Intellectualism and psychology

To begin with, I find Baehr’s insinuation against a desire for money problematic. Rather than being a ‘questionable end’, the acquisition of money is a very basic and, if you wish, honourable end, especially if it is motivated instrumentally. To realize our goals, we need to finance them, and to finance them, we need money in almost all cases.⁹ Without money, numerous ends that Baehr hesitates to set aside as ‘questionable’ will never be reached. Be that as it may, a second and more important objection can be levelled. I do not consider it to be particularly plausible that our attempts to gain information and form true beliefs are motivated primarily by our desire to reach the cognitive states of knowledge, understanding or wisdom for their own sake. Most of our epistemic urges seem to me to stem more instrumentally from our desires to obtain or realize certain non-epistemic goods. There are scarcely any non-epistemic goods that can be realized without knowledge, and that covers most of the cases in which we are engaged in the epistemic activity of inquiry. To realize our goals, we need knowledge, and to obtain knowledge, we need epistemic virtue. It is no indication of a lack of epistemic virtue if one acts virtuously primarily with these instrumental goals in mind. Scientists and saints may be primarily interested in knowledge for its own sake. Businesspeople, consumers and many others have a decidedly instrumental perspective on epistemic goods.

To be fair, a similar disdain for run-of-the-mill, instrumentally useful knowledge can be found in the writings of other virtue epistemologists. Discussing love of knowledge, Roberts and Wood, for instance, recommend the reading of such magazines as Atlantic Monthly or New York Review of Books.¹⁰ Nor is intellectualism the only difficulty we have to overcome in applying epistemic virtue theory more broadly. Another is that it remains somewhat unclear what precise mechanism secures the desired psychological orientation towards epistemic goods and bads. Epistemic virtues are claimed to determine a person’s psychological orientation towards epistemic goods and bads, and it is interesting to know how they accomplish this. Even though Baehr hints at various theoretical possibilities at various stages in his argument, he does not make an explicit connection to empirical work in psychology. Again, Baehr’s aim was not to develop a psychologically viable theory of epistemic virtues for business, so he can hardly be blamed for not having addressed this issue. Before applying an otherwise appealing theory to business, however, the problem of intellectualism and the problem of psychological mechanism must be solved first.

To solve the problem of intellectualism, I start with a view of epistemic virtues as contributing not so much intrinsically valuable personal intellectual goods, but rather instrumentally valuable epistemic goods as means to eudaimonia. In contrast to the personal intellectual worth view of epistemic virtue, I focus on instrumental epistemic value. To solve the problem of psychological mechanism, three further ingredients are needed: a notion of virtue as motivating and enabling people to act virtuously; a view of the sort of actions to which epistemic virtues apply; and a view derived from behavioural economics about what happens when epistemic virtues motivate and enable people to perform epistemic actions. I cover the first two ingredients in this chapter. I turn to behavioural economics in the next chapter, which examines individual epistemic virtues in more detail.

According to Baehr’s view of personal intellectual worth, people are intellectually good or better as far as they possess a positive orientation towards what it is intellectually good to have, become or do, and possess a negative orientation towards intellectual bads. Baehr makes clear that he adheres to the view that personal intellectual worth is largely dependent on a person’s being intrinsically motivated to reach epistemic goods. One loves knowledge, wisdom and understanding for their own sake, and despises ignorance, false belief and irrationality because they are bad in and of themselves.

Intrinsic or instrumental value?

It is not necessary, however, to attach intrinsic value to epistemic goods and intrinsic disvalue to epistemic bads. Such a position, as I have said, undermines the applicability of epistemic virtue theory. Reaching epistemic goods and avoiding epistemic bads is as important for businesspeople, for instance, as it is for those seeking to enlarge their personal intellectual worth, but unlike the latter, businesspeople have instrumental reasons to seek knowledge. True beliefs rather than false beliefs will further the development of new products and services. Beliefs based on evidence rather than mere speculation or guesswork will provide a firm foundation on which to implement business strategies. Understanding the characteristics of a particular market and its participants allows businesspeople to respond most effectively to market pressures and consumer demand. Wisdom, finally, allows managers to take a long-term perspective and to address the spiritual or religious concerns of their employees more adequately.

