Movatterモバイル変換

1.M. Kaplan 1985 and Dutant 2015 are helpful in laying outthe history here.

2. Hazlett takes this to motivate divorcing semantic considerationsabout the verb “to know” from knowledge, the state oftraditional epistemic interest. Even though “knows” is,according to Hazlett, not a factive verb, even Hazlett accepts thatknowledge itself is a state that can only obtain if its content istrue.

3. One possible objection to the truth condition (noted by NicholasMaxwell) is that we seem to have knowledge of false empiricaltheories, like Newtonian Physics. But this doesn’t show thatknowledge doesn’t require truth. Remember, the JTB theory is atheory ofpropositional knowledge, not knowledge byacquaintance. One might know Newtonian Physics (just as one might knowIsaac Newton)—it is much more dubious that one could know thatNewtonian Physicsis true.

4. Gettier 1963 considered three slightly different tripartite analyses,which differed in how the belief condition was articulated; one wasgiven in terms of “believes”, one in terms of“accepts”, and one in terms of “is sure that”.Epistemologists have tended to follow Gettier in treating these aseffectively equivalent. For further discussion, see McGlynn 2014:25–9 and Ichikawa forthcoming-a: §7.1

5. This is not necessarily to say that the subject must have engaged inanactivity of justifying, or attempted to show thatpis true. Rather, what the justification condition requires is merelythat a belief that qualifies as knowledge have thepropertyof being justified. It can have that property even ifS did notengage in the activity of justifying her belief thatp.Consider an ordinary person’s belief that five and five is ten.Most people have never attempted to justify this belief, and probablywould be at a loss as to how to go about justifying it. But for mostpeople, that belief would qualify as an instance of knowledge. Theimportance of the distinction between the activity of justifying and abelief’s property of being justified is emphasized by WilliamAlston in the following passage:

To turn to justification, the first point is that we will be workingwith the concept of a subjectS’sbeing justified inbelieving that p, rather than with the concept ofS’sjustifying a belief. That is, we will beconcerned with the state or condition of being justified, rather thanwith theactivity of justifying a belief. These concepts aresometimes conflated in the literature. The crucial difference betweenthem is that while to justify a belief is to marshal considerations inits support, in order for one to be justified in believing thatp it is not necessary that one have done anything by way of anargument forp or for one’s epistemic situationvis-à-visp. Unless one is justified in many beliefswithout arguing for them, there is precious little one can justifiablybelieve. (Alston 1991: 71)

For an alternative view, see Almeder 1999: 90 and 123. Almeder defendsthe view that,

as a matter of ordinary discourse, “being justified” isnot something we can always separate from the activity of giving, orbeing able to give, reasons when the question “How do youknow?” is appropriately asked. (Ibid: 92)

6.See Feldman & Conee 1985 and Conee & Feldman 2004 for ageneral presentation of their “evidentialist” view; seeFeldman & Conee 2001 for a distinctive focus on theirinternalism. For criticisms of evidentialism, see DeRose 2000, andPlantinga 1996a: 358–361, and the essays of Dougherty 2011; forcriticisms of internalism more generally, see Goldman 1999 &2009b.

7. This is a simplified statement of reliabilism; a more precise onedistinguishes conditionally reliable mechanisms (like inference) fromunconditionally reliable ones (like perception). See Goldman 1979.

8. Finer distinctions are sometimes drawn; for example, one mightconsiderex ante justification (in a position to have ajustified belief) as distinct from both doxastic (having a justifiedbelief) and propositional (having reason to believe) justification.For one such motivation, see Ichikawa & Jarvis 2013: 162–4.For another example, Lowy 1978 articulates a notion ofpersonal justification that is importantly distinct fromdoxastic justification.

9. This talk of “sufficient reason” is intended to beconsistent with the idea that in some cases, no reasons are needed tosuffice for propositional justification—that some justificationcomes “for free”, not dependent on the possession of anyreasons at all. See e.g., Lyons 2009, Wright 2004. In such cases, thenull set of reasons constitutes “sufficient” reason.

10. Externalists about doxastic justification, of course, will think thatwhether a belief is in the relevant sense “appropriate” issettled in part by external factors.

11. See Chisholm 1977, chapter 6. Chisholm’s strategy of building adegettiering clause into the justification condition is difficult tounderstand, given Chisholm’s deep commitment to internalism.Since degettiering is an external matter, this strategy makesjustification an external property.

12. See, for example, Armstrong 1973: 152 and Clark 1963. For furtherreferences, see Shope 1983: 24. This monograph provides acomprehensive discussion of the Gettier literature up to 1980. For ashorter helpful discussion of the Gettier problem, see the Appendix inPollock 1986.

13. This case is similar to Chisholm’s (1989) “sheep on ahill” case.

14. More sophisticated articulations of a sensitivity condition willrelativize it to bases or methods. See Nozick 1981: 179. The argumentagainst sensitivity given in the main text should apply equally wellto these subtler formulations.

