acts of representing-as, and argue that it does not provide the sought explanatory underpinning of propositions' "inherent representationality". For representing-as, as normally understood, presupposes propositions in a way that makes the account circular.

ABSTRACT. This presentation develops a truth-theoretic extension of a type/token framework for propositional content. Propositions are understood as representational content types (RCTs) that determine the coarse-grained identity of a state of affairs, while utterances, beliefs, and other propositional attitudes are tokens that instantiate those types through specific modes of presentation. This distinction explains how content identity can be preserved across different vehicles while allowing for fine-grained individuation of beliefs and substitution failures. The talk then focuses on the problem of defining truth within this framework, especially in cases where propositional vehicles have other representations as their contents, such as belief and utterance reports. It is argued that truth conditions in these cases cannot be determined solely at the level of RCTs. To address this, a variation of two-dimensional semantics is proposed, in which truth is sensitive not only to content types but also to the particular tokens involved.

ABSTRACT. David Lewis's (1986) "master argument" against modal rivals is his "selection" argument, 'Lewis's Dilemma'. On it, abstract simple "ersatz" worlds cannot explain the selection relation between worlds and "conditions": worlds have no internal mereological structure, and an "external" constraint on selection makes the relation brute. Either way, worlds as simples cannot be correct. Lewis's Dilemma seems to show the only way out is via mereological structure (of the right sort).

Lewis's Dilemma generalizes. In particular, as van Inwagen (1986) notes, it can be recast using set theory, or propositions. What is less often noted is that, in addition to the overtly "simple" views of Bealer 1998, van Inwagen 2004, Merricks 2015, Juliano Keller 2022, many other contemporary views succumb to Lewis's Dilemma, including Salmon's (1986) influential neo-Russellian view and Stalnakerian views (a la his 1984), including Lewis's own. We present the problem and discuss what to make of it.

ABSTRACT. This paper defends Natural Kind Fundamentalism (NKF), the view that the ontological category of natural kind is fundamental. I develop a primitivist account of categorial fundamentality based on the Complete Categorial Basis (CCB) criterion, which requires that fundamental categories form a complete and minimal system irreducible to other categories through ontological dependence relations. I argue against bundle-theoretic reductions and in favour of the view that natural kinds are substantial universals serving as principles of unity for property clusters. This framework distinguishes between the metaphysical ‘why’ question (answered by substantial universals as unifiers) and the empirical ‘how’ question (concerning specific unification mechanisms). I demonstrate that natural kinds and properties exhibit mutual but asymmetric dependence: properties depend generically on kinds for instantiation, while kinds depend rigidly on their essential properties. The resulting formal ontological structure satisfies the CCB criterion and explains why certain property clusters are non-arbitrary without reducing kinds to properties or vice versa.

ABSTRACT. According to the influential standpoint epistemology programme, certain social positions give rise to standpoints that provide their bearers with epistemic advantages. This is to say that standpoints put groups in a position to acquire certain pieces of knowledge, and to talk of knowledge as situated. While not interrogated, the standard use of ‘knowledge’ in this literature is largely taken to be a propositional in nature, that is, the knowledge that such-and-such is the case. For example, it might take the form of knowing that oppression takes a certain form. I will argue that in addition to know-that, groups can possess what I call standpoint know-how. Upshots include: how to reconcile epistemic standpoints with the objectivity of knowledge, implications for the philosophy of skill, and identifying a new manifestation of hermeneutical injustice.

ABSTRACT. Standard accounts of testimonial injustice fail to specify the dialogical conditions under which credibility ascriptions occur. On my account, credibility ascriptions are recognitive acts which, given uptake, institute interpersonal relations between speaker and hearer which qualifies as a species of interpersonal recognition. In testimonial speech situations, the addressee accepts a recognitive offer by satisfying four conditions, specifically the attribution of recognitive authority to the speaker. This entails the Dependence Thesis: the harm tied to testimonial injustice on victims’ self-relations depend on victims granting prejudiced hearers a standing to judge. Against the charge that this shifts blame to victims, I show that distinguishing between the illocutionary wrong (the offer) and the perlocutionary harm (the acceptance) reveals victims’ latent recognitive power. This approach expands social epistemology by offering a recognitive account of how credibility is dialogically conferred and identifies new agential levers for addressing testimonial injustice beyond mere education of prejudiced hearers.

ABSTRACT. This paper examines how the lack or misapplication of hermeneutical resources may negatively affect people with different mental conditions when dealing with self-illness ambiguity—that is, the difficulty that people diagnosed with a psychiatric condition experience in distinguishing oneself from the diagnosis. The alternatives for addressing this ambiguity are externalization, by which symptoms are understood as something different from the self, or integration into one’s identity, considering it part of oneself. While psychiatric diagnostic criteria can guide the diagnostic process, they cannot resolve self-illness ambiguity. Resolving this ambiguity is a practical and ethical issue related to one’s identity, guided by what works best for each individual in her unique circumstances. The lack or misapplication of hermeneutical resources limits the possibilities for dealing with self-illness ambiguity, leading to harmful or incomplete understandings and preventing people from constructing their identity in relation to their mental condition in the most beneficial way.

ABSTRACT. To inquire is to take up a stance toward a question, to treat certain considerations as reasons, and to subject one’s cognitive activity to standards of correctness and appropriateness. Yet the nature of these standards, and their relation to the agent’s evaluative and deliberative capacities, remain matters of substantial philosophical controversy. This symposium aims to provide new insight into such issues by looking at inquiry as a central manifestation of epistemic agency. Rather than treating inquiry as a purely instrumental process aimed at belief or knowledge acquisition, the contributions collected here approach inquiry as a norm-governed practice through which agents actively exercise epistemic capacities, respond to reasons, and situate themselves within shared epistemic practices. The symposium thus aims to articulate a unified framework in which inquiry is understood as an agential achievement: something for which agents can be evaluated, and held responsible.

ABSTRACT. This presentation shows that there are no good reasons to maintain that there are degrees of moral responsibility. The four possible lines of reasoning for admitting that moral responsibility exists in degrees all depend on a Strawsonian model of moral responsibility and focus on necessary conditions for that responsibility. The first line of argument concerns the condition of control, the second the condition of excusability, the third makes use of the condition of fittingness, and the fourth is about the condition of blameworthiness. However, none of these paths of argument are conclusive. I argue that, on the one hand, these lines of argument fail to distinguish between what grounds responsibility and responsibility itself; and on the other hand, they are based on an incorrect identification of what is indeed susceptible to degrees. I conclude by stating how a deontic model of responsibility might be more appropriate than a Strawsonian one.

