Philosophical Anti-Authoritarianism

FOR APPROVAL

Philosophical Anti-authoritarianism

Dylan B. Futter1

Received: 5 July 2016 /Revised: 2 September 2016 /Accepted: 30 September 2016# Springer Science+Business Media Dordrecht 2016

Abstract Unlike certain commentary traditions of philosophy in which defer-ence to an authoritative author was a central feature, there are within theanalytical tradition no recognised authorities to whom the reader is requiredto defer. This paper takes up the question of whether this anti-authoritarianposition in philosophy can be sustained. Three lines of argument are consid-ered. According to the first, there are no credible authorities in philosophy, or,even if there were, these authorities could not be identified by the non-expertreader. According to the second, since no philosopher is infallible, manyreaders have on many occasions epistemic grounds for non-deference to theauthor. According to the third, even if some readers have epistemic reason fordeference to some authors, an anti-authoritarian stance can be justified in termsof distinctively philosophical values such as conceptual understanding or intel-lectual autonomy. Although each of these lines of argument contains an elementof truth, a sufficient justification for philosophical anti-authoritarianism remainssurprisingly elusive.

Keywords Metaphilosophy. Authority . Disagreement . Analytical philosophy.

Commentary

1.

The norms governing the hermeneutical relationship between reader and philo-sophical author are not historically or contextually invariant. In some commen-tary traditions of philosophy, for instance, the reader was required to defer tothe authoritative author; he was not entitled to judge that the author had made a

PhilosophiaDOI 10.1007/s11406-016-9781-0

* Dylan B. [email protected]

1 Department of Philosophy, University of the Witwatersrand, Private Bag 3, Wits, 2050,Johannesburg, South Africa

FOR APPROVAL

mistake. By contrast, the style of reading characteristic of analytical philosophyis striking in its anti-authoritarianism. If a philosophical writer asserts some-thing that the reader judges to be false, or presents an argument that seems toher to be weak, she may reject it. Analytical readers do not take themselves tobe required to defer to any authoritative author.

Although the present-day anti-authoritarianism in philosophy seems perfectlynatural and appropriate to initiates,1 it would have been shocking to membersof those philosophical traditions in which deference to authority was a centralfeature. The governing question of this essay is therefore whether the anti-authoritarian stance in philosophy can be justified. In attempting to answer thisquestion, I begin by explaining in more detail what is meant by saying thatanalytical philosophy is anti-authoritarian in character; thereafter, in the remain-der of the paper, I consider some arguments that might be advanced in supportof the anti-authoritarian standpoint. In each case, what is found is that thereasons proposed fail to justify the position.

2.

The anti-authoritarian character of analytical philosophy is best understood byway of a contrast with certain commentary traditions of philosophy.2 The corepoint of distinction is manifest in diverging answers to the question of whetherthere are within the tradition any recognised authorities to whom the reader isrequired to defer. In this regard, some commentary traditions give an affirma-tive answer, whereas the analytical tradition gives a negative. There is, asjudged by the analytical tradition of philosophy, no author to whom the readeris required to defer. I shall refer to this hermeneutical standpoint asBphilosophical anti-authoritarianism^:

(PA) No reader is required to defer to any philosophical author on any philo-sophical claim.

Given PA, the analytical reader is permitted to attribute error to any author. Thereader’s taking herself to have good reasons for rejecting the author’s position issufficient for rejecting that position. In instances of disagreement with the author, thereader is entitled to assume that not she, but the author, is mistaken. By contrast, insome commentary traditions, the reader was not permitted to attribute error to the

1 See Zagzebski (2012, ch. 1) for an illuminating discussion of the historical roots of anti-authoritarianism.Compare also the following remarks by Hadot (1995, 76): BPhilosophers of the modern era, from theseventeenth to the beginning of the nineteenth centuries, refused the argument from authority and abandonedthe exegetical mode of thinking. They began to consider that the truth was not a ready-made given, but wasrather the result of a process of elaboration, carried out by a reason grounded in itself^.2 Although the genre of philosophical commentary is strongly associated with an attitude of deference toauthority, the account given in the text cannot be said to apply uniformly to all commentary traditions.References to Bthe commentator^ in the text below should be understood in relation to the model presentedrather than any specific historical practice. For more on the commentary traditions of philosophy, see Smith(1991), Hadot (1995, 73–76) and Futter (2016)

Philosophia

FOR APPROVAL

authoritative author. 3 The commentator’s taking himself to have good reasons forrejecting the author’s position was not thought to be sufficient for rejecting thatposition. In instances of disagreement with the author, the commentator had to assumethat he, rather than the author, was mistaken.

The anti-authoritarianism expressed by PA is an extreme hermeneutical position. PAapplies to every reader and author irrespective of their philosophical standing. It is notbelieved, for example, that the freshman philosophy student should accept what Aristotlesays about metaphysics because he is not yet in a position to disagree, this being permittedonly much later on, once he has thought about the subject for many years. Rather, everyreader is permitted to disagreewith and reject every claim by every philosophical author rightfrom the outset.4 Even the commentary traditions were not this radical; for in these traditionsthere were only some authors to whomdeferencewasmandatory. Similarly, in contemporaryscientific discourse the attitude of Bpeer review^ is limited to the specialists; the non-expert istaken to have compelling epistemic reason for deference to the scientific author.

I do not mean to suggest by this description that the anti-authoritarian position entitles thereader to dismiss any claim made by any author whenever she might feel like it. Rather, theprerogative to reject the author’s opinion expressed by PA is intended to allow the reader toreject the author’s opinion when she has good reasons, or at least, what she takes to be goodreasons.5 Moreover, the attribution of PA to the analytical tradition does not imply that thereis within this tradition no philosophical author whose opinion carries any weight. It can bereadily granted that in analytical philosophy some writers are treated as authorities in thesense that they are taken seriously. The seriousness with which some writers are taken is infact manifest in the application of another hermeneutical norm distinctive of philosophicalreading within the analytical tradition, viz. the so-called Bprinciple of charity .̂

Although the possibility that a philosophical tradition might embrace contradictoryhermeneutical norms should not be excluded, PA and the principle of charity are perfectlycomplementary. If it is good practice for the reader to reconstruct the author’s argument in itsstrongest form before criticising it, criticism of a fairly reconstructed argument must be

