Peter McCormick Certainty, Belief, and Rationality * [On the Lockean view] rationality is above all a property of the process of thinking, not of the substantive content of thought. Charles Taylor Those who say that Reason, in Locke, is purely procedural, not substantive, cannot have read Book TV. Nicholas Wolterstorff ... why then have to be human - and, escaping from fate, keep longing for fate? R. M. Rilke 1. Modernist Subjectivism The major development of modern culture has modernism in the arts at its center.1 This development, although a »many-faceted movement,« the Canadian philosopher, Charles Taylor, wants to call »subjectivation.« Sub- jectivation represents the move away from different norms and criteria seen as fundamentally independent from human beings to those that are funda- mentally dependent on the choices of individual human subjects. The movement of subjectivation has two major elements - one concerns the content of human action and the other the manner. Yet such a distinc- tion invites confusion because each element can be taken as self-referential in the same way. Thus, one can easily incline to the view that, just as the * Copyright © 1998 by Peter J. McCormick. All rights reserved. 1 See C. Taylor, The Ethics of Ambiguity (Cambridge, MA: Harvard University Press, 1992), Chapter Eight, »Subtler Languages,« pp. 81-91. Some of Taylor's more recent thoughts can be found in his replies to a series of critical papers about his various positions. SeeJ. Tully, ed., Philosophy in an Age of Pluralism (Cambridge: Cambridge University Press, 1994). See also Taylor's recent papers, Philosophical Arguments (Cambridge, MA: Harvard University Press, 1995), esp. pp. vii-xii. Filozofski vestnik, XIX (2/1998), pp. 53-79. 53 Peter McCormick manner in which I effect any one of my actions is necessarily mine in the sense that it cannot but refer to my ways of acting, so too the matter or con- tent of my actions must similarly refer to my goals or aims or whatever. The consequence of holding such a view comes to nothing more than the »ram- pant subjectivism« that characterizes much of the modernist period.2 And this subjectivism Taylor sees in turn as the root of both »instrumental rea- son and the ideologies of self-centered fulfillment« in so much contempo- rary reflection.3 In this paper I would like to examine critically Taylor's extended views with the hopes of clarifying our understandings of the seminal moments in the modern understandings of certainty. Now, there is more than one kind of self-referentiality at work in this picture of modern subjectivism. While granting that the manner of one's actions indeed refers necessarily to one's own ways of performing actions, to one's ownaims, desires, aspirations, or whatnot, Taylor argues that the matter of those actions need not. For one may just as well center the con- tent of one's actions on matters that owe their significance to something outside or beyond or independent of one's own interests. So subjectivism represents only one of the two kinds of subjectivation which, however easily confused, are nonetheless different. Taylor goes on to advance a second and more substantial claim, this time not about the two main features of subjectivation but about modern art in general and modernist poetry in particular. Without this distinction, Taylor thinks, one cannot understand either. »... the two kinds of subjectiva- tion have to be distinguished,« he writes, »if we are to understand modern art.«4 And to justify this claim Taylor turns to Rilke, to the very well-known opening of the first of the Duino elegies, »Who, if I cried out, would hear me among the angels' / hierarchies?«5 In order to understand what Rilke has in mind here in talking of »an- gels,« one has to see that the meaning of this crucial word, unlike say Mil- ton's use of the same word, cannot be established by any reference to »cer- tain publicly available orders of meaning« such as »a medieval treatise on the ranks of cherubim and seraphim.«6 For the modernist period charac- teristically lacks such agreed-upon, entirely external, public reference points. 2 C. Taylor, The Sources of the Self: The Making of the Modern Identity (Harvard University Press, Cambridge, M.A., 1989), p. 491. 3 Taylor 1992, p. 90. 4 Taylor 1992, p. 89. 5 Rilke, p. 151. r' Taylor 1992, p. 84. 54 Certainty, Belief, and Rationality Just as Baudelaire wrote about correspondences without anyone in his cul- ture, unlike anyone in the Renaissance era, believing any more in the exist- ence of such things, so too Rilke writes of angels without anyone in his cul- ture, unlike anyone in the Medieval period, believing any more in such doctrines. Consequently, in order to understand Rilke's use of a word like »an- gels,« one must forego any appeal to strongly external canons. Instead, we must work out the specific details of how Rilke uses this word in the chang- ing constellations of his own poetic practices. Rilke's meaning here is most directly tied to Rilke's language not to the cosmos. Thus, understanding Rilke's meaning entails understanding his uses of language.7 We may say accordingly that a strong contrast holds between an under- standing of interpretation, an externalist interpretation, that derives its basic canons from outside a particular human sensibility, and an internalist inter- pretation, one that does not. This second kind of interpretation is requisite for understanding much modernist poetry including Rilke's. »Rilke's 'order' can become ours only through being ratified afresh in the sensibility of each new reader. In these circumstances, the very idea that once such an order should be embraced to the exclusion of all the others - a demand that is virtually inescapable in the traditional context - ceases to have any force« (86-87). This distinction between internalist and externalist interpretation can now be linked with an earlier claim of Taylor's, namely that Rilke's poetry can be construed paradoxically as a non-subjective subjectivism. The poet- ry is subjective in the sense that it can exhibit a subjectivation of manner, yet not subjective in the sense that it does not entail a subjectivation of matter or content. Indeed, as the »thing poems« in Rilke's earlier New Poems show, much of Rilke's poetry is not subjective in content at all.8 The apparent paradox then in the passage from Rilke's Ninth Elegy, where the poet seems 7 »What could never be recovered,« Taylor summarizes, »is the public understanding that angels are part of a human-independent ontic order, having their angelic natures quite independently of human articulation, and hence accessible through languages of disruption (theology, philosophy) that are not at all those of articulated sensibility« (p. 89). 8 Relying on studies of Romantic poetry, Taylor summarizes his general view here in a phrase: »where formerly poetic language could rely on certain publicly available orders of meaning, it now has to consist in a language of articulated sensibility« (84). For Rilke's New Poems f rom 1907 and 1908 see the translations by E. Snow in two volumes (North Point Press, San Francisco 1984 and 1987 respectively). See also the less familiar but closely related poems from 1899 to 1906 in Rilke's The Book of Images, tr. E. Snow, revised edition (North Point Press, San Francisco 1994). 55 Peter McCormick to exhibit at the same time both a subjectivism tout court and a subjectivism that after all is non-subjective - this paradox dissolves. The passage is first interpreted internally as exhibiting a subjectivity in manner but not in con- tent. But, while the manner of expression necessarily refers to the subjec- tive sensibility of the poet, the content of these expressions refers, at least partly, to something external to the poet's sensibility. Correspondingly, the passage is then interpreted externally. The something external we may identify as we see fit, either with the »things« the passage mentions, or »the Earth« which the passage apostro- phes, or »the world« in some more elusive sense. The essential point remains that the nature of the epiphany Taylor takes this passage to celebrate is not exclusively a function of »our action,« but of »a transaction between our- selves and the world« (1989: 482). One final point needs mentioning. The distinction between two kinds of self-reference enables Taylor to discriminate within the general modern movement of subjectivation two different species of subjectivism, one closed the other open. More fundamentally however, the present distinction de- rives from a still larger view. For Taylor thinks that the movement of subjec- tivation that characterizes modern culture is part of a major shift at the end of the eighteenth-century from representation to creation. »The change ... here goes back to the end of the eighteenth century,« Taylor writes, »and is related to the shift from an understanding of art as mimesis to one that stress- es creation... . it concerns what one might call the languages of art, that is, the publicly available reference points that, say, poets and painters can draw on... . But for a couple of centuries now we have been living in a world in which these points of reference no longer hold for us« (82-83). The movement of subjectivation then is seen as flowing from a prior moment when a crucial shift takes place in the accessibility of the major points of reference for attempts to interpret poetry and works of art. More specifically, Taylor wants to interpret modernist poetry, as here in his inter- pretation of Rilke's paradox of subjectivism, with the help of a general view on modern culture from the standpoint of what he calls subjectivation. But that movement cannot be understood without explicitating its connection with a radical shift at the end of the eighteenth century. I would like now to situate more precisely just what Taylor takes to be the source of subjectivation. I will go on to show that, pace Taylor, this source can be located in a certain understanding of the Lockean framework of belief. 