Too Much Self-Control?

Abstract

Although it seems commonsensical to say that one cannot merely have too little, but also too much self-control, the philosophical debate has largely focused on failures of self-control rather than its potential excesses. There are a few notable exceptions. But, by and large, the issue of having too much self-control has not received a lot of attention. This paper takes another careful look at the commonsensical position that it is possible to have too much self-control. One key insight that will emerge is that there are certain important confusions surrounding this view. Once these are removed, however, we are led to the conclusion that there need not be anything intrinsically problematic about being a paragon of self-control.

1 Introduction

Commonsense thinking about self-control seems to display a fundamental ambivalence: on the one hand, there is the belief that self-control is “a good thing” and that failures of self-control are problematic. On the other hand, it also seems to be a widely shared belief that one can have “too much” self-control.

Until now, the philosophical debate has largely, and somewhat one-sidedly, focused on failures of self-control rather than on its potential excesses. There are a few notable exceptions to this (Kennett, 2001; Brownstein, 2018; Herdova et. al. 2023). But, by and large, the issue of having too much self-control has not received a lot of attention.

In this paper, I shall take another, careful look at the commonsensical position that one cannot only have too little, but also too much self-control. I specifically examine the view that having too much self-control is problematic because it manifests itself in various problematic forms of “overcontrol.” I will show that this “overcontrol worry” comes in several different versions and has considerable initial appeal. However, close inspection will reveal that it is largely unfounded. As I shall argue, standard everyday examples of overcontrol, rather than being manifestations of an excessive employment of the agent’s capacity for self-control, can best be accounted for without appealing to an agent’s self-control in any way. In an interesting turn-about, high levels of self-control can even be a mitigating factor.

An additional worry is that having too much self-control is problematic because it prevents an agent from making certain valuable “evaluative discoveries.” I shall argue, however, that self-control can either be an impeding or a facilitating factor when it comes to discovering the value of certain activities (depending on which further features and dispositions the relevant agent possesses). At the end of the paper, we are thus led to the conclusion that there need not be anything intrinsically problematic about being a paragon of self-control.

My paper is organized as follows. In Sect. 2, I offer a “minimalistic” conception of self-control and distinguish between two different ways of spelling out the view that an agent can have too much self-control. I then introduce the overcontrol worry, which will take center stage in the paper. In Sect. 3, I show that this worry comes in, at least, five different versions. In Sect. 4, I argue that, despite its initial appeal, the overcontrol worry proves largely unfounded. In Sect. 5, I present, and partly reject, the additional worry that having too much self-control is an obstacle to valuable evaluative discoveries. Section 6 concludes.

2 Too Much Self-Control: Preliminaries

Many different conceptions of the capacity for self-control have been offered in the literature (see, e.g., Mele, 1987, 1995, Ainslie, 2001, Kennett, 2001, Kennett & Wolfendale, 2019, Henden, 2008, Holton, 2009, Sripada, 2014, 2020, Herdova, 2017, Levy 2017, J.L. Bermúdez, 2018, Kalis 2018, Altehenger 2020, Debus 2020, J.P. Bermúdez, 2021, and Koi 2023). However, most theorists would be able to agree on the view that self-control is a capacity (or set of capacities) that is employed to bring one’s behavior into line with one’s goals in the face of motivational conflict.¹

This “minimalistic” conception of the capacity for self-control still leaves open many issues. Among other things, it leaves open which processes can realize an exercise of one’s capacity for self-control and whether there is a specific type of cognitive process that underlies all such exercises (see, e.g., Sripada, 2020). But the conception does come with an important theoretical commitment: it implies that whether a process counts as a successful exercise of one’s capacity for self-control depends on whether the process leads to behavior that is in line with one’s own goals (rather than with some external standard).

This assumption, in turn, points to a straightforward way of spelling out the view that an agent can have too much self-control, namely, that she might misuse her capacity for self-control in the service of goals that are harmful to others or to herself.

