Keynote speakers


Charles Travis: Changing Places

Frege, as is well known, drew something like a categorial distinction—perhaps even a more profound gulf—between concepts and objects. (Not that ‘concept’ has any unique agreed on denotation.) His way of drawing it has been inspirational along several lines in philosophy. It has, notably, inspired such outstanding philosophers as Cora Diamond and Tomas Ricketts, frst to a view of the (im)possibility of describing, or limning, the logical structure of thought, or its most general structure. One can grasp this structure, the idea is, in one’s capacity to recognise what clear and perspicuous expression of thought would be. But one cannot say what this is. More generally, on the view thus inspired, one, perhaps the main cause of philosophers so often end up just babbling, or talking pure Geschwätz, is that the nature of (traditional) philosophical questions forces one to try to ‘describe thought from outside it’, something comparable to jumping out of one’s own skin, or observing oneself rushing forward, or so on for a string of similar images.

The trouble is that on the concept-object distinction Frege was just wrong. He confused what are essentially, by his own lights, relational notions, or notions of roles, for notions of categories—as if in syntax one confused the notions noun phrase and subject. The confusion is easy to make, a point to be noted, less easy to unravel. Doing this will involve tracing the pedigrees of the two central notions in Frege’s thesis, objecthood and concepthood. On this last there are, of course, quite different philosophical notions of concept. The central one here will not quite be Frege’s (post 1892). But it will be the most ft to purpose. Working through the others—notably Frege’s own—can (or here will) be left as an exercise for the listener. As to object, the relevant notion is the one which, as Frege points out, is logikgemäss.

Blindspots are rare in Frege. Undoing this one matters to us for several reasons. One is that it is a way of undoing notions of the centrality of language in a general account of representing-as (specifically, the business of Wahrsein). These are not Frege’s own notions, but are taken as givens, or truisms, by many of his (would-be) followers. Another is that it can help us to insights as to what application logic actually has to thought and its expression. And finally it matters to our conception of philosophy and of what philosophy might do.

Anna Ciaunica: Self-Consciousness, Shared Embodiment and the Sense of Self

The question whether subjective conscious experience is a mere gloss on the top of a sophisticated bodily machinery designed to survive and reproduce, or something more, has long-time fascinated both philosophers and scientists. We typically experience a “real me” that resides in my body as the subject of experience and thought. Self-consciousness, the feeling that our experiences are bound to the self – as a unitary entity, the “I” – is thus a fundamentally first-personal subjective experience and is considered to be one of the most astonishing features of the human mind. Famously philosophers argue that it is possible to imagine a world in which my body behaves in exactly the same manner I behave in this world, yet without any subjective conscious experience inhabiting it. A “zombie” me, so to speak (Chalmers 2006). In this talk I will leave aside the philosophical implications of these thought experiments and simply focus on the mechanisms underlying the emergence of subjective conscious experience in the “real me” inhabiting my body in this particular world. Indeed, whilst is it relatively easy to imagine an “embodied zombie” (i.e. without subjective experience), imagining a conscious subjective experience without some kind of underlying embodiment seems less appealing.

One basic yet overlooked aspect of current debates on the nature of self-consciousness and perceptual experiences in this particular world is the fact that human beings start their journey into their experiential life within the body of another experiencing subject. Indeed, not only do my subjective experiences not occur in a vacuum – they are given to me through my body – but more crucially, my bodily experiences emerge from the outset within an other’s experiencing body. In this paper I discuss some implications of this overlooked relational and primitive aspect of our embodiment on current debates on the nature of self-consciousness. More specifically, I will focus on touch as paradigmatic example of perceptual awareness, i.e. as fundamental point of contact of an experiencing subject with the external environment. I suggest that endorsing a developmental perspective in addressing the nature of subjective conscious experiences might help us to: a) reframe current debates on the nature of self-consciousness from a more ecological perspective; b) better address some crucial challenges that contemporary the field of Developmental Robotics faces when it comes to defining “the sense of self” in a robot.