56 Certainty, Belief, and Rationality 2. From Procedural to Instrumental Rationality What drives this philosophical interpretation of representative modern- ist poetry like Rilke's poetry of suffering derives from an antecedent story about the characteristically subjectivist orientation of the modern period. But just when this new story begins is not clear. At times Taylor clearly suggests, as we have just noted, that the mod- ern period begins with the late eighteenth-century shift from an aesthetics of representation to one of creativity. This comes to specifying the Roman- tic era as the starting point for the modern period. But Taylor speaks at other times of the modern period beginning with Cartesian philosophy and sci- ence. Here the shift is f rom late medieval, scholastic views, some of which are still operative in Descartes's work, to the early modern views that crys- tallize in the seventeenth-century scientific revolutions. In this case Taylor construes the shift mainly in the epistemological terms of a change in the understanding of reason. At first glance then we seem to have a strong inconsistency in account- ing for the emergence of the modern period. But we need to recognize here a certain complexity. For Taylor is concerned to specify not just the emer- gence of any one thing called »the modern period.« He explicitly rejects any claims to be attempting either historical explanation or intellectual histo- ry.!1 Rather, he is at work on discriminating different strands within the com- plicated story of modernity. For one of these strands, the Cartesian moment is paramount; for another, the Romantic moment - two strands then and not two stories. In trying to understand just which strand is paramount in these views about the modern period as a movement of a double subjectivation that flows from the end of the eighteenth-century, we need to see that late eighteenth century moment itself against the major background of Locke's philosophy. For, implicated in this major strand in the modern period is a particular understanding of reason and rationality. And this understanding is indissol- ubly tied to a new notion of the self, »the ungrounded 'extra-worldly' states of the objectifying subject« (1989: 175). This idea of the self as, in some strong senses, disengaged and separat- ed from the world, is taken to be characteristic of the modern period. For in the modern period the self is understood mainly in just these terms. The shift to the modern then is not to be described in terms of mimesis and cre- ativity. Rather, the shift concerns the nature of reason itself. In particular, the shift is f rom »the hegemony of reason as a vision of cosmic order to the 9 Cf. Taylor 1989, pp. 202, 305. 57 Peter McCormick notion of a disengaged punctual subject exercising instrumental control.«10 And this is why Taylor sees the conflation of the manner and matter of sub- jectivation leading to a reduction of reason to its instrumental and functional capacities only. In other words, instrumental reason is »the kind of ration- ality we draw on when we calculate the most economical application of means to a given end.«11 Taylor describes at some length the beginnings of the modern period in the Lockean terms of an emergent understanding of the »punctual self« and its exercise of reason. The key element in this story is, of course, its continuity with the »procedural notion of rationality« that Taylor has already linked earlier to Descartes's notion of a »disengaged subject.« With the shift however from subject to self, a corresponding shift ensues from procedural to instrumental notions of reason and rationality. The question is how to specify the nature of this second shift. The procedural notion of reason and rationality linked to the disen- gaged subject is taken as the major result of a converging series of histori- cal studies on the spread of neo-Stoical disciplinary ideals to different types of seventeenth-century institutions. What these studies have brought into focus »is the growing ideal of a human agent who is able to remake himself by methodical and disciplined action. What this comes to is the ability to take an instrumental stance to one's given properties, desires, inclinations, ten- dencies, habits of thought and feeling, so that they can be worked on, do- ing away with some and strengthening others, until one meets the desired specifications.«12 This notion of the disengaged subject becomes, through the practices of instrumental reason, a procedural self. And this procedural self is the »familiar modern figure« who gains control through progressively objecti- fying successive domains by neutralizing their antecedent normative claims. What comes to define rationality in this picture is the idea of a self that prac- tices correct methods or procedures for constructing beliefs and attaining knowledge (162-63). The picture is Lockean. The Lockean picture of reason and rationality, as Taylor wants us to understand the matter, is then a radicalization of the Cartesian procedural notion into an instrumental one. Citing different sections of Locke's Essay, Taylor argues that Locke does far more than simply reject some of the un- derlying epistemological assumptions of Cartesian rationalism such as innate ideas. Rather, Locke goes on to articulate a »profoundly anti-teleological view 10 Taylor 1989, p. 174. 11 Taylor 1992, p. 5. 12 Taylor 1989, pp. 159-160. 58 Certainty, Belief, and Rationality of human nature, of both knowledge and morality« (164-65). And he does so largely by carrying out his self-described tasks of clearing away, as he writes memorably in the Epistle, »the rubbish that lies in the way to knowledge,« of demolition, and, only then, reconstruction. The key to this articulation is Locke's adoption of an atomistic theory of the mind. Understanding is based upon the »quasi-mechanical« process of assembling the inert imprints left on the mind by the senses, »the build- ing blocks of simple ideas,« by processes of association into complex ideas. This process enables the independent self to reconstruct a reliable and sound foundation for beliefs. This is the picture of the self refusing to con- template things but assuming responsibility and authority to construct ra- tional beliefs and knowledge. On this view, Taylor writes, »rationality is above all a property of the process of thinking, not of the substantive content of thought« (168). More specifically, Locke takes rationality as a process of thought in the sense that rationality must be essentially understood from a first-person and not from a third-person standpoint. The rational self is one that self-reflex- ively disengages from »spontaneous beliefs and syntheses, in order to sub- mit them to scrutiny. This is something which in the nature of things each person must do for himself« (168). So reason and rationality acknowledge no other authority than that of the not just disengaged subject but of the thoroughly independent , objectifying, anti-teleological self. This disengagement, moreover, is radically linked to the self-reflexive- ness that issues in self-objectification. The self-objedification is what allows of the piecemeal yet progressive reformulation of both habits of belief and habits of action. And this reformulation in turn leads to a self-reformulation that is a self-creation, a self-identification no longer with habits but with the process of objectification (171). Both belief and action are now articulated with a sharp and unyield- ing focus not on »relations of natural fact« but on connections that »are determined purely instrumentally, by what will bring the best results, pleas- ure, or happiness« (171). The basic link is thus forged between an idea of the self as a pure »detachable consciousness« existing »nowhere but in this power to fix things as objects« (172), and an idea of reason and rationality that consists not jus t in the exercise of current procedures but in the instru- mental and functional constructions of objectifications. And this is the strand of radical self-reflexivity that is woven together with others to initiate, on Taylor's view, the modern period. Locke will go on to nuance the relations between instrumental and procedural understandings of rationality. Two major areas of course are 59 Peter McCormick those of morality and natural theology. And much attention needs direct- ing to the connections between Lockean deism and the development of the Enlightenment. In theology, for example, Locke will not only reject origi- nal sin; he will put reason rather than faith at the center of things. But then Locke will claim that »instrumental