Abstract: Since a functional independence was set in the functional model for face recognition (Bruce & Young, 1986), many studies have been conducted to explore the relations between facial expression and facial identity recognition. Some studies supported the opinion of independent processing between facial expression and identity recognition, but the others provided evidences for the opinion of interdependent processing between them. One of the most important pieces of evidence that supported interdependent processing came from the behavior studies, which revealed the interconnection between recognition of facial identity and facial expression. However, in these studies the influence of facial expression on facial identity recognition is hard to be found compared to the influence of identity on expression recognition, especially in unfamiliar face recognition. Recently, Ganel et al. demonstrated the influence of expression on identity by reducing the discriminability of facial identity, and then explained the failure of finding this kind of influence in the past studies. The present study was conducted to challenge the argument that the discriminability of facial expression is lower than that of facial identity. The past studies all neglected the intensity variation information of facial expression although it should be very important to facial expression recognition. We addressed this issue in the present study. We hypothesize that when facial expressions were displayed with intensity varying, the influence of facial expression on identity recognition might disappear.
Method
To test the hypothesis, two experiments were conducted. In experiment 1, participants were instructed to judge identity or expression of randomly displayed morphing pictures, and the influence between facial identity and expression recognition was examined with reaction time and boundaries as dependent variation. In experiment 2, participants were asked to make judgment along morphing continua, and the influence between facial identity and expression recognition was measured with boundaries as indicator.
Results
The results of experiment 1 indicated that identity influenced expression recognition but expression did not affect identity judgment, which replicated the results of the previous studies. In experiment 2, the influence of identity on expression recognition disappeared. These results were consistent with our predictions: when facial expressions were displayed with intensity varying, the influence of facial expression on identity recognition would disappear.
Conclusions
Facial expression intensity cue enhanced the discriminability of facial expression but not that of identity, and then the influence of identity on expression recognition disappeared. Based on these results, we could draw a conclusion that the low discriminability of facial expression could cause the influence of identity on expression recognition. Therefore, the influence of identity on expression recognition must not necessarily support the opinion of independent processing. In addition, the results of present study also provided a possible way to dissociate information of facial identity and expression

Key words: facial identity recognition, facial expression recognition, intensity variation of facial expression, discriminabiltity, distributed representation