By Osvil Acosta-Morales

Winner of the Outstanding Graduate Paper Award at the 51st Annual Meeting of the Florida Philosophical Association

Osvil Acosta-Morales, University of Miami

Consider the claim that insofar as our interests are epistemic, what should guide our belief formation and revision is a strict adherence to the available evidence. The idea is that we should believe something if and only if we find that the evidence is sufficiently in its favor. Although some are opposed to this evidentialist approach, there have been strong supporters of it in the past, and plenty who endorse it in one way or another today. The position can be traced back at least as far as John Locke, when he argued that “assent ought to be regulated by the grounds of probability.”1David Hume2and W.K. Clifford3are also among the more easily recognized supporters of this view. Today it is not difficult to find defenders of a similar notion. A clear example is the defense of evidentialism by Richard Feldman and Earl Conee in their recent book.4 I will present here a case against the evidentialist approach within the ethics of belief, arguing that there are times when there is nothing epistemically wrong with believing something on admittedly “insufficient” evidence. Different attempts have been made, perhaps most famously by William James,5 to defend the idea that under certain circumstances believing something on insufficient evidence may be permissible. However, it is important to notice that my position differs in a significant respect from that of James and others like him. Instead of trying to justify or otherwise validate such beliefs on some sort of pragmatic or moral grounds, I argue that based on purely epistemic considerations it is a mistake to conclude that all beliefs based on insufficient evidence are in some way faulty or something that we ought rationally to avoid. Furthermore, I am inclined to make the stronger claim that some of these beliefs may be virtuous or praiseworthy, and an agent’s disposition to reach beliefs in such a manner is an epistemic virtue. I propose that we consider a form of courage to be an intellectual or epistemic virtue. It is through this notion of courage that we can see a weakness in the evidentialist position and find room for epistemically justified beliefs that are based on insufficient evidence. Of course, this all requires further explanation.

Lists of virtues usually include such things as temperance, justice, and courage. Most discussions of virtue focus on moral or ethical virtues. However, I am presently concerned with a different sort of virtue. The virtues that are relevant to the epistemic evaluation of an agent’s beliefs and belief revision are epistemic or intellectual virtues. Some examples of epistemic virtues might include open-mindedness or thoroughness of inquiry. There are a variety of ways that one can understand virtues, both epistemic and moral. There are many debates within the area of virtue theory in general and virtue epistemology6in particular that may be of interest, but we need not involve ourselves at this time with a detailed analysis of virtues in general or even epistemic virtue. It suffices for now to recognize that an essential part of what makes something a virtue is that it contributes to the attainment of valuable ends. I also believe that virtues involve a necessary internal or motivational element, but will not explore that here. Let us simply focus on the fact that moral virtues assist the agent in achieving valuable moral ends (e.g., fair treatment of others) while epistemic virtues assist the agent in achieving valuable epistemic ends (e.g., knowledge and truth).

Virtues are frequently understood to be favorable character traits or dispositions to behavior. I take this to be correct. However, I will work here with the notion that a particular choice or action may also be considered virtuous, or at least exhibiting virtue, even though the agent falls short of being virtuous because she lacks the appropriate character trait. For our present purposes at least, a choice or act that exemplifies a virtue may be understood as virtuous, independent of whether the agent making the choice or performing the act possesses a virtuous character. This helps the presentation of my primary argument by allowing us to focus on and evaluate particular beliefs and choices instead of keeping our attention and judgments confined to agents.

