Phenomenological Psychology

Phenomenological Psychology header image

Emotional Networks in the Brain

October 13th, 2010 by David Kronemyer · 1 Comment

Review of LeDoux, J. & Phelps, E. (2008).  Emotional networks in the brain.  In M. Lewis, J. Haviland-Jones & L. Barrett (Eds.), Handbook of emotions (pp. 159 – 179).  New York, NY: Guilford Press.

Summary of the Article

In this article, LeDoux and Phelps (2008) review the current state of research into neural and anatomical correlates of emotional experience.  The authors situate themselves in a tradition of theorizing about emotions dating back to Aristotle and Descartes and, in modern times, from Charles Darwin, William James and Sigmund Freud (for a brief history, see Oatley, Keltner & Jenkins, 2006).  LeDoux also is the author of The Emotional Brain (1996), one of the first popular works on the subject.

The authors hypothesize that the brain processes emotions using two separate neurophysiological circuits.  The first one primarily relies upon subcortical brain structures, such as the sensory thalamus and the amygdala.  The authors adopt a specialized, functional definition of the term “emotion,” using it to refer to the resulting output state.  Metaphorically, they characterize this as the “low road” (LeDoux & Phelps, 2008, p. 159).  In contrast, the second pathway primarily deploys higher-order cortical brain structures, such as the sensory cortex.  The authors define the outcome of this process as a “feeling,” and call it the “high road” (LeDoux & Phelps, 2008, p. 159).

In the case of an emotion, the brain receives a disruptive input stimulus.  A charging lion, for example, might activate “fear.”  This perception is routed directly to the amygdala through the sensory thalamus.  The amygdala in turn activates immediate output behavior in the form of an autonomic motor response – say, fleeing (LeDoux & Phelps, 2008).

Feelings, however, require higher-order cognitive processing.  Following receipt by the sensory thalamus, information about the stimulus is routed to the sensory cortex.  Using “working memory,” the brain then reviews the information and transmits a conclusion back to the amygdala (such as, “this is a charging lion, so I should be afraid”).  Describing this process, the authors state:

“Immediately present stimuli and stored representations are integrated in working memory by way of interactions among prefrontal areas, sensory processing systems (which serve as short-term memory buffers as well as perceptual processors), and the long-term explicit (declarative) memory system involving the hippocampus and related areas of the temporal lobe.  Working memory may involve interactions between several prefrontal areas, including the anterior cingulated and orbital cortical regions, as well as dorsolateral prefrontal cortex.” (LeDoux & Phelps, 2008, p. 167).

This compare-and-contrast review is more complex than a visceral emotional response.  It is a person’s “subjective experience” of the emotional state  (LeDoux & Phelps, 2008, p. 167).  Checks-and-balances between the amygdala and the cortex permit differential, contextual analysis of stimulus (such as, “it’s not really a threatening lion after all”) and subsequent mediation or modulation of emotional response (e.g., “I need not flee, because I am not actually in danger”), LeDoux and Phelps (2008).  It would be a mistake, however, to characterize the low road as “unconscious” and the high road as “conscious.”  “Instead, there is one mechanism of consciousness, and it can be occupied by either mundane events or emotionally charged ones” (LeDoux & Phelps, 2008, p. 168).

The authors oppose their theory to what they characterize as the received or “standard criterion” interpretation of emotions, as resulting from or being generated by distributed processing activity integrated throughout the limbic system (LeDoux & Phelps, 2008, p. 157).  The limbic system comprises a collection of anatomical regions in the brain, including not only the amygdala, but also the thalamus, the hippocampus and other subcortical structures regulating autonomic responses such as blood pressure, heart rate and attention.  The authors criticize the concept of the limbic system as being more of a theoretical construct, rather than an actual place in the brain where neural processing activity occurs.  According to the authors, the hippocampus (for example) actually is bypassed, as potentially emotion-causing stimuli are received directly by the amygdala, and then translated into output behavior (LeDoux & Phelps, 2008).