3 Since the commentator assumed that the authoritative author was a knower, he could not attribute error tohim. This sense of ‘could not’ is logical in character: if one regards an author as a knower with respect to agiven proposition one cannot consistently attribute error to him or her. To be sure, since the commentator wasrequired to assume that the author was a knower, there is another sense of ‘could not’ applicable here which isnot logical but social in character. In this regard, the commentator was not permitted to attribute error to theauthor in the sense that his forming the judgment of authorial error would open him up to sanction from histeachers or other members of the hermeneutical community. The relationship between the logical and socialsenses of the requirement to defer to the author is discussed further in §§7-8.4 If the reader disputes that undergraduate students are members of the analytical tradition, or that I haveaccurately characterised the norms for teaching philosophy within this tradition, or that the norms for teachingphilosophy within this tradition accurately reflect the practices of the tradition itself, then she or he shouldsubstitute Bphilosopher of average ability^ for Bfreshman philosophy student^. This last phrase is borrowedfrom Frances (2010b).5 I here understand rejection in a robust sense such that it implies the judgment that an author’s claim that p isfalse (or unjustified) or argument is invalid (or weak). In this regard, it is important to notice that since PApurports to describe the style of reading characteristic of analytical philosophy, its application will in someinstances be complicated by the specific meta-epistemic or meta-philosophical doctrines accepted by thereader. Hence it might be that some philosophers within the tradition believe that the reader cannot or shouldnot judge that a proposition is false or argument invalid because, e.g., all philosophical propositions aremeaningless (e.g. Wittgenstein (1921)) or philosophical reasoning should not be trusted (e.g. Kornblith(1999)). I do not think that this point fundamentally affects my basic claim that analytical philosophy is,when compared with some commentary traditions, rightly described as anti-authoritarian in its eschewal of anyrequirement for textual deference.

Philosophia

FOR APPROVAL

regarded as appropriate, something that would not make sense if PA were not correctlyascribed to the analytical tradition. A further and deeper point is that the application of theprinciple of charity positively implies that the author is not being treated as an authority inthe strong sense applicable to the commentary traditions. This can be shown as follows.Suppose that a reader assumed that he was not permitted to judge that some author waswrong. Then the reader’s judgment that the author’s position on a topic was implausiblewould be taken as evidence that he had failed to fully understand what was said. This oftenhappens in the reading of science when the lay person interprets the apparent absurditiesgenerated by (e.g.) quantum mechanics as a reason for thinking that he has failed tounderstand the theory. It follows that the author would not be in need of charity if he weretreated as an authority. For the question of how one should interpret so as to attribute to him abetter argument would not arise; the question would rather be why one had not properlyunderstood his meaning. Hence the perceived demand for hermeneutical charity presup-poses that the author is taken to be someone with whom one can disagree.

The need for hermeneutical charity arises in a particularly acute formwhen a philosopherwho is taken seriously is judged to havemade a simple error. The application of the principleof charity then involves the reader’s attempt to find an alternative interpretation of the claimor argument onwhich the claimwill seemmore plausible, or the argument, stronger. But thissuggests that rather than the principle of charity’s being a form of textual deference, it is infact a form of textual indeference. For in the application of the principle of charity, the readerBmakes^ the author agree with her assessments of what is plausible or cogent. She occupiesthe hermeneutical centre; she is, in effect, the authority. The reader does not think, e.g.,perhaps my judgments of what is plausible are in error. Rather, she tries to find a way ofhelping the author avoid what she takes to be a dubious claim or argument. Ironicallyenough, therefore, the reader displays charity by interpreting the author’s views as congruentwith her own. The application of a principle of charity therefore seems to involve taking up aposition of epistemic superiority.

3.

PA entitles every reader to go it on her own, without deference to any philosophicalauthority. To deny PA is then to claim that there are some readers and some authors forwhom it is appropriate that the reader defer to the author, that is, for the reader to preferthe author’s judgement over her own. Now, one reason for doubting the reasonablenessof relying on authority in philosophy is that the idea of philosophical expertise seemsproblematic.6 Many philosophical views held by some Bexpert^ or, at least, someonewell placed to be regarded as such, are contradicted by another Bexpert^, apparently justas deserving of the epithet.7 And this might suggest that there are no philosophicalauthorities. But if there are no philosophical authorities, then—so the argument goes—it would be unreasonable to require the reader to defer to one.

Although it should, I think, be accepted that the notion of philosophical expertise isproblematic, this line of response does not succeed. Everything depends on how therejection of philosophical authority is understood. If the fact of disagreement among

6 Cf. LaBarge (1997).7 Cf. van Inwagen (2004, 332).

Philosophia

FOR APPROVAL

experts is taken to imply that there are no knowable philosophical truths, then thereader would never be entitled to disagree with any author; she would never be entitledto judge that she was right and the author was wrong.8 Hence what might be called aBnon-factualist^ interpretation of philosophical authority is irrelevant to the justificationof the practice of analytical reading expressed by PA. On the other hand, if there areknowable philosophical truths, then the fact of disagreement will not exclude thepossibility that some people are better placed to acquire philosophical knowledge thanothers. It is difficult to deny that Aristotle is more of a philosophical authority than thephilosopher of average ability.9 So why shouldn’t she prefer Aristotle’s judgment to herown? PA seems irrational in many cases.

The above reasoningmight suggest that a Bno-expert^ defence of PA is better formulatedin epistemic terms. It could be argued that since there is pervasive disagreement in thephilosophical traditions, the reader cannot know that any given author is an expert; so she isentitled to make her own judgement, that is, to disagree with the author when she takesherself to have good reasons for disagreement. Although it is true that it is difficult for thenon-expert philosopher to identify the expert, this argument is seriously problematic. First,since PA expresses a general hermeneutical orientation, it is accepted prior to any knowledgeof whether or not there might be philosophical consensus on a point at hand. Hence specificknowledge of the philosophical literature will not normally be available to the reader injustifying PA. Secondly, and more importantly, even if the philosophical non-expert couldnever identify the expert, this would not seem to justify the prerogative to disagree with anyauthor (PA). A similar position in science would be absurd. If there were dissensus in ascientific community, the dilettante could not reasonably regard herself as entitled to disagreewith every author on account of her inability to identify the Breal^ experts.

4.