The personal intellectual worth view takes the intellectually good person as its point of departure; the approach I advance here starts from a person seeking not only an intellectually or epistemically good life, but a good life as a whole. Such a person also has a positive orientation to knowledge, understanding and other epistemic goods and a negative orientation to epistemic bads, but only as far as is necessary for the realization of non-epistemic goods that the good life requires. Looking at epistemic goods as means to eudaimonia entails that epistemic goods without such instrumental value are not the things that an epistemically virtuous person aspires to by necessity. No need exists to gain these epistemic goods whenever the knowledge, wisdom or understanding does not play a part in something else with value. This may sound like a severe restriction, but this view has considerable advantages here over the personal intellectual worth view.

For a start, it does not presuppose that epistemically virtuous people avoid epistemic goods if they are not instrumentally useful. The view is neutral with respect to these issues and one may still seek to follow epistemic pursuits for their own sake, unless they obstruct living well. Moreover, most knowledge is instrumentally useful in one way or another, so the restriction does not reach as far as one might think. It is not true, however, that any sort of epistemic behaviour sanctioned by way of its contribution to personal intellectual worth is acceptable from the perspective of instrumental epistemic value. This is explained by a second difference from Baehr’s view. Focusing on instrumental epistemic value allows a more ready explanation of the fact that the knowledge, wisdom or understanding that virtuous people aim at depends on what they want to do with it; and similarly, instrumental epistemic value makes it easier to understand why the levels of certainty, justification and warrant that epistemic virtue requires depend on what people want to do with the beliefs. Unlike personal intellectual worth, instrumental epistemic value can provide a precise explanation of why people may virtuously settle for a lower degree of justification when they set up surveys for marketing purposes than when they conduct survey research as academic marketing researchers, or when they investigate the potential side effects of a new medication or the risks of nuclear energy. Deploying high justificatory standards even as a marketer is virtuous for Baehr, but may be an unvirtuous waste of intellectual and other resources once we look at it from an instrumental point of view. In contrast to Baehr’s viewing the standards of justification as dependent on the epistemic goods sought, I propose to make the standards of justification dependent on the non-epistemic goods for which knowledge, truth and justification are means to an end.

Courage

Courage illustrates how virtues enable. Imagine that at some point in time person S has not yet acquired courage. S is a coward at that moment. He sees a child drowning in a raging river. He has his mobile phone ready, so he can ring the emergency number 999 (let us call this action A), and were it not for his cowardice, he could have jumped in the water and attempted to rescue the child (action B), or he could have called one of the tourists nearby and asked them to help (action C). But being the coward that he is, he neither jumps nor calls but only rings 999. The coastguard arrives only barely in time. Shocked by the sight of the guards’ resuscitation attempts and the child’s suffering, S decides to work on his lack of courage, and succeeds. At a later point in time he has acquired the virtue, and as though he were to be put to the test, he again sees a child drowning. He waits no longer, searches for a place where he can safely jump into the water, swims out and rescues the child.

Courage has enabled S to rescue the child and to perform other actions requiring courage by removing what one could call internal obstacles to the performance of such actions. The treatment of epistemic virtues in the next chapter shows that internal obstacles often arise from so-called behavioural biases leading us to behave suboptimally with respect to investigative activities and other forms of belief formation. For the purpose of illustration, however, I focus on a non-epistemic instance of courage. In the beginning, S was blocked by his cowardice from performing actions B and C; his choice situation was a singleton set containing action A only. Acquiring courage subsequently led to the removal of these internal obstacles, as a result of which his choice set at the later point in time contained the actions B and C besides A.

Generosity

Courage illustrates how virtues enable. Generosity shows how they motivate. S starts as a Scrooge spending nothing on anyone – ‘Bah, humbug!’ Haunted by the three ghosts of Christmas, he decides that it is time for a change and acquires liberality. It works. On Christmas we see him treating his relatives, neighbours and his clerk’s family with generosity and concern. Liberality has not so much removed obstacles to performing generous actions. It is wrong, for example, to describe S as initially incapable of giving. Rather, he initially had no preference whatsoever for giving; he was miserly in wanting to keep his money. What the three ghosts did was make him change his preferences to become motivated to be generous.

Two things have to be said about this very succinct virtue theory. I must say something about the theory of the mean (virtues lie in the middle of two extremes) and also about the idea that most virtues motivate and enable. First, the examples discussed so far only consider one vice, that is, one extreme of the virtue. I looked at a move from cowardice to courage, not a move to courage from the other extreme, recklessness, nor did I consider a move to liberality from prodigality. These moves can be described in exactly the same way, though. Interestingly, showing this also covers the second point about motivation and enabling.

