Free Will and Punishment: A Mechanistic View of Human Nature Reduces Retribution

Author(s): Azim F. Shariff, Joshua D. Greene, Johan C. Karremans, Jamie B. Luguri,

Cory J. Clark, Jonathan W. Schooler, Roy F. Baumeister and Kathleen D. Vohs
Source: Psychological Science , AUGUST 2014, Vol. 25, No. 8 (AUGUST 2014), pp. 1563-
1570
Published by: Sage Publications, Inc. on behalf of the Association for Psychological
Science

Stable URL: https://www.jstor.org/stable/24543587

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Sage Publications, Inc. and Association for Psychological Science are collaborating with JSTOR
to digitize, preserve and extend access to Psychological Science

This content downloaded from

103.19.199.29 on Sat, 17 Oct 2020 06:54:51 UTC
All use subject to https://about.jstor.org/terms
 J ASSOCIATION FOR
Research Article psychological science

Psychological Science
2014, Vol. 25(8) 1563-1570
Free Will and Punishment: A Mechanistic ) The Author(s) 2014
Reprints and permissions:
View of Human Nature Reduces sagepub.com/journalsPermissions. na ν
DOI: 10.1177/0956797614534693

Retribution pss. sagepub. com

®SAGE

Azim F. Shariff1, Joshua D. Greene2, Johan C. Karremans3,

Jamie B. Luguri4, Cory J. Clark5, Jonathan W. Schooler6,
Roy F. Baumeister7, and Kathleen D. Vohs8
'Department of Psychology, University of Oregon; department of Psychology, Harvard University;
'Department of Social and Cultural Psychology, Behavioral Science Institute, Radboud
University Nijmegen; department of Psychology, Yale University; 5School of Social Ecology,
University of California, Irvine; department of Psychology, University of California, Santa Barbara;
'Department of Psychology, Florida State University; and 8Carlson School of Management,
University of Minnesota

Abstract

If free-will beliefs support attributions of moral responsibility, then reducing these beliefs should make
less retributive in their attitudes about punishment. Four studies tested this prediction using both meas
manipulated free-will beliefs. Study 1 found that people with weaker free-will beliefs endorsed less retribu
not consequentialist, attitudes regarding punishment of criminals. Subsequent studies showed that learni
the neural bases of human behavior, through either lab-based manipulations or attendance at an undergr
neuroscience course, reduced people's support for retributive punishment (Studies 2-4). These results illustr
exposure to debates about free will and to scientific research on the neural basis of behavior may have con
for attributions of moral responsibility.

Keywords
free will, punishment, morality, responsibility, blame, open materials

Received 3/14/13; Revision accepted 4/14/14

Most people believe that humans have free will (Nahmias, supporting the mechanistic causes of human be
Morris, Nadelhoffer, & Turner, 2005). However, long- accumulates (Greene & Cohen, 2004; Wegner,
standing controversies remain as to what forms of free anti-free-will viewpoints reverberate beyond ac
will can actually exist alongside the known laws of nature and legal arguments and popular-press articles
(see Baer, Kaufman, & Baumeister, 2008). Although few ever-more mechanical causes for human be
people deny that humans regularly make uncoerced (Greene & Cohen, 2004; Nahmias, 2011; Wolf
choices and exercise self-control, many scientists and questions about the psychological and socie
philosophers have taken issue with the idea that con- quences of reduced free-will beliefs have e
scious humans can generate spontaneous choices and (Schooler, 2010; Shariff, Schooler, & Vohs, 2008
actions not fully determined by prior events (e.g., Bargh, Free-will beliefs underlie perceptions of mora
2008). According to this challenge, free will of this kind sibility (Eshleman, 2004; Nahmias et al., 2005). L
appears to be incompatible with a scientific understand- moral decisions often rest on whether one shou
ing of the world as a mechanical system in which all
events are fully determined by prior or random events. corresponding Author:
Many researchers have argued that this form of free will Azim F. Shariff, 1227 University of Oregon, Eugene, OR 974
is an illusion that grows less believable as research E-mail: shariff@uoregon.edu