For illustration, take the case of a torturer who effortfully suppresses the pity she feels for her victim and thereby enables herself to proceed with the torture.² This agent seems to exercise self-control and, more specifically, seems to do so in the service of a goal that is evidently harmful to others. Now, in view of this, one  might claim that the torturer has too much self-control. Or, maybe somewhat more naturally, one might claim that it would be better (for others) if she had less self-control. The same holds in cases where the agent exercises her capacity for self-control in a way that is harmful to herself. The phenomenon of “self-harm by self-control” has recently been stressed by Michael Brownstein (2018). Brownstein offers the example of a teen who effortfully overcomes her initial aversion to smoking so that she can fit in with the “cool kids” (2018, p. 590).³ At least on the face of it, the teen seems to exercise self-control and, more specifically, seems to do so in the service of a goal that is harmful to herself (to take up smoking). Again, one could say that the teen has too much self-control, or, maybe somewhat more naturally, that it would be better (for her) if she had less self-control.

The view that agents can employ their capacity for self-control to promote goals that are harmful to either others or to themselves seems highly plausible. In this respect, self-control just seems to be like numerous other capacities (e.g., intelligence, charm, empathy, creativity, etc.) that are generally thought to be “good things”, but that can also be misused. I am thus happy to grant that agents can have too much self-control in this specific and rather uncontroversial sense.⁴ But, this is not the sense that my paper will be about.

So let me now turn to the reading of the claim that one can have too much self-control that I will focus on in this paper. At its core is the following thought: there is something problematic about being someone who is very self-controlled, i.e., who frequently and successfully exercises their capacity for self-control,⁵ even if we disregard the fact that this capacity may be employed to promote harmful goals. A less precise (but more catchy) way of stating this idea is to say that being very self-controlled is intrinsically problematic.

What is supposed to be the problem here? One key worry has to do with the threat of “overcontrol.” Somewhat more precisely, the overcontrol worry (as I shall call it) can be stated as follows:

An agent who is very self-controlled, i.e., who frequently and successfully exercises her capacity for self-control, will manifest various problematic forms of overcontrol (or at least be at a high risk of doing so). This will have serious negative consequences for different domains of her life. In particular, such an agent will suffer a significant loss of freedom or spontaneity.

The overcontrol worry seems to be one key driving force behind the commonsense intuition that one can have too much self-control. As I shall show in the next section, the overcontrol worry comes in at least five different versions some of which also show up in the philosophical literature.⁶

3 Five Varieties of the Overcontrol Worry

3.1 Over-habitualization

The first version of the overcontrol worry starts from the assumption that someone who is very self-controlled will, at least typically, be someone with good habits (or routines). However, the thought continues, this is something that might easily be taken to extremes and lead to an “over-habitualization” of one’s life (“Saturday night is laundry night”; “Thursday Night is Thai food night”; “Each day at 5:30 p.m., I run 5,5 miles.”). On the view at issue, agents who are very self-controlled are thus at a significant risk of becoming a “prisoner” of their own habits and routines and of thereby leading a life that is utterly devoid of spontaneity. Arguably, such a life would be impoverished.

3.2 Over-planning

The second version of the overcontrol worry takes as its starting point the assumption that an agent who is very self-controlled will regularly engage in careful advance planning. However, as in the case of habits and routines, this, too, seems to be something that can easily be overdone, as Brownstein (2018) illustrates in the following passage:

As a child I made budgets to determine how to use my allowance. In junior high I wrote lists of my daily tasks that included ‘wake up’ and ‘eat breakfast.’ By college I began each semester by mapping out all of my assignments for the term. I treated deadlines like contracts punishable on pain of death. These days I try to be a little more spontaneous, but often this desire amounts to a self-defeating effort to ‘pencil in’ some time for spontaneity three Thursdays from now at 1:30 PM. I’m probably a paragon of self-control, in the sense psychologists mean it (…). In fact, I may be overcontrolled. (Brownstein, 2018, pp. 585–586)

According to the view at issue, someone who is very self-controlled will, thus, (likely) engage in over-planning. This will once more lead to a life that is utterly devoid of spontaneity and which therefore appears impoverished.⁷

3.3 Over-deliberation

Relatedly, the third version of the overcontrol worry claims that an agent who is very self-controlled will (likely) engage in over-deliberation. For illustration of this worry, here is how Jeanette Kennett (2001) characterizes an agent who has too much synchronic self-control, i.e., roughly speaking, someone who resists most (or even all) of the occurrent temptations she encounters:

[S]ynchronic self-control too can be taken to extremes, both in general and on particular occasions, to the detriment of the agent concerned. The earnest person who exhaustively searches through her reasons on each and every occasion (no matter how trivial) that her judgements are vulnerable to a motivational deficiency also fails to realize important values in her life, for example spontaneity, and might readily be accused of excessive seriousness—of lacking a sense of proportion about the importance of her own activities. There is a desirable mean of self-control which such a person fails to realize. (Kennett, 2001, p. 150, my emphasis)

To illustrate further, take Brownstein’s case of Alfred (2018, p. 586). In this case, Alfred’s wife makes the suggestion to spontaneously have ice cream for dinner. In response, Alfred keeps producing reasons against doing so (the groceries in the fridge will spoil, having ice cream for dinner is bad for one’s cholesterol level, the kids might be angry if their parents eat all the ice cream in the fridge, etc.). According to Brownstein, agents who are very self-controlled will routinely ruin occasions for spontaneity in this way.

3.4 Emotional overcontrol

A further popular worry is that an agent who is very self-controlled will be “emotionally overcontrolled.” Such an agent, the thought goes, will constantly suppress her emotions rather than express them. In view of this, she will suffer a substantial loss of “emotional spontaneity” and consequently behave in a somewhat bland and “robotic” manner. A striking expression of this idea can be found in the comparison of agents who are very self-controlled with the android character Data from Star Trek (see Zabelina et al., 2007, p. 471). However, a lack of emotional spontaneity might not only constitute a serious loss in itself but also compromise an agent’s ability to have close interpersonal relationships.

3.5 Over-frugality

Finally, there is the equally popular worry that someone who is very self-controlled will display what might be called over-frugality. More precisely, this view claims that when facing a choice between some future prudential good (e.g., better health, a completed work project, money saved for retirement) and some present enjoyment (e.g., eating a delicious cake, going out with friends, spending money on a sumptuous vacation), someone who is very self-controlled will always opt for the former. Consequently, the thought continues, such an agent will miss out on “all the fun” and lead a life that can only be described as joyless and grim.⁸On the view at issue, agents who are very self-controlled would thus be like the ant in the tale “The Ant and the Grasshopper”—except that, for them, winter never comes.⁹

4 Overcontrol and Self-Control: A Closer Look

4.1 Defusing the Overcontrol Worry

The overcontrol worry, as it was stated in the previous sections (Sect. 2 and 3), possesses considerable intuitive appeal. However, close inspection reveals that it rests on various confusions and is therefore largely unfounded.

To recapitulate, the overcontrol worry claims that there is something intrinsically problematic about being very self-controlled (or, more precisely, about being someone who frequently and successfully exercises their capacity for self-control), since an agent who is very self-controlled will display the overcontrolled behavioral patterns described before (or at least be at a high risk of displaying them).

To argue for the claim that the overcontrol worry is largely unfounded, let me begin with the following observation: all five versions of the overcontrol worry presented above involve agents who display certain rigid patterns of behavior. An agent who has “over-habitualized” her life is someone who invariably follows through with a great number of habits (or routines). Relatedly, Brownstein’s “over-planner” is someone who plans in advance almost each and every aspect of her life. The “over-deliberator”, as described by both Brownstein and Kennett, constantly engages in extensive searches through her reasons. The emotionally overcontrolled agent continuously withholds her emotions. And the overly frugal agent unfailingly choses future prudential goods over present enjoyment.

This raises the question of how we can best account for such rigidity. To answer this question, let me first offer another observation. Plausibly, for an agent who displays some or all of the overcontrolled behavioral patterns described above, the overcontrolled response is actually the easy option. The difficult option—the option whose realization would require effort on her part—, would be to depart from her habits (or routines), to refrain from engaging in extensive planning or deliberation, to express her emotions, or to forgo choosing the prudent option over the hedonic one.

This, however, suggests that for an agent who displays some or all of the overcontrolled behavioral patterns described above, the overcontrolled response is actually the option that is “motivationally downhill,” i.e., it is the option that the agent is most motivated to engage in anyway. Or, in other words, following through with her habits (or routines), engaging in extensive planning or deliberation, holding back her emotions or choosing the prudential option over the hedonic one is the response that such an agent will display even if she does not intervene in her own motivational condition.