Let us now look at the virtue that reveals how beliefs may be epistemically justified even when the evidence is admittedly lacking. I propose that we consider a form of courage to be an intellectual or epistemic virtue. The sort of courage that is directly relevant here is a doxastic courage, pertaining to beliefs. This differs from the sort of practical courage, pertaining to actions, that is usually discussed in the literature. However, the elements that make doxastic courage an epistemic virtue are the same sorts of things that normally make its practical counterpart a moral virtue. We need not concern ourselves with defending a special conception of courage to see this. The ordinary notion of a courageous act basically involves the agent exposing herself to significant harm in order to achieve some goal that is of significant value. The courageous person does not allow the risk of loss to keep her from taking action or making the appropriate choice. However, Aristotle7was right in pointing out that courage involves an appropriate sort of wisdom. Being courageous is more than just being willing to take risks. The courageous person refrains from being rash and making a decision without properly considering the level of risk involved and the value of the end she is seeking. The courageous believer I have in mind will also avoid the extremes. He will not let the fear of believing something that is false lead him to be too timid or overly cautious with what he believes. On the other hand, he will not disregard the available evidence entirely and believe anything at all.

There are different ways that a choice, belief, or action may be courageous. It seems that when most people think of courageous beliefs the sort of scenario that comes to mind is that of someone believing something that is contrary to popular opinion, exposing themselves to ridicule or perhaps even physical harm for believing as they do. I agree that in some sense a person’s belief is courageous when she believes something even though she knows it to be unpopular. However, this is not the sort of courage I am interested in exploring here. All acts of courage require a goal or end that the courage is meant to help the agent in achieving. Along with this there is always a “counter- goal” or danger the agent must face.8 In the situation described above the goal or aim of the belief seems to be truth or true belief, and the counter-goal or consequence of failure is something like ridicule or physical harm to the individual. I am interested in cases where both the goal and counter-goal are epistemic in nature, where the potential gain and the potential loss from the believer’s choice both involve epistemic value. This goes back to my interest in focusing on epistemic considerations, not pragmatic or moral ones. When an agent exhibits doxastic courage he recognizes that the evidence does not adequately support a claim, but he believes it, realizing that he is taking a significant risk. The risk to which he is exposing himself, at least immediately, is that of believing something false. A false belief has a negative epistemic value. If the agent is interested in promoting or maximizing epistemic value it may seem odd that doing something that he recognizes risks a significant loss of value could be justified. The answer is simply that the risk is a justifiable risk due to the probability of success and the positive epistemic value gained from a true belief if he is successful.

It is helpful here to consider a detailed example. But first it is important to remember that doxastic courage is not of a usual sort, and our normal manner of speaking, along with our intuitions, may not immediately recognize it as courage. There are different ways to be courageous in what one chooses to believe, and certainly a variety of things that might be gained or lost by choosing to believe or disbelieve something. However, the desired focus here is on the evaluation of beliefs when we concern ourselves with only epistemic goals and counter-goals.

Unfortunately, when we think of courageous choices we do not normally consider conditions in which what is at stake is purely epistemic in this way. It might be said that this new or different sense of courage I am suggesting is not actually courage at all if it does not adequately match-up with what we normally think of as courage, and so it should go by some other name. I am inclined to think of doxastic courage as genuine courage because of the strong structural similarities it has to practical courage, and doing so also helps us to more easily notice what allows doxastic courage to qualify as an epistemic virtue. Still, calling it by a different name should not ultimately detract from the substance of my view. Or, if it is preferable, we may understand this as a stipulative definition of courage. What is more important here is that we notice that the same things that lead us to consider an action courageous and virtuous are sometimes present in the realm of belief and this is good reason to consider such beliefs virtuous, and, I think , courageous.

When making a case for justified belief in the face of insufficient evidence, examples frequently involve belief in God or some other such thing. I think under the right conditions such a belief may demonstrate doxastic courage. However, such an example may be unnecessarily complicated and challenging in a number of ways. A simple example may work best here to shed light on the notion of doxastic courage. But before we proceed even with this, it is best briefly to consider a scenario that involves practical courage. Imagine a couple, Jack and Jill, who are involved romantically. At this time Jack is considering whether he should propose marriage to Jill. He realizes that they have not been seeing each other for very long, but he also has strong feelings for Jill, and he thinks she feels just as strongly about him. Still, he realizes that proposing at this point is risky. He thinks, and reasonably so, that while there is a chance that this act will bring them closer and strengthen the relationship there is also the chance that Jill might feel pressured and it will ultimately distance them. He sees that there is a significant chance of each outcome, but neither one is clearly more likely than the other. It seems right to think that in this scenario Jack acts courageously if he proposes. There is a real possibility that things will not go as he would like, but he does not allow the fear of rejection to keep him from missing this opportunity to take the relationship to the next level.

