This dissertation provides a new semantics for first-order modal logic. It is philosophically motivated by the epistemic reading of modal operators and, in particular, three desiderata in the analysis of epistemic modalities. (i) The semantic modelling of epistemic modalities, in particular verifiability and falsifiability, cannot be properly achieved by Kripke's relational notion of accessibility. It requires instead a more general, topological notion of accessibility. (ii) Also, the epistemic reading of modal operators seems to require that we combine modal logic with fully classical first-order logic. For this purpose, however, Kripke's semantics for quantified modal logic is inadequate; its logic is free logic as opposed to classical logic. (iii) More importantly, Kripke's semantics comes with a restriction that is too strong to let us semantically express, for instance, that the identity of Hesperus and Phosphorus, even if metaphysically necessary, can still be a matter of epistemic discovery. To provide a semantics that accommodates the three desiderata, I show, on the one hand, how the desideratum (i) can be achieved with topological semantics, and more generally neighborhood semantics, for propositional modal logic. On the other hand, to achieve (ii) and (iii), it turns out that David Lewis's counterpart theory is helpful at least technically. Even though Lewis's own formulation is too liberal---in contrast to Kripke's being too restrictive---to achieve our goals, this dissertation provides a unification of the two frameworks, Kripke's and Lewis's. Through a series of both formal and conceptual comparisons of their ontologies and semantic ideas, it is shown that structures called sheaves are needed to unify the ideas and achieve the desiderata (ii) and (iii). In the end, I define a category of sheaves over a neighborhood frame with certain properties, and show that it provides a semantics that naturally unifies neighborhood semantics for propositional modal logic, on the one hand, and semantics for first-order logic on the other. Completeness theorems are proved.

Most philosophers today assume what Kant might have called a 'material' conception of logic. According to the material conception, the laws of logic obtain independently of our consciousness of them, because they are either objective 'laws of truth', laws governing linguistic practices, or laws innate to our cognitive capacities. But it is often overlooked that this view of logic faces intractable difficulties in providing an adequate explanation of how these laws govern the mind. (Both rationalist and empiricist attempts to offer an explanation have been made.) The material conception immunizes logic from these problems, since it assumes that they do not concern logic, but merely concern epistemological views about what it is to have knowledge of logic. In this dissertation I argue that Kant avoids the epistemological difficulties because he has a 'formal' conception of general logic, according to which logical operations and rules articulate self-consciousness in any exercise of the understanding. That is, they are not rules or procedures for generating intellectual acts (such as judgments), nor are they products of intellectual acts. Instead, they bring to (self-) consciousness the necessity (or 'necessary synthetic unity') in the activity of the understanding itself. Logical cognition thus is not material cognition of that which is distinct from our cognition of it, but instead is formal cognition, or cognition that any act of cognition has of itself. I argue that we cannot fully appreciate these points if we assume an 'analytic approach' to Kant's logic, according to which logical operations consist in mere acts of comparison (or analysis) of representations. General logic must primarily concern itself with the understanding's acts of synthesis in cognition, acts that are directed at an inner telos or purpose (namely, systematic unity in the whole of cognitions). Kant's conception of logical form thus invokes an organic notion of 'form' that is linked to the teleological structure of our cognitive capacities.