Martine Nida-Rümelin: Pre-reflective Self-awareness and the Nature of Experiencing Subjects

To understand the nature of ourselves as experiencing subjects crucially involves understanding what our own identity and our own individuality consists in. I will argue that – according to our natural implicit understanding – the so-called simple view of identity (and a corresponding view concerning our individuality) is correct. This understanding is deeply incorporated in our conceptual architecture. It might still be inadequate. I will argue that it is not. That understanding of our own nature is based, as I will try to show, in the way in which we are constantly aware of ourselves as the one single subject undergoing changing experiences. According to the view I will present, our understanding of ourselves as experiencing subjects has its origin in what one may call pre-reflective self-awareness. Furthermore, pre-reflective self-awareness gives us access to what it is to be an experiencing subject. – My argument will crucially rely on a particular understanding of experiential properties, which can be characterized in the following way: any instantiation of an experiential property consists in the way it is like for the subject involved to have that property. My talk thus relates to the recent debate about so-called mine-ness. In passing, I will propose to distinguish three phenomena so-called mine-ness may be taken to refer to.

Andrea Giananti: The Intractable Problem of Perception and Reason

Does seeing that p entail knowing that p? This question has recently attracted a great deal of attention, especially in the context of the current lively debate surrounding epistemological disjunctivism (ED). However, the participants in the debate have failed to realize that the question has far-reaching implications concerning rationality and deliberation. In this paper, I illustrate these implications by setting up a puzzle about belief-suspension. The conclusion that I draw from the puzzle is mainly negative: the epistemic contribution of perception cannot be explained in terms of a warrant-conferring relation between perception and belief. However, toward the end I sketch a positive picture of the epistemological role of perception in terms of a direct explanatory relation between perception and knowledge.

Andrea Raimondi: Is Intentionality Sufficient for Mentality?

According to a familiar suggestion, intentionality is the mark of the mental. This suggestion is supported in terms of the intentionalist thesis (IT), according to which intentionality is a necessary and sufficient condition for mentality. IT has been rejected by many philosophers, who argue that there are non-intentional mental states, such as emotions and moods (Bordini 2017, Dretske 1995, McGinn 1996, Searle 1983) and purely qualitative states, like bodily sensations (Antony 1997, Voltolini 2014). Crane (1998, 2001) and Tye (1995, 2000) have sought to rebut these objections.

Another option – less discussed – for the opponent to IT is rejecting the thesis that intentionality suffices for mentality. Nes (2008) tries to reject this thesis, moving from Crane’s (1998; 2001) necessary and conjunctively sufficient conditions for intentionality, i.e. (a) directedness upon an object and (b) aspectual shape. The idea of Nes’s argument is that on a simpleminded but natural reading of (a) and (b), it comes out that certain non-mental states are intentional.

Firstly, I shall reconstruct Nes’s argument. Nes redescribes (a) and (b) in terms of two features of the intensionality of reports: (a*) a state exhibits the relevant feature of directedness only if the corresponding report suffers from failure of existential generalization; (b*) a state exhibits the relevant feature of aspectual shape only if the corresponding report suffers from failure of substitutivity. Then, Nes shows that there are some non-mental states whose reports suffer from failure of existential generalization and from failure of substitutivity, which makes them intentional states.

He then considers and rejects two different responses: (i) a weak response, according to which one should spell out intentionality in explicitly mental terms; (ii) a strong response, according to which one should ratchet up the necessary and conjunctively sufficient conditions for intentionality. After presenting a variety of strengthenings, I shall show that either fail to exclude all non-mental states or are so strong that they ground new challenges to the thesis that intentionality is necessary for mentality.
I shall consider a third response, Crane’s (2008) reply, trying to reconstruct it rigorously. He argues that the relevant sense of intentionality is that of intentionality-as-representation and, hence, (a) and (b) should be reformulated as follows: (a’) representativeness (the capacity of an intentional state to represent things) and (b’) perspectuality (the fact that an intentional state represents things in a certain way, from a certain perspective). I shall show how Crane’s proposal seems to provide a new form of intentionalism, which seems to respond to Nes’s objection (Nes’s examples of non-mental states cannot be said to represent anything at all). I shall argue that Crane’s representationalist proposal is not adequate: although it is true that it is not correct to attribute to non-mental states any representative feature, this is not enough to provide a satisfying reply to Nes, for the new account of intentionality-as-representation plays no relevant explanatory role, if compared with the “old” account of intentionality.

