How Emotions Are Made: The Secret Life of the Brain

The author recalls performing research at the University of Waterloo, where she studied anxiety and depression. After eight experiments, she failed to demonstrate that study participants were more depressed after failing their own standards and more anxious when letting down others. Instead, she found that her study subjects almost never distinguished between feeling anxious and depressed. Continuing her research, Barrett noticed a similar phenomenon in her months-long daily survey of hundreds of subjects’ emotions: Many participants struggled to distinguish one emotion from another. Others, however, were more specific about their feelings, which Barrett refers to as “emotional granularity” (3). At the time, Barrett perceived these subjects as having a more sophisticated self-awareness of their emotional states and wondered if she could teach this self-perception to others. To understand the accuracy of people’s judgment about their own emotions—for example, which feeling should be labeled joy, sadness, or excitement—Barrett needed a way to “measure an emotion objectively” (3). She was confident that she’d be able to analyze people’s facial expressions, bodily reactions, and brain activity to create a reliable “fingerprint” or identifier for each separate emotion; however, this wasn’t the case.

Barrett cites Silvan S. Tomkins’s experiment in the 1960s, which showed that people from a variety of cultural backgrounds could correctly match emotions with photos of facial expressions, suggesting that emotional responses are universal. Another technique, facial electromyography (EMG), allows scientists to precisely measure people’s facial muscles as they experience different emotions. The author notes that, unlike Tomkins’s study, EMG didn’t reliably predict which emotion participants were experiencing; it could only distinguish between positive and negative emotional experiences. She then cites studies that showed that people interpret others’ emotions based on context as much as facial expression. Barrett studied this interpretation herself; in one study, she showed two groups of participants a photo of a man and told one group that he’d just witnessed a shooting. The participants who thought the photographed man had seen a traumatic event were more likely to interpret the man’s expression as fearful, while the other group thought he looked surprised. The author claims that in interpreting facial expressions and emotional experiences, “variation is the norm” (11).

Because Barrett couldn’t reliably associate facial expressions with specific emotions, she attempted to measure bodily signals, such as blood pressure, to find emotions’ “unique fingerprint.” She cites a 1983 study by Ekman, Levenson, and Friedman that measured participants’ sweat, heart rate, and temperature while they made facial expressions associated with certain emotions. When the scientists observed people’s biological responses changing, they concluded that these were “objective, biological fingerprints” (12) for emotions in humans. However, Barrett points out that their method of using “facial feedback” isn’t a well-founded way to induce certain emotions, and this experiment didn’t yield the same results in a remote community in Sumatra. During the last two decades, several analyses of studies on physical reactions to emotion failed to identify “consistent and specific emotion fingerprints in the body” (14) but showed that bodily reactions to experiencing emotions can vary. Barrett explains that these studies don’t show that emotion isn’t “real” but instead demonstrate that emotions prompt different bodily reactions in different people.

Barrett next explores the Darwinian notion of “population thinking,” which suggests that a certain species is composed of unique individuals who don’t necessarily share an identical “fingerprint.” From this perspective, variation in emotions is normal, and Barrett decided to research the brain more closely.

Scientists who studied people with brain damage to try to understand the physical foundations of emotions identified the amygdala as the brain region responsible for processing fear. However, some studies raised questions about this claim; for example, twins who both had no amygdalas experienced fear very differently from each other. Barrett cites this example to support her claim that the brain must “have multiple ways of creating fear” (18), which include the actions of different kinds of neurons. Barrett argues that the brain has “core systems” that contribute in different ways to creating emotion, challenging the more simplistic classical perspective on emotion, which ties certain functions to specific regions of the brain. The author explains that activation of these core systems can be verified through functional MRI (fMRI) scans, which reveal brain activity. She notes that fMRI technology has been used in many brain studies to help identify the physical origin of emotions in the brain. However, Barrett argues that these studies only reinforced the notion that many parts of the brain collaborate to produce emotional responses. Barrett found that “[e]motions arise from firing neurons, but no neurons are exclusively dedicated to emotion” (22). Barrett summarizes studies and meta-analyses that produced much evidence against the classical view of emotion, and she encourages an acceptance that “variation is the norm” (23) in human emotional responses. As such, emotions such as anger, fear, and happiness can be understood as categories that prompt a variety of physical and brain-related activity, rather than specific, one-size-fits-all experiences.

