You are on page 1of 9
 
Power Changes How the Brain Responds to Others
Jeremy Hogeveen
Wilfrid Laurier University
Michael Inzlicht
University of Toronto
Sukhvinder S. Obhi
Wilfrid Laurier University
Power dynamics are a ubiquitous feature of human social life, yet little is known about how power isimplemented in the brain. Motor resonance is the activation of similar brain networks when acting andwhen watching someone else act, and is thought to be implemented, in part, by the human mirror system.We investigated the effects of power on motor resonance during an action observation task. Separategroups of participants underwent a high-, neutral, or low-power induction priming procedure, prior toobserving the actions of another person. During observation, motor resonance was determined withtranscranial magnetic stimulation (TMS) via measures of motor cortical output. High-power participantsdemonstrated lower levels of resonance than low-power participants, suggesting reduced mirroring of other people in those with power. These differences suggest that decreased motor resonance to others’actions might be one of the neural mechanisms underlying power-induced asymmetries in processing oursocial interaction partners.
Keywords:
 power, motor resonance, human mirror system, TMS, social cognitive neuroscience
The profound evolution of primate neocortex was influenced bythe computational demands of living in a complex social environ-ment (Dunbar & Shultz, 2007). For primates, a key factor creatingstructure within the social environment is power. In nonhumanprimates, an animal’s power is partly determined by the degree towhich they dominate conspecifics. Those that are able to exertdominance over others gain greater access to valuable resourceslike food and potential mates (Dunbar, 1980; Lewis, 2002; Watts, 2010). In human societies, power similarly creates “dependenceasymmetries,” wherein the powerless depend heavily on the pow-erful for resources, whereas the powerful enjoy relatively unabatedaccess to resources (Russell & Fiske, 2010). This asymmetryresults in differences in how the powerful and the powerlessprocess other individuals. Despite what we know about the effectsof power on social information processing, the majority of theevidence is indirect, and the mechanisms underlying power’s in-fluence remain a mystery. To begin to address this issue, we usedtranscranial magnetic stimulation (TMS) to provide a direct andonline measure of power’s impact on how the brain responds toobserved action.
The Psychological Impact of Power
The psychological literature on power indicates a reliable rela-tionship between power and information processing style (Ames,Rose, & Anderson, 2006; Fiske, 1993; Fiske & Dépret, 1996; Guinote, 2007a, 2007b; Obhi, Swiderski, & Brubacher, 2012; Smith & Trope, 2006; van Kleef et al., 2008). High-power indi- viduals are able to ignore peripheral information and focus on task relevant details, thereby improving goal pursuit (Guinote, 2007a,2007b), cognitive flexibility (Smith & Trope, 2006), and executive functioning (Smith, Jostmann, Galinsky, & van Dijk, 2008).Therefore, when powerful individuals ignore peripheral informa-tion during a nonsocial task, it may improve their performance.Conversely, when the powerful ignore “peripheral” information insocial settings, the outcome can be quite negative from the per-spective of the powerless.The powerful, because they already control resources, tend notto process individuating information about the less powerful. Incontrast, the powerless, because they do not control resources,
 are
motivated to process individuating information about the powerful(Fiske & Dépret, 1996; Goodwin, Gubin, Fiske, & Yzerbyt, 2000). Power-driven differences in the processing of others are alsoevident in the inability of high-power-primed participants to takethe visual, cognitive, and emotional perspectives of others, relativeto participants who feel relatively powerless (Anderson, Keltner,
 Editor’s Note.
 Mauricio Delgado served as the action editor for thisarticle.—IGJeremy Hogeveen, Centre for Cognitive Neuroscience and Departmentof Psychology, Wilfrid Laurier University, Waterloo, Ontario, Canada;Michael Inzlicht, Department of Psychology, University of Toronto Scar-borough, Toronto, Ontario, Canada; Sukhvinder S. Obhi, Centre for Cog-nitive Neuroscience and Department of Psychology, Wilfrid Laurier Uni-versity.This research was made possible by research grants from the NaturalScience and Engineering Research Council and the Social Science andHumanities Research Council (SSHRC), held by Sukhvinder S. Obhi, andan SSHRC Canada Graduate Scholarship awarded to Jeremy Hogeveen.Correspondence concerning this article should be addressed to JeremyHogeveen, Department of Psychology, Wilfrid Laurier University, 75University Avenue West, Waterloo ON N2L 3C5, Canada. E-mail: jhogeveen@wlu.ca
     T     h     i   s     d   o   c   u   m   e   n    t     i   s   c   o   p   y   r     i   g     h    t   e     d     b   y    t     h   e     A   m   e   r     i   c   a   n     P   s   y   c     h   o     l   o   g     i   c   a     l     A   s   s   o   c     i   a    t     i   o   n   o   r   o   n   e   o     f     i    t   s   a     l     l     i   e     d   p   u     b     l     i   s     h   e   r   s .     T     h     i   s   a   r    t     i   c     l   e     i   s     i   n    t   e   n     d   e     d   s   o     l   e     l   y     f   o   r    t     h   e   p   e   r   s   o   n   a     l   u   s   e   o     f    t     h   e     i   n     d     i   v     i     d   u   a     l   u   s   e   r   a   n     d     i   s   n   o    t    t   o     b   e     d     i   s   s   e   m     i   n   a    t   e     d     b   r   o   a     d     l   y .
Journal of Experimental Psychology: General © 2013 American Psychological Association2013, Vol. 142, No. 3, 000 0096-3445/13/$12.00 DOI: 10.1037/a0033477
1
AQ:au
tapraid5/zfr-xge/zfr-xge/zfr00313/zfr2440d13z
 xppws S
1 6/6/13 6:39 Art: 2012-1037
 APA NLM
 
