Skip to content

How Emotions are Made

My notes for the book How Emotions Are Made: The Secret Life of the Brain, as recommended by Binaural Histolog. I wanted something a little less challenging and I feel like this will give me 80% of the benefit of a new perspective without delving directly in to academic level reading. With that being said, take my digest of a digest with a grain of salt and as inspiration.

A few things about how I’m looking at the book:

  • I already didn’t subscribe to the “classical view of emotion” (see below)
  • They present solid evidence against the classical view (good)
  • Then, they present their own theory as a replacement (there’s got to be more than just the theory of constructed emotion as a counterargument)
  • If I’m incredibly skeptical, or just doing my due diligence, because I’m not familiar with research on affect, they could be making a straw man argument against the classical view of emotion. (Even if I do ironically intuitively agree with the theory of constructed emotion.)
  • This reminds me of state vs literally-everyone-else arguments in hypnosis - where useful and interesting findings are thrown out from the state theory side just because of the connotation.
  • This may come up later in the book - but I’m curious how they’d look for explicit evidence of construction. A new theory is going to appear stronger next to something you disprove.
  • It seems fine. Preferentially, I’d like something presented a bit more neutrally, but this is a pop-psy book, not academic research. (I’m also reminding myself of this so I don’t dig in to every damn rabbit hole that comes up.)

Introduction: The Two-Thousand-Year-Old Assumption

Section titled “Introduction: The Two-Thousand-Year-Old Assumption”

Our intuition suggests emotions are hardwired and universal. They present the classical view of emotion as:

  • We all have emotions from birth
  • “They are distinct, recognizable phenomena”
  • They happen quickly and automatically
  • Our voices reveal our emotions
  • Emotions are accurately represented by our body language
  • More intuitively, emotions are seen as a reflex, outside our cognition, attributed to a primitive part of our brain

I assume through foreshadowing, these are assumptions Lisa Barret, the author, intends to challenge and provide counterevidence to. They present this classical view as a culturally entrenched story about emotions, which continues to be pervasive even in much psychological literature.

As an appetizer - they provide:

  • Studies are unable to find physiological “fingerprints” of common emotions, citing replication issues as well as more nuanced studies finding these patterns are culturally influenced.

They explicitly state:

  • Emotions are componential in nature
  • Emotions vary from culture to culture
  • Emotions are not triggered, you create them
  • Emotions are real, but “a product of human agreement”

The author, for example’s sake, suggested they felt sadness in an instance because:

  • They had been raised in a certain culture.
  • Some bodily feelings go along with feelings of loss.
  • Her physiological responses (flushed face, knots in her stomach) were due to memories of a past sad event (in this case, loss of life)
  • Crying was a response to calm her nervous system
  • Meaning was made as “an instance of sadness”
  • It was constructed from context. Those same sensations could be attributed to joy (EG - when people cry at weddings.)

In spite of their scientific explanation, knowledge, and counterevidence, emotions are still subjectively experienced in line with the classical view of emotion.

1 - The Search for Emotion’s “Fingeriprints”

Section titled “1 - The Search for Emotion’s “Fingeriprints””

The author, starting her journey, experienced 3 years of replication issues examining the differences in feeling anxious and depressed. The intuitive emotional attribution did not match up with the participant’s interoception or physiological responses.

From her initial studies, she began to in to “emotional granularity,” so see if she could help individuals more precisely assess their internal states. She began to study alleged physiological responses to emotion (facial movements, body movements, interoceptive sensations, heart rate, etc) and found that the assumed ‘fingerprints’ of an emotion did not replicate.

Some notes:

