Colin Radford-Knowledge by Examples

Embed Size (px)

Citation preview

  • 8/3/2019 Colin Radford-Knowledge by Examples

    1/12

    Oxford University PressThe Analysis Committeehttp://www.jstor.org/stable/3326979 .

    Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

    JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

    content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

    of scholarship. For more information about JSTOR, please contact [email protected].

    Oxford University Press and The Analysis Committee are collaborating with JSTOR to digitize, preserve and

    extend access toAnalysis.

    http://www.jstor.org

    http://www.jstor.org/action/showPublisher?publisherCode=ouphttp://www.jstor.org/action/showPublisher?publisherCode=anacomhttp://www.jstor.org/stable/3326979?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/3326979?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=anacomhttp://www.jstor.org/action/showPublisher?publisherCode=oup
  • 8/3/2019 Colin Radford-Knowledge by Examples

    2/12

  • 8/3/2019 Colin Radford-Knowledge by Examples

    3/12

    2 ANALYSISFor although Miss Tercy's lack of certainty is, perhaps, neurotic andtreated as such, it is also treated as real. She really isn't sure. It isn'tthat she is simply prone to hedge.The next example is more ambitious.

    Example 3.Tom: Right. You won the noughts-and-crosses. Now we'll have aquiz: English history.Jean: Oh! No! I don't knowany English history.Tom: Don't be silly, everyonedoes. You musthave done some at school?Jean: They don't teach English historyat French-Canadianchools.Tom: Really? Well, this will be educational or you. And it's time I wonsomething. Ready?Jean: O.K. I'll just guess. Then I'll ask you some questions on French-Canadianhistory!Tom: Yes. Well: sixpence on the first question, shilling on the second,one and six on the third and so on up to five bob?Jean: Why not?Tom: Right. Firstquestion,for sixpence:whendid William he Conquerorland in England?Jean hesitantly): Ten sixty-six.Tom: There you are!Jean: Well, well! Ten sixty-sixand all that?Tom: Yes! Yes-for a shilling: whom did he defeat and kill when helanded?Jean: Oh! Oh. (Pause.) Oh-1 don't know.Tom: It's easy. Kids' stuff.Jean: I told you.Tom: Harold! Whom had Harold himself defeatedjust before the battleof Hastings?Jean: I'm glad this isn't for real money. (Pause). Frederick?Tom: Frederick? No. Harald Hadraga. That's-ah?-two bob. Foranother two bob-um?-um? Well-moving on a bit then: When didHenrythe Eighth die?Jean pause): He had six wives?Tom: Yes.Jean: Oh, I don't know. (Pause.) Fifteenseventy-seven?Tom: Bad luck! Fifteen orty-seven. Whatabout Elizabeth?Jean: Oh! Ah . . . Elizabeth. Elizabeth. Tsst! Ooh . . . Mmm . . .Sixteen-oh-threeTom: Yes! Now tell me you haven'tdone any history!jean: No, really.Tom sarcastically):That was just a guess, was it?Jean: Well, I don't know. Perhaps I picked that up on a Shakespearecourseor somewhere. We didn'tdo all these kingsandqueens. Anyway-youowe me ... ?Tom: You owe me-four bob take away two and six-one and six. Forthreebob, when did Jamesthe First die?Jean: Jamesthe First?

  • 8/3/2019 Colin Radford-Knowledge by Examples

    4/12

    KNOWLEDGE-BY EXAMPLES 3Tom: James Stuart,the Sixth of Scotlandand the First of England. Thefirst Stuart. He came after Elizabeth.Jean: Oh . . . Ah . . . James the First. So he's sixteen-oh-three to...

