The the past several decades, empirical findings from cognitive

evolutionary and cognitive role of emotion has long been a subject of inquiry
and debate. Within the past several decades, empirical findings from cognitive
neuroscience research have helped illuminate several testable theories as to
the purpose and importance of emotion. Naturally, such theories have given rise
to much debate in the scientific community. In this paper, I will first briefly
discuss the theme of the role of emotion as a device in decision-making
processes. Next, I will describe and explain the somatic marker hypothesis, and
provide empirical evidence that supports the theory. Then, I will discuss and
provide empirical evidence to support the competing theory that explicit
knowledge is enough for decision-making, and that reversal learning is
responsible for variation in gambling tasks. Finally, I will argue that the
evidence supporting the somatic marker hypothesis is convincing, and that
competing theories have several shortcomings and do not falsify findings in
support of the somatic marker hypothesis.

            For much of human history, the discussion of the role of
emotion as a part of decision-making has been reserved for philosophers. From
Descartes’ mind-body dualism to Hume proclaiming reason to be “slave of the
passions” (Hume, 1738) and beyond, there has existed a seemingly intuitive separation
between emotions and rational decision-making. The late-nineteenth and
early-twentieth century saw this debate shift from philosophy to psychology, as
several theories of emotion were developed. These include the James-Lange
feedback theory, which argues that a stimulus causes a bodily response that
then in turn feeds back to the brain to feel emotion, and the Cannon-Bard
theory, which posits that a stimulus elicits a parallel-processing of emotions
and bodily responses directed by the thalamus and hypothalamus. However, these
theories simply argued for different mechanisms by which emotion may be
produced, and did not suggest that emotions necessarily played a role in
decision-making. The role of emotions would be explored further by cognitive
neuroscientists studying fear, specifically fear conditioning and the role of
the amygdala and prefrontal cortex in learning (Maren & Quirk, 2004). From
this work, it was observed that damage to the orbital and ventromedial
prefrontal cortex not only impaired emotional responses, but also drastically impaired
decision-making. If emotions were truly isolated from processes of rational
reasoning, then emotional deficits should have improved, and not worsened,
decision-making ability. Ultimately, these findings led to the somatic marker

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

            Theorized by Antonio Damasio, the somatic marker
hypothesis (SMH) essentially argues that rational decision-making is not void
of emotion but rather is inseparable from emotional processing. The SMH argues
that emotions are a critical part of the systems neuroscience mechanism by
which all decisions are made. Specifically, the theory states that biological
marker signals, such as autonomic and endocrine changes, express themselves as
emotions and help quickly signal prospective consequences and outcomes of
actions (Bechara & Damasio, 2004). Thus, instead of a utility theory driven
cost-benefit analysis, decisions are made using an affective system heuristic which
simulates consequences via emotional markers.

further differentiates between so-called primary and secondary inducers of
emotion. Primary inducers of emotion are natural or learned stimuli signaling
pleasurable or aversive responses directly through a body loop, heavily
mediated by the amygdala (Bechara & Damasio, 2004). For example, seeing a
wasp automatically initiates fear. Alternatively, secondary inducers may
simulate the response through an “as-if” loop which activates a similar
somatosensory pattern in regions of the brain such as the insula, yet does not
necessarily result in the entire body undergoing the visceral response (Bechara
& Damasio, 2004). Damasio argues that this process is mediated by the
ventromedial prefrontal cortex (vmPFC), which links secondary inducers with somatic
state pattern responses. For example, imagining a wasp may induce a similar
activation pattern and emotional response as seeing a real wasp, without
resulting in the entire body undergoing the same emotional response. PET
studies have supported the idea that emotions are linked to somatic bodily
states, as when asked to recall happy, sad, or fearful life memories, subjects
activated regions such as the somatosensory cortices (Damasio et al., 2000).
This phenomenon is similar to somatic feedback in the James-Lange theory of
emotion, and suggests that emotions activate and take into account internal
bodily states.

