1. Aquinas on habits and virtues

Aquinas defines temperance as a type of virtue (virtus), and in turn defines virtue as a type of habit (habitus). It will be helpful, then, to begin by considering Aquinas’ understanding of the notions virtue and habit. It is worth noting at the outset that contemporary Thomists disagree over how the historical Aquinas defined these terms.Footnote ² One underlying cause of this state of affairs is Aquinas’ tendency to incorporate apparently conflicting definitions (from authorities as diverse as Augustine, Aristotle, and Averroes) into his own accounts of habit and virtue. Aquinas’ appeal to an Aristotelian metaphysics of substance and accident is seen, for instance, in the Prima Secundae, where he defines a habit (habitus) as “a disposition whereby someone is disposed, well or ill.”Footnote ³ Aquinas locates habits and dispositions within the category of quality (one of Aristotle's nine categories of accident),Footnote ⁴ and singles out habits and dispositions from other types of quality precisely on account of their alone being capable of being “disposed, well or ill,” i.e. their capable of being good or bad. This is because, in contrast with the other qualities,Footnote ⁵ habits and dispositions alone are “according to nature” (secundum naturam)Footnote ⁶, and to be “according to nature” involves being ordered with respect to a thing's end (rationem finis), i.e. being ordered according to the end (good) or failing to be so ordered (bad). Elsewhere, though, Aquinas invokes a narrower definition of habit, namely, the Avveroist definition of habit as a quality “whereby we act when we will”Footnote ⁷, i.e. a quality pertaining specifically to the will.Footnote ⁸

Aquinas’ definitions of habit pave the way for his definition of virtue in question 55 of the Prima Secundae, where Aquinas approves Augustine's definition of virtue as one which “comprises perfectly the essential notion of virtue”. Although Augustine's definition (“Virtue is a good quality of the mind, by which we live righteously, of which no one can make bad use, which God works in us without us.”)Footnote ⁹ does not explicitly mention ‘habit’, Aquinas goes on to state that “the definition would be more suitable if for quality we substitute habit, which is the proximate genus”,Footnote ¹⁰ and further goes on to summarize Augustine's definition of virtue as an “operative habit… productive of good works.”Footnote ¹¹ (By adding ‘operative’, Aquinas brings Augustine's definition of virtue into closer alignment with the Aristotelian definition of virtue as a “habit for choosing”).Footnote ¹²

In short, Aquinas understands virtues to be good habits. However, not all virtues play an equal role in the moral life. Following Aristotle, Aquinas rejects the Platonic conception of morality as a purely intellectual affair, making a distinction between intellectual virtues (such as the science of geometry) and moral virtues (among which resides the virtue of temperance).Footnote ¹³ Intellectual virtues perfect the intellect, but they do not always dispose a person to make good choices; it is only moral virtues (which perfect either the will or the sense appetite) which do so.Footnote ¹⁴ It is among the moral virtues that Aquinas places temperance; let us now turn, then, to Aquinas’ treatment of temperance.

2. Aquinas on temperance

Aquinas gives the name “temperance” (temperantia) to the virtue by which we gain proper control, or ‘moderation’, over the “desires and pleasures of touch”—by which he means the passions associated with food, drink and sex.Footnote ¹⁵ (In this, Aquinas follows Aristotle's claim that eating, drinking and sex are all based on the same sense, the sense of touch).Footnote ¹⁶ The reason why Aquinas regards temperance as a cardinal virtue is that the desires for food, drink and sex are the most basic (‘natural’)Footnote ¹⁷ to our animal nature, and hence move the appetite with greater force,Footnote ¹⁸ and longer duration, than all the other passions.Footnote ¹⁹ Since these most forceful and basic of the desires is moderated by temperance, the virtue of temperance is “one of those things that are requisite for the notion of virtue in general,” insofar as it is the virtue that moderates pleasure, par excellence.

