Comment: What’s Basic About the Brain Mechanisms of Emotion?

Psychologists often think that neuroscientists like me believe that emotions are burned into innate circuits by evolution—that there are innate circuits in the human brain that have been inherited from animal ancestors and that when activated lead to the expression of hard-wired emotional responses and the unleashing of a corresponding feeling of fear. But this is not an accurate representation of my views. In this commentary, I will outline my view of emotion and then show points of intersection and divergence with the four articles under consideration.

I have long maintained that while hard-wired circuits control defensive responses like freezing and associated physiological changes in the body in the presence of a threat, the same circuits are not directly responsible for consciously experienced feelings of fear elicited by the threat. Nevertheless, my work is often described as being about how the brain feels fear.

In an effort to figure out how to separately talk about this difference between circuits that detect and respond to threats and circuits that feel fear, I have, in the past, resorted to using the term emotion to refer to the hard-wired stimulus–response mechanisms and have reserved the term feelings for the conscious experiences that sometimes accompany these kinds of responses. But this strategy has not solved the problem. When you use a word like fear, most people, whether they are scientists or lay people, will think you are talking about the feeling of being afraid. We need another way to talk about this stuff.

I think the solution is to stop struggling over the meaning of emotion words like fear (LeDoux, 2014). Let’s let emotion be what we all think it is when we take off our scientific hats: the feeling that makes itself known to consciousness when we are in a situation in which we face a challenge or opportunity. So I am now treating emotions and feelings as the same thing—conscious experiences. I label innate circuits that control responses to significant stimuli (like stimuli related to danger, maintaining energy supplies and fluid balance, and reproduction) without reference to emotion—I call these survival circuits and argue that they did not evolve to make emotions but to give organisms behavioral tools to stay alive (LeDoux, 2012). This eliminates the confusion that results when emotion theorists try to figure out how emotions and feelings relate, whether emotions or feelings can be both conscious and unconscious, whether some of these states of consciousness are hard-wired and others not, and how conscious emotions/feelings relate to innate behaviors.

It is true that we often experience feelings of fear when we express defensive behaviors like freezing or flight in the presence of a threat. However, people can express responses to threatening stimuli without being aware of the stimulus and without reporting any particular feeling. If we don’t need emotional consciousness to explain emotional expressions in humans, we should not be calling upon emotional consciousness to explain behavior in animals, and should not assume that the circuits that control animal behavior are the same circuits that make human feelings (LeDoux, 2014).

This does not mean that animal research is meaningless for understanding human emotions. Through animal research we can, and have, identified circuits that account for a variety of behavioral responses and physiological adjustments in the brain and body. These survival circuits collectively give rise to global organismic states (LeDoux, 2012). In the case of a threat, a defensive survival circuit is activated and gives rise to defensive organismic states. Defensive kinds of behaviors are made more likely and other forms of motivation are suppressed. Sensitivity to incentives related to threats. Most if not all complex organisms, including many invertebrates, have some form of these global organismic states. But only organisms that can be consciously aware that these states exist can feel an emotion based on such a state (LeDoux, 2012, 2014). The extent to which these global states, as opposed to the component processes (such as defense circuit activity, body feedback, brain arousal) have a causal role is unknown.

I have studied defensive behavior using a task called Pavlovian threat conditioning (usually called fear conditioning) to implicate specific circuits. These are often thought of as THE defense circuits. When other tasks are used, other defense circuits are involved. There is no single module in the brain that accounts for all instances of defense (which is often taken to mean fear). One must be especially careful in interpreting human fMRI data, since this technique cannot make fine distinctions of this type. So evidence that “the amygdala” is involved in two tasks does not mean that the two tasks have the same neural requirements.

Not all emotions depend on global organismic states. But they all depend on some kind of conscious appreciation of events in the outer world and/or within organism. In this view, all emotions are fundamentally the same. They are products of cognitive processes. This is an unavoidable conclusion once it is realized that even fear, the most basic of basic emotions, does not require some prepackaged defensive motivational state. You cannot only be afraid of a snake at your feet but by also the possibility that you may die without having led a meaningful life. The only thing that is similar in these diverse states is that they involve some cognitive engagement in the situation.

