Elsevier

NeuroImage

Volume 50, Issue 1, March 2010, Pages 198-207
NeuroImage

The neural basis of agency: An fMRI study

https://doi.org/10.1016/j.neuroimage.2009.12.054Get rights and content

Abstract

Agency, a feeling that the self is the cause of action, has a strong relationship to the processing of discrepancies between the predicted multi-sensory feedback from one’s intended action and its actual outcome (hereafter, agency error). Although previous studies have explored the neural basis of agency by assessing the brain’s response to agency error, the effects found are confounded by two types of error irrelevant to agency: a mismatch between different sensory inputs in general (sensory mismatch, SM error) and a basic response to any type of prediction error (oddball error). In this functional magnetic resonance imaging study, we identified the neural response specific to agency error by dissociating it from responses to SM and oddball errors. Subjects played a game in which they controlled an on-screen character. Neural responses to rare events of violated control and congruency between types of audio–visual feedback were compared to dissociate agency from SM error. In a separate session, subjects viewed repetitive motions of the character, and neural responses to rare events of unpredictable change in movement were identified as related to oddball error. Agency-error-specific activation was observed in the supplementary motor area (SMA), left cerebellum, right posterior parietal cortex (PPC), and right extrastriate body area (EBA). Oddball errors also activated areas near the PPC and EBA peaks. SM errors activated the pre-SMA and the right posterior superior temporal sulcus. Our results suggest that the SMA, cerebellum, and some parts of the PPC and EBA serve as the neural bases of agency.

Introduction

Agency refers to the feeling, “I am the one causing or controlling a movement” (Gallagher, 2000). Defined as such, agency tells us whether an action is caused by ourselves or by other agents, and it plays an important role in self-consciousness (Newen and Vogeley, 2003).

To explain the mechanism of agency formation, several models have recently been proposed in the field of neuroscience, arising particularly from the observation of pathological agency states, in which patients claim that they have no feeling of control over their actions and feel as if their bodies are moving by their own agency or being controlled by other agents, such as anarchic hand syndrome or schizophrenic patients’ delusions of control (Biran et al., 2006, Asai et al., 2008). According to such proposed models, when we perform an action, our brain predicts the sensory consequences of that action as an assortment of multimodal sensory feedback, exemplified by visuo-proprioceptive association, and compares this prediction with actual sensory feedback (see Wolpert et al., 1995, Wolpert, 1997, Frith et al., 2000, Blakemore et al., 2002 for a review). If these match, then the action is judged as self-induced, giving rise to a sense of agency. On the other hand, if discrepancies between these emerge, then the action is attributed to others, disturbing the sense of agency. This type of prediction error is agency-specific, called “agency error” in this paper. To clarify the neural basis of agency, recent neuroimaging studies have sought brain regions that process this agency error, and have proposed candidate areas, including the posterior parietal cortex (PPC), posterior superior temporal sulcus (pSTS), supplementary motor area (SMA), insula, dorsolateral prefrontal cortex, and cerebellum (David et al., 2008 [Review]). A common method used in these studies is to induce agency error by altering the visual feedback that subjects receive as they control the movement of their own body (e.g., hand) or computer-graphic objects projected on the screen (e.g., virtual hand, cursor, cars) and to record the brain response to this violation of anticipated visuo-proprioceptive feedback (Leube et al., 2003a, Farrer et al., 2003a, David et al., 2007, Schnell et al., 2007).

However, when agency error was induced in previous studies, agency-irrelevant prediction errors were inevitably induced at the same time. Because previous studies have not dissociated brain responses to these errors from those to agency errors, some previous candidate regions might in fact process agency-irrelevant prediction errors. Such errors might take two forms. One is the inter-sensory matching error, which is not specific to visuo-proprioceptive conflict. When the visual feedback is distorted, as in tasks used in previous studies, violation of the predicted visuo-proprioceptive feedback association induces agency error. At the same time, however, it also evokes a more general inter-sensory matching error that is not specific to visuo-proprioceptive conflict and can be induced by any type of sensory mismatch, including audio–visual conflict. This modality-free inter-sensory matching error is irrelevant to agency, and we refer to this as “sensory matching error (SM error)” in this article. The other type of error involves the violation of prediction itself, irrespective of its content. This basic prediction error, described with terms like “oddball” or “mismatch negativity,” has been investigated using EEG (see Friedman et al., 2001, Picton et al., 2000 for a review). Here, we call this prediction error the “oddball error.”

In this study, we aimed to clarify which brain regions are specifically related to agency error by dissociating regions relating to agency error from regions relating to agency-irrelevant SM or oddball errors.

Section snippets

Subjects

Twenty-eight healthy right-handed volunteers with no significant psychiatric or neurological history were studied. Data from one subject were excluded due to head motion, and from three other subjects due to image artifacts. Thus, data from the remaining 24 subjects (mean age 21.04 years; range 18–24 years) are reported. Informed consent was obtained from all subjects prior to their participation. This study was approved by the Institutional Review Board of the Graduate School of International

Behavioral data

Overall good performance in both the agency and SM error task and the oddball error task showed that subjects were concentrating on each task during scanning. In the agency and SM error task, accuracy for the target was 97.7 ± 1.5% (mean ± SD), and response time was 598.9 ± 87.5 ms (mean ± SD). Similarly, accuracy for the target response in the oddball error task was 97.8 ± 6.8% (mean ± SD), response time was 509.4 ± 62.9 ms (mean ± SD). No statistically significant difference was found between the accuracy

Discussion

In this experiment, we aimed to clarify which brain regions specifically relate to agency error by dissociating regions relating to agency error from those relating to agency-irrelevant SM or oddball errors. We found that agency error activated the SMA, the left lateral cerebellum, the right PPC, and the right EBA. From these results, we suggest that these regions play an important role in agency error processing and work as the neural basis of agency, the feeling of “self-causality” in action.

