Frames, Biases, and Rational Decision-Making in the Human Brain

Abstract

Human choices are remarkably susceptible to the manner in which options are presented. This so-called “framing effect” represents
a striking violation of standard economic accounts of human rationality, although its underlying neurobiology is not understood.
We found that the framing effect was specifically associated with amygdala activity, suggesting a key role for an emotional
system in mediating decision biases. Moreover, across individuals, orbital and medial prefrontal cortex activity predicted
a reduced susceptibility to the framing effect. This finding highlights the importance of incorporating emotional processes
within models of human choice and suggests how the brain may modulate the effect of these biasing influences to approximate
rationality.

Full text

DOI: 10.1126/science.1128356
, 684 (2006); 313Science
et al.Benedetto De Martino,
the Human Brain
Frames, Biases, and Rational Decision-Making in
www.sciencemag.org (this information is current as of April 30, 2007 ):
The following resources related to this article are available online at
http://www.sciencemag.org/cgi/content/full/313/5787/684
version of this article at:
including high-resolution figures, can be found in the onlineUpdated information and services,
http://www.sciencemag.org/cgi/content/full/313/5787/684/DC1
can be found at: Supporting Online Material
found at:
can berelated to this articleA list of selected additional articles on the Science Web sites
http://www.sciencemag.org/cgi/content/full/313/5787/684#related-content
http://www.sciencemag.org/cgi/content/full/313/5787/684#otherarticles
, 8 of which can be accessed for free: cites 21 articlesThis article
7 article(s) on the ISI Web of Science. cited byThis article has been
http://www.sciencemag.org/cgi/content/full/313/5787/684#otherarticles
1 articles hosted by HighWire Press; see: cited byThis article has been
http://www.sciencemag.org/cgi/collection/neuroscience
Neuroscience
: subject collectionsThis article appears in the following
http://www.sciencemag.org/about/permissions.dtl
in whole or in part can be found at: this article
permission to reproduce of this article or about obtaining reprintsInformation about obtaining
registered trademark of AAAS.
c 2006 by the American Association for the Advancement of Science; all rights reserved. The title SCIENCE is a
CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005.
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the
on April 30, 2007 www.sciencemag.orgDownloaded from

and Ipl1 moves from kinetochores to spindle
microtubules shortly after the initiation of
anaphase (5, 27, 28). Microtubule attachment
to kinetochores in anaphase may be stabi-
lized by the loss of Ipl1, helping to keep the
checkpoint inactive. However, Ipl1 mutants
respond to treatment with nocodazole, whereas
anaphase-arrested cells do not (Fig. 1A), which
suggests that additional factors, such as Mps1
degradation, have turned off the checkpoint in
anaphase (29). The organization of other Bchro-
mosomal passenger proteins[ also changes as
cells enter anaphase (30), as do spindle micro-
tubule dynamics (31), and these factors may also
influence checkpoint behavior in anaphase.
Finally, the checkpoint destabilizes Cdc20, as
well as inhibits its activity, which reinforces the
mutual antagonism between the checkpoint and
APC
Cdc20
.
We have presented evidence for a mecha-
nism that inactivates the spindle checkpoint as
yeast cells enter anaphase. When mitosis starts,
the APC is off, the checkpoint is on, and check-
point proteins are stable. As long as one chromo-
some has not aligned, the checkpoint inhibits
the APC. When this chromosome biorients, a
threshold is crossed, the APC becomes active,
cells enter anaphase, and the destruction of Mps1
(and possibly other checkpoint proteins) perma-
nently inactivates the checkpoint. The opposing
activities of the checkpoint and the APC let cells
switch rapidly between prometaphase, when
they can sensitively monitor chromosome align-
ment, and anaphase, when they are irreversibly
committed to entering the next cell cycle, despite
the lack of tension at the kinetochores.
References and Notes
1. K. Nasmyth, Science 297, 559 (2002).
2. R. B. Nicklas, C. A. Koch, J. Cell Biol. 43, 40 (1969).
3. M. A. Hoyt, L. Totis, B. T. Roberts, Cell 66, 507 (1991).
4. R. Li, A. W. Murray, Cell 66, 519 (1991).
5. B. A. Pinsky, C. Kung, K. M. Shokat, S. Biggins, Nat. Cell
Biol. 8, 78 (2006).