In some commentary traditions, the hermeneutical attitude of textual deference was justifiedby an assumption of authorial infallibility.10 Given this assumption, the commentator wouldseem to have an overwhelming reason for deference to the author: for if the author were aknower, then disagreementwith the author would be disagreementwith the truth. It might be,therefore, and moving in the opposite direction, that the analytical tradition’s prioritisation ofthe reader’s prerogative to disagree with the author (PA) when he appears to be in error,expresses the thought that some readers aremore knowledgeable than some authors on sometopics. In short, some readers will have epistemic reason for disagreement with the author.

Although this line of reasoning has a certain degree of plausibility, it does notsucceed in justifying PA as a general hermeneutical principle. This can be shown by

8 My target here is someone who wants to defend the prerogative to disagree with the author expressed by PA.I do not mean to prejudge the question of whether there are in fact any true philosophical propositions (cf.Wittgenstein, Tractatus Logico-Philosophicus). See also note 5 above.9 The question of whether a member of the contemporary philosophical community knows something aboutthe world that Aristotle did not, and which entitles him to reject Aristotle’s philosophical rather than scientificopinion when it conflicts with his own, is not straightforward. I note in passing that in many branches ofphilosophy, e.g., metaphysics and ethics, Aristotelian or neo-Aristotelian positions are still defended. See alsothe discussion in §5. In any case, if the historical example is found problematic, then a more contemporaryphilosophical authority can be substituted for Aristotle.10 This is not to say, necessarily, that the assumption was believed. For discussion, see Futter (2016).

Philosophia

FOR APPROVAL

considering what the reader who exercises her right to disagree with the author mustassume about her own epistemic credentials.

To accept PA is to believe that one is entitled to reject any author’s claim or argumentif it should, after due reflection and charitable reconstruction, seem incorrect.11 There-fore, the affirmation of PA rationally requires that the reader hold every philosophicalauthor to be, in relation to a contested proposition, her philosophical equal. This can beseen as follows. Suppose that a reader encounters an author who asserts what she takesto be p. Since p seems to her to be false on the basis of apparent reasons (ex hypothesi),and since she believes that she need not defer to the author, she rejects the claim. Hencein rejecting the author’s claim the reader excludes the possibility that she, rather than theauthor, might be mistaken. She judges that her reasons for thinking that ~ p are weightierthan the author’s reasons for asserting what she takes to be p. Hence it is clear that thereader who rejects the author’s claim that p cannot regard the author to be her epistemicsuperior with regard to p, or at least, not by any significant degree. For if she assumedthat the author were by a significant measure her epistemic superior, she would have toseriously consider the possibility that p, or, that her judgment that ~ p, was in fact areason for thinking that she had not understood the author’s meaning. Hence the reader’sact of disagreement in regard to a given proposition, understood as the rejection of theauthor’s claim, requires that the reader regard the author as, at least roughly, herepistemic equal.12 Therefore, acceptance of PA implies acceptance of what I shall callphilosophical egalitarianism (PE).13

According to PA, no reader is required to defer to any philosophical author on anyphilosophical claim. Hence, every reader who accepts PA believes occurently ordispositionally that she is not required to defer to any author on any philosophicalclaim. Therefore, the acceptance of PA commits her to the view that every author is herphilosophical equal with respect to every proposition about which they disagree (PE).14

In fact, the reader can take herself to be rationally entitled to reject any claim that theauthor might make on the basis of her apparent reasons (that is, PA) only if she

11 The reader who accepts PA affirms the right to disagree with every philosophical author when she believesthat she has properly engaged with and understood his text. The important point of contrast with thecommentary tradition is that the commentator did not affirm this right: he was never entitled to disagree withthe authoritative author. For a discussion of how the commentator could deal with those instances in which theauthor had seemingly asserted something contradictory or absurd, see ibid., 106–112.12 By Bepistemic equal^ I am then referring to an all-things-considered judgment of philosophical equalitywith respect to a proposition, not necessarily equality of intelligence, intellectual conscientiousness, amount ofevidence considered, and so on.13 It might be that the above line of reasoning can be extended to show that the act of disagreement requires the readerto regard the author as her epistemic inferior. For if the reader assumed the author to be her epistemic peer on the pointat hand, and hence that she was not any more likely to have reached the truth than the author, she would have noreason to prefer her own judgment of ~ p over the author’s judgment that p. Hence itmight appear thatwhen the readerrejects the author’s view, thus concluding that she is right and he is wrong, she in effect assumes that she knows betterthan the author on the specific point of disagreement.Whatever the strengths andweaknesses of this line of reasoning,I shall set it to one side. It will be sufficient formy purposes that the reader regard the author as, at least, a philosophicalequal with respect to a contested proposition.14 In the vocabulary of the contemporary epistemology of disagreement literature, the argument is from PA tothe denial of the existence of philosophical superiors and hence to the claim that all philosophers are peers (cf.Frances 2010b). On peer disagreement, see Kelly (2005) and Frances (2010a). Compare also what NathanBallantyne calls the Bdogmatic dismissal^ strategy of Bpreserving rational belief in the face of disagreement^(2015, 143–145).

Philosophia

FOR APPROVAL

supposes that he is not any more likely—or much more likely—to have attained thetruth on the point at hand (that is, because PE).

If this reasoning is correct, then PA remains unsupported. Commitment to PA entailscommitment to PE and PE is manifestly implausible. Not all readers have the samephilosophical powers as all others; nor do all have the same opportunities for thedevelopment of their philosophical powers, nor even the same amount of time forserious philosophical thinking. Therefore, in many cases, the author and the reader willnot be philosophical equals in relation to contested propositions. A policy of textualindeference would seem unreasonable for many readers on many occasions; for itwould mean diminishing one’s chances of getting to the truth. Contrary to PA, it wouldseem that some philosophical readers have a strong epistemic reason for deference tosome authoritative authors.

5.

The activity of philosophy is a search for knowledge. Let us assume, to begin, that therelevant kind of knowledge is propositional knowledge, that is, propositional answers to avariety of philosophical questions about, e.g., the nature of justice and the relationshipbetween mind and body. It might be argued that even if some readers have epistemicgrounds for deference to some authors, PA is justified as a general hermeneutical principlebecause it can enable other readers, or the philosophical community of which these readersare members, to advance in philosophical knowledge. Perhaps Bimmodesty^ can helpphilosophy make progress. If every reader defers to the authority of the established author,things will remain the same; the false will never be replaced by the true. Philosophicalprogress, or so the argument goes, requires the overturning of wrong opinions.

