Modulation of the pre-supplementary motor area reduces the sense of agency (Commentary on Cavazzana et al.).

Abstract

Institute of Behavioural Neuroscience (IBN), Department of Experimental Psychology, University College London (UCL), London,UKThe sense of agency is the capacity to control one’s own actions and, through them, events in the outside world (Haggard & Chambon,2012). The proposition ‘I did that’ is an important aspect of our everyday lives. For example, this sense associates a light turning on toone’s action when pressing a switch. Abnormalities in this sense can lead to illnesses such as psychosis (increased sense of agency) anddepression (decreased sense of agency). Despite a large amount of theoretical work on the sense of agency, the neural correlates arepoorly understood. Only a handful of imaging and brain stimulation studies have attempted to investigate how different brain areas con-tribute to the emergence of this feeling. Cavazzana et al. (2015) describe evidence showing that modulation of activity in the pre-supple-mentary motor area (pre-SMA) using transcranial electrical brain stimulation can reduce the sense of agency. This is an important findingas it highlights a possible causal link between the activity of the pre-SMA and one’s sense of agency, and thus could have clinical impli-cations.One way of studying the sense of agency is through ‘explicit agency attribution tasks’. In these tasks a person judges whether they did ordid not cause a specific event. For example, a computer program presents a tone whenever the person presses a button. Immediate presenta-tion of the tone following the participant’s key press creates the sense that they caused the presentation of the tone (e.g. agency). If the toneis presented with a delay the sense of agency is reduced. This suggests that the sense of agency is computed by comparing the prediction ofaction outcome(s) with actual outcomes. Figure 1A is a diagram of how this process might work.In addition to explicit measures of the sense of agency there are also several implicit measures, such as intentional binding. Inten-tional binding refers to the temporal compression between a voluntary action and its sensory effect (Haggard & Tsakiris, 2009). In thetask of Cavazzana et al. (2015), an effect (e.g. a tone) is presented after a certain delay, rather than immediately, subsequent to the par-ticipant’s key press. The participant is then required to report the time on a clock indicating when they pressed the button or when thetone was presented. The sense of agency leads to an overestimation of time of action and an underestimation of time (Haggard & Tsak-iris, 2009). Cavazzana et al. (2014) adapted this task for children and presented a series of letters instead of a clock. Participants wererequired to report the letter presented on the screen at the time of their key press or at the time of the presentation of the tone. Cavazz-ana et al. (2015) used this novel method to test healthy adults in combination with a non-invasive brain stimulation paradigm. Anodaland cathodal transcranial direct current stimulation was used to modulate activity of the pre-SMA. Although it has been shown that ano-dal and cathodal stimulation cause excitation and inhibition of neuronal activity, respectively (Nitsche & Paulus, 2000), the results ofCavazzana et al. (2015) did not differentiate between the polarity of stimulation, i.e. both polarities reduced the sense of agency. Onepossible explanation for this is based on the contribution of the SMA and pre-SMA in the perception of time (Lewis & Miall, 2003;Allman et al., 2014). Transcranial direct current stimulation of the pre-SMA might have changed the expected duration of the delay andconsequently reduced the sense of agency. Figure 1 illustrates this possibility. The strongest sense of agency occurs when sensory feed-back and predicted outcomes match (Fig. 1B). Changes to the outcome prediction lead to a weakening of the sense of agency (Fig. 1C).As illustrated, this effect is not dependent on the direction of change of the predicted outcome. Moore et al. (2010) demonstrated a sim-ilar reduction in the sense of agency by stimulating the pre-SMA using transcranial magnetic stimulation. Their stimulation method,however, prevented them from studying both excitation and inhibition of neuronal activity. Application of transcranial direct current stim-ulation enabled Cavazzana et al. (2015) to extend their findings and show that both anodal and cathodal stimulation both lead to similarbehavioural outcomes. This is important as it shows that, regardless of the direction of the effect of stimulation, modulation of activityof the pre-SMA reduces the sense of agency.In summary, the findings of Cavazzana et al. (2015) constitute considerable progress towards understanding the brain network contributingto the sense of agency and highlight the important role of brain areas involved in time perception. More work, such as dynamic causalmodelling and psychophysiological interaction, will be required to elucidate the brain network(s) involved and also studies using behaviouralmanipulations to disentangle the effects of time perception from the sense of agency.