Jonathan Vogel

The chapter takes structuralism to be the thesis that if F and G are alike causally, then F and G are the same property. It follows that our beliefs about the world can be true in various brain-in-a-vat scenarios, giving us (some) refuge from skeptical arguments. The trouble is that structuralism doesn’t do justice to certain metaphysical aspects of property identity having to do with fundamentality, intrinsicality, and the unity of the world. A closely related point is that the relation…lies-at-some-spatial-distance-from…obeys necessary truths that need not apply to other relations with the same causal profile. This observation is especially important if, as David Lewis argued, the only alternatives to skepticism are structuralism and an anti-Humean stance toward modality. Some pertinent views of David Chalmers’s are discussed, and parallels are drawn between the structuralist response to skepticism and functionalism in the philosophy of mind.

Suppose you have evidence E for H. What reason do you have for believing that your evidence isn’t misleading? That is, what reason do you have for believing (E & H)? Two very plausible, related principles imply that E itself can’t provide empirical justification for believing (E & H). The _Entailment Principle_ says that if Y entails X, X can’t justify Y. The _Confirmation Principle_ says that X can’t justify Y unless X raises the probability of Y. The chapter argues that E can indeed justify (E & H), and that both principles are false. Further conclusions are: Epistemic closure withstands recent criticisms due to Fred Dretske; we don’t have a priori reasons for rejecting (E & H); the dogmatist reply to scepticism is unscathed by a challenge posed by Roger White; and there is a promising response to the New Riddle of Induction.

I explore and develop the idea, due to Peter Unger, that knowledge is non-accidentally true belief. Non-accidental truth is different from the absence of epistemic luck, as discussed by Pritchard. The original analysis faces two counterexamples, the Meson Case and the Light Switch Case. The former concerns knowledge of nomological necessities; the latter turns on the direction-of-fit between a belief and the facts. I propose: (ENA) S knows that P when S’s belief that P is non-accidentally true because (i) it is based on good evidence, and (ii) in and of themselves, beliefs based on good evidence tend to be true. ENA gets the two examples right, and compares favorably with safety, defeasibility, and knowledge-as-credit accounts. Lackey has claimed that the credit-based approach mishandles knowledge via testimony. I critically examine her objection, and show that ENA faces no difficulty of that sort.

Philosophical understanding of the problem of skepticism and its sources has grown in recent years, but important questions remain about the contribution of underdetermination and closure principles to skeptical arguments. My aim here is to improve upon this situation. Sections _1, 2 compare a closure principle I call Weakening_ to a principle I call _Underdetermination_. It appears that the former doesn’t follow from, and is less plausible than, the latter. Section _3 examines Dretske’s Zebra Case as a putative counterexample to Weakening. The New Zebra Case, a variant of the original, indicates that Weakening is preserved after all. Sections 4, 5 establish that Weakening and Underdetermination converge in a canonical skeptical argument, the Misleading Evidence Argument_. Such convergence is significant, because it is widely thought that abandoning Weakening allows us to escape skepticism. Section 6 takes up Martin Smith’s claim that underdetermination principles don’t underwrite a skeptical argument of any interest. I argue that this assessment is mistaken.

Important conceptual and metaphysical issues arise when we try to understand the mental lives of “split-brain” subjects. How many distinct streams of consciousness do they have? According to Elizabeth Schechter’s partial unity model, the answer is one. A related question is whether co-consciouness, in general, is transitive. That is, if α and β are co-conscious experiences, and β and γ are co-conscious experiences, must α and γ be co-conscious? According to Schechter, the answer is no. The partial unity model faces some serious objections. Its underpinnings are suspect, and ways of working out the model fall into trouble of various sorts: incoherence, emptiness, or unacceptable indeterminacy about the identity of (token) conscious states. One apparent lesson is that a subject can host two distinct conscious experiences of the same sort. Another is that the co-consciousness relation is transitive after all.

A prominent response is that Cartesian skepticism is too Cartesian. It arises from outmoded views in epistemology and the philosophy of mind that we now properly reject. We can and should move on to other things. §2 takes up three broadly Cartesian themes: the epistemic priority of experience, under-determination, and the representative theory of perception. I challenge some common assumptions about these, and their connection to skepticism. §3 shows how skeptical arguments that emphasize causal considerations can avoid some suspect Cartesian commitments. §4 explores the related idea that perceptual knowledge is what I call instrumental knowledge. Drawing on that connection, §5 examines a skeptical argument based on minimal assumptions, and explains why a Moorean response to it is unsatisfactory. The problem of skepticism is still with us, and it requires a substantive solution

The chapter takes structuralism to be the thesis that if F and G are alike causally, then F and G are the same property. It follows that our beliefs about the world can be true in various brain-in-a-vat scenarios, giving us refuge from skeptical arguments. The trouble is that structuralism doesn’t do justice to certain metaphysical aspects of property identity having to do with fundamentality, intrinsicality, and the unity of the world. A closely related point is that the relation…lies-at-some-spatial-distance-from…obeys necessary truths that need not apply to other relations with the same causal profile. This observation is especially important if, as David Lewis argued, the only alternatives to skepticism are structuralism and an anti-Humean stance toward modality. Some pertinent views of David Chalmers’s are discussed, and parallels are drawn between the structuralist response to skepticism and functionalism in the philosophy of mind.