The argument from philosophical progress can be understood in relation to individualphilosophers or whole philosophical communities. With regard to the individual, it is notclear that PA can be justified as conducive to philosophical progress. For it would seembetter that the person of limited philosophical powers not reject Aristotle’s opinions if hedesires to make philosophical progress. He is more likely to advance in propositionalknowledge by deferring to Aristotle than by going it alone. In fact, the philosopher ofaverage ability would seem to be defeating himself by assuming PA. For although he desiresphilosophical knowledge (ex hypothesi), he commits himself to thinking that every author hereads has philosophical opinions that are no better than his own (PE).15

To be sure, the actual epistemic relationship between reader and author will notalways be one of inequality; it will sometimes be one of equality or even readersuperiority. Therefore, even if some readers would do better to trust Aristotle thanthemselves, there are other readers who would do better to follow their own philo-sophical judgment. But this does not justify PA as a general hermeneutical principle,

15 It might be objected here that I have vastly underestimated the role of argument in philosophical reading.Many philosophical arguments begin with premises that are widely accepted and then proceed to deriveconclusions that extend the reader’s knowledge in some way. Although this point is right, I believe, it isirrelevant to the question at hand. The claim is not that philosophical progress cannot occur when one agreeswith the premises of the arguments presented, but that it cannot occur when one does not agree. The samepoint applies at the level of the arguments themselves. The claim is not that progress cannot occur when oneagrees with the arguments, but when one does not agree.

Philosophia

FOR APPROVAL

that is, a methodological precept applicable to all philosophers working within theanalytical tradition. The choice is not between a blanket assumption of equality and ablanket assumption of inequality. More moderate positions can be formulated. Even ifthe young Descartes should not defer to Aristotle, this does not mean that the philos-opher on the Clapham omnibus should not, if, that is, he wants to attain knowledge.

The argument for PA based on philosophical progress seems stronger when focussedon the philosophical community. For it might seem that even if there are some readersfor whom adherence to PA results in an epistemic loss, the community as a whole willbe better off if everyone accepts PA. For the community might include philosopherswho are capable of surpassing Aristotle, or whoever the accepted authorities might be.And when this happens, everyone in the community will be better off on account of theprogress achieved by their philosophical superiors. This mechanism for the transmis-sion of knowledge seems quite close to that which operates in the hard sciences.

The above line of reasoning can, however, be questioned at several points. First, ifphilosophical progress—advancement in propositional knowledge—is possible, thenso is philosophical regress. If someone who disagrees with Aristotle fashions a theorythat is weaker than Aristotle’s, then the whole community will lose ground if the newtheory comes to be accepted. So the mechanism whereby new Bknowledge^ is trans-mitted to the rest of the community might in fact result in the community’s acceptanceof a substantially false or impoverished account of the phenomena.

Secondly, if PA is accepted by everyone in the community, then new knowledge willnot be easily disseminated. PA will be used to preserve certain wrong opinions by (exhypothesi) those who are not the philosophical equals of their respective authors, nowunderstood as those authors who have improved upon Aristotle. At a superficial level,anti-authoritarianism prevents the reader from knowing because it prevents her fromtaking herself to know what the author has said.

This last point can be elaborated by attention to the hermeneutical authoritarianism ofthe sciences. If an astronomical text says that Alpha Centauri is 4.37 light years from thesun, then the lay reader will accept this claim on an assumption of authorial expertise.16

Although there may be good reasons for her to make this assumption, it is clear that this isnot something that she undertakes to justify. Moreover, if she were to attempt to justify herattribution of knowledge to the author, she would find that a number of philosophicalquestions arise, which have no easy solution. Thus it seems that scientific reading iscapable of transmitting Bknowledge^ from expert to lay reader only because the lay readerassumes that the author is her epistemic superior. Likewise, at the most superficial level,philosophical anti-authoritarianism (PA) enables the reader not to know, or to take herselfnot to know, what she reads. Hence the anti-authoritarian principle PA actually preventsthe transmission of new knowledge within the philosophical community.

Although the above responses to the attempt to justify PA on the grounds of epistemicprogress in philosophy seem to be sufficient, it is worth considering briefly the question ofwhether philosophy does make progress in propositional knowledge. Here it seems to methat the answer is ‘no’. Most, if not all, of the basic questions of philosophy remainunanswered.17 This is evident from the simple fact that the present positions on fundamental

16 I have benefitted from the discussion of this point in Jones (unpublished). However, I have drawn ratherdifferent conclusions from those drawn by Jones.17 In support of this judgment, see Chalmers (2015).

Philosophia

FOR APPROVAL

philosophical questions are verymuchwhat theywere in the time of Plato. Of course, it doesnot follow that there are not better and worse answers to philosophical questions: clearlythere are. But this does not support the view that philosophical progress can be understoodon the model of scientific progress. Philosophical ideas never seem to die completely; thereis no even approximately linear progression in propositional knowledge in philosophy.

6.

It must be admitted that there is something disconcerting about the idea of formingphilosophical opinions on the basis of testimony. A person who believes on theprinciple Bbelieve what Aristotle says about philosophy^ might come to many truebeliefs, but would seem to be making a mistake about what philosophy is. By contrast,a person who believes on the principle Bbelieve what Aristotle says about biology^would likely come to many false beliefs, but would not seem to be making the samekind of mistake. She is not, one wants to say, confused about what she is doing, but hasonly adopted ineffective means to her ends.

Given this line of reasoning, it might be that PA expresses the view that even if there areepistemic reasons for a reader to defer to a philosophical authority, she should not do so. Shehas what might be termed a philosophical reason for indeference. The reader should notaccept philosophical opinions on the basis of authority no matter who might be assertingthem. A philosophical reason for textual indeference trumps whatever epistemic reason shehas for deference.18 I shall formulate this refashioned anti-authoritarianism as follows:

PA′: No reader should defer to any philosophical author on any philosophicalclaim even if she has an epistemic reason to do so.

In evaluating PA′, it is important to notice that postulating an over-riding philosoph-ical reason for not deferring to the author, whatever that might involve, does not byitself undo any of the reasoning above (§§3-5). If the author were an expert on the topicunder consideration, some readers would become worse off epistemically bydisagreeing, or at least, lose out on the opportunity for becoming better.19 Whateverreasons there are for Bgoing it alone^ do not seem to justify a complete disregard forauthority in philosophy. Hence, even if the philosophical reasons for indeferenceoverride the value of the epistemic states potentially achieved by textual deference,the latter might retain their significance. It is difficult to determine what the appropriateresponse to such philosophical residue might be.20 I shall set the question to one side.