This content downloaded from

103.19.199.29 on Sat, 17 Oct 2020 06:54:51 UTC
All use subject to https://about.jstor.org/terms
 1564 Shariff et al.

acted differently, which p

acted differently. When g
sible, condemnation is less
nomena, such as viruses a
morally responsible for t
they are not perceived as
The rejection of free wil
undermine attributions of
and for others, renderin
natural phenomena. effect, it is not threatened by mechanistic arguments
As a consequence, diminishing people's belief in free against free will. In contrast, retributivism
will may likewise weaken their belief in moral responsi- transgressors being blameworthy and deser
bility, and potentially license them to transgress. Indeed, because of their transgressions, so anyt
prior research found that participants whose free-will ishes that sense of deservingness—such as
beliefs were experimentally diminished were less helpful have freely chosen not to commit the tr
and more likely to lie, cheat, steal, and act aggressively diminishes the justification for retributiv
than were participants in control conditions (Baumeister, Though these are theoretical connection
Masicampo, & DeWall, 2009; Vohs & Schooler, 2008). dence to suggest that people do base th
Such antisocial outcomes lend support to concerns about sion making, in part, on their philosophi
the negative social consequences that could follow a (e.g., Rai & Holyoak, 2013; Vohs & Scho
broad erosion of free-will beliefs. Therefore, we predicted that people's free-will beliefs
In the current work, we considered the opposite side should be related primarily to their support for retribut
of that coin. We tested whether reduced belief in free will punishment, and further, that these beliefs are causal
would lead people to see others' bad behavior as less tors. Diminished beliefs in free will should dimini
morally reprehensible, resulting in less retributive blame, which should, in turn, diminish the endorsem
punishment. of retributive punishment.

Free Will, Retributivism, and The Present Research

Consequentialism In Study 1, we tested the relationship between free-w
Humans respond to transgressions with an urge to exact beliefs and attitudes toward consequent
punitive costs on the transgressor (e.g., Buckholtz et al., tivist punishment. We then moved to
2008; Nelissen & Zeelenberg, 2009; Smith, 1759). One ods, testing whether punitive attitudes
theory of punishment, labeled retributivism, holds that diverse manipulations aimed at diminis
punitive urges reflect normative moral principles based beliefs. In Study 2, we had participants
on universal norms of reciprocity and fairness: People lenges to free will, whereas in Studi
who harm others should be harmed themselves. In this pants were exposed to research a
view, reciprocity is moral justification for punishment, mechanisms underlying human ac
independent of any other benefits the punishment may measures in these experiments were re
bring. Retributivism, in other words, holds that the point tudes toward criminality. We predic
of punishment is to extract suffering from the transgres- belief in free will would predict we
sor as "just deserts." Retributivism is contrasted with con- punishment,
sequentialist justifications for punishment, which depend

on the utilitarian benefits of punishment (e.g., rehabilita- Study 1; Correlations Betw

tion or deterrence). According to consequentialism,
will Beliefs and Retribution
selection of punishment should be based on what has
the best social consequences, regardless of how much or Study 1 tested how free-will beliefs relate to
how little the transgressor suffers (if at all), or indeed about both retributive and consequentialist pu
whether the transgressor even deserves punishment. Given that retributive punishment springs
Research has provided some evidence for a correla- blame (Greene & Cohen, 2004) whereas conse
tion between beliefs about punishment and beliefs in punishment is intended to benefit society, we pr
free will. For instance, Krueger, Hoffman, Walter, and that support for retributive punishment woul
Grafman (in press) found that people with greater belief tively related to free-will beliefs, whereas su
in free will also tended to be more punitive, though only consequentialist punishment would be unrela

This content downloaded from

103.19.199.29 on Sat, 17 Oct 2020 06:54:51 UTC
All use subject to https://about.jstor.org/terms
 Free-Will Beliefs and Punishment 1565

Table 1. Results of Regression Analyses Predicting Support for R

Predictor
Predictor Retributive
Retributive Consequentialist
Consequentialist
punishment punishment punishment punis

Free-will
Free-willbelief (no controls)
belief (no controls) 0.242**
0.242** -0.024

Free-will belief (controlling for the variables below) 0.164* 0.018

Age 0.088 0.156*

Gender (male) 0.050 0.045
Education -0.139* 0.056
Religiosity -0.158* 0.034
Political ideology (social) 0.323** -0.215*

Political ideology (economic) 0.003* -0.013

Note: The table reports standardized regression coefficients. Higher values for political ideology indicate a more conservative position.
*p < .05. **£> < .001.