The authors’ model makes it possible to explain phenomena such as timing differentials between emotions and feelings.  While the routing is different, the amygdala is the destination in either case.  According to the authors, however, the upwards neural circuitry from the amygdala to the cortex is better-developed than the downwards neural circuitry from the cortex back to the amygdala.  From a clinical standpoint, the timing differential between response and appraisal often may seem minute.  However there is a significant class of cases where both their temporal sequence and duration is significant.  For example, LeDoux and Phelps’ (2008) model helps explain the persistence of emotional activity (generated by the amygdala), despite cognitive inhibition (e.g., “I know I shouldn’t feel this way, even though I do”).  And, recent research in decision theory has shown that the outcome of a decision is encoded in the prefrontal and parietal cortex up to ten seconds before it enters conscious awareness (Soon, Brass, Heinze & Haynes, 2008).  The “outcome of a decision” may be analogous to the cognitively-mediated “feeling” hypothesized by LeDoux and Phelps (2008).

How the Article Can Be Useful to a Mental Health Professional

LeDoux and Phelps’ (2008) article might be useful to three different categories of mental health professionals: the neuropsychologist; the psychodynamic psychotherapist; and the cognitive behavioral therapist.

The neuropsychologist.  For the neuropsychologist, LeDoux and Phelps (2008) present a straightforward way of examining the neuroanatomical and neurochemical activity involved in human emotional response.  Other authors in the field, such as Antonio Damasio (1999) and Candace Pert (1997), have relied extensively on LeDoux’s work.  LeDoux and Phelps’ (2008) article has implications for other vexing neuropsychological phenomena, such as the characteristics of attention and intentionality; the relationship between brain chemistry and phenomenological experience; and even the nature of human consciousness.

The psychodynamic psychotherapist.  LeDoux and Phelps’ (2008) work also provides a challenge for the psychodynamic psychotherapist, who primarily is concerned with a client’s phenomenological experience.  The main issue with the two-pathway theory, and its related distinction between “emotions” and “feelings,” is that it denies the subjective mental reality of psychological phenomena (such as feelings), instead seeking to reduce them to something materialistic (neural events or anatomical features of the brain), or preferably eliminate them altogether.  It is important for the psychodynamic psychotherapist to understand this dynamic and its clinical implications.

For the psychodynamic psychotherapist, properly considered, the two-pathway theory rests on a basic category mistake.  LeDoux and Phelps (2008) assume science only can study “objective” phenomena, such as the neurophysiology underlying emotions.  Science is “epistemically objective” in that it uses replicable methods and procedures to obtain and verify results.  Brain anatomy and brain chemistry in turn are “ontologically objective” in that they are about discoverable facts, which exist independently of any particular person’s mental or psychological state.

In this sense, LeDoux and Phelps (2008) imply any systematic study of “feelings” is unscientific.  Feelings are “epistemically subjective” because the person who is having them is the only one who can experience them.  They also are “ontologically subjective” because they are real only for that person; my feelings aren’t yours, and vice versa.  They might be revealed, for example, by a person’s behavior, or described by a person’s self-report.  They cannot, however, be independently accessed or observed.  [This schematic comes from Searle (1997), who used it however in a different context..]  This being so, there is no reason to conclude there cannot be an epistemically objective science about mental phenomena that are ontologically subjective.  That science is called “psychology,” which is something the two-pathway theory completely overlooks.

Here is an example of the problem.  Although they do not address it, LeDoux and Phelps (2008) would be committed to the proposition one could not “have” a feeling (ontologically subjective), unless it is possible to explain (epistemically objective) the correlative neurophysiological state (ontologically objective).  Thus, they might describe “pain” as a certain class of sensory signals (say, c-fiber firings), which in principle can be identified and described, and which in principle are the same for everyone.  They might contrast this with (say) “suffering,” which is the (ontologically subjective) feeling associated with pain.  Then, they might say, “you cannot actually be suffering, unless you are in that certain neurophysiological state, which we call pain, and we can tell that is the case.”  This is dubious because, for the person who is in it, pain is just pain; it has a qualitative aspect, which does not depend for its sufficiency or necessity on an accompanying neurophysiological explanation.  The reciprocal also might be true; one might feel as if one is in pain, without being in the requisite underlying neurophysiological state.

