Book Review INFORMAL LOGIC XV.I, Winter 1993 The Intellectual Virtues and the Life of the Mind by Jonathan L. Kvanvig JULIA DRIVER Brooklyn College, CUNY K vanvig, Jonathan L. The Intellectual Virtues and the Life of the Mind. (Savage, MD: Rowman & Littlefield, 1991), xii+ 198pp, $39.50. In The Intellectual Virtues and the Life of the Mind Jonathan Kvanvig attempts to spell out what a virtue oriented epistemol- ogy would be like; what its strengths and weaknesses would be. He sees this project as an extension of a similar project in eth- ics which attempts to develop ethical theo- ry in terms of virtue evaluation, and focuses on agent rather than act evaluation. His overall strategy is to show how, even given the current 'Cartesian perspec- tive' around which much modern episte- mology revolves, intellectual virtues can playa crucial role (the 'Cartesian perspec- tive' is the perspective which focuses on concepts like 'justification' and 'knowl- edge' with respect to particular beliefs and propositions). This in turn is crucial to his argument that such a perspective is an in- adequate one when it comes to accounting for "epistemology as a whole" (ix), be- cause it fails to provide an account of all that we find intellectually excellent and desirable-such as the the intellectual vir- tues. Thus, the goal of the book is a rather ambitious, though not uncommon, attempt at undermining Cartesian epistemology. However, Professor Kvanvig's approach in trying to achieve this goal is original. The first part of his project relies on de- veloping an analogy between epistemology and ethics, so that he may exploit work done in virtue ethics in developing his own position. One obvious connection is that virtue ethics criticizes theories like Utili- tarianism and Kantianism for being too much focused on particular actions and problems. Ethics should be reoriented in such a way as to provide methods of evalu- ating persons as a whole, and their lives. But, as Kvanvig notes himself, there are a variety of ways in which one could devel- op a virtue ethics. One could be an elmina- tivist and argue that notions of 'obligation' and 'duty' and 'forbidden' should be purged from ethics and replaced with the richer vocabulary of virtue. Such a strategy is, however, " ... extreme, implying that standard moral notions are to be excluded from one's moral theory .... in order to justi- fy [this claim], one would have to show that the notions in question are incoherent, never instanced, or out of order in some other way." (p. 3) It is far easier to argue that virtue evaluation is somehow funda- mental to ethics, and that notions of obli- gation etc. are to be understood in terms of virtues. Kvanvig does not commit himself to specific versions of virtue ethics. He only wants a rough understanding of its form to function as a guide to the develop- ment of a virtue epistemology. Even though Kvanvig thinks, ultimate- ly, that virtue epistemology fails, he argues that his book will demonstrate how virtues do play a role in epistemology. This in turn, he argues, will lead to a total revision of traditional epistemology_ His conclusion 74 Julia Driver will turn out to be that the traditional Cartesian perspective cannot give us an ad- equate account of "the cognitive life of the mind." Thus, this perspective must be abandoned in favor of one that can. How- ever, other than offering a few brief sug- gestions about what this new perspective would be like, K vanvig defers further investigation of ii. Like the virtue ethicists, virtue episte- mologists find their source of inspiration in Aristotle. The difference he outlines be- tween the classical modem epistemolo- gists, such as Locke, and Aristotle are the following: the classical modern writers, when discussing the nature of knowledge itself, do not appeal to a person's capaci- ties, "Instead, the nature of knowledge is grounded in the evidential relations be- tween beliefs or the propositional contents of belief...." (p. 15), whereas the Aristote- lian epistemologist characterizes knowl- edge as the product of characteristics of the mind. It is exercising one's intellectual ca- pacities which constitutes, somehow, knowledge. So, in order to characterize a particular belief as knowledge, on this view, we first evaluate the capacity that gave rise to it. If the capacity or character- istic of the person is accurate, then the be- lief passes the test. An example of this might be perceptual acuity. [f a belief is due to this feature of a person, then it passes. Of course, there is a tremendous varie- ty of ways for the view to be spelled out, and one of Kvanvig's primary tasks in the book is to flesh out these varieties, and dis- cuss their problems and inadequacies. Space precludes a thorough discussion of the alternatives here; however, they are meant to reflect recent trends in epistemol- ogy and ethics. One type of virtue episte- mology he rules out is motivational virtue epistemology which holds that justified or 'properly' based beliefs need not merely be caused by evidence. Rather, it holds " ... that the causal route must be described as going through some of the intellectual virtues to belief.. . .it is because a belief is classified correctly as, say, a perceptual be- lief that it is justified doxastically, and not because it was caused to be held by some body of evidence. t. (p. 40) On this view a belief is justified if the cognitive character- istics of the individual, which are responsi- ble for the belief, count as intellectual virtues. This account calls for an internal constraint on determining when a belief is properly based. This constraint will ex- plain " ... how a belief passes from being merely propositionally justified to being doxastically justified." (p. 44) This con- straint is what K vanvig has problems with because he feels that it will not adequately accommodate the justified beliefs of small children and animals who lack an aware- ness of, for example, whether or not their beliefs are quite perceptive, or due to their diligence. Certainly in the case of small children, it would seem that the virtues have not yet been developed. After reject- ing motivational virtue epistemology K vanvig considers and rejects two other formulations of virtue epistemology- eliminative and