ng

rDaniel

Keywords:Consciousness

fMRIienceto the

unexplained neural phenomena to defend dualism and a spiritual perspective on the mind.

l essehiloso005;

1949). Although the physical origin of the body is

gments (Greene,, 2001), e(Farrer &he brain l

physical. Moreover, the appeal of fMRI research extendsbeyond academia and has captured the attention of thegeneral public. Laypersons express greater interest andbelief in psychological research when it also containsneuroscience information (McCabe & Castel, 2008), evenif that information does not provide additional support

0010-0277/$ - see front matter 2012 Elsevier B.V. All rights reserved.

Corresponding author. Address: Department of Psychology, Univer-sity of Illinois at Urbana-Champaign, 603 E Daniel St., Champaign, IL61820, USA. Tel.: +1 217 333 0344.

E-mail address: jlp@illinois.edu (J.L. Preston).

Cognition 127 (2013) 3137

Contents lists available at SciVerse ScienceDirect

Cognit

journal homepage: www.elsehttp://dx.doi.org/10.1016/j.cognition.2012.12.003and the physical, and so it feels as though we are madeof two parts: mind and body (Descartes, 1641; Ryle,

up during mental activity, fMRI research appears to nallyprovide hard evidence that the mind is grounded in theemotion, or exercises free will, subjective experienceseems to magically occur and is not obviously tied to anyphysical event (Wegner, 2003). The very act of introspec-tion suggests a qualitative difference between the mental

human experience, including moral judSommerville, Nystrom, Darley, & Cohen(LeDoux, 1996), and personal agency2002). Accompanied by vivid images of tmotionFrith,

ighting2000). Some scholars have recently argued that belief insouls is culturally universal and hard-wired in cognitiveprocesses used in agency detection (Bloom, 2004; Bering,2006). But more than just a way to understand otherminds, belief in the soul also helps people to explain theexperience of their own mind. Whenever one thinks, feels

physical explanation for the mind, this belief may be threa-tened by physical explanations for the mind. The presentresearch examines how belief in the soul is affected byneuroscience research that implies a physical origin ofthe mind. fMRI studies have uncovered neural correlatesfor many psychological phenomena seen as central to theExplanatory gapNeuroscienceSoulMindbody dualism

1. Introduction

Belief in the soula non-physicahas been an important subject of pfor thousands of years (e.g., Plato, 2In two experiments we examine how exposure to neuroscience research impacts belief inthe soul. We nd that belief in soul decreases when neuroscience provides strong mecha-nistic explanations for mind. But when explanatory gaps in neuroscience research areemphasized, belief in soul is enhanced, suggesting that physical and metaphysical explana-tions may be used reexively as alternative theories for mind. Implications for the future ofbelief in soul and neuroscience research are discussed.

2012 Elsevier B.V. All rights reserved.

nce of a beingphy and scienceDescartes, 1641/

intuitively understood, the origin of the mind is less clear;indeed, the mind appears to arise from some extra-physical force, and the concept of the soul is commonlyevoked as the source of this ineffable essence of self.

To the extent that belief in the soul is used as a meta-neuroscience has reignited old debates over mindbody dualism and the soul. While somescientists use neuroscience to bolster a material account of consciousness, others point toBrief article

Neuroscience and the soul: Competihuman experience

Jesse Lee Preston , Ryan S. Ritter, Justin HepleDepartment of Psychology, University of Illinois at Urbana-Champaign 603 E

a r t i c l e i n f o

Article history:Received 8 May 2012Revised 13 December 2012Accepted 16 December 2012

a b s t r a c t

The development ofthe psychological scena seen as centralexplanations for the

St. Champaign, IL, 61820, USA

techniques has generated a boom of neuroscience research acrosss, and revealed neural correlates for many psychological phenom-human experience (e.g., morality, agency). Meanwhile, the rise of