& John, 2003; Galinsky, Magee, Inesi, & Gruenfeld, 2006). Sim- ilarly, socioeconomic status (SES) has been linked to empathicaccuracy, with high-SES individuals making less accurate judg-ments about others’ affective states than low-SES individuals(Kraus, Côté, & Keltner, 2010). As a result, the powerful oftenform a relatively shallow understanding of others, compared to theless powerless.Despite the strong evidence that high power leads to reducedprocessing of others’ actions and emotions, there are conflictingfindings in the literature. For example, Schmid Mast, Jonas, andHall (2009) found that high power actually improves empathicaccuracy, and Côté et al. (2011) have also shown that high power,combined with a prosocial orientation, leads to improved empathicability. Therefore, the relationship between power and the degreeto which people process their social interaction partners is notstraightforward. In the present investigation, we begin to addressthis ambiguity using a direct index of the degree to which peopleprocess others’ actions.
The Neural Representation of Observed Actions
In recent years, researchers have shown that the human brain isexquisitely tuned to the perceptual and cognitive demands of processing others (Hari & Kujala, 2009). One reliable finding fromthis work that appears to be important for human social perceptionis resonant or vicarious activity, whereby perceiving an interactionpartner automatically activates neural circuits that would underlietheir experience (Keysers & Gazzola, 2009). For example, withrespect to action observation, neural circuits that are related toaction execution become active when the person observes some-one else making the same action; in other words, the observer’sbrain resonates with the model’s motor behavior (Hogeveen &Obhi, 2011; Iacoboni, 2009; Oberman & Ramachandran, 2007; Obhi & Hogeveen, 2010; Rizzolatti & Sinigaglia, 2010). We refer to the network of brain regions involved in this process as themotor resonance system (cf. Hogeveen & Obhi, 2012). Motorresonance includes the human parietofrontal mirror system, andmany believe that resonance reflects mirror system activity (Fa-diga, Craighero, & Olivier, 2005; Fadiga, Fogassi, Pavesi, &Rizzolatti, 1995).A reliable index of resonance is the amplitude of motor-evokedpotentials (MEPs) recorded from a specific muscle via electro-myography (EMG), while a person observes another person acting.An MEP is elicited by applying a single, fixed intensity TMS pulseover an area of the motor cortex that corresponds to a muscleunderlying the observed action. For a given intensity of stimula-tion, changes in MEP amplitude reflect changes in the excitabilityof motor cortical representations (see Figure 1; for a review, seeFadiga et al., 2005).
Power and Motor Resonance: The Present Study
Researchers suggest that motor resonance provides a scaffoldfor understanding the actions of our interaction partners (cf. Brass,Ruby, & Spengler, 2009; Decety & Sommerville, 2009; Grafton, 2009; Spunt & Lieberman, 2012), and those actions are frequently less important to those with high-power status (Fiske, 1993; Fiske & Dépret, 1996; Goodwin et al., 2000; Russell & Fiske, 2010). Yet, previous investigations of power and the processing of others’actions and emotions have yielded conflicting results, sometimessuggesting an increase (e.g., Côté et al., 2011), and elsewhere adecrease (e.g., Galinsky et al., 2006), in interpersonal sensitivity.In the present study, we examine whether power can increase ordecrease interpersonal sensitivity by examining the effects of power priming on motor resonance.The present study had participants write an essay documentinga high-, neutral, or low-power experience, and then used a directand online technique to index motor resonance during a passiveobservation task. The power priming procedure—recalling a mem-ory with or without power—is a well-established technique thathas demonstrated a wide range of downstream effects, with thehigh-power condition often found to decrease interpersonal sensi-tivity relative to low-power priming (Galinsky, Magee, Gruenfeld,Whitson, & Liljenquist, 2008; Galinsky et al., 2006). To the extent that resonance is an automatic response when observing the ac-tions of others, any changes in resonance as a function of powercan be construed as a “default” effect of power on the brain. Again,as previous researchers have suggested, it is reasonable to expectthat such differences in resonance may contribute to the differ-ences in how high- and low-power individuals process other peo-ple. Specifically, given the balance of the literature suggests thatpeople in positions of power tend to act in a self-interested mannerand display reduced interpersonal sensitivity to their powerlesscounterparts (Fiske, 1993; Galinsky, Gruenfeld, & Magee, 2003; Galinsky et al., 2008, 2006; Keltner, Gruenfeld, & Anderson, 2003; Russell & Fiske, 2010), we hypothesized that high-power
Figure 1.
 Schematic of the experiment. TMS
transcranial magnetic stimulation.
     T     h     i   s     d   o   c   u   m   e   n    t     i   s   c   o   p   y   r     i   g     h    t   e     d     b   y    t     h   e     A   m   e   r     i   c   a   n     P   s   y   c     h   o     l   o   g     i   c   a     l     A   s   s   o   c     i   a    t     i   o   n   o   r   o   n   e   o     f     i    t   s   a     l     l     i   e     d   p   u     b     l     i   s     h   e   r   s .     T     h     i   s   a   r    t     i   c     l   e     i   s     i   n    t   e   n     d   e     d   s   o     l   e     l   y     f   o   r    t     h   e   p   e   r   s   o   n   a     l   u   s   e   o     f    t     h   e     i   n     d     i   v     i     d   u   a     l   u   s   e   r   a   n     d     i   s   n   o    t    t   o     b   e     d     i   s   s   e   m     i   n   a    t   e     d     b   r   o   a     d     l   y .
2
 HOGEVEEN, INZLICHT, AND OBHI
F1
tapraid5/zfr-xge/zfr-xge/zfr00313/zfr2440d13z
 xppws S
1 6/6/13 6:39 Art: 2012-1037
 APA NLM
 