Camilla Colombo: The loss/no gain and the doing/allowing distinctions

In 1979, Daniel Kahneman and Amos Tversky discussed their experimental findings for a famous case-study known as the “Asian flu” case. According to their interpretation, decision-makers are affected by how the choice problem is framed; specifically, different framings select different reference points as the salient baseline, and induce a perception of the same amount of a currency as a potential loss rather than a potential gain. As people are more risk-averse when it comes to pursue gains and more risk-seeking when it comes to avoid losses, two extensionally equivalent decision problems can trigger different choices, depending on the framing. The psychological motivation behind these risk attitudes, they further argue, is the so-called “endowment effect”, a reasoning bias which amounts to the tendency to ascribe more value to things merely because we own them.

In section 1 of this paper, I describe how, according to the authors, the endowment effect of the baseline explains the empricial findings in the “Asian flu” case, thus acknowledging for the preferences reversal which characterises this example. I then argue that these same experimental results are consistent as well with the following interpretation: shifts in the baseline induce shifts in the agents’ classification of the same action as an instance of “doing harm” rather than “allowing harm to occur” (section 2). My hypothesis is that people are risk-seeking when it comes to avoid causing extra deaths– doing harm– and risk-averse when it comes to preventing more deaths– by the means of allowing other deaths to occur. I suggest the same risk attitude patterns analysed by Kahneman and Tversky call for a different psychological explanation: in this “ethically sensitive” example, the preference reversal reflects the fact that people have different moral intuitions over the permissibility of doing harm and allowing the same harm to occur. As a result, when the reference point triggers different interpretation of the same action as “doing” rather than “allowing”, this brings about different preferences over the decision problems.

The relation between the doing/allowing distinction and the loss/no gain effect has been widely discussed in the moral literature. Tamara Horowitz and Walter Sinnot-Armstrong, for instance, argue the “Asian flu” case shows that the distinction between doing and allowing collapses into loss/no gain perceptions, and could thus be expressed in purely non-moral terms as the result of individuals’ different risk attitudes with respect to losses and gains, together with agents’ sensitivity to framing. Arguably, as Frances Kamm notices, this conclusion would threaten the moral significance of the doing/allowing distinction. Against this, she claims that a) the loss/no gain and doing/allowing distinctions do not coincide; and b) doing/allowing classifications are not subject to framing effects as loss/no gains perceptions are. In section 3, I clarify my interpretation with respect to Kamm’s account: while my hypothesis is consistent with a), I argue that Kamm’s analysis does not successfully support b). Alternatively, my interpretation acknowledges for the possibility that the doing/allowing distinction is still subject to framing; I explore the bearing of this conclusion on the stance of the doing/allowing distinction in section 4.

Francesco Praolini: No Closure without Truth

It is well-known that the lottery paradox and the preface paradox both show that the following three principles are inconsistent: (Sufficiency) very probable propositions are justifiably believable; (Conjunction Closure) justified believability is closed under conjunction introduction; (No Contradictions) propositions known to be contradictory are not justifiably believable. This paper shows that there is a hybrid of the lottery paradox and the preface paradox that does not require Sufficiency to arise, but only Conjunction Closure and No Contradictions; and it argues that, given any plausible solution to this paradox, the acceptance of Conjunction Closure surprisingly implies the acceptance of the thesis that justified believability is factive. Obviously, this conclusion is troublesome for all philosophers defending views of justification that deny the thesis that justified believability is factive, while accepting Conjunction Closure.

Giacomo Giannini: Resemblance, Representation, and Counterparts

The thesis of this paper is that David Lewis’ Genuine Realism cannot provide an account for de re sentences. I first show that Counterpart Theory is relevant to de re sentences if and only if there is a connection between i) something’s being true at a world and it being represented at that world, and ii) between counterparthood and representation. I argue that this second condition cannot be met, because counterparthood is a kind of resemblance, and resemblance is neither a sufficient nor a necessary condition for representation. I then show that the extra features of counterparthood, uniqueness and the peculiar high degree of resemblance, do not bridge the gap. Finally, I examine three strategies to amend Lewis’ theory: i) the adoption of a new representation relation, ii) the use of conventions and prescription-based theories of representation, such as Kendal Walton’s, and iii) a weakening of the link between representation and counterparthood, and conclude that none of these succeed in salvaging Counterpart Theory. I then conclude that this flaw is fatal to Lewis’ theory.