The transition from PA to PA′ is meant to capture the idea that there is something aboutphilosophy which makes deference to authority illegitimate, even if it were to bring about

18 See Hopkins (2007) who makes a related suggestion in regard to moral testimony. See also the discussion ofmoral deference in McGrath (2009).19 This qualification is necessary because, as will be argued below, a requirement of textual deference does notcompel belief.20 What I have in mind here is a case where, e.g., the value of coming to believe a true philosophicalproposition on the basis of testimony does not coincide with the value of having understood and assimilatedthis proposition for oneself. To imagine such a case, one need only think of an instance where someone gets tothe truth by believing Aristotle and would not get to this truth by relying on her own philosophical reasoning.

Philosophia

FOR APPROVAL

advancement in propositional knowledge. In the remainder of the paper, I will considerthree candidates for the explaining the nature of this potentially over-riding philosophicalreason: first, what I shall call epistemic intimacy,21 secondly, conceptual understanding,and thirdly, intellectual autonomy. Since each proposal is compatible with the other two,the justification for PA′ could in principle incorporate all three.

7.

In Plato’s Charmides, Socrates suggests that philosophical inquiry is a kind of self-inquiry (166c-d). It is difficult to understand precisely what he means. Nevertheless, heis surely right in holding that philosophical thinking is personal; philosophical theoriesmust connect with the way the reader sees the world.22 Presumably this is one reasonfor the ad hominem structure of Socratic dialectic. Socrates always tries to meet theinterlocutor where he stands; to allow philosophy to grow from his point of view,however confused and distorted.23

This point can be elaborated by considering briefly the nature of sophistry asrepresented in the Platonic dialogues.24 The sophistic interlocutor is someone whohas moved so far beyond the intuitive and experiential grounding of his philosophicalopinions that he is constrained by propositional inconsistency alone. He has lost theability to know whether the proposition that he defends expresses what he actuallythinks. Hence sophistry is a mode of disconnection from the pre-theoretical groundingin which philosophy takes root. In this respect, sophistry is philosophy floating-free-of-the-self. The association of sophistry with hypocrisy—an incongruity of word anddeed—can then be explained in terms of the sophist’s deeds better reflecting hisBphilosophy^ than his discourse.

There is clearly much more to be said on this difficult topic. Nevertheless, the abovesketch will be sufficient for considering whether PA′ can be justified by philosophy’sepistemic intimacy. The thought is that since philosophical theories are, unlike scien-tific theories, answerable to the thinker’s view of the matter, the reader is justified inassuming PA′. The revised anti-authoritarian principle (PA′) expresses the requirementthat philosophical discourse be epistemically intimate.

One response to this argument is to notice that epistemic intimacy does not requireanti-authoritarianism. To deny PA′, and to claim that some readers should defer to theopinion of the author, is not to say that these readers should believe the opinion of theauthor. If someone encounters a philosophical proposition that seems false, but onwhich he accepts that the author is an authority, he cannot believe it. This is because hisacceptance that the author is an authority does not change the fact that the propositionstill seems false and he does not understand how it could be true. Hence to require thatthe reader abandon his philosophical belief in favour of the author’s without substantialconceptual therapy is an impossible demand.

21 This useful phrase is borrowed from Jones (unpublished, ch. 2).22 This way of expressing the point owes much to Jones (ibid., 47–52).23 Cf. Aristotle’s account of peirastic dialectic (De Sophisticis Elenchis 2.165 b4-6). For discussion, seeBolton (1993).24 The contrast between ordinary and sophistic discourse is vividly drawn in Laches and Charmides. Fordiscussion, see Gonzalez (1998).

Philosophia

FOR APPROVAL

A strong requirement for textual deference implies that the reader will not be entitled toreject the author’s opinion; but this does not require him to believe the opinion on the basis ofauthority. To be sure, in circumstances when the reader disagrees with the author, therequirement for textual deference, understood as the requirement not to reject the author’sview, will bring about a kind of cognitive dissonance. The reader will believe that p, whilstalso believing that he should not reject what he regards as the author’s claim that ~ p.Although this is not a comfortable state to be in, this does not mean that it is not possible, ordesirable. In fact, the Socrates of the Platonic dialogues seems to regard aporia of a relatedkind to be the best possible state for a human being.25

The attempt to justify PA′ on the grounds of epistemic intimacy seems to rest on theerror of thinking that if the reader had a strong epistemic reason for deference to the author,then he would have to accept philosophical opinions on the basis of the author’s testimony.But since this is not so, there is much room for philosophical thought that is epistemicallyintimate within the boundary of the proposition with which the reader disagrees and yetholds the author to know. The kind of inquiry that deference to the author will inspire isself-inquiry in the relevant sense. For the reader would be trying to make sense of—fromher perspective—what the author is, or might be, getting at. The attempt to assimilate theauthor’s view to one’s own occurs within a space that is personal. It is the search for anunderstanding of the claim that can be accepted by the reader.

The attempt to ground PA′ on philosophy’s epistemic intimacy involves a mistake inthe other direction too. The motivation to defend epistemic intimacy by prescribingtextual indeference seems incompatible with philosophy’s self-conception as a pursuitof truth or knowledge.26 PA′ permits the reader to remain as he is, rather than openinghimself up to the possibility of being transformed by a truth-directed dialogue with theauthor. As Gadamer says:

To reach an understanding in a dialogue is not merely a matter of putting oneselfforward and successfully asserting one's point of view, but being transformed intoa communion in which we do not remain what we were.27

Philosophy’s epistemic intimacy should not be thought of as believing what onewants to believe, or coming to believe in a way that is comforting. True epistemicintimacy, as it were, should allow for the possibility of a transformation in one’sphilosophical outlook. Even if philosophical discourse must connect with the thinker’sintuitions and experiences, one would be giving up on the very idea of philosophy ifone preferred one’s own viewpoint to philosophical truth.

8.

