Research Report

Inducing Disbelief in Free Will Alters Brain Correlates of Preconscious Motor Preparation: The Brain Minds Whether We Believe in Free Will or Not
Davide Rigoni1,2, Simone Khn1,3, Giuseppe Sartori4, and Marcel Brass1
1

Psychological Science 618 22(5) 613 The Author(s) 2011 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0956797611405680 http://pss.sagepub.com

Department of Experimental Psychology, University of Gent; 2Department of Developmental Psychology and Socialization, University of Padova; 3Charit University Medicine, St. Hedwig Krankenhaus, Clinic for Psychiatry and Psychotherapy, Berlin, Germany; and 4Department of General Psychology, University of Padova

Abstract The feeling of being in control of ones own actions is a strong subjective experience. However, discoveries in psychology and neuroscience challenge the validity of this experience and suggest that free will is just an illusion. This raises a question: What would happen if people started to disbelieve in free will? Previous research has shown that low control beliefs affect performance and motivation. Recently, it has been shown that undermining free-will beliefs influences social behavior. In the study reported here, we investigated whether undermining beliefs in free will affects brain correlates of voluntary motor preparation. Our results showed that the readiness potential was reduced in individuals induced to disbelieve in free will. This effect was evident more than 1 s before participants consciously decided to move, a finding that suggests that the manipulation influenced intentional actions at preconscious stages. Our findings indicate that abstract belief systems might have a much more fundamental effect than previously thought. Keywords motor processes, electrophysiology, consciousness, cognitive neuroscience, free will
Received 9/6/10; Revision accepted 1/22/11

Recent discoveries in psychology and neuroscience suggest that the perception of free will is illusory (Wegner, 2002) and just arises from unconscious brain activity (Hallett, 2007; Libet, Gleason, Wright, & Pearl, 1983; Soon, Brass, Heinze, & Haynes, 2008). Despite the scientific and philosophical relevance of these findings, people seem to naturally believe in free will (Baumeister, Masicampo, & DeWall, 2009). Furthermore, this belief seems to be essential for peoples well-being (Leotti, Iyengar, & Ochsner, 2010). But how would people behave if free will really turned out to be an illusion? Previous research has demonstrated that control and agency beliefs affect peoples intention to act, their motivations, and their behavior (e.g., Ajzen, 2002; Bandura, 1982). For instance, when people believe that they cannot exercise control over their behavior and the events that affect their lives, they perform poorly, even though they have the skills to do better (Bandura, 1989). In addition, recent studies in social psychology have shown that believing in free will also has implications for social behavior: Weakening beliefs in free will increases cheating (Vohs & Schooler, 2008) and aggressiveness, and reduces prosocial behavior (Baumeister et al., 2009).

Although this research shows that undermining beliefs in free will influences performance and social behavior, the question arises whether such an influence can be demonstrated on a basic motor level as well. In the study reported here, we investigated whether inducing the idea that people do not have free will changes the brain correlates of voluntary motor preparation. Our hypothesis was that denying free will would deplete participants intentional involvement in the preparation of voluntary motor actions. Specifically, we predicted that a disbelief in free will affects brain correlates of volitional processes at the earliest stages of voluntary actions. We recorded event-related potentials while participants were performing the Libet task (Libet et al., 1983). In this task, participants execute voluntary movements and then report the time at which the intention to move occurred to them (W judgment). It has been demonstrated that voluntary movements are
Corresponding Author: Davide Rigoni, Department of Developmental Psychology and Socialization, University of Padova, Via Venezia 8, 35131 Padova, Italy E-mail: davide.rigoni@unipd.it