6. C. L. Rieder, A. Schultz, R. Cole, G. Sluder, J. Cell Biol.
127, 1301 (1994).
7. X. Li, R. B. Nicklas, Nature 373, 630 (1995).
8. W. A. Wells, A. W. Murray, J. Cell Biol. 133, 75 (1996).
9. A. Musacchio, K. G. Hardwick, Nat. Rev. Mol. Cell Biol. 3,
731 (2002).
10. L. H. Hwang et al., Science 279, 1041 (1998).
11. S. L. Jaspersen, J. F. Charles, R. L. Tinker-Kulberg,
D. O. Morgan, Mol. Biol. Cell 9, 2803 (1998).
12. O. Cohen-Fix, J. M. Peters, M. W. Kirschner, D. Koshland,
Genes Dev. 10, 3081 (1996).
13. M. Shirayama, A. Toth, M. Galova, K. Nasmyth, Nature
402, 203 (1999).
14. O. Cohen-Fix, D. Koshland, Genes Dev. 13, 1950 (1999).
15. F. M. Yeong, H. H. Lim, C. G. Padmashree, U. Surana,
Mol. Cell 5, 501 (2000).
16. E. Weiss, M. Winey, J. Cell Biol. 132, 111 (1996).
17. K. G. Hardwick, E. Weiss, F. C. Luca, M. Winey,
A. W. Murray, Science 273, 953 (1996).
18. C. A. Ball et al., Nucleic Acids Res. 28, 77 (2000).
19. K. G. Hardwick, A. W. Murray, J. Cell Biol. 131, 709 (1995).
20. R. Li, Proc. Natl. Acad. Sci. U.S.A. 96, 4989 (1999).
21. M. Glotzer, A. W. Murray, M. W. Kirschner, Nature 349,
132 (1991).
22. Single-letter abbreviations for the amino acid residues
are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe;
G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro;
Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; Y, Tyr; and
X, any amino acid.
23. S. L. Jaspersen et al., Dev. Cell 7, 263 (2004).
24. See supporting data on Science Online.
25. M. H. Jones et al., Curr. Biol. 15, 160 (2005).
26. J. Pan, R. H. Chen, Genes Dev. 18, 1439 (2004).
27. S. Buvelot, S. Y. Tatsutani, D. Vermaak, S. Biggins, J. Cell
Biol. 160, 329 (2003).
28. G. Pereira, E. Schiebel, Science 302, 2120 (2003).
29. S. Biggins, N. Bhalla, A. Chang, D. L. Smith,
A. W. Murray, Genetics 159, 453 (2001).
30. P. Vagnarelli, W. C. Earnshaw, Chromosoma 113, 211
(2004).
31. T. Higuchi, F. Uhlmann, Nature 433, 171 (2005).
32. O. Cohen-Fix, D. Koshland, Proc. Natl. Acad. Sci. U.S.A.
94, 14361 (1997).
33. We thank S. C. Schuyler and B. Stern for their support and
technical assistance, the members of the Murray and
Winey labs for their helpful comments and suggestions,
K. Hardwick for providing antibodies, and A. Wenger for
his efforts. Supported by NIH grants GM43987 (A.W.M.)
and GM51312 (M.W.). S.L.J. is a Leukemia and Lymphoma
Society Special Fellow.
Supporting Online Material
www.sciencemag.org/cgi/content/full/1127205/DC1
SOM Text
Figs. S1 to S6
Table S1
9 March 2006; accepted 30 May 2006
Published online 6 July 2006;
10.1126/science.1127205
Include this information when citing this paper.