rationality, properly conducted, is of the essence of our service of God« (243). Here however the distinctive relation between procedural and the more radical instrumental rationality we have seen at work in the epistemology is somewhat blurred. Thus Taylor seems to go back on his earlier distinction when he writes in a theological context, »the rationality in question is now procedural: in practical affairs, instrumental; in theoretical, involving the careful, disen- gaged scrutiny of our ideas and their assembly according to the canons of mathematical deduction, and empirical probability« (243). But this foun- dation, when taken over into the larger epistemological and metaphysical contexts of Locke's quarrels with Descartes, is not sharp enough to capture Locke's radical differences with Descartes. To do so we need the idea of a radicalization of the procedural reason of the disengaged Cartesian subject into the instrumental rationality of the autonomous and objectifying Lock- ean self, »the new stance,« as Taylor himself puts it later on, »which Descartes inaugurates and Locke intensifies« (177). One now apparent gap in this story is the role of Montaigne's reflec- tions specifying the beginnings of the modern era. Despite his great emphasis on Descartes and Locke, Taylor does not overlook Montaigne. But however different in their understandings of reason and rationality, Descartes and Locke are for Taylor part of one strand only in the emergence of the mod- ern. And that strand unwinds from their opposition mainly to Platonic and later versions of Neo-platonism but also to Aristotelian philosophies. This opposition is what unites Descartes and Locke in their related but different attempts to articulate specific senses of reason's autonomy from nature and God in an expanded notion of the subject and the self. Despite his similar concerns with subjectivity and the turn inward, Montaigne is part of a separate strand. For Montaigne needs to be under- stood as part of the Augustinian world view rather than in opposition to Platonic notions of form or Aristotelian conceptions of nature. The Greek heritage of Montaigne is new scepticism; but the even greater influence is the Augustinian models of inwardness - a self-exploration rather than a dis- engagement from the self. Montaigne's is then a different form of self-re- flexivity than that of either Descartes or of Locke. »Rather than objectify- ing our own nature and hence classifying it as irrelevant to our identity,« Taylor writes, »it [this stance of radical reflexivity] consists in exploring what 60 Certainty, Belief, and Rationality we are in order to establish this identity, because the assumption behind modern self-exploration is that we don't already know who we are« (178). Interestingly, Taylor sees this second strand at the origins of the mod- ern finding expression in some of Rilke's work also. For the move to self- exploration in Montaigne issues in an experience of inner instability, uncer- tainty, and impermanence. And this exposure in turn give rise to »an accept- ance of limits.« However different the spirit of this acceptance - in Mon- taigne's case Christian and Epicurean, in Rilke's perhaps neo-Lucretian and almost pagan - Taylor interprets passages from Rilke's Second Elegy about the figures on the Attic gravestones in the terms of such an acceptance. Remember the hands, how weightlessly they rest,though there is power in the torsos. These self-mastered figures know: »We can go this far, this is ours, to touch one another this lightly; the gods can press down harder upon us. But that is the gods' affair.«13 Nevertheless, the spirit of Montaigne's self-exploration differs strong- ly from that of the self-disengagement of Descartes and Locke. Taylor sees this difference as one of both aim and method. For, unlike Descartes's, the aim of Montaigne's self-exploration is »to identify the indi- vidual in his or her unrepeatable difference, where Cartesianism gives us a science of the subject in its general essence; and it proceeds by a critique of first-person self-interpretations, rather than by proof of impersonal reason- ing ... its aim is not to find an intellectual order by which things in general can be surveyed, but rather to find the modes of expression which will al- low the particular not to be overlooked« (182). The one strand at the be- ginning of the modern then is, as we have seen, one of radical disengage- ment; the other, and very different, strand in Montaigne is one, just as rad- ical, of »engagement in our particularity.« Thus, without slighting Montaigne's role at the outset of the modern era, Taylor insists on emphasizing the other strand of »radical reflexivity,« the strand of disengagement of the subject in Descartes and the atomiza- tion of the self in Locke. For this is the strand that gives major form to both the inchoative modern ideas of reason and rationality. 3. The Origins of the Modern and the Transformation of the Self This detailed and persuasive picture of the emergence of the modern era in Locke's transformations of the Cartesian idea of a disengaged subject 13 Rilke, pp. 158-61; cited in Taylor 1989, p. 346. 61 Peter McCormick whose rationality is mainly procedural to a punctual self whose rationality is mainly instrumental is not the only picture contemporary philosophers have developed. Indeed, Taylor's own view is controversial. Consider a re- cent, largely implicit critique of Taylor's picture and especially the alterna- tive picture of the origins of modernity that generates this critique. In »Tradition, Insight and Constraint,« his 1992 Presidential Address to the Central Division of the American Philosophical Association,14 Nicho- las Wolterstorff carries several steps fur ther a reading of the origins of the modern era on which he has been at work for some years. The basis of this reading is an argued rejection of Hegel's understanding of the genesis of modernity. And Hegel's view, when filtered through the work of Max We- ber on the transformation of a religious view of the world into a disenchant- ment from which both strongly differentiated social and cultural spheres have emerged,15 is Taylor's inspiration for construing the origins of the modern in terms of a transformation in the notion of the self. Taylor's earlier works on Hegel, Wolterstorff thinks, strongly support this genealogy. For there Taylor first expounds Hegel's ideas, especially in the Philosophy of Right and Lectures on the History of Philosophy, that the »mo- dernity« of modern philosophy is to be understood in terms of the central role of subjectivity in modern philosophy. And this role is inaugurated in Descartes's articulation of a new conception of the subject that breaks with the previous era.10 Taylor centers this Hegelian story on the idea that Des- cartes's new subject is best understood not with respect to the putative con- nections between the subject and an external cosmic order, but as a »self- defining subject.«17 Descartes's cogito thus conceals a radically new idea of the self as in- dependent of any connection with the freshly discovered causal contingen- cies of the cosmic order, the modern self as self-defining subject. And, as we have already seen, it is this construal of the Cartesian self-defining subject with its associated ideas of reason and rationality that, according to Taylor if not Hegel, Locke radicalizes as a punctual self incorporating not jus t pro- cedural but instrumental notions of reason and rationality. 14 Proceedings of the American Philosophical Association, 66 (November 1992), 43-57. Cf. Wolterstorff's distinction between Locke's »descriptive epistemology« of knowledge and his »regulative epistemology« of belief in his »Lockean Philosophy of Religion,« in V. Chappell, ed., The Cambridge Companion to Locke (Cambridge University Press, Cambridge 1994), pp. 184-5. 15 Wolterstorff, pp. 55-57. 10 Cf. Wolterstorff's citations from Hegel, p. 55. 17 See Taylor 1989, p. 6; cited in Wolterstorff, p. 55. Further references in the text are to the Wolterstorff article. 