614 preceded by a readiness potential (RP), a slow negative-going wave that can be divided into an early component (early RP), which reflects movement preparation, and a late component (late RP), which is involved in the specific programming for movement execution (Kornhuber & Deecke, 1965; Shibasaki & Hallett, 2006). The early RP starts up to 2,000 ms before movement execution, and it has been proposed that the early RP is generated by the pre-supplementary motor area. The late RP begins around 500 ms before movement onset and has been associated with the activity of the primary motor cortex and supplementary motor area proper (Shibasaki & Hallett, 2006). The early RP is modulated by the level of intentional involvement (Keller & Heckhausen, 1990; Kornhuber, 1984; Libet, Wright, & Gleason, 1982), and it is restricted to movements that are executed with the introspective feelings of the willful realization of the intention to move at a particular time (Lang, 2003, p. 20). Conversely, the late RP is influenced by factors concerning motor execution (Shibasaki & Hallett, 2006). Our hypothesis was that inducing disbelief in free will would lead to a reduction of early-RP amplitude. Prior to the task, participants were randomly assigned to one of two groups: The no-free-will group read a text claiming that scientists now recognize that free will is an illusion; the control group read a text on consciousness that did not mention free will at all. This procedure has been shown to affect peoples belief about free will (Vohs & Schooler, 2008).

Rigoni et al. had normal or corrected-to-normal vision, provided informed consent, and were paid 15. The study was conducted according to the Declaration of Helsinki, and approval of the local ethical committee was obtained.

Experimental design
Participants read one of two passages from Cricks book The Astonishing Hypothesis (1994; see the Supplemental Material available online for the excerpts): The no-free-will group read a passage claiming that scientists now recognize that free will is an illusion; the control group read a passage on consciousness that did not mention free will. To encourage participants to read the material carefully, the instructor stated that a comprehension test would be administered at the end of the experiment. After reading the text, participants performed a variant of the Libet task (Fig. 1; Banks & Isham, 2009; Libet et al., 1983; Rigoni, Brass, & Sartori, 2010). In this task, participants sat in front of a monitor and decided when to press a button. At the start of each trial, a clock appeared on screen. After 500 ms, a red cursor appeared at a random position on the clock and started to rotate clockwise. Participants were instructed to press the button with their right index finger whenever they chose, after at least one rotation of the cursor. Auditory feedback followed each button press. The feedback could be delivered simultaneously or delayed by 20 ms, 40 ms, or 60 ms. (The rationale behind the feedback manipulation is outlined in the Supplemental Material.) After a random interval of 800 ms to 1,500 ms after the button press, the cursor stopped. Participants were then asked to report the position of the cursor at the time they initially decided to press the button. Participants first practiced the task and then started the experimental session (120 trials administered in two blocks).

Method Participants
Data were recorded from 30 participants (20 females, 10 males; age range = 1824 years). All participants were righthanded according to the Edinburgh inventory (Oldfield, 1997),

0 ms

500 ms

> 2,560 ms

8001,500 ms 060 ms

Response-Dependent Interval

Trial Starts

Intention to Press

Button Press

Cursor Stops

Intention Judgment

Fig. 1. Illustration of the task. At the start of each trial, a clock appeared. After 500 ms, a red cursor appeared at a random position on the clock and started to rotate clockwise. Participants were instructed to press a button with their right index finger at a time of their own choosing, after at least one rotation of the cursor (i.e., > 2,560 ms). Auditory feedback followed each button press. The feedback was delivered simultaneously or delayed by 20 ms, 40 ms, or 60 ms. After a random interval of 800 ms to 1,500 ms after the button was pressed, the cursor stopped. Participants then used a mouse to position the cursor where it had been located when they initially decided to press the button.

The Brain Minds Whether We Believe in Free Will or Not Afterward, we administered the Free Will and Determinism (FWD) scale (Rakos, Laurene, Skala, & Slane, 2008) to assess whether the reading manipulation affected beliefs about free will. Fourteen items of the FWD scale assess beliefs about whether other people have free will (e.g., Human beings actively choose their actions and are responsible for the consequences of those actions); these items form the General FWD subscale. The other eight items assess the free-will beliefs of the individual taking the assessment (e.g., My decisions are influenced by a higher power); these items form the Personal FWD subscale. Participants also completed the Tangney SelfControl scale (Tangney, Baumeister, & Boone, 2004) and the Marlow-Crowne Social Desirability scale (Tatman, Swogger, Love, & Cook, 2009).
40 30 20 10 0 Control No Free Will p < .01

615

Personal FWD Score

Group
Fig. 2. Mean score on the Personal subscale of the Free Will and Determinism scale (FWD; Rakos, Laurene, Skala, & Slane, 2008) as a function of group. Error bars show standard deviations.