To start with recklessness, a reckless person is, one might say, imprudently brave, or ‘too courageous’. A reckless S seeing a drowning child dives into the river without thinking, and is injured because the water is too shallow. One might think that for a reckless person to learn how to steer the middle course between cowardice and recklessness requires a form of ‘disenabling’. On that count, S has acquired internal obstacles to the performance of reckless acts. A reckless S may well learn, however, to change preferences and acquire a motivation for more careful and considerate, but not cowardly, behaviour. Courage is a virtue that both enables and motivates. This is not true of all virtues. The move from the extreme of prodigality to the mean of liberality only constitutes a preference change (roughly, a change to give less and keep more). When an excessively generous person learns to acquire the right attitude to getting and giving, this does not mean disenabling certain prodigal actions but only demoting these actions in the preference ordering.

Even though I have not yet fully introduced the concept of epistemic virtue, this is the appropriate moment to say something about the way epistemic virtues motivate and enable their possessors to act in epistemically virtuous ways. I deal with the details in Chapter 3, but for now it is interesting to point out that research in behavioural economics has revealed a number of biases that human beings are prone to suffer when processing information. This research is increasingly also attracting the attention of ethicists.¹⁴ A significant number of these biases have to do with belief formation. I take the confirmation bias as an example. People vulnerable to this bias stick to their beliefs too tightly even when they face significant counterevidence. They have difficulties abandoning their beliefs even when the available evidence gives reason to believe them to be wrong. What epistemic virtues do is decrease the influence of these biases by motivating and enabling people to do what it is epistemically virtuous to do. To continue the example, epistemic justice makes a person open-minded with respect to various sorts of evidence, and epistemic humility leads people to be aware of their own fallibility, helping them to realize that their beliefs may be wrong after all. These virtues motivate and enable people to take counterevidence seriously and not to stick to their beliefs too long.

Doxastic voluntarism

Questions such as these land us, first, in a debate about doxastic voluntarism, which is the thesis that one is free to form any belief at will. Its strong form, direct doxastic voluntarism, implies that one can now decide to believe any proposition p, or at least a sufficiently large number of propositions. A weaker form, indirect doxastic voluntarism, maintains that one can bring it about that one believes any proposition p by soaking one’s mind, say, with books supporting p, and by avoiding what refutes p, in the expectation that in the end one has formed the beliefs aimed at. It appears at first sight that both positions give us too much power actively to influence our beliefs. Both seem to go against the fact that in typical cases of belief formation we feel compelled by the evidence to believe what we believe rather than free to believe anything at will. How can one decide to adopt the belief that York is the capital of the United Kingdom or that water is poisonous? What sorts of web sites or brochures can one consult in order to form that belief? Will they be available, in the first place?

For a long time in the history of epistemology either form of doxastic voluntarism had few adherents. Under the influence of recent work in philosophy on epistemic agency and doxastic responsibility, however, the popularity of certain versions of doxastic voluntarism is on the rise. Epistemic agency here stands for the idea that believing and knowing are activities that share many features with ordinary actions such as walking or making investment decisions, and if plausible, this idea might lend support to the view that freedom applies to belief and knowledge in the same way as it applies to actions. Doxastic responsibility, moreover, is the idea that we bear responsibility for what we believe and know, and that our beliefs and knowledge are not things that merely ‘happen’ to us. Since responsibility implies freedom on many counts, this idea again supports the thesis of doxastic voluntarism.

It is important to stress that epistemic virtues would not be particularly interesting to ethicists if believing and knowing were not sufficiently similar to ordinary actions that we have some freedom to perform and for which it makes sense to ascribe some responsibility to us. Whenever one has no influence on one’s beliefs at all, or cannot be held responsible for holding them, it is senseless to examine epistemic virtues from an ethical point of view. Although a thorough treatment of doxastic voluntarism is beyond the scope of the present chapter, it is important to see how belief formation and knowledge acquisition can be seen as forms of acting.

Investigation

Knowledge acquisition has to begin with what I call investigative actions. Prospective house owners visit web sites of banks to search for the best mortgage deals. Marketing researchers set up surveys and experiments. Rating agencies develop models to estimate the probability that a company cannot repay its debt. Surely these actions often do not lead to knowledge. Mortgagors may find out that they are borrowing money under different conditions than they initially assumed. Marketers may find that their results are not statistically significant. Rating agencies may fail to predict bankruptcy. These activities are not a sufficient condition for gaining knowledge, but they are certainly necessary.

