Social Influence on "Self "-Control

Abstract

As Duckworth and Kern (2011) note, currently over 1% of the abstracts in PsycInfo are indexed by "selfcontrol" or one its synonyms. As part of this widespread interest, cognitive and neural scientists are debating the psychological mechanisms of self-control (Ainslie, 1975; Metcalfe & Mischel, 1999; Muraven & Baumeister, 2000), and the implementation of these mechanisms in the brain (Figner, et al., 2010; Hare, Camerer, & Rangel, 2009; Hare, Malmaud, & Rangel, 2011; Kable & Glimcher, 2007, 2010; McClure, Ericson, Laibson, Loewenstein, & Cohen, 2007; McClure, Laibson, Loewenstein, & Cohen, 2004). These efforts, however, currently proceed without much agreement on a theoretical or operational definition regarding what constitutes "self-control" (Duckworth & Kern, 2011). Definitions have been offered, of course, but one gets the sense that many investigators are content defining self-control in much the same manner that American courts define pornography "I know it when I see it" ( Jacobellis vs Ohio, 1964). Just as our intuitions regarding physics can be mistaken, so too can our intuitions regarding psychology (Stanovich, 1985). This essay argues that an over-reliance on "intuitive psychics" is hindering efforts to identify the cognitive and neural processes involved in self-control. Specifically, current theories tend to underemphasize or ignore completely a factor of critical importance – the social world. Yet, "self-control" is a concept that only emerges at the level of the person in society: it is the social world that defines what is and is not a self-control problem. This realization has important implications for people interested in cognitive and neural mechanisms: it suggests that self-control is unlikely to be a single process; that the computation of social norms is an understudied process that is likely critical for self-controlled behavior; and that interventions that target the social context to increase the influence of norms may prove the strongest way to increase self-controlled behavior. Disciplines Other Psychiatry and Psychology | Psychology This conference paper is available at ScholarlyCommons: https://repository.upenn.edu/psychology_papers/29 As Duckworth and Kern (2011) note, currently over 1% of the abstracts in PsycInfo are indexed by ―self-control‖ or one its synonyms. As part of this widespread interest, cognitive and neural scientists are debating the psychological mechanisms of self-control (Ainslie, 1975; Metcalfe & Mischel, 1999; Muraven & Baumeister, 2000), and the implementation of these mechanisms in the brain (Figner, et al., 2010; Hare, Camerer, & Rangel, 2009; Hare, Malmaud, & Rangel, 2011; Kable & Glimcher, 2007, 2010; McClure, Ericson, Laibson, Loewenstein, & Cohen, 2007; McClure, Laibson, Loewenstein, & Cohen, 2004). These efforts, however, currently proceed without much agreement on a theoretical or operational definition regarding what constitutes ―self-control‖ (Duckworth & Kern, 2011). Definitions have been offered, of course, but one gets the sense that many investigators are content defining self-control in much the same manner that American courts define pornography – ―I know it when I see it‖ (Jacobellis vs Ohio, 1964). Just as our intuitions regarding physics can be mistaken, so too can our intuitions regarding psychology (Stanovich, 1985). This essay argues that an over-reliance on ―intuitive psychics‖ is hindering efforts to identify the cognitive and neural processes involved in self-control. Specifically, current theories tend to underemphasize or ignore completely a factor of critical importance – the social world. Yet, ―self-control‖ is a concept that only emerges at the level of the person in society: it is the social world that defines what is and is not a self-control problem. This realization has important implications for people interested in cognitive and neural mechanisms: it suggests that self-control is unlikely to be a single process; that the computation of social norms is an understudied process that is likely critical for self-controlled behavior; and that interventions that target the social context to increase the influence of norms may prove the strongest way to increase self-controlled behavior. Self-control at the Behavioral Level: A Descriptive List What is ―self-control‖? We begin by looking at the answer to this question at a more holistic level, in terms of the specific behaviors people engage in that exemplify self-control or the lack of self-control. Ultimately, this essay is focused on the answer at a different level of analysis, that of the cognitive and neural mechanisms involved in self-control. However, it is easier to start by considering specific behaviors that exemplify self-control, a question on which there seems to be more agreement, before turning to the debate over what the underlying mechanisms of those behaviors are. So, what behaviors get labeled ―self-controlled‖? Though in general we should have a healthy suspicion of our psychological intuitions as explanations, here those intuitions are the topic