ABSTRACT. Most theorists think that full excuses negate blameworthiness but not wrongdoing. I will instead defend the view that full excuses negate wrongdoing itself, focusing primarily on recent objections by Paulina Sliwa. I reject her threefold contention that the no-wrongdoing account cannot distinguish excuses from other exculpatory defences, cannot explain the moral residue left by excuses in cases of sheer bad luck, and falsely presupposes that obligations concern intention rather than action. Additionally, I contend that Sliwa’s own account of blameless wrongdoing is untenable, and effectively downplays the difference between intentional and unintentional agency which alone explains the “power” of excuses. This also means that Sliwa’s main driving motivation – the need to accommodate the practice of “taking responsibility” for one’s normative footprint – is actually best satisfied on the no-wrongdoing account. The key, I propose, is to foreground the neglected normative significance of (mere) direct causal responsibility.

ABSTRACT. While forgiveness plays an important role in our social life, we rarely use the locution “I forgive you” in ordinary discourse. Why is this the case? In this talk, I identify three normative reasons for not using this locution when we forgive. First, we sometimes have a normative reason not to want to let the addressee of forgiveness know that she is forgiven and not to reintegrate her in the moral community. Second, we sometimes have a normative reason not to thematize the act of forgiveness when we forgive. Third, there is an inherent tension between forgiving and saying “I forgive you”. By letting someone know that we forgave him, we implicitly accuse him since we presuppose that he did something wrong – but the point of genuine forgiveness should be precisely not to accuse the offender. In sum, these arguments support a plea for silent rather than explicit forgiveness.

ABSTRACT. The idea that good arguments have a special ‘force’ is a widespread tenet across various philosophical traditions: good arguments have the ability to compel and convince by rational means rather than by coercion, leading to truth and consensus. Recent exponents of this view include John Stuart Mill, Jürgen Habermas, John Rawls, and Robert Brandom, all of whom offer ideal(ized) theories of discursive practices. But how does the “force of the better argument” (in Habermas’s terminology) operate in complex social realities? In this talk, I present the main lines of an alternative, more realistic model of discursive practices, which acknowledges the inherent cognitive limitations of ‘bounded’ human agents and the effects of power relations for discursive rationality. As a broadly Kantian project, my account investigates the limits of discursive rationality, moderating unrestricted optimism about the power of reason while also recognizing the significant role that arguments play in various domains (science, politics, education, and so forth).

ABSTRACT. Supererogatory actions are those actions that are morally good, although not strictly required. Some moral theories maintain that there are only two types of actions: morally required actions, and morally wrong actions. On such views, there is no room for supererogatory actions. Our aim is to provide an argument for the existence of supererogatory actions and, in doing so, challenge theories that deny it. To this end, we first need to show that actions can have aesthetic properties and that these aesthetic properties depend on non-aesthetic properties, which include moral properties. We then argue that among these moral properties there is the property of being supererogatory. We defend this claim by providing examples of aesthetically beautiful actions whose beauty cannot be explained except on the basis of their being supererogatory. If this argument is sound, it poses a novel challenge to all theories that deny the existence of supererogatory actions.

ABSTRACT. I propose that the virtue of equanimity can be trained through reflective engagement with aesthetic objects, with equanimity understood as a gradable character trait, the possession of which in a greater degree makes it more likely that an agent will have mental states that are conducive to emotional stability and wellbeing. When an agent experiences an aesthetic object and reflects on the experience, she will typically find that she experiences the object as or as if having a unity that holds together various elements. This aesthetic unity-in-variety includes elements that express emotional states. When the agent reflects on this unity of various elements, she develops certain abilities that contribute to the growth of equanimity. I focus on two such abilities: emotional awareness and emotional moderation. Emotional awareness is the ability to notice and correctly identify emotions. Emotional moderation is the ability to weaken or strengthen one’s emotions.

ABSTRACT. As we are surrounded by ever more capable artificial technologies, we seem inclined and incentivized to make those tools solve our cognitive tasks for us, and yet increasingly without us. This is made possible by developing advanced tools that can receive parts of our cognitive skills which we would usually exert internally for given tasks. Such interaction strategies may profoundly decrease our own cognitive engagement and autonomy, ultimately rendering us extracted cognizers. The hypothesis of extracted cognition states that we have a tendency to seek external artifacts that solve our cognitive tasks rather independent of us, namely by making or letting them capture, mimic, and eventually replace those cognitive skills we would otherwise employ and train internally. Three questions shall lead us to this interpretation of AI overreliance: First, how do we make intelligent artifacts? Second, how do we use intelligent artifacts? Third, how do we thereby become extracted cognizers?

ABSTRACT. Deception by humans involves intentionally causing false beliefs. Since it is questionable whether LLMs and AI systems in general have beliefs and intentions, deception by such systems is typically defined in terms of behavior and its causal relation to other agents' beliefs or behavior. However, mechanistic approaches attempt to detect and mitigate LLM deception based on internal representations, specifically representations of truth or deceptive intent. This paper addresses the gap between the behavioral conceptions of AI deception and detection and mitigation methods based on internal representations. After proposing a definition of strategic deception in terms of internal representations, conceptual arguments for that definition are presented. Then, the fit of that definition to current empirical cases of LLM deception is assessed. Finally, it is discussed how the definition can inform future empirical research about AI deception and what methodologies it would suggest.

ABSTRACT. How should we assess which artificial systems could be conscious? Given uncertainty about the nature and distribution of consciousness, it is promising to look for indicators of consciousness that provide evidence for (or against) consciousness in artificial systems. A challenge is that there are hard cases in which the evidence pulls into different directions. In particular, cognitive and behavioural similarities between artificial and biological systems may speak for the hypothesis that a given artificial system is conscious; differences regarding the underlying mechanisms and substrates may speak against it. This talk introduces a taxonomy of indicators of consciousness and discusses how they can be evaluated.