reductive virtue epistemol- ogy. Elminative virtue epistemology can it- self take a variety of forms, some more radical than others. The one of most inter- est here would be a partial form, much like the partial form of eliminative virtue ethics which holds that the work done by abstract concepts of duty and obligation can be done with richer virtue concepts, such as honesty, integrity, and so on. Thus, the ab- stractions may be done away with in the interests of parsimonious theory construc- tion. Kvanvig rightly points out that even this partial form of virtue ethics is far too extreme. It would be a mistake to jettison useful concepts such as ·obligation.' Analogously, the corresponding version of virtue epistemology is too extreme. Harder to criticize is the reductive virtue epistemology, which K vanvig finds the most plausible, though still false. Can the traditional notions of justification and/or knowledge be better understood as attaching to intellectual virtues? K vanvig argues not. For space considerations I wiJI only consider here the version of reductive virtue epistemology with respect to knowl- edge. On this view, it must be shown that intellectual virtues are somehow required for the agent to possess knowledge. Some connection can be established, because in order to know the agent has to have proper- ly functioning cognitive equipment. How- ever, this connection cannot form the basis for any kind of necessary and sufficient condition for knowledge based on virtue. In any given instance of perceptual acuity, for example, there may still be room for er- ror. Further, the perceptually inept may on a given occasion attain knowledge. So while a connection can be established, it is not one strong enough to base a reductive virtue epistemology on (p. 107). After discussing these various virtue epistemologies, and why they fail, Kvanvig attempts to come up with at least a definition of 'intellectual virtue.' The fi- nal version of the definition, which he ar- gues can weather all of the difficulties posed for previous definitions, is: V6: C is an intellectual virtue of S of kind K =df (i) c is a contributory member of a maximal and nonredundant set of nomo- logically possible and independent charac- teristics R that is necessarily such that were S to exemplify each member of R, S would be disposed toward epistemically warrant- ed belief, (ii) C is an epistemically signifi- cant characteristic of S, and (iii) the strength of the disposition in question is notable for kind K. of which S is a member. (p. 140) This requires significant unpacking. The first condition simply locates the virtue within a class of characteristics such that intellectually ideal person possessing such characteristics would be disposed to justi- fied belief. The second condition is intend- ed to rule out trivial characteristics (like 'having fingernails') as being intellectual virtues. The third condition specifies that a virtue must be an excellence relative to a Review of Kvanvig 75 kind of being-and S is that type of being. This means that the trait, for it to count as a virtue, must be 'pretty good' or somehow better than normal, for that type of being. He doesn't mean to use a purely statistical notion of what it is for a trait to be superior or abnormally good-determinations of superiority are made relative not only to actual populations but to hypothetical ones as well. Thus, if all people, just by some accident or nuke happened to be meticu- lous, let's say, that trait is still a virtue be- cause relative to possible people the trait is still unusally good. This condition he ad- mits is quite vague (p. 129), and I feel that this is one area where K vanvig could have been clearer in spelling out his views. But I think an analogy with artifacts may make the point clearer. Suppose that I design a car that will normally get 33 miles per gallon of gas. Due to the fact that the first parts used were extremely well made, all of the test cars get 45 miles to the gallon. These cars are superior, even though in the actual population of cars they are the norm. This is because judgements of their quality must be relativized to the potential cars-the ones that I plan on building in the future, let's say. One feature of this definition which he devotes a great deal of space to defending is that it represents a corporate rather than an atomistic understanding of virtues. The intellectual virtues, as dispositions towards justification, cannot be understood as such atomistically-that is no one particular virtue is connected to justification. Rather, a trait is an intellectual virtue as part of a collection, or set of traits which as a whole are conducive to justification (as opposed to truth). K vanvig feels that he needs to make this move in the definition in order to avoid problems faced by an atomistic con- ception " .. .in which each virtue individual- ly aims at either the good or the truth ... " (p. 121). This is because some intellectual virtues-just as some moral virtues--can be used for 'bad' ends. In the case of the intellectual virtues, they may be used in 76 Julia Driver such a way as to not be aimed at the truth, or justification. To use K vanvig's example-a person may be methodical, which is intellectually admirable, even though he is not aiming at the truth or justi- fied belief, because. perhaps, he is 'de- fending the party line.' (p. 121) The parallel with virtue ethics is that K vanvig would maintain that courage, for example, can be used for a bad end, and still be a virtue. This runs counter to a popular view about virtue-that a trait is only truly a virtue when it aims at a good end; when aimed at a bad end, courage does not operate as a virtue. Given that epistemology cannot be successfully based on virtue evaluation, how is it that the intellectual virtues are important to epistemology? Kvanvig early on rejects the option that virtue evaluation plays no role in epistemology as far too radical. This leaves him with two options: (i) the traditional epistemological option, where the importance of the virtues is un- derstood in terms of their being conducive to knowledge and justification, where this in turn is understood by reference to the "time-slice" conception of epistemology embraced by the Cartesian perspective; and (ii) the