In my dissertation, I examine the best arguments for external world skepticism. This project is important because in their discussions of arguments for external world skepticism, philosophers often focus on one or two particular arguments without taking the time to formulate the best versions of such arguments or to consider the full range of them. To maintain focus, I limit my discussion to arguments that target at least one of the following six epistemic properties: one's belief being probable, the truth of one's belief not being due to luck, the truth of one's belief being creditable to one's cognitive faculties, one's belief being sensitive, one's belief being deontologically justified, and one's knowing. Ultimately, I conclude that the best arguments for external world skepticism are underdetermination arguments and lottery arguments. More tentatively, I also conclude that the best arguments for external world skepticism target properties that are internal (i.e. they supervene on one’s phenomenal states) or complete (i.e. they are held to the highest degree possible.)

Here is a breakdown of the dissertation chapter by chapter. In my first chapter, I introduce the six types of epistemic property I will be discussing in this thesis. I also offer some support for the claim that each is necessary for knowledge. In the course of doing so, I make the case that insofar as the properties are necessary for knowledge, they are best understood as conditional properties. For instance, consider the following two properties: (i) a belief’s being probable (ii) a belief’s being probable given one’s evidence. Arguably the latter is more plausibly a necessary condition on knowledge. Keeping this point in mind allows one to get around a number of potential counterexamples to claims that various properties are necessary for knowledge. In addition, I make two distinctions that affect the six properties, viz that they each come in internal and external varieties, and likewise they come in complete and incomplete varieties.

In my second chapter, I first say what counts as an argument for external world skepticism and discuss how I will be classifying arguments and evaluating argument strength. Next, I introduce the four types of argument for external world skepticism that I will be discussing in subsequent chapters: regress, underdetermination, lottery and linking. Next, I argue that every other type of argument for external world skepticism either fails to be plausible or is equivalent to one of these four types of argument. This involves an examination of a number of skeptical arguments, both contemporary and historical.

In my third chapter, I discuss regress arguments. In particular, I examine a number of particular regress arguments and conclude that each is either weak or equivalent to an underdetermination argument. In the course of doing so, I make a distinction between higher-order and first-order regress arguments and argue that many of the most important regress arguments are higher-order. I then argue that all higher-order regress arguments are weak -- in particular, that the premise that there can be no circular chains is false. Failure to see this has resulted from a failure to distinguish between higher-order and first-order regress arguments. Finally, I argue that first order regress arguments are either weak or equivalent to underdetermination arguments.

In my fourth chapter I discuss underdetermination arguments. My main conclusion is that there are strong underdetermination arguments for each of the six epistemic properties. In the course of arguing for this, I offer an extended argument that there are strong underdetermination arguments that target knowledge via targeting deontological justification. A second, more tentative conclusion, is that the strongest underdetermination arguments are first-order.

In my fifth chapter I discuss lottery arguments. My main conclusion here is that these arguments are successful in showing that one lacks certain epistemic properties. In the course of doing so, I examine in detail certain arguments dealing with quantum phenomena and criticize attempts by David Lewis and Timothy Williamson, among others, to deny their conclusions. I also have a discussion of the threshold problem and attempts by fallibilists to solve it, arguing that their attempts are unsuccessful. In addition, I offer a tentative argument that the strongest lottery and underdetermination arguments target properties that are internal or complete.

In my sixth chapter I discuss linking arguments. My main conclusion is that each linking argument is either weaker than or equivalent to an underdetermination or lottery argument. This is significant because much energy has been focused on linking arguments -- in particular, on an argument called “the argument from ignorance.’’ But if my conclusion is right, that energy would be better focused elsewhere.