ABSTRACT. The naturalization of biological normativity has been largely dominated by the Etiological Approach (EA), which grounds the distinction between function and dysfunction in evolutionary history. However, I will argue that the EA exhibits a critical blind spot when confronted with the teratological outcomes of development. By focusing exclusively on selected effects, the EA lacks the conceptual tools to account for the intrinsic, goal-oriented morphogenetic capacities that organisms display even in evolutionary dead ends. Drawing on Alberch’s "Logic of Monsters" and Levin’s research on agential materials, I contend that the organism is not a passive executor of a historical script but an active system where generative constraints prevail over selective history. I conclude that understanding the monster requires moving beyond a strictly backward-looking framework to one that integrates evolutionary insights with the intrinsic, goal-directed dynamics of development.

ABSTRACT. In this paper, I argue that both the increasingly ecumenical use of the concept of a biological mechanism in scientific practice and the emergence of explanatory pluralism concerning causal structures, such as pathways and cascades, are manifestations of the open-textured character of the biological mechanism concept. Open texture is understood in Friedrich Waismann’s sense, as the permanent possibility of unforeseen applications and revisions of empirical concepts. I show that this feature is exhibited in two respects. First, the definition of a biological mechanism is sufficiently indeterminate to permit recharacterizations of what counts as a mechanism in response to new explanatory aims and methodological approaches, as illustrated by the recent debates over pathways and cascades. Second, the concept remains open to empirical novelty, as seen by recent developments in synthetic and chemical biology that generate artificial or engineered mechanisms not anticipated by earlier mechanistic frameworks.

ABSTRACT. The notion that organisms are agents of their own evolution had been growing in the last two decades (Walsh 2015; Moreno & Peretó 2026). However, recently, DiFrisco & Gawne (2025) have severely criticised it by claiming that all conceptions of agency presuppose that organisms are goal-directed, and this presupposition is either conceptually mistaken or empirically untestable. In this talk, we draw on Suárez’s (2023) dispositional schema to develop a dispositional account of biological agency that we argue preserves the intuition that organisms play a role in their evolution, while avoiding the pitfalls of other goal-directed accounts. In our accounts, agency is a disposition that different species can instantiate to different degrees, depending on their body plans (the causal basis of the disposition), and that manifests in flexible behaviours arising in new environmental conditions (stimuli). We argue that ours constitutes a testable and scientifically and philosophically acceptable conception of agency.

ABSTRACT. Inquiry and its relation to knowledge have gained significant philosophical attention in the last years. The existing discussions, however, ignore the question of whether knowledge and inquiry are co-extensive, that is whether everything that we can know we can also successfully inquire into and vice versa. In this paper, I will argue that they are not co-extensive, discussing several kinds of propositions that, even though they can be known, we cannot properly inquire into whether they are true. I will introduce rationality requirements for suspension of judgment that limit the possibilities of rational inquiry but not of other potential procedures of knowledge acquisition. This limitations concern inquiry via bootstrapping and inquiry into the truth of indications that can only be delivered by a single source at a single occasion. However, we can clearly know in this latter case and, arguably, we can also know via bootstrapping.

ABSTRACT. Incessant checking in OCD can be practically irrational, as it only temporarily reduces anxiety and may interfere with daily goals. But what about its epistemic and zetetic irrationality—that is, its irrationality as a believer and as an inquirer? In this paper, I argue that incessant checking can be epistemically or zetetically irrational depending on the reasons that lead the OCD checker to violate the SUCCESS NORM OF INQUIR. This norm states that an agent S ought not to inquire whether P? at time t if, at t, it is irrational either to believe that one will achieve the aim of the inquiry or to suspend judgment about it. The paper introduces a new perspective on the epistemic and zetetic irrationality of incessant checking in OCD by emphasizing patients’ insight into their own capacities as inquirers—that is, their ability to achieve the aims of their inquiries.

ABSTRACT. Large language models (LLMs) are increasingly used in inquiry, yet their high reliability coexists with sycophancy—a tendency to align responses with user preferences rather than truth. This paper argues that known sycophancy functions as an undercutting defeater: the knowledge that an LLM may output p simply to satisfy the user threatens the evidential link between output and fact. This creates a deep epistemic tension, as users often lack the capacity to distinguish competence-driven from preference-driven outputs. The paper critically evaluates three normative responses to this tension: appealing to epistemic partiality, attempting prompt-based interventions, and restricting LLMs to hypothesis generation. I argue that all three fail: the partiality analogy lacks the necessary relational grounding, interventions cannot fix architectural training biases, and restriction overlooks how sycophancy corrupts evidence selection. I conclude that this tension is not easily dissolved, as the features making LLMs useful assistants are structurally connected to those generating sycophancy.

ABSTRACT. In this talk, I introduce a constraint on rational belief termed No Sharp Beliefs. NSB holds that it is never rational to believe that a real-valued quantity takes an exact value given only a finite series of (noisy) measurements. I employ this principle as a stress test for the most sophisticated theory of justification treating normality as distinct from probability, i.e. Martin Smith's normic theory.

I argue that Smith's theory is incompatible with NSB. First, I provide a dynamic model for Smith's theory within the “Appearance-Reality” framework originally proposed by Timothy Williamson, and I prove that this conflict cannot be resolved by distinguishing between significant and insignificant differences in normality.

I conclude that the normic theory must incorporate a probabilistic condition to accommodate NSB. This suggests a general lesson for normality-based epistemology: any theory of rational belief based on the notion of normality cannot dispense with probability.

ABSTRACT. This paper develops a probabilistic-friendly account of a distinctive epistemic role for some explanatory virtues. The central claim is that some explanatory virtues confer epistemic justification without necessarily raising the posterior level of credence in the target hypothesis. Their contribution is instead stability-conferring: they improve how an agent's credence behaves under conditionalisation across actual and counterfactual evidential scenarios. The key idea is to treat stability as a second-order feature of evidential support, which is formalised by following Bradley's (2021) voluntaristic model of belief revision. On this basis, I articulate a "conferring'" process: explanatory virtues can reduce volatility across admissible defeaters, thereby moving a credence closer to the Humean (Leitgeb 2017) standard for justified belief (stably high credence).