Let us now examine a very similar case where the choice involves belief instead of action, and where the loss and gain considered are the loss from acquiring a false belief and the gain from acquiring a true belief. Barney has fallen deeply in love with a woman, Betty, who he has known for only a few days. Although he has known her for only a short time, Barney felt a special connection with Betty right from the beginning. Neither one has explicitly professed love for the other, but there is some reason to think that Betty does love Barney as much as he loves her. Still, sometimes Betty acts in a way that suggests that she actually does not love Barney. There might be an explanation for such things, but Barney must admit that it seems just as likely that she does not love him as it does that she loves him. A strict adherence to the evidence would undoubtedly suggest that Barney refrain from making a judgment on the matter since the evidence favors neither belief nor disbelief here. Still, it seems that Barney could exhibit genuine courage by choosing to believe that she does love him. By doing so Barney exposes himself to a potential false belief, and other loss as well, but at the same time he allows for the possibility of a true belief, and other gain as well. I propose that if in the earlier example of Jack and Jill we conclude that Jack is exhibiting courage, then in the parallel case of Barney and Betty, courage and virtue are also being exhibited.

In order to see how such a risk can be a justified one, it is helpful to examine a position that explicitly denies that any belief based on insufficient evidence can ever be epistemically justified. So, we turn here to evidentialism. I will focus on the position as it is presented by Conee and Feldman, since they are currently some of the most active defenders of the view, and they provide a clear and adequately representative analysis of evidentialism. The basic claim of this view is that epistemic justification “is determined entirely by the person’s evidence.”9 Directly linked to this is the further claim that in so far as we are interested in promoting epistemic value or having epistemically rational beliefs we ought to believe something if and only if it is epistemically justified.10 It a mistake to disregard the evidence entirely when deciding what to believe. It is also wrong to believe only those things that we take to be indubitable. At least when our interests are epistemic, we do best to avoid these strategies. Avoiding error is important, but achieving a comprehensive system of beliefs is a valuable epistemic goal that has a place alongside our interest in possessing an accurate set of beliefs. The evidentialist recognizes that the best strategy for maximizing epistemic value lies somewhere between these extremes, but he misses the mark because he does not properly account for the virtue of courage.

Justified belief will always involve a level of risk if we allow for epistemic justification with anything less than certainty. The evidentialist rightly sees that some risk is acceptable. Feldman agrees that even though “believing on only modest amounts of evidence involves taking some epistemic risk, … you should believe when your evidence is supportive rather than neutral, even if the evidence is not at all decisive.”11 The problem is that the evidentialist position is too strict. It allows for too little epistemic risk. In at least some cases we are epistemically justified in taking more risk than evidentialism allows. We see this in the cases where the evidentialist suggests that we suspend judgment.

In a situation where there is some evidence in support of a proposition and some evidence in support of its negation so that neither seems to have more support than the other, evidentialism tells us that we are epistemically unjustified if we do not suspend judgment on the matter. The claim is that, “neither belief nor disbelief is epistemically justified when our evidence is equally balanced.”12 However, if we are interested in maximizing epistemic value or utility, taking a risk where there is a 50% chance of gain and an equal chance of equal loss does not appear to be any more unjustified or irrational than not taking any risk at all, leaving no chance for loss and no chance for gain. A straight-forward matrix where we evaluate estimated value tells us that the courageous strategy is as effective as the conservative evidentialist one at maximizing epistemic value. We can see this in Table 1 below.

