Science Briefs

Emotion and Cognition: The Case of Automatic Vigilance

When confronted with a threatening stimulus, people typically devote increased attentional resources to that stimulus, raising the accessibility of evaluatively-similar information in memory, and biasing subsequent perceptions and judgments toward a threatening evaluation

By Randy J. Larsen, PhD

In St. Louis we are celebrating the bicentennial of the Lewis and Clark expedition. Captain Lewis made an entry in his journal that nicely illustrates the interaction between emotion and cognition. He describes how he was traveling alone one day, well ahead of his corps, to determine the best route. Suddenly he was surprised by an aggressive grizzly bear charging at him from out of the bush. Lewis narrowly escapes by jumping into a river. After the bear withdraws, Lewis makes his way back to his troops, a distance of about 12 miles. Along the way, he notices a variety of other animals, most of which he perceives as threatening and several of which he shoots preemptively. In his journal he describes feeling surrounded by danger: "It now seemed to me that all the beasts of the neighborhood had made a league to destroy me." (Bakeless, 2002, p. 187). The editor of the Lewis and Clark journals, in a footnote to this passage, notes that the animals Lewis encountered along the way were not typically considered aggressive or dangerous, and opines that Lewis was probably nervous after his frightening encounter with the grizzly.

Automatic Vigilance Following Threatening Information
The example from Lewis' journal illustrates the phenomenon of automatic vigilance, where emotional cues in the environment bias subsequent information processing. More precisely, the detection of threatening information can interrupt ongoing cognitive activity in ways that tune subsequent perception, attention, judgment, and even memory towards threat-related outcomes. One experimental analog of automatic vigilance is affective priming (Klauer, 2003), particularly priming with threatening stimuli. Here a threatening image or word is briefly presented (the prime) and quickly followed by another stimulus (the target) to which the subject responds (e.g., makes a lexical decision, categorizes as a good or bad object, etc.). Automatic vigilance occurs when a negatively valenced target stimulus (e.g., an image of a COCKROACH) is categorized faster and/or more accurately when it is preceded by a threatening prime stimulus (e.g., the word DISEASE) than a hedonically neutral prime stimulus (e.g., the word DISHPAN) (Hermans, DeHouwer, & Eelen, 2001).

Researchers suggest that the presentation of an evaluative or threatening prime may automatically activate biased perceptions of emotionally-congruent targets (Fazio, Jackson, Dunton & Williams, 1995). The explanation for this effect is that, when confronted with a threatening stimulus, people typically devote increased attentional resources to that stimulus, raising the accessibility of evaluatively-similar information in memory, and biasing subsequent perceptions and judgments toward a threatening evaluation (Klauer, 2003; Wentura & Rothermund, 2003).

Other Examples of Automatic Vigilance Effects
Some researchers have identified the emotional Stroop task as an example of automatic vigilance (e.g., Pratto & John, 1991; Wentura, Rothermund, & Bak, 2000). In this task, subjects are asked to quickly name the colors of various words, some of which are threatening (e.g., DISEASE) and others are neutral (e.g., DISHPAN). In general, people are slower to name the colors of threatening words than the neutral words. However, a crucial problem with many of these studies is that the two word lists - threat and control words - often differ with respect to critical linguistic parameters known to contribute to reaction time differences in word recognition. For example, Larsen, Mercer, and Balota (2004a) showed that, across 34 emotion Stroop studies, the threatening words used were more infrequent, of greater length, or had a larger orthographic neighborhood than the control words. All of these purely linguistic features contribute to slower recognition of the threatening words, casting doubt on the validity of the emotional Stroop effect being due to automatic vigilance to the threat value of the word.

In a recent paper, Algom, Chajut and Lev (2004) reasoned that automatic vigilance should not be limited to color naming of words but should apply to any cognitive activity. In a series of very carefully done experiments they demonstrated that both color naming and word reading were slower for threatening than control words. Larsen, Mercer, and Balota (2004b) recently analyzed lexical decision time and word reading time for a list of over 1,000 words that had been previously normed for valence (Bradley & Lang, 1996). After controlling for important linguistic parameters (e.g., frequency, length, orthographic neighborhood), Larsen et al. (2004b) found that word negativity was still a significant predictor of longer reaction times, both for lexical decisions and word reading time. Cothran, Larsen, Zelenski, Prizmic, & Chein (2004) demonstrated automatic vigilance effects in the recognition of facial displays of emotion.

There is converging evidence from a number of literatures on the existence of automatic vigilance effects. The general interpretation is that a dedicated preattentive system operates in an automatic fashion to screen the perceptual stream for threatening information (Ohman, 1993). When such information is detected, ongoing cognitive activity is interrupted (accounting for the generic slowing) and reprioritized to be biased for future threatening information (accounting for the priming effects). Such a system would have obvious evolutionary advantages, in that humans who lacked such a system would be less likely to become ancestors.