  • Given a story, emotions are easier to attribute when observing a facial expression. However, when removing the story, accuracy dropped dramatically, mostly reducing down to positive or negative valence.
  • Using technology instead of people to measure facial movements did not provide a reliable ‘fingerprint’ to emotional expression.
  • When individuals were asked to watch angry or scared babies, they were able to identify the emotion from body language, even when the faces were blanked out. However, newborns don’t show unique facial expressions similar to adults, suggesting it’s culturally influenced.
  • TLDR - removing the story behind the emotion of the facial expression severely deteriorated the ability for adults to granularly recognize emotions. The story changes our attribution.
  • In a study where subjects were asked to hold a facial expression for an emotion, they did observe repeatable physiological responses. However, this study was inconclusive - they could not prove that these physiological responses were either unique to an emotion or could not be attributed to anything else. (Do you hold your breath a bit when scowling or grimacing?) Further counterevidence was found when people removed from western culture were asked to hold similar poses, but did not have the same physiological responses.
  • Different studies found different patterns when presenting emotionally evocative material. They did not replicate.
  • Changing our body’s position (leaning forward, back, etc) can alter our physiological response. However, we cannot generalize emotions to certain physiological responses.
  • Although the amygdala is important to fear responses, it does not dictate the experience of fear. (A woman with a damaged amygdala was still able to identify fearful expression, they could not be conditioned to experience fear through classical conditioning, but they did still panic in a life threatening situation.) The amygdala is also involved in non-emotional events.

The author presents a single ambiguous image, then explains what the image is. Check https://www.behaviouralbydesign.com/post/neuroscience-of-strange-and-beautiful-experiences for an example. They call your first state experiential blindness. They highlight (and intend to generalize) that:

  • The experience of [thing] being present in the image is constructed.
  • Past experiences and knowledge influence your observations.
  • “The process of construction is invisible to you.”
  • Un-seeing this construction is either difficult or impossible.

They call the process of hallucinating these experiences simulation.

Moreover, when you imagine consuming an apple, or read the word apple - some physiological responses happen automatically (salivation, preparing for digestion) as well as some experiential memories (taste, smell, feel.) They categorize all this as simulation.

Context influences experience. At a party for their kid, they smeared some baby food on a diaper. Even though the kids knew it was essentially apple sauce, some kids still wretched in response.

Without concepts, you are experientially blind, as you were with the blobby bee.

Page 5 of Ch2 of How Emotions are Made

They suggest your mind makes emotional sense of your interoceptive awareness in the same way. (If your stomach is tied up in knots, you could be flirting in love, or about to have an uncomfortable conversation… or even just have the flu.)

They’ll use neuroscience to back the construction theory up later. This is only for an intuitive understanding.

The theory of constructed emotion could be coming from multiple layers:

  • Social circumstances and cultural understanding

  • Psychological. (EG - someone could experience either anger or euphoria, depending on the context after being unknowingly injected with adrenalin.)

  • For example, construction of visual experience with honey bee, experiential and blindness

    • past experiences
    • construction is “invisible” to you - you can’t observe yourself constructing this image

(They highlight core systems in the brain often have similar functions between people, brain regions and networks often do not serve only one function. This is evidence that it’s unlikely there is a core region of the brain solely responsible for emotion.)

Analogizing, muffins and cupcakes are made with the same ingredients. You can’t analyze the chemistry of these pastries and figure out which one is which. The difference is a cultural understanding. We have the same difficulties analyzing the components of our neural and interoceptive responses when trying to attribute them to an emotion.

They highlight, when adopting the constructivist view:

  • Facial expression is a misnomer, we should use facial construction.
  • We don’t detect emotion in other people through expression alone - this is through context. Perception is often an internal (and often opaque) experience.
  • They’d prefer to avoid language of emotions “happening to you.” Avoid language similar to something trigger an emotion, or having an emotional reaction.
  • Accuracy in perception is subjective.

Here’s what they want to get across:

No scientific innovation will miraculously reveal a biological fingerprint of any emotion. That’s because our emotions aren’t built-in, waiting to be revealed. They are made. By us. We don’t recognize emotions or identify emotions: we construct our own emotional experiences, and our perceptions of others’ emotions, on the spot, as needed, through a complex interplay of systems.

The book provides an example where you can experience a constructive shift in your interpretation of someone’s facial expression. (EG - captioning a cropped image of someone as screaming in terror, aaaaand then revealing they just won a tennis match. ) This highlights the cultural and contextual understanding of facial configurations.

Most of this chapter is about how the classical understanding of emotional expression subtly shifted how research was done, creating poor evidence in support of expression and experience being universal.

  • Interoception is one of the core ‘ingredients’ to a constructed emotional experience.

(I’m going to start summarizing this in my own way. I don’t really want to CliffsNotesify the book - I want this for my own reference and don’t mind sharing.)