ion

vier .com/locate /COGNIT

but recreate an exact copy of it on Mars (see Supplemen-

32 J.L. Preston et al. / Cognition 127 (2013) 3137for the theory beyond what the behavioral data demon-strate (Weisberg, Keil, Goodstein, Rawson, & Gray, 2008).Widespread acceptance of fMRI as an explanatory toolmay also have an important effect on general beliefs aboutthe soul. Work on causal discounting demonstrates thatalternate explanations for the same phenomenon can com-pete with each other on a cognitive level, such that increas-ing belief in one diminishes belief in the other (Morris &Larrick, 1995; Sloman, 1994). For example, reading scien-tic explanations for important phenomena (e.g., evolu-tion) reduces belief in religious explanations (e.g.,creationism) (Lawson & Weser, 1990; Shariff, Cohen, &Norenzayan, 2008), but when scientic explanations areframed as weak, they can actually bolster belief in super-natural explanations (Preston & Epley, 2009). We proposethat a similar reexive relationship may also occur forphysical vs. metaphysical explanations of the mind, withimplications for neuroscience research on the belief inthe soul. As neuroscience continues to discover neural cor-relates of more and more essential psychological pro-cesses, the brain could edge out the soul as the prevailingexplanation for the mind (Clark, 2010; Farah, 2005).

On the other hand, if neuroscience seems limited in itsability to explain psychological experiences, exposure tothat research could enhance belief in the soul. Despitethe many impressive breakthroughs of fMRI studies, thereremains one epistemological issue of the mind thatneuroscience may not be able to solve, dubbed the hardproblem of consciousness (Chalmers, 1996) or theexplanatory gap (Levine, 1983). In sum, although neuro-scientists can identify neural correlates associated withmental processes, they are still unable to explain preciselyhow activity in the brain creates the experience of thesemental phenomena. This issue can have some importantimplications for belief in the soul. If the neural activity cap-tured by fMRI serves to demystify the mind, awareness ofan explanatory gap may only re-mystify the mind. Indeed,while some scientists use neuroscience to bolster a mate-rial account of consciousness (e.g., Crick, 1994), otherspoint to unexplained neural phenomena to defend dualismand a spiritual perspective on the mind (e.g. Schwartz,Stapp, & Beauregard, 2005). Whether a legitimate concernor not (for a discussion, see Dennett, 1991; Nagel, 1974), anapparent explanatory gap leaves some aspects of the mindunexplained and so re-opens the intuitive plausibility ofmetaphysical explanations.

1.1. Research overview

The present research examines how neuroscienceexplanations for psychological phenomena can impact laybelief in soul, guided by two complementary hypotheses:

H1. Exposure to neuroscience research with strong mech-anistic explanations for psychological experience willdecrease belief in the soul as an alternative explanationfor the mind.

H2. Exposure to neuroscience research with weak mecha-nistic explanations for psychological experience (i.e.,research highlighting the explanatory gap) will increasetary materials). Dilemma 2 described a scenario in whichparticipants were asked to imagine they had a fatal illness,and that although scientists were working on a cure, itbelief in the soul as an alternative explanation for themind.

2. Experiment 1

2.1. Method

One hundred fty-one undergraduates (95 women) vol-unteered to participate for partial course credit. Partici-pants were randomly assigned to one of three conditions:psychology information, neuroscience information, or acontrol condition. All participants were told that theywould complete three unrelated studies for credit. The rsttask was framed as a study to measure students judg-ments about college course descriptions. Subjects readbrief course descriptions, selected the best title for thecourse between two options we provided, and rated theirinterest in the course. All participants rst read a coursedescription on introduction to geology, then a surveycourse on Shakespeare. In the psychology and neurosci-ence conditions, subjects then read two additional coursedescriptions: one on the study of love and one on morality.Both the Love and Morality course descriptions raised anumber of questions about the psychological phenomenon(e.g.,Why do people fall in love at rst sight? Are some peopleinherently good or bad?). In the neuroscience condition, thedescriptions also referred to recent evidence for the neuralbasis of love/morality (see Supplementary materials). Thepsychology description therefore raised the same ques-tions regarding the phenomenology of the mind withoutproviding any physical mechanism for the psychologicalexperiences. As with the two previous course descriptions,subjects selected the best name for each course descriptionbetween two options. For the Love course, two optionswere: Mechanisms of Love or Mystery of Love. Forthe Morality course, the two options were: Moral Mecha-nisms or The Moral Compass. These items thereforeserved as a manipulation check, as the rst options sug-gested a mechanistic explanation for the phenomena,whereas the second options suggested the presence ofsome explanatory gap in the understanding of thephenomena.