Utility of Automatic Vigilance for Studying other Phenomena
As an aspect of psychological functioning, the automatic vigilance effect is interesting in its own right. However, it also has utility for studying other psychological phenomena. In the remainder of this article I describe two areas where the concept of automatic vigilance may contribute to our understanding of other phenomena.

Understanding stereotype activation. In a series of important experiments, Payne (2001; Payne, Lambert, & Jacoby, 2002) tried to clarify what factors contributed to the killing of Amidou Diallo, an unarmed Black immigrant from West Africa who was shot 19 times by several White New York police officers one night as he was retrieving his wallet from his pocket. In Payne's studies he used a priming procedure, where participants were primed for 200 ms with a photo of either a Black face or a White face, then immediately shown a drawing of a handgun or a hand tool for 100 ms. Participants then have 400 ms to decide if the second object they saw was a hand tool or a handgun. Of course they make a lot of errors due to the speeded nature of the response. However, among predominantly White participants the errors are not random. Instead, participants are more likely to confuse the hand tool for a handgun following the Black face prime than the White face prime. The dominant interpretation of this finding is that the prime activates stereotypes beliefs about what is associated with being Black or White (Judd, Blair, & Chapleau, 2004). Consequently, this stereotype makes it more likely to make a "gun" response following the Black prime than the White prime.

Recently, we (Larsen, Chan, & Lambert, 2004) reasoned that automatic vigilance may have played a role in both Payne's results as well as the shooting of Amidou Diallo. If Blacks (or other outgroup members) are threatening to majority participants, then priming with a Black face may activate automatic vigilance for future threat, making the gun response more likely. In an extension of Payne's gun/tool paradigm, we (Larsen, Chan, & Lambert, 2004) replaced the Black and White primes by photos of threatening animals (snakes, spiders) and non-threatening animals (bunnies, kittens). We found the same pattern of gun/tool bias; after being primed with a threatening animal, participants were more likely to confuse hand tools as handguns compared to being primed with the non-threatening animal.

In a second experiment we put the Black and White faces back in as primes, and moved the good and bad animals to the target position. Participants were given 400 ms to categorize the animals as either good or bad. We again found a bias consistent with an automatic vigilance effect; participants were more likely to confuse a good animal as a bad one following a Black prime than a White prime. Obviously, the animals have nothing to do with stereotype associations to Blacks or Whites. However, they do have threat value and so our subjects' biased processing following Black facial primes is consistent with an automatic vigilance effect. Besides clarifying the underlying mechanism for the gun/tool bias, the results have implications about interventions to counteract such biases. Efforts to change stereotyped beliefs about outgroup members would be a lot different than efforts to change prejudiced emotional reactions toward them.

Understanding why bad is stronger than good. From a number of quite different literatures there is converging evidence that stimuli of equal hedonic weight, but opposite in hedonic sign, will evoke non-equivalent affective reactions (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001)). For example, people are more distressed by the loss of $50 than they are made happy by finding $50. A few years ago I applied a psychophysics framework to this issue (Larsen, 2002). After all, emotion is a lot like perception, where some aspect of the outer world (an emotional event, a sensory stimulus) is transformed to an inner representation (an affect, a sensation). Applying this framework to several data sets, I reviewed evidence that unpleasant events or stimuli, compared to equivalently pleasant events or stimuli, evoke larger emotional responses, longer duration responses, and have a broader impact on the cognitive system. Moreover, I estimated the impact of negative stimuli as being about three times that of the impact of positive stimuli.

Why should bad be stronger than good? It seems likely that one function of the automatic vigilance system is to act as a signal gain mechanism for threatening information. The automatic vigilance system functions to amplify threatening information by directing cognitive resources, such as perception and attention, toward such information. There is no specialized counterpart that acts in such an automatic and preattentive fashion for positive stimuli. In fact, in studies of cognitive interference from affective meaning using the affective Simon task, we found that the interference effect size for negative stimuli was approximately three times as large as the interference effect size for positive stimuli (Larsen & Yarkoni, 2004). The automatic vigilance system may be an explanation for the ubiquitous finding that bad is stronger than good.

Cognition and emotion interact in various ways, and one of the more interesting and increasingly documented ways is the automatic vigilance effect. This phenomenon highlights differences between automatic and controlled psychological processes, in that the effect is purely automatic. Much like a reflex, it occurs very fast, happens without our awareness or effort, and runs to completion without conscious monitoring. And yet the effects may be far-reaching, as when automatic vigilance impacts on cognitive resources such as attention and memory. And the effects may be especially far-reaching when the elicitors of the vigilance, or the objects of its effect, are other people.


Preparation of this article, and some of the research reported, was supported in part by grant RO1-MH63732 from the National Institute of Mental Health.


Algom, D., Chajut, E., Lev, S. (2004). A rational look at the emotional Stroop phenomenon: A generic slowdown, not a Stroop effect. Journal of Experimental Psychology: General, 133, 323-338.

