The typical effect-size of depletion experiments (~d = .6; Hagger et al. Reference Hagger, Wood, Stiff and Chatzisarantis2010a) is not the right order of magnitude to go with the idea that there is a literal store of willpower that is being “used up.” Nor is the fact that post-depletion, self-control can be restored by receipt of a gift, or by a self-affirmation (Schmeichel & Vohs Reference Schmeichel and Vohs2009). Compare self-control depletion effects with the actual muscle fatigue from repeatedly lifting a heavy weight – here total muscle failure can easily and reliably be produced, and any effect of self-affirmation is likely to be modest. Although it remains to be seen whether Kurzban et al.'s model has it just right, it is on its face a more plausible account.
However, we think neither Baumeister et al. (Reference Baumeister, Sparks, Stillman and Vohs2008) nor Kurzban and colleagues characterize willpower usefully (although in fairness, only Baumeister and colleagues seem to want to use this term). The phenomenon that both deal with is variously referred to as “executive function,” “conscious processing,” and as the output of “System 2” (Baumeister et al. Reference Baumeister, Sparks, Stillman and Vohs2008). The incongruent condition of the Stroop fits well, and is a standard depletion paradigm. The task requires color naming, which competes with the automatic tendency to read lexical items. There is no question that the Stroop Task is an interesting example of an important category of mental functioning. But the term “willpower” has a more specific meaning – it is not a synonym for “executive function”. In particular, we believe that the mechanisms of willpower are directed at reducing the otherwise marked tendency most people have to systematically change their preferences over time.
The case of addiction is illustrative. The criteria used in the United States for “substance dependence” and for “substance use disorder” include (paraphrasing from the Diagnostic and Statistical Manual of Mental Disorders IV-R) “failed attempts to quit or moderate use” and “repeated episodes in which the individual uses more than she originally planned.” In everyday use of the terms, these central features of addiction are considered struggles of “willpower” and of “self-control.”
What does the self-control struggle of the addict have in common with an executive function task such as the Stroop? This is a point of some disagreement, but we suspect that the answer is not very much. First, peak performance on executive control tasks is observed in early adulthood; performance declines dramatically with aging. If this type of functioning were synonymous with “willpower,” one would have reason to expect addiction to be a rare problem among young adults, but to increase in prevalence as people age. But the opposite is observed, with prevalence highest in early adulthood, and with a large percentage of addicts “aging out” in mid to late life (Anglin et al. Reference Anglin, Brecht, Woodward and Bonett1986). Second, there is a mismatch between the timescale of failures on the incongruent Stroop and the self-control failures of an addict. When a person makes a mistake on the Stroop, she is succumbing to a reflexive tendency to read lexical items. If she is given an opportunity to slow down, she will fix her error. By contrast, when the cocaine addict who has been clean for a month “falls off the wagon,” she may have to go to some lengths to get cocaine. It is not a momentary “oops” that is reliably corrected if she is given a moment to collect her thoughts. Unlike the participant performing the Stroop, the individual looking for cocaine is engaged in sustained goal-directed action, and even complex problem solving.
Of course, the situation of the addict falling off the wagon is so interesting because it is goal-directed action she previously dismissed as undesirable, and which she likely will later regret. Indeed, she may even believe she will regret it, even as she currently devotes herself to obtaining the drug. And these are the critical features that define the domain of willpower/self-control struggle – they involve the recognition of systematic changes in preference over time, and responses to the anticipated inconsistency. The “resolutions” people make are central. The “resolution” is more than a plan; it anticipates a future in which some alternative to that current plan might be more attractive. Moreover, the resolution seems to preemptively apply some force to oppose the foreseen reversal. The nature of that force is not completely understood, but Ainslie (Reference Ainslie1975; Reference Ainslie1992; Reference Ainslie2001) productively suggested that when remembered, a past resolution makes the otherwise appealing plan – for example, to binge today but diet tomorrow – less plausible. I could say, “Just this last one, and then I will be good.” But if I made a similar resolution yesterday, I have reason to believe I cannot both binge today and be confident my resolution about tomorrow will fare any better than yesterday's. In other words, an interest in a particular future behavior, and uncertainty that the interest will be realized, indirectly gives force to resolutions, since they cause current behavior to have added importance beyond what is literally at stake (Monterosso & Ainslie Reference Monterosso and Ainslie1999).
This sort of internal dialog – the resolution, transgression, regret, and back around again to another resolution – is, we think, familiar to most people. To the extent that this process yields regularity in outcomes, the regularities diverge from the patterns highlighted in the will-as-muscle literature. Most notably, a single failure appears to often turn into a protracted run of failures (known as the “abstinence violation effect”; Shiffman et al. Reference Shiffman, Hickcox, Paty, Gnys, Kassel and Richards1996). There is nothing within the depletion account that predicts this phenomenon, but it follows naturally as a collapse of confidence, if the force of a resolution rests in part on the belief that continued resolve is possible if the resolution is kept.
The willpower-as-muscle metaphor has brought attention within the behavioral sciences to willpower struggle. If Kurzban and colleagues are successful in casting doubt on the metaphor's usefulness, then it will be a good time to consider alternative positive accounts of willpower. Intrapersonal bargaining provides, we think, a promising framework.