According to PA′, every philosophical reader has an over-riding philosophical reasonfor indeference to the author. The question is what the nature of this philosophical

25 See Futter (2013).26 The concepts of truth and knowledge are here used as placeholders. It should not simply be assumed thatphilosophical truth and knowledge are propositional. See also the discussion of conceptual understanding in §8 below.27 See Gadamer (2004, 371). I am grateful to Gary Beck for drawing my attention to this passage. See alsoGadamer (1997, 36).

Philosophia

FOR APPROVAL

reason might be. In this section I consider the possibility that PA′ can be justified by thevalue of conceptual understanding.

What I mean by conceptual understanding will be familiar to everyone who has engagedin philosophy. After a class in epistemology, say, one rarely—in fact I would say never—hasanswers to the questions that govern the discussion; but one has a command of a conceptualdomain that one did not have before. Before the class, onemight not even have distinguishedknowledge from belief, let alone true belief, justified true belief, warrant, and so on. Hence itseems that conceptual understanding of a philosophical domain can improve withoutaccumulating propositional knowledge. By learning how to navigate one’s way, see simi-larities and differences, draw distinctions, and advance the discussion in a certain manner,one’s understanding of the topic becomes clearer.28 The Socrates of Theaetetus is a brilliantillustration of this point, since his Bknowledge^ of knowledge—that is, conceptual under-standing of knowledge—is manifest in the knowledgeable way he conducts the inquiry.29

This remains the case even though the dialogue ends in aporia.The question before us is whether PA′ can be justified by the value of conceptual

understanding. An inquirer, who arrives at a set of reasonable philosophical beliefs onthe basis of testimony, e.g., Bknowledge is justified true belief^, or Bjustice is equalityof rights and resources^, but without any understanding of the relationships betweenthe relevant concepts, or the reasons in favour of the position, seems to be lackingsomething of philosophical importance. So the idea is that PA′ can be justified as anattempt to prevent the reader from accepting philosophical propositions on the basis ofauthority, thus forcing him to think for himself and improve his understanding. Thepossible advantages of deference to authority in bringing about propositional knowl-edge are outweighed by the value of conceptual understanding.

Since conceptual understanding is a valuable state or ability, the question is whetherPA′ facilitates its development. One question relevant to determining this is whethertrue beliefs about a topic are necessary for understanding it.30 Although this is a difficultquestion, it is not dialectically necessary to resolve it, since the defender of PA′ wouldseem to be on strongest ground if she denied that conceptual understanding involvestrue belief. For if understanding requires true belief and something more, viz. Bseeing^connections between concepts,31 then the objections from previous sections will apply.Deference to authorities would help some thinkers to acquire the true propositionalbeliefs necessary for conceptual understanding.

While PA′ cannot directly produce understanding, it might encourage the reader tothink through the relationships between the relevant propositions and concepts forherself. I shall transform the question of whether PA′ is effective at stimulating philo-sophical thought into the comparative question of whether it would be, in this regard,

28 Compare Catherine Elgin’s description of un-tethered understanding in the sense of knowing one’s wayBaround the field^ (2007, 4–5). Although knowing one’s way around epistemology is not fundamentally amatter of propositional knowledge, I would not regard it as un-tethered.29 For an account of Socrates’ dialectical ability in terms of knowing how to conduct a philosophical inquiry,see Gonzalez (1998).30 This will require answering, at least, the question of whether Bintuitions^ generated by the possession ofconcepts are to be understood as beliefs. For an argument in favour of a negative answer, see Bealer (1998).31 True beliefs can be acquired by testimony; but one cannot teach someone else to see logical relationshipsbetween ideas. See Plato, Meno 82a-85e, and Burnyeat (1987).

Philosophia

FOR APPROVAL

superior to a corresponding assumption of authority. Although different authoritarianprinciples can be formulated, the following will be sufficient for present purposes:

D: Every non-expert philosophical reader must defer to the opinion of therecognised philosophical authority.32

If PA′ does no better than D at promoting conceptual understanding, then it cannotbe justified on these grounds.

Consider two non-expert readers, A and B, each with a set of philosophical opinions,occurent or dispositional, on a topic of concern. When, in the course of their reading onthis topic they are presented with claims asserted by the author, they find themselves eitherin agreement or disagreement, or are unsure where they stand. 33 Person A reads onprinciple PA′ and person B reads on principle D, and the author is for D a recognisedphilosophical authority. Suppose, in the first instance, that both readers come upon someclaim that strikes them as true. This is a context of agreement. What occurs? Reader Aaccepts the proposition and continues reading. Reader B accepts the proposition andcontinues reading. In a circumstance of agreement, there is no difference between theirepistemic responses. Both individuals follow to see where their agreement with the authormight lead them. Suppose now, in the second instance, that the readers come upon someclaim that seems to them to be false. Reader A makes the judgement that the author iswrong, and moves on to another proposition, argument, or author.34 By contrast, B is notpermitted to reject the author’s claim or argument, no matter how strong he thinks hisreasons might be. Since he is required to defer to the author, he must assume that his ownjudgment is incorrect, or that he has failed to grasp the author’s meaning. Hence B cannotmove past the claim or argument. He cannot accept or believe it because it seems to him tobe false. But he cannot reject it because of D. He is in aporia.

If reader B is to dissolve the perplexity, hemust continue to philosophise. The fact that heis not permitted to reject the apparently false claim of the philosophical authority forces himto rethink the matter. The kind of thought that he is required to engage in is not the kind ofthought that can be borrowed from a superior. He must ask such questions as Bwhy do Ithink what I think such that I disagree with the author?^, Bhave I understood the author’sreasoning?^, Bis the author really saying what I think she is saying?^ Hence D generatesphilosophical activity of the sort that can improve understanding. Therefore, in a context ofdisagreement, D seems better than PA′ at promoting conceptual understanding.

Now consider a case in which readers do not have prior occurent or dispositionalopinions on the subject matter. It is likely in such a circumstance that the readers willupon sufficient reflection form Bintuitions^ relevant to judgement.35 In this regard,

32 The concepts can be filled out in different ways. For example, let a non-expert be someone who is notAristotle or Kant.33 The third case of uncertainty could refer to an informed suspension of belief or a failure to understand thematter sufficiently so as to be able to make up one’s mind.34 The judgment that the author is wrong need not be hasty or ill informed; rather, it is, or could be, justified bystrong apparent reasons.35 See Bealer (1998). If the reader does not form any intuitions about the truth or falsity of the proposition, butremains in a state of honest perplexity, D does not do any worse than PA’ in encouraging the philosophicalactivity necessary for conceptual understanding. The psychological position of one who thinks I can’t rejectthis claim because someone who knows thinks it is true but it does not seem true to me is unstable in a way thatwould seem likely to inspire inquiry.

