Elsevier

NeuroImage

Volume 25, Issue 1, March 2005, Pages 133-140
NeuroImage

An fMRI study of joint attention experience

https://doi.org/10.1016/j.neuroimage.2004.10.047Get rights and content

Abstract

Although much is now known about eye movement detection, little is known about the higher cognitive processes involved in joint attention. We developed video stimuli which when watched, engender an experience of joint attention in the observer. This allowed us to compare an experience of joint attention to nonjoint attention within an fMRI scanning environment. Joint attention was associated with activity in the ventromedial frontal cortex, the left superior frontal gyrus (BA10), cingulate cortex, and caudate nuclei. The ventromedial frontal cortex has been consistently shown to be activated during mental state attribution tasks. BA10 may serve a cognitive integration function, which in this case seems to utilize a perception–action matching process. The activation we identified in BA10 overlaps with a location of increased grey matter density that we recently found to be associated with autistic spectrum disorder. This study therefore constitutes evidence that the neural substrate of joint attention also serves a mentalizing function. The developmental failure of this substrate in the left anterior frontal lobe may be important in the etiology of autistic spectrum disorder.

Introduction

Joint attention is a process whereby two individuals attend to the same object because one is checking, monitoring, or following the focus of attention of the other. It is an early cognitive skill to emerge; infants will look in the direction of other people's head and eye movements at about 6 months of age (Scaife and Bruner, 1975), and by the age of 18 months they will reliably follow the gaze of others (Butterworth and Jarrett, 1991). It may therefore be an important developmental precursor to other social and cognitive abilities.

In particular, joint attention processes may constitute cognitive precursors to higher level mental state attribution such as that concerned with understanding false belief (Baron-Cohen, 1994, Gopnik et al., 1994, Mundy, 1993). In a longitudinal study, joint attention at 20 months positively correlated with measures of theory of mind collected between 3 and 4 years of age (Charman et al., 2001).

A number of lines of evidence also suggest that joint attention may be an important precursor to the development of language abilities: onset consistently precedes the emergence of referential language in the second year of life; the ability to engage attention jointly during naturalistic infant–mother interactions predicts infants' word comprehension and word production (Carpenter et al., 1998); joint attention bids measured between 6 and 18 months make a unique contribution to language development at 30 months (Morales et al., 2000); and gaze following, which is an important component of joint attention, predicts vocabulary between 1 and 2 years (Morales et al., 1998).

Joint attention skills are notably poor in autism and severity of joint attention impairment predicts the outcome of this disorder (Dawson et al., 2002). In one study (Charman et al., 1997), 12 children with symptoms of autism at 20 months (whose diagnostic status was subsequently confirmed) switched gaze less often between a toy and an experimenter to share aspects of a situation with the experimenter. Children with autism are impaired in their comprehension and production of protodeclarative pointing (pointing aimed at sharing interest with another rather than aimed at requesting objects), and they also seem unable to use the direction of gaze of a person to infer what the person wants (Baron-Cohen, 1989, Baron-Cohen, 1995). They are significantly less likely to check the focus of attention of an experimenter in ambiguous situations compared to control groups (Phillips et al., 1992). However, it appears that these children do possess the ability to compute the direction of another person's gaze and thus to understand what a person is looking at (Leekam et al., 1997, Leekam et al., 2000). Recently, children with autism have been shown to be able to use eye direction cues to equal advantage to controls (Kylliainen and Hietanen, 2004, Senju et al., 2004) in reducing their reaction times in responding to a lateralized cue. Rather, it seems they fail to spontaneously follow the direction of gaze of another person with eyes and head turned toward an object (Leekam et al., 2000)—they fail to join in with the other person's attentional focus. This experiment has been largely incorporated into the Autism Diagnostic Observation Schedule (Lord et al., 2000), where joint attention items discriminate well between children with autism and those with other clinical problems.

If gaze detection is functioning normally in autism, it may be that it is another component of the joint attention process that is functioning abnormally in autism. As well as being involved with the detection of another individual's attentional direction, joint attention is also concerned with the tendency to switch gaze between an object and a person, and to direct the attention of another (Leekam et al., 1997, Phillips et al., 1992, Swettenham et al., 1998). Notably, the tendency to direct attention and the ability to (spontaneously) monitor attentional direction are both impaired in autism, suggesting that they could both be affected by a common mechanism, possibly that are concerned with integrating them in the formation of a representation of joint attention. Baron-Cohen, 1994, Baron-Cohen, 1995 proposed the existence of a modular ‘Shared Attention Mechanism’ (SAM), which could serve such an integration function, suggesting that this would be an integral part of a ‘theory of mind’ mechanism.