62 Certainty, Belief, and Rationality Now Wolterstorff concedes that taking the critical measure of the many complexities in such a story, only summarized here very briefly, would in- volve a very extended discussion. In his most recent comments he confines himself to indicating one major line of argument only. Wolterstorff thinks Taylor »has lumped together two quite different ideas in his notion of the self-defining subject. One is the idea of the will as central to the self, the other is the idea of the self as autonomous« (55). Wolterstorff objects here not to the idea that subjectivity is character- istic of both societies in the modern era and the modern philosophies which to some degree are their reflections. Rather, he contests Taylor's Hegelian view that, since this philosophical shift occurs in Descartes's work, Descartes's work marks the origins of the modern. »1 submit,« he writes, »that the self was not yet the center of attention in Descartes's philosophy, neither the volitional self nor any other... . Descartes's project was the practice and grounding of scientia« (56). When viewed from the perspective of Descartes's own project, the central work was not, pace Hegel, the Meditations on First Philosophy, especially Meditation Two, but the Principles of Philosophy. Far from being the crucial figure in the genesis of the modern, Descartes is »a transi- tional figure« whose project, whatever features we now take as modern, was not distinctively modern« (53) .1S The basic justification Wolterstorff offers for this unconventional read- ing is that the historical, cultural, and theological contexts of Descartes's era, its pervasive and growing »cultural anxiety« that Wolterstorff sees as mark- ing the beginnings of that era in which we ourselves still live, »goes virtually unacknowledged in Descartes« (53). More specifically, »the fact that these cataclysmic events at the founding of modernity go almost unacknowledged in his philosophy,« Wolterstorff argues, »indicates that the project in which he was engaged did not call for their acknowledgement« (53). Thus, very much like Toulmin's careful reflections on the social and theological con- texts of the shift from the humanistic revolution of the 1580's to the scien- tific revolutions of the 1640's, Wolterstorff insists on making room for his- tory. As the Latin text of his writings with their repetition of the crucial terms shows, Descartes's project was the construction of scientia by starting »with certitude and by demonstrative inference [to] arrive at certitude« (53). Whatever the modern interest in Descartes's therapy of doubt versus the medieval »dialectical appropriation of the textual tradition« as a way of car- rying out the project of scientia, »nothing at all of modernity is reflected in 18 The point may be ironic when we recall the similar strategy Taylor adopted with respect to Montaigne. 63 Peter McCormick his reason« for recommending this therapy (54). Scientia, its grounding and its construction, remains the center of Descartes's project not the doubting subject. The subject as the self, as Wolterstorff sees things, is indeed at the ori- gins of the modern. But these origins are not to be located in Descartes's philosophical projects but in those of Locke. »Only when we come to Locke,« Wolterstorff writes, »does the self occupy center stage« (56). This self however is taken not as the autonomous self but as »the de- ciding self,« the deciding subject whose center is volition not autonomy. »There is not a whiff in Locke of the autonomous self. Of the deciding self, there is much; of the autonomous self, nothing. The reality with which our Reason puts us in touch is a reality laced through with meaning, for it is a reality [for Locke's Puritan vision] created by God and under the law of God. It is when that conviction decays in philosophers after Locke, that the no- tion of the self as autonomous emerges« (56). So Wolterstorff contests Taylor's claims that Descartes's understanding of the subject marks the beginnings of the modern on the grounds of the nature of rational understanding, its role in Descartes's philosophical project, and the lack of any intrinsic connection between that philosophi- cal project and its cultural contexts. By contrast, his alternative claim is that Locke's project arises directly out of the need to fashion a philosophical response to the cultural anxieties of his time. That project places the self at the center, and construes the self and its at tendant conceptions of reason and rationality not, pace Taylor, as autonomous and self-defining, but as volitional only. »My case for Locke as the first of the great modern philosophers,« Wolterstorff summarizes, »is based on the claim that he was the first to ad- dress himself head on to this anxiety [i. e., »the breaking apart of the moral and religious traditions of Europe into warring, partisan fragments«] at the founding of modernity. It is this that made him place the volitional self on center stage. The growing disenchantment of the world would lead yet lat- er philosophers to make of that volitional self an autonomous self. Only then does the self in modern philosophy become a self-defining self« (57). 4. Lockean Belief and Substantive Reason Now, if these are the main lines in a recent critique of a contemporary Hegelian reading of the genesis of the modern, what are the salient details of Wolterstorff's own reading of Locke? How does Wolterstorff justify his counter-claims not so much about Descartes, but about Locke, specifically 64 Certainty, Belief, and Rationality his claim that Locke not Descartes centerstages the self not as autonomous but as volitional only?19 Wolterstorff reads Locke's philosophy as fundamentally responsive to the pervasive »cultural anxiety« of his times. This anxiety was the new and widespread incapacity to resolve religious, political, and moral quandaries at a time of radical historical revolutions. And this incapacity itself is seen to follow from the outspoken disavowal of the efficacy of the medieval dia- lectical appropriation of both texts and traditions for resolving conflicts of such large extent. Neither individuals nor societies, in Locke's view, could turn any long- er to texts and traditions for the resolution of their deepest problems. »The practices which European humanity had cultivated for resolving its moral and religious quandaries,« Wolterstorff writes, »had proved profoundly unsatisfactory: the textual tradition was fractured and seen as fractured, and scriptural interpretation was riven with controversy« (49-50). These were the basic contexts of Locke's philosophy. In responding to these contexts, Locke inaugurates the modern. The basic issue at stake in the anxieties of the times, as Locke under- stood the matter, came to the question of how one was to form one's most fundamental beliefs in a reliable way for grounding one's life rationally at a time of such fragmentation and strife when both appeals to tradition and reappropriation of texts were rightly discredited. »How are we to go about forming our beliefs, especially on matters of morality and religion, when the old way has been rendered irrelevant« (45)? Locke's response was to formu- late a philosophical project centered on »overcoming« the profound cul- tural anxieties of his own times. That project, on Wolterstorff s as well as Taylor's reading, was essentially epistemological. But, unlike Taylor, Wolterstorff sees the governing idea here as a radically novel »proposal for a new practice of belief-formation, a new doxastic practice - not only, though especially, on moral and religious mat- ters« (45). Much more specifically, Locke's new doxastic practice was a »foun- dationalist procedure - for doing one's best to bring it about, for proposi- tions which one does not know, that one believes them if and if only if they are true« (49). The key to this reading of Locke's epistemology as a foundationalist procedure rather than just a foundationalist criterion for warranting or ''' See especially N. Wolterstorff , John Locke and the Ethics of Belief (Cambridge University Press, Cambridge 1996), pp. 227-246, and his Divine Discourse: Philosophical Reflections on the Claim that God Speaks (Cambridge University Press, Cambridge 1995), pp. 261- 266. 65 Peter McCormick entitling reliable beliefs is a very sharp focus on Book IV of Locke's Essay, especially on the end of Book IV. By contrast, in developing his reading of Locke's »quasi-mechanistic« epistemology, Taylor relies almost exclusively, as most contemporary philosophers still do, on Book II. For, while referring to The Epistle, Book I and even Book IV, Taylor centers his own account on Book II. But Wolterstorff argues persuasively that the aim of the Essay is to pro- vide the details of Locke's proposal for a new doxastic procedure. And these details, after the endless preparations, Locke provides in Book IV only where he specifies the nature of belief formation and especially its governance. Moreover, these details come at the end of Book IV once Locke has shown that, unlike belief, genuine knowledge is »short and scanty« (cited in W 45). So Locke's epistemology is not only a response to the cultural anxieties of his times; it is mainly a matter of providing an analysis of procedures for ensuring the proper governance of belief formation rather than an analysis of the nature of knowledge itself. His epistemology thus is mainly a regula- tive epistemology of belief rather than a descriptive epistemology of knowl- edge. In Book IV, knowledge while related to belief is very importantly dis- tinct from belief. Knowledge is direct awareness of facts. These facts are ideas which are »directly present to the mind.