The recognition of threatening facial stimuli.

Abstract

Two studies examined the information that defines a threatening facial display. The first study identified those facial characteristics that distinguish between representations of threatening and nonthreatening facial displays. Masks that presented either threatening or nonthreatening facial displays were obtained from a number of non-Western cultures and scored for the presence of those facial features that discriminated between such displays in the drawings of two American samples. Threatening masks contained a significantly higher number of these characteristics across all cultures examined. The second study determined whether the information provided by the facial display might be more primary nonrepresentational visual patterns than facial features with obvious denotative meaning (e.g., diagonal lines rather than downturned eyebrows). The subjective response to sets of diagonal, angular, and curvilinear visual stimuli revealed that the nonrepresentational features of angularity and diagonality in the visual stimulus appeared to have the ability to evoke the subjective responses that convey the meaning of threat. A century-long debate was stimulated by Darwin's (1872) intriguing speculation that emotional expression in human beings, as in other animal species, might rest upon a common neuromuscular foundation. One line of reasoning (e.g., Birdwhistell, 1963; Klineberg, 1938; LaBarre, 1947), primarily based on anthropological evidence covering a wide array of human activity, has maintained that emotion is expressed through highly variable behaviors, cross-culturally, and so is best understood as the result of a culture-specific cognitive labeling process. Following more closely Darwin's interest in the stimulus configurations generated by facial displays, a second line of reasoning (Ekman, 1971,1982;Izard, 1971, 1977; among others) has provided impressive support for the position that speciescharacteristic features underlie facial displays of emotion. Typically, in this work, photographs of individuals experiencing (or actors simulating) a set of emotions are shown to members of different cultural groups who are then asked to identify the emotion displayed. These studies have shown that a wide range of emotional representations can be discriminated correctly (Ekman, 1971; Izard, 1971), with the results from less Western cultures parallel to those obtained from more Western cultures. More recently, such research has been attentive to the specific features that compose a type of facial display. Coding systems of emotionally related facial displays, such as Ekman and Friesen's (1978) Facial Action Coding System, are more than conve-We gratefully acknowledge the assistance of William Crano, Lester Hyman, Lawrence Messe, and Eileen Thompson in the design and execution of this study. Valuable comments by Thomas Carr, Hiram Fitzgerald, Lawrence Messe, and Rose Zacks on an earlier draft of this paper are much appreciated.

Differential attentional guidance by unattended faces expressing positive and negative emotion

“…This is consistent with a number of findings in the literature. For example, schematic faces appear to communicate emotional meaning effectively (Aronoff, Barclay, & Stevenson, 1988;Cuceloglu, 1970;McKelvie, 1973). Schematic faces also show disruptions in perception when inverted (Farah et al, 1998;Moscovitch et al, 1997;Yin, 1969) similar to those found with photographed faces.…”

Anger superiority effect for change detection and change blindness

“…Furthermore, the V-shaped form of the eyebrows is suspected to potentially attract attention so that the threatening expressions themselves are not necessarily configurally perceived, and the perception of threatening faces may be based on perception of single features (Aronoff, Barclay, & Stevenson, 1988). …”

Preattentive face processing: What do visual search experiments with schematic faces tell us?

“…Either way, the results of Eastwood et al (2001) It can be argued that the research by Lundquist and colleagues (Lundquist, Esteves, andÖhman, 1999, 2004;Öhman et al 2001) suggests an a priori reason for regarding one face pair as most representative. These authors have argued for the importance of eyebrows in the attribution of anger or threat (see also Aronoff, Barclay, & Stevenson, 1988), implying that, when in doubt, the stimuli from Experiment 2 would be preferred. However, Fox et al (2000) have defended their use of brow-less faces by proposing that an evolved mechanism for threat detection should be biased towards false alarms (in contrast to misses), and should therefore respond to ambiguous stimuli.…”