EEG recordings
EEGs were recorded using a 30-channel elastic electrode cap (Easycap, Herrsching, Germany), on which electrodes are placed according to a modified 10-20 setting. The activity of left and right mastoids was also recorded. Electrode impedance was kept below 5 k. Amplified voltages were sampled at 512 Hz. Electrooculograms were recorded with bipolar montage. Continuous EEG data were analyzed off-line using Brain Vision Analyzer software (Brain Products, Munich, Germany). The signal was rereferenced to the average signal of the mastoids and filtered (notch filter of 50 Hz, band-pass filter from 0.016 to 70 Hz, 24 dB/octave attenuation), and epochs were time-locked to participants button press (using a time window between 2,500 ms prior and 500 ms after the button press, with the interval between 2,500 ms and 2,300 ms as a baseline). Ocular movement artifacts were corrected using a semiautomatic blink-detection algorithm (Gratton, Coles, & Donchin, 1983). Epochs containing artifacts were rejected on the basis of visual inspection and automatic artifact detection (peak to peak < 100 V). On average, there were 101 artifactfree trials per participant (SD = 16.62; no-free-will group: M = 104 trials, SD = 13.26; control group: M = 98 trials, SD = 19.06). Data from 1 participant in the no-free-will group was excluded from further analysis because it did not show a clear negativity prior to the actual movement.

control group did, but the effect was only marginally significant (control group: M = 80.93, SD = 10.13; no-free-will group: M = 86.13, SD = 9.27), t(27) = 1.45, p = .08. Scores on the Tangney Self-Control and the Marlow-Crowne Social Desirability scales were not significantly different between the two groups, all ps > .1.

Reported time of intentions (W judgments)

To test whether the experimental manipulation had an effect on the reported time of intentions to respond, we performed an ANOVA with group (control, no free will) as the independent variable and W judgment as the dependent variable. The analysis revealed no significant effect, p = .41, W judgments control group: 242 ms before button press; no-free-will group: 223 ms before button press; thus, it appears that the free-will manipulation did not influence the reported time of conscious intention to move (see the Supplemental Material available online for the analysis of action-monitoring processes).

EEG data
In a first analysis, the epoch from 2,300 to 100 ms before the participant pressed the button was divided into 11 nonoverlapping 200-ms windows. The mean amplitude across all time points in each window was then entered into a mixed ANOVA with a 2 (group: control, no free will) 11 (time window) 3 laterality (3, z, 4) 3 region (centro-parietal, central, fronto-central) design. When appropriate, Greenhouse-Geisser correction was used. The RP differed between the two groups, as revealed by the Time Group interaction, F(1, 10) = 2.353, p = .011, p2 = .086; the RP had a larger amplitude in the control group than in the no-free-will group. The analysis revealed a main effect of time, F(10, 270) = 47.834, p < .0001, p2 = .639; this effect reflected the slow negative-going increase in the amplitude of the RP prior to the motor response. The increase of the RP was not uniform across the scalp, as reflected by the interactions between time and laterality, F(20, 540) = 8.029, p < .0001, p2 = .229, and between time and region, F(20, 540) = 3.551,

Results Self-report scales

Scores on the Personal FWD subscale revealed a significant difference between the two groups (no-free-will group: M = 29.79, SD = 3.95; control group: M = 33.6, SD = 3.22), t(27) = 2.86, p < .01 (Fig. 2); this indicates that the manipulation was effective in weakening beliefs in free will related to oneself in the no-free-will group. Conversely, scores on the General FWD subscale did not differ between the two groups, p = .26; the no-free-will group scored lower on this scale than the

616
Button Press Control Group No-Free-Will Group

Rigoni et al.