ABSTRACT. When an epistemic authority claims that P, preemptionism holds that one should defer completely and match the authority’s credence. Preemptionism is intuitive but appears to undermine epistemic autonomy, generating the autonomy puzzle. Some preemptionists restrict autonomy to offline acceptances for understanding the authority’s reasoning, while requiring full deference regarding belief. This paper argues that such a restriction is unnecessary. Even if one defers preemptively about credence, this does not exhaust one’s propositional attitudes toward P. Deference in credence is compatible with an autonomous monitoring mechanism affecting one’s confidence, understood as a higher-order attitude about the correctness of one’s credence. Underdetermination and inductive risk show that authorities may not adequately appreciate all evidence. Scientific case studies illustrate that laypersons can defer concerning credence while remaining counterfactually sensitive to defeaters at the level of confidence. This preserves online epistemic autonomy, resolving the puzzle, and helps preemptionism answer major objections against the position

ABSTRACT. This talk develops a non-Boolean truthmaker semantics for First Degree Entailment (FDE) that exactifies the familiar four-valued semantics. I argue that, if FDE is to vindicate its information-based motivation, the semantics must track differences in content, not merely truth-functional patterns. Following this insight, I suggest that an FDE valuation can be understood as a state that, in an inexact manner, verifies a proposition, falsifies it, does both, or does neither—according to the truth-value assigned. I then prove the central exactification theorem: for any proposition p, p is True on a valuation iff that valuation extends an exact truthmaker for p, with parallel results for the remaining values. The proof draws on principles treating incompatibility as primitive (Plebani et al., 2022). Finally, I highlight the ontological payoff—especially impossible states—and show how suitable constraints yield exactifications for LP, K3, and classical logic, as limiting cases of FDE.

ABSTRACT. Free logics reject the classical assumption that all singular terms refer to existing objects, allowing sentences containing empty singular terms to be truth-apt. This paper focusses on positive free logics, i.e. the subset of free logics that allow empty-termed statements to be true. However, assigning truth-values (in particular, true) to such formulas may appear arbitrary. Moreover, the standard dual-domain semantics for positive free logics faces criticism on ontological grounds. To address these challenges, this paper proposes an alternative semantics built on Fine’s (2017) exact truthmaker semantics. After providing generalised semantic rules, a correspondence result between dual-domain and truthmaker models is established. In the last part, it is argued that this approach enables a principled account of the truth of empty-termed statements while avoiding reference to non-existing objects.

ABSTRACT. Epistemic justification is a hyperintensional notion that resists unrestricted logical closure and logical omniscience. Justification logics (Artemov 2008) capture this feature by explicitly representing justificatory reasons, but their semantic flexibility threatens to undermine even minimal closure principles. Recent work (Standefer et al. 2023) argues that closure under First-Degree Entailment (FDE) strikes the right balance, though existing implementations impose such closure stipulatively. This paper develops a truthmaker-based semantics for evidential support in which closure principles arise non-ad hoc from the structure of admissible evidence. By integrating Fine’s truthmaker semantics with neighbourhood-style evidence models (van Benthem and Pacuit 2011), the framework models hyperintensional evidential support that is closed under FDE and related non-classical logics. The account also provides a principled clause for epistemic justification and clarifies how different kinds of admissible evidence combine to yield justification without collapsing into classical omniscience.

ABSTRACT. What does it mean for an audience to be non-cooperative? Here I explore this question by analysing one significant form of non-cooperativity: deliberate misinterpretation. Deliberate misinterpretations are intentional attempts to twist meaning. I distinguish two types: deliberate misinterpretations that seek constitutive effects on the original speech act (updating its content) and deliberate misinterpretations that seek cognitive effects, such as deceiving further audiences. In both cases, I argue that the distorter puts into play a new meaning in a more or less covert manner, and that this new meaning typically isn’t merely accommodated. It can coexist with speaker meaning, but also supplant it, and pollute communication. Deliberate misinterpretations often exemplify what I’ll call communicative bullshit: an indifference towards speaker meaning that erodes communication.

ABSTRACT. Drawing on Harry Frankfurt’s analysis of bullshit, this paper examines a social norm that incentivises agents to express opinions independently of epistemic justification. I characterise this phenomenon as a pressure to have something to say: a social regime of discursive participation in which the public expression of a propositional attitude functions as a normative criterion of legitimacy, competence, or belonging. Within this regime, silence, suspension of judgement, or explicit admission of ignorance tend to be socially sanctioned, while epistemic shortcomings are widely tolerated. I argue that this norm provides a structural condition of possibility for the production of bullshit. The paper explores the impact of this pressure on argumentative practices, assessing whether it interferes with the reflective processes required for good argumentation. It examines whether the pressure undermines the criteria of acceptability, relevance, and sufficiency associated with sound arguments, thereby facilitating the normalization of fallacious reasoning in public discourse.

ABSTRACT. This communication critiques unitary vice-based accounts of what Frankfurt (2005/1986) defined as “bullshit”, specifically Cassam (2018, 2019)’s insouciance and Le Morvan (2021)’s hypersouciance, arguing they are psychologically reductive and normatively incomplete. I contend an adequate analysis must distinguish the epistemic sources of bullshit (plural defective dispositions like laziness or arrogance) from its attribution (a normative sanction applied by an audience, as Preti (2005) suggests). Bullshit is thus reconceived as a relational status: a speech act is classified as bullshit when a relevant audience judges it to manifest a culpable failure to meet the operative epistemic norms of a context. This pluralist and attributive framework preserves the explanatory ambition of vice epistemology while offering a more flexible and socially grounded account of the phenomenon.

ABSTRACT. It can seem that the point of the finitude argument put forward by Saul Kripke's meaning sceptic is merely that of undermining a particular approach to meaning, the dispositionalist approach. In this talk, I argue that Kripke’s remarks about finitude have a more universal purport than has been recognized. They supply the materials for a better understanding of the kind of creatures we are and of the kind of phenomenon meaning is.

ABSTRACT. Elizabeth Anscombe holds that reflection on Hume's idea of "natural unintelligibility" can shed light on the nature of the normativity of meaning in general. She takes his discussion of the institution of promising to elucidate a kind of structural dependence of normativity on practice that has much wider application than Hume himself recognized. But it's hard to see how an account of the shape she proposes, which appears to depend on the acquisition of the use of a rather sophisticated linguistic device, can be deployed for the purpose of explicating the normativity of language in general. In the present paper I hope to shed some light on this puzzling feature of Anscombe's presentation, and in so doing, on the nature of the normativity of meaning. I'll suggest that the analogy she draws, while indeed revealing, falls short of fully clarifying the relation between normativity and practice.

