(1991)

who found a larger effect for emotion than gender. Luh et al. paired a neutral face half with a happy face half, as in the current study. Still, the quality of the pictures could have affected the size of the left visual hemispace bias. The latter might be especially true for the gender test, which is heavily depended on the number and quality of the feminine characteristics in the photos. As left-held infants have a better view of their mother’s most expressive left face half (Hendriks et al., in press), this finding suggests that a reduced left-bias is caused by poorer exposure to faces during infancy. Whether this would be the result of face perception per se could be studied in future research by also assessing perception for stimuli that have been proven not to be sensitive for the left visual hemispace bias, such as assessing object form (Luh H 89 et al., 1991), books and bags (Harris et al., 2010) or by presenting stimuli that normally result in a right visual Tanespimycin hemispace bias, such as speech reading (Burt & Perrett, 1997). A reduced leftward bias has also been found in left-handed individuals (Harris et al., 2001, Levy et al., 1983 and Rueckert,

2005). One might argue, therefore, that the reduced left-bias in right-held participants (with left-handed mothers) was caused, not by their suboptimal view of their mother’s face, but by their own atypical pattern of lateralisation resulting from their genetic predisposition. Although this possibility DNA ligase cannot be ruled out, there are two arguments against it. First, the right-held participants were all strongly right-handed (on the handedness test, they were right-handed on 10 out of 10 items), which makes atypical lateralisation due to genetic factors perhaps not so likely for other functions.

Second, even truly left-handed individuals (which the present participants, being strongly right-handed, were clearly not) usually show the typical right-hemisphere lateralisation for faces, and, correspondingly, mostly prefer to cradle an infant on the left-arm, similar to the right-handed population (e.g. Salk, 1960: 78% of left-handers cradle on the left-arm). In other words, the present results seem difficult to explain with a genetic predisposition account. There is another potential problem for the interpretation that the present results are caused by impoverished face exposure. That is, if one’s holding bias is related to one’s own bias on face chimera tests, as has been indicated by some studies (e.g. Bourne and Todd, 2004 and Vauclair and Donnot, 2005, but see Donnot & Vauclair, 2007), a mother with a rightward bias might prefer to hold an infant on her right-arm because that would agree with her own lateralisation for the perception of faces and emotions. Consequently, the mother’s face half most visible to the infant might be her most expressive face half, even in right-held infants, making it less likely that the present results are attributable to differences in face exposure.