12 10

Voltage (V)

8 6 4 2 0 2,500

2,000 1,300

1,500 1,100

1,000 900

500 700

0 500

500 300 Time (ms)

3.5 V 1,500 0 V 3.5 V

Fig. 3. Average event-related potential (ERP) amplitude across all electrodes as a function of time and group. The gray area indicates the time range during which the two groups started to differ significantly and the late readiness potential began. The red and the blue circles in the ERP signals indicate the average times at which the intention to respond occurred in the two groups. The topographical maps beneath the graph show the difference between the two groups in premovement brain activity.

2 = .116. The RP amplitude was larger over the p < .019, p midline and over the sites contralateral to the hand movement than over the ipsilateral sites, and it was more pronounced over the central and fronto-central scalp locations than over the centro-parietal scalp locations. These results are consistent with findings reported in the literature on RPs (Shibasaki & Hallett, 2006). The lack of significant Time Laterality Group and Time Region Group interactions revealed that the RP had a similar scalp distribution in the two groups. No other effects reached or approached significance, all ps > .1. To determine the time at which RPs in the two groups started to differ, we performed t tests on the 11 time windows averaged over the electrodes in which the RP was maximal (FCz, FC3, Cz, and C3). This analysis revealed that smaller RP amplitudes in the no-free-will group than in the control group appeared around 1,300 ms before button press (see the topographical maps in Fig. 3), t(27) = 2.45, p = .01. The early RP was then obtained by defining a window of interest from 1,300 ms to 600 ms prior to the button press; this range captured the activity reflecting only early motor preparatory stages (Shibasaki & Hallett, 2006). As predicted, the amplitude of the early RP was lower in the no-free-will group than in the 2 control group (Fig. 3), F(1, 28) = 4.43, p < .05, p = .136. To test whether the disbelief in free will was associated with changes in the earlier stages of motor preparation, we separately correlated the Personal FWD score with the earlyRP amplitude and the late-RP amplitude, which was measured

as the peak amplitude in the time window from 600 ms to 100 ms (Shibasaki & Hallett, 2006). Our hypothesis was that disbelief in free will was associated with the amplitude of the early RP, but not with that of the late RP. The analysis across all participants confirmed this hypothesis: The Personal FWD score correlated negatively with the early-RP amplitude, r(29) = .404, p < .05, but showed no correlation with the late RP, p = .173.

Discussion
In the study reported here, we demonstrated that weakening the belief in free will affects brain processes underlying early stages of motor preparation. In particular, the early-RP amplitude was lower in the no-free-will group than in the control group (Fig. 3). Furthermore, the no-free-will group scored lower on the Personal FWD subscale than the control group did (Fig. 2). The degree of disbelief in free will correlated with the amplitude of the early RPsa finding indicating that RP amplitudes were smaller for participants with weaker beliefs in free willbut the degree of disbelief in free will showed no correlation with late-RP amplitudes. This suggests that beliefs in free will are related to early motor preparation rather than to processes associated with motor execution. Furthermore, we excluded the possibility that other brain processes unrelated to volition and preparation to action, such as sensory processes, were attenuated in the no-free-will group (see the