Method Study 1 supported the hypothesis that free-will beliefs

positively predict punitive attitudes, and in
Two hundred forty-four Americans (147 female; mean retri5utivist attitudes Th
age = 36.81 years) participated online via Amazon's third.variable explanation
Mechanical Turk. Respondents completed the seven-item correlational desig
Free Will subscale of the Free Will and Determinism Plus whether a reduction o
scale (FAD+; Paulhus & Carey, 2011; α = .88. Items (e.g., resultant reduction in re
"People have complete control over the decisionswe moved they 2 thrQugh 4>methods t
to experimental
make") were rated from 1 (strongly disagree) to 5 {strongly the causal relationship
agree). FAD+ scores have been shown to be moderately poft for retributive pun
positively associated with religiosity, belief in a just world, beliefs directly
and locus of control (Paulhus & Carey, 2011).
In order to measure attitudes toward retributivist and
consequentialist motivations for punishment, we had Study 2: Manipulated Free-Will Beliefs
participants read descriptions of retributivism and conse- and Retributivism
quentialism as motivations for punishment and then indi- Method
cate on two separate Likert scales (1 = strongly disagree,
7 = strongly agree) how important retributivism and con- Participants. Forty-six students (30 female; mean
sequentialism should be in determining motivations for age = 20.44 years) participated for partial course credit,
criminal punishment. Participants also completed ques
tions about their education, religiousness, political ideol- Procedure. Told that they were taking part in unrelated
ogy (including separate ratings for economic and social experiments, participants first completed a free-will
issues), and demographics (for the materials used in manipulation. They were randomly assigned to read one
Study 1, see Methodological Details and Materials in the of two passages from Crick's (1994) The Astonishing
Supplemental Material available online). Hypothesis. In the anti-free-will condition, the passage
rejected free will and advocated a mechanistic view of
Results human behavior. In the neutral condition, the passage
was unrelated to free will. This task has been previously
Zero-order correlations indicated that the retributivism validated by Vohs and Schooler (2008), and test
and consequentialism scales were moderately negatively manipulation in an independent sample revealed
correlated, K243) = -.36, p < .001. passages led to the expected differences in free-will
In order to test for relationships with free-will beliefs, beliefs, as measured by a single item (see Methodological
we regressed retributivism and consequentialism scores Details and Materials in the Supplemental Material),
on Free Will subscale scores. As predicted, stronger belief t{205) = 2.55, p = .01.
in free will predicted greater support for retributive pun- Next, participants read a fictional vignette involving an
ishment, β = 0.24, p < .001, but was not predictive of offender who beat a man to death. Acting as hypothetical
support for consequentialist punishment, β = -0.02, p = jurors, participants recommended the length of the prison
.72. Effects remained significant when we statistically sentence (if any) that this offender should serve follow
controlled for age, gender, education, religiosity, and ing a 2-year, nearly 100%-effective, rehabilitation treat
economic and social political ideology (Table 1). ment. The notion that the offender had been rehabilitated

This content downloaded from

103.19.199.29 on ffff:ffff on Thu, 01 Jan 1976 12:34:56 UTC
All use subject to https://about.jstor.org/terms
 1566 Shariff et al.