and colleagues have used the Iowa Gambling Task (IGT) as a paradigm by which to
test the SMH. The IGT consists of subjects having to choose between two decks
of cards, one with initial high gains but ultimately high losses, and another
with initial low gains, but ultimately low losses and higher gains overall. The
findings reveal that relative to control subjects who perform increasingly well
as more cards are selected by choosing more from the advantageous deck, vmPFC
and amygdala lesion patients fail to avoid the bad deck and perform poorly
overall (Bechara & Damasio, 2004). In order to determine the cause of this
phenomenon, skin conductance responses (SCRs) were taken while the subjects
were involved in the task. After a few rounds of card selection, patients began
to form anticipatory SCRs prior to card selection. Notably, amygdala and vmPFC
patients had practically absent anticipatory SCRs compared to control subjects.
Thus, Damasio hypothesized vmPFC and amygdala patients were unable to make
advantageous decisions because they lacked the emotional somatic markers
necessary to guide their decision (Bechara & Damasio, 2004).

other studies further corroborate Damasio’s initial findings. A follow-up study
revealed a dissociation between conscious knowledge and decision-making, as
vmPFC patients were shown to sometimes know and acknowledge that they were
about to make a risky or poor decision, yet would continue to do so anyway
(Bechara et al., 1997). This demonstrates how conceptual knowledge is not
enough to guide decision-making and shows that the vmPFC lesion is not just
impairing general task understanding. Additionally, this further suggests that
emotional signaling is truly necessary to making advantageous choices. Variants
of the IGT that increased future punishment or decreased future reward in
disadvantageous decks also failed to change vmPFC patient behavior, further
supporting the notion that such patients exhibit a “myopia for the future” due
to their lack of consequence assessment (Bechara et al., 2000). Brain imaging
studies using fMRI during the IGT were also consistent with the SMH, as the
dlPFC (working memory), insula and PCC (emotional patterns), OFC and vmPFC
(matching memories with emotional patterns), and SMA (executing actions) were
each selectively active (Li et al., 2011). Thus, the neural circuitry suggested
by the SMH appears to be activated during the IGT, where working memory and
somatic states are integrated by the PFC and then executed by the motor system.
More generally, event-related imaging studies have also bolstered the idea that
risk anticipation is essential to performance on the IGT, and that even in
healthy subjects, increased activity during risky decisions is correlated with
better performance on the task (Fukui et al., 2005). Ultimately, these studies
each suggest that emotions play a vital role in decision-making, and likely
evoke somatosensory patterns that help guide decision-making processes in
accordance with the SMH.

there exist several critiques and alternative explanations to the SMH. Maia and
McClelland performed a study in which they asked participants in the IGT more
specific questions as to their awareness of the expected outcomes of the decks
they were to choose from (Maia & McClelland, 2004). They found that when
participants behaved advantageously, they not only consciously knew about the
relative expected outcomes of the decks, but also reported quantitative
knowledge that would have been sufficient for making their decision. Therefore,
they could have relied on explicit knowledge rather than implicit somatic
markers to guide their decisions. Maia and McClelland argue that this
alternative explanation of using overt, explicit knowledge to guide decisions
is particularly likely in the self-paced and numerical IGT, as there would be
no need for fast-paced implicit heuristics (Maia & McClelland, 2004). They
further argue that the SCRs present in control patients and absent in vmPFC
patients in Damasio’s study are simply a result of emotions elicited by
conscious knowledge, and there is no evidence to suggest that the SCRs are
indicative of a causal role of emotion (Maia & McClelland, 2004). Thus,
Maia and McClelland argue that conscious knowledge guides behavior, rather than
unconscious implicit processes such as somatic markers.