Doxastic stance

On the whole, it is unproblematic to see these activities as genuine actions. They may require such commonplace things as going to the bank, handing out surveys, programming computers. What makes them special is that they are performed to provide evidence for or against adopting a belief. One might think that despite the fact that investigative actions themselves are quite ordinary types of actions, the need to include belief formation in the account dashes our hopes of viewing knowledge acquisition as an ordinary form of human action too. Yet it is relatively easy to see that belief formation is a form of acting. To approach the issue slightly formally, assume I want to gain knowledge about a proposition p. I carry out a number of investigative actions and now I ponder the belief to adopt. I can select one of three possible doxastic actions. I can, first, adopt the belief that p is true. I can, secondly, adopt the belief that p is false (disbelieve the proposition, as it is called). Thirdly, feeling that the investigative actions have not delivered sufficient evidence to justify either of these doxastic actions, I can perform a third sort of doxastic action and suspend belief. Suspending belief means that I neither believe that p is true nor believe that p is false. What exactly the attitudes of belief, disbelief and suspension of belief amount to psychologically is not essential to the analysis offered here. Believing a proposition may come down to giving full credence to it, but it may also mean accepting it, rather than its negation, for the sake of argument or for the sake of deliberation. This need not detain us here.

Justification

An extra condition has to be introduced. One performs investigative actions and subsequently selects a doxastic attitude justified by the outcomes of the investigative actions. Here things start getting tricky as normativity trickles into the argument. I approach these considerations from various angles. First, I consider the concept of justification from the perspective of the standard philosophical analysis of the concept of knowledge, which I very briefly introduce. Then, in the next section, I approach justification from the perspective of epistemic virtues. Together this leads to a defence of the view that epistemic virtues help people to perform investigative actions, and subsequent doxastic actions, in such a way that the latter are justified by the outcomes of the former. Moreover, I describe epistemically virtuous people as people whose goal it is to perform a third class of actions, epistemic actions, where an epistemic action is a combination of investigative actions and justified subsequent doxastic actions with the added condition that the doxastic actions are the right ones; this means that there is belief when the proposition is true, and disbelief when it is false.

The standard analysis of knowledge sees knowledge as a form of belief. A person knowing that Tegucigalpa is the capital of Honduras believes that Tegucigalpa is the capital of Honduras. Moreover, the belief has to be true. One cannot, for instance, know that Tegucigalpa is the capital town of Nicaragua, but one can hold that proposition as a belief. True beliefs are not necessarily knowledge. If, when quizzed about the capital of Honduras, one were to make a wild guess and answered Tegucigalpa, it would be wrong to claim knowledge; for to know it, one must have looked it up previously in an atlas, learned it from a geography teacher and so forth. One needs, that is, evidence that justifies the belief.¹⁵ How justification should be analysed is beyond the scope of the present discussion. Suffice it to say that epistemologists, quarrelling about the exact specification of the justification condition, agree that a bit more is needed than the belief being merely justified.¹⁶ Referring to what is needed in addition to mere justification with the symbol +, knowledge is thus sometimes characterized as true belief that is ‘justified+’. These technicalities are unimportant here, and in what follows I therefore use justification where others perhaps use justification+.

It is crucially important in the present context, however, to point out that both the truth condition and the justification condition contribute instrumental value. For truth this is plain. One is not helped much by the false belief that house prices will keep on rising if one considers buying a house that is excessively priced. For justification, however, it is less clear. In some sense, it even sounds totally counterintuitive. Assuming that a belief is true, why does it matter whether one adopted it on the basis of justifying evidence or just by sheer luck? The insight that justification does matter goes back to Plato. In the Meno Socrates considers the difference between a person S, who merely possesses a true belief about the way to the town of Larissa, and a person T, who possesses genuine knowledge about the way to Larissa. Suppose S embarks on a journey to Larissa, but after a while finds that the road is gradually going in what she believes to be the wrong direction. Then she will probably return. Moreover, imagine that T is going exactly the same way. Of course she too notices that the road is going in the wrong direction, but this does not cause her to give up her belief that she is going to Larissa. The evidence justifying T’s belief allows her to see it as a small detour only and, unlike S, she knows she will ultimately arrive in Larissa. From this observation Socrates derives the famous claim that knowledge is more ‘fastened’ or ‘secured’ than true belief; knowledge is more stable in that it ‘remains’ longer. One does not lose knowledge as easily as one loses true belief, and as a result, knowledge is to be preferred over mere true belief, even from the instrumental point of view that sees it as a means to an end only.

