 There's nothing in Reddit's help section about the emotional distress caused by their awful redesign. I guess facts don't care about my feelings. Feelings are hard. Even if you're not in therapy trying to untangle a wild snarl of confusing or conflicting emotions about some traumatic event, you've probably experienced one of those days where every single person you interact with is clearly an idiot doing their absolute best to piss you off. And then you realize that you just forgot to drink your morning coffee. The traditional way we think about emotions casts them as fundamental building blocks of our mental landscape. Raw, unprocessed, automatic features of the human psyche. Tap someone's knee with a mallet and their leg jumps. Cut them off in traffic and they'll feel angry. Jump out of a bush at them with a scary mask on, they'll feel afraid. Fear, anger, happiness, lust, all of these emotions are automatic, pre-programmed responses to appropriate environmental stimuli. You can sometimes do funny things with those circuits, but when everything is working normally, stuff happens and sometimes that stuff trips some switch in your head that makes you feel feelings. Along with emotions, psychologists sometimes highlight a related phenomenon they call affect, which is a sort of general vibe on two axes, pleasant to unpleasant and excited to calm. Imagine being content, pleasant calm, versus ecstatic, pleasant excited, versus bored, unpleasant calm, versus murderously angry, unpleasant, excited. Many emotions are associated with certain physiological responses according to where they land in this schema, a display of affect caused by the triggered emotion. If you're feeling anxious or guilty, that causes you to sweat or knit your eyebrows together. If you're angry or jealous, you might narrow your eyes or grit your teeth. The emotions as discrete modules in the brain model is very useful. It readily describes the subjective experience of feelings as we feel them, but as with many models, there's some weirdness around the edges that maybe cause for some skepticism that that's really the beginning and end of the story. For example, there doesn't really seem to be anything like a rigorous objective way to determine someone's emotional state from the outside. It's no surprise that different people express the same emotions in different ways. Someone scared silly in a haunted house might whimper, scream bloody murder, go deathly quiet, or maybe even laugh hysterically in terror. Those differences are compounded when you compare emotional reactions across cultures. But even with very sensitive measurement tools like fMRI data, galvanic skin response, or electrodes on the face or scalp, there's a lot of evidence suggesting that no two people have enough brain architecture in common to verify anything but affect. Forget looking in a brain scan and saying, this person is anxious, that person is depressed. The best you can really do is say, these people are not super excited and not loving it. To make things even more complicated, the same person can experience the same emotion with an entirely different network of neurons firing. Being angry that you burnt dinner might look totally different than being angry that your laptop died. Even for gross evaluations of emotional states that we'd imagine ought to be dead simple to identify, humans are very easy to trick if given the wrong context clues for what that feeling might be. Take the same face with the same expression and change the background, maybe frame it with a colorful story or some moon music, and you can easily get someone to confidently identify the wrong emotion. That's kind of fun, right? But even more interestingly, it's not particularly difficult to get people to misidentify their own emotions. Forgetting your morning coffee and being pissed off at the world is an example of what's called a misattribution of arousal, confusing a couple of emotions that happen to share the same general affect. Maybe you felt like the world is ending when you're really just hungry, or if you drink too much coffee, mistaking your racing heartbeat for a feeling of imminent doom. It takes many years to teach children who are otherwise inclined to throw a tantrum that they're really just tired if the primitive emotional response model were true, if our brains contain hard-wired circuitry for feeling depressed or anxious or delighted, and our bodies responded to these emotions in reliable ways, we'd have to find some way to excuse or discount these discrepancies. Or, if you're up for it, we could try something a little more unconventional. Lisa Feldman Barrett's Theory of Constructed Emotion flips the script on the traditional interpretation of how emotions work. Despite how instantaneous and direct the things that we call anger and sadness and awe feel, according to the theory, they aren't really raw elements of our psychological landscape. They're interpretations, conceptual labels that we hang on certain combinations of affect in certain contexts. According to Barrett, when you feel something like terror, what you're actually feeling is a specific set of physiological sensations tied to a certain affect. Your muscles are tense, your body is quivering in a state of high arousal, maybe flooded with adrenaline, your heart is pounding. That affect might be due to your situation, you see a person with a knife and believe you're in danger, or maybe something else, like maybe you're breathing a higher than usual concentration of carbon dioxide. To make sense of those sensations, your brain is combining them in real time with context clues about what might be causing them, like features of the environment or memories of similar events. Feeling out what makes the most sense in that situation, then stamping the whole mess with some emotional label as a description or name. That doesn't imply that terror doesn't exist, but