Bakeless, J. (2004). The journals of Lewis and Clark. New York: Signet.

Baumeister, R. F., Bratslavsky, E., Finkenauer, C., and Vohs, K. D. (2001). Bad is stronger than good. Review of General Psychology, 5, 323-370.

Bradley, M. M., & Lang, P. J. (1996). Picture media and emotion: Effects of a sustained affective context. Psychophysiology, 33, 662-670.

Cothran, D. L., Larsen, R. J., Zelenski, J., Prizmic, Z., & Chien, B. (2004). Do Emotion Words Interfere with Processing Emotion Faces? Stroop-Like Interference versus Automatic Vigilance for Negative Information. Manuscript under review.

Fazio, R. H., Jackson, J. R., Dunton, B. C., & Williams, C. J. (1995). Variability in automatic activation as an unobtrusive measure of racial attitudes: A bona fide pipeline? Jounral of Personality and Social Psychology, 69, 1013-1027.

Hermans, D., De Houwer, J., & Eelen, P. (2001). A time course analysis of affective priming task. Cognition and Emotion, 15, 143-165.

Judd, C. M., Blair, I. V., & Chapleau, K. M. (2004). Automatic stereotypes vs. automatic prejudice: Sorting out the possibilities in the Payne (2001) weapon paradigm. Journal of Experimental Social Psyhcology, 40, 75-81.

Klauer, K. C. (2003). Affective priming: Findings and theories. In J. Musch and K.C. Klauer (Eds.), The psychology of evaluation: Affective processes in cognition and emotion (pp. 7-50). Mahwah, NJ: Erlbaum.

Larsen, R. J. (2002). Differential contributions of positive and negative affect to subjective well-being. In J. A. Da Silva, E. H. Matsushima, and N. P. Riberio-Filho (Eds). Annual meeting of the International Society for Psychophysics (vol. 18, pp. 186-190). Rio de Janeiro, Brazil: Editora Legis Summa Ltda.

Larsen, R. J., Chan, P. Y., & Lambert, A. (2004). Perceptual consequences of threat and prejudice: Misperceiving weapons and other dangerous objects. Manuscript under review.

Larsen, R. J., Mercer, K., & Balota, D. (2004a). Lexical characteristics of words used in emotion Stroop studies. Manuscript under review.

Larsen, R. J., Mercer, K., & Balota, D. (2004b). Lexical Characteristics and Word Recognition Parameters for Emotion Words: Effects of Word Negativity. Manuscript under review.

Larsen, R. J. & Yarkoni, T. (2004). Negative stimuli cause more interference than positive stimuli in the affective Simon task. Manuscript under review.

Ohman, A. (1993). Fear and anxiety as emotional phenomena: Clinical phenomenology, evolutionary perspectives, and information-processing mechanisms. In M. Lewis & J. M. Haviland (Eds.), Handbook of emotions (pp. 511-536). New York: Guilford Press.

Payne, B. K. (2001). Prejudice and perception: The role of automatic and controlled processes in misperceiving a weapon. Journal of Personality and Social Psychology, 81, 181-192.

Payne, B. K., Lambert, A. J., & Jacoby, L. L. (2002). Best laid plans: Effects of goals on accessibility bias and cognitive control in race-based misperceptions of weapons. Journal of Experimental Social Psychology, 38, 384-396.

Pratto, F. & John, O. (1991). Automatic vigilance: The attention-grabbing power of negative social information. Journal of Personality and Social Psychology, 61, 380-391.

Wentura, D. & Rothermund, K. (2003). The "meddling-in" of Affective Information: A general model of automatic evaluation. In J. Musch and K.C. Klauer (Eds.), The psychology of evaluation: Affective processes in cognition and emotion (pp. 51-86). Mahwah, NJ: Erlbaum.

Wentura, D., Rothermund, K. & Bak, P. (2000). Automatic vigilance: The attention-grabbing power of approach- and avoidance-related information. Journal of Personality and Social Psychology, 78, 1024-1037.

About the Author
Randy J. Larsen earned his PhD in Personality Psychology from the University of Illinois-Champaign in 1984. He has served on the faculty at Purdue University (1984-1989), the University of Michigan (1989-1998), and Washington University in St. Louis (since 1998), where he is currently the William R. Stuckenberg Professor of Human Values and Chairman of the Psychology Department. His research interests focus on personality and emotion, with particular interests in emotional reactivity and mood regulation, process models of daily mood, and cognitive consequences of affective states. He has published over 80 scientific articles and book chapters and has co-authored (with David Buss) a text in Personality Psychology. Professor Larsen was awarded the 1991 APA Distinguished Scientific Award for Early Career Contribution to Personality Psychology, received a Research Scientist Development Award from the National Institute of Mental Health, and is a past president of the Midwestern Psychological Association.