Notice, though, that if the explanation just offered is on the right track, then the overcontrolled responses are not manifestations of a successful exercise of the capacity for self-control.¹⁰ This is because we employ our capacity for self-control precisely because we seek to bring about a response that is different from the response which we would display if we did not intervene in our own motivational condition.¹¹ Indeed, the claim that self-control is a form of “motivational self-intervention”¹² can be regarded as a platitude.¹³

However, this raises the question of who the “real culprit” is. If I am correct in claiming that the overcontrolled responses at issue are not manifestations of a successful exercise of the agent’s capacity for self-control, then how else can we account for the fact that an agent invariably follows through with a great number of habits (or routines), constantly engages in extensive planning or deliberation, continuously withholds her emotions, or unfailingly choses future prudential goods over present enjoyment?

On reflection, there seem to be two major alternative candidates: obsessive-compulsive tendencies on the one hand and various sorts of negative affective states on the other.

To illustrate this point, let us once more consider Brownstein’s case of Alfred (see Sect. 3.3). In Brownstein’s construal of the case, Alfred’s “overcontrolled behavior”—the fact that he responds with over-deliberation to his wife’s suggestion to spontaneously have ice cream for dinner—is a manifestation of his having “too much” self-control. In contrast, I would argue that Alfred’s behavior is rather to be explained by the fact that he is somewhat obsessive–compulsive (and his wife’s suggestion to drastically depart from their usual way of having dinner thus makes him uncomfortable) or by the fact that he is plagued by health-related fears (and the thought of consuming a large amount of fat and sugar thus genuinely frightens him) or by a combination of both factors.¹⁴ Moreover, note that very similar explanations could be offered to account for cases of over-habitualization and over-planning.

Turning next to emotional overcontrol, negative affective states once more seem to play a key role. In particular, it seems plausible to assume that agents who constantly withhold, rather than express, their emotions will often be plagued by social anxiety, i.e., fear of the negative reactions others will display towards them if they openly express their emotions. Another possibility is that the constant withholding of emotions is a quasi-automatic response that, for instance, may simply be the result of growing up in an environment where emotions tended to remain unexpressed.

The phenomenon of constantly choosing future prudential goods over present enjoyment might seem the most puzzling. However, close inspection reveals that negative affective states are, once again, a key factor. Turning to a certain body of empirical evidence will prove helpful here. As George Loewenstein (2018, pp. 97–98) has convincingly argued, an over-emphasis on future prudential goods often results from various future-biased affective tendencies, which he also describes as “hyperopic” tendencies.¹⁵ One example of such a tendency is the “pain of spending” (Prelec & Loewenstein, 1998), i.e., negative feelings related to the spending of money which may lead to “underspending” rather than “overspending.” Other examples include various negative emotions related to self-indulgence or, more generally, present enjoyment (see Kivetz & Simonson, 2002; Haws & Poynor, 2008), such as a feeling of guilt that is triggered by the thought of engaging in some self-indulgence (e.g., a gourmet dinner), or a feeling of fear that engaging in some present enjoyment (e.g., going to a party tonight) will prevent the procurement of some future prudential good (e.g., a completed work project by the end of the month). It is easy to see how having strong tendencies of this sort can lead to a rigid over-emphasis on future prudential goods, and, more specifically, to the kind of grim and joyless life described before (Sect. 3).

This alternative way of accounting for the overcontrolled behavioral patterns at issue has an interesting consequence: self-control may actually be part of the remedy (rather than being the culprit). To see this, note that both obsessive-compulsive tendencies (or, more precisely, the urges such tendencies give rise to) and negative affective states, such as fear or guilt, are among the “target states” of self-control, i.e., the kind of states that the capacity for self-control can be employed to regulate. Accordingly, the fact that an agent is very self-controlled (i.e., that she is someone who frequently and successfully exercises her capacity for self-control) may actually bring it about that tendencies to display the above forms of overcontrol shape her behavior less rather than more.