In my dissertation I show that we cannot conceive of ourselves as embodied beings unless we know some of our physical features without observation or inference. I also argue that we have the requisite sort of self-knowledge, and that it consists in our knowledge of ourselves as intentional agents. Descartes claimed that when one is self-consciously aware of oneself, one is aware of oneself as a purely psychological being. In chapter two I argue that if his claim were correct, it would be unclear what it could mean for one to identify oneself with a human being. I then argue that self-conscious beliefs about oneself are beliefs about oneself that are not grounded on observation or inference. In chapter three I take up the task of making it plausible that we do possess the required sort of self-knowledge. I offer a novel interpretation of Anscombes thesis that we know what we are doing intentionally without observation or inference. The key lies in the Aristotelian doctrine that action itself can be the conclusion of practical reasoning. In chapter four I reply to two objections to my account. The first is an argument for volitionalism, or the thesis that events that are describable as an agents moving her body are acts of trying that occur prior to her bodily movements. In response, I argue for an alternative, according to which bodily action is a temporally extended process that is complete only when ones body has moved. The second argument begins from the premise that we can act intentionally without knowing that we are succeeding. I argue that this shows only that our self-conscious capacity to act intentionally is fallible in a certain respect. Conditions which potentially inhibit the success of ones doing such-and-such intentionally also inhibit ones capacity to know that one is doing so when the action is successful. Finally, in chapter five I defend a non-reductionistic account of intentional action in contrast to dominant reductionistic models. I conclude that an intentional action is simply an exercise of a rational agents will, described as such.

In this work, I have three related aims. First, I attempt to show that none of the popular responses to the Knowledge Argument works. The first chapter presents the Argument as it is originally proposed by Jackson and Nagel, clarifies the distinctions between two proposals, and identifies the core of the challenge it raises for physicalism. In the following two chapters, I show that no-new-knowledge and new-knowledge/old-facts responses to the Argument fail. My second aim is to develop an adequate account of phenomenal concepts. A widely held and influential idea is that phenomenal concepts, concepts under which we grasp the qualitative properties of our experiences, are epistemically and semantically special. In the fourth chapter, I provide an account that, I believe, captures the core of this special nature. On my account, phenomenal concepts are epistemically special in that our beliefs formed by the application of phenomenal concepts to our experiences are non-inferentially justified, and they are semantically special in that they are simple concepts that represent certain phenomenal properties without using any property other than those properties. In the fourth chapter, I also draw the logical implications of my account of phenomenal concepts with respect to the conceptual/non-conceptual distinctions, the Myth of the Given, and the infallibility of phenomenal beliefs. My third aim, which is addressed in chapter five, is to provide a novel argument for property dualism on the basis of my account of phenomenal concepts. Instead of asking how are phenomenal experiences possible in an entirely physical world?, I raise the question how are phenomenal concepts possible in an entirely physical world? This question accepts the invitation of our physicalist to ascend from the level of properties to the level of concepts but inquires whether such an ascent is of any help to the physicalist. I argue that ascending to higher levels only displaces the problem: phenomenal concepts are as problematic for physicalism as are phenomenal properties. I conclude that phenomenal concepts are not possible in an entirely physical world and, since we have phenomenal concepts, our world cannot be entirely physical.

I maintain that consequentialism is committed to the view that there must be a fundamental bifurcation between sorts of value possessed by inner and outer elements of any exercise of agency, and I argue that this view is false because its adherents are unable to avoid global skepticism about the possibility of agentive efficacy. This result obliges us to articulate a concept of a thing that an agent can do in which her efforts are robustly united, rather than merely conjoined, with the achievement of her goals. I find such a concept in the notion of a task as it figures in the concept of multitasking. A task is a kind of unity that subsumes everything an agent is doing in order to achieve a certain end. I argue that empiricism about practical optionsthe view according to which a practical option is just given to our practical cognition rather than being somehow the work of our practical cognitive facultiescannot account for the role that this sort of unity plays in our ordinary concept of multitasking. I also present an argument against a command picture of ethics according to which the ethical value of an exercise of agency can be determined by considering exclusively what its agent has wrought thereby, rather than necessarily by also considering the reasons from which she wrought it. Finally, I offer an alternative to this picture in the form of a series of adaptations and adjustments of Kantian moral doctrine: first, that respect for the moral law is a determinable having many determinationshonesty, benevolence, etc.and that only actions performed from one of the various motives that are respectful of the moral law possess moral worth; and second, that the moral law is also a determinable having many determinations and that an action can possess moral worth if the motive from which it is prompted is an attunement to the determination of this determinable that she bears. My most novel claim is that an agents attunement to her determination of practical rationality is secured by the activity of a faculty of practical imagination.