15.The term “abominable conjunction” is from DeRose 1995:27–28.

16. Note that a sensitivity condition, being only a necessary conditionon knowledge, does not itself imply the nonskeptical claim. A skepticcould commit to a sensitivity condition without admitting abominableconjunctions.

17.See Williamson 2000: ch. 7 for a more detailed discussion.

18. This is one of several formulations of safety Sosa offers.

19. Ichikawa 2011 offers a semantics for counterfactuals in which,pace Sosa, sensitivity and safetyareequivalent.

20. SupposeS truly believes thatp. Then the uniquelynearest possible world in whichS believesp is theactual world; by stipulation,p is true in that world.Therefore, ifS were to believe thatp,p wouldnot be false.

21. Some contemporary contextualists, including Lewis 1996, Schaffer2004, Blome-Tillmann 2009b, and Ichikawa forthcoming-a, are naturally thoughtof as endorsing versions of the relevant alternatives theory (wherewhich alternatives are “relevant” for a given use of“knows” depends on the context of utterance). See§13.

22. E.g., Goldman 1967, 1976; Armstrong 1973; Dretske 1981. See alsoSturgeon 1993. Kornblith (2008, 7) suggests that the decline inattention for such views is attributable to Goldman’s 1979 shiftfrom his previous position to one in which knowledge requiresjustification, where the latter is understood as reliability, butquestions the motivation for Goldman’s move.

23. For examples of reliabilist approaches to knowledge, see: Armstrong1973, Dretske 1981, and Nozick 1981.

24. Note that process reliabilism is an example of a view that requires a“causal condition” in a weaker sense: features of theprocess that caused the belief are important for justification. As theterm is used here, “causal theories” require a causalconnection between the belief and the truth of the propositionbelieved.

25. Gaṅgeśa Upādhyāya articulated an early versionof the causal theory in the 14^th century CE. Phillips &Ramanuja Tatacharya 2004.

26. Dretske’s own information-theoretic account of knowledge, givenin ch. 4 of his 1981, is more complicated than Simple K-Reliabilism.Dretske claims (1981: 97) that his view avoids Gettier problems, butthis is controversial. For some discussion of what Dretske would sayabout the barn facades case, see Dretske 2005: 24, note 4.

27. For an example of a reliability condition amended with an eye towardthe Gettier problem, see Goldman 1976 and Goldman 1986:46–7.

28. Scott Sturgeon 1993 madea similar observation, pointing out that any non-conclusivejustification will permit Gettier cases. He endorsed an externalistconception of justification that entailed truth, defending a simpler“JB” theoryof knowledge.

29. Whether it also entails the third will depend on one’sparticular theory of justification. Alternatively, another analysis inthis spirit, following the lead of the views discussed in§6, could simply omit the justification condition.

30. This example was first brought to the attention of one of the authors(Jonathan Ichikawa) by Joshua Schechter, who suggested it may be dueto Gilbert Harman.

31. For more on distinct notions of epistemic luck, see Pritchard 2005,Steglich-Petersen 2010, and McKinnon 2013.

32. Sosa characterizes the relevant skill in reliabilistic terms; abelief is adroit if and only if it is produced by a mechanism thattends to produce true, rather than false, beliefs. But the AAA model,and the corresponding account of knowledge, does not obviously dependon this reliabilism. One could characterize skill differently, andstill define knowledge in terms of skill as Sosa does.

33. This is how Sosa characterizes a particular kind of knowledge, whichhe calls “animal knowledge”, distinguishing another, moreambitious state of “reflective knowledge”. See Sosa 2007:24.

34. Alan Millar seems to be defending one or both of these strategies inPritchard, Millar, & Haddock 2010: 129–30. Greco (2009)seems to be defending the first.

35. Sutton 2007 and Bird 2007 each offer an approach to justificationgiven in terms of knowledge.

36. In favour of something like the knowledge-first stance, seeWilliamson 2000, Nagel 2013, Miracchi 2015, or Ichikawa forthcoming-a;against, see McGlynn 2014 and many of the essays in Greenough &Pritchard 2009.

37. The term “pragmatic encroachment” was first used by JonKvanvig on the Certain Doubts weblog:http://certaindoubts.com/?p=13 (“Pragmatic aspects of knowledge?” June 12, 2004)

38. Owens (2000) gives an early defense of pragmatic encroachment; seehis 2000: 29 et seq. Note that talk of “sameness of epistemicposition” must be understood carefully. There is a sense ofcourse in whichknowing that p is part of one’sepistemic position; when we say that Sandra and Daniel are in the sameepistemic position, this cannot be the sense that we intend. Onemight—and some epistemologists do—say that Sandra andDaniel have the same (or equivalent) evidence. This will suffice onlyif pragmatic encroachment does not extend to evidence as well; it willnot if, for example, Williamson (2000) is right that a subject’sevidence consists in one’s knowledge, or if Stanley (2005: 124)is right that encroachment extends to all interesting epistemicnotions.