priming would decrease resonance, whereas low-power primingwould increase it.
MethodParticipants
Forty-five participants took part in the experiment for financialremuneration or partial course credit. On the basis of previousbetween-group MEP studies, it was determined that 36 participantswould be sufficient to achieve statistical power of 80% (
1.19;Fitzgibbon et al., 2012; Fourkas, Bonavolonta`, Avenanti, & Agli-oti, 2008). However, since we were combining the MEP measurewith a priming technique that typically demonstrates slightlysmaller effects (
0.80; Galinsky et al., 2008, 2006), we ran nine additional participants to attain sufficient statistical power. Partic-ipants were assigned to the high-power (
n
18; 11 female, 7 male;
 M 
20.59 years,
 SD
2.12), low-power (
n
17; 13 female, 4male;
 M 
20.65 years,
 SD
2.12), or neutral (
n
10; 8 female,2 male;
 M 
18.63 years,
 SD
0.81) condition. All participantshad normal or corrected-to-normal vision, and all but five partic-ipants were right-handed. Participants were screened for contrain-dications to TMS prior to participation, and informed consent wasobtained from all participants.
Apparatus and Stimuli
The TMS experiment was programmed with SuperLab (Version4; Cedrus Corporation, San Pedro, California) and run on a Delldesktop computer with stimuli displayed on a 20-in. (50.8-cm)LCD monitor. TMS was carried out with a Magstim Rapid
2
system(Magstim Company Ltd., Wales). EMG data was recorded with aBiopac psychophysiological recording system (Biopac SystemsInc., Goleta, California). MEPs were measured with pairs of 8-mmsurface electrodes placed in a belly-tendon arrangement over theabductor pollicis brevis muscle of participants’ right hand. TheEMG signal was acquired with a 1,000-Hz sampling rate, ampli-fied (to 5.0 mV), and filtered (bandpass 10–500 Hz), and sent toa laptop computer for offline analysis.Stimuli were videos depicting a right hand (palm facing down)squeezing a rubber ball between the thumb and index finger, suchthat the ball was substantially deformed (see Figure 1). The videosconsisted of a single squeeze repeated three to seven times. Videoediting was performed with Adobe Premiere Pro CS4 (AdobeSystems Inc., San Jose, California). All inferential statistical anal-ysis was performed with SPSS statistics (Version 17; SPSS Inc.,Chicago, Illinois).
Procedure
Participants entered the laboratory, were seated in front of thecomputer, and asked to write an essay. Participants in the low-power group wrote about an experience where someone had powerover them; those in the neutral group wrote about what happenedthe day before they came in for the study; and lastly, participantsin the high-power condition described an experience where theyhad power over someone else. In the literature, this essay writingprocedure for priming power has demonstrated a profound andfar-reaching impact on a variety of downstream behavioral mea-sures, affecting participants’ ability to recognize emotional facialexpressions (Galinsky et al., 2006), their tendency to conform toothers’ behavior (Galinsky et al., 2008), and even something asfundamental as their sense of agency (Obhi et al., 2012). There-fore, we reasoned that it would be an apt technique for activatinghigh, neutral, or low power prior to measuring motor resonanceduring a TMS/action observation session.For the TMS setup, participants were seated in a Brainsight Gen3 TMS chair (Rogue Research, Montréal, Canada). The experi-menter located the vertex and hand area of left primary motorcortex (M1) using a standard landmark technique (Hogeveen &Obhi, 2012). A Brainsight neuronavigation system (Rogue Re-search, Montréal, Canada) ensured stable coil positioning through-out the experiment (Lepage, Tremblay, & Théoret, 2010). Lastly,akin to several studies in the literature, stimulator output waslowered to determine the minimum intensity capable of elicitingvisible MEPs (
1 mV peak to peak) on more than 50% of TMSpulses, which was used as the protocol intensity rather than stim-ulating at 110%–120% of resting motor threshold (Enticott et al.,2012; Hogeveen & Obhi, 2012; Lepage et al., 2010). Stimulation intensity ranged from 49% to 75% (
 M 
 
