Defensive Motivation and the Error-Related Negativity

The error-related negativity (ERN) is a negative deflection in the event-related potential that is maximal approximately 50 ms after the commission of an error. The ERN is generated in the anterior cingulate cortex, a region of the medial prefrontal cortex implicated in both cognitive and emotional processing. Affective and motivational variables influence the magnitude of the ERN, which suggests that the ERN may relate to emotional or motivational aspects of error detection. In the present study, we evaluated the possibility that errors prime defensive motivational responses, and that the ERN may predict the magnitude of defensive reactivity after errors. We found that (a) the defensive startle response was larger following errors than following correct responses, and (b) the magnitude of the ERN predicted the degree of startle potentiation following errors. Thus, response errors prime defensive motivation—and the ERN predicts individual differences in the aversive response to errors. Research on error processing has grown over the past 15 years, in part because of the discovery of the error-related negativity (ERN; Falkenstein, Hohnsbein, Hoormann, & Blanke, 1991; Gehring, Coles, Meyer, & Donchin, 1995). The ERN is a sharp negative deflection in the event-related potential (ERP) that peaks approximately 50 ms after an unintended response (Falkenstein, Hoormann, Christ, & Hohnsbein, 2000). The ERN is maximal over fronto-central recording sites, and evidence from source localization (Holroyd, Dien, & Coles, 1998), intracerebral recording (Brazdil, 2005), and magnetoencephalography (Miltner et al., 2003) indicates that the ERN is generated in the anterior cingulate cortex (ACC). The ACC is richly interconnected with both limbic and prefrontal areas of the brain, and is ideally situated to respond to, and integrate, cognitive and affective information (Bush, Luu, & Posner, 2000). Although the ERN is known to occur immediately following the commission of an error, there remains considerable debate as to its functional significance and the specific process it indexes. Computational models suggest that the ERN reflects either the coactivation of erroneous and error-correcting responses (conflict theory; Yeung, Cohen, & Botvinick, 2004) or the evaluation, based on expectations developed during learning history, that events are worse than anticipated (reinforcementlearning theory; Holroyd & Coles, 2002). Both the conflict and reinforcement-learning theories, then, highlight specific cognitive functions subserved by the ACC in their conceptualization of the ERN. Given the involvement of the ACC in pain, depressed mood, and other distressed states, Luu, Collins, and Tucker (2000) proposed that the ERN may signify affective processing in response to errors. In fact, there is ample evidence that the ERN can be modulated by motivational and affective variables (cf. Hajcak, Moser, Yeung, & Simons, 2005). In between-groups designs, individuals with obsessive-compulsive disorder (Gehring, Himle, & Nisenson, 2000; Hajcak, Franklin, Foa, & Simons, in press; Hajcak & Simons, 2002), depression (Chiu & Deldin, 2007), and more generalized anxiety (Hajcak, McDonald, & Simons, 2003a, 2004) have exhibited increased ERNs compared with control participants—a finding consistent with the notion that affective psychopathology is associated with hyperresponsivity to negative information. Other studies have suggested a possible relation between the ERN and individual differences in punishment sensitivity (Boksem, Tops, Wester, Meijman, & Lorist, 2006; Dikman & Allen, 2000). It is important to note that most of these results cannot be accounted for simply by differences in error rates and reaction time, and they suggest that the ERN varies as a function of motivational states and traits. In a broader context, errors are maladaptive responses that may place an organism in danger and threaten its safety. It stands to reason that error detection should prompt defensive motivation—and rapidly activate those reflexes and dispositions toward action that protect the organism (Gray, 1994). To date, however, there has been no demonstration of a direct link Address correspondence to Greg Hajcak, Department of Psychology, Stony Brook University, Stony Brook, NY 11794-2500, e-mail: greg. [email protected]. PSYCHOLOGICAL SCIENCE Volume 19—Number 2 103 Copyright r 2008 Association for Psychological Science between errors and defensive motivation. Although errors prompt heart rate deceleration and skin conductance responses (Hajcak, McDonald, & Simons, 2003b, 2004), this pattern of autonomic reactivity is consistent with either an orienting or a defensive response to errors (Cook & Turpin, 1997). The human startle response, in contrast, is a well-validated and specific measure of defensive activation (Lang, 1993; Lang, Davis, & Öhman, 2000). The startle response is a reflex in which the body contracts into a defensive posture and is typically measured in humans by recording blink magnitude from the obicularis oculi in response to a sudden and loud acoustic probe. The magnitude of the startle response can be modulated by aversive stimuli and settings: Startle is potentiated when participants view threatening stimuli (Bradley, Codispoti, & Lang, 2006; Bradley, Moulder, & Lang, 2005; Lang et al., 2000) and when they anticipate an electric shock (Curtin, Patrick, Lang, Cacioppo, & Birbaume, 2001; Grillon, Ameli, Merikangas, Woods, & Davis, 1993). These findings are consistent with the notion that aversive stimuli prime defensive reflexes. In the present study, we sought to determine whether defensive responding would similarly be increased following errors, compared with correct responses. To evaluate this possibility, we measured the human startle response 300 ms following errors and correct responses. We hypothesized that startle responses would be larger after errors than after correct responses. To evaluate the role of startle predictability, we included both predictable and unpredictable startle probes on correct trials. Additionally, we simultaneously measured error-related brain activity to determine whether individuals with larger ERNs would demonstrate greater startle potentiation following errors. If the ERN reflects a motivationally relevant correlate of error processing, larger ERNs should predict greater startle potentiation following errors.