Truth and justification are not part of the concepts of investigative and doxastic action. One can perform investigative actions that do not yield evidence justifying any particular doxastic action. Moreover, even if inquiry provides evidence, one may still perform the wrong doxastic action. It is therefore useful to introduce a third concept to capture it all. An epistemic action is a combination of investigative and doxastic actions satisfying truth and justification conditions. The investigative part of the epistemic action may take the form of inquiry, observation, experimentation, asking questions and getting answers, and many other investigative actions. I use the plural here, but it may be convenient to speak about one large investigative action with respect to gaining knowledge about one single proposition p, where this large investigative action may contain many rather different sorts of inquiry. The doxastic part of the epistemic action is one of only two doxastic actions. Having researched p, one may decide one has not obtained sufficient evidence to support adopting a belief or disbelief; belief will then be suspended. This cannot be part of a genuine epistemic action, however, because the concept of epistemic action is intended to capture an action resulting in knowledge. As a result, the doxastic part of an epistemic action is either to believe the proposition or to believe its negation. Two conditions are in place, finally. The chosen doxastic action has to be the right one, because the belief or disbelief has to be true; and the chosen doxastic action has to be justified.

To the extent that philosophers disagree about the precise analysis of justification they may disagree about whether a particular person has performed an epistemic action in the sense in which I have introduced it here. This may look like a drawback of the concept, but the situation is no different from many other disciplines. Competent judges or legal scholars may disagree about whether a person has committed manslaughter or murder; competent accountants may disagree about whether to subsume an item under cost of sales in the profit and loss account, or to capitalize it on the balance sheet; and similar examples can be found in medicine, engineering and other fields. This does not make concepts of manslaughter, murder, cost of sales or capitalization meaningless or arbitrary.

In the context of a theory of epistemic virtues, however, an additional pressing issue arises, for some writers see the theory as entailing a novel view of justification. It all depends on how we read the following statement:

One way to read the statement is that it defines justification. Virtue epistemology, then, provides a novel view of justification in competition with existing views. Advocates of this position have to argue that virtue epistemology offers a genuine concept of justification that, it is hoped, does better than the incumbents. Another way to read the statement is to see the statement as expressing a conceptual link between a theory of epistemic virtues and a particular theory of justification. To defend this second reading, one has to show that whenever epistemically virtuous behaviour results in beliefs, these beliefs are justified in the sense of the particular theory of justification. The difference between the two positions is that the former gives a new definition of justification whereas the latter leads to new insights about an existing definition or theory of justification. Both approaches may be attractive. A theory of epistemic virtues that is serviceable to genuine normative issues that people face in their everyday lives must, however, have a broader orientation than knowledge alone. Financial markets are full of situations where people have to form beliefs that fail to qualify as knowledge by any standard. People typically do not know whether one insurance policy is best suited to their situation, not to speak about investment and other decisions with even greater degrees of uncertainty and scarcity of evidence. Add to that, in the next section I argue that there are also more fundamental reasons why virtuous behaviour need not result in knowledge in all cases.

The alternative view of justification that I develop is sensitive to these concerns and sees justification as coming in degrees. It sees epistemic virtues as contributing to increased degrees of justification. They maximize the likelihood that their possessor ends up with knowledge, but they do not guarantee knowledge. Exploring that view further is part of the aim of the next section. The idea is that notwithstanding the fact that epistemically virtuous people ceteris paribus perform epistemic actions and adopt beliefs that are genuine knowledge, they may depart from this ideal for several reasons. One is that insufficient evidence is available; another is that carrying out the investigative actions leading to sufficient evidence takes too much time, given the other goals one may have.

Courageous soldiers and evil demons

Weakening the success condition somewhat, Julia Driver proposes that epistemic virtues are character traits that systematically produce true beliefs, which means that they tend to produce more true than false beliefs in the circumstances in which they are possessed.¹⁸ This may help overcome the courageous soldier objection. Courage does not guarantee success, but across the board courageous people tend to be more successful than people who are not. Yet Jason Baehr has criticized consequentialist approaches to epistemic virtue such as Driver’s on the grounds that they founder on another objection, the evil demon objection. Baehr invokes Descartes’s evil demon manipulating the world in such a way that most of the beliefs we adopt are false, despite our using the highest standards of epistemic virtue. Baehr believes that the situation of the person manipulated by an evil demon can be described in only one of two ways. Either we must say that there is an ‘intuitively and reasonably pretheoretical’ view of epistemic virtue such that the person has ‘satisfied only one half of its requirements’.¹⁹ If we accept this description the person does not possess the virtue. Or we must say that the person is virtuous ‘in one legitimate sense’ of virtue (i.e., Baehr’s personal intellectual worth sense), but that in ‘another distinct but still legitimate’ sense (i.e., a consequentialist sense such as Driver’s) the person is unvirtuous.²⁰ Baehr then states that the second description is the most plausible. In the presence of an evil demon, consequentialist approaches fail to see virtue where Baehr himself sees virtue.

