Professor Montmarquet defends a version of virtue theory which maintains that virtues are linked with particular true beliefs about the world. For example, an agent's virtue of generosity is linked with the belief that "life will provide adequately for the agent." At minimum, professor Montmarquet is suggesting that these true beliefs are to be seen as necessary conditions of virtues, and that the absence of such beliefs will be at the heart of corresponding vices. His essay also implies that a proper understanding of the catalog of virtues requires knowledge of the true beliefs or belief systems which are linked with the virtues. Professor Montmarquet's theory is classified as cognitivist since the virtues are grounded in the truth value of our beliefs, and these, in turn, have reference to facts about the world. His theory is bold in its optimist view of morality, and to my knowledge is thoroughly unique. In the following I will discuss possible counter examples to his theory which suggest that true beliefs are not necessary conditions for virtues. My arguments are not intended to refute professor Montmarquet's theory, but only to highlight areas which may be further clarified.

Consider the following scenario. Suppose that a certain religious cult mandates that true believers need to acquire morally proper dispositions, and those who don't will be drowned in the Putrid Swamp in the after life. Although the religion is a sham, Jones is unaware of that fact and has become a devout believer. In accord with this religion, he develops proper dispositions which a spectator would identify as charity, courage, temperance, justice, and patience. In each case Jones's dispositions are linked with his erroneous belief system that unbelievers (or the unvirtuous) will be drowned in the Putrid Swamp in the afterlife. If we asked Jones about a specific disposition of his, such as temperance, he might argue that the Lord of the Putrid Swamp (who drowns the unvirtuous) can only see those who are intemperate. The temperate will be invisible to him, and thus escape being drowned. On the surface, then, we have a situation where apparent virtues are linked with false beliefs. Although I have fabricated the particulars of Jones's religion in this example, it clearly follows the structure of a typical divine command virtue theory (complete with sanctions). Thus, with some variation, this is a counter-example which many of us may be familiar with.

Two other types of counter-instances also suggest themselves. First, suppose that an agent initially develops apparent virtues based on a specific belief system, but as time passes she can no longer articulate the relevant belief system of her youth; nevertheless she continues to act virtuously. Here we would have a virtue which is not linked with a relevant belief system. Second, suppose that an agent initially develops an apparent virtue based on belief system A, but later rejects belief system A for belief system B (where A and B are incompatible); however, she retains the same virtue, but now links this virtue with belief system B. Here we would have a virtue which at different times is linked with incompatible belief systems, both of which cannot be true.

There are two possible responses which the defender of the cognitivist theory might present to the above counter-instances. For simplicity I will focus only on the first religion-based counter-instance. First, the defender might argue that a kind of de-mythologizing is necessary to discover the core belief system which is most closely linked with an agent's virtues. Although Jones may claim that his temperance is linked with the belief that he will escape detection by the Lord of the Putrid Swamp, his de-mythologized belief might be that "moderatation will minimize conflict." The problem with the de-mythologizing approach is that it assumes a privileged position of interpretation which is not justified. For, it improperly presumes that although a person claims to believe X, she really believes Y. But, it is conceivable that if an agent was presented with the de-mythologized account of her belief, she would reject it as an inaccurate representation of her true belief. Further problems are created if more than one de-mythologized interpretation suggests itself. Finally, de-mythologized statements risk sounding like aphorisms from fortune cookies, such as the above, "moderation will minimize conflict." And these tend to be less like cognitive statements of fact, and more like noncognitive expressions of optimistic feelings.

A second possible response to the above counter-example is to simply deny that the dispositions which Jones develops are genuine virtues. One might argue that although to us Jones appears to have the virtues of charity, courage, temperance, justice, and patience, there are basic flaws with his dispositions that keep them from being true virtues. But what keeps these dispositions from qualifying as virtues? It would beg the question to argue that they do not qualify as virtues merely because they are linked with false beliefs. For at this stage it has yet to be established that true beliefs are indeed necessary conditions for virtues. Instead, it would have to be shown that something else about the disposition does not match the necessary qualifications of a virtue. For instance, we may note that the agent has the wrong kind of attitude, or that her disposition produces the wrong kind of behavior. These flaws, then, might be correlated with the presence of a relevant false belief. The difficulty, though, is in showing that for each agent we select who has a relevant false belief, that agent will also have a flawed disposition.

Another problem with this approach is that it seems too unfriendly to the holder of false belief systems. If Jones appears genuinely virtuous to us in spite of his relevant false beliefs, do we really want to say that he is unvirtuous? It seems that we should be entitled to our false belief systems without putting our moral status at risk for that reason alone.

There are, then, several obstacles for a cognitivist theory of virtue. In making a final assessment, though, it is important to see if rival theories of virtue fare any better. Professor Montmarquet discusses two alternative approaches. The most popular is a passion-based theory of virtue which sees virtues as dispositions which regulate how we respond to particular passions. Courage, for example, is a regulated response to fear. Briefly, the problem I find with this approach is that most virtues have no obvious passion with which they are associated. The obvious courage-fear association is an exception, rather than the rule. It thus seems forced to insist that all virtues follow the courage-fear model. The second alternative approach sees virtues only at the level of behavior (i.e. action and thoughts). This view might be called a behaviorist theory of virtue since what is relevant is that a virtuous person displays the characteristic and observable marks of a virtue. Professor Montmarquet criticizes this approach since it fails to explain why the agent chooses these dispositions. But, what is so important about why an agent chooses a disposition? I have suggested above that a range of disparate and mutually incompatible beliefs may give rise to the same dispositions which from the outside we would identify as virtues. If this is so, then what is central is the displayed behavior, and not why the agent choose a disposition (motivated by one belief vs. another belief). Accordingly, the behaviorist theory of virtue best reflects what is central to a virtue: the displayed behavior.