cognitive ideal option, which maintains that the intellectual virtues are important not solely because of their con- nection to knowledge and justification; rather, " ... the having and displaying of the intellectual virtues are themselves taken to be distinct and separate aspects of the cog- nitive ideaL" (p. 150) Thus, what makes them valuable is not simply that they have certain effects-they are intrinsically valu- able as part of living a good life 'of the mind.' This view would resemble the view in virtue ethics that virtues are valuable 'for their own sake' and not simply for the sake of the goods that they bring about. K vanvig argues that the Cartesian perspec- tive cannot give us an adequate account of the good intellectual life, since it focuses on time slices of persons, and not whole persons. On the Cartesian view, the good life is constructed by 'gluing together' the time-slices to form a whole. However, an epistemology that makes room for intellectual virtues has certain su- periorities over the Cartesian perspective. The traditional view of epistemology- that the goal is to seek the truth, and avoid error or what is false-fails to account for other things that we strive for in account- ing for intellectual excellence. The time- slice perspective fails to account for certain things that we want to know: "We want to know which ways of organizing, structuring, or arranging this cognitive life will be most fruitful from an epistemic point of view." (p. 165) The Cartesian per- spective advocates, in Kvanvig's view of it, that the individual focus on distinct time- slices " ... and whether knowledge or justifi- cation is possessed at each such time-slice, and let the totality of that life get generated by cementing together these time-slices." (ibid) This perspective also fails to explain things like why we value the breadth and depth of ones knowledge as opposed to its mere truth. Further, the traditional view fails to explain why certain features of the- ories are desiderata-like simplicity. It is not at all clear that simplicity is correlated with truth. K vanvig explains that these things display or mirror 'intellectual virtu- osity' (p. 184) and thus are aims of the life of the mind, even though these aims are not recognized by the Cartesian perspec- tive which simply focuses on clearing up notions like 'justification' and 'knowl- edge' with respect to discrete beliefs and propositions. A focus on intellectual vir- tues takes us from the abstract to the con- crete. This is a move many may find familiar from virtue ethics. The charge against traditional theories is that they are too abstract, and focus on accounting for the rightness and wrongness of actions ab- stracted from the concrete. A virtue ethics, on the other hand, is supposed to remain in the concrete. We evaluate persons as a whole, based upon the features they exem- plify as a whole. We may evaluate actions too, but not in isolation from the qualities of the individual performing the actions. This outline of Kvanvig's project does not do justice to the arguments and consid- erations he presents in its favor. Neverthe- less, I feel that I can make a few critical comments based on what I have pointed out about his account. My first is that I find the analogy between virtue ethics and vir- tue epistemology confusing. I didn't get a very clear picture of what exactly a virtue ethics is from his characterization on pp. 1-3. Usually the point of using an analogy is to try to make one side of it clear by comparison to the other side, which is in- dependently clear to those to whom one is presenting the analogy. Kvanvig-by try- ing to articulate a virtue epistemology based on an analogy with a not very clear virtue ethics-has an extremely difficult task to perform in this regard. I also have problems with the defini- tion of 'intellectual virtue' itself. My main problem is with condition (iii)-that the trait be notable for the kind. I find this too high a standard. Suppose that honesty, for example, was the norm for humans and this was not due to some accident. It is still a moral virtue, because of the contribution it makes to human flourishing. Likewise, suppose that intellectual honesty was the norm for humans-it is still a virtue, be- cause of the contribution it makes to intel- lectual flourishing (however that is to be understood). In this respect the definition seems to narrow. It also seems too broad, because I don't see how, in principle, it would rule out physical traits which must be conducive to intellectual flourishing. Biologists often point out that certain physical traits of human beings facilitated the development of rationality: ego the opposable thumb, standing upright. Of Review of Kvanvig 77 course, these traits are normal for human beings, so K vanvig could rule them out with condition (iii). However, you could certainly imagine someone, lets say, who had a particularly dexterous opposable thumb which enabled her to manipulate experiments, etc., in particularly creative ways. Is this an intellectual virtue? I think not. K vanvig could, of course, say that in- tellectual virtues must be features of psychology only. Then the problem is ex- plaining why this must be the case, and is not just an ad hoc condition. The conclusion of the book was not as radical as I had anticipated. The traditional perspective needs supplementing, certain- ly. But this falls short of transformation. The Cartesian can point out that intellectu- al virtues are worth studying as traits con- ducive to knowledge (or warranted belief, etc.). However, how we figure out whether or not someone is disposed to warranted belief is to figure out what a warranted be- lief is. We need to be able to evaluate those things in order to evaluate the agents that embody them. Kvanvig's project, however, is quite interesting because it extends a compari- son between epistemology and ethics to ar- gue for a new perspective in epistemology. Comparisons like this have been made before--e.g. there is noncognitivism in ethics and epistemology. Kvanvig's com- parison is both timely and informative, coming as it does when virtue ethics is re- ceiving great attention in the philosophical literature. JULIA DRIVER DEPARTMENT OF PHILOSOPHY BROOKLYN COLLEGE. CUNY BROOKLYN. NY 11210 0