Exp Brain Res (2010) 201:155–165
DOI 10.1007/s00221-009-2019-7RESEARCH ARTICLE
Eye movements aYrm: automatic overt gaze and arrow cueing
for typical adults and adults with autism spectrum disorder
Gustav Kuhn · Valerie Benson · Sue Fletcher-Watson ·
Hanna KovshoV · Cristin A. McCormick · Julie Kirkby ·
Sue R. Leekam
Received: 24 June 2009 / Accepted: 12 September 2009 / Published online: 2 October 2009
 Springer-Verlag 2009
Abstract People with autism spectrum disorder (ASD)
show reduced interest towards social aspects of the envi-
ronment and a lesser tendency to follow other people’s gaze
in the real world. However, most studies have shown that
people with ASD do respond to eye-gaze cues in experi-
mental paradigms, though it is possible that this behaviour
is based on an atypical strategy. We tested this possibility
in adults with ASD using a cueing task combined with eye-
movement recording. Both eye gaze and arrow pointing
distractors resulted in overt cueing eVects, both in terms of
increased saccadic reaction times, and in proportions of
saccades executed to the cued direction instead of to the tar-
get, for both participant groups. Our results conWrm previ-
ous reports that eye gaze cues as well as arrow cues result
in automatic orienting of overt attention. Moreover, since
there were no group diVerences between arrow and eye
gaze cues, we conclude that overt attentional orienting in
ASD, at least in response to centrally presented schematic
directional distractors, is typical.
Keywords Gaze following · Gaze cueing · Oculomotor
inhibition · Autism · Eye movements · Social attention
Introduction
Autism spectrum disorder (ASD) refers to a range of
neuro-developmental conditions characterized by impair-
ments in social interaction and communication, and the
presence of repetitive behaviours and restricted interests
(American Psychiatric Association, APA 1994). One of
the earliest manifestations of autism is a deWcit in joint
attention behaviours with social partners, and this is
thought to restrict the child’s opportunities for social
learning, impairing subsequent social and communication
development (Baron-Cohen et al. 1996; Klin et al. 2003;
Mundy 1995; Mundy and Burnette 2005). One component
of joint attention is gaze following whereby the observer
traces the line of sight of another person towards the
object/target/person that is being looked at (Emery 2000;
Frischen et al. 2007).
In typical development spontaneous gaze following
behaviour develops during infancy (for reviews see Nation
and Penny 2008; Senju and Johnson 2009) and it has been
suggested that gaze following may be innate (Baron-Cohen
1994; Hood et al. 1998; Tantam 1992). Several behavioural
studies have found spontaneous gaze following impair-
ments in children with ASD (Leekam et al. 1998; Leekam
et al. 2000) although they have an intact ability to compute
gaze direction geometrically (Leekam et al. 1997). The rea-
sons for this spontaneous gaze following impairment are far
G. Kuhn
Department of Psychology, University of Durham,
South Road, Durham DH1 3LE, UK
e-mail: Gustav.Kuhn@Durham.ac.uk
Present Address:
G. Kuhn (&)
Department of Psychology, Brunel University,
School of Social Sciences, Uxbridge UB8 3PH, UK
e-mail: Gustav.Kuhn@brunel.ac.uk
V. Benson · H. KovshoV · C. A. McCormick · J. Kirkby
Southampton University, Southampton, UK
S. Fletcher-Watson
Newcastle University, Newcastle, UK
S. R. Leekam
CardiV University, CardiV, UK123

156 Exp Brain Res (2010) 201:155–165from clear, and several diVerent explanations have been
proposed (see Nation and Penny 2008). Individuals with
ASD may fail to interpret gaze or head movement as an
index of the other person’s state of attention towards a par-
ticular location (Baron-Cohen 1995). This interpretation
coincides with the general Wnding that individuals with
ASD typically show deWcits in their ability to read the men-
tal states of other people (theory of mind deWcits) (Baron-
Cohen 2000; Baron-Cohen et al. 1996). However, gaze fol-
lowing does not necessarily require intact theory of mind
skills. For example, gaze following is often observed in
macaque monkeys, who are not thought to have representa-
tions of other people’s states of mind (Deaner and Platt
2003). Others suggest that infants learn to predict where an
object will appear in response to head turn which in later
development results in automatic gaze following (Corkum
and Moore 1998). Individuals with ASD generally pay less
attention to social stimuli, and in particular may Wxate less
than usual on the eye-region (Dalton et al. 2005; Klin et al.