it does back us away a step from the traditional view that terror is a hardwired emotional circuit in the brain waiting to be activated by a dude with a knife. Fight or flight is certainly still happening, your body will get itself revved up to deal with a potential threat, but this emotion called terror is a conceptual framing of the situation, an interpretation, a useful way to think and talk about feeling revved up and bad in a specific context of mortal peril. In constructed emotion theory, that label is, at least in part, a cultural phenomenon, a feature of your language and thought, picking out that particular expression of affect in that particular situation. There's a large amount of overlap in these emotional labels between cultures, as we might expect, but there's also some differences. English doesn't have a word for a pleasant calm affect in the situation that something bad is happening to someone who deserves it, but Germans know when they're feeling Schadenfreude as a specific emotion, and once you incorporate it into your vocabulary and way of thinking, you might feel it too. We probably wouldn't go delving into neuroscience looking for a Schadenfreude circuit in a brain, but oh, they did do that. The predictive processing model of cognition also comes into play here. As we described in episode 130, there are good reasons to suspect that a large part of how our brains make sense of the world is through successive layers of prediction, with noisy sense data at lower levels and abstract concepts at higher ones, each level trying its damnedest to only communicate the most attention-worthy stuff up the chain, and directing lower levels in what to expect. If your visual cortex just expects the image it's getting from your eyes to blur when they snap from focusing on one thing to focusing on something else, you're not actually going to see that blurred image, because it's going to get processed out of the signal before you're aware of it. Barrett asserts that emotions are subject to the same prediction-noise-reduction sort of process, that the reason they feel so immediate and urgent is because your brain is scrambling to slap the most likely label on a wave of affect before it gets to your conscious mind, so it's not unexpected or weird when it gets there. Oh man, there's a puppy. We always get an excited positive affect with puppies. The emotional state we associate with that affect is adoration. Okay, cue it up. Three...aww. Interestingly, this framework implies something that has been part of clinical psychology for a long time, the power of reframing. When someone gets into a rut of negative emotions like sadness, anxiety, stuff like that, their brain is doing its best to anticipate and categorize the bodily sensations that accompany, or in some sense, cause those emotions. By paying conscious attention to those sensations, it might be possible to cast them in a slightly different light and adjust that interpretation to something more useful. Are you nervous about an upcoming test? Well, you're bouncing your leg a bunch. You're certainly full of energy. Rather than nervous, could it be that you're excited? Eager? You feel miserable. This day just couldn't get any worse. You're moving slow, not a lot of energy. Maybe you're really just tired. Of course, clinical depression and anxiety are a different can of worms, but if Barrett's theory is right, the only actual difference between these states is which predictive model your brain is using to explain your affect. And if you can encourage it to use a different model, you're literally changing what emotion you're feeling. You're not at the mercy of some hard-coded anger response or shame circuitry in your brain, the right interpretation of the situation and your body's response to it can shunt you into a totally different emotional state. Barrett's model also has implications for how we conduct research about emotions. There are mountains of data that have been gathered in search of some objective neurological basis for emotional circuits, the seemingly elusive signatures in the brain of happiness, fear, and so on. There's also a great deal of work built on models of the emotional experiences of animals, attempts to develop useful insights and interventions for humans by observing animals and characterizing their behavior as signaling something analogous to human emotions. If emotions are constructed, as Barrett suggests, this kind of research is kind of doomed to failure from the outset. No matter how hard you search, you won't be able to find a configuration of neurons or a mouse model that will give you insight into any individual human's anger or anxiety, because their emotions are unique to them. Of course, as with many high-level psychological models, the theory of constructed emotion is fascinating, potentially very useful in many situations, and exceptionally difficult to either verify or falsify. The evidence gathered so far around the idea doesn't disprove it. Some studies support his claims, some undermine them. Also, as we discussed in episode 190, link in the description, a theory developed to address the deficiencies in some standard theory is always going to start off with an advantage in explaining those holes it's supposedly patching, provided it's not transparently stupid. It can take a while to analyze a new idea to figure out if it fully explains all the evidence better than what we already have. But it's certainly something I'm going to think about the next time I forget my morning coffee. What about you? Does Barrett's theory about how feelings are built out of affect and predictive interpretation ring true to you? Where can you see a model like this being useful for working with emotions? Please leave a comment below and let me know what you think. Thank you very much for watching. Don't forget to blah blah subscribe, blah share, and don't stop thunking.