Frames, Biases, and Rational
Decision-Making in the Human Brain
Benedetto De Martino,
*
Dharshan Kumaran, Ben Seymour, Raymond J. Dolan
Human choices are remarkably susceptible to the manner in which options are presented. This
so-called ‘‘framing effect’’ represents a striking violation of standard economic accounts of human
rationality, although its underlying neurobiology is not understood. We found that the framing
effect was specifically associated with amygdala activity, suggesting a key role for an emotional
system in mediating decision biases. Moreover, across individuals, orbital and medial prefrontal
cortex activity predicted a reduced susceptibility to the framing effect. This finding highlights the
importance of incorporating emotional processes within models of human choice and suggests
how the brain may modulate the effect of these biasing influences to approximate rationality.
A
central tenet of rational decision-making
is logical consistency across decisions,
regardless of the manner in which avail-
able choices are presented. This assumption,
known as Bextensionality[ (1)orBinvariance[
(2), is a fundamental axiom of game theory (3).
However, the proposition that human decisions
are Bdescription-invariant[ is challenged by a
wealth of empirical data (4, 5). Kahneman and
Tversky originally described this deviation from
rational decision-making, which they termed
the Bframing effect,[ as a key aspect of pros-
pect theory (6, 7).
Theories of decision-making have tended to
emphasize the operation of analytic processes in
guiding choice behavior. However, more intui-
tive or emotional responses can play a key role in
human decision-making (8–10). Thus, when
taking decisions under conditions when availa-
ble information is incomplete or overly com-
plex, subjects rely on a number of simplifying
heuristics, or efficient rules of thumb, rather
than extensive algorithmic processing (11).
One suggestion is that the framing effect
results from systematic biases in choice
behavior arising from an affect heuristic under-
written by an emotional system (12, 13). How-
ever, despite the substantial role of the framing
effect in influencing human decision-making,
the underlying neurobiological basis is not
understood.
We investigated the neurobiological basis of
the framing effect by means of functional mag-
netic resonance imaging (f MRI) and a novel
financial decision-making task. Participants (20
university students or graduates) received a
message indicating the amount of money that
they would initially receive in that trial (e.g.,
BYou receive U50[). Subjects then had to choose
between a Bsure[ option and a Bgamble[ option
presented in the context of two different frames.
The Bsure[ option was formulated as either the
amount of money retained from the initial
starting amount (e.g., keep U20 of the U50;
BGain[ frame) or as the amount of money lost
from the initial amount (e.g., lose U30 of the U50;
BLoss[ frame). The Bgamble[ option was
identical in both frames and was represented as
a pie chart depicting the probability of winning
or losing (Fig. 1) (14).
The behavioral results indicated that sub-
jects_ decisions were significantly affected by
our framing manipulation, with a marked dif-
ference in choices between the two frames (Fig.
2A). Specifically, and in accordance with pre-
dictions arising from prospect theory, subjects
were risk-averse in the Gain frame, tending to
choose the sure option over the gamble option
Wellcome Department of Imaging Neuroscience, Institute of
Neurology, University College London, 12 Queen Square,
London WC1 3AR, UK.
*To whom correspondence should be addressed. E-mail:
b.martino@fil.ion.ucl.ac.uk
4 AUGUST 2006 VOL 313 SCIENCE www.sciencemag.org
684
REPORTS
on April 30, 2007 www.sciencemag.orgDownloaded from

Egambling on 42.9% of trials; significantly dif-
ferent from 50% (P G 0.05, t
19
0 1.96)^,and
were risk-seeking in the Loss frame, preferring
the gamble option E gambling on 61.6% of
trials; significantly different from 50% (P G
0.005, t
19
0 3.31)^. This effect of frame was
consistently expressed across different proba-
bilities and starting amounts (fig. S1).
Reaction times for decisions were not af-
fected by frame EGain frame, 1895 ms; Loss
frame, 1884 ms (P 9 0.1)^; this result provides
evidence that difficulty was well matched be-
tween the two frames. Moreover, subjects
performed highly accurately on Bcatch[ trials
(14) (fig. S2) where the expected outcomes of
the sure and gamble options were unbalanced,
indicating their continued engagement with
the task throughout the experiment. Despite
the marked though variable impact of the
frame on subjects_ choice behavior (Fig. 2B),
the majority (16/20) of subjects seemed un-
aware of any biasing effect when specifically
questioned in a debriefing session that fol-
lowed the experiment.