vmPFC does not ubiquitously impair decision-making, further calling into
question the SMH. Fellows and Farah employ a variant of the IGT in which they
shuffle the advantageous and disadvantageous decks, so that there will be no
initial preference for the disadvantageous deck. The results of the shuffled
IGT were that vmPFC patients performed almost identically to control subjects,
while dlPFC patients continued to perform poorly (Fellows & Farah, 2005).
This differs significantly from Damasio’s findings of poor performance by vmPFC
patients in the original IGT, and is unexpected, as vmPFC patients would still
be expected to perform badly due to their lack of somatic markers. Fellows and
Farah argue that the shuffling task was able to eradicate differences between
the vmPFC patients and the controls because vmPFC patients were affected not by
an inability to form somatic markers, but by a deficiency in reversal learning.
Reversal learning is when an individual must learn to reverse a learned
association. Fellows and Farah suggest that the vmPFC is essential to reversal
learning, and that its damage results in poor decision-making when individuals
must reappraise learned decisions, such as in the original IGT. Maia and
McClelland further argue that reversal learning itself does not need to be
mediated by somatic markers (Maia & McClelland, 2005). Instead, they argue
that the vmPFC has direct neural connections to the striatum, and that lesions
to these regions impact reversal learning identically to lesions of the vmPFC.
Therefore, the vmPFC controls reversal learning directly via connection to the
striatum rather than by generating somatic markers. Instead, Fellows and Farah
argue that decision-making in the shuffled IGT is truly impaired by the dlPFC, which
is central to working memory. This idea is corroborated by the work of Hinson
et al., who demonstrates that IGT performance declines as working memory load
increases. While this does not falsify the SMH, it suggests that
decision-making variations in the IGT by vmPFC patients may be explained by
reversal learning independent of somatic markers, and that decision-making
deficits may be linked more closely to the ability to access explicit,
conscious, non-emotional information such as that in working memory.

opposing theories, I believe the SMH to be a convincing theory regarding the
role of emotion in decision-making. First, it is clear from Bechara’s 1997
follow-up study of the original IGT experiments that Maia and McClelland’s
theory that overt, conscious knowledge drives decision-making is likely
misguided. In the follow-up study, it was demonstrated that even though
disadvantageously playing vmPFC patients knew that they were going to perform
poorly, they continued to make disadvantageous decisions. Thus, simply knowing
explicitly or stating the advantageous decision to make is not enough to
actually make the advantageous decision.

the use of emotions as a decision-making tool makes sense in an evolutionary
sense. In order to maintain learned associations, using an “as-if” loop to
simulate consequences would be both a fast and salient method by which to
effectively drive decision-making. For example, it would certainly be
advantageous to feel disgust when thinking about a poisonous fruit by
simulating the feeling of nausea, rather than having to actually see or eat
that fruit. In a general sense, it simply would not make sense for emotions to
either be outside of or a hindrance to advantageous decision-making processes.
Neuroscience research on the amygdala, prefrontal cortex, insula, and various
other regions of the brain all demonstrate how processing of emotions such as
fear and disgust may lead to effective learning and thus advantageous
decision-making. Thus, it would follow that emotions should be an integral part
of decision-making.

response to the Fellows and Farah study, it may be reasonable to question the
role of the vmPFC in integrating working memory with internal bodily states, as
it appears that the vmPFC is more directly utilized in reversal learning in the
IGT. This work reveals that there are undoubtedly still many unanswered
questions about the SMH. In order to make the theory more parsimonious, it
would be helpful to try and identify if different neural systems are used under
different conditions, such as having to make fast versus slow decisions. In
addition, one could test if some emotions are conducive to advantageous
decision-making, while others could be harmful. Although the SMH is a good idea
in theory, it will garner wider support once more discrete neural pathways and
systems are identified for specific situations and emotions under which
individuals make decisions. However, the overall theory that emotions are
utilized by triggering somatic patterns that reflect prospective consequences
of choices is not only viable, but also a likely role that emotions play in

the role of emotion in decision-making remains a highly tested and debated
theme in cognitive neuroscience. The somatic marker hypothesis and empirical
evidence from the Iowa Gambling Task seems to suggest that emotions are used to
simulate consequences, and thus are integral to and guide decision-making.
Competing theories suggest that implicit somatic markers are unnecessary, and
that decisions can be made by simply using explicit conscious knowledge.
Additionally, evidence suggests that the vmPFC is not essential to integrating somatic
states and working memory as hypothesized by the SMH, but rather is used in
reversal learning. Finally, I argue that the SMH is a convincing theory, as it
provides an evolutionarily beneficial role for emotion in decision-making, and
explicit advantageous knowledge fails to directly link to advantageous choices,
thus suggesting that implicit processes must play a role in decision-making.