A big part of the problem in emotion research, in my opinion, is that we start from our introspections about emotions and then go looking for states and processes in the brains of humans and animals. This folk psychology approach, while not ideal, has some value in human work, but is actually detrimental to progress in understanding what animals can teach us about human emotion. We should not be asking whether we can find human emotions in animals but instead we should be asking which neural processes in animals are conserved in the human brain.

With regards to conscious emotions (feelings), my view has been pretty consistent for a long time (LeDoux, 1996, 2002, 2008, 2012, 2014). This can be summarized by way of analogy to the way the character of a soup arises from its ingredients. None of the ingredients are soup ingredients. They are things that exist in nature and that can be used in soups of various kinds, and in other kinds of dishes as well. But the particular combination of ingredients gives the soup its character. Similarly, nonemotional ingredients (sensory activation of a survival circuit that produces arousal in the brain and body responses that feed back to the brain; retrieval of relevant implicit and explicit memories; cognitive processes such as attention, monitoring, appraising, prediction) come together to give rise to an emotion, a feeling. The feeling can be imposed on consciousness in bottom-up fashion or can be based on top-down interpretation of what is going on. This view is consistent with cognitive theories of consciousness and with cognitive theories of emotion.

I think that the picture I’ve painted here resonates well with the four target articles. I obviously would not use the term “emotion” the way these authors do. But I’ll overlook the semantic difference for now and just comment on the conceptual linkages.

There is a great deal of overlap between my views and Barrett’s (2014) conceptual act theory. For example, we agree that emotions like fear are not prepackaged entities in the brain but are cognitively assembled from other processes. This is why there can be so many kinds of fear and why there is no one physiological maker in the brain or body that defines fear. But where we may differ is on the degree to which the feeling of fear can be imposed on consciousness in a bottom-up way from implicit ingredients, as opposed to always being constructed through conceptual acts. Another issue is whether fear is a physical state of the brain that can have a causal role once it exists. I go that way. Barrett claims it is a functional rather than a physical state.

I have often written about my work in terms of appraisal. Brain circuits that detect and respond to threats must have some filter (appraisal mechanism) to decide what is threatening. Where things get fuzzy is in the role of “mental” states in appraisal. I prefer an expression like “threat detector” than “fear appraisal” to break with the folk psychological implications. My global organismic state idea was influenced by Scherer’s (2014) idea that emotions play a coordinating role in organizing component processes. Moors’ (2014) idea that conscious experience is determined by components is compatible with my idea that feelings are amalgams assembled from nonemotional ingredients.

As my previous comments indicate, I believe that evolutionary-based functions contribute to emotions, but not in the way that basic emotions or other evolutionary-based theories assume. I argue that evolution has not given us emotions, but instead behavioral survival tools built into circuits. So I disagree with ideas such as those by Tracy (2014) that emphasize innate modules for fear and other emotions. This may in part be a semantic difference, but that is why we need a different language for circuits that give rise to felt experiences as opposed to circuits that control bodily responses in the face of threats or other significant experiences. However, Tracy’s idea about processes that provide contingent control of behavior is compatible with the global organismic idea. I view these states as permissive rather than causal. When a threat occurs, resources are marshaled in an effort to stay alive, raising the likelihood that defensive motivational states will be expressed and decreasing the likelihood of behaviors related to other forms of motivation.

Some may be surprised that I also find my views to be compatible with the sociodynamic model of Mesquita and Boiger (2014). We obviously work at different levels of analysis—I deal with the role of neurons, synapses, and molecules in animal behavior, and they with social factors in human behavior and experience. But like them, I agree that emotions do not always have rigidly fixed expressions, and that expressions are tailored to situations—that the way you act when you are afraid varies depending on the context in which you find yourself. But I believe this applies more to the second wave of behavior (the instrumental responses) than to the initial and fairly automatic responses that are more or less hard-wired (though the latter may be regulated to some extent by preexisting context before they are expressed, or short-circuited in a top-down fashion as they are being expressed).

As psychologists we all know the power of biases that can arise from implicit or explicit presuppositions. I hope psychologists who have supposed that I think that fear is burned into the brain will see my work in a new light.