It should be noted that, as with the terms “habit” and “virtue”, Aquinas sometimes uses “temperance” in more than one way. In particular, Aquinas sometimes uses “temperance” in a broader sense, referring not just to the virtue dealing specifically with the strongest desires of the concupiscible appetite (i.e. for food, drink and sex), but rather, to the moderation of the concupiscible appetite in general.Footnote ²⁰ In this latter, broader sense, temperance is understood as one of four subcategories encompassing all the moral virtues. However, in the narrower sense, temperance is distinct from a number of other virtues residing in the concupiscible appetite, which do not take the passions for food, drink, and sex as their object. Thus, Aquinas distinguishes temperance from virtues that moderate senses other than the sense of touch,Footnote ²¹ as well as from virtues that moderate passions weaker than the passions of touch,Footnote ²² such as the desire for knowledge (moderated by studiositas).Footnote ²³ In what follows, we will focus on this narrower understanding of temperance, i.e. the virtue dealing solely with the pleasures associated with food, sex and drink.

According to Aquinas, temperance moderates the pleasures for food, sex and drink by aligning our desires for these things with a mean between two vicious extremes:Footnote ²⁴ On the one hand, an insufficient desire, for food, sex and drink (insensibilitas),Footnote ²⁵ and on the other hand (and far more commonly), an excessive desire for food (gluttony),Footnote ²⁶ for drink (drunkenness),Footnote ²⁷ or for sex (lust).Footnote ²⁸ This conception of temperance might appear puzzling insofar as it seems difficult to make sense of things such as fasting or virginity (both of which Aquinas regards as parts of the temperate life) as falling within a ‘mean’—after all, fasts typically involve eating “too little”, and virginity involves the permanent renunciation of sexual pleasure. To see why Aquinas regards both virginity and fasting as compatible with temperance understood as a mean state, it is necessary to elaborate on Aquinas’ conception of the ‘rational mean’ (medium rationis).

Following Aristotle, Aquinas regards the rational mean as a mean “through conformity with right reason”.Footnote ²⁹ As Aristotle put it, such a rational mean is “relative to ourselves” and “determined by a rational principle… that principle by which a man of practical wisdom would determine it.”Footnote ³⁰ Two observations can be made about this conception of the rational mean. First, it is a mean determined not by quantitative but rather by prudential considerations about what is morally appropriate in a given circumstance. It is the prudential character of the rational mean that explains why Aquinas regards fasting and virginity as characteristic of temperance. While they lie on an ‘extreme’ from a quantitative perspective (i.e. involving excessively little quantities of food, or sexual pleasure), they do not lie on a prudential extreme, because, as Aquinas argues elsewhere, fasting and a commitment to virginity are both praiseworthy and morally good actions, which can therefore lie within the dictates of prudence.Footnote ³¹

Second, the rational mean is “relative to ourselves”, i.e. what counts as the mean is not the same for every person, but is sensitive to contextual facts about each individual. For instance, chastity may call for an absence of sexual passion in certain circumstances (e.g. if one is married but is temporarily absent from one's spouse)Footnote ³² but not in others (Aquinas notes, for instance, that chastity in the married state may even require a greater sexual desire than one is inclined to, given that prudence dictates it is good to, at least sometimes, “pay the marriage debt”).Footnote ³³

3. Can temperance be misused for evil?

It will be helpful at this point to consider a common objection to temperance, my reply to which will help to highlight the nature of the rational mean in Aquinas’ account of temperance. The objection, in short, is that temperance seems capable of being misused for evil, and thus cannot be considered truly good. The objection is a serious one, since Aquinas holds to the Augustinian definition according to which virtue is a habit “of which no one can make bad use.”Footnote ³⁴ If this is right, and if temperance can be abused for evil, then, by Aquinas’ own lights, temperance would fail to truly be a virtue. However, the objection goes, temperance (understood as the moderation of pleasures) does appear to be capable of abuse. As Kant put the objection:

Kant's implicit argument is that a scoundrel's ‘temperance’ would, if anything, make him more dangerous than an intemperate one (his self-restraint might make him capable of pulling off crimes that intemperate scoundrels would lack the self-discipline to carry out).