Subjects performed the behavioral task
inside an fMRI scanner, allowing us to obtain
continuous measures of regional brain activ-
ity. The subjects_ individual decisions during
the entire fMRI experiment were recorded and
used to construct four regressors of interest:
sure decisions in the Gain frame (G
sure
), gam-
ble decisions in the Gain frame (G
gamble
),
sure decisions in the Loss frame (L
sure
),
and gamble decisions in the Loss frame
(L
gamble
).
Given that the frame effect relates to
subjects_ asymmetrical pattern of decisions
across frames, the key experimental contrast of
interest is the interaction between the decision to
gamble (or not) and the valence of the frame:
E(G
sure
þ L
gamble
)–(G
gamble
þ L
sure
)^.Itis
noteworthy that this interaction contrast is
balanced with respect to both decision type
and frame valence. Consequently, we could
identify brain areas that were more active when
subjects chose in accordance with the frame
effect (i.e., G
sure
þ L
gamble
), as opposed to
when their decisions ran counter to their
general behavioral tendency (G
gamble
þ L
sure
).
This contrast revealed significant activation in
the bilateral amygdala (Fig. 3, A and B). To
ensure that this activation in the amygdala was
not being driven by a significant effect in one
frame alone (e.g., Loss frame), we conducted
an independent analysis for each frame. This
confirmed that robust activation in the amyg-
dala was equally observed for simple effects of
decision type (sure or gamble) in each frame
separately. Thus, amygdala activation was
Fig. 1. The financial
decision-making task. At
the beginning of each
trial, participants were
shown a message indicat-
ing the starting amount
ofmoneythattheywould
receive (e.g., ‘ ‘Y ou receive
U50’’) (duration 2 s).
Subjects were instructed
that they would not be
able to retain the whole
of this initial amount,
but would next have to
choose between a sure
option and a gamble op-
tion (4 s). The sure option
was presented in the Gain
frame trials (A)asan
amount of money re-
tained from the starting
amount (e.g., keep U20
of the U50) and in the
Loss frame trials (B)as
an amount of money
lost from the starting
amount (e.g., lose U30
of the U50). The gamble
option was represen ted
as a pie chart depicting
the probabili ty of win-
ning (green) or losing
(r ed ) all of the star ti ng mon ey. The expected outcomes of the gamble and sure options wer e equivalent.
Gain frame trials were intermixed pseudo-randomly with Loss frame trials. No feedback concerning trial
outcomes was given during the experiment.
Fig. 2. Behavioral results.
(A) Percen tages of trials in
which subjects chose the
gamble option in the Gain
frameandtheLossframe.
Subjects showed a signifi-
cant increase in the per-
centage of trials in which
the gamble option was
chosen in the Loss frame
with respect to the Gain
frame [61.6% 9 42.9% (P G
0.001, t
19
0 8.06)]. The
dashed line represents risk-
neutral behavior (choosing
thegambleoptionin50%
of trials). Error bars denote
SEM. (B) Each bar represents,
for each individual subject, the percentage difference between how often subjects chose the gamble option in the Loss frame as compared to the Gain frame. A hypothetical
value of zero represents a complete indifference to the framing manipulation (i.e., fully ‘ ‘rational’ ’ behavior). All participants, to varying degrees, showed an effect of the
framing manipulation.
www.sciencemag.org SCIENCE VOL 313 4 AUGUST 2006
685
REPORTS
on April 30, 2007 www.sciencemag.orgDownloaded from

significantly greater when subjects decided to
choose the sure option in the Gain frame EG
sure
–
G
gamble
^EMontreal Neurological Institute (MNI)
space coordinates (x, y, z) 18, –4, –24; Z score 0
4.0^, and the gamble option in the Loss frame
EL
gamble
–L
sure
^EMNI space coordinates –16, 0,
–26; Z score 0 3.80; 12, 2, –22; Z score 0 4.67^,
in keeping with a central role in mediating the
frame effect.