θ3
TD
Œ
CD
Ε
Ε
ο
ο
CD
OC

=3
Ο

Neutral Anti-Free-Will Neutral Neuroscience

Passage Passage Articles Articles
Study 2 Study 3

Fig. 1. Mean recommended prison sentence for a hypothet

free-will conditions) and Study 3 (neutral and neuroscience

was used in order to isolate participants' desire for pu

ishment as retribution.1 The passage further focused
ticipants on retributive, rather than consequential
punishment by noting that the prosecution and defen
had agreed that the rehabilitation would prevent recid
vism and that any further detention after rehabilitati
would offer no addition deterrence of other poten
criminals. order to directly test the hypothesis that beliefs about
Participants then chose among seven punishment free will undergird judgment
options: treatment time only, with no imprisonment (1);
2 years of imprisonment posttreatment (2); 5 years of
Study 3: Systematic Exposure to
imprisonment posttreatment (3); 10 years of imprison
Neuroscience and Retributivism
ment posttreatment (4); 25 years of imprisonment post
treatment, with the chance of parole after 15 years (5); 25 in Study 2, we diminished participants' f
years of imprisonment posttreatment, with no chance of by having them read a quite opinionated e
parole (6); and life imprisonment, with no chance of a Nobel laureate (Crick, 1994). In Stud
parole (7). Finally, participants completed a demograph- subtler manipulation that may represent h
ics questionnaire and suspicion probe (for the materials free-will beliefs are more commonly chall
used in this experiment, see Methodological Details and pants read articles made to look like they w
Materials in the Supplemental Material). science magazine articles on findings
neuroscientific research. These articles did, in fact,
Results describe actual neuroscience findings implying that
human behavior is caused mechanistically. Crucially, the
As predicted, participants who read the anti-free-will pas- passages made no mention of free will, thereby allowing
sage recommended significantly lighter prison sentences participants to draw their own conclusions about its
than participants who read the neutral passage (M = 2.91, relevance.
SD = 1.08, vs. M = 3.96, SD = 1.49), *(44) = 2.71, p < .05, We hypothesized that relative to exposure to scientific
Cohen's d = 0.82 (Fig. 1). Specifically, participants whose views on other topics, exposure to research implying a
free-will beliefs had been experimentally diminished mechanistic view of human action would reduce belief in
recommended roughly half the length of imprisonment free will, and thereby retributive tendencies in sentencing

This content downloaded from

103.19.199.29 on Sat, 17 Oct 2020 06:54:51 UTC
All use subject to https://about.jstor.org/terms
 Free-Will Beliefs and Punishment 1567

Blameworthiness Positive and Negative Affect Schedule (PANAS; Watson,

Blameworthiness
Clark, & Tellegen, 1988) so that we could account for
mood effects that might emerge from challenges to intu
itions about the mechanisms behind human action.
-0.55* (a) -0.50** (b)
Finally, participants completed a demographics question
naire and suspicion probe. (For the materials used in this
Exposure to -0.72* (c)
Punishment experiment, see Methodological Details and Materials in
Neuroscience -0.45 (C)
the Supplemental Material.)