In examining the neural correlates of joint attention, there is now strong evidence for specialized neural mechanisms serving gaze and eye movement perception (Baron-Cohen et al., 2001, Hoffman and Haxby, 2000, Jellema et al., 2000, Perrett et al., 1989, Puce and Perrett, 2003, Puce et al., 1998, Wicker et al., 1998). At a higher cognitive level of processing, Wicker et al. (2003) investigated the neural correlates of attributing emotion secondary to gaze processing. Yet, to our knowledge no functional neuroimaging paradigm has been used to investigate integrating information from gaze perception with executed gaze direction at the level of forming a representation of joint attention itself. If joint attention is a cognitive precursor to ‘theory of mind’ capacity, this may involve similar neural substrate such as ventromedial frontal cortex. In this study, we developed video clips (Fig. 1) that engendered an experience of joint attention in observers who watched them. The joint attention condition was contrasted with a very similar but nonjoint attention condition, where the model's gaze moved equally as often but was always directed elsewhere. This has allowed us to investigate the neural substrate of joint attention at the level concerned with integrating gaze-direction information from the self and another individual.

Section snippets

Materials

We hypothesized that there would be neural structures more active during joint attention than during a control situation when attention was directed to differing locations. To create these situations, we constructed a number of video clips by positioning a model behind a silk screen, with the video frame showing their head in the top two thirds and the screen in the bottom third. A red dot projected onto the screen moved horizontally and randomly between four points spaced evenly apart. The

Results

First we identified areas of activation in the joint condition compared to rest and nonjoint condition compared to rest (Fig. 2; Table 1, Table 2). Both conditions, as expected, caused activation of the posterior aspect of the superior temporal gyrus and frontal eye fields. These are likely to be associated with detection of attentional direction and the preparation and execution of eye movements, as the moving dot changed position. Other areas were also activated, including subcortical areas,

Discussion

One of the largest clusters of activation specific to joint attention was centered around the ventromedial frontal cortex in an area associated with ‘theory of mind’ function. The 265-voxel cluster identified with a center at coordinates x = 2, y = 33, and z = 40 is in a region consistently associated with ‘theory of mind’ (Castelli et al., 2000, Castelli et al., 2002, Fletcher et al., 1995, Frith, 2001, Frith and Frith, 2000, Gallagher and Frith, 2003, Gallagher et al., 2000) as well as

Acknowledgments

We are especially grateful for financial support from the National Alliance for Autism Research, to Stuart Duncan and Iain Harold for their work on stimulus preparation, and to David Lovell, Duncan Stewart, and the other volunteers for their help with conducting this experiment.

References (50)

  • M. Morales et al.

    Following the direction of gaze and language development in 6-month-olds

    Infant Behav. Dev.

    (1998)
  • M. Morales et al.

    Responding to joint attention across the 6- through 24-month age period and early language acquisition

    J. Appl. Dev. Psychol.

    (2000)
  • R.C. Oldfield

    The assessment and analysis of handedness: the Edinburgh Inventory

    Neuropsychologia

    (1971)
  • G.D. Waiter

    A voxel based investigation of brain structure in male adolescents with autistic spectrum disorder

    NeuroImage

    (2004)
  • B. Wicker et al.

    Brain regions involved in the perception of gaze: a PET study

    NeuroImage

    (1998)
  • B. Wicker et al.

    Being the target of another's emotion: a PET study

    Neuropsychologia

    (2003)
  • J.H. Williams et al.

    Imitation, mirror neurons and autism

    Neurosci. Biobehav. Rev.

    (2001)
  • M.A. Arbib

    The mirror system, imitation and the evolution of language

  • S. Baron-Cohen

    Joint-attention deficits in autism: towards a cognitive analysis

    Dev. Psychopathol.

    (1989)
  • S. Baron-Cohen

    How to build a baby that can read minds: cognitive mechanisms in mindreading

    Cah. Psychol. Cogn./Curr. Psychol. Cogn.

    (1994)
  • S. Baron-Cohen

    Mindblindness: An Essay on Autism and Theory of Mind

    (1995)
  • S. Baron-Cohen et al.

    The “Reading the Mind in the Eyes” test revised version: a study with normal adults, and adults with Asperger syndrome or high-functioning autism

    J. Child Psychol. Psychiatry

    (2001)
  • G. Butterworth et al.

    What minds have in common in space-spatial mechanisms serving joint visual–attention in infancy

    Br. J. Dev. Psychol.

    (1991)
  • M. Carpenter et al.

    Social cognition, joint attention, and communicative competence from 9 to 15 months of age

    Monogr. Soc. Res. Child Dev.

    (1998)
  • Cited by (149)

    • Infants Oscillatory Frequencies change during Free-Play

      2021, Infant Behavior and Development
    • Beta oscillations precede joint attention and correlate with mentalization in typical development and autism

      2019, Cortex
      Citation Excerpt :

      One of the sources of the beta power was the right TPJ, which is suggested to be part of the mentalization network (Saxe, Whitfield-Gabrieli, Scholz, & Pelphrey, 2009). This area shows consistent activation during JA tasks in adults (Caruana, de Lissa, & McArthur, 2015; Redcay et al., 2010; Schilbach et al., 2010; Williams et al., 2005), as well as during development from childhood to adolescence (Oberwelland et al., 2016). This network is consistently more active in JA as compared to nJA behaviors (Williams et al., 2005).

    View all citing articles on Scopus
    View full text