A different pattern of brain activation was
identified when subjects made decisions that ran
counter to their general behavioral tendency. In
this reverse interaction contrast E(G
gamble
þ
L
sure
)–(G
sure
þ L
gamble
)^, we observed en-
hanced activity in the anterior cingulate cortex
(ACC)(Fig.3,CandD)(andtoalesserextent
in the bilateral dorsolateral prefrontal cortex at
an uncorrected threshold of P G 0.005; fig. S3)
when subjects chose the gamble option in the
Gain frame and the sure option in the Loss
frame.
In light of the substantial intersubject varia-
bility in behavioral susceptibility to the frame,
we next identified subject-specific differences in
neural activity associated with their decision bias
(that is, the decision  frame interaction) (Fig.
2A). Using the overall susceptibility of each
subject to the frame manipulation as a between-
subjects statistical regressor, operationalized as
a Brationality index[ (14), we found a signifi-
cant correlation between decreased susceptibil-
ity to the framing effect and enhanced activity
in the orbital and medial prefrontal cortex
(OMPFC), specifically in the right orbitofrontal
cortex (R-OFC; r 0 0.8, P G 0.001) and the
ventromedial prefrontal cortex (VMPFC; r 0
0.75, P G 0.001) (Fig. 4). In summary, those
subjects who acted more rationally exhibited
greater activation in OMPFC associated with
the frame effect.
Our data provide a neurobiological account of
the framing effect, both within and across
individuals. Increased activation in the amygdala
was associated with subjects_ tendency to be risk-
averse in the Gain frame and risk-seeking in the
Loss frame, supporting the hypothesis that the
framing effect is driven by an affect heuristic
underwritten by an emotional system. The amyg-
dala plays a key role in value-related prediction
and learning, both for negative (aversive) and
positive (appetitive) outcomes (15–17). Further-
more, in simple instrumental decision-making
tasks in animals, the amygdala appears to me-
diate decision biases that come from value-
related predictions (18). In humans, the amygdala
is also implicated in the detection of emotionally
relevant information present in contextual and
social emotional cues (19). It was previously
shown that activation in the amygdala during the
passive viewing of surprised faces is significantly
modulated by the valence of preceding verbal
contextual information (20). Our data extend the
role of the amygdala to include processing the
type of contextual positive or negative emotional
information communicated by the frame in the
context of a decision-making task.
In our study, activation of the amygdala was
driven by the combination of a subject_s decision
and the frame in which it took place, rather than
by the valence of the frame per se. Consequent-
ly, our findings indicate that frame-related
Fig. 3. fMRI results. (A) Interaction contrast [(G
sure
þ L
gamble
)–(G
gamble
þ L
sure
)]: brain activations
reflecting subjects’ behavioral tendency to choose the sure option in the Gain frame and the
gamble option in the Loss frame (i.e., in accordance with the frame effect). Bilateral amygdala
(Amyg) activation [MNI space coordinates (x, y, z)]: left hemisphere, –14, 2, –24 (peak Z score 0
3.97); right hemisphere, 12, 2, –20 (Z score 0 3.82). (C) Reverse interaction contrast [(G
gamble
þ
L
sure
)–(G
sure
þ L
gamble
)]: brain activations reflecting the decision to choose counter to subjects’
general behavioral tendency. Anterior cingulate cortex (ACC) activation: 2, 24, 44 ( Z score 0 3.65);
–2, 8, 56 ( Z score 0 3.78). Effects in (A) and (C) were significant at P G 0.001; for display purposes
they are shown at P G 0.005. (B and D) Plots of percentage signal change for peaks in right
amygdala (12, 2, –20) (B) and ACC (2, 24, 44) (D). Error bars denote SEM.
Fig. 4. Rationality across
subjects: fMRI correlational
analysis. Regions showing
a significant correlation
between rationality index
[between-subjects measure of
susceptibili ty to the framing
manipulation; see (14)] and
the interaction contrast image
[(G
sure
þ L
gamble
)–(G
gamble
þ
L
sure
)] are highlighted. (A)
Orbital and medial prefron tal
cortex (OMPFC) [MNI space
coordinates (x, y, z)]: VMPFC
(left panel), –4, 34, –8
(Z score 0 4.56); OMPFC
andR-OFCcircledinright
panel [R-OFC: 24, 30, –10
(Z score 0 5.77)]. Effects
were significant at P G
0.001; for display purposes
they are shown at P G
0.005. (B)Plotofthecorre-
lation of parameter estimates
for R -OFC wit h the rationality
index for each subject (r 0
0.8, P G 0.001).