Fig. 2. Effect of exposure to neuroscientific research on punishment

recommendations
recommendations as as mediated
mediated by perceived
by perceived blameworthiness
blameworthiness of the of the Results
offender in Study 3. Standardized regression coefficients are shown along
thepaths;
the paths;c' c' represents
represents the mediated
the mediated effect
effect of of condition
condition on punishment Mood. Neither positive nor negative mood differed sig
on punishment
recommendations,
recommendations, taking intointo
taking account the indirect
account effect of
the indirect blameworthi
effect of blameworthi- nificantly by Condition (pS > .4). Higher levels of positive
ness. Asterisks indicate significant effects (*p < .05, **p < .001).
ness. Asterisks indicate significant effects (*p < .05, **p < .001). affect corresponded to lower attributions of blame,
K86) = -.34, p = .002. No other relationships involving
mood were found.
recommendations. Moreover, we predicted that exposure
to mechanistic neuroscience would reduce the extent to Manipulation check. Results for the manipulation
which a transgressor was perceived as blameworthy, and check were consistent with the independent test of the
that this assessment would mediate the effect of condi- manipulation; participants who read the neuroscience
tion on sentencing recommendations. articles reported marginally lower belief in free will com
pared with participants who read the neutral articles,
Method Ki, 87) = 3.46, .07.
Participants. Ninety-one undergraduates participated Pu
for partial course credit. Four participants were excluded parti
(3 for suspicion, 1 for admitted intoxication), which left a mend
final sample of 88 participants (6l female; mean age = parti
20.81 years). 3-10, SD = 1.48, vs. M= 3.83, SD = 1.77), K86) = 2.09, p =
.04, d = 0.45 (Fig. 1) and blamed the transgressor less
Procedure. Told that they were taking part in unrelated (M = 5.48, SD = 1.17, vs. M = 6.03,
experiments, participants first read two popular-science 2.40, p = .02, d = 0.52. Recommended pu
articles made to look as if they were from the Scientific perceived blameworthiness were signific
American and New Scientist Web sites. Participants in the K88) = .37, p < .001. Controlling for m
neuroscience condition read articles on brain-imaging edly change the main results.2
studies that showed dissociations between motor actions
and people's perception of conscious intention (Osborne, Mediation. We predicted that reading the neuroscience
2003; Soon, Brass, Heinze, & Haynes, 2008). Participants articles would decrease participants' perception that the
in the neutral condition read articles on nuclear power offender was morally blameworthy, and that this effect
and natural headache remedies. In neither condition did on perceived blameworthiness would account for the
the material mention free will, morality, or responsibility. lighter sentencing recommendations in the neuroscience
As in Study 2, pretesting with an independent sample condition than in the neutral condition. Consistent with
revealed that the neuroscience passages led to lower predictions, bootstrapping analyses (10,000 resamples;
free-will beliefs than did the neutral science passages, Preacher & Hayes, 2004) indicated that perceived blame
/(198) = 2.35, p = .02. This confirms that relative to other worthiness mediated the effect of condition on sentenc
scientific descriptions, descriptions of scientific research ing recommendations (95% confidence interval = [-0.68,
promoting a mechanistic view of human behavior dimin- -0.03]; Fig. 2).
ish belief in free will. Study 3 bolstered the results of Study 2 in several
For the ostensible second study, participants read and ways. First, it replicated the finding that
responded to the vignette from Study 2. In addition to manipulations that decrease free-will belief
recommending a prison sentence, they rated the offend- retributive punishment. Second, the mediat
er's blameworthiness (1 = not at all deserving of blame, perceived blameworthiness made a strong c
7 = completely deserving of blame). Participants then role of moral responsibility in the effect of
completed the FAD+ as a manipulation check and the free-will belief on retribution. Third, St

This content downloaded from

103.19.199.29 on Sat, 17 Oct 2020 06:54:51 UTC
All use subject to https://about.jstor.org/terms
 1568 Shariff et al.

manipulation that did no

instead influenced them
scientific findings sugg
mechanistic. Were changes in retributivism related to students' learn
ing outcomes? In neither class did grades predict
in sentence recommendations. However, for student
Study 4: Neuroscience Education and
neuroscience course, decrease in the length of s
Retribution
recommendations was strongly correlated with incr
In Study 4, we delved further into the issue of naturalistic students' self-reported know
manipulations of free-will beliefs. Instead of using a labo- first to the last class Ο = -.45,
ratory-based manipulation, we employed a naturalistic course, there was no correla
method by which people learn about mechanistic causes changes in brain knowledge an
underlying human action: university neuroscience classes. tions (r = .02, p = .90). The st
Participants in an introductory cognitive-neuroscience differed significantly between th
class judged the appropriate punishment for a hypotheti- p = 05), which suggests that th
cal criminal. They did so on both the first and last days of directly corresponded to what
class, 10 weeks apart. We predicted that learning about learned in the neuroscience clas
the brain would reduce support for retributive punish
ments, a result that would conceptually replicate the prior General Discussion
studies. We further predicted that students' attitude change
across time would correspond to measures of learning in Convergent results across a co
the class. We also included students from a nonneurosci- experiments with diverse manip
ence class as a neutral comparison group. onstrated that shifting from a bel
mechanistic view of human behavior reduces supp
retributive punishment. Study 1 found that individu
Method
ences in free-will beliefs predict retributive, but no
jParticipants. Students from an introductory cognitive- quentialist, motivations for pun
neuroscience class in = 34) and an introductory geogra- experimentally diminishing fre
phy class (n = 36) participated in exchange for the chance free-will arguments diminishe
to win a raffle prize. Two students who were simultané- suggesting a causal relationship
ously taking both classes were dropped from analysis, exposure to neuroscience implyi
which left 68 participants (39 female; mean age = 20.44 human action—either reading p
years). taking an introductory neuroscience class in college—simi
larly produced a reduction in retributivism. (For
Procedure. On the first day of class, students completed table of all results, as well
a shortened version of the punishment scenario from not mentioned here,
Studies 2 and 3· They were also asked about their per- Materials in the Supplemen
ceived knowledge of the brain (relative to other students gest that shifts in people
from the university), age, sex, and current classes. Ten free-will beliefs, even th
weeks later, at the courses' final class, students completed brain, can affect people's
another survey with the same measures, followed by a ity, with potentially broa
question about their anticipated class standing. Instruc- Retributivism plays an im
tors for both classes were blind to the hypothesis. (For tern. Historically, much of
the materials used in this experiment, see Methodological ment has been an instit
Details and Materials in the Supplemental Material.) public's retributive desire
Stephen (1883) famously wrote that "the sent
law is to the moral sentiment of the public w
Results
to hot wax" (p. 423). In recent years, justice r
Supporting predictions, a paired-sample comparison and advocates have argued fo
showed that prison-sentence recommendations decreased restorative justice—a con
from the start to the end of the neuroscience class, from repairing the moral imb
3.41 to 2.91 on the 7-point scale, i(33) = 2.15,p= .04, d= (Braithwaite, 2002). Th
0.44. No change was found for students in the geography changing attitudes about
class (dis = 3.32 on the first day vs. 3.08 on the last day), be important to this evol