Giulia Martina: Illumination and Colour Appearance

Objects can look different in different conditions of illumination while retaining their colour: for instance, a directly illuminated part of a uniformly white wall and a shadowed part of that wall look different in spite of being, and looking to be, the same colour. These very familiar cases of changing appearances have been used to motivate substantive philosophical theories of appearances. On these views, in order to explain how things look we need to posit some special subject- or context-dependent properties – appearance properties – objects possess in addition to their objective, stable properties, such as colours, shapes and sizes (e.g. Shoemaker 2000, 2006, Noë 2004, Schellenberg 2008, Hill-Bennett 2008, Genone 2014). In this paper, I argue that changes in colour appearance under different illuminants can be explained in terms of ordinary, objective properties we visually perceive and do not motivate a metaphysical commitment to properties of a different kind. I first appeal to evidence from theories of colour constancy, colour matching experiments and phenomenological considerations to show that visual appearances exhibit a structural complexity that can only be explained by acknowledging that properties other than colours contribute to the way things visually appear. I then argue that the best explanation for this complexity is that we can visually perceive the illumination of surfaces and that illumination distinctively contributes to determining visual appearance. I conclude that cases of changing appearances involving colours under different illuminants do not support the appearance properties view. Finally, I discuss an alternative proposal to mine, on which illumination is a dimension of colour appearance (Hilbert 2005, Jagnow 2010), and argue that we have no reason to accept its counter-intuitive consequences. Changing appearances do not requires us to abandon the naïve view on which the way the objects we perceive appear is explained by their perceivable, objective properties.

Gretchen Ellefson: Cooperation and the Conversational Record

For several decades, philosophers of language have widely assumed that conversations are cooperative activities between participants. Yet it has become increasingly difficult to understand conversations as standardly cooperative as we are presented with new data about how difficult it can be for many people—women and gender non-conforming folks, people of color, trans women and men, religious minorities and members of other underrepresented groups—to make their voices heard. In this paper, I appeal to three cases to argue that many conversations are not necessarily cooperative; that is, that many conversations do not have the features specified by or derivable from Grice’s (1975) Cooperative Principle. In some cases, conversations lack a shared purpose or goal between participants, in others, the speaker does not make cooperative contributions, and in still others, the audience in a conversation does not interpret the speaker cooperatively. Any conversation in which one of these three things take place should be understood as a non-cooperative conversation. Although Grice’s Cooperative Principle cannot account easily for these cases, David Lewis’s (1979) notion of conversational scorekeeping can help to explain how participants can engage in a non-cooperative way in a conversation. As we will see, Lewis’s picture makes space for participants to have beliefs about the conversation or to update the conversation in ways that are contrary to the Cooperative Principle.

I begin by spelling out Grice’s Cooperative Principle, arguing that we can derive three distinct requirements from his formulation: one on the speaker, one on the audience, and a coordination requirement between participants. Next I give three examples of non-cooperative conversations, illustrating a violation of each of the requirements of the Cooperative Principle. Next I show how Lewis’s notions of scorekeeping can help explain non-cooperative conversations.

James Brown: Ecumenical Expressivism and Normative Propositions

Metaethical expressivists have traditionally rejected the claim that normative assertions and attitudes have normative propositional content. This is because they claim that normative thought and discourse is not fundamentally in the business of attributing normative properties or relations to things. Whatever propositions are, the thought goes, they are the sorts of things that involve the attribution of properties and relations. However, some philosophers have recently argued that expressivists can maintain their nonrepresentational commitments without rejecting normative propositions. What expressivism needs, they argue, is a theory of normative propositions that explains how normative propositions can play the object of attitude role without representing substantive normative properties, relations, or states of affairs.

This paper examines this suggestion in the context of Ridge’s (2014) expressivist extension of Soames’ (2015) theory of cognitive propositions, the view that propositions just are types of cognitive acts or events. As well as being an interesting application of Soames’ theory, the possibility of an expressivist-friendly account of normative propositions has deep implications for metaethics (and elsewhere). For it challenges the widely held assumption that normative propositions are available only to cognitivist or descriptivist theories of normative thought and discourse. While this is a challenge that should be taken seriously, this paper argues that Ridge’s theory fails. However, discussion of why it fails provides some general constraints that any expressivist-friendly account of propositions must meet.