of study. That is, what does the concept of ―self-control‖ refer to, as used in the natural language? Before considering the validity of intuitive or other explanations for what self-controlled acts share, it is necessary to know what the category of self-controlled acts is, as concretely as possible. The following examines how and when the term ―self-control‖ is used to describe behavior, including the examples psychologists and neuroscientists use, as well as systematic investigation of the acts people nominate as representing self-control (Tsukuyama, Duckworth, & Kim, in press). One category of acts that get called self-controlled involves avoiding vices. These vices include foods that are bad for you, cigarettes, alcohol, drugs, gambling, impulse buying, pornography, and cheating on one‘s romantic partner. The dieter, teetotaler, and tightwad are held up as having good self-control (in some cases too good). Controlling one‘s anger might also be placed in this category, since it also involves avoiding an action that, at least in the short-term, feels pleasurable. Interestingly, most of the acts in this category often get couched in the language of addiction. So one might summarize this category by saying that, if you can ―seek help‖ or go to rehab for it, it‘s a behavior that someone has ascribed to a lack of self-control. A second category of acts that get labeled self-controlled involves hard work – grit, persistence, effort, etc. This includes exercising to stay fit, practicing to master a new skill, studying to do well in school, and completing one‘s jobs at work. The A-student, the worker who meets deadlines, and the athlete who puts in long hours at the gym are held up as having good selfcontrol. By contrast, procrastination and laziness are ascribed to a lack of self-control. Rather than avoiding an action that in the short-term feels pleasurable, these behaviors involve persisting in an action that in the short-term feels painful. Already, this suggests a broad division between acts of self-control that involve reigning in approach tendencies and acts of self-control that involve reigning in avoidance tendencies. There are other behaviors that get ascribed to a lack of self-control that do not fit neatly into one of these two broad categories. These include criminal behavior, dishonest behavior such as lying, and socially inappropriate behavior such as speaking without thinking or being overly revealing. Self-control at the Mechanistic Level: Three Prominent Accounts and Their Weaknesses This list of concrete, observable behaviors serves as the target for potential explanations of selfcontrol at the level of cognitive and neural processes. Several accounts have been proposed, each seeking to explain the commonality across all of the self-controlled behaviors in this list. The term ―self-control‖ implies that there is something being controlled and something doing the controlling. Most mechanistic accounts follow this implication, positing a process that implements control and a process that requires controlling. In this respect, it is important to note that we can be mislead by the implications bundled into the term itself. The three prominent accounts discussed below are all framed as competing systems – one a controller and one a controlee. But nothing about the self-control dilemmas these accounts seek to explain necessitates an assumption of competing systems or processes. All self-control dilemmas necessarily entail is a conflict between two actions – eating dessert or sticking to your diet, procrastinating or studying, breaking the law or following it. A theory of self-control only needs to explain the source and essence of this conflict. This conflict could arise from competing processes, but it need not do so. 1 There are three prominent accounts of the conflict inherent in self-control dilemmas. While each pinpoints important insights, all three have significant flaws as a complete mechanistic explanation. Further, these accounts also fall short of characterizing the necessary and sufficient conditions for an act of self-control. 1 As one example, Joe McGuire and I (2011) have proposed an explanation for failures to persist in delay-ofgratification that does not involve competing systems. Rather, there is only a growing conflict as one waits between the value of continued waiting and value of abandoning the delayed reward. A first account characterizes the self-control conflict as requiring the inhibition of automatic responses to achieve one‘s goals (or equivalently, the inhibition of ―dominant responses,‖ ―urges,‖ ―desires,‖ or ―impulses‖) (for example, Muraven & Baumeister, 2000). This description nicely captures that there must be an inclination to engage in the activity that is being avoided. We do not typically call someone who avoids drinking alcohol because they don‘t like the taste self-controlled. However, while the inhibition of an automatic response could be a necessary condition, it is not sufficient for an act to qualify as ―self-control‖. When someone leaves the house Saturday morning to drive to the store, and successfully catches themselves before driving on autopilot to work, they are certainly inhibiting an automatic response, but this seems quite far from the prototypical self-controlled behaviors listed above. Behaviors like this one have been widely studied under the rubric of ―cognitive control‖ (for example, Miller & Cohen, 2001), a concept that seems to be overlapping with but not identical to self-control. There is a bigger issue with this inhibition account, though. What does the inhibiting? Often, the inhibiting force is called ―controlled cognition,‖ the ―will,‖ or the ―self.