This content downloaded from

103.19.199.29 on Sat, 17 Oct 2020 06:54:51 UTC
All use subject to https://about.jstor.org/terms
Free-Will Beliefs and Punishment 1569

That mere exposure to modern neuroscience can be the mechanistic worldview espoused by many scientists,
sufficient to reduce retributivist motivations may be par and particularly psychologists, gains attention (e.g.,
ticularly relevant to court cases. The explicit existence of Gazzaniga, 2011; Monterosso & Schwartz, 2012; Nichols,
free will may be rarely debated in court, but neuroscien 2011), the impact of these trends—good, bad, or both—
tific evidence often is. Indeed, recent research showed calls for understanding.
that judges afforded shorter sentences to hypothetical
psychopathic criminals when the description of the crimi Author Contributions
nals' psychopathy included a biomechanical component, A. F. Shariff, J. D. Greene, J. C. Karremans, J. W. Schooler, and
compared with when it did not (Aspinwall, Brown, & K. D. Vohs developed the study concept. A. F. Shariff and J. B.
Tabery, 2012). Our findings likewise suggest that merely Luguri conducted the studies. A. F. Shariff analyzed and inter
presenting such a perspective may move judges and jurors preted the data. A. F. Shariff drafted the manuscript, and C. J.
toward being less punitive and less retributive in general. Clark, J. D. Greene, J. C. Karremans, J. B. Luguri, J. W. Schooler,
Whereas previous research showed that diminished K. D. Vohs, and R. F. Baumeister provided critical revisions. Al
beliefs in free will encourage antisocial, immoral behav authors approved the final version of the manuscript for
submission.
ior, the current findings expand this story. One explana
tion for the prior findings is that participants may have
used the anti-free-will arguments as an excuse for moral Declaration of Conflicting Interests
laxity—taking advantage of apparently scientifically valid The authors declared that they had no conflicts of interest with
justifications in order to abandon self-control (Baumeister respect to their authorship or the publication of this article.
et al., 2009; Vohs & Schooler, 2008). In contrast, our stud
ies offered no immediate benefit to participants for being Funding
more punitive. Although our data do not negate the idea This project was supported in part by the John D. and Catherine
that lowered free-will beliefs provide an excuse for self T. MacArthur Foundation (Award 07-89249-000-HCD), by the
interested behavior, they suggest that diminished free Regents of the University of California, and by the John
will beliefs are more than excuses for selfishness: They Templeton Foundation. The content of this publication does
appear to provoke a genuine decline in belief in moral not necessarily reflect the views of the funders.