« And the awareness, insight into, and apprehension of the »connection and agreement, or disagreement and repugnancy, among one's ideas« is knowledge. Belief on the other hand does not consist in direcdy apprehending facts but in »taking some proposition to be true« (47). Belief may accompany knowledge, but need not. » ... belief is often present where knowledge is lacking. And even when belief does accompany knowledge [this is the nor- mal situation], it remains distinct from it. Taking a proposition to be true is distinct from apprehending some fact« (47). Knowledge for Locke remains very limited. And that is because the familiar difficulties with his very narrow construal of knowledge - difficul- ties with memory, for example, which show that some knowledge is of facts of which we are not presently aware - »forces Locke to choose between fac- tual awareness and certitude of belief as definitive of knowledge, and he chooses certitude of belief« (46). The result is that there are few facts of which we can have knowledge in this sense of certain belief. Consequently, the Cartesian project of scientia is radically circumscribed — «some pure mathematics, some pure ethics, some logic; but no scientia of nature« (46). Yet the compensation for this strictly limited scope of knowledge is the far 66 Certainty, Belief, and Rationality larger scope of belief. Thus, for Locke, where we lack knowledge, as in most of the spheres of life, we are to rely on beliefs. But of course belief has its own difficulties. Wolterstorff summarizes these difficulties succinctly. »Some knowledge, on Locke's official definition, is awareness of some facts, where there is not fact, there is not knowledge. Belief is not so fortunate. It is true that where there is no proposition, there is no belief. But all too often the proposition believed is false« (47). In these cases Locke thinks we have an obligation to find good reasons for our be- liefs, a doxastic obligation. What turns out however as alone capable of making reasons »good reasons« is beyond our capacities - «nobody could possibly, for all her beliefs, do what is necessary to hold them for good rea- son. «20 Unable then finally to provide a prima facie criterion for what is to count as having »good reason« to believe a particular proposition, Locke turns instead to making a proposal for a new way to form one's beliefs. »Rath- er than proposing a general criterion of entitled belief,« Wolterstorff holds, »Locke was urging a reform in the doxastic practices of European humani- ty - a new way of using our indigenous belief-forming dispositions. His claim was that following his proposed practice amounts to doing one's best, for some proposition, to bring it about that one believes it if and only if it is true. And it was his conviction that, for each of us, there are some matters of such high 'concernment ' that we are obligated to try seriously to do our best« (48). In short, Locke's response to the cultural anxieties of his times was to propose a radically new picture of which beliefs can be taken as entitled when knowledge is so starkly circumscribed. Wolterstorff takes the cardinal passage in Book IV to be xv. 5.This pas- sage, which I will cite in a moment in recapitulation, is about the specific nature of this new doxastic practice, of what doing one's best comes to. The passage can be taken as comprising three rules for mediate (not immedi- ate) beliefs, that is, for doing one's best in believing a proposition when one does not know the corresponding fact. These rules Wolterstorff calls the rule of evidence, the rule of appraisal, and the rule of proportionality. When applied successively, they run as follows: 1. The Rule of Evidence. »Acquire evidence for and against the propo- sition such that each item of evidence is something that one knows and such that the totality is a reliable indicator of the probability of the proposition.« 2. The Rule of Appraisal. »Examine the evidence one has collected so as to determine its logical force, until one has 'perceived' what, on that ev- idence, is the probability of the proposition.« 20 Wolterstorff, p. 47; cf. Locke, Book IV. xx. 2. 67 Peter McCormick 3. The Rule of Proportionality. »Adopt a level of confidence in the proposition which is proportioned to its probability, on that evidence« (49). These rules are taken as the gist of the key passage mentioned above which now can be quoted for comparison: »the mind if it will proceed rationally, ought to examine all the grounds of probability, and see how they make more or less, for or against any probable proposition, before it assents to or dissents from it, and upon a due balancing of the whole, reject, or receive it, with a more or less firm assent, proportionably to the preponderancy of the greater grounds of probability on one side or the other.« Thus, Locke's proposal is a foundationalist procedure for arriving at reliable mediate beliefs in matters of high »concernment« to which one is rationally entitled. Or, as we noted in Wolterstorff's summary formulation earlier on, a »foundationalist procedure - for doing one's best to bring it about, for propositions which one does not know, that one believes them if and only if they are true« (49).21 This reading of Locke's understanding of reason is not without serious consequences. For while clearly situating Locke in »that long tradition, be- ginning with Plato, of thinking of Reason as a faculty of awareness, of ap- prehension, of insight, of 'perception, '« Wolterstorff is careful to note that Locke's conception of the scope of reason and »the ontological status of its objects« is very narrow. But, however narrow reason's scope, reason is more than procedural and instrumental. For, when scrutinized in the contexts of Book IV, reason, as one of the sources of genuine knowledge, allows direct awareness of ap- prehension of facts. But, more specifically, the facts reasons apprehends are »the logical relations among propositions« (46). This reading is what ena- 21 Wolterstorff has modified his formulations of these rules in the 1994 paper I cited in note 19. There he speaks of »four principles« rather than of »three rules.« For the sake of comparison, the »four principles« read as follows: »Principle of Immediate Belief. One is to believe something immediately only if it is certain for one—that is, only if one knows it« (p. 182). »Principle of Evidence. One is not to believe something mediately until one has acquired evidence for it such that each item of evidence is something that one knows and such that the totality of one's evidence is satisfactory« (p. 183). »Principle of Appraisal. One is not to believe some proposition mediately until, having satisfactory evidence, one has examined that evidence to determine its logical force and one has 'seen' what, on that evidence, is the probability of the proposition« (p. 183). »Principle of Proportionality. Having determined the probability, on one's satisfactory evidence, of the proposition in question, one ought to adopt a level of confidence in it which is proportioned to its probability, on that evidence« (p. 184). 68 Certainty, Belief, and Rationality bles Wolterstorff to return to the initial quarrels we saw with the Hegel-Tay- lor version of the origins of the modern. Without mentioning Taylor Wol- terstorff clearly has him in mind when he states categorically, »those who say that Reason, in Locke, is purely procedural, not substantive, cannot have read Book IV« (46). 5. Substantive Reason and Doxastic Obligations When we review carefully the details of these two competing accounts, we need to distinguish immediately a series of different issues. Moreover, among these different issues we also need to specify just which considera- tions are most important for our concerns with the emergence in the mod- ern era of a particular understanding of the rational nature of interpreta- tion. One major issue the accounts address is the origins of the modern era, especially the respective roles of Descartes and Montaigne. The difference was a matter of several generations only, although these generations saw the shift from a humanistic revolution in the work of Montaigne to a scientific one in that of Descartes. Here, the difference is between the philosophical project of Descartes and, again almost two generations later, the philosoph- ical project of Locke. For Taylor the modern era begins with Descartes, for Wolterstorff only with Locke. (In neither case, I would add, is there room for Pascal's general rationality of probability.) This first issue however comes to something more than simply a disa- greement about where the modern era begins. For, as we have seen, that decision itself seems to be a function of some antecedent notion of what constitutes the modern. For Taylor, what is most characteristic of the mod- ern era is its implication of at least three different strands in the understand- ing of what it means to be an individual subject, self, or person.22 And since Taylor believes that Descartes disengages the first of these strands, a new understanding of the nature of the subject, he strongly leans to situating the beginning of the modern in Descartes's philosophy. Wolterstorff however has a different understanding of what character- izes the modern, especially under the sign of modern philosophy. For him a philosophy begins to be modern not just when it first polarizes its concerns in terms of the subjective, but when it articulates its basic philosophical project as a response to both individual and societal anxieties about the reliability of basic beliefs. Since, on Wolterstorff's reading, Locke is the first 22 Taylor 1989, pp. 177-78, 185, 305, 308. 69 Peter McCormick philosopher after the breakdown of the medieval syntheses to structure his philosophical project in just this way, Locke's philosophy and not Descartes's is seen to mark the beginnings of the modern. This first issue then comes down to adjudicating a disagreement about the satisfactoriness of two dif- ferent but equally implicit understandings of the modern. But I do not think either Taylor or Wolterstorff has given us the requi- site detail to allow of any satisfactory appreciation of why they think of the modern in such different terms. Nor is it evident that the notion of »the modern« has the same extension for each one. Wolterstorff clearly specifies that he takes »the beginnings of the modern« to refer to »the beginnings of modern philosophy.