1
Motivation and Cognitive Control
Introduction
Todd S. Braver
| Experimenter: | So what did you think of the task? |
| Participant #1: | Well, I tried my best, but I got two or three wrong, and it made me mad. Would you let me try it again? Iām pretty sure I could do better next time. |
| Experimenter: | Excuse me, but Iām noticing that you arenāt really looking at the display when responding, and you seem to be pushing the buttons randomly. |
| Participant #2: | Yeah, well, Iām only doing this experiment as a course requirement, and itās pretty boring. I guess I just zoned out there for a while. |
The contrast between the motivated and unmotivated participants illustrated in the foregoing two scenarios is likely to be very familiar to any investigator who has conducted human experimental research. Indeed, the ubiquitous nature of these types of scenarios makes them seem almost trivial. Classically, in cognitive studies, participant motivation is treated as a source of noise or measurement error that the investigator attempts to minimize. A standard approach is to just assume that providing participants with performance instructions (e.g., respond as quickly but as accurately as you can) will be sufficient. This assumption is a tenable one, because most experimental participants do seem to perform even very challenging tasks highly successfully, just from instructions alone. Yet it is also very well known that providing additional constraints, such as feedback or other forms of incentive, especially concrete ones, such as monetary bonuses, can have a significant impact on task performance. This truism underlies the widespread use of incentives and performance feedback in a variety of educational, business, and other real-world settings (Bettinger, 2012; Bonner & Sprinkle, 2002; Fryer, 2011; Garbers & Konradt 2014; Smith & Walker, 1993). Yet such practices appear to beg the question: Given that participant motivation can strongly contribute to cognitive task performance, how and why does this occur?
It is my own fascinationāand struggleāwith this question that led to the genesis of this book. Specifically, this book arises from the belief that further progress in our understanding of cognition, and in particular the control processes that regulate cognitive processing, will critically depend upon our making progress in understanding the nature of motivation-cognition interactions. The present volume serves as a snapshot of the current state of this progress. In this introductory chapter, I give an overview of the key themes and complexities that have arisen in the study of motivation-cognition interactions, while also providing a roadmap to the various perspectives and approaches taken by the bookās contributors in their respective chapters.
Motivation and Cognitive Control
Over the last decade, there has been a rapidly growing interest in the investigation of motivational influences within cognitive psychology and cognitive neuroscience. In these investigations, rather than trying to minimize motivational contributions to cognitive task performance, the experimental design instead tries to assess these contributions directly, through motivational manipulations and/or assessment of motivation-related individual differences. The focus is also somewhat different from the types of performance incentive studies conducted in industrial/organizational psychology and behavioral economics (Bonner & Sprinkle, 2002; Garbers & Konradt, 2014; Smith & Walker, 1993). In particular, the goal is not primarily to establish whether a given motivational manipulation affects task performance, but rather to use the tools of modern cognitive psychology and neuro-science to better understand which components of performance are affected, and to link these components to specific cognitive processes and neural mechanisms.
Out of this work there has been a growing list of cognitive processes that appear to be specifically enhanced by the presence of motivational incentives, including active maintenance in working memory, task-switching, selective attention, response inhibition, episodic memory encoding, and decision making (Locke & Braver, 2010; Maddox & Markman, 2010; Pessoa, 2009; Pessoa & Engelmann, 2010; Shohamy & Adcock, 2010). These findings are well reviewed by many of the contributors to this volume. Yet a common theme running through much of the work has been an emphasis on the particularly strong relationship between motivation and cognitive control. This emphasis has been clear ever since the first high-profile studies in this area, as even a cursory scan of the literature makes clear. For example, in a seminal study from the behaving primate neurophysiological literature, cueing of reward incentives was found to selectively sharpen working memoryārelated activation for the upcoming target in dorsolateral prefrontal cortex (PFC) neurons, suggesting a mechanism of control enhancement (Leon & Shadlen, 1999). Another study using the n-backāa paradigmatic task of working memory and cognitive control in human cognitive neuroscienceāfound that parametric increases in motivational value impacted the same lateral PFC regions modulated by parametric increases in cognitive control demands (Pochon et al., 2002). Finally, in a cognitive study of task-switching, it was found that adding reward incentives specifically reduced task-switch costs, by increasing the likelihood of preparatory control (Nieuwenhuis & Monsell, 2002).