Before I reply on behalf of Aquinas, it is worth noting one avenue of reply to Kant taken by some virtue theorists, namely, that of conceding that temperance is indeed capable of being abused for evil.Footnote ³⁶ This concession, it is argued, does not pose a serious problem for virtue ethics: We simply need to realize that ‘un-abusable virtue’ is not the correct moral ideal, and that virtues are still good and worth striving for despite their capacity for abuse.Footnote ³⁷ Neither Aquinas nor Kant would find this reply satisfying, since they both regard ethics as concerned with what is unqualifiedly good, and not merely with what is good in a merely qualified, or instrumental, sense. Indeed, Aquinas’ agreement with Kant on this score is illustrated in his example of the so-called “temperate miser” who plays a similar role to that of Kant's scoundrel in appearing to abuse virtuous qualities for base ends. Aquinas's verdict is that the miser is not truly virtuous:

What is noteworthy for present purposes is that, while Aquinas does not think the miser's ‘virtues’ (such as the capacity for curbing short-term desires for the sake of long-term monetary gain) are true virtues (i.e. moral virtues) at all, he does not take the miser to be a counterexample to temperance's status as an unqualifiedly good virtue. Rather, he simply uses it to illustrate that the habit of temperance is in a different order from habits such as self-discipline, self-denial, or self-restraint, precisely because the latter are merely instrumental habits, and can be used for both good and bad ends. As such, they can be possessed independently of prudence,Footnote ³⁹ because acting for bad ends involves a failure of prudential judgment (to judge what is good according to right reason in a given situation). Temperance, by contrast, is by its nature inseparable from prudence, since it involves the conformity of one's desires to the rational mean, a mean determined by prudence. Kant's critique is thus instructive: it reveals the importance of Aquinas’ understanding of temperance as determined by a rational mean in distinguishing temperance from merely instrumental, non-moral (and indeed, potentially evil) habits for moderating sensual desires in a merely routine, or quantitative, way.

4. Does temperance devalue effort?

Our discussion of temperance and the rational mean has shown that Aquinas regards temperance as crucially determined by right reason. One consequence of this we have just seen is that, for Aquinas, temperance is (pace Kant) incapable of being abused for evil, or used to make immoral decisions. A second consequence is that, for Aquinas, temperance is distinct not only from intemperance, but also from two other states: continence and incontinence. The difference between the three states can be stated (in brief) by considering the following three elements in temperance:

1. (i) passions and desires rightly ordered to reason

2. (ii) the right judgment of reason

3. (iii) right action

A continent person has (ii) and (iii) but not (i); he lacks rightly ordered desires. Thus, the continent person, like the temperate, still acts in accord with right reason, but he finds it difficult (e.g. to refrain from an alcoholic binge) whereas the temperate person, with rightly ordered desires, finds it easy. On the other hand, the intemperate and incontinent both lack (iii) and (i): Like the continent, they lack rightly ordered desires, but they furthermore allow those disordered desires to lead to disordered action. However, as Aquinas puts it, “the incontinent man repents at once, as soon as the passion has passed,”Footnote ⁴⁰ while the intemperate person “rejoices in having sinned, because the sinful act has become connatural to him by reason of his habit.” Aquinas describes this difference between the incontinent and the intemperate as a difference in ‘choice’: the incontinent acts “without choice”,Footnote ⁴¹ because his sin does not proceed so much from his will or intellect, but rather from his disordered passions (which overcome his will and intellect in the heat of the moment). The intemperate, on the other hand, acts “from choice”, since his intellect regards the disordered actions as good not just in the heat of the moment, but in a more permanent way. Thus, the incontinent might be said to retain (ii) to a greater degree than the intemperate, being able to judge rightly when not overcome by the ‘heat of passion’.