« Taylor is not so explicit. At times he seems to take the crucial phrase here in the same narrow sense as Wolterstorff does; more often however he seems to have in mind something that includes philoso- phy but is broader, namely the beginnings of modern culture. So, even were one to have more detail about their respective qualifications for these dif- ferent antecedent understandings of »the modern,« actually deciding be- tween the two would not clearly be a matter of deciding between accounts of the same thing. Regardless of these matters, just where we situate the beginning of the modern does not bear directly enough yet on our interests here in under- standing more fully the nature of rational interpretation. For whether we take the modern era to begin with either Descartes or Locke (or indeed following Toulmin with Montaigne or with someone else), we would still have to specify just which elements of their respective projects are the pertinent ones for elucidating the background understandings of rational interpre- tation. So, whatever the respective merits of each of these accounts of the origins of the modern, we have no central interest in trying to explore the matter far enough in order to judge the issue. A second issue that separates these two accounts however is more in- triguing. This issue has to do with how we are to understand Descartes's philosophical project. For Taylor, this projet turns on a revolutionary con- strual of the rational subject which finds its canonical formulation in Med- itation Two, whereas for Wolterstorff the Cartesian philosophical project turns on refining the late medieval notion of scientia whose canonical for- mulation is to be found in the Principles. Descartes's philosophy for Taylor is a new philosophy of rational subjectivity, whereas for Wolterstorff Des- cartes's philosophy is a transitional philosophy of »science.« At first glance then it seems we are being asked to decide between com- peting readings of Descartes's philosophy. The issue we might think comes to the now familiar concerns as to whether we are to read Descartes's phi- 70 Certainty, Belief, and Rationality losophy of science f rom the perspective of his metaphysical and not just epistemological reflections on the cogito, or, conversely, whether we are to read those metaphysical reflections from the perspective of the scientific works and, especially, of the mathematics. Here again however I think we need to be circumspect. Wolterstorff writes from a different standpoint than Taylor. Like Taylor, he has a much larger story to tell than just the story of how philosophy develops from the later medieval period through the Renaissance and into the onset of the seventeenth-century scientific revolution. For he carries his story beyond Descartes and Locke to Hume and then on to both Kant and especially Reid before he takes up Taylor's affinities for Hegel (cf. 50-52). More important- ly, as he writes Wolterstorff has Taylor's story before him. Yet Wolterstorff's own larger story, while independent of Taylor's account (cf. his earlier work), partly depends on a protracted critique of Taylor's reading of Descartes. In this sense Wolterstorff is offering an alternative account of Descartes to the one already before us, namely Taylor's and of course behind Taylor's - and this is part of the great interest in the contrast — to Hegel's own reading. This peculiar asymmetry I think explains in part some of the differences between these two accounts of Descartes. That difference is more finely- grained than just where one's reading of Descartes is to start, whether from the mathematical or from the metaphysical. Rather, the distinction is be- tween Wolterstorff's attempts to characterize the rationality of a philosoph- ical project of Descartes (as he does with Locke) and then judge whether the project exhibits what he takes to be characteristic of modernity, and Taylor's attempt to highlight the new construal of rationality as one only of several salient features of Descartes's philosophical xuork. When viewed from the standpoint of his philosophical project, Des- cartes's philosophy shows both strong continuities and strong discontinui- ties with later medieval notions of philosophy's task as scientific. To confirm this point we need only underline such central terms in Descartes's Latin writings as »idea.« But when viewed from the standpoint of Descartes's met- aphysical reflections on the nature of the cogito, then Descartes's work as opposed to his philosophical project shows a striking and perhaps revolu- tionary innovation in what was initially an Augustinian theme of inwardness which then filtered through Montaigne's neo-Stoical humanism. This move does not reconcile the two readings, nor does it try. What is of concern rather is to see more clearly what is the issue that separates these two different readings of Descartes. In the light of this recognition that Tay- lor and Wolterstorff are reading from different standpoints, I think we can now see that the focus has to be sharpened if we are to appreciate the dif- 71 Peter McCormick ferences intelligently enough. Regardless then of whether there are com- pelling philosophical arguments available for preferring one standpoint on Descartes's philosophy to another, we have the more specific question of whether Wolterstorff has gotten the description of Descartes's philosophi- cal project historically and conceptually right, and whether Taylor has gotten the description of Descartes's notion of the subject right. Just here it seems to me is where the discussion comes together. For by reason of his commenting on Taylor Wolterstorff moves to Taylor's own chosen ground of the cogito and challenges Taylor's reading of the cogito. So the issue now seems to be whether Taylor's construal of the cogito as comprising a distinctively novel (if not modern) understanding of the ra- tional subject as self-defining and in that sense autonomous squares with both Descartes's writings and the conceptual implication of his central claims about the rational nature of the cogito. Yet Wolterstorff does not address this issue as directly as he should. For, while agreeing that »the self in general, and the deciding self in particular, was moved to the center of attention in early modern philosophy,« he goes on to assert, without direct argument, that »the self was not yet the center of attention in Descartes's philosophy, neither the volitional self nor any other« (56). And of course this leaves Taylor free to underline his concern with the rational subject rather than the self, and indeed with Descartes's cogito as implying substantive claims about the subject's definition of itself as in at least some sense from its place in a network of orderly cosmic rela- tions. The second issue then, in a similar but different way to the first, is not joined. And once again there can be no rush to judgment on the »correct- ness« of the respective accounts of Descartes. Besides the first issue of where we are to situate the beginnings of the modern and this second issue of whether Descartes philosophical work if not his project inaugurates in the metaphysics of the cogito a revolutionary understanding of the rational subject as self-defining, a third issue needs attention. This issue turns on the details, not of Descartes's modernity or his theme of rational subjectivity, but of Wolterstorff's apparent counter-claim that »only when we come to Locke does the self occupy center stage,« not the autonomous self but the deciding self, »the idea of the will as central to the self« (56; 55). Suppose, in the light of the considerations that have surfaced in dis- cussing the first two issues, we set aside the historical claim here that in ear- ly modern philosophy the self first comes to center stage in Locke's philos- ophy. Whether or not this is the case - and we would certainly have to scru- tinize the implicit distinctions all along between subject and self - we need 72 Certainty, Belief, and Rationality not investigate. For what is more important is the complex second claim that the self in question here is not the autonomous self but the deciding self. And specifically what holds our attention is not the fate of the autonomous self and just where historically that notion arises. Rather, the key idea here is the construal of the deciding self as a self in which the central component is the will, what Wolterstorff also calls »the volitional self« (57). One important feature of this idea is that it arises in the course of Wolterstorff s development of his alternate account of the origins of the modern. There is no direct engagement here however with Taylor's read- ing of Locke. For Taylor's reading, as Wolterstorff points out repeatedly, centers on Book II of the Essay where Locke's accent falls strongly on the nature of knowledge rather than on Book IV where the accent falls on be- lief. And while Wolterstorff certainly wants to reject Taylor's construal of the Lockean self as mainly a punctual self that includes a creating self-defining autonomy, he nevertheless is much more interested in detailing his own reading of just what the Lockean volitional self comprises. Rather than di- rectly opposing to Taylor's picture of a Lockean punctual self his own sketch of a Lockean volitional self, Wolterstorff is at some pains to show exactly where the volitional self fits in his understanding of Locke's philosophical proposal to institute a new doxastic practice. We recall that Wolterstorff takes Locke as »urging the institution of a new doxastic practice« rather than »proposing a criterion of entitled belief« (48). What needs underl ining here however is not the idea of a doxastic practice whose successive steps require careful formulation in terms of pro- cedural rules, but the associated idea of a doxastic obligation. This notion can be seen at work in a passage from Book IV that Wolterstorff cites. »He that believes, without having any reason for believing, may be in love with his own fancies; but neither seeks truth as he ought, nor pays the obedience due his maker, who would have him use those discerning faculties he has given him, to keep him out of mistake and error...