2002). On the other hand, abnormalities in gaze following
could reXect more general attentional problems with disen-
gaging attention from one stimulus and shifting it to
another location (e.g. Hill 2004; Landry and Bryson 2004;
OzonoV et al. 1991).
Evidence for a gaze following deWcit in ASD is based on
studies involving interpersonal interaction within real-
world experimental paradigms, but attentional orienting in
response to gaze cues is more often investigated using a
Posner type cueing task (Posner 1980). In this paradigm
participants are presented with pictures of faces where the
gaze is either averted to the left or the right (e.g. Friesen
and Kingstone 1998). Following this shift in eye gaze, a tar-
get appears either at the gazed at (valid) location, or at the
non-gazed at (invalid) location. Participants typically detect
targets appearing at the valid location more rapidly than tar-
gets appearing at the invalid location, thus demonstrating
that attention has been oriented towards the gazed at loca-
tion. The diVerence in reaction time (RT) for detecting tar-
gets at valid versus invalid locations provides an index of
gaze cueing (for reviews see Frischen et al. 2007; Langton
et al. 2000). Importantly this gaze cueing eVect occurs even
when the gaze cues are non-predictive (Friesen and
Kingstone 1998) or even counter predictive of the target
location (Driver et al. 1999; Friesen et al. 2004), thus
suggesting that this type of attentional orienting is reXexive.
Given that a lack of spontaneous gaze following is one
of the hallmarks of ASD, it has been rather surprising that
few studies have found abnormalities in gaze cueing in Pos-
ner-type tasks (see Nation and Penny 2008 for a review). At
Wrst sight, children with ASD seem to have comparable
cueing eVects for eye gaze to that of typically developing
(TD) children (Kylliainen and Hietanen 2004; Swettenham
et al. 2003). Nevertheless, there is evidence of subtle atypi-
calities in the way in which children with ASD respond to
gaze cues. For example, Ristic et al. (2005) found that chil-
dren with ASD were more reliant than TD children on the
presence of motion in the eye-gaze cue to trigger attentional
shifts, when gaze-direction was not predictive of the target
location. Senju et al. (2004) found that children with ASD
responded reXexively to both arrow and eye-gaze cues,
while TD children were able to inhibit a response to a non-
predictive arrow cue. This suggests that children with ASD
use the same, learnt and non-social, mechanism for respond-
ing to both social (eye-gaze) and non-social (arrow) cues.
This interpretation is supported by the work of Chawarska
et al. (2003) who demonstrated, by recording eye-move-
ments during an eye-gaze cueing task, that toddlers with
ASD responded more quickly to an eye-gaze cue than their
TD peers. This group diVerence disappeared when a non-
social cue was used, leading to the suggestion that the TD
children were slower to respond to eye-gaze cues because of
the extra time they spent processing this social stimulus.
This paper will explore responses to eye-gaze cues in
adults with ASD using a novel, cueing paradigm which mea-
sures overt attention using eye-movements as well as RT.
Such tasks oVer a more direct way of investigating atten-
tional abnormalities, as they are more closely related to the
deviations in gaze following typically observed in ASD.
Gaze cueing in TD adults has been investigated by mea-
suring overt attention (i.e. eye movements) (Kuhn and Benson
2007; Kuhn and Kingstone 2009; MansWeld et al. 2003;
Ricciardelli et al. 2002). In one of these paradigms partici-
pants are required to Wxate a central Wxation point and are
instructed to look at targets on either side of the screen as
indicated by a colour change of the Wxation point. During
Wxation the eye gaze of a centrally presented face shifts
either to the right or the left. Participants are therefore
required to saccade either in the same direction (congruent
trials), or the opposite direction (incongruent trials) to
which the distractor eyes are pointing. Results have shown
that participants are signiWcantly slower at initiating eye
movements on incongruent than on congruent trials, even
when the gaze cue is non-predictive (Kuhn and Benson
2007; Ricciardelli et al. 2002) or counter predictive (Kuhn
and Kingstone 2009) of the saccade instruction. Moreover,
participants typically make more errors on incongruent than
on congruent trials thus demonstrating that in TD adults,
gaze cues result in automatic gaze following.
Another opportunity aVorded by this study is to explore
the responses of adults with ASD to a diVerent attentional
cue: arrows. Initially it was thought that attentional orient-
ing in response to gaze cues may be unique to biologically
relevant stimuli (Friesen and Kingstone 1998). However,
numerous recent studies have shown that symbolic cues,
such as arrows, result in automatic shifts in both covert
(Bonato et al. 2009; Hommel et al. 2001; Ristic and Kingstone123