    to ... sixteen-oh-three. Sixteen twenty-five?Tom: Yes! Look here, you must have done these people!Jean: Well, I certainly don't remember. As far as I can tell I'm justguessing. And don't think you're going to get me to double up or anythinglike that!Tom: I wouldn't dreamof it; I owe you one-and-six. Well-this is givingyou a chance: Charles he First?Jean: When did he get on and off this English throne?Tom: Yep. He's the next. This is for three-and-six.Jean: Ah... the next? Charles he First? So that'ssixteen... Whatwasit, sixteen twenty-five? to . . (Pause). Sixteen-oh-three,sixteen twenty-five-sixteen forty-nine?Tom: Well, I-Jean: Is that right?Tom: You wouldn't like to double your stake on that?Jean: Oh no! Is it right, though?Tom: Yes. No more clues. Um (thinks). Oh! Easy: Victoria. Whendid she ascend the throne?Jean: Ah ... Victoria? Victoria. Victoria. About eighteen twenty?Tom: What date?Jean: Eighteen twenty.Tom: Hmm. When did she die?Jean: Ah! Ah ... the VictorianAge. Um ... Eighteen ninety-eight?Tom: No. She's eighteen thirty-seven to nineteen-oh-one. You'reslipping. All right then. Last question. For five shillings: Edward theSeventh came to the throne in nineteen-oh-one. When did he die?Jean thinking): Nineteen-nineteen?Tom: No. Nineteen-ten. That's the last three wrong. Let's see, that'sfour bob, four and six, five bob, take away five bob-nine and six.Jean: Well, thereyou are. Now do you believe me?Tom: Well, no. You don't know much, that's true. But besides tensixty-six, you got all the questions right about the Tudor and Stuart kings-apart from Henry the Eighth.Jean reflectively): Yes.Tom: And even thereyou got a kind of nearmnemonicmiss. You knowyou must have done them at some time. You couldn't justhave been guessing,Jean, could you?Jean: No, I don't suppose .... Yes, you know-come to think of it-Ithink I rememberI didonce have to learnsome dates ..Tom: Ah, yes!Jean Some kings and queens. Perhaps it was these. As a punishmentI think it was. But I'd quite forgotten about it, really.Tom: Oh-Jean: Yes, I think it was these-but really-Tom: No, no, no-I believe you. Freudianforgetting, I expect.Like the others this is not of course a real-life example, and thistime we should consider whether it is a possible one, i.e., whether it isa conversation that could take place, and, if it could and did, whether

  • 8/3/2019 Colin Radford-Knowledge by Examples

    5/12

    4 ANALYSISits participants would be right in concluding, as they do, that theJean-figure did know some English history.

    Clearly such a conversation might take place, and I shall temporarilyassume that the participants' conclusion would be correct if it did.I shall also assume for the sake of simplicity that Jean was sincere ineverything he said and that the questions about the Tudor and Stuartmonarchs that he got right he would have got right without anyprompting, cues, or clues at all. For if we allow this as a possibility,then we should have to say about our hypothetical example that priorto the quiz, or at very least during the quiz but before hearing Tom'scomments on each correct answer, Jean did know some English history,viz. that William landed in 1066, Elizabeth died in 1603, etc. In particular,he knew, e.g., the date of James I's death, viZ. 1625; that is to say, heknew that James I died in 1625, that P.And yet of course, although in this situation Jean knew that P,he was not certain, or sure, or confident that P. Indeed he was fairlycertain that his answer to the question was wrong, i.e. that not-P, sincehe believed it to be a pure guess in a situation where only one of manysuch guesses could be correct.Moreover, though he was not sure that P, Jean would not have hadany grounds for being sure-or, at least, as he was not aware of them,i.e. of having learned that P, etc.-he would not have been justified inbeing sure, etc., that his answer was right, viz. that P, had he been sure.For although he had at some stage learned that P, he had quite forgottenthat he had done so, and was, indeed, quite sure that he had not. (Ofcourse, when we arequite sure about such matters, but cannot rememberlearning about them, we characteristically infer that, since we are sure,we musthave learned though we have forgotten doing so. But if Jeanwere both sure, e.g., that James I had died in 1625, and yet sure thathe had never learned or heard of or read the date, he would have noright, no good or adequate reason or justification for being sure thatit was 1625 or whatever-unless, for whatever reason, his 'intuitions'about such matters invariably turned out to be right, and he knew this.)So if Example 3 is a possible one it shows that neither being surethat P nor having the right to be sure that P, can be necessary conditionsof knowing that P.1 Indeed, it shows that a man may know that P eventhough he is neithersure that P, and indeed fairly sure that not-P, norjustified in being sure, etc., that P!

    This perhaps is surprising. Certainly it contradicts most of whatphilosophers have had to say about knowledge. Moreover it raises1 That is not to say that one is not characteristicallyor paradigmaticallysure of what oneknows. But if being sure is a necessary condition, one cannot know unless one is sure.But cf. Cohen's remark, 'Claims to Knowledge', Arist. Soc., Supp. Vol. XXXVI (1962),'. .. if... confidence that p is nevera necessary condition of knowledge that p. . .' (p. 46,my italics).