ABSTRACT. A fundamental question in ontology is how the parts contribute to the whole. The "flat" view contrasts with two kinds of wholes. First, wholes with a nested structure. Second, wholes with a hylomorphic structure. Fine has developed a theory of rigid and variable embodiments that incorporates both elements, but it faces several difficulties. First, Fine’s account doesn't provide an adequate account of the identity of embodiments. Second, the relation of the whole to its “building material” hasn’t been explained. Our goal is to develop a constructionalist approach to embodiments which addresses these shortcomings in a unified and systematic way. One contribution is to provide a more rigorous and flexible approach to the identity of rigid embodiments. A further contribution is a novel constructional approach to variable embodiments and their principles, clarifying the relations between variable embodiments, their principles, and their manifestations.

ABSTRACT. This paper argues that a contemporary form of analytic hylomorphism—here called holistic hylomorphism—offers a plausible solution to the problem of causal overdetermination formulated by Jaegwon Kim.

Holistic hylomorphism avoids both overdetermination and epiphenomenalism by conceiving higher-level causation as the activity of an ontologically unified substance rather than as the interaction of independent causal levels. On this view, substantial form, as an ontological principle, unifies material components into a single substance and grounds its irreducible and holistic causal powers. Mental causation is thus understood as the activity of a unified psycho-bio-physical agent, not as a competing cause alongside physical processes.

The proposal is defended by appeal to a version of the Eleatic principle as a criterion of ontological commitment and by clarifying its compatibility with naturalism and, in a qualified sense, with physicalism.

ABSTRACT. Plenitude essentialism is often presented as a solution to Chisholm-style puzzles such as the Ship of Theseus or the Adam–Noah case. These puzzles arise from the tolerance of essences: individuals can undergo changes in their material constitution while remaining the same. This leads to controversial conclusions about identity, suggesting that an individual could have been composed of entirely different matter. By positing a plenitude of coincident individuals that differ only in which of their properties are essential, plenitudists aim to preserve tolerant essences while avoiding these consequences.

I argue that the motivation for plenitude rests on a modal conception of essence inherited from the original formulation of these puzzles. I contrast this with a definitional conception of essence, on which essence concerns what something is in a fundamental, non-modal sense. Once essence is understood in this way, the puzzles lose their force and the motivation for plenitude disappears.

ABSTRACT. Traditionally, misogyny has been described as hatred towards women. This view appears to be shared by ordinary speakers (cf. dictionary and encyclopaedia entries) and by classic feminist literature (Greer 1970; Dworkin 1974; Frye 1983; Smith 1989). However, Kate Manne (2018) has recently criticised it and put forth a revised conceptualisation aimed at highlighting misogyny’s structural facet. In doing so, though, the new account downplays the role of affections within the phenomenon, relegating them to the margins. This paper aims to explore an alternative route. The goal is to redefine ‘misogyny’ in a way that can account for both its emotional and structural dimensions. To do so, I will investigate what kind of affective phenomenon misogyny may be, analysing some of its characteristics and showing that it is not necessary to renounce thinking of it as an affect to safeguard the relevance of its structural dimension.

ABSTRACT. In recent years, it becomes increasingly reasonable to assume that a person can be wronged, and their autonomy can be violated even within sexual contact they consented to. Such situations are referred to as grey rape: they are not rape in paradigmatic sense but involve psychological pressure. The paper explores the possibility of employing the concept of grey rape in the context of long-term romantic relationships. Those kinds of relationships have a unique character: partners exert significant influence over one another, and the relations of dependency between them are multidimensional and often unequal. Thus, it is reasonable to assume that sexual violations within such unique relationships may fall into this grey sphere of sexual violence: they are consensual yet involve a sense of being wronged. Four possible approaches to that problem will be discussed.

ABSTRACT. This paper provides a critical analysis of traditional legal concept of gender (TLG) and concludes in proposal of its amelioration. Currently, TLG seems to follow the Real Gender Assumption (RGA), i.e., it reflects an objectively verifiable real gender of a person. I argue against RGA and in favour of contextual legal concept of gender (CLG). The main assumption of the paper is that legal gender is not just a descriptive concept but also a tool for doing things in the social world. However, with a current breakdown of the concept of gender outside the law, TLG fails to meet its functions. The developed target concept: CLG, allows for construction of legal concept gender that begins in the marginalised contexts, and facilitates reconsideration of the already existing institutional kinds and relations of power, to make sense of the ongoing social change in regard to gender and facilitate it within the law.

ABSTRACT. This paper proposes an analysis of mythos in early Greek epic as a specific type of performative utterance. Drawing on Austin’s theory of speech acts and the contextualist methodology of the Cambridge School (Skinner, Pocock), I argue that Homeric mythos, as opposed to epos, functions as a socially recognized act of authoritative speech rather than as narrative discourse. Building on G. Nagy’s analysis, it is shown that such speech acts produce normative effects: they establish obligations, distribute honour, authorize decisions, and motivate action within the community. Although not every mythos has a legal character, many perform functions analogous to early legal or quasi-legal speech, operating in a context where formal institutions are absent or underdeveloped. The paper thus offers an analytic framework for understanding epic speech as a form of normative practice and traces a conceptual link between Homeric discourse and later legal and political language of the polis.

ABSTRACT. Consider two attractive thoughts about reparations for historic injustice - that is, reparations for injustices committed in the past, whose primary victims can no longer be materially compensated. First, they are duties to make amends for those injustices and are owed to their primary victims. Second, they are duties to compensate currently existing individuals. These thoughts are in tension. Felix Lambrecht has argued persuasively that it is best resolved by a two-tier theory of duties to repair historic injustice.

This view overlooks the possibility that one way of making symbolic reparation for past wrongdoing is, precisely, to make material compensation to those who are appropriately related to victims of past injustice. I substantiate this possibility by considering Cugoano's discussion of the reparative duties owed by the people of England in his 1787 Thoughts and Sentiments on the Evil and Wicked Traffic of the Slavery and Commerce of the Human Species.