This remarkably clear and comprehensive account of empirical knowledge will be valuable to all students of epistemology and philosophy. The author begins from an explanationist analysis of knowing—a belief counts as knowledge if, and only if, its truth enters into the best explanation for its being held. Defending common sense and scientific realism within the explanationist framework, Alan Goldman provides a new foundational approach to justification. The view that emerges is broadly empiricist, counteracting the recently dominant trend that rejects that framework entirely. Topics treated include the Gettier problem, the nature of explanation and inductive inference, the justification of foundations for knowledge in terms of inference to the best explanation, the possibility of realist interpretations of contemporary science, reference, and the relations between empirical psychology and epistemology. Professor Goldman defends the need for a foundational theory of justification and presents a version that refutes standard criticisms of that doctrine. His defense of realism takes into account contemporary advances in semantics and philosophy of science. It attempts to clarify the kinds of skeptical argument the philosopher must take seriously, without succumbing to them. While recent epistemology has tended to dismiss the traditional foundational approach, it has not provided a suitable alternative. Goldman breaks new ground by adapting that approach within his explanationist, inductive theory.

I explore and develop the idea that (NA) knowledge is non-accidentally true belief. The applicable notion of non-accidentality differs from that of ‘epistemic luck’ discussed by Pritchard. Safety theories may be seen as a refinement of, or substitute for, NA but they are subject to a fundamental difficulty. At the same time, NA needs to be adjusted in order to cope with two counterexamples. The Light Switch Case turns on the ‘directionof-fit’ between a belief and the facts, while the Meson Case concerns knowledge of nomological necessities. A proposed revision to NA is: (ENA) S knows that P when S’s belief that P is non-accidentally true because (i) it is based on good evidence, and (ii) in and of themselves, beliefs based on good evidence tend to be true. ‘knowledge-as-credit’ accounts have been offered as an alternative way of sharpening NA. I argue that such accounts face serious objections that don’t apply to ENA. Finally, I take up Lackey’s worry that the credit-based approach mishandles knowledge via testimony. ENA faces no difficulty on that score, which counts in its favor.

This dissertation begins with a general discussion of the role of epistemic principles in arguments for and against Cartesian skepticism . The skeptic may be viewed as trying to establish that, according to non-arbitrary epistemic principles we ordinarily accept, we have no knowledge of the external world. So construed, the skeptic's challenge cannot be dismissed. It can, however, be refuted, particularly if the epistemic principles invoked by the skeptic prove to be invalid. ;One epistemic principle is that of the closure of knowledge under known logical implication: that if a person knows A, and knows that A entails B, then that person knows, or can know, B. ;The validity of the closure principle has been challenged recently by Fred Dretske and Robert Nozick. Dretske's case against the principle rests on some apparent counterexamples. I maintain that these examples don't have the force Dretske thinks, and that other, even stronger examples fail to undermine the closure principle. Nozick bases his criticism of the closure principle on a plausible assumption about the conditions for knowledge. I argue that this assumption is deeply flawed, especially in light of certain facts about inductive knowledge. ;I next consider and defend the claim that justification is closed under known logical implication. On the assumption that knowledge does require justification, this result supports two further conclusions: that knowledge itself is likely to obey closure, and that, even if it doesn't, the skeptic could still proceed with his argument unimpeded. In connection with these issues, I discuss some relevant views of Peter Klein, and argue that his attempted refutation of skepticism isn't fully successful. ;A principal result is that rejection of the closure principle doesn't provide a legitimate way to avoid Cartesian skepticism. The discussion also indicates that a convincing refutation of the skeptic may be found in an appeal to principles of inference to the best explanation. I distinguish various skeptical hypotheses and show that they may be rejected on grounds of explanatory inadequacy. When faced with this sort of response, Cartesian skepticism either becomes untenable or collapses into skepticism about induction.

For some time, it seemed that one had to choose between two sharply different theories of epistemic justification, foundationalism and coherentism. Foundationalists typically held that some beliefs were certain, and, hence, basic. Basic beliefs could impart justification to other, non-basic beliefs, but needed no such support themselves. Coherentists denied that there are any basic beliefs; on their view, all justified beliefs require support from other beliefs. The divide between foundationalism and coherentism has narrowed lately, and Susan Haack attempts to synthesize these competing accounts into a view she calls "foundherentism."

Subjunctivitis is the doctrine that what is distinctive about knowledge is essential modal in character, and thus is captured by certain subjunctive conditionals. One principal formulation of subjunctivism invokes a ``sensitivity condition'' (Nozick, De Rose), the other invokes a ``safety condition'' (Sosa). It is shown in detail how defects in the sensitivity condition generate unwanted results, and that the virtues of that condition are merely apparent. The safety condition is untenable also, because it is too easily satisfied. A powerful motivation for adopting subjunctivism would be that it provides a solution to the problem of misleading evidence, but in fact, it does not.

Laurence BonJour, among others, has argued that inference to the best explanation allows us to reject skeptical hypotheses in favor of our common-sense view of the world. BonJour considers several skeptical hypotheses, specifically: our experiences arise by mere chance, uncaused; the simple hypothesis which states merely that our experiences are caused unveridically; and an elaborated hypothesis which explains in detail how our unveridical experiences are brought about. A central issue is whether the coherence of one’s experience makes that experience more likely to be veridical. BonJour’s recent treatment of “analog” and “digital” skeptical hypotheses is also discussed. I argue that, although there are important lessons to be learned from BonJour’s writings, his use of inference to the best explanation against skepticism is unsuccessful.Keywords: BonJour; Explanation; Inference to the best explanation; Skepticism; Confirmation; Probability; Coherence; External world

In an important study, Lawrence Sklar has defended a doctrine of methodological conservatism (very roughly, the principle that a proposition derives some sort of epistemic warrant from being believed). I argue that Sklar's careful formulation of methodological conservatism remains too strong, and that a yet weaker version of the doctrine cannot be successfully defended. I also criticize Sklar's argument that the rejection of methodological conservatism would result in total skepticism. Finally, I turn to a closely related issue, and try to show why different people with the same evidence should be justified in believing the same things.