Copyright by Ivan J Hu 2015 The Dissertation Committee for Ivan J Hu certifies that this is the approved version of the following dissertation: Epistemicism Committee: ____________________________________ Richard M. Sainsbury, Supervisor ____________________________________ Johan A. Kamp, Co-Supervisor ____________________________________ Daniel A. Bonevac ____________________________________ Robert C. Koons ____________________________________ Joshua Dever ____________________________________ Diana Raffman Epistemicism by Ivan J Hu, B.A. DISSERTATION Presented to the Faculty of the Graduate School of the University of Texas at Austin in Partial Fulfillment of the Requirements for the Degree of DOCTOR OF PHILOSOPHY The University of Texas at Austin May 2015 ! For my parents Acknowledgements The greater part of this work grew out of many memorable hours of rich and lively conversation with Daniel Bonevac and Hans Kamp. I am indebted to them both. My best times of productivity regularly came from talking with Dan. Over these years, the extent to which I have benefited from those moments—learning from his intellectual creativity, sheer breadth of knowledge, and contagious enthusiasm when discussing all things philosophical and more—is beyond repayment. My debt to Hans is equally deep. His uncanny ability to quickly get to the heart of the issue when discussing any philosophical matter, with all seriousness and rigor yet always with a constructive attitude (a rare quality), set for me an example of true profundity tempered with intellectual humility. I am privileged to have a dissertation committee consisting of so many teachers whose own contributions to the philosophy of vagueness are each noteworthy in their own right. Much of my thinking on vagueness is owed to reading Mark Sainsbury’s writings on the subject. His insightful feedback has always challenged me to fine tune my own work while achieving more in less through leaner economy of expression. I owe much to Josh Dever, whose support and availability proved invaluable on many occasions and whose brilliance in pedagogy never ceases to amaze. Diana Raffman was kind enough to join my committee at a late stage, though I had already learned much from reading her work on vagueness in previous years. Rob Koons was the first professor I met as a graduate student. I am grateful for all the advice and encouragement he has given me over the years, especially regarding the importance of knowing one’s own broader philosophical outlook. v I have also enjoyed many discussions about vagueness with other faculty members not on my committee, including David Sosa, Cory Juhl, Ray Buchanan, Galen Strawson, and Jon Litland. Thanks also to my fellow graduate student colleagues who have offered comments and feedback related to this work: Fatema Amijee, Kate Ritchie, Richard Davis, and Brian Cutter. I should also thank John Campbell, whose supervision during my undergraduate years did so much to sharpen my philosophical writing and instill within me the discipline of being thorough. I probably would not have pursued scholarship in philosophy, or even the greater humanities, were it not for Gary Hendrickson, whose encouragement in high school was pivotal when I first encountered philosophy through literature. For my dearest family and friends: I thank my parents for being among the few Chinese parents on this earth who would ever let their son major, never mind pursue a doctorate, in the humanities. Their love and support over the years in allowing me to find my own path has been invaluable. Thanks to my Berkeley CKG1 peers for believing in me, especially Brian Chang. Thanks to Greg and Mary Jane Grooms, whose home has been a haven of constancy and rest during this otherwise often disjointed journey of graduate school. Finally, thanks to Janie for supporting me while I got this done, through both sacrifice and celebration. Ivan J Hu The University of Texas at Austin May 2015 vi Preface Back in 2011, while reading some of the latest on vagueness—by David Barnett (2009), Brian Weatherson (2006), Patrick Greenough (2003), Cian Dorr (2003), Crispin Wright (2001)—I was shocked to discover that people were denying what seemed the most plausible part of epistemicism: the claim that vagueness entails ignorance, or what I call the UNKNOWN principle—e.g. if it’s vague whether somebody is bald, you don’t know whether he’s bald. Such objections went either undiscussed or unnoticed. Evidently no one deemed it worthwhile to defend the epistemicist theory, even against criticisms that (I felt) were surely wrong. So I took the task upon myself. What emerged was my prospectus thesis: a longer, unwieldier version of Chapter 1 (“Epistemic Gaps”) written in response to Barnett’s (2010) claim that UNKNOWN fails to capture anything essential about vagueness: in a hypothetical scenario where our cognitive faculties were improved so as to enable knowledge, vagueness might still be possible, since people’s intuitions