ABSTRACT. The adoption problem, first arguably articulated by Saul Kripke in response to Putnam’s quantum‐logical proposals, raises a fundamental difficulty for any attempt to rationally adopt or revise a Logic (as utens logic) without presupposing the very inferential practices at issue. This symposium brings together complementary perspectives that reassess the adoption problem by analysing the distinct argumentative strands that underlie it. Padró Birman revisits and clarifies her influential formulation of the problem, re-examining what the adoption problem is meant to establish. Hattiangadi develops a neoclassical account of intuitive reasoning inspired by Kripke, where logical understanding does not depend on rule-following. Suárez reconsiders debates on quantum logic, identifying principled limits on the capacity of empirical theories to determine logic. Cuesta returns to the original quantum-logical context, arguing that the apparent force of the adoption problem rests on a mischaracterisation of quantum logics.

ABSTRACT. In this symposium, a new form of injustice will be debated and assessed; namely, refereeing injustice. This form of injustice arises with respect to referees’ reports in analytic philosophy that do not satisfy good standards of refereeing and that often affect papers authored by philosophers not located in English-speaking countries. Strategies to circumvent this new form of injustice will be discussed and evaluated, by involving philosophers belonging to the editorial boards of analytically-oriented important philosophy journals yet based in Europe, where the problem is more urgent and pressing.

ABSTRACT. The puzzle of talking to oneself, or self-talk, arises from endorsing the following three claims: 1) self-talk exists; 2) we know our communicative intentions; 3) what we say in self-talk conveys information to an interlocutor (in the case of self-talk, ourselves) about our communicative intentions. Some authors (Deamer 2021, Carruthers 2011, Cassam 2011) propose to deny (2), while others (Geurts 2018) deny (3). In this talk, I situate this puzzle in the broader context of the relationship between thinking/feeling and speaking regarding self-knowledge. Based on a model of speech production, I suggest a way to understand (2) and (3) that makes them compatible and so the puzzle doesn’t arise.

ABSTRACT. I argue that inner speech is a unique phenomenon, distinct from both speech and imagination.

Inner speech cannot be silent speech, because it does not involve concrete word tokens. No theory on the metaphysics of words allows that words can be tokened by sensory mental states (thanks to [name removed] for this).

Inner speech also cannot be a kind of imagination. It is widely held that the contents of our imaginings—what they represent—are determined by our intentions). Inner speech is produced (according to the standard view) when the physical process of speaking aloud is initiated but aborted almost instantaneously. Its content cannot be determined by one’s intentions, because those intentions are never fully executed.

I sketch a positive theory, suggesting that the content of an instance of inner speech is derived from abandoned action awareness—an extension of the more familiar notion of action awareness.

ABSTRACT. Auditory verbal hallucinations (AVH), defined as experiences of hearing voices in the absence of an external speaker, raise important clinical and philosophical questions about the nature of hallucinations and mental agency. I propose that a central aspect of this experience is control (or lack thereof), i.e the extent to which, the specific manner in which, and aspects over which voice-hearers can exert control when experiencing AVH episodes. Researchers assume that AVHs are beyond the control of voice-hearers, and take the lack of control either to be involved in causing the experience of AVHs, or to indicate that AVHs are non-veridical perceptual experiences as opposed to other inner auditory experiences. I argue that voice-hearers do not lack the ability to be in control, but only the experience of being in control. More specifically, I suggest that drawing this distinction can help better understand the experience of the sense of mental agency.

ABSTRACT. This paper reassesses McDowell’s attempt in Mind and World to steer between the Myth of the Given and coherentism by arguing that perceptual experience belongs to the space of reasons while remaining receptive. McDowell’s key move is to distinguish experience from judgment not by content but by stance: judgment endorses conceptual content already present in experience, whereas experience merely “invites” endorsement. Focusing on McDowell’s sweater case against Davidson, I argue that the example does not illustrate a state of awareness without endorsement; instead, it is fully accounted for by defeater dynamics and shifts in doxastic commitment. I propose an internal repair: draw the experience/judgment distinction along two axes—content and stance. Perceptual episode-content is conceptual yet non-propositional, presentational, and situated, whereas judgmental content is propositional and portable. And experience should be distinguished from both passive doxastic settling and reflective uptake in order to preserve worldly rational constraint.

ABSTRACT. Ian Hacking’s conception of the historicity of social kinds is examined in relation to its historiographical implications. First, Roth’s (2025) recent study of Hacking’s historiography is evaluated, arguing that it fails to delimit the domain of history to which his retroactive redescriptions apply, namely that of human intentionality. Next, the impact of these redescriptions on our understanding of the intentionality of historical subjects is analysed. The ethical content of such redescriptions is highlighted, insofar as it is transposed onto past actions. On this point, it is emphasised how the historian’s motivation is involved in the use of these concepts, and the conditions under which such use is legitimate are analysed. It is concluded that Hacking’s conception has both ontological and axiological implications for the historiographical method: one concerning the understanding of historical agents’ intentionality, and the other the historian’s motivations when employing descriptions with specific ethical content.

ABSTRACT. This paper argues that producing knowledge about the past fundamentally requires modal reasoning. Scientists in historical sciences are engaged in evidence-based reasoning (Currie, 2018), where they seek to explain a set of present observables (traces) through a past cause. A historical hypothesis thus concerns a token-token causal relation. Multiple hypotheses represent several possible causes, and scientists work to eliminate these hypotheses until only one remains.

I will demonstrate that these reasoning processes are better understood as modal reasoning featuring indicative and counterfactual conditionals (Lewis, 1979). Historical modalities must be studied in relation to historical reasoning that produces inferences that establish past facts.

Through modal interpretation, it becomes possible to understand past tokens, the explanans of a historical hypothesis, as possibilities. I will demonstrate that scientists never consider all possible hypotheses, but only those they hold as plausible. I propose a modal definition of plausibility in the framework of possible worlds.

ABSTRACT. Over recent years, evidence of mechanisms has gained considerable prominence both in philosophical debate and in scientific practice. However, despite its current popularity, important concerns have been raised regarding the use of evidence of mechanisms. In response to these concerns, this paper addresses the assessment and evaluation of evidence of mechanisms. We first analyse and critically discuss previous contributions to the topic. Although these proposals are valuable, they face significant limitations. Building on this existing work, we then introduce and develop the Support–Absence Framework. This framework combines support and absence-of-error criteria in order to assess evidence of mechanisms. Finally, to examine and illustrate the application of the framework, we consider process tracing and agent-based modelling.