The Brain Minds Whether We Believe in Free Will or Not Supplemental Material available online). Finally, we showed that inducing disbelief in free will does not affect peoples estimates of when the intention to move entered their awareness. This result is consistent with our observation that the belief manipulation affected early-RP amplitudes but not late-RP amplitudes. Indeed, other studies have reported that the W judgment is associated with late stages of motor preparation, rather than with the initial preparation for action (Haggard & Eimer, 1999; Haggard, Newman, & Magno, 1999). Previous research has revealed that weakening beliefs in self-efficacy, control, and agency influence complex behavior (Ajzen 2002; Bandura, 1982, 1989). In particular, self-efficacy beliefs that is, peoples beliefs about their capabilities to exercise control over events that affect their livesdetermine peoples level of effort: The stronger their belief in their capabilities, the stronger and more persistent are their efforts (Bandura, 1989). More recent research has tested whether inducing disbelief in free will also affects how people behave in social situations. These studies have demonstrated that disbelief in free will leads to antisocial behavior (Baumeister et al., 2009; Vohs & Schooler, 2008). The current study extends these findings by showing that manipulating beliefs in free will also influences low-level brain processes associated with the preparation of intentional action. Furthermore, our study shows that disbelief in free will affects intentional action at a preconscious level of motor preparation, because the effect of our belief manipulation was evident more than 1 s before participants consciously decided to act (Fig. 3). Currently, we can only speculate about the precise mechanisms that mediate the influence of high-level beliefs on basic motor preparation. Reading deterministic messages may induce participants to modify their high-level beliefs on free will because they were swayed by a rational argument. Alternatively, people may be emotionally impressed by the rhetoric or by the message conveyed by the text. In any case, undermining the idea that we are the masters of our own houses (der Herr im eigenen Haus; Freud, 1917, p. 11) presumably reduces the intentional effort we put into action. How disbelief in free will affects intentional effort is an open question. One likely possibility is that this effect is mediated by self-efficacy and perceived control (Ajzen, 2002; Bandura, 1982, 1989). However, it is crucial to note that selfefficacy and control beliefs refer to peoples beliefs about the capability to control their own lives, but the anti-free-will manipulation concerned a more metaphysical and deterministic idea of control (i.e., people are not free because their genes and the environment determine their behavior). It is important to note that the influence of high-level beliefs on basic motor processes might also help us understand why such beliefs lead to antisocial and irresponsible behavior (Baumeister et al., 2009; Vohs & Schooler, 2008). Putting less effort into an action might weaken our sense of agency for these actions and lead to a reduced feeling of responsibility. This reduced feeling of responsibility would very likely result in more careless and irresponsible behavior. The basic assumption

617 of this explanation is that disbelief in free will influences peoples sense of agency. It is interesting that authorship indicators (e.g., action-effect consistency) have been shown to affect selfreported agency (Ebert & Wegner, 2010). From this perspective, we could hypothesize that dismissing the idea that people can control their own actions acts as a nonauthorship indicator, thereby decreasing peoples sense of authorship. In sum, our results indicate that beliefs about free will can change brain processes related to a very basic motor level, and this suggests that abstract belief systems might have a much more fundamental effect than most people would expect. Declaration of Conflicting Interests
The authors declared that they had no conflicts of interest with respect to their authorship or the publication of this article.

Funding
This work was supported by the Bijzonder Onderzoeksfonds of Ghent University, Belgium, and the Centro Universitario Internazionale, Monte San Savino, Italy.

Supplemental Material
Additional supporting information may be found at http://pss.sagepub .com/content/by/supplemental-data

References
Ajzen, I. (2002). Perceived behavioral control, self-efficacy, locus of control, and the theory of planned behavior. Journal of Applied Social Psychology, 32, 665683. Bandura, A. (1982). Self-efficacy in human agency. American Psychologist, 37, 122147. Bandura, A. (1989). Human agency in social cognitive theory. American Psychologist, 44, 11751184. Banks, W.P., & Isham, E.A. (2009). We infer rather than perceive the moment we decided to act. Psychological Science, 20, 1721. Baumeister, R.F., Masicampo, E.J., & DeWall, C.N. (2009). Prosocial benefits of feeling free: Disbelief in free will increases aggression and reduces helpfulness. Personality and Social Psychology Bulletin, 35, 260268. Crick, F. (1994). The astonishing hypothesis: The scientific search for the soul. New York, NY: Touchstone. Ebert, J.P., & Wegner, D.M. (2010). Time warp: Authorship shapes the perceived timing of actions and events. Consciousness and Cognition, 19, 481489. Freud, S. (1917). Eine schwierigkeit der psychoanalyse: Gesammelte werke band (Vol. 12). Frankfurt am Main, Germany: Fischer Verlag. Gratton, G., Coles, M.G.H., & Donchin, E. (1983). A new method for off-line removal of ocular artifact. Electroencephalography & Clinical Neurophysiology, 55, 468484. Haggard, P., & Eimer, M. (1999). On the relation between brain potentials and the awareness of voluntary movements. Experimental Brain Research, 126, 128133. Haggard, P., Newman, C., & Magno, E. (1999). On the perceived time of voluntary actions. British Journal of Psychology, 90, 291303.