Table 1


1 -1
-1 1
0 0

Believe that p Believe that ~ p Suspend Judgment

For simplicity, we can take the epistemic value of a true belief to be 1, and a false belief as –1. When one suspends judgment on a matter, no epistemic value is gained or lost. There may be reason to think that the epistemic value of true belief and false belief is not symmetrical in this way for every proposition or hypothesis considered.13 Perhaps believing some proposition will produce only half as much epistemic value as believing its contrary. If so, this makes things more complicated, but a case for doxastic courage can still be made in a very similar fashion. Let us take the epistemic values to be those described in Table 1, at least for now, and see how this shows the error in evidentialism. It is not difficult to notice that if there is an equal chance that the proposition p is true as there is that it is false, the expected epistemic value for each row of the matrix will be the same. Suspending judgment is no better at maximizing epistemic value than believing that p or believing that not-p. Yet, the evidentialist position tells us that the only epistemically justified option is to suspend judgment. According to this view we are therefore epistemically obligated to suspend judgment since they take being epistemically justified as equivalent to being epistemically obligated.14 The evidentialist better- safe-than-sorry approach claims to be the best way to promote epistemic value, but it apparently has no advantage over the more epistemically courageous approach I am suggesting.

One way of impeding the argument for doxastic courage is simply to deny that true beliefs have any epistemic value. Feldman presents an argument along these lines. He says that although it seems right that something like true belief is a proper epistemic goal and something of epistemic value, this cannot be correct. Feldman argues that if merely true belief did have epistemic value then you might be gaining epistemic value just through luck and you might lack epistemic value even if you are being rational.15 Neither one of these consequences sits well with Feldman. He considers the possibility that knowledge might possess epistemic value, but rejects this as well. He says this seems plausible, but there are problems with this position.16 The problems with this view emerge in situations where you have strong evidence for a false proposition. The problem here is that if adopting an attitude that will yield knowledge is what is valuable, then believing the negation of the proposition will do equally well because it helps achieve the truth condition of knowledge. So, it seems this strategy suggests equally that we believe a proposition that is not supported by the evidence. Another difficulty Feldman mentions involves situations where there is strong evidential support for a proposition, but it is insufficient for knowledge. Evidentialism suggests we should believe the proposition, but if knowledge is the proper bearer of epistemic value, then evidentialism finds itself disconnected from the promotion of epistemic value. Feldman concludes that only rational belief has epistemic value.17

The fundamental weakness with Feldman’s arguments is that he neglects the possibility that true belief, knowledge, and rational belief all possess epistemic value. His arguments against true belief and knowledge as bearers of epistemic value work only when we assume that each is to be the sole bearer of epistemic value. Once we realize that each can have epistemic value there does not seem to be any difficulty. I agree with Feldman that true belief and knowledge alone cannot be the bearers of epistemic value. There certainly seems to be epistemic value in rational or reasonable belief that is neither true nor qualifies as knowledge. However, this only means that rational belief has epistemic value, not that true belief and knowledge lack it, as Feldman suggests.

Adopting a doxastically courageous approach serves best to maximize epistemic value. It acknowledges the role that evidence has in epistemic justification and its connection with promoting epistemic value, but it avoids the narrowness of an evidentialist position. The courageous believer I have described is not rash and one who disregards the evidence, but she also realizes that taking epistemic risks will not necessarily keep her from her epistemic goals. Doxastic courage is an epistemic virtue that serves the agent in achieving her epistemic ends. Believing something in the face of insufficient evidence is not always an epistemic vice to be avoided, and the agent who recognizes this is better off for it.


Works Cited

Aristotle. Nicomachean Ethics. Tr. David Ross. Oxford: Oxford UP, 1998.

Axtel, Guy. Knowledge, Belief, and Character: Readings in Virtue Epistemology. Lanham: Rowman and Littlefield Publishers, 2000.