Barrett explains that some software programs can correctly identify emotional reactions by analyzing patterns from brain imaging. She concludes that this “pattern classification” or “neural mind reading” shows that artificial intelligence (AI) can accurately summarize and interpret people’s varied brain activity. She argues that this is an example of “population thinking” at work, in the same way that despite differences among all Labradors, we can distinguish retrievers. She concludes her chapter by reiterating that by closely examining the evidence on the origin of emotions, we can better comprehend ourselves.

The author shares a visual exercise: First, she presents a partial image with black and white shapes that don’t reveal a specific whole. Then, she reveals the full photograph that contains the shapes, revealing that the image is of a bee. She points out that given the context of the whole image, the shapes in the original image are clearly part of a “bee” shape. Barrett notes that this exercise prompts a form of “hallucination” in the mind, which shows that experiences color current perceptions. This phenomenon, called “simulation,” occurs in an instant and is responsible for helping create associations and even sing-songs in the head. Barrett recalls hosting a children’s birthday party during which she playfully triggered simulations by associating tasty food with disgusting associations; for example serving yellow juice in medical urine cups and brown baby food in diapers. The author explains that simulations are the brain’s way of helping us navigate the world and prepare for experiences. She expands on this idea by referring again to the example of a bee; people’s brains produce different simulations depending on their past experiences with bees, which may prompt positive associations like flowers and gardens, or negative ones like annoyance or bee stings. Barrett writes that everyone has a certain “concept” of a bee, which is “actually a collection of neural patterns in your brain, representing your past experiences” (28). These concepts apply to everything, and cultural influence on them is significant. For example, people accustomed to listening to Western music and the 12-tone scale often interpret gamelan music from Indonesia as “noise” rather than music.

The author argues that our brains may produce “emotion concepts” that depend on our circumstances. For example, the brain may interpret an achy stomach at the dinner table as hunger but interpret the same kind of ache as nausea during flu season or as a lovesick feeling when in a relationship: “Your brain makes meaning from your aching stomach, together with sensations from the world around you, by constructing an instance of that concept. An instance of emotion” (30).

Barrett builds on this idea to claim that, in contrast to the classical view of emotions, they’re “not reactions to the world” (31) since we don’t just passively receive input and generate an automatic response but are “active constructors” of our emotional responses. She notes that her theory of constructed emotion is more “challenging” for people to understand and support, as it’s more complex than the classical view and doesn’t have such a “linear, cause and effect” (32) explanation for emotion. She credits other scientists with creating the school of thought called “constructionism,” which argues that your brain and body construct your actions and experiences in the moment. According to the constructionist view, aside from some basic neural wiring we share with all humans, our emotions aren’t automatic responses based on our genetics.

Two schools of thought inform the constructionist view: Social Construction, which emphasizes societal influences, and Psychological Construction, which argues that the brain shapes your experiences through the function of its “basic parts.” While these approaches don’t always agree, both posit that emotions are made rather than predictably prompted. Neuroconstruction also supports this view, because although we can identify neurons and synapses, they may be part of whole systems or smaller circuits—and operate differently depending on our genes and previous experiences. When we’re born, the “macrostructure” of our brains is guaranteed, but the “microwiring” can vary greatly among people. Barrett credits scientist William James with early discussions of the “variability of emotional life” (35) and reiterates that Charles Darwin also considered species as categories containing varied individuals.

Barrett compares people to cookies: While we can differ greatly in our looks and behavior, we’re still recognizable as a distinct category. She builds on this comparison to show how the brain’s core systems are like “ingredients” that each individual uses in different ways to produce varied results, a phenomenon Barrett calls “degeneracy.” The inherent connections among the brain’s components precludes neatly separating them from each other for analysis. Their interactions create “emergent properties” in the brain that behave differently than the core systems that made them. The author argues that, as such, the brain’s different “ingredients,” or core systems, must be studied within their contexts, an approach she calls “holism.” Barrett reiterates that emergent properties, degeneracy, and holism are all “the very antithesis of fingerprints” (38).

Barrett argues that neurons’ flexibility is an evolutionary asset that results in a more changeable and adaptable brain. While we don’t share universal inborn emotions, we do share the ability to form these concepts, which Barrett says are highly informed by our “social reality” of cultural influences. She then introduces new terms she uses in later chapters, such as “facial configuration” (as opposed to the more biased “facial expression”) and “instance of emotion” (in place of simply “emotion”). Barrett posits that speaking of emotions happening to you is inaccurate given that we generally participate unconsciously in their generation. The author acknowledges that much of her argument may seem “unintuitive” but notes that its lack of obviousness doesn’t mean it’s inaccurate. She compares the classical view of emotion to seeing the sun rise and set and assuming that it revolves around the earth