Now, few would challenge Aquinas’ claim that intemperance is worse than incontinence.Footnote ⁴² However, Aquinas’ claim that temperance is a better state than continence is more controversial. Indeed, a well-known objection (again found in Kant) is that, if we regard temperance as greater than continence, we thereby devalue the importance of effort in the moral life. In a well known example, Kant compares persons who protect themselves from threats to their life out of “an immediate inclination to do so,” with a suicidal man for whom “disappointments and hopeless misery have quite taken away the taste for life”, such that he “longs for death,” yet “still preserves his life without loving it—not from inclination or fear but from duty”.Footnote ⁴³ Both the ordinary persons and the suicidal man are striving to protect their lives, but Kant takes it that only the latter has acted from a maxim with “moral content”, since the ordinary persons act only from inclination and not from duty. The objection to Aquinas can be put as follows: If temperance is valued as a virtue, doesn't this lead us to falsely praise those ordinary persons who act merely from inclination, while devaluing the truly moral soul who acts out of duty and not inclination?

As Jean Porter has noted,Footnote ⁴⁴ a difficulty in answering this objection is that it highlights a deep divide between Kantian and Aristotelian approaches to morality. While the Kantian approach emphasizes duty and effort in the moral life, the Aristotelian approach emphasizes human flourishing in a more general sense. The two are not necessarily in conflict, and indeed, there is a degree of common ground. Hence, even though Aquinas regards temperance as better than continence, he nevertheless regards continence as having “something of the nature of a virtue”.Footnote ⁴⁵ Nevertheless, the Kantian objector would be right in identifying Aquinas’ understanding of temperance as one that does not value ‘moral struggle’ or ‘effort’ in and of itself. For the Aristotelian, there is good reason for this position, since such struggle reflects the existence of disordered appetites within the soul, and disordered appetites are incompatible with complete human flourishing. While this assessment of moral struggle might not convince a Kantian, it highlights what is distinctive about Aquinas’ ethics. Here we will note two such features. First, the importance of desires in the moral life. Whereas the Kantian focuses solely on the “good will”, and hence finds more to admire in the continent's exercise of willpower than in the prompt and easy actions of the temperate, Aquinas regards desires and passions as themselves important factors in moral evaluation. Second, Aquinas’ optimistic perspective on passions. Whereas the Kantian is suspicious of acts proceeding from inclination or emotion, Aquinas has an outlook on human passions and desires that is fundamentally positive. For although Aquinas regards passions as morally neutral in themselves, they are necessary components of moral perfection in good acts, and hence in the virtuous life.Footnote ⁴⁶ To conclude this section, I should re-iterate that, in replying to the objection that temperance devalues effort, my purpose has not been to provide a dialectically effective response to a Kantian objector. Rather, it has been to clarify what I take to be an important feature of Aquinas’ account of temperance. I have suggested that, by contrasting temperance with continence, Aquinas should be seen not so much as devaluing effort, but rather, as emphasising the importance of passions and desires (and their potential for good) in the moral life.

5. Does temperance devalue strong passions?

We have so far discussed the role of the rational mean in temperance and the fact that passions and desires (rightly ordered) are essential to the life of the temperate person. However, what has not yet been discussed is the precise nature of the relationship between the passions on one hand, and reason on the other, in the life of the temperate person. Is it possible for passions to exhibit ‘spontaneity’ and independence from reason (while remaining in harmony with reason) in the temperate person's life? Or rather, must they always be fully subjugated to reason, ‘at the beck and call’ of the judgment of reason? While some contemporary Thomists have portrayed Aquinas as allowing a degree of spontaneity in passion, Giuseppe Butera has recently shown that the latter, ‘subjugated’ view is Aquinas’ own.Footnote ⁴⁷ On Aquinas’ view, the temperate person has no strong passions for food, drink and sex that are involuntary or that arise independent of the judgment of reason. The reason for this is twofold. First, Aquinas holds that an act is virtuous only insofar as it follows from the judgment of reason, as opposed to a strong passion. This point is most clearly made in De Veritate q. 26, ad 7:

It would follow from Aquinas’ position that the stronger a passion is, the more it would appear to detract from the moral worth of the act when occurring prior to, or ‘antecedent’, to an action. Hence, Aquinas says that antecedent passions detract from one's culpability in committing sin, but also that when doing a good act, “such a passion diminishes the goodness and merit of the act.”Footnote ⁴⁹