Acknowledgments

This study was supported by the JST/RISTEX, JST/CREST. Y.Y. was supported by a Grant-in-Aid for JSPS fellows (20 · 8136). M.S. was supported by KAKENHI (21119504).

References (77)

  • FarrerC. et al.

    Modulating the experience of agency: a positron emission tomography study

    NeuroImage

    (2003)
  • FarrerC. et al.

    The role of proprioception in action recognition

    Conscious. Cogn.

    (2003)
  • FriedmanD. et al.

    The novelty P3: an event-related brain potential (ERP) sign of the brain's evaluation of novelty

    Neurosci. Biobehav. Rev.

    (2001)
  • FristonK.J. et al.

    Detecting activations in PET and fMRI: levels of inference and power

    NeuroImage

    (1996)
  • GallagherS.

    Philosophical conceptions of the self: implications for cognitive science

    Trends Cogn. Sci.

    (2000)
  • GalleaC. et al.

    Error processing during online motor control depends on the response accuracy

    Behav. Brain Res.

    (2008)
  • GaravanH. et al.

    Dissociable executive functions in the dynamic control of behavior: inhibition, error detection, and correction

    NeuroImage

    (2002)
  • GaravanH. et al.

    A midline dissociation between error-processing and response-conflict monitoring

    NeuroImage

    (2003)
  • GiummarraM.J. et al.

    Mechanisms underlying embodiment, disembodiment and loss of embodiment

    Neurosci. Biobehav. Rev.

    (2008)
  • GrefkesC. et al.

    Crossmodal processing of object features in human anterior intraparietal cortex: an fMRI study implies equivalencies between humans and monkeys

    Neuron

    (2002)
  • HaggardP.

    Conscious intention and motor cognition

    Trends Cogn. Sci.

    (2005)
  • HaggardP. et al.

    Intentional action: conscious experience and neural prediction

    Conscious. Cogn.

    (2003)
  • HaggardP. et al.

    Supplementary motor area provides an efferent signal for sensory suppression

    Cogn. Brain Res.

    (2004)
  • JacksonS.R.

    ‘Action binding’: dynamic interactions between vision and touch

    Trends Cogn. Sci.

    (2001)
  • KaoK.L.C. et al.

    Enhanced detection of visual targets on the hand and familiar tools

    Neuropsychologia

    (2009)
  • LeubeD.T. et al.

    The neural correlates of perceiving one's own movements

    NeuroImage

    (2003)
  • LeubeD.T. et al.

    Observing one's hand become anarchic: an fMRI study of action identification

    Conscious. Cogn.

    (2003)
  • MaravitaA. et al.

    Tool-use changes multimodal spatial interactions between vision and touch in normal humans

    Cognition

    (2002)
  • MiallR.C. et al.

    Forward models for physiological motor control

    Neural Netw.

    (1996)
  • MullerB.W. et al.

    Cortical activation to auditory mismatch elicited by frequency deviant and complex novel sounds: a PET study

    NeuroImage

    (2002)
  • NaatanenR. et al.

    ‘Primitive intelligence’ in the auditory cortex

    Trends Neurosci.

    (2001)
  • NewenA. et al.

    Self-representation: searching for a neural signature of self-consciousness

    Conscious. Cogn.

    (2003)
  • PicardN. et al.

    Imaging the premotor areas

    Curr. Opin. Neurobiol.

    (2001)
  • SchnellK. et al.

    An fMRI approach to particularize the frontoparietal network for visuomotor action monitoring: detection of incongruence between test subjects' actions and resulting perceptions

    NeuroImage

    (2007)
  • ScholzV.H. et al.

    Laterality, somatotopy and reproducibility of the basal ganglia and motor cortex during motor tasks

    Brain Res.

    (2000)
  • SpenglerS. et al.

    Was it me or was it you? How the sense of agency originates from ideomotor learning revealed by fMRI

    NeuroImage

    (2009)
  • WolpertD.M.

    Computational approaches to motor control

    Trends Cogn. Sci.

    (1997)
  • WolpertD.M. et al.

    Internal models in the cerebellum

    Trends Cogn. Sci.

    (1998)
  • Cited by (90)

    • Sense of agency as synecdoche: Multiple neurobiological mechanisms may underlie the phenomenon summarized as sense of agency

      2022, Consciousness and Cognition
      Citation Excerpt :

      The 32 neuroimaging studies of SoA employed different reports to measure SoA, both explicit and implicit (See Tables 1 and 2 and Supplementary Table S2A & S2B). SoA was measured implicitly in eight Feedback Manipulation (Di Plinio et al., 2020; Jardri et al., 2011; Kontaris et al., 2009; Macuga & Frey, 2011; Nahab et al., 2011; Schnell et al., 2007; Spaniel et al., 2016; Tsakiris, Longo, & Haggard, 2010) and three Action-Effect studies (Kikuchi et al., 2019; Matsuzawa et al., 2005; Yomogida et al., 2010). Some studies used explicit reports (graded or dichotic), to measure the self/other dimension of SoA (Feedback Manipulation: Balslev, Nielsen, Lund, Law, & Paulson, 2006; David et al., 2007; Farrer & Frith, 2002; Farrer et al., 2008; Ohata et al., 2020; Action-Effect: Fukushima, Goto, Maeda, Kato, & Umeda, 2013; Renes et al., 2015, 2016; Spengler, von Cramon, & Brass, 2009) or the experience of control (Feedback Manipulation: Miele et al., 2011; Action-Effect: Chambon, Wenke, Fleming, Prinz, & Haggard, 2013, de Bezenac, Sluming, Gouws, & Corcoran, 2016).

    View all citing articles on Scopus
    View full text