4 AUGUST 2006 VOL 313 SCIENCE www.sciencemag.org
686
REPORTS
on April 30, 2007 www.sciencemag.orgDownloaded from

valence information is incorporated into the
relative assessment of options to exert control
over the apparent risk sensitivity of individual
decisions. The observation that the frame has
such a pervasive impact on complex decision-
making supports an emerging role for the
amygdala in decision-making (21, 22).
When subjects_ choices ran counter to their
general behavioral tendency, there was en-
hanced activity in the ACC. This suggests an
opponency between two neural systems, with
ACC activation consistent with the detection of
conflict between predominantly Banalytic[ re-
sponse tendencies and a more Bemotional[
amygdala-based system (23, 24).
Previous descriptions of the frame effect
have been predominantly confined to between-
subjects investigations. Our experimental design
allowed us to distinguish the anatomical bases of
the frame effect, both within and between
subjects. Interestingly, amygdala activity did
not predict the substantial intersubject difference
in terms of susceptibility to the frame effect.
Instead, subjects_ tendency to be susceptible to
the frame showed a robust correlation with
neural activity in the OMPFC. It is noteworthy
that there are strong reciprocal connections
between the amygdala and the OMPFC (25),
although each may contribute to distinct func-
tional roles in decision-making (26). Lesions of
the OMPFC cause impairments in decision-
making; these are often characterized as an in-
ability to adapt behavioral strategies according
to the consequences of decisions, leading to im-
pulsivity (27, 28). It is thought that the OMPFC,
incorporating inputs from the amygdala, rep-
resents the motivational value of stimuli (or
choices), which allows it to integrate and eval-
uate the incentive value of predicted outcomes
in order to guide future behavior (29, 30). Our
data raise an intriguing possibility that more
Brational[ individuals have a better and more
refined representation of their own emotional
biases that enables them to modify their be-
havior in appropriate circumstances, as for
example when such biases might lead to
suboptimal decisions. As such, our findings
support a model in which the OMPFC eval-
uates and integrates emotional and cognitive
information, thus underpinning more Brational[
(i.e., description-invariant) behavior.
Our findings suggest a model in which the
framing bias reflects an affect heuristic by which
individuals incorporate a potentially broad range
of additional emotional information into the de-
cision process. In evolutionary terms, this mech-
anism may confer a strong advantage, because
such contextual cues may carry useful, if not
critical, information. Neglecting such informa-
tion may ignore the subtle social cues that com-
municate elements of (possibly unconscious)
knowledge that allow optimal decisions to be
made in a variety of environments. However, in
modern society, which contains many symbolic
artifacts and where optimal decision-making
often requires skills of abstraction and decon-
textualization, such mechanisms may render
human choices irrational (31).
References and Notes
1. K. J. Arrow, Econ. Inq. 20, 1 (1982).
2. A. Tversky, D. Kahneman, J. Bus. 59, S251 (1986).
3. J. von Neumann, O. Morgenstern, Theory of Games and
Economic Behavior (Princeton Univ. Press, Princeton, NJ,
1944).
4. D. Kahneman, A. Tversky, Choices, Values, and Frames
(Cambridge Univ. Press, New York, 2000).
5. B. J. McNeil, S. G. Pauker, H. C. Sox Jr., A. Tversky,
N. Engl. J. Med. 306, 1259 (1982).
6. D. Kahneman, A. Tversky, Econometrica 47,263
(1979).
7. A. Tversky, D. Kahneman, Science 211, 453 (1981).
8. H. Damasio, T. Grabowski, R. Frank, A. M. Galaburda,
A. R. Damasio, Science 264, 1102 (1994).
9. J. Greene, J. Haidt, Trends Cogn. Sci. 6, 517 (2002).
10. G. F. Loewenstein, E. U. Weber, C. K. Hsee, N. Welch,
Psychol. Bull. 127, 267 (2001).