It is undoubtedly true that Driver’s approach fails to see virtue in the evil demon case; in a manipulated world epistemic virtue does not help people to produce more true than false beliefs. To state his case, Baehr could equally well have taken a real-life example (epistemic virtue does not inoculate consumers against deceptive marketing strategies) or have used Aristotle’s own example of Priam, king of Troy during the Trojan wars and most unsuccessful because of great misfortune. Aristotle did not call Priam unvirtuous, however, nor should we necessarily call deceived consumers so.²¹ Aristotle allowed for an empirical connection between eudaimonia and fortune. Good fortune may not be a necessary condition of the good life, but we do not describe as happy or living well a person stricken by serious setbacks.

This point can be adequately captured by a consequentialist approach to epistemic virtue. A simple model makes this clear. In order to gain knowledge about an issue, people select investigative actions. The best choice is to select activities that, given a number of constraints that I shall deal with shortly, maximize the probability of obtaining evidence concerning the proposition at stake. The likelihood of epistemic success depends on the likely consequences of the investigative actions. Suppose one has a choice either to make a phone call to a service desk or to consult a web site concerning information about, say, the departure of a bus. Given that the service desk is staffed with people getting real time information about delays and detours, the probability of epistemic success is greater if one calls the desk than if one consults the web site. But selecting investigative actions is only where things start. Having called the service desk, one can decide to perform a doxastic action and adopt a belief about the departure time, but all the same one can postpone this and also check the Internet, just in case.

A consequentialist might embrace the view that an epistemic virtue is an acquired disposition to select investigative and doxastic actions in ways that maximize the likelihood of obtaining knowledge. This definition will not do. To begin with, the investigative actions that people select depend, quite trivially, on what investigative actions they believe they can select. I may lack information about the availability of very good investigative actions. I am unlikely to call a service desk if I do not know I can call a service desk, in spite of this being the best way to gain knowledge about the bus. Or I may not have adequate beliefs about the likely consequences of investigative actions. I may have inaccurate beliefs about the ‘informational’ or ‘evidential’ value of investigative actions. I know I can call the service desk and check the web, but I do not know that the service desk is exquisitely staffed with friendly and competent people. That I do not select this action, however, is no indication of a lack of epistemic virtue. (At least, it is no such indication as long as the lack of knowledge about the availability of the investigative actions or their likely consequences is not a result of epistemic vice.) This is similar to the non-epistemic case. When soldiers lack knowledge about ways to access a building in which the enemy is hiding it is no indication of cowardice if they do not enter the building.

First improvement

A first improvement is to define epistemic virtue as an acquired disposition to select investigative actions (from the set of investigative actions of which the person is aware) and subsequently to perform doxastic actions (based on the outcomes of the investigative actions) in ways that maximize the likelihood of obtaining knowledge, where the likelihood has to be determined relative to what people believe about the likely outcomes of the investigative activities of which they are aware. But this is not sufficient. Suppose I am in a hurry to get home by bus. To do that I need to know when and where the bus departs. It does not foster my goal of getting home quickly to call the service desk several times and check the Internet in addition. Certainly doing so increases the likelihood of ending up with full-blooded knowledge, but I am probably going to miss the bus. The investigative actions I carry out and the amount of time I allow to pass before I decide which doxastic action to perform have to depend, for an epistemically virtuous person, on the goals that the knowledge is supposed to help realize.

This may seem to be a radical, perhaps undesired consequence of my adopting an instrumental view of epistemic value rather than Baehr’s personal intellectual worth view. That I opt for an instrumental view of epistemic value is motivated by a desire to apply the theory of epistemic virtue to less intellectualist pursuits, and this view unmistakably entails a different view of the value of justification. But there are independent and more theoretical reasons why the degree of justification that epistemically virtuous people aim at must depend on the broader set of goals they want to realize. This is no different from the non-epistemic case. Take precision. This is an ‘instrumental’ virtue. Good bakers and pharmacists have to be precise because it allows them better to realize their goals. One could think that the more fine-grained their weighing instruments are the better they realize this virtue. What precision amounts to for a baker, however, is crucially different from what it means to a pharmacist. A baker using a pharmacist’s tools to weigh ingredients is not precise, but overly precise, which is a vice.