The potentially special relationship between motivation and cognitive control has not only been a prominent feature of experimental studies but also has been noted in theoretical accounts of the literature (Botvinick & Braver, 2015; Pessoa, 2009; Sarter, Gehring, & Kozak, 2006). The relationship is an intuitive one, in that cognitive control is often defined as the set of processes that regulate thought and action based on internally maintained task goals. As such, it seems natural that motivational signals might serve a prioritization function in biasing the selection, activation, and intensity level of such task goals. Moreover, control modulation might serve as an important mediating route by which motivation has an influence on a wide range of cognitive processes and behavioral performance metrics. For example, my colleagues and I have suggested that motivational signals could drive a shift towards proactive cognitive control, a computationally (and potentially metabolically) expensive mode of control, in which sustained maintenance of goal-relevant information (within lateral PFC) is utilized to optimize goal-driven biasing of attention, perception, and action systems (Braver, 2012; Braver, Gray, & Burgess, 2007; Jimura, Locke, & Braver, 2010; Locke & Braver, 2008). Other accounts have emphasized the role of motivational processes in modulating cognitive control through a shift in the allocation of general processing resources (Pessoa, 2009) or by increasing the expenditure of attentional effort (Sarter et al., 2006).
Broader Perspectives
The cognitive psychology and cognitive neuroscience literature provides an attractive and potentially informative perspective on the nature of interactions between motivation and cognitive control. However, it is not the only one. Indeed, there are much longer traditions of motivationally focused research within other areas of psychology and neuroscience. For example, in the animal learning and behavioral neuroscience tradition, there is a rich body of data that addresses mechanisms of incentive learning with associated experimental paradigms. A particular focus is on the distinction between Pavlovian and instrumental incentive learning, with a further instrumental distinction between habitual and goal-directed behavioral control, each of which shows dissociable effects of motivational manipulations (Dickinson & Balleine, 2002). These conceptual distinctions are reflected in the reinforcement learning computational framework, which postulates specific mechanisms for learning the reward value of stimuli and actions (i.e., reward prediction errors; Sutton & Barto, 1998). The reinforcement learning framework has become a highly influential point of contact between the computational and cognitive neuroscience literatures, particularly in the domain of decision making (Daw & Shohamy, 2008; Niv, 2009).
At the other end of the spectrum, social, affective, and personality psychology have also had a long tradition of examining motivational influences on behavior. In this tradition, there is a strong focus on trait-like properties of the individual, such as his or her beliefs and motives (i.e., the class of incentives the individual finds intrinsically attractive; McClelland, 1987), the situational variables that modify incentive values of desired outcomes (e.g., by altering the feasibility and expectancy that they will be attained; Heckhausen, 1977), and the reciprocal relationships between motivation and affect (Buck, 1985; Carver, 2006; Carver & Scheier, 1990). Another important emphasis, which suggests a potential point of contact with cognitive neuroscience conceptualizations, is on the role of goal representations as the mediating variable for motivational influence (Bargh, Gollwitzer, & Oettingen, 2010; Elliot & Fryer, 2008). However, in this literature, a key distinction is made between the motivational factors that influence how goals are selected and how they are pursued (Gollwitzer, 2012). Indeed, the more recent work has also suggested that motivational factors can modulate goal selection and pursuit even outside of conscious awareness, prompting investigations into subliminal and indirect motivational manipulations (Custers & Aarts, 2010).