could still be vague over vague matters; it just wouldn’t entail ignorance. It took 70 pages (including eight whole pages of endnotes) to respond to Barnett’s one counterexample. The reason being: it took quite some work to figure out the logical structure underlying Barnett’s arguments. It wasn’t trivial, for instance, spelling out how the determinacy (clarity) operator used to express claims of vagueness was supposed to interact with a knowledge operator and with conditionals. But that’s what’s exactly at issue because UNKNOWN involves all three. What I did was take some of the methods of reasoning about vague claims, which I had picked up from Greenough and Barnett, and apply these exhaustively to a whole range of claims—response-dependence conditionals (if it seems that p, one can know that p), ambivalence (if it both vaguely seems that p and vaguely seems that ¬p), intuition-completeness (if p, it will vii seem that p), connections between clarity and knowledge (if p is knowable, p is clear, or vice versa)—tracing out all their possible interrelations, not just when things were vague, but (more crucially) when things were higher-order vague. What I discovered was that you can derive quite a lot just by using the minimal resources of a KT modal logic for determinacy (clarity) and knowledge. For instance, the type of response-dependence Barnett wanted, where one is never quite determinately ignorant about vague matters, can be shown to be compatible with UNKNOWN and any broader encompassing epistemicist theory. It merely imposes an extra constraint: you never have definite borderline cases. That’s enough to get you the same results, without meddling with the intuitive principles already in place linking vagueness and ignorance. I took this as evidence for Greenough’s (2003) fundamental insight that some principle connecting vagueness to unknowability—contrary to the skeptical attitude of recent work—really could serve as a minimal theory of vagueness, or something like it. This was the seemingly undeniable part of epistemicism. To that extent, challenging epistemicism on epistemic grounds struck me as unthinkable. I found no reason to believe that any of the objections to UNKNOWN posed by Wright (2001), Dorr (2003), or Barnett (2010) were insurmountable. For their alternative proposals for what the relation between vagueness and unknowability should look like instead all represented variants of the same general type of view concerning the epistemology of vagueness, the overall structure of which was something that seemed easily recoverable on epistemicism—one merely had to deny the existence of knowable or definite borderline cases. I later came to think there were independent reasons to question, on epistemic grounds, this entire family of views advanced by Wright, Dorr, Barnett, and others. In Chapter 2 (“Vague Unknowns”), I go through each of their alternative proposals one by one and show why there’s independent reason to find them implausible. What’s more, I came to find what Williamson himself had to say about the relations between vagueness and ignorance deeply unsatisfactory, or viii downright puzzling at best. Since then I’ve come to doubt whether Williamson’s account should even be taken to be representative of the epistemicist perspective on vagueness at all (see Chapters 2 and 4; see also discussion in Outlook). That said, I should pause here to pay my intellectual respects. It was while reading Timothy Williamson’s Vagueness (1994) in a seminar on vagueness co-taught by Hans Kamp and Mark Sainsbury in the spring of 2010 that I realized I had to work on vagueness. No other philosophical topic held so great an aura of enigma, for so ostensibly simple a problem as the Sorites paradox. Williamson’s book made this salient. The unparalleled rigor and clarity found in his discussion of the subject was infectious. The first paper I ever wrote on vagueness was for that seminar: an embryonic version of Chapter 4 (“Margins for Error in Meaning”). I wrote it five years ago, while still very much within a Williamsonian frame of mind and earnestly trying to press out the wrinkles in his theory. Although I’ve since then come to think there are some fundamental, irremediable differences in my approach in addressing the epistemology of vagueness or even what it means to be an epistemicist theory, I still find it valuable to confront Williamson’s theory on its own terms, if not because it raises a host of fundamental questions about the epistemology of meaning. My general line of argument there is this. There’s a real worry about whether the margin for error principles, so crucial to Williamson’s explanation of vagueness-related ignorance, won’t also rule out semantic knowledge, if the meanings of our vague expressions are themselves in constant flux. Even Sainsbury in his illuminating (1997) comment on Vagueness tries to come to the rescue by developing Williamson’s externalist idea of semantic knowledge as linguistic induction.