To further illustrate this point, note first that one’s capacity for self-control cannot only be employed to establish a new habit. It can also be employed to suppress a habitual response, i.e., to not go through with the response. Consequently, an agent may employ her capacity for self-control to fight her tendency to over-habitualize her life (should she have such a tendency).¹⁶ Relatedly, although plan-making arguably plays an important role for successfully managing temptations, it is equally plausible to assume that an agent can employ her capacity for self-control to suppress an urge to plan, and hence to fight her tendency to engage in over-planning. Analogous points can be made for the case of over-deliberation. Moreover, given that negative affective states like fear seem to be a key contributor to emotional overcontrol, an agent can, arguably, employ her capacity for self-control to fight her tendency to withhold her emotions (as illustrated, e.g., by the case of making an effort to express one’s gratitude or, conversely, moral indignation toward another person). Finally, the reasoning just offered also carries over to the case of over-frugality: if my above claim that a tendency to rigidly choose future prudential goods over present enjoyment is crucially fueled by emotional states like guilt and fear is correct, then we can  plausibly assume that an agent can employ her capacity for self-control to fight such a tendency.¹⁷ There is even some empirical evidence in support of the claim that agents sometimes do employ their capacity for self-control in this way, i.e., to increase the amount of present enjoyment in their lives (e.g., Kivetz & Simonson 2002).¹⁸

There is, of course, no automatism here. An agent also must be aware of displaying the relevant form(s) of overcontrol, judge this to be problematic, and decide to employ her capacity for self-control to reduce their influence. But this does not undermine my general point that self-control (or, more precisely, being someone who is very self-controlled, i.e., who frequently and successfully exercises their capacity for self-control) may actually be a mitigating factor when it comes to the five forms of overcontrol at issue.¹⁹

4.2 The Residue

I have argued for the claim that the five types of overcontrolled behavioral patterns described above (see Sect. 3) can best be accounted for in a way which entails that self-control is not the culprit but may even be part of the remedy. That said, I might not yet have defused all concerns one may have about high levels of self-control involving problematic forms of overcontrol. In this section, I shall therefore take a closer look at what might remain of the overcontrol worry even if we  concede the points made in the preceding section (see Sect. 4.1).

4.2.1 The Misuse View Reloaded

A proponent of the overcontrol worry might raise the following objection: even if the alternative explanations offered before can account for many (if not most) cases of overcontrol, there is still a type of case they cannot account for. For illustration, take the example of an agent who deliberately sets out to establish a great number of very strict habits, or become someone who never shows her emotions, or become someone who always chooses work over pleasure. Here, even if the overcontrolled response may become the “motivational downhill” option with time, an agent surely must repeatedly and successfully exercise her capacity for self-control to get there. Thus, in the kind of cases at issue, self-control (or, more precisely, the fact that an agent has exercised her capacity for self-control too frequently and “too successfully”) is clearly a key causal factor in accounting for the overcontrolled behavior displayed by her.²⁰

We should note, however, that, in the kind of cases just described, it is not the repeated and successful exertion of the agent’s capacity for self-control per se that leads to overcontrol. Instead, the resulting overcontrol is crucially due to the fact that the agent has a specific and rather peculiar goal, which she then tries to achieve by repeatedly employing her capacity for self-control. But this means that the objection at hand ultimately collapses into the misuse worry (see Sect. 2). In essence, it describes just one more way in which agents may employ their capacity for self-control to promote problematic goals. However, it provides no support for the view that there is something intrinsically problematic about the frequent and successful employment of one’s capacity for self-control (or about being someone who frequently and successfully employs their capacity for self-control).

4.2.2 “Self-Control Pleasure” and Obsessions

Another worry is that there might be something about certain forms or techniques of self-control that can lead to the development of overcontrolled behavior irrespective of the agent’s goals. This worry is fueled by the observation that there can be a certain pleasure to specific forms or techniques of self-control, such as, e.g., thrift or carefully crafted to-do lists. But if this is the case, then one might wonder whether there is a risk that things could get out of hand and develop into an obsession.²¹

In response, I first want to point out that we would, again, have to take a very careful look at whether the displayed behavior really is self-controlled behavior (i.e., behavior that is brought about by a successful exercise of the agent’s capacity for self-control) or rather driven by other factors. Excessive thrift, for instance, can also be driven by affective states and thus be a manifestation of a lack of self-control (see Sect. 4.1).²² Likewise, while carefully crafted to-do lists may be measures of self-control, they need not be. Instead, they may simply be the result of acting on certain urges, such as an “urge to plan” (see Sect. 4.1).