LeDoux and Phelps (2008) might continue, “On the other hand, you might just think you’re in pain” (epistemically subjective), without conceding you actually are (ontologically subjective).  They might even go so far as to say, “you can’t actually be suffering” (ontologically subjective), “unless you can cognitively identify the state you are in” (epistemically subjective), “as the one uniquely resulting from that certain neurophysiological state” (ontologically objective), “which we call pain” (epistemically objective).  After all, this is the whole point of the cortex-to-amygdala feedback loop.  This also is suspect, because being in pain does not depend on being able to recognize (much less explain) one’s status as pain-experiencer.  The psychodynamic psychotherapist will need to devise a schematic to address these apparent contradictions, in order to be as useful as possible to one’s clients, all of who initially present with what essentially are cognitive issues.

The cognitive behavioral therapist.  The cases discussed above do not rely on sleight-of-hand or obscure semantic distinctions.  They also underlie the usefulness of proven clinical methods, such as cognitive behavioral therapy (CBT).  The basic premise of CBT is that thought processes precede emotions and then subsequent behavior.  Thus, one experiences an activating event as a trigger or real-world adversity.  One then has, or develops, an orientation or perspective towards the activating event.  Negative feelings or behavior in turn result from, or are a consequence of, this belief-set (not from the activating event itself).  It follows that if the dysfunctional belief can be cognitively restructured, then the negative feelings or behavior will disappear, or at least diminish.  Therapeutically, this process requires identifying the belief operationally; disputing or refuting it; creating alternatives to it; collecting evidence against the original belief and in favor of the alternative formulation; and then modifying behavior to support this replacement.

Different CBT theorists have contrived different names for the “cognitive” part of this equation.  For Ellis it was “irrational;” Beck, part of a dysfunctional “schema;” Burns, a “cognitive distortion;” Seligman, an aspect of “explanatory style”.  What they share is the concept that once one has become aware of or achieved some degree of insight into one’s thoughts and thinking patterns, then they can be cognitively restructured to eliminate adverse emotional associations and correlative behavior.

The reason why this presents an issue is because it is the exact reverse of the two-pathway theory, which holds that emotions precede thoughts.  If LeDoux and Phelps’ (2008) position were correct, then CBT would flounder, because there would be nothing to be cognitive about.  One constantly would be adrift on a sea of emotions, each competing to be understood qua “feelings” as they swim upstream from the amygdala to the cortex, and then back down again.  While these timing differentials may be short, there nonetheless is a wide class of cases in which they may be significant, particularly when it comes to the long-term gestation of feelings, in turn generated by emotions (supra, page 3).  It is something of an embarrassment for such a significant discrepancy to exist at the junction of neuropsychology and CBT, particularly considering CBT’s demonstrated effectiveness in treating a wide range of psychopathologies.


Damasio, A. (1999).  The feeling of what happens.  San Diego, CA: Harcourt.

LeDoux, J. & Phelps, E. (2008).  Emotional networks in the brain.  In M. Lewis, J. Haviland-

Jones & L. Barrett (Eds.), Handbook of emotions (pp. 159 – 179).  New York, NY: Guilford Press.

LeDoux, J. (1996).  The emotional brain.  New York, NY: Simon & Schuster.

Oatley, K., Keltner, D. & Jenkins, J. (2006).  Understanding emotions.  Malden, MA: Blackwell.

Pert, C. (1997).  Molecules of emotion.  New York, NY: Scribner.

Searle, J. (1995).  The mystery of consciousness.  New York, NY: New York Review.

Soon, C., Brass, M., Heinze, H. & Haynes, J. (2008).  Unconscious determinants of free decisions in the human brain.  Nature Neuroscience, 11(5), 543 – 545.  doi:10.10.1038/nn.2112