After the course description task, participants com-pleted some ller tasks (spatial reasoning items, and four4 items taken from Reading the mind in the eyes test),which they were told would measure visual processingas the second part of the study.

2.1.1. Bodysoul dilemmasIn the third part of the study, participants played a

game called Staying Alive with two hypothetical dilem-mas involving trade-offs between different forms of theself (materials adapted from www.philosophersnet.com/games/). Dilemma 1 described a scenario in which partici-pants could choose to travel to Mars by spaceship, or to bereplicated by a transporter that would destroy their body

the three items. The 11-item Dualism scale showed good

J.L. Preston et al. / Cognition 127 (2013) 3137 33would not be ready in time to save them. Participants wereasked to decide to either freeze their body to be laterthawed and cured when the cure was discovered (how-ever, the freezing process would likely kill their soul), orto let their body die from the disease but preserve theirsoul. Participants responded on a six-point scale, with end-points: 1 = I would strongly prefer to let my body die, 6 = Iwould strongly prefer to be frozen. While both Dilemmas 1and 2 measure participants relative value of their physicalbody, only Dilemma 2 contrasts the physical body with thesoul. Responses to Dilemma 2 serve as our dependent var-iable because they indicate whether participants valuetheir soul more or less than their physical body.

2.1.2. Other measuresAt the end of the study, participants completed an 11-

item Dualism scale comprised of questions on belief insoul, separation between mind and body, and materialismof the mind (e.g., The true self is not governed by thebrain, but by a persons soul, endpoints: 1 = Strongly Dis-agree; 7 = Strongly Agree), and a three-item religiosity in-dex that measured frequency of religious attendance(1 = never, 5 = frequently), religiosity (1 = not at all,5 = Strongly), and belief in God (1 = not at all, 5 = Strongly).Participants then reported the number of neuroscience andpsychology courses previously taken, and completeddemographic items on gender, age, ethnicity, and alsoand political ideology (1 = Strongly liberal, 5 = Stronglyconservative). Finally, participants were probed for suspi-cion and debriefed.

2.2. Results

2.2.1. Manipulation checkNo participants correctly guessed the hypothesis of the

study. We compared the course titles selected for the Psy-chology and Neuroscience courses to test whether thecourse descriptions effectively manipulated participantsperceptions of the mechanistic nature of these phenomena.Title selections for each course were dummy-coded as 1 or2, with 1 reecting the mechanistic title, and 2 reectingthe explanatory gap title. Choice of the mechanistic-expla-nation vs. explanatory-gap title was correlated betweenthe Love and Morality courses, r(101) = .30, p = .002, andthe mean of the two titles was computed. An independentsample t-test conrmed that subjects were more likely toselect the mechanistic course titles in the neurosciencecondition relative to the psychology condition, t(99) =5.09, p < .001. Participants who answered both manipu-lation checks incorrectly were removed from furtheranalyses, leaving n = 134 in the total sample includingparticipants in the control condition.