As an aside - while Skinnerian operant and classical conditioning is a real thing, this does not hold at a neurological level. The brain has constant, intrinsic activity, as well as constantly running intrinsic networks. Intrinsic brain activity is not only responsible for keeping your basic bodily functions running, but also construction as well. (I’m curious to see evidence of this later.) It makes predictions from seeing the bee in the earlier image, all the way to figuring out what’s subjectively going on.

Prediction is such a fundamental activity of the human brain that some scientists consider it the brain’s primary mode of operation.

They also note that the illusion of free will is about the subjective experience of free will, but you do still have control, since you are your brain. They suggest a brain that was purely reactive would be too physiologically expensive to run all the time, so the hack is to run predictive processing to be more efficient. Our predictions are ‘tuned’ by our sensory inputs, not completely resynthesized moment to moment.

There’s some stuff here on how visual information is mostly predicted. For more on predictive processing, check out Nadine Dijkstra on youtube.

In predictive processing, it’s theorized that we respond to prediction errors, and try to process them. In this theory, the following happens in loops at the neurological (not just subjective) level.

  1. Predict
  2. Simulate (construction, if I’ve got this right)
  3. Compare
  4. Resolve errors
  5. GOTO 1

This sort of prediction happens out of awareness - for example when playing catch with someone and the motor responses required for catching the ball.

Two things can happen when a prediction error occurs:

  • Change the prediction. (Try to improve at catching the ball.)
  • Screw it, just keep the old prediction. (That’s not baby-food in the diapers, that’s filth.)

Usually, you experience interoception only in general terms: those simple feelings of pleasure, displeasure, arousal, or calmness that I mentioned earlier. Sometimes, however, you experience moments of intense interoceptive sensations as emotions.

![[Pasted image 20250520181056.png]] Our experiences shift from predictive processing to sensory input, as well as from less learning to more learning, depending on what we’re doing.

They say we can predict the “future” at will through simulation. (I’m curious to see the evidence suggesting this only happens in humans.)

The interoceptive explanation we experience is simplified. (We don’t think, “wow dang, my liver’s producing a lot of bile today,” we might just experience an upset stomach.)

A friend provided a handy cheat sheet for some terminology she simplified:

  • Body Budget: Allostasis
  • Costs: Allostatic Load

Continuing the simplification - the network for interoception includes two major sets of brain regions:

  • Sending predictions to control the “internal environment”: (limbic and visceromotor regions), (heart rate, breathing, cortisol, metabolism)
  • Internal sensations: primary interoceptive cortex

These two sets of regions work together in a predictive processing loop, creating interoceptive sensations.

They argue that these regions are claimed to also be responsible for emotion. Continuing that, they highlight that they are not reactive, but predictive. Therefore, emotions attributed to these regions are predictions.

These changes to allostasis happen without conscious awareness - EG - when approached by your boss at work. (Cortisol is released to provide extra energy.) These responses happen whether or not they are real or imagined, the simulation driving the response.

When you interact with your friends, parents, children, lovers, teammates, therapist, or other close companions, you and they synchronize breathing, heart beats, and other physical signals, leading to tangible benefits. Holding hands with loved ones, or even keeping their photo on your desk at work, reduces activation in your body-budgeting regions and makes you less bothered by pain… In contrast, when you lose a close, loving relationship and feel physically ill about it, part of the reason is that your loved one is no longer helping to regulate your budget.

It’s generally scientifically accepted that affect is present from birth. Affect (a general sense of feeling, not emotion) has two components:

  • Valence: How pleasant/unpleasant you feel
  • Arousal: How calm/agitated you feel

These are both predictions/representations based on interoception. (EG, the same way you optically perceive light and dark, you interoceptively perceive valence and arousal. )


This is only touched on in the book - but they go into James Russel’s circumplex model of valence and arousal, creating affect. You can check more out on it here: https://psu.pb.unizin.org/psych425/chapter/circumplex-models/ On this tangent - Russel’s model differentiates between core affect and PEE. 😅

  • Core affect is the most basic feeling at any point in time.
  • PEE is a prototypical emotion episode. They are created by a specific eliciting event, an emotion. It includes affect, behavior changes, “cognitive appraisal, subjective feelings,” physiological changes/responses, and “a clear, short time period during which PEE occurs”

(I just think PEE is funny.)


When people make decisions based on their gut feelings, they call it affective realism. So… bribe your DM with snacks before you play a game. They also suggest our affective responses are related to how it’d affect our body budget. (Photos of kittens are calming, corpses are less so for our body’s safety.)

