Kurzban et al. run a fruitful attempt to explain why engaging in certain tasks (e.g., self-regulatory tasks) causes aversive experiences of mental effort and drops in task performance. Though Kurzban et al.'s model offers a potentially powerful explanation for a considerable set of empirical findings, at the same time it seems unable to accommodate other empirical results without relying on extra assumptions. We offer an addition to Kurzban et al.'s model in an attempt to explain these seemingly inconsistent findings in a more parsimonious way.
The gist of Kurzban et al.'s argument is that usage of executive function (inevitably related to self-regulatory exertion; cf. Schmeichel Reference Schmeichel2007) carries an opportunity cost, resulting in the subjective experience of effort and, ultimately, task switching. Opportunity costs are derived from comparing benefits linked to engaging in a given executive task with costs associated with not engaging in other tasks requiring the very executive functions already occupied by the initial task. Though Kurzban et al.'s model assumes independence between cost-benefit algorithms and the behavior it attempts to explain, we propose that cost-benefit analyses depend on executive function as well. Indeed, a link between complex thinking and executive function has been demonstrated empirically for tasks involving logic and reasoning or cognitive extrapolation (Schmeichel et al. Reference Schmeichel, Vohs and Baumeister2003). Those tasks seem quite similar to the cost-benefit analyses described by Kurzban et al. This potentially poses a problem, as it implies that the quality of the cost-benefit analysis could depend on how much executive function is available for the cost-benefit calculation itself, which ultimately relies on the demandingness of the initial executive task. Indeed, Kurzban et al. argue rather convincingly that decision makers cannot successfully engage simultaneously in multiple executive tasks competing for executive resources. Research on cognitive load effects supports this claim: When the brain simultaneously engages in two tasks, both requiring executive function (or some minimal level of cognitive involvement), executive function mainly gets devoted to one of the tasks, while the other task gets attended to more automatically (cf. Hinson et al. Reference Hinson, Jameson and Whitney2003). This has implications for the task outcomes: The literature suggests that whether a task or a decision problem is attended to in a relatively more cognitive and effortful way, or a relatively more automatic and effortless way, may substantially change behavior or decisions (e.g., Lee et al. Reference Lee, Amir and Ariely2009).
We want to apply this insight to one particular type of setting studied by Kurzban et al., namely, a so-called depletion setting, in which two different self-regulatory tasks are performed in sequence (Baumeister et al. Reference Baumeister, Bratslavsky, Muraven and Tice1998; Muraven et al. Reference Muraven, Tice and Baumeister1998). Specifically, “depletion” studies typically use a two-task paradigm requiring participants to either engage in self-regulation (i.e., depletion condition) or in a neutral control task (i.e., neutral condition) first. Then, all participants engage in a seemingly unrelated task requiring self-regulation. “Neutral” participants typically obtain higher self-regulation scores in this second phase than do “depleted” participants (cf. Hagger et al. Reference Hagger, Wood, Stiff and Chatzisarantis2010a, for a recent meta-analysis). If our reasoning with respect to the cost-benefit analysis proposed by Kurzban et al. holds, there are two theoretical possibilities when using cost-benefit algorithms to explain depletion effects: Relying on the assumption that executive function faces simultaneity problems (cf. Kurzban et al.) and cost-benefit analyses depend on executive function (our proposition), either the executive task will be attended to in a relatively more effortless way and the cost-benefit calculation in a relatively more effortful way, or the executive task will be attended to in a relatively more effortful way and the cost-benefit calculation in a relatively more effortless way. Depletion conditions typically are relatively more challenging than neutral conditions (i.e., depleted participants engage in self-regulation in the first phase of the experiment already, whereas neutral participants do not), implying that depleted participants will invest relatively more executive resources in the experimental task at hand, and thus will have less to spare for cost-benefit trade-offs compared to neutral participants. Behaviorally, these less thorough cost-benefit analyses in depletion, compared to neutral conditions, translate into reduced self-regulatory performance in the second phase of the study.