2.2.2. Bodysoul dilemmasOne-way ANOVAs on condition were conducted on both

hypothetical dilemmas. We did not expect any effect onDilemma 1 (teleportation), and no effect was observedfor the omnibus ANOVA, F(2,131) = 1.96, p = .14. But con-sistent with predictions, the omnibus ANOVA was signi-cant for Dilemma 2 (bodysoul tradeoff), F(2,131) = 3.47,p = .03, g2P = .05, and on the planned linear comparisonscale reliability (Cronbachs a = .84). We calculated the cor-relations between Dualism scales, religiosity, Dilemma 1(teleportation) and Dilemma 2 (freeze body/kill soul). Thereligiosity index was positively correlated with Dualism(r = .50, p < .001), and negatively correlated with the choiceto freeze body/kill soul in Dilemma 2 (r = .36, p < .001).Belief in dualism was also negatively associated with thechoice to freeze body/kill soul (r = -.49, p < .001). None ofthe variables were correlated with Dilemma 1 (teleporta-tion) (all ps > .23), suggesting that value of the physical selfmeasured by this dilemma was conceptually distinct frommindbody dualism and other spiritual/metaphysical be-liefs. One-way ANOVA on condition was not signicanton either the Dualism scale (F < 1, p = .43) or on the religi-osity index (F < 1, p = .66).

2.2.4. DiscussionIn Experiment 1, belief in the soul was measured by a

hypothetical tradeoff between body and soul. Reading psy-chological descriptions about love and morality accompa-nied by neuroscience explanations increased preferencefor body-continuity over the soul. Conversely, reading thepsychological descriptions without neuroscience explana-tions (i.e., raising questions about the phenomenologywithout providing answers) increased preference forsoul-continuity over the body. Manipulations did not affectpreferences in a teleportation dilemma, suggesting the ef-fect is not merely due to increased value for the physicalbody. Rather, the ndings support our argument that beliefin the soul and belief in mechanistic explanations of expe-rience are reexive explanations for understanding thepsychological experience of the mind.

3. Experiment 2

In Experiment 1, exposure to a neuroscience coursedescription increased relative value of the physical bodyover the soul, whereas reading a psychology coursedescription on the same topics increased value of the soulover the physical body. Experiment 2 extended the nd-ings of Experiment 1 to a behavioral domain. Drawing frompopular culture, we examined participants willingness tosell their soul after being exposed to neuroscience infor-mation. In addition, Experiment 2 manipulated the explan-atory strength of the neuroscience information (strong orweak explanations). This allowed us to test whether the(weights: neuroscience = 1, control = 0, psychology = 1)was also signicant F(1,131) = 7.11, p < .01. Participantsin the neuroscience condition showed strongest preferenceto freeze their body and let their soul die (M = 3.59,SD = 1.77) compared to the control condition (M = 3.34,SD = 1.69), and participants in the psychology condition re-ported lowest preference (M = 2.57; SD = 1.74) to freezetheir body and let their soul die.

2.2.3. Other measuresReliability on the three religiosity items (attendance,

religiousness, belief) was strong, (Cronbachs a = .85), wetherefore created a religiosity index using the mean of

34 J.L. Preston et al. / Cognition 127 (2013) 3137mere presence of neuroscience information would de-crease belief in soul, or whether the neuroscience informa-tion must serve as a strong explanation for psychologicalphenomena while weak neuroscience information (i.e., anexplanatory gap) may enhance lay belief in soul.

3.1. Method

Seventy-ve undergraduates (37 women) volunteeredto participate for partial course credit. Participants wereassigned to one of three conditions: strong neuroscience,weak neuroscience, or a control condition. In both neuro-science conditions, participants read brief descriptions ofneuroscience research on the experience of free will, love,and moral judgment. In the strong neuroscience condition,each description concluded with the statement that thephenomenon was completely explained by neural activity.In the weak neuroscience condition, each description con-cluded with a statement that the research could not ex-plain the source of the phenomenon or how brainactivity translates into the phenomenal experience (seeall neuroscience descriptions in Appendix B). Followingeach passage, participants were asked to select one oftwo titles that best described the passage. In the controlcondition there was no passage.