Finally, a third domain of psychology and neuroscience, which is also becoming increasingly influenced by motivation-focused perspectives, is that of aging and development. Within the aging literature, there has been a newfound appreciation of the fact that age-related cognitive decline might be importantly moderated by motivational factors (Nielsen & Mather, 2011). In particular, there is a strong focus on motivational reprioritization, indicating that older adults may emphasize and select behavioral goals that are associated with positive affect, have high self-relevance, and center on maintenance or loss prevention, rather than acquisition or growth (Carstensen & Mikels, 2005; Hess, 2014). On the developmental side, the emphasis has been on the maturational trajectories of cognitive vs. affective/motivational neural circuits. A primary focus has been on the adolescent period, which may be marked by oversensitivity to appetitive goals and reward incentive cues (Luciana & Collins, 2012; Somerville & Casey, 2010). This developmental work has a clear point of contact with the cognitive neuroscience literature, in that they both have employed similar experimental tasks, designs, and motivational manipulations.
The Need for This Book
The initial impetus for this book arose out of the realization that, despite the rapidly growing interest in motivational processes from within cognitive psychology and neuroscience, there was still an important gap in the literature. In particular, there was a strong need not only to synthesize the accumulating findings, but also to properly integrate and situate them with the broader perspectives taken on motivation-cognition interactions from the animal learning tradition; social, affective, and personality psychology; and the aging and development literatures. My colleagues and I have undertaken a number of synthetic and theoretical reviews of the literature on motivation and cognitive control, which attempt to incorporate such perspectives (Botvinick & Braver, 2015; Chiew & Braver, 2011; Krug & Braver, 2014; Locke & Braver, 2010), but the need is broader than what can be accomplished by single papers and a small set of authors. A first step towards filling this gap was MOMCAI (Mechanisms of Motivation-Cognition-Aging Interactions), a small conference organized in May 2013 in Washington, DC, and sponsored by the National Institute on Aging (NIA). As the name indicates, MOMCAI brought together researchers and trainees working on motivation-cognition interactions from the various disciplines described earlier. The meeting was a great success and highly productive, leading to the publication of a special issue of the journal Cognitive, Affective, and Behavioral Neuroscience (CABN; Volume 14, Issue 2; June 2014). This special issue included articles from a number of conference attendees, and also featured a group-authored introductory article (Braver et al., 2014), which laid out a number of key themes, definitions, conceptual distinctions, and unresolved issues within this domain.
Even though I believe that these outcome products of the MOMCAI conference make a strong contribution, they also do not completely fill the gap in the literature. Specifically, there were still a number of important researchers working in the area who either were not in attendance at MOMCAI or did not contribute to the special issue, or both. Likewise, the journal article format and audience provide constraints that limited the work that could be included, even for special issue contributors. In contrast, the book and chapter format provides complimentary advantages, such as the opportunity to include more synthetic reviews, and to highlight important conceptual frameworks and theoretical models. Thus, I am extremely gratified that this current volume does complete the effort begun by our initial reviews, by the MOMCAI conference, and by the CABN special issue, in providing a relatively comprehensive snapshot of the current state of the field of motivation-cognition interaction studies. Moreover, the contributions by chapter authors significantly advance this effort, by providing a range of different disciplinary perspectives, by authoritatively summarizing the available literature in these perspectives, and in some cases by putting forward new and innovative theoretical models and syntheses. Finally, the book does an excellent job of highlighting not only progress in the study of motivation-cognition interactions but also some of the important challenges that remain for theoretical and disciplinary integration.
Key Themes and Organizational Structure
The scope of this book is relatively broad, with a number of interweaving themes, questions, and issues that are addressed from various perspectives across the chapters. However, it can also be seen as loosely structured into three distinct sections. In the first section, the focus is on reward incentives and how they influence a range of cognitive processes and neural systems. One of the central themes is how to conceptually organize these reward influences. In particular, chapters in this section focus on the relationship between reward motivation and other constructs, such as attention and reinforcement learning, and on the different potential mechanisms by which reward motivation can modulate cognitive control. In the second section, the focus expands to examine more closely the relationship between motivation and affect, and on aversive signals as well as positive ones. One of the central themes is that cognitive control is effortful and may also be associated with aversive feedback signals, such as performance errors, response conflict, and fatigue. Thus, the chapters in this section explore the ideaāfrom various perspectivesāthat there may be an intrinsic bias to avoid or conserve control that can be overcome by positive motivational value associated with the cognitive task and outcome. In the last section, the focus further broadens to examine how motivational influences on cognition might change across the life span. There are two central themes of this section. The...