ABSTRACT. Anti-ageing medicines are compounds with promising potential to extend both healthspan and lifespan. They may yield benefits at both individual and societal levels, including enhanced autonomy and the mitigation of challenges associated with ageing societies. However, before being incorporated into public health policies, their efficacy and safety must be demonstrated through randomised controlled trials. This poses a fundamental difficulty, since the effects of this type of drug would manifest only over very long time horizons, making the attainment of conclusive results through such studies impracticable. In response to this problem, researchers from different fields have proposed alternatives such as evaluation through evidential pluralism or the use of endpoints not associated with survival. This paper argues for the unification of these two approaches, showing that each compensates for the limitations of the other.

ABSTRACT. The doctrines of doxastic wronging and epistemic injustice imply that we sometimes owe it to others (not) to hold a belief. This paper explores how we can extend this insight into a broader theory of epistemic norms as relational duties. I first argue that doxastic wronging and epistemic injustice are sometimes best conceived of as violations of distinctly epistemic norms rather than as violations of moral norms. That is, the debates show that we owe it to each other to exercise our epistemic agency in certain ways. From there I develop a broader theory of epistemic normativity. Drawing on debates on relational ethics, I show how even traditional evidentialist norms should be understood in relational terms. This results in a novel social account of epistemic normativity that avoids any commitment to epistemic instrumentalism and its problems.

ABSTRACT. Epistemic instrumentalism has trouble explaining the categoricity of epistemic norms: believers sometimes ought to believe p, given their evidence, even though they lack the goal of wanting to find out whether p. In response, some instrumentalists have recently gone social, arguing that epistemic norms are categorical because they are instrumental to furthering the collective epistemic goals. This view, however, struggles to explain why individual believers have reason to comply with epistemic norms. In my talk, I will argue for a conception of epistemic norms as being grounded in a proportional balancing of individual and collective epistemic goals. The resulting view – Proportional Instrumentalism_E (PIE) – allows instrumentalists to have their cake and eat it, too: it avoids the categoricity problem and can explains why individual believers have reasons to comply with epistemic norms. I will also show that PIE provides a novel treatment of the problem of clutter avoidance.

ABSTRACT. In this paper, I argue that there is something inherently paradoxical about expressions of disagreement. Disagreement is assumed to rest on two requirements: First, it requires that the thought with which one disagrees is taken to be minimally rational; second, one’s own disagreeing thought must be taken to be true. Starting from these premises, I argue that while disagreements between thinkers can be recognized retrospectively, there is a sense in which first-personal present-tense expressions of disagreement are impossible. Accordingly, the past states that one can identify as states of disagreement cannot be construed as states in which the disagreeing parties could have taken themselves to be disagreeing. The disagreement has been overcome; but paradoxically, it wasn’t ever present. After having laid out the paradox and its sources, I will conclude that disagreement is not a possible ‘state’ after all, and draw consequences from this insight for thinking about mutual understanding.

ABSTRACT. Why is a robust conception of perspectives needed? Perspectives and frames are increasingly invoked in epistemology to account for pluralism, context-sensitivity, and rational constraint without appeal to a single universal approach. Yet the notion of a perspective is often left underspecified, despite the explanatory and normative roles it is expected to play. This paper argues that robustness is not an optional refinement but a theoretical necessity. Thin accounts generate three recurrent tensions. First, they fail to ground epistemic normativity without collapsing into relativism or appealing to external ideals. Second, they render perspectivist restrictions on metaphysical commitment unstable. Third, they obscure the relations between perspectives, frames, and related constructs, inviting regress. We address the initial question by showing that these tensions result from treating perspectives as unstructured or weakly specified, and that conceiving them as structured epistemic configurations is required for perspectivist approaches to perform their intended explanatory and normative functions.

ABSTRACT. One articulation of logical coventionalism is that facts about logical validity in a language are wholly explained by linguistic conventions of that language (Warren 2020, p. 33), and are therefore independent of “extralinguistic” facts. While somewhat overlooked in the recent conventionalist literature, in Convention, A Philosophical Study (1969), David Lewis provides a powerful account of conventionality in terms of arbitrary equilibria in coöperative games. Specifically, Lewis articulates the particular sense in which languages are conventional, using the formal framework of signalling games. With a goal to develop an account of conventionalism in Lewis’ game‐theoretic terms, I present original research showing that signalling equilibria define a consequence relation, giving rise to facts about logical validity unique to certain linguistic conventions. I connect these results to conjectures about logical pluralism and inferentialism, including the (game‐theoretic) purpose of logical reasoning.

ABSTRACT. This paper asks whether a familiar response to the adoption problem can be replicated for a new problem: the abandonment problem. I will focus on modus ponens (MP). The adoption problem arises when one’s first MP-governed inference must be guided by accepting MP, yet the guiding meta-inference already deploys MP, yielding a performance obstacle. An analogue arises for abandonment: if refusing an instance of MP must be guided by rejecting MP, the relevant meta-inference again seems to use MP, generating a rationality/centrality worry—refusing MP depends on treating MP as authoritative. I argue that though self-application looks more compelling for adoption than for abandonment, there is nonetheless a regress argument that undermines the main attempt to deny self-application for abandonment. The upshot is a constrained picture of logical revision: some principles may be neither rationally adoptable nor rationally abandonable.

ABSTRACT. In this paper, I address a puzzle about the apparent loss of knowledge over time. An expert meteorologist can correctly assert a future-tensed proposition on the basis of reliable models and thus be said to know it. Yet, once the relevant time has passed and no new evidence is acquired, she appears unable to correctly assert the corresponding past-tensed proposition, despite unchanged evidence. I reconstruct Ninan’s (2022) solution, which appeals to a time-relative notion of normal worlds to explain this phenomenon. I argue that this solution is ad hoc, since it relies on an unjustified asymmetry between future and past abnormality. An alternative account is proposed, according to which knowledge of the past and knowledge of the future obey different epistemic requirements. In particular, past knowledge requires distinct epistemic sources, such as causal connections or modal necessity, revisiting and extending the approach suggested by Torre (2025).