3.1.1. Soul cardsPrevious research has shown that most undergraduate

participants refuse to sell their soul as part of a psychologyexperiment even if they do not believe in souls (Haidt &Bjorklund, 2000). Instead of buying participants souls,we made the experimental paradigm less threatening tosubjects by using a novelty card to represent the soul. Fol-lowing the course description manipulation, participantswere brought into a separate room and were seated at adesk across from the experimenter. Each participant re-ceived a novelty soul card as a free gift for participatingin the experiment. The front of the card was labeled SoulID card, with a 35 digit ID code and a picture of a humanform. The back of the card had spaces that read: initials ofcarrier, and date of birth for the cardholder to complete.The experimenter placed the soul card face down on the ta-ble and asked participants to write their date of birth andinitials on the card. The card was turned over and partici-pants were told it represented their soul and were givena moment to examine the card.

3.1.2. BargainingParticipants were told that the card was theirs to keep if

they wished or that they could try to sell it back to theexperimenter for up to $10. The rules of the bargainingprocedure were then explained as follows: If participantschose to sell their soul cards, they would write down aminimum asking price (up to $10) on a piece of paper, thatrepresented the smallest amount of money for which theywould be willing to sell the card. After writing down theirprice, participants randomly drew a dollar amount be-tween $1 and $10 (written on slips of paper) from a box.If the amount randomly drawn was equal to or greaterthan the participants asking price, the participant wouldhave to sell the soul card in exchange for the amountdrawn from the box. However, if the participant drew anumber that was less than the asking price, the participantwould keep the soul card and would not get any money.Thus, participants should be motivated to generate a min-imum price for the card that reects its subjective value.Once the bargaining procedures were clear, participantsdecided to either keep or try to sell their soul card in thebargaining game. Those participants who chose to keeptheir card (i.e., not play the bargaining game), were thenasked for a hypotheticalminimum price (above $10, no lim-it), that they would accept to sell the card. Special care wastaken during debrieng to reassure subjects that the cardwas not a true embodiment of their soul, and that no realpurchase of their soul had transpired.

3.1.3. Other measuresAfter the bargaining procedure, participants completed

a 13-item Dualism scale (see Supplementary materials).Participants reported their current college major and num-ber of science classes previously taken. Finally, participantscompleted demographic items on gender, age, ethnicity,and also frequency of religious attendance (1 = never,5 = frequently), religiosity (1 = not at all, 5 = strongly),and belief in God (1 = not at all, 5 = strongly).

3.2. Results

3.2.1. Manipulation checkChoices of the passage title (mechanistic-explanation

vs. explanatory-gap titles) in the Psychology/Neuroscienceconditions were summed across the three passages, and anindependent sample t-test conrmed that subjects weremore likely to select the mechanistic titles in the Strongvs. Weak neuroscience condition, t(50) = 5.06, p < .001. Allparticipants answered at least one manipulation checkitem correctly and were retained in the subsequentanalyses.

3.2.2. BargainingParticipants who played the bargaining game were

coded as willing to sell the soul card, regardless of whetherthe bargain resulted in a successful sale of the card. Theproportion of participants who chose to sell their soul cardin each condition was analyzed by chi-square (see Table 1).As predicted, the linear by linear association on conditionwas signicant, v2 (1, N = 75) = 5.02, p = .03. Relative tothe control condition (52% sell soul card), participants ex-posed to strong neuroscience explanations were morelikely to sell their soul card (72%), whereas participants ex-posed to weak neuroscience explanations were less likelyto sell their soul card (40%).

We also examined the reported minimum prices to sellthe card. For those who opted to engage in the bargaininggame, prices ranged from 1 to 10 dollars. But those who didnot play the game reported a hypothetical minimum pricethat had no upper limit. The hypothetical minimum priceswere not recorded for eight participants who chose to keepthe card. For the remaining subjects who kept the card,minimum prices ranged between one thousand dollarsand priceless (see Fig. 1). The skewed variance andlimitless upper range of this data made it unfeasible to

analyze the differences statistically, but simultaneously re-veals the powerful difference between those who werewilling to sell the card vs. those who were not. Subjectswho chose not to play the bargaining game valued the soulcard considerably more than those who chose play the bar-gaining game, suggesting it took on some important valuebeyond just a piece of paper.

for important psychological phenomena, suggesting anexplanatory competition between the soul and brain as dif-ferent explanations of mind. On the other hand, weak neu-roscience information decreased willingness to sell thesoul card. This suggests that mere awareness of neurosci-ence does not diminish belief in the soul, rather it is theapparent power of neuroscience as an explanation for psy-

Table 1Observed and expected cases of participants who sell vs. keep the soul card in Experiment 2, by condition.