Clifford, W.K. “The Ethics ofBelief” in The Ethics ofBeliefand Other Essays. Amherst: Prometheus Books, 1999.

Conee, Earl and Richard Feldman. Evidentialism: Essays in Epistemology. New York: Oxford UP, 2004.

Fairweather, Abrol and Linda Zagzebski. Virtue Epistemology: Essays on Epistemic Virtue and Responsibility. Oxford: Oxford UP, 2001.

Hume, David. An Enquiry Concerning Human Understanding. Great Books of the Western World, Vol. 35. Ed. Robert Maynard Hutchins. Chicago: Encyclopedia Britannica, 1952.

James, Willaim. “The Will to Believe” in Essays on Faith and Morals. New York: Longmans Green & Co., 1947.

Levi, Isaac. Gambling With Truth. Cambridge: MIT P., 1980.

Locke, John. An Essay Concerning Human Understanding. Great Books of the Western World, Vol. 35. Ed. Robert Maynard Hutchins. Chicago: Encyclopedia Britannica, 1952.

Pears, D.F. “Aristotle’s Analysis of Courage.” Midwest Studies in Philosophy 3 (1978): 273-85.

Rescher, Nicholas. “Peirce and the Economy of Research.” Philosophy of Science 43 (1976): 71-98.


  1. John Locke, An Essay Concerning Human Understanding, Great Books of the Western World, Vol. 35, Robert Maynard Hutchins, ed. (Chicago: Encyclopedia Britannica, 1952), Book IV, Ch. 16, Sect. 1.
  2. David Hume, An Enquiry Concerning Human Understanding, Great Books of the Western World, Vol. 35, Robert Maynard Hutchins, ed. (Chicago: Encyclopedia Britannica, 1952), Section X.
  3. W.K. Clifford, “The Ethics of Belief,” in The Ethics of Belief and Other Essays (Amherst: Prometheus Books, 1999).
  4. Earl Conee and Richard Feldman, Evidentialism: Essays in Epistemology (New York: Oxford UP, 2004).
  5. Willaim James, “The Will to Believe,” in Essays on Faith and Morals (New York: Longmans Green & Co., 1947).
  6. We can see some of the epistemological issues addressed from differing perspectives within Guy Axtel, Knowledge, Belief, and Character: Readings in Virtue Epistemology (Lanham: Rowman and Littlefield Publishers, 2000), and Abrol Fairweather and Linda Zagzebski, Virtue Epistemology: Essays on Epistemic Virtue and Responsibility (Oxford: Oxford UP, 2001).
  7. Aristotle, Nicomachean Ethics, David Ross, trans. (Oxford: Oxford UP, 1998), esp. Books II and VI.
  8. This is explored at length within D.F. Pears, “Aristotle’s Analysis of Courage,” Midwest Studies in Philosophy 3 (1978): 273-85.
  9. Conee and Feldman, 1.
  10. Conee and Feldman, 177, 182.
  11. Conee and Feldman, 180.
  12. Conee and Feldman, 83.
  13. See Nicholas Rescher, “Peirce and the Economy of Research,” Philosophy of Science 43 (1976): 71-98 and Isaac Levi, Gambling With Truth (Cambridge: MIT P., 1980).
  14. Conee and Feldman, 88.
  15. Conee and Feldman, 185-6.
  16.  Conee and Feldman, 183-4.
  17. Conee and Feldman, 184-5.

Osvil Acosta-Morales

Osvil Acosta-Morales specializes in epistemology and ethical theory. His research interests also include philosophy of religion, aesthetics, and philosophy of education. His current research is focused on a philosophical analysis of courageous belief, and he is presently completing work on his dissertation at the University of Miami under the direction of Professor Michael Slote. He is the recipient of the 2005 Florida Philosophical Association Outstanding Graduate Essay Award for his present essay, “Courage, Evidence, and Epistemic Virtue.” Recently, he has accepted a Visiting Assistant Professor position at the University of Toledo. [email protected]