Less thorough cost-benefit analyses might cause self-regulatory performance reductions for various reasons. One theoretical possibility could be that, akin to ease-of-retrieval effects (e.g., Winkielman et al. Reference Winkielman, Schwarz and Belli1998), engaging in less thorough cost-benefit analyses and presumably coming up with fewer alternative uses of executive function makes these alternative uses subjectively more obvious and hence more pressing, causing one to cease engagement in the ongoing task sooner. Such a mechanism would also be in line with research on goal dilution, showing that increasing the number of goals that can be fulfilled by a single means (e.g., investing executive function) reduces perceived instrumentality with respect to each goal (Zhang et al. Reference Zhang, Fishbach and Kruglanski2007). This also seemingly suggests that switching to alternative tasks will be more likely when fewer goals are activated. The potential mechanism through which thoroughness of cost-benefit analyses impacts ongoing task performance, however, remains to be tested empirically.
An example of behavioral findings that our adapted model can explain without relying on additional assumptions like practice effects (cf. Kurzban et al.) is: increasing self-regulation in sequential self-regulation tasks that are similar in nature. Indeed, previous research found depletion effects only when regulatory response conflicts differed across sequential self-regulation tasks (e.g., restraining food intake in one task and anagram solving in another), but not when regulatory response conflicts were similar (e.g., restraining food intake in both tasks; Dewitte et al. Reference Dewitte, Bruyneel and Geyskens2009). Our adapted model can explain this pattern as task similarity frees up mental resources (i.e., engaging in two different self-regulatory tasks is more taxing than engaging in one self-regulatory task), causing cost-benefit analyses to be more thorough in high-similarity conditions. Behaviorally, these thorough analyses translate into better (compared to regular depletion conditions) self-regulation in the second phase of the study (cf. previously developed reasoning as to why less thorough cost-benefit analyses might cause self-regulatory performance reductions).
Until now, depletion literature mainly provided motivational (as opposed to cognitive) explanations for the effect, as reflected in Kurzban et al.'s article. However, the field needs a cognitive perspective on depletion also (cf. Inzlicht & Schmeichel Reference Inzlicht and Schmeichel2012). It is our hope that our adapted model can support such initial cognitive steps.
Kurzban et al. run a fruitful attempt to explain why engaging in certain tasks (e.g., self-regulatory tasks) causes aversive experiences of mental effort and drops in task performance. Though Kurzban et al.'s model offers a potentially powerful explanation for a considerable set of empirical findings, at the same time it seems unable to accommodate other empirical results without relying on extra assumptions. We offer an addition to Kurzban et al.'s model in an attempt to explain these seemingly inconsistent findings in a more parsimonious way.
The gist of Kurzban et al.'s argument is that usage of executive function (inevitably related to self-regulatory exertion; cf. Schmeichel Reference Schmeichel2007) carries an opportunity cost, resulting in the subjective experience of effort and, ultimately, task switching. Opportunity costs are derived from comparing benefits linked to engaging in a given executive task with costs associated with not engaging in other tasks requiring the very executive functions already occupied by the initial task. Though Kurzban et al.'s model assumes independence between cost-benefit algorithms and the behavior it attempts to explain, we propose that cost-benefit analyses depend on executive function as well. Indeed, a link between complex thinking and executive function has been demonstrated empirically for tasks involving logic and reasoning or cognitive extrapolation (Schmeichel et al. Reference Schmeichel, Vohs and Baumeister2003). Those tasks seem quite similar to the cost-benefit analyses described by Kurzban et al. This potentially poses a problem, as it implies that the quality of the cost-benefit analysis could depend on how much executive function is available for the cost-benefit calculation itself, which ultimately relies on the demandingness of the initial executive task. Indeed, Kurzban et al. argue rather convincingly that decision makers cannot successfully engage simultaneously in multiple executive tasks competing for executive resources. Research on cognitive load effects supports this claim: When the brain simultaneously engages in two tasks, both requiring executive function (or some minimal level of cognitive involvement), executive function mainly gets devoted to one of the tasks, while the other task gets attended to more automatically (cf. Hinson et al. Reference Hinson, Jameson and Whitney2003). This has implications for the task outcomes: The literature suggests that whether a task or a decision problem is attended to in a relatively more cognitive and effortful way, or a relatively more automatic and effortless way, may substantially change behavior or decisions (e.g., Lee et al. Reference Lee, Amir and Ariely2009).