Explanation condition

Weak neuroscience Control Strong neuroscience Total

Choice Observed (Expected) Observed (Expected) Observed (Expected)

Keep 16 (12.2) 11 (10.4) 7 (11.3) 34

Sell 11 (14.8) 12 (12.6) 18 (13.7) 41Total 27 23 25 75

J.L. Preston et al. / Cognition 127 (2013) 3137 353.2.3. Other measuresThe three religious items (attendance, religiousness, be-

lief) showed good scale reliability, Cronbachs a = .90. Wetherefore computed the mean of these items to create areligiosity index. The Dualism scale also showed good reli-ability (Cronbachs a = .74), and was positively correlatedwith the religiosity index (r = .49, p < .001). One-wayANOVA on condition was not signicant on either theDualism scale (F (2,72) = 1.84, p = .17) or on the religiosityindex (F < 1, p = .57). However, willingness to sell the soulcard was negatively correlated with both the religiosity in-dex (r = .25, p = .01), and the Dualism scale (r = .29,p = .01).

3.3. Discussion

In Experiment 2, willingness to sell a soul card varied asa function of exposure to strong versus weak neuroscienceexplanations for psychological phenomena. Exposure tostrong neuroscience explanations (for the experience oflove, moral judgment, and conscious will) increased partic-ipants willingness to sell their soul card. This neuroscienceinformation provided an alternative material explanationFig. 1. Minimum selling prices for the Soul Card, by subjecchological phenomena. Indeed, weak neuroscience expla-nations further diminished willingness to sell the soulcard, evidence that an apparent explanatory gap can en-hance metaphysical explanations as an alternative meansof understanding the experience of mind.

It is also worth noting that in both Experiments 1 and2, manipulations did not impact responses in a self-re-ported Dualism scale, but the scale was correlated withreligiosity and the key dependant measures. It is possiblethat the Dualism scale was too far removed from themanipulation to be affected, but another possibility isthat the Dualism scale has better validity as a measureof individual differences in belief (or trait beliefs),rather than temporary changes in belief, (or state be-liefs). People can be reluctant to alter their explicit state-ments about religious and spiritual belief after amanipulation of this kind, but behavioral measures of be-lief are more sensitive to gut-level feelings that can be af-fected by explanatory information. Unlike the explicitstatements in the Dualism scale, the key dependent mea-sures in both studies (the hypothetical dilemma and sell-ing the soul-card) used a symbolic action on the soul.This may be essential to their effectiveness, because thesemeasures allow psychological leeway to alternativelyts who opted to sell or keep in the bargaining game.

ourish in the age of fMRI research. There are also other

extensions of the evolutionary model. Behavioral and Brain Sciences,

Farah, M. J. (2005). Neuroethics: The practical and the philosophical.

36 J.L. Preston et al. / Cognition 127 (2013) 3137reasons and motivations to believe in the soul (e.g., fearof death), and potential social costs to losing belief in thesoul (e.g., decreased moral responsibility; Vohs & Schooler,treat the soul as either real or ction, without violatingexplicit religious convictions.

4. Conclusions

People have long believed in the soul, in part as a way tounderstand the source of the mind and consciousness. Butrecently neuroscience research has made important dis-coveries about the brains involvement in essential psycho-logical experiences that may impinge on these long-heldbeliefs. The present research found that belief in the soulwas inversely related to the availability of neuroscienceexplanations of the mind. Importantly, the direction ofchange depended on the perceived explanatory power ofthe neuroscience information: whereas strong neurosci-ence explanations decreased belief in the soul, weak neu-roscience explanations increased belief in the soul.Consistent with a causal discounting account, these nd-ings suggest that acceptance of physical vs. metaphysicalkinds of explanation are inversely related, and may be usedreexively as a means to understand the mind.