The structure of this paper is as follows. First, I outline Ridge’s Ecumenical Expressivism and his account of normative propositions (§1). Second, I raise an objection to the account (§2). Third, I consider an amendment to the account that avoids the previous objection but raises problems of its own (§3). Finally, I argue that the objections raised suggest a deeper structural problem underlying the attempt to unite the respective explanatory frameworks of metaethical expressivism and cognitive act theories of propositions (§4). Particularly, I argue that any expressivist theory of propositions needs a unified answer to the general question, ‘What determines the content of our propositional acts and attitudes?’ So while I conclude that the Ecumenical Expressivist account of normative propositions fails, the discussion suggests some general constraints that apply to any expressivist theory of normative propositions, providing direction for further investigation.

Jiwon Kim: Taking Metaphysical Indeterminacy Seriously for Moral Realism

I argue that metaphysical indeterminacy is compatible with moral realism. Indeterminacy is, roughly put, there is no fact of the matter. At first glance, indeterminacy is incompatible with moral realism because for moral realists there are determinate, objective moral facts. I introduce one of the attempts from a moral realist, especially that of Russ Shafer-Landau (1995), to make metaphysical indeterminacy compatible with moral realism. Shafer-Landau argues that we have to give up on the principle of bivalence (i.e. a sentence must be either true or false) for noncomparative moral sentences in order to make moral realism compatible with indeterminacy. A noncomparative moral sentences have a form of “a is F”. For example, “Paul is pious” or “Telling ten white lies is morally permissible.” Consequently, I argue that Shafer-Landau also has to also give up on the law of excluded middle (i.e. P ∨ ~P) if he gives up on bivalence. I find his attempt problematic because Shafer-Landau does not have to reject the principle of bivalence and the law of excluded middle for noncomparative moral sentences. It is because giving up the essential rules of classical logic is a serious cost for moral realists. When one gives up on the rules of classical logic, some of our intuitive judgments regarding noncomparative moral sentences become counterintuitive. Therefore, I suggest a way of making metaphysical indeterminacy compatible with moral realism without violating the rules of classical logic. My solution is to introduce Elizabeth Barnes’s (2010) account of metaphysical indeterminacy. Her account is advantageous over Shafer-Landau’s for two reasons. First, her model of indeterminate possible worlds does not give up on the principle of bivalence and the law of excluded middle. Second, Barnes’s metaphysically indeterminate possible worlds are compatible with moral realism, even with the most robust, non-naturalist kind of moral realism. Therefore, I conclude that metaphysical indeterminacy is compatible with moral realism without giving up on any of the rules of classical logic.

Jonathan Egeland Harouny: Accessibilism and Ideal Rationality

This paper argues for accessibilism: one is always in an a priori position to know which doxastic attitude one now has justification to hold toward any proposition. Whereas most (if not all) of the arguments for accessibilism rely on our judgments about what rationality requires of non-ideally rational agents (NRAs), the arguments offered in this paper rely on our judgments about what rationality requires of ideally rational agents (IRAs). Accessibilism must be true in order to account for the rational abilities of IRAs.

Matheus Valente: Communicating and disagreeing with distinct concepts: a defense of semantic internalism

This paper attempts to solve a conflict between semantic internalism and the principle that successful communication and genuine agreement presuppose identity of concepts between the relevant parties. My solution of the conflict rests on the thesis that there can be successful communication and genuine agreement between thinkers employing distinct concepts as long as there is a certain relation (of conceptually guaranteed co-reference) between them. In section 2, I present what I take to be the core thesis of semantic internalism and show how it conflicts with another independently plausible thesis. In section 3, I carve the logical space of possible solutions to the conflict into two halves: the liberal and conservative solutions. Section 4 assesses Wikforss’ conservative solution to Burge’s arthritis thought-experiment and concludes that it fails for more than one reason. Section 5 introduces a new case study involving a deferential concept. This case then serves as the backdrop for the account offered in section 6, whose main idea is that successful communication and genuine (dis)agreement depend on the concepts employed (or the representations used to express them) being guaranteed to converge on one and the same thing. I conclude, in section 7, by considering the similarities and differences between my account and a recent one presented by Recanati.