It’s unlikely that affect depends on “predictive brain circuitry.”

While predictive brain circuitry is important for affect, it likely is not necessary. Consider the case of Roger, a fifty-six-year-old patient whose relevant circuitry was destroyed by a rare illness… Nevertheless, Roger experiences affect. Most likely, his affect is driven by actual sensory inputs from his body; other brain regions could be supplying the predictions, an example of degeneracy (different sets of neurons producing the same outcome).

It seems the point they want to get across, despite the tangents they’re going on, is…

Interoception… is more influential to perception… than the outside world is.


Wow, also seems like they have it out for the “triune brain.” Ironic, since The Triune Brain, Hypnosis and the Evolution of Consciousness is on my reading list. Shots fired I guess!

(Edit - I’m back. I glossed through the intro of the aforementioned book and it seems a bit out there bonkers! Maybe less something for edification and more for entertainment. Also, demoted off my reading list.)

The model begins with ancient subcortical circuits for basic survival, which we allegedly inherited from reptiles. Sitting atop those circuits is an alleged emotion system, known as the “limbic system,” that we supposedly inherited from early mammals. And wrapped around the so-called limbic system, like icing on an already-baked cake, is our allegedly rational and uniquely human cortex. This illusory arrangement of layers, which is sometimes called the “triune brain,” remains one of the most successful misconceptions in human biology.

  • Stimulus perception is consolidated into ‘categories,’ and is largely done outside of your awareness. (EG, these aren’t just pixels on a screen, words require context to be understood, the perception of these things automatically constructed.)
  • Common meanings of a word for an emotion are likely culturally constructed.
  • They suggest there are goal based concepts. EG: “Things that will protect you from this hailstorm.” Therefore, our experience is not just VAKog, but connotates goal and context.

Emotion concepts are goal-based concepts. Instances of happiness… are highly variable… Perhaps your goal is to feel accepted, to feel pleasure, to achieve an ambition, or to find meaning in life. Your concept of “Happiness” in the moment is centered on such a goal, binding together the diverse instances from your past.


  • Some of our earliest learning is done with statistical inference.
    • We also learn at an early age that others may have their own preferences.
  • Reportedly, words are required for concept learning.

A word tells the infant, “Do you see all these objects that look different physically? They have an equivalence that is mental.” That equivalence is the basis for a goal-based concept.

(The author talks about her child’s development and provides anecdotal evidence for the cultural definition of emotion. Author - if you’re reading: 1. I’m sorry for being so harsh. 2. I’m on board, just not engaged! )


This chapter contradicts itself here, IMO.

Most of us who speak English were able to enjoy someone else’s misfortune long before the word schadenfreude entered our language. All you need is a concept. (The author talks about how concepts can be combined and modified, creating feelings like schadenfreude.)

And then…

(Immediately after talking about having a strong or weak vocabulary for emotion.) When a mind has an impoverished conceptual system for emotion, can it perceive emotion? From scientific experiments in our own lab, we know that the answer is generally no.

They do have some interesting and digestible information on alexithymia, though.

The way around this conundrum is to study people who have a naturally impoverished conceptual system for emotion, a condition called alexithymia… Its sufferers do have difficulty experiencing emotion, as the theory of constructed emotion would predict. In a situation where a person with a working conceptual system might experience anger, people with alexithymia are more likely to experience a stomachache… If a person with a working conceptual system saw two men shouting at each other, she might make a mental inference and perceive anger, whereas a person with alexithymia would report perceiving only shouting…

I might be losing a bit of detail here, but developmentally predictive processing works to simplify interoceptive and sensory input. Leading with that - predictive processing is also responsible for labeling and categorizing emotion.

Oh. Yup. This is just all predictive processing, interoception and sensory input effects predictive processing, aaaand I’m guessing this is heading in the direction of emotion has a story component (as an impurity to the simplicity of affect/activation/valence.)

If you’re not me - just go here and read https://binauralhistolog.com/newbie/theory/components-of-hypnosis#predictive-processing. It’ll save you a lot of time.

Interestingly, within the predictive processing model, they suggest that more granularity in our emotional understanding is more efficient.