We want to apply this insight to one particular type of setting studied by Kurzban et al., namely, a so-called depletion setting, in which two different self-regulatory tasks are performed in sequence (Baumeister et al. Reference Baumeister, Bratslavsky, Muraven and Tice1998; Muraven et al. Reference Muraven, Tice and Baumeister1998). Specifically, “depletion” studies typically use a two-task paradigm requiring participants to either engage in self-regulation (i.e., depletion condition) or in a neutral control task (i.e., neutral condition) first. Then, all participants engage in a seemingly unrelated task requiring self-regulation. “Neutral” participants typically obtain higher self-regulation scores in this second phase than do “depleted” participants (cf. Hagger et al. Reference Hagger, Wood, Stiff and Chatzisarantis2010a, for a recent meta-analysis). If our reasoning with respect to the cost-benefit analysis proposed by Kurzban et al. holds, there are two theoretical possibilities when using cost-benefit algorithms to explain depletion effects: Relying on the assumption that executive function faces simultaneity problems (cf. Kurzban et al.) and cost-benefit analyses depend on executive function (our proposition), either the executive task will be attended to in a relatively more effortless way and the cost-benefit calculation in a relatively more effortful way, or the executive task will be attended to in a relatively more effortful way and the cost-benefit calculation in a relatively more effortless way. Depletion conditions typically are relatively more challenging than neutral conditions (i.e., depleted participants engage in self-regulation in the first phase of the experiment already, whereas neutral participants do not), implying that depleted participants will invest relatively more executive resources in the experimental task at hand, and thus will have less to spare for cost-benefit trade-offs compared to neutral participants. Behaviorally, these less thorough cost-benefit analyses in depletion, compared to neutral conditions, translate into reduced self-regulatory performance in the second phase of the study.
Less thorough cost-benefit analyses might cause self-regulatory performance reductions for various reasons. One theoretical possibility could be that, akin to ease-of-retrieval effects (e.g., Winkielman et al. Reference Winkielman, Schwarz and Belli1998), engaging in less thorough cost-benefit analyses and presumably coming up with fewer alternative uses of executive function makes these alternative uses subjectively more obvious and hence more pressing, causing one to cease engagement in the ongoing task sooner. Such a mechanism would also be in line with research on goal dilution, showing that increasing the number of goals that can be fulfilled by a single means (e.g., investing executive function) reduces perceived instrumentality with respect to each goal (Zhang et al. Reference Zhang, Fishbach and Kruglanski2007). This also seemingly suggests that switching to alternative tasks will be more likely when fewer goals are activated. The potential mechanism through which thoroughness of cost-benefit analyses impacts ongoing task performance, however, remains to be tested empirically.
An example of behavioral findings that our adapted model can explain without relying on additional assumptions like practice effects (cf. Kurzban et al.) is: increasing self-regulation in sequential self-regulation tasks that are similar in nature. Indeed, previous research found depletion effects only when regulatory response conflicts differed across sequential self-regulation tasks (e.g., restraining food intake in one task and anagram solving in another), but not when regulatory response conflicts were similar (e.g., restraining food intake in both tasks; Dewitte et al. Reference Dewitte, Bruyneel and Geyskens2009). Our adapted model can explain this pattern as task similarity frees up mental resources (i.e., engaging in two different self-regulatory tasks is more taxing than engaging in one self-regulatory task), causing cost-benefit analyses to be more thorough in high-similarity conditions. Behaviorally, these thorough analyses translate into better (compared to regular depletion conditions) self-regulation in the second phase of the study (cf. previously developed reasoning as to why less thorough cost-benefit analyses might cause self-regulatory performance reductions).
Until now, depletion literature mainly provided motivational (as opposed to cognitive) explanations for the effect, as reflected in Kurzban et al.'s article. However, the field needs a cognitive perspective on depletion also (cf. Inzlicht & Schmeichel Reference Inzlicht and Schmeichel2012). It is our hope that our adapted model can support such initial cognitive steps.