  • 8/3/2019 Colin Radford-Knowledge by Examples

    6/12

    KNOWLEDGE-BY EXAMPLES 5furtherproblems. So, beforeconcludingby discussinga coupleof theseproblems, I want to considerwhether and, if so, why, the conclusionthat Jeandid know some English historyis correct.Since the exampleis a fabricatedone we may properlyassumethatthe participants rerightas to thefacts,i.e. thattheyarerightin thinkingthat, e.g., Jamesdied in 1625, that Jeanhad learnedthis at some time,etc. Given this, then, if the conclusion that Jean knew is one thatEnglish speakerswho shared this information would generally tendto come to, Jeandid indeed know and this conclusionis correct. For,ultimately,whether he 'knows' is a question of what 'know' means,which in turn is very mucha matterof when, in what situations,Englishspeakers say or would say that someone knows or does not know.Tom's conclusion,which Jeanhimselffinallyaccepts,viz. that Jeandidknow some English history, did know, e.g., the date of James'death,is one, I think, that similarlyplaced English speakerswould make,andis thereforecorrect.

    However, whether we sayJeandoes or does not know is not simplya matter of appealingto one's intuition and then checkingthis againstthe result of some Naessian survey. We can provide reasonsfor ourjudgmentthat he knows-and weigh considerationswhich seem to tellagainstit-and this is what Tom and Jean do. After some discussionthey agree

    (i) that Jean's answers show that he has-he musthave-learnedsome English history, viz. that which constitutes or, moreprobably, includes the answers to those questions he got right,which seems to be, almostexclusively,the dates of the Tudor andStuartmonarchs. I.e., at some stage he must have learned thatJamesI died in 1625, etc. He did not get these answersright bysheer fluke or chance(or-a possibility they did not even botherto rule out-as a result of some mysterious ntuition).(ii) They also show that he has notforgottenall the history, allthe dates,that he musthave learned-even thoughhe hasforgottenthat he has learned them (it). For he produces various correctdateswhen asked,and does so in such a way and sufficientlyoftento precludethe possibility that he might simply be guessing andnot remembering-even though he is inclined to think that he isguessing.(iii) So he remembersome history, and hence he knows somehistory, including,e.g., thatP.

    Of course, Jean's knowledge of English history is a poor thing,sparse,uncertain,unwitting, and thereforeunimpressiveand of little

  • 8/3/2019 Colin Radford-Knowledge by Examples

    7/12

    6 ANALYSISuse. But thatis not to saythat,at the timeof the quiz, it is wholly gone,totally forgotten, i.e. that it does not exist.

    Although the quiz reveals that Jean does know some Englishhistory,he does not know that he knows any until after Tom has toldhim that certainof his answers are correct.1E.g., when asked, Jeanknew the date of JamesI's death,viZ. hatP, but he did not know thathe knew this. For he did not think that he knew the date of James'death and was indeed quite sure that he did not and that he wouldthereforehave to makea guess at it. Moreover,had he been sure thathe knew the date,andyet still sure,as he was, that he had never learnedit (and sure that if he hadever seen or heardor readit, it had left no'impression',had not 'registered',etc.),he would certainlyhave had noright to be surethathe knew the date. But this lastpoint is a complica-tion. Jeanwas not sure that he knew the date of James'sdeath for hewas sure that he did not know it, and, having forgotten that he hadlearnedit, and indeed being quite sure that he had never learnedit,he did not have the right to be surethathe knew this date.But although this conclusion is correct,the accountabove of whyJeandid not know thathe knew,e.g., thatP, is incompleteandin sucha way as to seem paradoxical. For it appearsto reintroduce at thesecond level (knowingthat one knows thatP) precisely hose conditionsfor knowing which, I have argued, are not necessaryconditions forknowing at the firstlevel (knowing thatP) or at any level at all.In fact, of course, no paradoxis involved here. For to say thatbeing sure and having the right to be sure arenot necessaryconditionsof knowing that . . . is not to say that it is possible to know that 0without satisfyingthese conditionsfor any value of '0', but only thatone can know that b without satisfyingthese conditions for at leastone value of '0'. (Cf. Footnote 1, p. 4.)Even so, the account of why Jean does not know that he knowsany English history, including,e.g., that P, is incomplete,and we canremovethe airof paradoxby seeingboth why we want to saythatJeanknows at the firstlevel-even though he is not sureand does not havethe right to be sure(that P) and that these considerationsdo not existor operateat the second level in this particular ase.We concludethat Jean knows some English history, e.g., the datesof some of the Tudorand Stuartmonarchs,becausehis answers,though

    1Hence a gap can appearbetween knowing that P and knowing that one knows that P.That it does sometimes appear is suggested by remarks made when the gap closes, cf. 'Ididn't know I knew that-you know, that the molecular weight of oxygen was sixteen'.But that it can has been denied e.g. by RichardTaylor, 'Knowing what one knows', ANALYSIs,16.2 (December 1955, p. 65), and queried e.g. by Michael Clark (ibid., December 1963p. 48).

