Results and Discussion

Observers did not significantly differ on the related trials compared with the unrelated trials: i(17) = 0.41, ns. Error rates on the two sets of trials also did not differ: related error rate = 2.8 percent; unrelated error rate = 1.9 percent; i(17) = .78, ns. The lack of positive priming suggests two possibilities. First, the visual system does not use the same neural mechanisms to automatically code upright and upside-down

Figure 11.9

Upside-down distractor faces were presented along with upright targets.

Figure 11.9

Upside-down distractor faces were presented along with upright targets.

faces. These findings agree with the physiological recordings of face-selective cells in sheep (Ken-drick and Baldwin 1987). Second, time (Cochran et al. 1983; Perrett et al. 1988; Rock 1973; Valentine and Bruce 1988) or attentive processing is necessary to transform a representation of an upside-down face into its upright counterpart. In either case automatic face processing of upside-down faces without attention does not result in representations that match upright versions of the same face.

General Discussion

In sum, a negative priming procedure showed that response latencies were longer when one or both target faces had appeared as distractors on the immediately preceding trial. Thus, never-before-seen faces are represented and require inhibition. Response latencies were shorter when target faces had appeared as distractors with high-frequency noise or were contrast-inverted. Thus, underlying representations may in fact be facilitated. Response latencies were unaltered when target faces had appeared as upside-down distractors; not all distractor representations afford response priming (negative or positive). It is proposed that the visual system indeed represents unfamiliar faces without focused attention, but blocks them only if they vie with targets for the control of action (Khurana 2000).

At the simplest level of analysis, the above experiments show that neither negative nor positive priming of human faces requires a network of preexisting face representations. The present findings suggest that the visual system creates representations of human faces "on the fly" in the absence of focused attention. It is possible that all the perceptual mechanisms which process objects present in a negative priming display are activated and primed, regardless of whether they are used to process distractors or targets. If the distractors compete with the targets in the selection of a response, their representations must be suppressed. Whether a distractor competes with targets depends upon the nature of the representations that task performance relies upon (Khurana 2000; Tipper et al. 1994). In the absence of inhibition, if the pathways that rendered the internal representations are accessed again, they will result in faster or more accurate task performance.

Finally, the priming results reported here are largely in accord with response profiles of "face" cells (Desimone 1991; Kendrick and Baldwin 1987; Perrett et al. 1982; Perrett et al. 1987; Rolls et al. 1993; Rolls et al. 1989; Rolls et al. 1985; Rolls and Baylis 1986). This form of face priming could specify the structural descriptions achieved by gnostic cells (Konorski 1967) selectively tuned to faces in humans. However, the present priming results make it unlikely that these cells are the neural correlates of the explicit representations of faces (Crick and Koch 1998). If anything, the present findings suggest that the cells in the superior temporal sulcus are involved in the implicit processing of faces. Explicit face representation may require the involvement of the recently reported face cells in the inferior prefrontal cortex that receive direct input from the superior temporal sulcus (Scalaidhe et al. 1997). The activity of these cells either alone or in conjunction with face-sensitive cells in the superior temporal sulcus might render representations that correspond to the experience of seeing someone.

Was this article helpful?

0 0

Post a comment