responsibility. That such changes were shown to occur

over the duration of a university neuroscience course Supplemental Material
(Study 4) suggests that as more people inside and outside Additional supporting information may be found at http://pss
of academia learn about neuroscience, moral attitudes . sagepub .com/ content/by/supplemental-data
may shift in response.
Whether these shifts are desirable is open to debate. Open Practices
Clearly, punishment serves important functions. Indeed,
the presence of norms ensuring the punishment of trans
gressors is essential for group cohesion (Fehr & Gâchter, All materials have been made publicly available via Open
2002). Free-will beliefs may, therefore, serve an important Science Framework and can be accessed at osf.io/dy3pm. The
cultural function in both encouraging the feelings of complete Open Practices Disclosure for this article can be
found at http://pss.sagepub.com/content/by/supplemental
responsibility that motivate people to behave ethically
data. This article has received the badge for Open Materials.
and producing the moral outrage and retributive desires
More information about the Open Practices badges can be
necessary to motivate costly but necessary punishment. found at https://osf.io/tvyxz/wiki/view/ and http://pss.sagepub
On the other hand, although diminished free-will ,com/content/25/l/3.full.
beliefs reduce retributivist motivations, Study 1 suggests
that the motivation to punish in order to benefit society
Notes
(consequentialist punishment) may remain intact, even
while the need for blame and desire for retribution are 1. Though 100%-effective treatments do not yet exist, no p
forgone. Thus, a societal shift away from endorsingticipant
free noted explicit doubts regarding the effectiveness of th
treatment. Although including this treatment in the vignette w
will could occur without disrupting the functional role of
necessary to circumvent rehabilitative motivations in this study
punishment. Society could fulfill its practical need for law
future research might include a more generalizable dependen
and order, leaving the social benefits of punishment
measure.

intact while avoiding the unnecessary human suffering

2. Analyses controlling for PANAS scores co
and economic costs of punishment often associated observed
with effects were due to mood. The effects of condition on
retributivism (Greene & Cohen, 2004; Tonry, 2004). FAD+ score, punishment, and blameworthiness in these analy
What is clear is that the belief in free will is inter ses remained significant, HI, 87) = 5.98,p = .02; HI, 84) = 5.06,
p = .03; As
twined with moral, legal, and interpersonal processes. and HI, 87) = 5.77,p = .04, respectively.

This content downloaded from

103.19.199.29 on Sat, 17 Oct 2020 06:54:51 UTC
All use subject to https://about.jstor.org/terms
 1570 Shariff et al.