11. T. Gilovich, D. W. Griffin, D. Kahneman, Eds., Heuristics
and Biases: The Psychology of Intuitive Judgment
(Cambridge Univ. Press, New York, 2002).
12. P. Slovic, M. Finucane, E. Perers, D. MacGregor, in
Heuristics and Biases: The Psychology of Intuitive Judgment,
T. Gilovich, D. W. Griffin, D. Kahneman, Eds. (Cambridge
Univ. Press, New Y ork, 2002), pp. 397–421.
13. X. Gabaix, D. Laibson, in The Psychology of Economic
Decisions, Vol. 1: Rationality and Well-Being, I. Brocas,
J. D. Carrillo, Eds. (Oxford Univ. Press, Oxford, 2003),
pp. 169–183.
14. See supporting material on Science Online.
15. M. G. Baxter, E. A. Murray, Nat. Rev. Neur osci. 3, 563 (2002).
16. J. E. LeDoux, The Emotional Brain (Simon & Schuster,
New York, 1996).
17. J. J. Paton, M. A. Belova, S. E. Morrison, C. D. Salzman,
Nature 439, 865 (2006).
18. L. H. Corbit, B. W. Balleine, J. Neurosci. 25, 962 (2005).
19. R. Adolphs, Brain Res. 1079, 25 (2006).
20. H. Kim et al., J. Cogn. Neurosci. 16 , 1730 (2004).
21. M. Hsu, M. Bhatt, R. Adolphs, D. Tranel, C. F. Camerer,
Science 310, 1680 (2005).
22. B. W. Balleine, S. Killcross, Trends Neurosci. 29, 272 (2006).
23. M. M. Botvinick, T. S. Braver, D. M. Barch, C. S. Carter,
J. D. Cohen, Psychol. Rev. 108, 624 (2001).
24. E. K. Miller, J. D. Cohen, Annu. Rev. Neurosci. 24, 167
(2001).
25. D. G. Amaral, J. L. Price, A. Pitkanen, S. T. Carmichael,
in The Amygdala: Neurobiological Aspects of Emotion,
Memory, and Mental Dysfunction, J. P. Aggleton, Ed.
(Wiley-Liss, New York, 1992), pp. 1–66.
26. C. A. Winstanley, D. E. Theobald, R. N. Cardinal,
T. W. Robbins, J. Neurosci. 24, 4718 (2004).
27. A. Bechara, A. R. Damasio, H. Damasio, S. W. Anderson,
Cognition 50, 7 (1994).
28. E. T. Rolls, J. Hornak, D. Wade, J. McGrath, J. Neurol.
Neurosurg. Psychiatry 57, 1518 (1994).
29. G. Schoenbaum, B. Setlow, M. P. Saddoris, M. Gallagher,
Neuron 39, 855 (2003).
30. G. Schoenbaum, M. R. Roesch, T. A. Stalnaker, Trends
Neurosci. 29, 116 (2006).
31. K. Stanovich, F. West, in Heuristics and Biases: The
Psychology of Intuitive Judgment, T . Gilovich, D. W. Griffin,
D. Kahneman, Eds. (Cambridge Univ. Press, New York,
2002), pp. 421–440.
32. Supported by a Wellcome Trust Programme Grant (R.J.D.)
and a Wellcome Trust studentship (B.D.M.). We thank
H. Spiers, P. Sterzer, and J. Hughes for helpful discussions
during the analysis of the study and P. Bossaerts for
useful comments on the manuscript.
Supporting Online Material
www.sciencemag.org/cgi/content/full/313/5787/684/DC1
Materials and Methods
Figs. S1 to S3
Tables S1 to S3
References
5 April 2006; accepted 12 June 2006
10.1126/science.1128356
www.sciencemag.org SCIENCE VOL 313 4 AUGUST 2006
687
REPORTS
on April 30, 2007 www.sciencemag.orgDownloaded from