Face recognition : laterality effects and the extraction of configural information

Bertelson, Van Haelen, & Morais (1979) have suggested that a right hemisphere advantage (RHA) will only emerge under conditions where the extraction of physiognomic information is demanded, for example, when the face to be recognised differs in its angle of view, expression, age or state of health. This process constitutes face recognition proper and requires that invariant properties of the face be extracted. They argued that decisions about identical pictures seen at initial presentation and later test may .tell us as much about recognition memory for pictures as about memory for faces. This process represents what Bertelson et al. (1979) call 'stimulus recognition'. A left visual field advantage (LVFA) was reported in their study for the 'facial identity' condition only, that is, when the faces to be compared differed in their viewing angle. No visual field aymmetry emerged for the 'physical identity' group. This present study was designed to investigate the results of Bertelson et al. (1979) employing a within-groups design. Thirty-two subjects were shown pairs of successively presented faces, with the second face shown randomly to either the LVF or RVF, and required to indicate whether this face belonged to the same or different person. All the subjects were exposed to faces in both the varied presentation ('facial identity') and identical presentation ('physical identity') conditions. In addition, two types of faces were used in the present study. Orientation based (OB) faces focused on the viewing angle of faces whereas Component based (CB) faces emphasised feature comparison. The main hypothesis that a LVFA would emerge for OB faces when they varied in their presentation was not supported. The prediction of a RVFA for CB faces when same responses were required was also nonsignificant. These results are discussed in relation to methological problems and difficulties in manipulating faces