Second improvement

What epistemic virtues help people to accomplish is, in my view, not so much to maximize the chance of obtaining knowledge; it is that they help people to select investigative actions and adopt beliefs in ways that maximize the likelihood of forming beliefs and gaining knowledge inasmuch as this is necessary for reaching other goals. This may seem a capitulation to lesser epistemic standards than most virtue epistemologists accept: for if knowledge is not needed, epistemic virtues do not force you to pursue it. This way of looking at things has two advantages, however. First, it offers a view of epistemic virtue that is more broadly normative than, for example, the personal intellectual worth view. It offers normative guidance in situations where a view that singles out knowledge as the sole epistemic ideal cannot. Unlike others, I defend a view that is sensitive to the fact that people often act on beliefs that are more or less justified, but are far from knowledge. One hardly ever knows the exact departure time of an aeroplane; things change too quickly there. Yet it makes a lot of sense to say that one can act more or less epistemically virtuously when it comes to acquiring information about departures. One might suspect that knowledge-centred virtue epistemologies will easily accommodate this objection and that recommendations for non-ideal circumstances can easily be derived; for instance, that in non-ideal circumstances one must do one’s utmost to come as close as possible to gaining knowledge. This, however, is unpersuasive when coming as close as possible to knowledge stands in the way of realizing other goals. One can certainly get more reliable information about departure times by checking the screens in the waiting area more frequently, but one may also want to have lunch or read a book. The view I defend allows for a possibility where epistemically virtuous people do not carry out any further investigative actions, knowing full well that they have not reached certainty, yet knowing too that they are fully justified in doing so because getting more evidence would hamper the realization of other goals. Such situations are certainly less prominent than it may seem. This point constitutes a fundamental difference from other approaches, however.

Secondly, the view defended here sees epistemic and moral virtues in a much more unified way. Ultimately, in order to accomplish the goal of getting home in time, one has not only to select a mode of transportation (including a departure time), but also to select a way of finding out which modes of transportation one can choose. One confronts a large decision problem with actions that are partly epistemic and partly non-epistemic; and epistemic virtues and non-epistemic virtues both help to maximize expected utility.

This should not be taken to entail that I agree with the view that epistemic and non-epistemic virtues are more or less similar.²² Aquinas, who had an intricate theory of epistemic virtues avant la lettre, observed that though the usual moral virtues aim at the good, epistemic virtues aim at the truth, even if aspiring to the truth is subservient to reaching the good.²³ This is not by itself a knockdown argument for a distinction between non-epistemic and epistemic virtues, particularly because following Aristotle, Aquinas’s approach is still very much wedded to the idea that epistemic virtues are innate. It might have been the case that even if epistemic aims are categorically different from other aims, similar methods are suitable to reach these aims. A second distinction between non-epistemic and epistemic virtues, however, is that they apply to different kinds of actions. Performing investigative actions is performing ordinary activities, and because one needs motivation or enablement here, epistemic virtues are like non-epistemic virtues. The courage I need to work as a war reporter near the battlefields is the courage the soldier needs too. It is, one could say, a form of non-epistemic courage set to use for epistemic purposes. Another form of courage is concerned with essentially epistemic issues. An example is the courage to ‘face the truth’, that is, to adopt a particular belief knowing that this is going to hurt one’s self-image. Portfolio managers require such courage if they have long placed all their confidence and money on particular investments but gradually obtain increasingly strong information that their investments are not going to pay off. Take Thierry de la Villehuchet, CEO of Access International Advisors, a feeder fund in the Madoff scam, who stuck to his belief that Madoff was quite all right even though very close colleagues possessed overwhelming evidence to the contrary. He said:

De la Villehuchet found it difficult to part company with his received views. Similarly, whereas non-epistemic temperance is typically concerned with tactile pleasures only, epistemically temperate people are sufficiently reticent not to adopt any belief on the basis of insufficient evidence. It is natural, I believe, to call these virtues (epistemic) courage or temperance rather than use a different terminology. They describe forms of courage and temperance after all.