The typical effect-size of depletion experiments (~d = .6; Hagger et al. Reference Hagger, Wood, Stiff and Chatzisarantis2010a) is not the right order of magnitude to go with the idea that there is a literal store of willpower that is being “used up.” Nor is the fact that post-depletion, self-control can be restored by receipt of a gift, or by a self-affirmation (Schmeichel & Vohs Reference Schmeichel and Vohs2009). Compare self-control depletion effects with the actual muscle fatigue from repeatedly lifting a heavy weight – here total muscle failure can easily and reliably be produced, and any effect of self-affirmation is likely to be modest. Although it remains to be seen whether Kurzban et al.'s model has it just right, it is on its face a more plausible account.
However, we think neither Baumeister et al. (Reference Baumeister, Sparks, Stillman and Vohs2008) nor Kurzban and colleagues characterize willpower usefully (although in fairness, only Baumeister and colleagues seem to want to use this term). The phenomenon that both deal with is variously referred to as “executive function,” “conscious processing,” and as the output of “System 2” (Baumeister et al. Reference Baumeister, Sparks, Stillman and Vohs2008). The incongruent condition of the Stroop fits well, and is a standard depletion paradigm. The task requires color naming, which competes with the automatic tendency to read lexical items. There is no question that the Stroop Task is an interesting example of an important category of mental functioning. But the term “willpower” has a more specific meaning – it is not a synonym for “executive function”. In particular, we believe that the mechanisms of willpower are directed at reducing the otherwise marked tendency most people have to systematically change their preferences over time.
The case of addiction is illustrative. The criteria used in the United States for “substance dependence” and for “substance use disorder” include (paraphrasing from the Diagnostic and Statistical Manual of Mental Disorders IV-R) “failed attempts to quit or moderate use” and “repeated episodes in which the individual uses more than she originally planned.” In everyday use of the terms, these central features of addiction are considered struggles of “willpower” and of “self-control.”
What does the self-control struggle of the addict have in common with an executive function task such as the Stroop? This is a point of some disagreement, but we suspect that the answer is not very much. First, peak performance on executive control tasks is observed in early adulthood; performance declines dramatically with aging. If this type of functioning were synonymous with “willpower,” one would have reason to expect addiction to be a rare problem among young adults, but to increase in prevalence as people age. But the opposite is observed, with prevalence highest in early adulthood, and with a large percentage of addicts “aging out” in mid to late life (Anglin et al. Reference Anglin, Brecht, Woodward and Bonett1986). Second, there is a mismatch between the timescale of failures on the incongruent Stroop and the self-control failures of an addict. When a person makes a mistake on the Stroop, she is succumbing to a reflexive tendency to read lexical items. If she is given an opportunity to slow down, she will fix her error. By contrast, when the cocaine addict who has been clean for a month “falls off the wagon,” she may have to go to some lengths to get cocaine. It is not a momentary “oops” that is reliably corrected if she is given a moment to collect her thoughts. Unlike the participant performing the Stroop, the individual looking for cocaine is engaged in sustained goal-directed action, and even complex problem solving.
Of course, the situation of the addict falling off the wagon is so interesting because it is goal-directed action she previously dismissed as undesirable, and which she likely will later regret. Indeed, she may even believe she will regret it, even as she currently devotes herself to obtaining the drug. And these are the critical features that define the domain of willpower/self-control struggle – they involve the recognition of systematic changes in preference over time, and responses to the anticipated inconsistency. The “resolutions” people make are central. The “resolution” is more than a plan; it anticipates a future in which some alternative to that current plan might be more attractive. Moreover, the resolution seems to preemptively apply some force to oppose the foreseen reversal. The nature of that force is not completely understood, but Ainslie (Reference Ainslie1975; Reference Ainslie1992; Reference Ainslie2001) productively suggested that when remembered, a past resolution makes the otherwise appealing plan – for example, to binge today but diet tomorrow – less plausible. I could say, “Just this last one, and then I will be good.” But if I made a similar resolution yesterday, I have reason to believe I cannot both binge today and be confident my resolution about tomorrow will fare any better than yesterday's. In other words, an interest in a particular future behavior, and uncertainty that the interest will be realized, indirectly gives force to resolutions, since they cause current behavior to have added importance beyond what is literally at stake (Monterosso & Ainslie Reference Monterosso and Ainslie1999).
This sort of internal dialog – the resolution, transgression, regret, and back around again to another resolution – is, we think, familiar to most people. To the extent that this process yields regularity in outcomes, the regularities diverge from the patterns highlighted in the will-as-muscle literature. Most notably, a single failure appears to often turn into a protracted run of failures (known as the “abstinence violation effect”; Shiffman et al. Reference Shiffman, Hickcox, Paty, Gnys, Kassel and Richards1996). There is nothing within the depletion account that predicts this phenomenon, but it follows naturally as a collapse of confidence, if the force of a resolution rests in part on the belief that continued resolve is possible if the resolution is kept.
The willpower-as-muscle metaphor has brought attention within the behavioral sciences to willpower struggle. If Kurzban and colleagues are successful in casting doubt on the metaphor's usefulness, then it will be a good time to consider alternative positive accounts of willpower. Intrapersonal bargaining provides, we think, a promising framework.