But let us grant for the sake of argument that there are some cases where agents initially engage in thrift or certain forms of planning behavior as a self-control measure (i.e., as part of a “motivational self-intervention”, see Sect. 4.1), discover that they find the respective activity (highly) pleasurable, and consequently engage in it to a degree that could be labelled  obsessive. Would such cases of “self-control obsessions” (as I shall put it) lead to a resurrection of the overcontrol worry (see Sect. 2)?

I believe that the correct answer to this question is “no”. To see why, it is important to be very clear about what is supposed to happen in such cases: first, an agent’s repeatedly engaging in a certain self-control measure is assumed to create a disposition to experience “self-control pleasure.” This disposition, then, is supposed to increase her motivation to engage in that measure to the point that the relevant behavior (e.g., thrift behavior, certain forms of planning behavior, etc.) becomes obsessive.

Thus, all that such cases describe is a possible indirect effect of repeatedly engaging in certain forms of self-control (“indirect” because the assumed effect is mediated by an agent’s disposition to experience “self-control pleasure”). However, what they do not show is that there is anything intrinsically problematic about being very self-controlled (or, more precisely, about being someone who frequently and successfully exercises their capacity for self-control). Nor does it seem especially plausible to assume that they describe a typical effect of being very self-controlled. It is, of course, ultimately an empirical question whether and how often “self-control obsessions” occur.²³ My conjecture is, though, that, if they occur at all, their occurrence requires that there is already an obsessive streak in the agent’s personality on which the deployed self-control measure can “piggyback” (such as, e.g., a strong need for controlling one’s environment or a strong need for orderliness). But if I am correct about this, then engaging in certain forms or techniques of self-control can  at best be regarded as a contributory factor for the development of obsessions, and whether an agent is at risk for this at all depends on what else is true about her personality.

In sum, even if we concede that “self-control obsessions” occasionally occur, such cases merely illustrate that there might be an indirect causal link between high levels of self-control and certain forms of overcontrolled behavior, or, more precisely, that the former might be a contributing cause of the latter. However, what the supposed phenomenon of “self-control obsessions” does not show is that there is anything intrinsically problematic about high levels of self-control. Thus, although certainly very interesting, it does not lead to a resurrection of the overcontrol worry as it was stated above (see Sect. 2).

Finally, it should (once again) be pointed out that an agent could employ various other forms of self-control to fight a “self-control obsession” (see Sect. 4.1). An agent may, for instance, employ willpower to resist the temptation to craft yet another meticulous to-do-list or use social support to fight excessive thrift behavior. Hence, even if there should be some cases where certain forms of self-control causally contribute to the development of certain forms of overcontrolled behavior, self-control would (again) be among the remedies.

5 Self-Control and Evaluative Discovery

So far, I have focused on the overcontrol worry and argued that it is largely unfounded. Yet, even if one agrees that the threat of overcontrol has been dispelled, a further concern about having “too much” self-control might still remain. This concern is rooted in the widely held view that exercises of (one’s capacity for) self-control aim at promoting conduct which is in line with one’s best judgment.²⁴ More precisely, the concern is that, due to this feature, being someone who exercises their capacity for self-control too frequently and too successfully will prevent one from making certain valuable “evaluative discoveries,” since the making of such discoveries requires occasionally acting against one’s best judgment.

The concern that being “too self-controlled” might hinder valuable evaluative discoveries comes in several version. A first version can be found in Kennett (2001). In a nutshell, Kennett’s argument proceeds as follows: sometimes an agent’s emotional responses, rather than her best judgments, are “indicators of what is really valuable” (as illustrated by cases of inverse akrasia) (Kennett, 2001, p. 148). Hence, occasional failures to bring one’s behavior in line with one’s best judgments in the face of such recalcitrant emotions provide one with a valuable source of “evidence of evaluative error” (p. 148)—a source which agents who exercise their capacity for self-control too frequently and too successfully will lack.

I fully agree with Kennett’s claim that one’s recalcitrant emotions can be more reliable indicators of what is really valuable than one’s best judgments. That said, I think that Kennett is too quick in claiming that this evidential source is not available to an agent who only rarely (or even never) suffers a failure of self-control. After all, acting on one’s recalcitrant emotions against one’s best judgment is not the only way to use one’s “affective resistance” as a source for identifying evaluative error. Agents who encounter affective resistance may alternatively simply take this as an opportunity to step back and re-open their deliberative process which may then lead to a revision of their best judgment.