Nemesio García-Carril Puy: Defending the type/token theory against Hazlett’s argument in the ontology of music

The aim of this paper is to offer a defence of the type/token theory in the ontology of music against the argument presented by Allan Hazlett (2012). Hazlett’s argument is originally addressed against the idea that there are repeatable artworks. It will be shown that this original version of the argument is not methodologically admissible in the ontology of music. A methodological acceptable version of the argument will be provided. This remodelled version concludes that musical works are not abstract objects, under the assumption that the modal inflexibility of abstract is incompatible with the modal flexibility of musical works. However, it will be shown that this argument does not constitute an objection against the type/token theory. It will be argued that the strategy to be adopted by the type/token theorist requires the adoption of the counterpart theory and Lewis’ contextualism about our modal talk. This strategy will show that the modal inflexibility of abstract objects is not incompatible with the modal flexibility of musical works, and thus that we are free of assigning to musical works the ontological category of types.

Robert Vinten: Wittgenstein and Freedom of the Will

At first glance it may seem as though Wittgenstein’s philosophical remarks are of no relevance to political philosophy. He does not engage in debates about major political ideologies in his later work and his work has no clear ideological implications. However, in this paper I will argue that his grammatical remarks about psychological concepts as well as his remarks about philosophical methodology can help to dissolve conceptual problems that are clearly relevant to political philosophy. My focus will be on Patricia Churchland and Christopher Suhler’s paper ‘Control: conscious and otherwise’, where they formulate what they think of as a neurobiological account of control. They do so in an attempt to tackle problems about the extent to which we ought to hold people responsible in cases where they are not conscious of the way in which circumstances affect their choices. Some philosophers and cognitive scientists have argued that empirical research shows that circumstances have such a large impact on people’s choices that we ought to say that a person’s control over what they do in many cases is very limited. Given the lack of control we ought not to hold people responsible for their actions to the extent that we do. This is known as the ‘Frail Control hypothesis’ and Churchland and Suhler think that their account of control undermines it.

The debate clearly has implications concerning questions of justice in society; implications concerning the way in which we ought to hold people accountable for the things they do. It is also clearly a version of old problems about freedom of the will. Wittgenstein’s philosophical remarks can help clarify the terms in which the debate is conducted and to untangle some of the conceptual confusions involved. Churchland and Suhler are right to challenge the Frail Control hypothesis and some of their conclusions are correct. However, the arguments they use to get to their conclusions are confused in various ways. The aim of the paper is to suggest that Wittgenstein’s remarks can help us to dissolve confusions surrounding problems about freedom of the will; help us to achieve clarity. A better understanding of Wittgenstein’s philosophy can help us achieve a better understanding of political philosophy.

Simon Wimmer: Knowledge-first belief: counterfactuals and variables

What is it to believe a proposition? Several recent, interesting additions to the literature answer this question by appeal to knowledge. Here I focus on an issue concerning how such knowledge-first accounts of belief should be formulated. Various such accounts, for instance those suggested in Hyman (2017), Nagel (2017), and Williamson (2000), feature a would-counterfactual and appeal to one proposition variable only. I argue that this combination of features is problematic. §3 argues that, since the suggested accounts feature would-counterfactuals, we can generate two kinds of counterexamples to them. And §4 argues that the most promising response to those counterexamples is to appeal to two proposition variables with different domains, rather than one. So, it’s time to revise the accounts suggested in Hyman (2017), Nagel (2017), and Williamson (2000).

Thomas J. Breed: Inferentialism in Context

Over the last 30 years, Charles Travis has been waging a full-scale attack on orthodox semantics, arguing that the meaning of a sentence is not, and does not determine, truth-conditions. His arguments are driven by evocative and memorable examples—Travis cases—that purport to show that a seemingly unambiguous sentence can be used by a speaker to describe the very same state of affairs on two different occasions and, due to variations in the wider conversational context, say something true on one occasion and false on the other. These examples get their bite from the fact that the apparent context-sensitivity exhibited in the example sentences does not seem to derive from the presence of a familiar context-sensitive expression. Such cases have therefore been used by Travis and others to argue that natural language exhibits widespread and seemingly unsystematic context-sensitivity. In this paper, I present and defend an inferentialist explanation of this phenomenon. The central semantic thesis of inferentialism is that the meaning of a sentence is given by citing the role that sentence plays as a premise and conclusion in inferences. However, what was actually said by an assertion of that sentence—its inferential significance—depends not just on the sentence’s meaning but also on the collateral premises used to determine its inferential consequences. I argue that the conversational context makes salient a set of collateral premises relative to which the inferential consequences of an assertion are to be determined, and that this feature of inferentialism can be used to explain Travis cases.