  • 8/3/2019 Colin Radford-Knowledge by Examples

    8/12

    KNOWLEDGE-BY EXAMPLES 7different,are rightsufficientlyoften and in such a way as to persuadeus that he has learned these dates and not simply guessed them, i.e.they persuadeus that he has learnedsome history and has rememberedsome. And, sincehe remembers ome, he knows some, even though heis not sure,etc. In contrast,had the questionsin the quiz been framedslightly differently, 'Do you knowwhen ... (or are you going to havea guess)?', Jeanwould have consistentlyand no doubt wearilyreplied'No, I don't know. I'll guess. Was it ... ?', evenwhendid he know thedate.Of course Jean would talk in this way preciselybecausehe is surethat he has never come into contact with any English history and is,therefore,quitesurethathe does not know any. But this does not meanthat a man could not know that he knew that . . . unless he was surethat he knew and he had the right to be sure. For let us consideraslightly different case in which a man is not sure that he knows anyEnglish history and does not have the right to be sure that he knowsany as it is yearssince he did any history. Despite this, he says at thebeginningof a historyquiz, quite modestlybut firmly,thathe doesknowa littlehistory. He is then askedten questions,andon thefouroccasionsthat he does know the date he says, sometimes aftera little hesitation,that he does know it and gives the correctdate,and when he does notknow he saysthathe does not. I thinkwe should have to say of suchaman that he did know a little history, and that he knew that he knewa little history. (We could hardly say of him that he didn'tknow thathe knew anyhistory.)So it is perhapsnot merelyJean'snot being sure,or his not havingthe right to be sure, or even his not believing that he has the right tobe sure,thathe knowsthe answers o anyof the questionsthatTom askshim, thatdebarshim from knowing thathe knows any Englishhistory,or knowing that he knows, e.g., that P. It is ratherthat (being quitesure he does notknow any)he saysat the beginningof the quiz thathedoes not, and would say that he did not know the answer to anyparticularquestionin the quiz even when he did. It is becausehe getsor would get the answersto the 'Do you know ... ?' questionswrong(andcertainof the datesright)that we saythathe is not aware,does notrealize,i.e. he does not know, that he knows anyhistory.'Secondly:althoughat the time of the quiz Jeanknew some Englishhistory, vik. a few dates, before he realised--beforehe knew--that heknew some, he hadno right o saythat he knew. It would, in someway,have been improperfor him to say, e.g., that he knew that Jamesdiedin 1625, prior to his learningthat this was indeed the case, i.e. that heknew. That is not to say that differentcircumstances ould not excuse,

    1But could a man know that he knew some history if he were not merely unsure that hedid, but pretty sure that he did not? Apparently not. But why not?