However, this response may not yet dispel all concerns one might have about self-control’s potential to prevent valuable evaluative discoveries. Consider the following line of thought: sometimes an agent might simply need to do something to realize the value of it. But to gain such direct experience, she might sometimes need to act against her best judgment and hence display a failure of self-control.²⁵  To illustrate, think about a person belonging to a strict religious cult who, due to a failure of self-control, discovers that going to a dance is actually innocent fun (rather than a deeply sinful act).²⁶ Or think of someone who, due to a failure of self-control, discovers that she enjoys going out on school nights much more than going out on weekends. It seems that these agents would never have discovered the value of the respective activities if they had not been “less than perfectly self-controlled.”

Let me make two points in reply. First, I agree that there are cases where an agent fails to make a valuable evaluative discovery partly because her capacity for self-control is working “too well.” The qualification “partly” is important here, however, because, at least in the above examples, evaluative discovery is also crucially hindered by the fact that the agent has overly rigid values. Indeed, one might even maintain that the latter is the decisive factor , while the agent’s “perfect” self-control is merely a contributary one.

Second, self-control can also promote direct experiences and thereby valuable evaluative discoveries. We can see this once we take into account that self-control is also often needed to fight anxiety, shyness and similar phenomena, which, arguably, can be significant obstacles when it comes to having certain direct experiences and thus valuable evaluative discoveries. For further illustration, think about a person with (moderate) social anxiety, who, due to an exercise of self-control, enables herself to participate in a local mushroom-hunting event and thereby discovers that she thoroughly enjoys this activity. Here, it seems that the relevant discovery might never have been made if the person had not engaged in a successful exercise of her capacity for self-control. In sum, once we take a closer look, we see that self-control can both hinder and promote valuable evaluative discoveries.

For this reason, we cannot say that being very (or even perfectly) self-controlled is generally problematic because it precludes one from making valuable evaluative discoveries. Rather, when it comes to such discoveries, self-control can either help or hinder, depending on which further features and dispositions the relevant agent possesses.

6 Conclusion

In the paper, I have discussed the commonsensical position that one cannot only have “too little”, but also “too much” self-control. In particular, I examined in detail the view that someone who is very self-controlled (or, more precisely, someone who frequently and successfully exercises their capacity for self-control) will manifest various problematic forms of overcontrol. However, despite the overcontrol worry's considerable initial appeal, it proved largely unfounded: as I have argued, standard everyday examples of overcontrol can best be accounted for in a way that, rather than appealing to an excessive employment of the agent’s capacity for self-control, does not refer to an agent’s self-control at all. Moreover, in an interesting turn-about, I have shown that self-control may even play an important role for reducing overcontrolled behavior.

As then became clear, there may still be some cases where high levels of self-control act as a contributory factor in the development of certain forms of overcontrolled behavior. Specifically, an agent may deliberately seek to establish certain forms of overcontrolled behavior by the repeated (and successful) exercise of her capacity for self-control. I argued, though, that cases of this kind ultimately describe just one more way in which an agent might “misuse” her capacity for self-control, i.e., employ it in the service of problematic goals. However, they do not show that there is any intrinsic link between high levels of self-control and problematic forms of overcontrol. The same, I maintained, held for cases of “self-control obsessions”: such cases merely support the view that there may sometimes be an indirect causal link between the frequent and successful deployment of certain forms of self-control and certain forms of overcontrolled behavior, depending on what else is true about the relevant agent’s personality.

Finally, I discussed the additional worry that being very self-controlled is problematic because it is an obstacle to valuable evaluative discoveries. In particular, I focused on the view that being a paragon of self-control will prevent an agent from discovering the value of certain activities because it will prevent her from having certain direct experiences. In response, I pointed out that, in some such cases, the main culprit is actually an agent’s overly rigid values rather than her overly high level of self-control. Moreover, I also showed that self-control cannot just be an impeding, but also a facilitating factor when it comes to direct experiences and, consequently, valuable evaluative discoveries, depending on which further features and dispositions the relevant agent possesses.

To sum up, the various strands of reasoning offered in this paper seem to converge on the following view: while an agent may certainly have “too much” self-control in the sense that she may employ her capacity for self-control to promote problematic goals (or values), there need not be anything intrinsically problematic about being a paragon of self-control.