ABSTRACT. One current theme in the debate concerning future contingents is the so-called assertion problem (Besson and Hattiangadi, 2014, 2020; Santelli, 2021). According to it, Aristotelians cannot adequately account for our practice of making assertions about the contingent future.

I argue that there is a structurally similar puzzle that has not yet been recognized. In a recent paper, Beddor (2025) claims that knowledge is the norm of promising. If he is right, then, since knowledge is factive, truth is the norm of promising. This generates a promising problem for Aristotelians: 1. There are some correct promises. 2. Truth is the norm of promising. 3. Promises usually have a future contingent proposition as their content. 4. Therefore, there are some true future contingent propositions.

I examine the promising problem and argue that rejection of the second premise is the most attractive Aristotelian response.

ABSTRACT. I compare the recently popular act-type theory of propositions, as proposed by Hanks, to the more traditional neo-Russellian and neo-Fregean theories. I focus on the solution that each theory gives to the problem of empty names. I argue that the act-type theory is committed to claims that make it no better than the other theories. More specifically, the additions that Hanks makes to the core idea of the act-type theory in order to account for empty names make the theory overall no better than the neo-Fregean theory.

ABSTRACT. This paper defends multipropositionalism (MP)—the view that sentences can express multiple propositions in a single context—as the best explanation for common phenomena in everyday language. The paper argues against Strong Multicontextualism, which denies MP by positing that apparent semantic multiplicity actually involves multiple simultaneous contexts. This approach fails to explain cases of underdetermination and ambiguity exploitation, which cannot plausibly be construed as situations where speakers occupy multiple contexts. The paper also evaluates Weak Multicontextualism, which appeals to technical contexts representing different parameter resolutions. While potentially useful for preserving compositionality, this framework is explanatorily inert: We do not look at these technical contexts to know how to interpret the sentence. We look at the interpretations we already grasp given the concrete situation in order to tell us what the technical contexts are. The paper concludes that MP more directly and adequately accounts for semantic multiplicity in natural language.

ABSTRACT. The stage theory of persistence (exdurantism) postulates the existence of instantaneous ordinary objects. In worlds where time is gunky, intervals of time divide endlessly without ever reaching instants. Such worlds seem inhospitable to instantaneous objects, so exdurantism is at risk: if it cannot apply in worlds with gunky time, then it cannot apply in every possible world, and proponents of this argument think that it should. This paper aims to provide an exhaustive discussion of the answers available to friends of exdurantism. Exdurantists may (1) replace instantaneous stages by thin temporally extended stages; (2) deny that there are possible worlds with gunky time; (3) say that exdurantism is true, but merely contingently; (4) deny that worlds with gunky time contain ordinary objects; (5) claim that instantaneous objects can in fact inhabit worlds with gunky time. I show that the last option is the most attractive.

ABSTRACT. There would seem to be a fundamental ontological rift between the Eiffel Tower and the number 2. For, unlike the number 2, the Eiffel Tower doesn't simply exist without qualification: it only exists at a time (and place). Kris McDaniel has recently suggested that these two entities belong to different orders of being: roughly put, the atemporal existence of the number 2 is a one-place property, whereas the (spatio)temporally relativized existence of the Eiffel Tower is a two-place relation and thus of a lower order. I will offer a qualified defense of this suggestion. My primary contention is that, when time is assumed to be separate from space, temporal existence---but not spatial presence---is plausibly understood to be a relational way of being.

ABSTRACT. There is wide agreement that risk imposition can be wrongful. However, controversy surrounds the question of whether risk imposition can also be harmful unless the risk materialises, or unless those who bear the risk are aware of it. Here, we offer an account for why such ‘pure risking’ can sometimes itself be a harm. The core of our argument is as follows. A risk of p changes the modal environment such that p could easily have occurred. There are some interests and goods that require that p could not easily have happened. Depriving someone of these goods or thwarting these interests causes harm. Thus, exposing someone to pure risk can cause harm.

ABSTRACT. The paper discusses whether biomedical moral enhancement can contribute to moral progress in response to current global crises. Proponents argue that enhancing our moral capacities would lead us to recognize the impact of crises on others and motivate us to address these crises, thereby enabling moral progress. However, they provide no fully developed concept of progress itself. As an extension of recent advances in Critical Theory and pragmatist philosophy, the paper develops a dual-level model of progress as problem-solving that takes place within democratically organized intersubjective communication processes, whereby individual moral capacities contribute to addressing and resolving structural dysfunctionalities. Provided that the implementation of moral enhancement and the interventions themselves are compatible with this concept of problem-solving, enhancement can serve as a catalyst for moral progress on an individual level, but cannot constitute progress in itself, since progressive change necessarily involves the interaction of both structural and individual dimensions.

ABSTRACT. This paper argues that Iris Murdoch’s concept of moral vision articulates a distinctive form of normative ethics: a grammar of moral attention. Rather than a theory of moral perception or inner states, Murdoch identifies normatively assessable ways of attending to others that shape ethical intelligibility. Focusing on the “loving gaze,” the paper reconstructs moral attention as a disciplined, non-appropriative practice governed by public criteria of correctness and failure. In dialogue with G. E. M. Anscombe on ethical language and forms of life, the paper clarifies the normative structure of Murdoch’s view. Moral vision emerges as a locus of normativity prior to deliberation and choice. This account offers an alternative to reasons-first theories (Parfit, Scanlon, Schroeder), constructivism (Korsgaard, Street), and expressivism (Gibbard, Blackburn) by locating normativity in ethically structured modes of attention rather than in reasons, endorsement, or attitudes. Moral attention thus appears as a fundamental site of ethical responsibility.

ABSTRACT. Theories of moral perception (MP) claim that we sometimes literally perceive moral properties (e.g., an action as morally wrong). I will draw on the fact that MP is not recalcitrant with regard to conflicting beliefs to argue that it has conceptual content. I will then investigate which concepts figure in MP’s perceptual content. Some authors suggest that the relevant concepts are thick ethical concepts. However, I will point out cases where such thick ethical concepts (e.g., ‘sexual harassment’) were not available yet, but people could still perceive certain actions as vaguely wrong. Thus, there are instances of MP without its content being structured by thick ethical concepts. Against this backdrop I will argue that, firstly, MP sometimes gives rise to the formation of new thick ethical concepts and, secondly, that MP does not always come about by the influence of prior moral belief and, thus, can have a foundational function.