‖ As many scholars have noted, these terms seems an awful lot like homunculi (Kurzban, 2010; Wegner, 2002). Although homunculi can have a role in psychological theorizing, that role is to ―stand-in‖ for a process or mechanism that we do not yet understand (Attneave, 1960; Baddeley, 2001). A homunculus is a placeholder for a theory, not a theory in itself. A second description of self-control dilemmas focuses on the timecourse of costs and benefits, and locates the conflict between immediate and delayed considerations (for example, Ainslie, 1975). 2 The addict chooses an immediate reward, the high of drugs, over a larger delayed reward, better long-term health. The procrastinator avoids an immediate pain, the drudgery of work, but incurs a larger delayed pain, the costs of failing to complete that work. In contrast, self-controlled behaviors usually involve choosing a long-term gain at the expense of a shortterm pain. This account nicely captures the fact that most self-controlled acts have a similar timecourse of costs and benefits. However, while conflict between the short and long-term is a frequent characteristic of selfcontrol dilemmas, patience alone (i.e., choosing delayed over immediate rewards) does not seem to be the essence of self-control. In many cases, choosing immediate outcomes at the expense of delayed ones is not seen as a failure of self-control. For instance, we do not begrudge the lottery winner who buys their parents a new house rather than invests all of their winnings, nor do we criticize the aging retiree who moves their accounts into cash and bonds. Correspondingly, acting in one‘s long term interests is not always attributed to self-control. For example, consider the small businessperson who sells you an initial product at a loss to gain you as a customer, or the bird who hides away seeds and comes back to eat them months later. 2 This conflict is often described as one between a system that values only immediate outcomes and a system that is more future-oriented (McClure, et al., 2007; McClure, et al., 2004). But, as I have argued elsewhere (Kable & Glimcher, 2007, 2010), this need not be the case. Both the behavioral and neural evidence is consistent with a single system that evaluates both immediate and delayed outcomes. The conflict arises, then, when that system evaluates two actions as being quite close in overall value, despite having different temporal profiles. There is also a further issue with the impatience account, which is that only some immediately available outcomes are associated with self-control problems, while others are not. When I walk through the corner market, I may be ―tempted‖ by the chocolate ice cream but not by the vanilla right next to it, despite both being available immediately. One of the important findings of Mischel‘s delay-of-gratification experiments was that performance was not determined by immediacy per se. The marshmallow hidden under a bowl was easier to resist that than the one in plain sight, despite the fact that both were available with the same immediacy (Mischel, Shoda, & Rodriguez, 1989). A third explanation situates the self-control conflict as one between ―hot‖ emotional processes and ―cool‖ cognitive ones (for example, Metcalfe & Mischel, 1999). Here self-control involves the recruitment of cognitive processes to blunt the influence of emotional cues on behavior. This account nicely points attention towards the aspect of ―temptation‖ missed by purely time-based accounts. However, the ―hot‖ system metaphor seems better suited for approach dilemmas, where a positive emotion motivates you towards an activity you would like to avoid, than for avoidance dilemmas, where a negative emotion motivates you away from an activity you would like to engage in. For someone exercising or studying or practicing, quitting that activity does not seem like a particularly ―hot‖ option. Often, cognitive rationalizations – ―I‘ll run tomorrow instead‖ – are a more salient feature. There are also a few larger theoretical issues with the ―hot‖ versus ―cool‖ systems account. One issue is that this account relies on a distinction between cognition and emotion that many scholars argue cannot be cleanly drawn (Lindquist, Wager, Kober, Bliss-Moreau, & Barrett, in press). In addition, this account seems to elide any role for emotions in adaptive decisionmaking, despite several strong theoretical arguments that emotions function to improve decisions (Cosmides & Tooby, 2000; Damasio, 1994). In the specific context