Secondly, Aquinas holds that if the soul is strongly drawn to the operation of some power, the activity of other powers will be diminished: “when one power is intent in its act, another power becomes remiss, or is even altogether impeded, in its act.”Footnote ⁵⁰ The result is that if one's passion (e.g. for food) is excessively strong, one will be strongly hindered from judging according to reason. It should be noted that Aquinas holds that antecedent passions exist in everyone (even temperate persons) in light of the Fall, because one of the results of the Fall was that mild inordinate antecedent passions (fomes peccati) are always present in everyone (except Mary and Jesus). As a result, Aquinas holds that reason's control over the concupiscible powers in a post-Fall world is never complete, or “despotic”, but rather partial, or “political,”Footnote ⁵¹ i.e. the sense appetites always have power to move the will with some independence from reason. (In contrast, “in the state of innocence the inferior appetite was wholly subject to reason: so that in that state the passions of the soul existed only as consequent upon the judgment of reason”).Footnote ⁵² However, Aquinas nonetheless accepts that these antecedent passions do not prevent us from gaining temperance, because in a temperate person these passions will be so mild as to pose no threat to reason's control (in contrast to ‘vehement’ antecedent passions which impede reason, as mentioned above).Footnote ⁵³ In sum, the temperate person, while having mild antecedent passions like everyone else, will ideally tend towards the sort of ‘despotic’ control over his passions through reason, akin to that exercised by Adam prior to the Fall. The temperate lack strong antecedent passions, and their passions are (by and large) voluntary: they arise “either from being commanded by the will,” or from “not being checked by the will”Footnote ⁵⁴ (i.e. permitted to continue, but remaining within one's power to eliminate if required). In any case, they are not ‘spontaneous’ or independent of reason.

Aquinas’ portrayal of the temperate person as (ideally) exercising ‘despotic’ control over his passions may appear more Kantian than Aristotelian, insofar as it appears to devalue any kind of good action that does not immediately proceed from a ‘dispassionate’ rational judgment. In fact, we will now consider an objection that Aquinas’ account of temperance, while emphasizing rightly ordered passions, at the same time unacceptably eliminates an important category of passions from the moral life, namely, strong, yet good, antecedent passions.Footnote ⁵⁵ One paradigmatic case of such a passion is the passion involved in sexual intercourse, which contemporary philosophers and theologians have tended to treat in a much more positive light than has been in the past, given contemporary theological reflections on the good of marriage and spousal union. The objection can be put like this: The passions involved in the sexual act are no doubt strong and vehement ones. However, they surely seem to be capable of being regarded as good, indeed virtuous and temperate, if enjoyed in the proper context of mutual marital self-giving and a proper appreciation of one's spouse and the nuptial union.Footnote ⁵⁶ But Aquinas seems unable to follow this development in moral theology, in light of his firm view that strong passions diminish (indeed, perhaps eliminate) the judgment of reason in an act, and his view that the judgment of reason is key to an act's goodness. Since intercourse involves strong passions, it would seem that Aquinas can never regard marital intercourse as truly good or virtuous, unless it were engaged in without any strong sexual pleasure, or with sufficiently ‘moderated’ pleasure. But surely decreased pleasure should not determine the goodness of marital intercourse.

The above criticism can be extended from the case of sexual passion to other cases of apparently strong yet morally virtuous passion. Consider, for instance, a musician's highly emotionally charged performance; an artist's creative inspiration; spontaneous emotions of grief over the death of a loved one. In each case, a similar accusation can be made: Aquinas seems categorically incapable of acknowledging that such actions have anything of virtue, unless the emotions and passions are moderated so as to provide more room for rational judgment.

In evaluating whether this objection is a fair one to make against Aquinas, three points can be made. First, it should be clarified that Aquinas does not take any kind of negative view toward sexual pleasure as such. In a well-known passage, Aquinas says:

This passage shows that Aquinas does not regard sexual pleasure as (in itself) bad in any way. However, it does not yet respond to the objection, since the claim that pleasure is not contrary to virtue does not show whether Aquinas can hold that an act involving strong pleasure itself could be virtuous.