  • 8/3/2019 Colin Radford-Knowledge by Examples

    9/12

    8 ANALYSISjustify, or even make praiseworthy Jean's saying this, but only that,whatever the circumstances, a Irima facie objection remains, i.e. toclaiming that one knows that P when one is not sure that P or does notbelieve that one knows that P.Cohen (loc. cit.), who distinguishes statements as acts, statements asthe contents of a sub-class of such acts, and propositions (I am unableto make the latter distinction), argues that a man's lack of confidencethat P does not eoizso render his act of 'making the statement' that heknows that P unjustifiable, i.e. morallyunjustifiable (p. 39), which ofcourse is correct. And for him the only other question of justifiabilitythat can arise is whether the proposition that is 'uttered' when the state-ment is made is true or not. But although special circumstances could,e.g., make Jean's claiming that he knew that P morally justifiable, thatis not to say that the Irima facie objection would not remain to hisdoing so or that it would not have to be met and overcome if his actionwere to be morally justified. It is this prima facie impropriety, whichCohen misses or dismisses and which remains even when what Jeansays happens to be true, that I want to explain.Those who claim that being sure and having the right to be sureare necessary conditions of knowledge can give a clear explanationthough, I have argued, an incorrect one. They can say that had Jeanclaimed that he knew that P, what he said would have been improperin that it would have been false. But, on my thesis, had he said this,it would have been true; Jean did know the date of James' death, andyet it would nonetheless have been improper for him to say this. Why?The answer is of course simply that, although Jean did know thatP, he neither knew that he knew that P nor did he believe that he knew.In exactly the same way, Jean would have had no right frima facie tostate, assert, claim-or, for that matter, agree, admit, concede, etc.-that P either, since although he did know that P he did not know thathe knew that P nor did he believe that he knew that P (for he was neithersure that P, nor did he have the right, or believe that he had the rightto be sure that P).This account presupposes that it is frima facie improper for a manto state, etc., that he knows that P, or that P, or whatever, unless hebelieves that he does know that what he says is true, and his beliefthat he knows this is confident, sure, certain, etc., and, he believes,well-grounded or reasonable, etc. This may be true, but how so? Well:

    (i) if a man states, etc., that P, or that he knows that P, or what-ever, i.e. states or concedes, etc., whatever he does state or concede(without qualification, 'as if it were a fact'), his doing this impliesin some way that he believes that what he says is true, that he isconfident that this is so, and that he believes that he has the right

  • 8/3/2019 Colin Radford-Knowledge by Examples

    10/12

    KNOWLEDGE-BY EXAMPLES 9to be confidentthatthis is so. (For if he is not sure, etc.,he shouldqualifywhat he says.) I.e., it implies thathe at least believeshat heknows that what he says is true. But Jean did not believe that heknew that P, and so his stating, etc., that P would have impliedsomethingfalse, and hence would have beenfrimafacie improperin this sense.'(ii) if a man does not believe that he knows that P, this may bebecausehe is not sure that P, or not sure that he has the right tobe sure that P, or sure that he does not know that P, or sure thathe knowsthatnot-P,etc. So if a man does not believethathe knowsthat P, then, ceteris aribus, it is not likely that he doesknow thatP and entirely problematic for most values of 'P') that P. So ifhe says that he knows that P, what he says is likely to be falseeither because it is not the case that P or because he does notknow thatP. Hence, what he saysis likely to be improper n thissense. Moreover, and as we have alreadyseen, if a man states,etc.,thathe knows thatP, or P, or whatever,his doing this impliesthat he at least believes that he knows that what he says is true.So not only does he say something that may very well be false,his doing this implies that he at least thinks he knows that whathe says is true, and this is certainlyfalse. But not only is what isimplied by his stating that he knows that P, or P, or whatever,false, it is something which, more than anything else he couldimply by what he does, would tend to suggest to others thatwhathe saysis true. (Comparewhat he does, which is to state,etc.,thathe knows that P, with someone's guessinghat he, the guesser,knows that P. His doing this, viz. guessing that he knows thatP, does not imply that he, the guesser,knows or thinkshe knowsthat what he guesses is true. Au contraire.So his guessing wouldnot tend to persuadeor suggest to a listenerwho knew what theguesser was doing, vig. guessing, that the guesser did know thathe knew thatP. It would suggest insteadthat the guesserwas notat all surethat he knew thatP.) Thus a man'sstating,etc.,thatheknows that P when he does not believe that he knows this hasimplications that are themselves false and which suggest to thenaive hearerthat he may accept as true what may very well befalse, viz. that the speaker knows that P. Such behaviour isintentionallymisleading,and, of course, remainsso even if, likeJean,the speakerdoes know thatP.

    I think this does explain how Jean's stating, etc. that he knewthat P, when he did not believe this, would have been 1rima facie1 Of course,what is implied n this kindof way is not alwaysbelievedby a hearer,noris it alwaysmorallywrongto mislead rto intend o misleadhim. Butit isprimafacie rong.

  • 8/3/2019 Colin Radford-Knowledge by Examples

    11/12

    10 ANALYSISimpropereven though he did know that P. The explanation,whichhas the advantageof allowing that Jeancould know without believingthathe knew,alsohas theadvantageof offeringanaccount,andpreciselythe same account, of the precisely similar impropriety of a man'sstating,etc.,thatP when he does not believe thathe knows this. Nowas in this case the man does not state or claim or assert, etc. that heknowshatP, the rivalexplanationhas no applicationhere and so cannotpossiblyexplainthe impropriety.'But', it may now be asked, 'even if it is true that a man's stating,etc. thatP somehowimpliesthat he is confident thathe knows, and, hebelieves,properlyconfidentthathe knows thatP, how does it do this?'