References Nelissen, R. Μ. Α., & Zeelenberg, M. (2009)· Moral emotions as

determinants of third-party punishment: Anger, guilt, and
Aspinwall, L. G., Brown, T. R., & Tabery, J. (2012).the The
functions of altruistic sanctions. Judgment and Decision
double-edged sword: Does biomechanism increase or
Making, 4, 543-553.
decrease judges' sentencing of psychopaths? Science, Nichols,
33 7,S. (2011). Is free will an illusion? Scientific American
846-849. Mind, 22, 18-19.
Baer, J., Kaufman, J. C., & Baumeister, R. F. (Eds.). (2008).Osborne, Are L. (2003, June 22). Savant for a day. New York Times.
we free? Psychology and free will. Oxford, England: Oxford Retrieved from http://www.nytimes.com/2003/06/22/maga
University Press. zine/22SAVANT.html?pagewanted=all
Bargh, J. A. (2008). Free will is un-natural. In J. Baer, Paulhus,
J. C.D. L., & Carey, J. M. (2011). The FAD-Plus: Measuring
Kaufman, & R. F. Baumeister (Eds.), Are we free? Psychology lay beliefs regarding free will and related constructs.
and free will (pp. 128-154). Oxford, England: Oxford Journal of Personality Assessment, 93, 96-104.
University Press. Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures
Baumeister, R. F., Masicampo, E. J., & DeWall, C. N. (2009)· for estimating indirect effects in simple mediation models.
Prosocial benefits of feeling free: Disbelief in freeBehavior will Research Methods, Instruments, & Computers, 36,
increases aggression and reduces helpfulness. Personality 717-731.
and Social Psychology Bulletin, 35, 260-268. Rai, T. S., & Holyoak, K. J. (2013). Exposure to moral relativ
Braithwaite, J. (2002). Restorative justice & responsive regula ism compromises moral behavior. Journal of Experimental
tion. Oxford, England: Oxford University Press. Social Psychology, 49, 995-1001.
Buckholtz, J. W., Asplund, C. L., Dux, P. E., Zald, D.Schooler,
H., Gore, J. W. (2010). What science tells us about free will. In
J. C., Jones, O. D., & Marois, R. (2008). The neural corre R. F. Baumeister, A. R. Mele, & K. D. Vohs (Eds.), Free will
lates of third-party punishment. Neuron, 60, 930-940. and consciousness: How might they work? (pp. 191-218).
Crick, F. (1994). The astonishing hypothesis. New York, NY: England: Oxford University Press.
Oxford,
Touchstone. Shariff, A. F., Schooler, J., & Vohs, K. D. (2008). The hazards of
Eshleman, A. (2004). Moral responsibility. In Ε. N. Zalta (Ed.),claiming to have solved the hard problem of free will. In J.
The Stanford encyclopedia of philosophy. Retrieved fromBaer, J. C. Kaufman, & R. F. Baumeister (Eds.), Are we free?
plato. Stanford. edu/ entries/moral-responsibility/ Psychology and free will (pp. 181-204). Oxford, England:
Fehr, E., & Gâchter, S. (2002). Altruistic punishment in humans. Oxford University Press.
Nature, 415, 137-140. Smith, A. (1759). Theory of moral sentiments. London, England:
Gazzaniga, M. S. (2011). Who's in charge? Free will and the sciA. Miller.
ence of the brain. New York, NY: Harper (Ecco). Soon, C., Brass, M., Heinze, H., & Haynes,J. (2008). Unconscious
Greene, J., & Cohen, J. (2004). For the law, neurosci determinants of free decisions in the human brain. Nature

ence changes nothing and everything. Philosophical Neuroscience, 11, 543-545.

Transactions of the Royal Society B: Biological Sciences, Stephen, S. J. F. (1883). A history of the criminal law of England
359, 1775-1778. (Vol. 3)· New York, NY: Macmillan.
Krueger, F., Hoffman, M., Walter, H., & Grafman, J. (in press). Tonry, M. (2004). Thinking about crime: Sense and sensibil
An fMRI investigation of the effects of belief in free will ity in American penal culture. New York, NY: Oxford
on third-party punishment. Social Cognitive and Affective University Press.
Neuroscience. Vohs, K. D., & Schooler, J. W. (2008). The value of believing
Monterosso, J., & Schwartz, B. (2012, July 27). Did your in free will: Encouraging a belief in determinism increases
brain
make you do it? New York Times. Retrieved from http:// cheating. Psychological Science, 19, 49-54.
www.nytimes.com/2012/07/29/opinion/sunday/neurosci Watson, D., Clark, L. Α., & Tellegen, A. (1988). Development
ence-and-moral-responsibility.html?_r=l and validations of brief measures of positive and negative
Nahmias, E. (2011, November 13). Is neuroscience the death affect: The PANAS scales. Journal of Personality and Social
of free will? New York Times. Retrieved from http://opinion Psychology, 54, 1063-1070.
ator.blogs.nytimes.com/2011/ll/13/is-neuroscience-the Wegner, D. M. (2002). The illusion of conscious will. Cambridge,
death-of-free-will/ MA: MIT Press.

Nahmias, E., Morris, S., Nadelhoffer, T., & Turner, J. (2005).

Wolfe, T. (1997, February 2). Sorry, but your soul just died. Forbes.
Surveying freedom: Folk intuitions about free will Retrieved
and from http://90.146.8.18/en/archiv_files/19971/
E1997_236.pdf
moral responsibility. Philosophical Psychology, 18, 561-584.

This content downloaded from

103.19.199.29 on Sat, 17 Oct 2020 06:54:51 UTC
All use subject to https://about.jstor.org/terms