618
Hallett, M. (2007). Volitional control of movement: The physiology of free will. Clinical Neurophysiology, 118, 11791192. Keller, I., & Heckhausen, H. (1990). Readiness potentials preceding spontaneous motor acts: Voluntary vs. involuntary control. Electroencephalography & Clinical Neurophysiology, 76, 351361. Kornhuber, H.H. (1984). Attention, readiness for action, and the stages of voluntary decision: Some electrophysiological correlates in man. Experimental Brain Research Supplement, 9, 420429. Kornhuber, H.H., & Deecke, L. (1965). Hirnpotentialnderungen bei Willkurbewegungen und passive Bewegungen des Menschen: Bereitschaftspotential und reafferente Potentiale. Pflgers Archiv fr die Gesamte Physiologie des Menschen und der Tiere, 284, 117. Lang, W. (2003). Surface recordings of the Bereitschaftspotential in normals. In M. Jahanshahi & M. Hallet (Eds.), The Bereitschaftspotential: Movement-related cortical potentials (pp. 1934). New York, NY: Kluwer Academic/Plenum. Leotti, L.A., Iyengar, S.S., & Ochsner, K.N. (2010). Born to choose: The origins and value of the need for control. Trends in Cognitive Sciences, 14, 457463. Libet, B., Gleason, C.A., Wright, E.W., & Pearl, D.K. (1983). Time of conscious intention to act in relation to onset of cerebral activity (readiness-potential): The unconscious initiation of a freely voluntary act. Brain, 106, 623642. Libet, B., Wright, E.W., & Gleason, C.A. (1982). Readiness potentials preceding unrestricted spontaneous pre-planned voluntary acts. Electroencephalography & Clinical Neurophysiology, 54, 322325.

Rigoni et al.
Oldfield, R.C. (1997). The assessment and analysis of handedness: The Edinburgh inventory. Neuropsychologia, 9, 97113. Rakos, R.F., Laurene, K.R., Skala, S., & Slane, S. (2008). Belief in free will: Measurement and conceptualization innovations. Behavior and Social Issues, 17, 2039. Rigoni, D., Brass, M., & Sartori, G. (2010). Post-action determinants of the reported time of conscious intentions. Frontiers in Human Neuroscience, 4, Article 38. Retrieved May 14, 2010, from http://www.frontiersin.org/human_neuroscience/10.3389/ fnhum.2010.00038/full Shibasaki, H., & Hallett, M. (2006). What is the Bereitschaftspotential? Clinical Neurophysiology, 117, 23412356. Soon, C.S., Brass, M., Heinze, H.-J., & Haynes, J.-D. (2008). Unconscious determinants of free decisions in the human brain. Nature Neuroscience, 11, 543545. Tangney, J.P., Baumeister, R.F., & Boone, A.L. (2004). High selfcontrol predicts good adjustment, less pathology, better grades, and interpersonal success. Journal of Personality, 72, 271322. Tatman, A.W., Swogger, M.T., Love, K., & Cook, M.D. (2009). Psychometric properties of the Marlowe-Crowne Social Desirability Scale with adult male sexual offenders. Sex Abuse: A Journal of Research and Treatment, 21, 2134. Vohs, K.D., & Schooler, J.W. (2008). The value of believing in free will: Encouraging a belief in determinism increases cheating. Psychological Science, 19, 4954. Wegner, D. (2002). The illusion of conscious will. Cambridge, MA: MIT Press.