Yet because they act on different sorts of actions, the way they motivate and enable is different from the case of the non-epistemic virtues. I tend to disagree therefore with virtue epistemologists putting too much stress on the similarity between epistemic and non-epistemic virtues. At the same time, the view defended here gives, I hope, a more convincing picture of how epistemic and non-epistemic virtues ‘interact’ when people make decisions, and in that sense my view stresses the unity of the virtues. To solve a decision problem, people need epistemic and non-epistemic virtues simultaneously. A view of epistemic virtue that requires too much from epistemically virtuous people risks being incompatible with the demands placed on decision makers by non-epistemic virtues. The risk is particularly present when the fact that epistemic virtues aim at the truth is seen as something that has non-instrumental intellectual value that may more readily clash with other values. When aiming at the truth is seen as ancillary to realizing the good, epistemic and non-epistemic virtues are more likely to place consistent demands on people.

Courageous villains

Before I turn to an examination of individual epistemic virtues in the next chapter, one potential worry needs to be addressed. It concerns the problem of the courageous villain. A villain by definition has an unvirtuous character. All the same, he can display courageous behaviour, and perhaps we can even attribute courage to him. Courageous villains pose a problem to virtue theorists, in particular those embracing the view that virtues form a unity. Robert Adams, for instance, has argued that a person may possess courage and use it in the pursuit of some evil ends, but that such a person lacks what he calls ‘capital V virtue’.²⁵ Jason Baehr rejects Adams’s approach, and defends a view according to which a courageous villain displays courageous behaviour even though he is not courageous:

Baehr, I believe, conflates here a question about whether the villain should be called courageous with a question about whether we should admire him for being courageous if he is. What matters for the villain to be called courageous is, first, whether he displays ‘courageous looking’ behaviour rather than, say, recklessness, rashness, faked fearlessness and so on; or in Baehr’s words, what matters is whether courage rather than something else is ‘manifested’ in his actions and attitudes. For the villain to be courageous, it is not sufficient, however, that he manifest courageous behaviour only once. It has to be a stable disposition for him to act in courageous ways. To ascertain whether the villain should be called courageous consequently requires us to probe deeper into his character. As long as we know the villain only very superficially, we can only describe him as manifesting courageous behaviour, not as a courageous villain.

But suppose we are sufficiently familiar with the villain to know that he has a stable disposition to act courageously. Then, I claim, if we admire him for being courageous, we admire him not for the actual way in which courage is ‘manifested’ in him, as Baehr writes, because the way he makes use of his courage is by definition villainous. We admire him for possessing a character trait that would, counterfactually, enhance his overall virtue if he also possessed other virtues. This fits well with the fact that virtue theory accords considerable room to moral education and character improvement. A virtue is an acquired character trait. One learns to be courageous by acting in ‘courageous looking’ ways in circumstances that require courage, and despite the fact that this may initially require a lot of willpower, the idea is that after some time practising one will have acquired the trait, which thenceforth needs maintenance only.

Consider this argument. Some people, let us suppose, have more or less finished the acquisition of courage and a number of other virtues, but are still struggling with another virtue: generosity, say. Living extravagantly, they are prone to give away too much and exemplify one of the extremes of generosity: prodigality. Must we wait before admiring their courage until they have given up the wasteful life and learnt to be generous? That is not just unnecessarily harsh but highly artificial. We should rather say that we admire them for their courage, even if they sometimes use courage in the pursuit of extravagance. If we do that, then we admire them precisely because of a character trait that, ‘if put in the service of the good, would enhance [their] overall orientation to the good’.²⁷

This is important for courage and other moral virtues, but it is likely to be even more important for epistemic virtues. Many atrocities are committed because of ignorance, stupidity, gullibility and other epistemic bads and vices. But just as many would not have been committed had the perpetrators been lesser epistemic agents. Cooking the books as creative accountants do, manipulating the bonuses as clever bankers do, money laundering, deceptive sales techniques, marketing products to vulnerable consumers, misleading the tax office: without epistemic virtues these forms of misconduct are less common. Does this mean that we should describe the perpetrators as epistemically unvirtuous? Not at all. As long as people lack certain virtues and possess certain vices, the virtues they do possess may be applied in ways that support the vices. Hardly anyone exudes all virtues, so being too rigid risks ascribing courage only to saints and angels. Everyone falling short of that ideal is a coward, then, because courage may be used in tandem with one or more vices. That makes for much too ethereal and esoteric a view of virtue to be applicable in real life.