of self-control, this account seems to ignore a whole host of ―hot‖ emotions – shame, embarrassment, guilt, pride – that more likely act to promote self-control than to thwart it. 3 What All of These Accounts Miss: The Social World This last point provides a critical clue about what is missing in all three of these process-level accounts. Shame, embarrassment, guilt and pride have been called the ―social‖ emotions, because they function to promote socially appropriate behavior, behavior in line with social or cultural standards (Adolphs, 2003). It is the existence of such a social or cultural standard, which behavior lacking self-control violates, that is deemphasized or ignored in the above process-level accounts. In fact, the essence of the conflict in self-control dilemmas may be between what you want to do (that is, what you have an inclination or impulse to do) and what you should do – or, at least, what other people think you should do. 3 These points all argue that the ―hot‖ system construct is too broad. Most of what it seeks to explain, however, seems well captured by the narrower construct of ―Pavlovian‖ processes, as described for example by Dayan and colleagues (Dayan, Niv, Seymour, & Daw, 2006). A couple of examples illustrate the necessary role that social norms play in what gets labeled self-controlled behavior. Consider a Japanese Samurai in the middle ages, who commits seppuku (a type of ritual suicide) because of shame he has brought onto himself. While in many cases suicide is considered impulsive, I think seppuku would in contrast be considered self-controlled. The act is premeditated, occurs in a ritualized public performance, and is considered (in that time and place) honorable. There is no clear cognition-emotion conflict, since emotions such as fear and shame factor on both sides, and there is no conflict between the shortand long-term, since clearly the action is costly in both. What is shared with other self-controlled actions, though, is that despite a strong inclination to do otherwise, the actor follows through on the behavior required by his cultural norms. Seppuku may be a heavy-handed example. Consider a light-hearted one, contained in a personal anecdote. When my daughter was around eighteen months old, we would reenact the same situation every evening. We would come home and enter the kitchen. There would be a bowl of water put out for our cats, and if I did not remember quickly enough to move the bowl out of reach, my daughter would walk over and dump the water – usually soaking herself as well as the floor. After this had happened probably a dozen times, I pleaded, ―Why is the cat‘s water so tempting?!?‖ Notice the implicit inference, couched in the language of temptation and selfcontrol. Could my daughter have had an irresistible impulse to turn over water bowls? Perhaps, though that seems a little silly. Was there a long-term cost to such behavior? Not really. But she was engaging in behavior that I, at least, had deemed inappropriate, and when she did not respond to me telling her so my first linguistic impulse was to blame this on a lack of selfcontrol. 4 The place of social norms in defining what is self-controlled explains that third, uncategorized group of behaviors listed above. What do committing a crime, lying, and speaking inappropriately all share with each other and with the other behaviors listed? The foremost thing these behaviors share is that they violate social norms and standards. In addition, the necessary role of social standards in delineating what is self-controlled helps to explain an interesting fact about self-control – that we are reluctant to attribute it to other animals. For neurobiologists, the idea that self-control is unique to human beings is very puzzling if self-control is simply choosing patiently or overriding automatic behaviors. Other animals also plan and sacrifice for the distant future, such as the scrub jays that cache food based on their future (rather than present) needs (Clayton, Bussey, & Dickinson, 2003). Other animals also exhibit goal-directed behavior, behavior that is not simply ―automatic‖ or ―habitual‖ (Balleine & Dickinson, 1998). Yet such animal behaviors are not typically considered examples of ―self-control.‖ What these behaviors lack, though, is that they do not fall in line with a clear social or cultural standard. Although there is some evidence for culture in other primates (de Waal, 2001), there is still a large gulf in cultural sophistication and elaboration between humans and other animals. And it this gulf – that humans are steeped in social norms while other animals 4 A similar kind of issue has been raised by Rachlin (2000) regarding Mischel‘s famous marshmallow test. He argues that the important motivation controlling delay behavior is not the desire for two marshmallows versus one, but rather the desire to comply with an adult‘s (implied) request to wait. are not – likely explains why we attribute the capacity for self-control to humans and not other

Topics

    0 Figures and Tables

      Download Full PDF Version (Non-Commercial Use)