When your brain constructs multiple instances of “Happiness” at seeing Uncle Kevin, it must sort out which one best resembles your current sensory input, to become the winning instance. This is a big job for your brain with some metabolic cost. But imagine if the English language had a more specific word than “happiness” for feeling attachment to a close friend, such as the Korean word jeong (정)… On the other hand, if you were constructing with the very broad concept “Pleasant Feeling” rather than “Happiness,” your brain’s job would be harder. Preciseness leads to efficiency; this is a biological payoff of higher emotional granularity.

(They also denote that with ‘population thinking,’ your brain generates many instances of happiness, and then picks the more correct ones, culling out the options that are less so.)

They also suggest your control network controls context, affecting predictive processing. (They contend that there is more nuance to the process, though.) Clearing the terminology up, the control network is more of an optimizer. In layman’s terms, it’s what’s responsible for your “attentional spotlight.” They also say it’s the “emotional regulation” (self control) network.

They highlight that perception is separate from reality, including colors and sound. How we categorize color is a cultural phenomena.

They suggest two things allow emotions to be real in a social (folk) reality:

  • Collective intentionality - a shared understanding for an intention. “Categorization as a cooperative act.”
  • Language. We can quickly talk about emotion and the stories around it.

Emotions serve the following functions:

  • They make meaning (a story)
  • They prescribe action
  • They provide predictions for allostasis

Socially - you can:

  • Communicate.
  • Influence other people’s allostasis. (Time to haul ass in a group?)

As part of the book’s theory - they assert that without a concept of an emotion, you cannot experience that emotion. (Which… I’m not sure I’m sold on personally. It presents the chicken and the egg issue. Continuing that, nonverbal individuals and animals certainly are not emotionless. I could absolutely see something like - “if an emotion is a culturally generated concept, you need to understand it before experiencing it.” I guess I’ll check the chapter about whether or not a dog can be angry later. This assertion also suggests someone that grew up isolated would be emotionless.)

Aaaaand the rest of this chapter talks a lot about how we disagree about emotion from culture to culture.

The author would like us to:

  • Shift away from the classical view of emotion. (Emotional reaction, facial expression, emotional circuits of the brain)
  • And move towards their understanding (interoception, prediction, body budget, social reality)

I’m… not particularly interested by this chapter.

TLDR:

  • Eat Healthy
  • Exercise
  • Sleep Well

In addition:

  • Massage
  • Yoga
  • Houseplants (will allegedly improve your allostasis)
  • Friends
  • Pets

There’s more here, but I mostly skimmed it. It’s frankly discouraging for someone covering up that they’re having a rough time.

Not terribly interested in this chapter either.

Not at all interested in this. Go read Hypnosis and Forensic Psychology as a treat instead.

The theory of constructed emotion requires:

  1. Interoception
  2. Emotional concepts
  3. A social reality

Cool - we kick it off by agreeing that most animals experience interoception, and therefore affect.

Animals reportedly have a smaller affective niche. Grabbing that from the glossary…

Affective niche: Everything that has any relevance to your body budget in the present moment.

https://how-emotions-are-made.com/notes/Glossary

And with that…

Humans construct goal-based concepts, and a macaque brain simply lacks the necessary wiring to do so. It’s the same lack of wiring that accounts for their smaller affective niche.

Their take, more directly is…

Any concept can be goal-based—recall that “Fish” can be a pet or a dinner—but emotion concepts are only goal-based, so it seems very likely that chimps cannot learn emotion concepts like “Happiness” and “Anger.”

Because of this likelihood, and the lack of symbolic communication, they argue that they cannot create an emotional social reality.

Because of this, they suggest it’s unlikely they categorize emotion, therefore do not experience emotion. (But they do still experience affect.) They offer explanations and counterevidence for things that look like emotion in dogs, but are not conclusive.

On the flip side…

Dogs may well have some emotion-like concepts. For example, a number of scientists now suspect that very social animals, such as dogs and elephants, have some concept of death and can experience some kind of grief. This grief need not have exactly the same features as human grief, but both could be rooted in something similar: the neurochemical basis of attachment, body budgeting, and affect. In humans, the loss of a parent, lover, or close friend can wreak havoc with your budget… When one creature loses another who helped to keep its body budget on track, the first creature will feel miserable from the budget imbalance.

Another TLDR I’ll provide - behavioral response is not equivalent to emotional expression in their model.

Skipping this chapter. It’s rhetoric and a recap.