Philosophia

FOR APPROVAL

reader Awill accept the proposition if it seems true, and reject the proposition if it seemsfalse, where seeming arises from reflection. No further inquiry in either case. On theother hand, while reader B will accept the proposition if it seems true, he will not rejectit if it should seem false, because he reads on the assumption that the author is hisepistemic superior. And since he is not entitled to reject the proposition because hemust assume that it is true, or very likely to be true, he has little choice but to inquirefurther, that is, to try to refashion his understanding of what the proposition means, orwhy the author has put it forward as true. Hence it seems that D is superior to PA′ inpromoting active philosophical thought, and hence, by assumption, promoting the goodof conceptual understanding. If PA′ cannot be justified over its (near) contrary in regardto the goal of understanding, then it cannot be defended on these grounds.

9.

Philosophical anti-authoritarianism in the form of PA′ cannot be justified on the basis ofepistemic intimacy or conceptual understanding. Are there any other philosophicalgoods that could justify it? One further possibility is that PA′ expresses an aspiration tointellectual autonomy. The philosophical reader should not rely on the opinions ofothers but must in some sense become capable of thinking for herself.

The meaning of autonomy in the present context is not the right to intellectualautonomy or the freedom to determine one’s own research agenda.36 Neither is it thesense of autonomy relevant to political theory in the liberal tradition, that is, a set ofrational capacities common to all or most human beings. Plainly the anti-authoritarianstance cannot be justified as conducive to the development of a set of human powersthat everyone already has. Rather, the relevant sense of autonomy must be developed,not given. Intellectual autonomy is, in this sense, an ideal.37

If political autonomy is a community’s actualisedcapacity togovern itself politically, thenintellectual autonomy is a person’s actualised capacity to govern herself intellectually. Thiscaptures at least part of the Kantian thought that autonomy is rational self-governance. Thispoint can be extended by noticing that rational self-governance is undermined by bothexternal conditions that threaten a person’s independence, and by internal impediments thatinhibit her ability to exercise her capacity for rationality.38 Neither a person who believesunder hypnotic suggestion, nor someone whose thinking is systematically biased, haveactualised the capacity for rational self-control.

The proposal for evaluation is whether PA′ can be justified as a methodologicalprecept designed to enable readers to develop intellectual autonomy. This turns into thequestion of whether PA′ encourages readers to become independent in philosophicalthought and to carry out their thinking rationally without bias and prejudice. I shallargue that PA′ cannot be justified in these terms because it reinforces intellectualunselfconsciousness.

36 The concept of intellectual autonomy is difficult to disentangle from related notions such as self-reliance,authenticity and integrity. On the relationship between epistemic autonomy and self-reliance, see Zagzebski(2012). On autonomy more generally, see Christman (2015).37 See Christman (2015) and Feinberg (1989).38 Cf. Zagzebski (2012, 25).

Philosophia

FOR APPROVAL

The first point to be made in developing this argument is that philosophicalreading within the analytical tradition takes place from a position of hermeneuticalattachment. Unlike the reader of a fiction, who does not expect to form beliefs onthe basis of what is said, and hence feels no impulse to the evaluation of theepistemic credentials of the claims put forward,39 the analytical reader believes thatshe must take up an epistemic position. Whether reading an article on the concept ofsupervenience, or the nature of hermeneutical egalitarianism, she must evaluate thearguments presented by asking whether the reasoning is valid, or strong, andwhether the premises are true, or justified.40

The attached hermeneutical standpoint is in a relevant sense unselfconscious. Thisclaim can be illustrated with a simple example. Suppose that a given reader of Rawlshas, by intelligence and endurance, overcome adversity and achieved success in herprofessional life. In her reading of A Theory of Justice, she is troubled by those passageswhich appear to downplay personal responsibility and desert. 41 In her view everyindividual is responsible for what she makes of herself. Now, the present issue is notwhether Rawls is or is not correct about desert; the point has to do with theunselfconsciousness of the attached attitude. In the present example, the reader asdescribed has many pragmatic reasons for her belief in the significance of personalresponsibility. But, on account of attachment, she does not think about them. Forexample, she does not think: since I am so deeply committed to seeing the world interms of personal responsibility, it is likely that I find it difficult to see the world fromthe perspective of one who experiences himself as powerless. Nothing of this sortoccurs to her. Instead she feels called on to take up a position in response to Rawls’claims, in this instance judging that he is wrong for such and such, and indeed, possiblycompelling, reasons. Her concern with judgment formation, understood as concentra-tion on the Bis this true or justified?^ question, diverts her attention away from anotherquestions about the self, viz. Bwhat are the sources of my inclinations to judge in theway that I do?^42

The argument thus far deals with attachment, not PA′. Nevertheless, if theabove reasoning is correct, then intellectual autonomy can only be cultivated bya set of hermeneutical principles that encourage a breakdown or suspension ofthe attached attitude and a redirection of the reader’s attention onto herself asmoral and epistemic agent. Hence the argument from attachment can be ex-tended by asking whether PA′ does a better of job than its authoritariancontrary, D, in developing in the reader the capacity for detached contemplationof herself qua hermeneutical agent. In this regard, consider again two readers,A and B, where A reads on principle PA′ and B reads on principle D, andwhere the author is again, for B, a relevant authority. In cases of agreementwith the author, there is little difference between the responses of the tworeaders; both will, irrespective of whether they read on PA′ or D, remain in the

39 For example, if the author of a fiction asserts that Alpha Centauri is 4.37 light years from the sun, the readerwill not experience the claim as one on which she is required to make an epistemic evaluation. This would beso even if the writer of the fiction happened to be a well known scientist expressing in his fiction a true andjustified opinion.40 See Futter (2015). The argument of the next few paragraphs draws on this article.41 Relevant passages include Rawls (1971, 72, 104, 311–312); cf. Nozick (1974, 214).42 Cf. Kornblith (1999) on the rationalisation of belief.

