Why do some people commit serious violent crimes? Why do they use brutal force to injure and kill other people? One idea is that violent aggression results from malfunctioning of a violence inhibition system. In normal, non-violent people perception of distress in others triggers arousal and freezing, inhibiting violent impulses. The most effective cues for activating this violence inhibition system are fearful and sad facial expressions.
According to this model, aggressive behavior and violent crimes may result from a failure to perceive distress cues such as fearful faces. If true, this would open a whole range of novel treatment options. Imagine a future in which violent crime will be eradicated by simply training people on perceiving fear in others' faces?
Indeed, violent offenders have difficulty categorizing fearful facial expressions as displaying fear. However, this could reflect a problem with linking a percept (of a fearful face) with the correct category label (fear) rather than a problem with the basic, "raw" perception of fear per se. In other words, does the deficit emerge at an early visual processing stage or at a later conceptual processing stage? Answering this question is important for designing effective training regimes.
In collaboration with Aiste Jusyte and Michael Schönenberg from the University of Tübingen, we tested 45 violent offenders (in jail for violent crimes) and 46 control participants in two tasks. One task measured implicit perceptual biases for facial expressions, the other task measured explicit categorization of facial expression. In both the perceptual and the categorization task, participants saw an array of eight face photographs on the screen. Seven of the eight faces were the same and had neutral expressions (the distracters), and participants responded to the odd one out (the target).
In the perceptual task, the target was a different face than the distracters. Participants indicated the gender of the target. In some trials, the target also had an emotional expression, such as fear, making it stand out among the seven distracters with neutral expressions. Participants were faster to indicate the gender of the target when it had a fearful expression. Thus, although the expression of the target was irrelevant to the task, fear was grabbing attention.
If violent offenders had a basic perceptual deficit for fear, they should show less of a fear advantage. For example, if they had difficulty distinguishing fear from neutral, a fearful face would not be grabbing attention among the neutral distracters. However, we found a full-strength fear advantage in violent offenders in this perceptual task. There seemed to be no problem with implicit perception of fear at this early processing stage.
We next tested for a fear deficit in the categorization task. Here, all eight faces in the array showed the same person. The target was the one face with an emotional expression, and participants indicated whether this expression was happy, angry, or fearful. Violent offenders were slower for fearful (compared to happy) expressions than control participants. Thus, violent offenders seem to have difficulty with the explicit labeling rather than with the basic perception of fearful expressions.
What are the implications of these findings? It seems unlikely that violent behavior reflects a failure to automatically perceive distress cues in others. Rather, violent offenders have problems categorizing and interpreting these distress signals. Instead of training perceptual discriminability of fear, therapeutic interventions should therefore support aggressive individuals to link their percepts to the correct linguistic labels and mental concepts.
Link to the manuscript PDF
Blogging about my research.