«23 The seeking of truth Locke writes of here is one of those matters of high »concernment« human beings have to worry about. So the first point here is the existence of an obligation. But this obligation concerns specifically the truth of certain matters, a truth which requires the use of discernment, a non-evident truth therefore of which one cannot have certain knowledge in all instances. And hence one which requires the discernment of just which of those among our beliefs are reliable, a discernment of opinion, of judg- ment, of doxa. Thus the obligation here is doxastic obligation. And although 23 Locke, Book IV, xvii, 24; cited in Wolterstorff, p. 47. 73 Peter McCormick these obligations will vary somewhat, Locke certainly reserves a msyor place for moral and religious matters as well as political and educational ones. Still, what exactly is the nature of this doxastic obligation? As Wolter- storff reads Locke, persons have an obligation to go through certain evalu- ative procedures for testing, supporting, and ensuring the reliability of just those moderate beliefs that are central to the issues of »concernment.« Locke believes, we remember Wolterstorff asserting, that »for each of us there are certain issues of such 'concernment ' ... as to place us under obligation to try seriously to do our best to bring it about that our beliefs conform with the facts on those issues« (48). What the doxastic practice comes to is an explicitation, such as Wolterstorff's three rules, of »doing our best.« The doxastic obligation concerns the other of the two major components here, the »trying seriously.« This serious trying, I think, is where the idea of the rational self comes in as volitional, as having at its core the notion of the will, as deciding. For Locke's doxastic obligation of its nature requires a response. And that re- sponse for Locke has to be a matter of the person »trying seriously,« that is to say both wanting reliable beliefs and deciding to secure them. What is striking of course is that this wanting and deciding comes to constitute the very nature of the Lockean self. But this claim, whatever its merits, remains independent of the prior claim that the doxastic practices Locke's philosophical project is designed to elaborate and institutionalize themselves derive from doxastic obligations. And these in turn make sense for Locke - given his antecedent Puritan convictions about the existence of God, the nature of creation, the obedience human beings owe to their cre- ator, etc. - only on the assumption that part of what it means to be a ration- al person in such a context entails wanting and deciding on the securing of reliable beliefs about high matters of »concernment.« In this light we might then construe the third issue that emerges from the confrontation of two related but very different accounts of the origins of the modern as whether indeed there is a difference between Wolterstorff's notion of the Lockean self as primarily a volitional self and Taylor's notion of the Lockean punctual self as a self-deciding-subject. This may be a dis- tinction without a difference. We see the room for different emphases here - is the Lockean ego mainly a self or a subject? But does Taylor stress the »deciding« element in his expression »self-deciding-subject« enough to bring the expression's significance close enough to Wolterstorff's talk of a »voli- tional self«? If so, then where exactly is the issue? If not, then granted the difference; but then, how interesting is this difference philosophically? 74 Certainty, Belief, and Rationality Whether we choose to pursue these matters or not, the third issue here about the links between the nature of the rational self, the putative exist- ence of doxastic obligations, and the procedures of doxastic practice bring us to the most important issue for our concerns with the nature of rational interpretation in the modern era. That issue turns on the connections be- tween the understanding of reason at work in the details of a Lockean dox- astic practice such as Wolterstorff has expanded them, and the understand- ing of reason at work in the Lockean account of knowledge such as Taylor works out. We recall Wolterstorff's earlier expostulations: »Those who say that Reason, in Locke, is purely procedural, not substantive, cannot have read Book IV« (46). I turn to this matter in the next and concluding sec- tion. 6. Rationality and Interpretation Besides the three issues we have been looking at so far, a fourth and central issue between Taylor and Wolterstorff concerns the nature of rea- son and rationality at the beginnings of the modern era. Initially, this issue might be put in the form of a question. Given the radically influential role of early modern reflection on the development of Enlightenment, Roman- tic, and late modern understandings of interpretation, what are the central connections in that reflection among reason, rationality, and interpretation? More specifically, how is reason to be understood inside the revolutionary Lockean framework of belief? Very generally, Taylor thinks that the movement of Western philosophy from the Greeks to the early modern era involves a gradual and thorough- going change in the understanding of reason. Thus, were we to focus these matters temporarily on one theme only such as the theme of »self-control,« Taylor thinks that we would find »a very profound trans-mutation, all the way from the hegemony of reason as a vision of cosmic order to the notion of a punctual disengaged subject exercising instrumental control« (1989: 174). The important point to underl ine in this general view is the opposition between reason as a vision and reason as a procedure. Taylor goes on to make these general comments more specific. Thus he reminds us that for both Plato and the Stoics neither introspection nor self-examination were important for understanding the cosmos and human goals. The classical ideal was rather a matter of taking reason as »the cru- cial capacity ... of seeing the order - in the cosmos (for Plato) or in the pri- ority of human goals (for the Stoics)« (174). 75 Peter McCormick The modern ideal in Descartes and especially in Locke, according to Taylor, is completely different. For reason is to be understood as a capacity for »disengagement and objectification.« And this capacity requires both self- examination and introspection. »The modern ideal,« Taylor writes, »... re- quires a reflexive stance. We have to turn inward and become aware of our own activity and of the processes which form us. We have to take charge of constructing our own representation of the world ... we have to take charge of the processes by which associations form and shape our character and outlook« (174-75). So a modern ideal of reason as reflexively construct ing an o rder through disengagement and objectification is opposed to a classical ideal of reason as non-reflexively seeing an order through focus on the essence and the substance of things. Where reason on the one model provided ac- cess to the way things are, on the other reason is preeminently an instrument, a set of procedures, at the service of a disengaged self. Taylor's contrast between the classical and the modern centers on his extended discussions of Locke's epistemology. We recall that Taylor describes that epistemology largely although not exclusively in terms taken from Book II of the Essay. There Locke proposes that knowledge be understood in the contexts of a radically new picture of the mind. This picture makes room for neither the essences and the forms of things nor for innate ideas of rea- son. Rather, the Lockean picture is to be taken as a reified and quasi-mecha- nistic portrait of the most fundamental atoms or elements of mind. These elements are the simple ideas which are the residue of the senses' »quasi- mechanical« impact on the mind. In turn, out of these simple ideas com- plex ideas are then constructed again by a process of association understood in quasi-mechanical terms (167). The result is a construction that provides a new and reliable representation of both the world and ourselves to replace the old and unreliable traditional representations. The key characteristic of this process, as Taylor expands it, is the inde- pendent, self-responsible nature of reason, »a notion of reason as free from established custom and locally dominant authority« (167). And although Taylor is careful to reserve a place for mathematical truths, deductive knowl- edge, and even the rules of »probable evidence« in Book IV, he character- izes the nature of reason mainly in the contexts of the atomistic and mech- anistic account he describes Locke developing in Book II. Taylor now identifies this Lockean picture as »the modern conception of reason.