This brings us to the second point, which concerns the distinction between antecedent and consequent passions. A key assumption made by the objection is that in each example (e.g. marital intercourse), there are only strong antecedent passions, and not strong consequent passions. However, it is not difficult to see that the passions involved in marital intercourse can be strong consequent passions, i.e. passions that proceed from a prior act of will (e.g. the decision to engage freely in intercourse, or the decision to engage in certain intimate acts that one foresees will lead to intercourse). Since Aquinas holds that consequent passions “help more than hurt” acts, this is one way in which Aquinas can legitimately be understood as regarding strong passions as good. However, this point still does not fully satisfy, insofar as the sexual passions, on this view, are understood in a somewhat counterintuitive way, i.e. only as passions which are subsequent to a prior decision, as opposed to passions which in many cases do precede a decision to consummate the marital act. Insofar as we want to make room, at least for the potential for a positive moral evaluation of sexual passions in this latter scenario, the problem re-emerges. For nothing we have said so far helps us to see how, in the latter scenario, Aquinas might regard the sexual passions as virtuous.

This brings us to my third point, which is inspired by Robert Roberts’ recent defenceFootnote ⁵⁸ of the virtue of temperance. There, Roberts defends a conception of the rational mean similar to the one we have discussed earlier in this essay: In virtue of having a rational mean, temperance consists not merely in ‘restraining’ or ‘stoking’ our desires for food, drink and sex so that they become desires for an appropriate quantity of those objects (e.g. one serving of dinner and not two), or for an appropriate type of object (e.g. not just meat but also vegetables; or, not underage or homosexual union, but union with adult members of the opposite sex). Rather, temperance has a further component, namely, a cultivation of the right understanding, or perception, of the relevant objects of desire. According to Roberts, a temperate person who is told that this wine costs $80 a glass will lose his very desire for that wine, precisely because the information he has received changes his perception of the wine: It is not just that the person knows that he lacks $80 to spare from his monthly budget, and is able to restrain his desires appropriately; rather, he perceives the wine as ‘too-expensive wine’ and his desires are promptly transformed according to this new perception. If Roberts is correct, the relationship between reason and passion becomes intertwined, and inseparable from, the act of perception. Arguably, such a picture of perception accords with Aquinas’ view.Footnote ⁵⁹ In any case, it provides us with a way of deflecting the brunt of the present objection, because acts such as marital intercourse (or grief over the death of a loved one, etc.) undeniably are shaped by perception: Sexual passion is passion for a specific person, overwhelming grief is grief for the loss of a specific person, and so on.Footnote ⁶⁰ If temperance can be understood as the proper shaping of one's perceptions through reason, whereby a spouse's sexual passion crucially involves e.g. perceiving the object of their desire as their spouse in sacramental marriage, who is currently willing and able to have conjugal relations (for example), it follows that ‘strong passions’ such as those involved in the sexual act can indeed be subject to judgments of reason, understood as perception. (E.g. a spouse whose sexual passion involved the appropriate perception of their spouse, would ‘spontaneously’ lose their passion in circumstances where intercourse were rendered imprudent, e.g. if the spouse became unwilling or sick, etc.).

If this suggestion is accepted, it follows that Aquinas can recognize the goodness of strong passions like sexual desire without significantly modifying his conception of the relationship between passion and reason. After all, what is crucial to Roberts’ conception of temperance as proper perception, is that reason can still be understood as having despotic control over passion insofar as reason is presupposed in correct perception. What needs to be changed in Aquinas’ picture, is merely the claim that all instances of strong sexual passion are cases of strong antecedent passion. Instead, it can be recognized that many cases of strong sexual passion are not truly strong antecedent passions, but in fact consequent to a person's rational judgment insofar as their passion presupposes a proper, rational perception of their spouse as the object of their desire. On this view, ‘vehement’ passions remain incompatible with temperance, not on account of strength or vividness, but rather on account of being so strong as be, by their very nature, an obstacle to the proper perception of the object of desire. Insofar as sexual passions or spontaneous acts of grief do not obscure such perception of their objects (and in at least some cases of marital union, we conjecture, they do not), we are thus able to say that Aquinas could in theory accept these passions as proper parts of a virtuous act.