    This is not perhapsa questionthat I ought to try to answer withinthe confinesof this paper,but since it is interestingand difficult,and aproblem that the paperleaves me with, I shall very briefly try to saysomethingaboutit.It is tempting o saythat there s aconvention,alinguisticconvention,in English and perhaps n other languagestoo, that one does not state,etc., that P unless one is confident and, one feels, properlyconfidentthat one knows that P. So anyone who breaks this convention sayssomething or, perhaps,does something, that is misleading,and if hebreaks t deliberately, hen he is being deliberatelymisleading.But this sketchof anaccount,thoughattractive,s not justinadequatebut, I suspect,fundamentallyncorrect. For it entails that there couldbe a languagewith a convention which allowed one to state, etc., thatP, even though one did not believe thatone knew thatP, i.e.a languagein which someone's stating that P did not imply that the speakeratleastthoughthe knew that whathe saidwas true,andhencea languagein which, if someone did state that P without believing that he knewthatP, his doing so would not be regardedas misleadingor frimafacieimproper.The difficultyhere is that if there were such a language,we couldnot understand ts users' 'stating', 'asserting','claiming', 'conceding',etc. thatP, i.e. their sayingthatP without qualification nd as if it werea fact, as that. We should ratherunderstandt, if we could understandit at all, as their analogueof our saying 'I don't know whether P ornot', but this of courseis not a way of sayingthatP.The reasonfor this is that statingetc.,that P, i.e. sayingthat some-thing is the case, is essentiallysomething that we do to inform. Butwe cannot hope or try or intend to inform unless we at least believethat we are ourselvesinformed,i.e. have knowledge. Thus we cannotconstruesomethingthata persondoes as his stating,asserting,etc.,i.e.,as his sayingthat somethingor otheris the case,without thinkingthateither he at least believes he knows what he is talking about and soknowsthatP, if that s what he says,orhe is misleadingus and,therefore,his sayingthatP is at leastprimaacieimproper.

  • 8/3/2019 Colin Radford-Knowledge by Examples

    12/12

    MORE ABOUT KNOWING AND FEELING SURE

    MORE ABOUT KNOWING AND FEELING SUREByL. JONATHAN COHENS LET it be granted that there is a range of situations in whichstatements f the form'X know(s) hatf' arenot true unlessat least

    (i) it is true thatP(ii) X believe(s)that5f, and(iii) X's belief that is adequatelygrounded.The problemI wish to discuss is whether there is a sub-classof thosesituations such that

    (ii') X feel(s)surethat]could be written in place of (ii) above. The problemwhether(iii) alsorequires,or admits of, modificationhas been extensivelydiscussedinANALYSISecently', but I shall say nothing about it here.Radfordclaims,however,thateven(ii), let alone(ii'),is notanecessarycondition of X's knowing thatf in at least some situations,e.g. whereX is an examinee and it is clear that X has learnedthatf5, though Xthinks he does not know thatf, yet when questioned he eventuallyreplies that]. But this thesis is not tenable. No doubt the examineethinks he's merely guessing that ]f, and therefore thinks he doesn'tbelieve that]. But if he reallyknows thatf, he's wrong to think he'sjust guessing and wrong therefore to think he doesn't believe thatf.Confusionor errorcaneasilyarisehere if we aretemptedto applylong-term criteria of knowing and short-term criteria of believing. In1Cf. Keith Lehrer, 'Knowledge, Truth and Evidence', ANALYSIS xxV (1956) p. 168ff. andthe references cited there.

    11To brieflysummarisemy negative conclusions:neitherbelievingthat P nor, afortiori,being confident,sure,quite sure,or certainthat Pis a necessary onditionof knowingthatP. Nor is it anecessary onditionof knowing that P that one should have the right to be, or be justifiedin being, or have adequategrounds for being sure that P. Nor is it anecessarycondition that one should believehat one has the right tobe, etc., sure that P. It is, perhaps,rather that being sure that P, andbelievingthat one has the rightto be surethatP, arenecessary onditionsof believinghat one knows, and hence of having theIrimafacieright tosay that one knows thatP.

    Universityof Kentat Canterbury