Although these studies demonstrated a hydraulic rela-tion between neuroscience evidence and belief in soul,we should be clear that people do not necessarily chooseone explanation over another. Indeed, in the absence ofstrong evidence for one kind of explanation, it may be mostcommon for people to seek a synthesis between physicaland metaphysical folk theories and use aspects of each tosatisfy different explanatory needs (Legare, Evans, Rosen-gren, & Harris, 2012). Descartes himself argued that thepineal gland was the seat of the soul a physical struc-ture where the soul interacted with the body (Descartes,1641). Today many people subscribe to a kind of moderndualism that accepts the brain as responsible for computa-tional cognitive functions (i.e., ability to think), but re-serves the soul as an explanation for personal aspects ofmind, such as emotions (Richert & Harris, 2008). As thepublic love affair with neuroscience continues, the feelingof a personal essence of self that we all experience will cer-tainly remain. However, the present results do illustratethat relative belief in physical versus metaphysical descrip-tions of mind change in response to new neuroscienceinformation, and the trajectory of neuroscience researchmay forecast a shift in the relative emphasis on the soulas an explanatory tool.

Finally, these ndings have some interesting implica-tions for the study of neuroscience itself. It may be tempt-ing to conclude that neuroscience research will replacebelief in the soul as it uncovers more information on thebrains role in our most essential human experiences.Important however, neuroscience research still does notseem to fully capture the subjective experience of con-sciousness that compels us to feel a mind separate fromour bodies (the root of the hard problem of consciousness).Indeed, the present ndings also imply that limitations ofneuroscience provide explanatory space for the soul toTrends in Cognitive Sciences, 9, 3440.Farrer, C., & Frith, C. D. (2002). Experiencing oneself vs. another person as

being the cause of an action: The neural correlates of the experienceof agency. NeuroImage, 15, 596603.

Greene, J. D., Sommerville, R. B., Nystrom, L. E., Darley, J. M., & Cohen, J. D.(2001). An fMRI investigation of emotional engagement in moraljudgment. Science, 293, 21052108.

Haidt, J., Bjorklund, F., & Murphy, S. (2000). Moral dumbfounding: Whenintuition nds no reason. University of Virginia (Unpublishedmanuscript).

Lawson, A. E., & Weser, J. (1990). The rejection of nonscientic beliefsabout life: Effects of instruction and reasoning skills. Journal ofResearch in Science Teaching, 27, 589606.

LeDoux, J. E. (1996). The emotional brain: The mysterious underpinning ofemotional life. New York, NY: Simon & Schuster.

Legare, C. H., Evans, E. M., Rosengren, K. S., & Harris, P. L. (2012). Thecoexistence of natural and supernatural explanations across culturesand development. Child Development, 83, 779793.29, 486498.Bloom, P. (2004). Descartes baby: How the science of child development

explains what makes us human. New York, NY: Basic Books.Chalmers, D. J. (1996). The conscious mind: In search of a fundamental

theory. New York, NY: Oxford University Press.Clark, A. (2010). There is no non-materialist neuroscience. Cortex: A

Journal Devoted to the Study of the Nervous System and Behavior, 46,147149.

Crick (1994). The astonishing hypothesis. New York, NY: Simon & Schuster.Dennett, D. C. (1991). Consciousness explained. Boston, MA: Little, Brown

and Co..Descartes, R. (1641). Meditations on rst philosophy. In R. Ariew (Ed.),

Ren Descartes Philosophical essays and correspondence (pp. 97141).Indianapolis, IN: Hackett, Publishing Company, Inc.found, in the online version, at http://dx.doi.org/10.1016/j.cognition.2012.12.003.