« And he summarizes his understanding of this conception with the expression »reason is procedural« (168). Reason is procedural in the sense that it is neither a seeing nor a completing of any independent order 76 Certainty, Belief, and Rationality of things. Procedural reason is rather a construction of »a picture of things following the canon of rational thinking.« Locke takes these canons differently from Descartes. Yet for Taylor Descartes shares with Locke the modern conviction that following some such procedures of reason is the most reliable way of getting as close as we can to the way things are. Reason then is construction on the basis of detailed canons and procedures for the processes of thinking. A central property of these processes is rationality. »Rationality is above all a property of the proc- ess of thinking, not of the substantive content of thought« (168). But, on Taylor's view, what exactly do these thinking processes comprise besides the following of certain canons and procedures? These processes comprise a »radically reflexive« procedure that »essentially involves the first- person standpoint. It involves,« as we noted earlier, »disengaging from my own spontaneous beliefs and syntheses in order to submit them to scruti- ny« (168). Moreover, these thinking processes also comprise the particular constraint that they must be carried through independently in the sense of being carried through »radically and intransigently exclusive of authority« (168). Finally, the aim of these thinking processes is, as we have seen, to construct and to remake our fundamental representations, but to do so in such a way that the central connections among our ideas »are determined purely instrumentally, by what will bring the best results, pleasure, or hap- piness« (171). In short, Taylor takes the modern understanding of both reason and rationality to arise, whatever the role of Descartes, mainly from Locke's ar- ticulation of reason as preeminently procedural, instrumental, and non- substantive. And, as we have already noted briefly, Wolterstorff disagrees. The question now comes to what exactly is the disagreement about? I think the disagreement is about whether reason and rationality at the beginnings of the modern era are in any way »substantive.« Taylor says no, Wolterstorff says yes. But do they disagree about the same thing? To answer this question I do not think that we have to mount a major study of all the nuances that make up the somewhat different Lockean accounts of knowl- edge and belief in Books II and IV. Rather we need to specify the role of reason in the different spheres of knowledge. Both Taylor and Wolterstorff, I take it, share the view that reason is »substantive« in the domain of our knowledge of some logical relations, of mathematical truths, and of deductive truths. For in these different realms Locke thinks that reason can be based upon direct awareness. Reason here, as Wolterstorff writes, is a faculty of direct awareness, »of apprehension, of insight, of 'percept ion '« (46). So, despite appearances, when Wolterstorff 77 Peter McCormick goes on to assert very strongly that »those who say that Reason, in Locke, is purely procedural, not substantive, cannot have read Book IV,« I don' t think either he or we should include Taylor. For while insisting on the preeminent- ly procedural character of Lockean reason, Taylor does not hold that Lock- ean reason is »purely« procedural. This brings us then to the much different - broader, variegated, and difficult - domain of belief. On the basis of the close examination we have been conducting here, I think Taylor wants to hold that the practices of thinking in the domain of belief when explicitated show Locke's understand- ing of reason here, if not in the other domains, to be procedural and instru- mental. Since however even Wolterstorff agrees that the extent of this do- main is very large in Locke's philosophy as a consequence of construing knowledge so narrowly in terms of immediate direct awareness, one can justify the generalisation that Lockean reason is procedural. Wolterstorff on the other hand is most impressed by two things. First, that even in the initial construals of knowledge in Book IV, knowledge is defined in terms of both insight and facts. »Let the main point ... not be missed,« he writes; »knowledge is insight; knowledge is awareness, or appre- hension, of facts« (46). Reason as a faculty for yielding such insight into facts is, at least in this sense, more than procedural; it is substantive. The nature of these facts may of course be controversial. As Wolterstorff remarks, »if Plato were to read Locke's account of the scope and ontologi- cal status of the objects of Reason's awareness, he would feel profoundly claustrophobic« (46). Still, Locke's claim is that reason does more than pro- vide procedures for connecting simple ideas instrumentally; reason also provides access to facts. The second point that strikes Wolterstorff is the extent and subtlety of the analyses of doxastic procedure in Book IV. Before he gets to attempting an articulation of just what the rules are for making our beliefs about mat- ters of concernment reliable, however, Wolterstorff acknowledges an impor- tant distinction in Locke's account. The distinction is between immediate and mediate beliefs. Knowledge for Locke, we remember, is »direct appre- hension of facts,« whereas belief is »taking some proposition to be true« (47). A belief is immediate when one takes some proposition to be true on the basis of the accompanying direct apprehension of the fact that corresponds to the proposition; otherwise, the belief is mediate. Locke's proposal of a new doxastic practice is a way of discharging one's doxastic obligation to attend to those cases only where beliefs are mediate; these are cases where either belief does not accompany knowledge, or where »the proposition believed is false« (47). 78 Certainty, Belief, and Rationality So in those cases where belief does accompany knowledge and where belief is immediate I don' t think Taylor can have any quarrel with includ- ing these cases among the others as instances being more than just proce- dural. For, just as in the cases of our knowledge of logical truths, mathemat- ical truths, and deductive conclusions, so too here, reason is substantive. Reason in each case is a faculty which yields insight into facts. We arrive then at the nub of the disagreement - the Lockean concep- tion of reason in the most extensive and important cases of establishing the reliability of one's mediate beliefs about matters of high concernment. And here, once again on the evidence of the extensive details in Wolterstorff's account of Locke's new doxastic practice, I think we have to conclude that reason is more than procedural and instrumental. It is true that Locke explicitly says that »the mind if it will proceed ra- tionally, ought to examine all the grounds of probability...»(IV. iv. 5). And Wolterstorff himself talks about unravelling »the practice Locke has in mind into three rules to be applied in succession« (49). But, as his formulation of the rules of evidence, appraisal, and proportionality explicitly shows, two of these formulations involve knowledge. »Acquire evidence ... such that each item of evidence is something that one knows... .« And »examine the evidence ... until one has 'perceived« ... .« (49). Thus we have a foundationalist procedure. But this procedure is clear- ly directed to establishing »for propositions which one does not know, that one believes them if and only if they are true« (49). Reason then as the fac- ulty that enables us to arrive at this knowledge, through the practices of this new doxastic procedure, is certainly procedural and instrumental in the case of mediate beliefs. But reason is also much more - Lockean reason is also substantive. More generally, we may conclude by saying that rational interpretation at the beginnings of the modern era cannot be construed as procedural and instrumental only. Rather, reflection on the complicated issues at stake in two contrasting accounts of the origins of the modern shows that an under- standing of reason as substantive is just as much part of the origin of the modern as an unders tanding of reason as instrumental and procedural. Construing the rationality of interpretation in the modern period then in the overly narrow terms of procedural reason is unjustified. To the contra- ry. Our understanding of interpretation needs to retrieve the genuine role of a substantive construal of reason if it is to merit any thoughtful claim on the nature of modern interpretation. And that substantive construal must make new room for a much more nuanced account of the complex relations holding between certainty, belief, and rationality. 79