 61%) of stimulatoroutput.In the first part of the TMS experiment, baseline corticospinalexcitability was determined by delivering 30 TMS pulses whileparticipants viewed a fixation cross. Next, participants began theaction observation block, which contained 75 total trials, contain-ing a fixation cross (2,000 ms), followed by videos of the hand-squeezing action (3,750–8,750 ms). TMS pulses were delivered atpoints of maximum squeeze intensity on 30 of the trials. Duringthe baseline block, participants counted the number of seconds thefixation was presented. Similarly, during the action observationblock, participants counted the number of squeezes contained ineach video. TMS pulses in both blocks occurred 3,128, 4,328,5,494, or 6,728 ms after trial onset. Therefore, the task andtemporal information during baseline and action observation werematched, making fixation cross versus action videos the onlydifference between the two blocks.
Results
For each participant, raw MEPs greater than 3 standard devia-tions from their mean were omitted from analysis. This resulted inthe removal of 1.38% of the data. Our main dependent measure of motor resonance was the change in MEP amplitude between thebaseline block and the action observation block—henceforth re-ferred to as MEP facilitation. For this measure, participants with amean change falling outside 2.5 standard deviations of the groupaverage for each experimental condition (high power, neutral, lowpower) were excluded. This procedure resulted in removal of oneparticipant in the high-power condition.
MEP Facilitation: Power Groups
Our main analysis was a one-way analysis of variance with onefactor at three levels (i.e., power: high, neutral, or low). Since wehypothesized that higher power would lead to a reduction in MEPfacilitation, we ran a linear contrast to test this prediction. Accord-ingly, we obtained a significant linear effect of power on MEPfacilitation,
 
(1, 42)
 
 5.44,
 p
 
 .03,
 
 
 0.72. Crucially,
     T     h     i   s     d   o   c   u   m   e   n    t     i   s   c   o   p   y   r     i   g     h    t   e     d     b   y    t     h   e     A   m   e   r     i   c   a   n     P   s   y   c     h   o     l   o   g     i   c   a     l     A   s   s   o   c     i   a    t     i   o   n   o   r   o   n   e   o     f     i    t   s   a     l     l     i   e     d   p   u     b     l     i   s     h   e   r   s .     T     h     i   s   a   r    t     i   c     l   e     i   s     i   n    t   e   n     d   e     d   s   o     l   e     l   y     f   o   r    t     h   e   p   e   r   s   o   n   a     l   u   s   e   o     f    t     h   e     i   n     d     i   v     i     d   u   a     l   u   s   e   r   a   n     d     i   s   n   o    t    t   o     b   e     d     i   s   s   e   m     i   n   a    t   e     d     b   r   o   a     d     l   y .
3
POWER AND MOTOR RESONANCE
tapraid5/zfr-xge/zfr-xge/zfr00313/zfr2440d13z
 xppws S
1 6/6/13 6:39 Art: 2012-1037
 APA NLM

Reward Your Curiosity

Everything you want to read.
Anytime. Anywhere. Any device.
No Commitment. Cancel anytime.
576648e32a3d8b82ca71961b7a986505