References

Bering, J. M. (2006). The cognitive science of souls: Clarications and2008). It is therefore possible that advances in neurosci-ence could create a backlash for threatening the valued be-lief in the soul. What then, should be neuroscientistsresponsibility in framing their results? To protect the cher-ished beliefs of others by diminishing the explanatorypower of neuroscience ndings? Or to take a confronta-tional stance against the soul by framing results as power-ful explanations, and so dissuade the public from the perilsof magical thinking? Sciences responsibility must be onlyto the truth, of course, but given the serious issues of bothunderselling and overclaiming the implications of neuro-science research, the need for clear and unbiased represen-tation of results is ever more important.

Acknowledgments

We thank Evan Branson, Da Yae Keum, Claire Hou-tsma, Regina Mendez, Tessa Rainey, Agnes Strojewska,and Pin-ya Tseng for assistance in data collection. Thispublication was made possible from grant support fromthe John Templeton Foundation, awarded to Jesse Preston.The opinions expressed in this publication are those of theauthors and do not necessarily reect those of the JohnTempleton Foundation.

Appendix A. Supplementary material

Supplementary data associated with this article can be

Levine, J. (1983). Materialism and qualia: The explanatory gap. PacicPhilosophical Quarterly, 64, 354361.

McCabe, D. P., & Castel, A. D. (2008). Seeing is believing: The effect of brainimages on judgments of scientic reasoning. Cognition, 107, 343352.

Morris, M. W., & Larrick, R. P. (1995). When one cause casts doubt onanother: A normative analysis of discounting in causal attribution.Psychological Review, 102, 331355.

Nagel, T. (1974). What is it like to be a bat? The Philosophical Review, 83,435450.

Plato. (2005). Phaedo. In S. M. Cohen, P. Curd, & C. D. C. Reeve (Eds.),Readings in Ancient Greek Philosophy: From Thales to Aristotle (pp. 229-281). Indianapolis, IN: Hackett Publishing Company, Inc.

Preston, J., & Epley, N. (2009). Science and god: An automatic oppositionbetween ultimate explanations. Journal of Experimental SocialPsychology, 45, 238241.

Richert, R. A., & Harris, P. L. (2008). Dualism revisited: Body vs. mind vs.soul. Journal of Cognition and Culture, 8, 99115.

Ryle, G. (1949). The concept of mind. Oxford, England: Barnes & Noble.

Schwartz, J. M., Stapp, H. P., & Beauregard, M. (2005). Quantum physics inneuroscience and psychology: A neurophysical model of mind-braininteraction. Philosophical Transactions of the Royal Society B, 360,13091327.

Shariff, A. F., Cohen, A. B., & Norenzayan, A. (2008). The devils advocate:Secular arguments diminish both implicit and explicit religious belief.Journal of Cognition and Culture, 8, 417423.

Sloman, S. A. (1994). When explanations compete: The role ofexplanatory coherence in judgments of likelihood. Cognition, 52, 121.

Vohs, K. D., & Schooler, J. W. (2008). The value of believing in free will:Encouraging a belief in determinism increases cheating. PsychologicalScience, 19, 4954.

Wegner, D. M. (2003). The minds best trick: How we experienceconscious will. Trends in Cognitive Science, 7, 6569.

Weisberg, D. S., Keil, F. C., Goodstein, J., Rawson, E., & Gray, J. R. (2008).The seductive allure of neuroscience explanations. Journal of CognitiveNeuroscience, 20, 470477.

J.L. Preston et al. / Cognition 127 (2013) 3137 37

Neuroscience and the soul: Competing explanations for the human experience1 Introduction1.1 Research overview

2 Experiment 12.1 Method2.1.1 Bodysoul dilemmas2.1.2 Other measures

2.2 Results2.2.1 Manipulation check2.2.2 Bodysoul dilemmas2.2.3 Other measures2.2.4 Discussion

3 Experiment 23.1 Method3.1.1 Soul cards3.1.2 Bargaining3.1.3 Other measures

3.2 Results3.2.1 Manipulation check3.2.2 Bargaining3.2.3 Other measures

3.3 Discussion

4 ConclusionsAcknowledgmentsAppendix A Supplementary materialReferences