Most of us believe that induction and perception have some normative status that counter-induction and crystal gazing lack: the former are correct, but the latter are not. How are such beliefs about rationality justified? My dissertation examines two skeptical arguments that contend the answer is: theyre not. The first skeptical worry centers on circularity. The only defense I can give for the claim that induction will mostly lead me to true beliefs will invoke induction it will point out that induction has been reliable in the past and thus conclude (via inductive inference) that induction will be reliable in the future. Much the same applies to perception: I can give a story about why I expect it to be reliable, but only by citing perceptual beliefs. These defenses seem worryingly circular. Non-skeptical responses to this puzzle fall into two camps: Mooreans embrace the circular defenses of perception and induction; rationalists say that justification to believe that perception and induction are reliable is apriori. I defend Moorean responses to skepticism: the most plausible accounts of why the aforementioned reasoning is viciously circular fail. In addition, I argue that rationalismwhile perhaps trueis insufficient to deflect the skeptical worry. It turns out that even rationalists need to embrace Moorean circular reasoning. The second skeptical worry focuses on the etiology of our faculties of reason. There is some causal story about why I am inclined to engage in certain patterns of normative reasoning: roughly, evolution by natural selection. Selection pressures favored norms that helped our ancestors find food and show off to potential mates. A puzzle arises because correctness does not appear well-positioned to provide an adaptive edge. The correct ways of reasoning about normative matters might have aided survival, but only as a fortuitous side effect - so getting it right would be a fluke. I show that this puzzle yields a serious skeptical worry. We ought to doubt that we are trustworthy normative reasoners unless there is an explanatory connection between the normative facts and our faculties for normative reasoning.

In these studies I investigate paradigmatic empiricist accounts of three notions of moral philosophy: desire for happiness, moral approbation, and rational choice. In the first chapter I situate John Lockes account of the desire for happiness in his general account of the mental faculties. I argue that in Lockes Essay the uneasiness of desire is to be interpreted neither as a perception of an idea nor as a volition, but as an act of a separate faculty of feeling. Only if the uneasiness of desire is understood in this way, will it be possible to make sense of Lockes claim that it constantly accompanies the perception of ideas. Understanding desire as an act of feeling will also clarify what kind of knowledge of happiness Locke assumes we have when we desire happiness. In the second chapter I examine David Humes account of the origin of the sentiment of moral approbation. Hume seems to give a general empirical explanation of this sentiment; but this explanation of the origin of moral approbation faces apparent counterexamples: the approbation of what Hume calls useless or monkish virtues. I argue that Humes own treatment of these counterexamples demands a restrictive interpretation of what he labels his experimental method, and an understanding of his moral philosophy as a self-enforcing genealogy of morals. Taking as a starting point a thesis of David Gauthiers about the status of expected utility theory, I discuss in the third chapter whether an empiricist and subjectivist theory of value is compatible with an account of rational choice that leaves room for some form of autonomy. I argue that if autonomy presupposes an activity of practical reason, the maximization of expected utility cannot be the principle of rational choice. In each of these studies I attempt to bring into the open insufficiently acknowledged elements in empiricist moral philosophy: the role of non-experiential consciousness in Lockes account of the universal desire for happiness, the restriction of the experimental method in Humes genealogy of moral approbation, and the assumption of the determinacy of the notion of expected utility maximization in Gauthiers theory of rational choice.

