Developmental Science 11:5 (2008), pp 761–770 DOI: 10.1111/j.1467-7687.2008.00726.x © 2008 The Authors. Journal compilation © 2008 Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA. Blackwell Publishing Ltd PAPER To see or not to see: infants prefer to follow the gaze of a reliable looker Virginia Chow, Diane Poulin-Dubois and Jessica Lewis Department of Psychology, Concordia University, Canada Abstract In two experiments, we examined whether 14-month-olds understand the subjective nature of gaze. In the first experiment, infants first observed an experimenter express happiness as she looked inside a container that either contained a toy (reliable looker condition) or was empty (unreliable looker condition). Then, infants had to follow the same experimenter’s gaze to a target object located either behind or in front of a barrier. Infants in the reliable looker condition followed the experimenter’s gaze behind the barrier more often than infants in the unreliable looker condition, whereas both groups followed the experimenter’s gaze to the target object located in front of the barrier equally often. In the second experiment, infants did not generalize their knowledge about the unreliability of a looker to a second ‘naïve’ looker. These findings suggest that 14-month-old infants adapt their gaze following as a function of their past experience with the looker. Introduction Gaze following occurs when one person focuses his/her visual attention to where another person is looking. The ability to follow the gaze direction of others is considered a critical component in social interactions (Argyle & Cook, 1976; Kleinke, 1986; Langton, Watt & Bruce, 2000) and is posited to be a developmental precursor to children’s later theory of mind abilities (e.g. Corkum & Moore, 1998). The capacity to follow another person’s line of sight is relevant for a number of abilities including understanding the meaning of an emotional display (Moses, Baldwin, Rosicky & Tidball, 2001; Repacholi, 1998), language acquisition (Baldwin, 1995; Bloom, 2002; Tomasello, 1995), and inferring a range of mental states that include intentions, beliefs, and desires (Baldwin & Moses, 1994; Lee, Eskritt, Symons & Muir, 1998; Meltzoff & Brooks, 2001; Onishi & Baillargeon, 2005). The importance of gaze following as a developmental milestone can be inferred from the case of children with autism. Specifically, autistic children often lack the ability to use gaze direction as a cue to understand a speaker’s referential intent (Baron- Cohen, Baldwin & Crowson, 1997). Research on infants’ ability to follow the gaze direction of others has revealed that from 3 months of age, infants can follow gaze to a highly visible object that is within their immediate visual field (Caron, Caron, Roberts & Brooks, 1997; D’Entremont, 2000; D’Entremont, Hains & Muir, 1997) and later, at 12 months of age, to targets outside their visual field (Carpenter, Nagell & Tomasello, 1998; Moll & Tomasello, 2004; Morissette, Ricard & Gouin-Décarie, 1995). However, there is an ongoing debate in the literature concerning the meaning of infants’ gaze following and whether it should necessarily be construed as being mentalistic. Specifically, there is uncertainty as to whether infants understand adults’ looking as directing their attention towards an object in the environment (Bretherton, 1991; Baron-Cohen, 1995; Carpenter et al., 1998; Caron, Kiel, Dayton & Butler, 2002) or whether infants’ attention is drawn to a location as a result of a built-in orienting response (Langton et al., 2000; Moore & Corkum, 1994; Moore, 1999; Povinelli, 2001). Experimental studies that have provided support for a ‘richer’ or mentalistic interpretation have typically used one of two basic experimental paradigms to demonstrate that infants understand the link between visual perception and objects in the environment. First, the Eye Status paradigm has been used wherein a person systematically moves just his/her eyes, head, or both toward an object in order to identify the cues that elicit gaze following in infants. Using this approach, evidence suggests that it is not until 18 to 19 months that infants are able to follow gaze on the basis of eye movement (head remains frontal; Corkum & Moore, 1995; Moore & Corkum, 1998). However, a recent ERP study has reported the encoding of gaze–object relations on the basis of gaze alone in 9-month-old infants (Senju, Johnson & Csibra, 2006). Before this age, when head and eye movements are discordant (head and eyes turn in different directions), infants seem more inclined to follow the head direction than the eye direction (Corkum & Moore, 1998; Caron, Address for correspondence: Virginia Chow, Department of Psychology, Concordia University, 7141 Sherbrooke Street West, PY 276–3, Montréal, Québec, Canada, H4B 1R6; e-mail: vchow@alcor.concordia.ca