Philosophia

FOR APPROVAL

attached attitude. However, in cases of disagreement, as the example of thereader of Rawls shows, PA′ entitles A to reject the author’s claim, and by sodoing, reinforces the attached attitude. A does not attain a position of self-consciousness from which it is possible to understand the sources of herepistemic responses to the author, because she is permitted to reject the author’sview whenever it conflicts with her own. 43 Hence PA′ allows the reader toremain exactly as she is. It allows her, in a way, to remain blind to the biasesand prejudices that prevent her from becoming intellectually autonomous.

Things are rather different for B, who reads on the authoritarian principle.This principle (D) is capable of manufacturing a breakdown in the attachedattitude reminiscent of the Socratic reduction of the interlocutor to aporia. Inthe context of disagreement, acceptance of principle D prevents reader B fromactively rejecting the apparently problematic claim. Although the claim seemsfalse to him, he cannot dismiss it. He must assume that he is not seeing thingscorrectly. In addition, principle D prevents B from ascribing error to the author,since the author is, after all, taken to be a knower, or at least, to a considerableextent his epistemic superior. So B must assume that he is himself to blame forhis lack of comprehension. Hence the locus of the responsibility for the failureto understand is shifted from author onto self. In this way, B can begin to payattention to himself as hermeneutical agent. This kind of self-attention isprecisely what is necessary for the development of intellectual autonomy. Soit appears that the authoritarian principle, D, is preferable to the anti-authoritarian principle, PA′, in destabilising the attached attitude from withinwhich it is impossible to take responsibility for oneself as epistemic agent. Thedevelopment of intellectual autonomy by philosophical reading would seem torequire inequality in the epistemic relationship between reader and author.44

10.

In the Republic, Thrasymachus claims that it Bis easier to ask questions thananswer them^ (336c; Grube trans.). Whether or not this is so in general, it iscertainly true in the present case. For having asked the question of whether theanti-authoritarian standpoint characteristic of philosophical reading in the ana-lytical tradition can be justified, I have failed to provide an answer. What isclear is that none of the arguments I have thought of justify philosophical anti-authoritarianism. If there are other ways of supporting this hermeneuticalstance, these will have to be developed and evaluated.45

43 This is not in any way to deny the possibility of rational responsiveness, or that the reader might encounter anovel argument. The point is that belief revision of this type occurs within the attached attitude. If the reader isprovoked to reflect on herself as moral and epistemic agent when reading this is not because of any normsinherent in the contemporary hermeneutical practice.44 See Futter (2016).45 I would like to express my thanks to an anonymous referee for Philosophia for constructive commentary onan earlier version of this paper, and to Jill Mortillero for kind assistance during the editorial process.

Philosophia

FOR APPROVAL

References

Ballantyne, N. (2015). Debunking biased thinkers (including ourselves). Journal of the AmericanPhilosophical Association, 1(1), 141–162.

Bealer, G. (1998). Intuition and the autonomy of philosophy. In M. De Paul & W. Ramsey (Eds.), Rethinkingintuition: the psychology of intuition and its role in philosophical inquiry (pp. 201–239). Lanham Md.:Rowman & Littlefield.

Bolton, R. (1993). Aristotle’s account of the Socratic elenchus.Oxford Studies in Ancient Philosophy, 11, 121–152.

Burnyeat, M.F. (1987). The Inaugural Address: Wittgenstein and Augustine De Magistro. Proceedings of theAristotelian Society, Supplementary Volumes, 61, 1–24.

Chalmers, D. J. (2015). Why Isn’t there more progress in philosophy? Philosophy, 90(1), 3–31.Christman, J. (2015). Autonomy in moral and political philosophy. In Edward N. Zalta (Ed.), Stanford

Encyclopedia of Philosophy. http://plato.stanford.edu/archives/spr2015/entries/autonomy-moral/.Elgin, C. (2007). Understanding and the Facts. Philosophical Studies, 132(1), 33–42.Feinberg, J. (1989). Autonomy. In J. P. Christman (Ed.), The Inner Citadel: Essays on Individual Autonomy

(pp. 27–53). New York: Oxford University Press.Frances, B. (2010a). Disagreement. In D. Pritchard and S. Bernecker (Eds.), Routledge Companion to

Epistemology (pp. 68–74). New York: Routledge.Frances, B. (2010b). The reflective epistemic renegade. Philosophy and Phenomenological Research, 81(2),

419–463.Futter, D. B. (2013). Socrates’ human wisdom. Dialogue, 52(1), 61–79.Futter, D. B. (2015). Variations in philosophical genre: the platonic dialogue.Metaphilosophy, 46(2), 246–262.Futter, D.B. (2016). Commentary, authority, and care of the self. Philosophy & Rhetoric, 49(1), 98–116.Gadamer, H. G. (1997). The philosophy of Hans-Georg Gadamer. Chicago: Open Court.Gadamer, H.-G. (2004). Truth and method. Translated by Joel Weinsheimer and Donald G. Marshall. Second,

revised ed. London & New York: Continuum.Gonzalez, F. J. (1998). Dialectic and dialogue: Plato’s practice of philosophical inquiry. Evanston, Ill.:

Northwestern University Press.Hadot, P. (1995). In A. I. Davidson (Ed.), Philosophy as a way of life: spiritual exercises from Socrates to

Foucault. Oxford: Blackwell.Hopkins, R. (2007). What is wrong with moral testimony? Philosophy and Phenomenological Research,

74(3), 611–634.Kelly, T. (2005). The epistemic significance of disagreement. Oxford Studies in Epistemology, 1, 167–196.Kornblith, H. (1999). Distrusting reason. Midwest Studies in Philosophy, 23(1), 181–196.LaBarge, S. (1997). Socrates and the recognition of experts. Apeiron, 30(4), 51–62.McGrath, S. (2009). The puzzle of pure moral deference. Philosophical Perspectives, 23(1), 321–344.Nozick, R. (1974). Anarchy, state, and utopia. New York: Basic Books.Rawls, J. (1971). A theory of justice. Cambridge: Belnap Press.Smith, B. (1991). Textual deference. American Philosophical Quarterly, 28(1), 1–13.van Inwagen, P. (2004). Freedom to break the laws. Midwest Studies in Philosophy, 28(1), 334–350.Wittgenstein, L. (1921). Tractatus Logico-Philosophicus. Translated by D.F. Pears and B.F. McGuiness.

London: Routledge & Kegan Paul.Zagzebski, L. T. (2012). Epistemic authority: a theory of trust, authority, and autonomy in belief. Oxford:

Oxford University Press.

Philosophia