How may we try to answer the central question of ethics, the question how one should live? Understood as concerning the good lives of rational agents qua rational, the question concerns the standards of practical reason. How may we vindicate a view about those standardsan ethical view, for short? This dissertation examines whether it is possible to vindicate an ethical view without begging any first-order normative questions against skeptics in the process. I argue that it is not. If there are sound arguments for ethical views, they must rely on premises that, while true, beg some first-order normative question against a possible skeptic. I call this thesis the autonomy of ethics. The result is that sound ethical argumentation is disturbingly partisan: sound arguments in ethics cannot be seen to be sound by anyone who does not already share the right first-order view to at least some extent. I argue for the autonomy of ethics by examining attempts to avoid it. Constitutivism seeks to ground ethics in the metaphysics of agency. Metasemantic strategies seek to ground ethics in the conditions of concept-possession, and in the implicit normative commitments that such conditions purportedly involve. Closely related metapragmatic strategies seek to ground ethics in the conditions of using concepts in judgments or in reasoning. Against each strategy, I argue that the relevant conditionsthe conditions of agency, of concept-possession, and of concept useare normatively neutral. I further argue that, given the failure of these strategies, there is no further way to avoid the autonomy of ethics. The only possible sound arguments in favor of ethical views are ethically partisan in the way outlined. One way of putting this conclusion is that there is no purely metaethical way of vindicating any ethical view. If there can nonetheless be objective truths in ethics, their possibility cannot depend on their having a purely metaethical grounding.

I argue for three main conclusions. First, we should adopt a "probability first" approach to epistemology, which takes facts about justification for outright belief to supervene on facts about rationally permissible credence distributions. Such an approach is plausible even though standard accounts that reduce belief to credence above a threshold or invariance in conditional preferences are vulnerable to intuitive counterexamples. Second, I argue that a dogmatist response to skepticism about inferential justification is false if we adopt a probability first approach. Dogmatists hold that we might gain justification to believe E -> H for the first time when we learn E (and nothing stronger). I show that only a dynamic Keynesian model is compatible with dogmatism about inferential justification. But the main virtue of the dynamic Keynesian model--it allows for learning about fundamental evidential relationships--is by no means unique to it. I conclude that a rationalist liberalism, which holds that we are independently justified in believing E -> H whenever we are inferentially justified in believing H on the basis of E, is the best anti-skeptical account of inferential justification on most probabilistic models. Finally, I argue that the compatibilist approach to the conflict between Bayesian conditionalization and inference to the best explanation (IBE) fails. However, we anyway need to impose constraints on rational credence other than conditionalization, and we should take explanatory considerations to constrain the rationally permissible prior credence distributions. I present an account of IBE such that we should give higher conditional prior credence to H, given E, when H is the most intellectually satisfying explanation of E, and defend this account against the objection that the subjectivity of intellectual satisfaction will lead to an unacceptably permissive epistemology.

Despite a growing appreciation in recent decades for the significance of the social in many areas of philosophy, most philosophers today have not adequately examined their assumptions about how human beings are fundamentally social, in particular, how they are socially constituted. This dissertation argues that the human individual is socially constituted because her very capacity to be a self and agent must draw on a shared public understanding of the interwoven practices, norms, and roles that enables her to exercise this capacity in general. In Part I of the dissertation, I explicate and adopt Philip Pettits suggestion about how to define the thesis of the social constitution of the individual and the general form that the argument for this thesis should take, even though I find Pettits own argument for this thesis to be wanting. I then consider how Martin Heideggers conception of human social existence in Being and Time when properly understood can significantly improve Pettits argument. I elaborate and defend the view that the human individual is socially constituted because she always initially and mostly shares a public understanding of the world, including of herself and her relations with others, that is (in the first instance) normalized. In Part II of the dissertation, I make explicit and criticize the dominant understanding of human sociality in many strands of contemporary philosophy. This understanding assumes (roughly speaking) that the fundamental or primary way in which human beings are social consists in modes of interpersonal interactions (IPIA). I critically engage three varieties of IPIA in contemporary philosophy: (1) prominent theories of collective intentionality; (2) Donald Davidsons conception of social interaction in successful linguistic communication and of triangulation as a necessary condition of the objectivity of thought; and (3) accounts of normativity that stem from standard communalist readings of Ludwig Wittgensteins Philosophical Investigations. I argue that these versions of IPIA are problematic not only in their own terms, but also inadequate precisely because they fail to take into account the social constitution of the individual.