Language and Emotion

I am going to start with a dedication to Paul Ekman. Paul is the person who got me interested in the study of emotion, got me involved in the International Society for Research on Emotion, where I’ve been listening in quietly and learning a lot for many years. Paul’s contributions are reflected everywhere in the study of emotions, whether you agree with them or not, and I particularly want to honor not only his scholarship and his contributions but his generosity.

I want to talk about why a linguist is studying emotion.

I started to study the relationship between language and meaning in 1963. What I was trying to do then was to unify generative linguistics with formal logic. At that time, logic was assumed to be the study of meaning and thought, and I had found lots of ways in which meaningful concepts and inference based on them enter into grammar.

Both generative linguistics and formal logic are disembodied theories that use inherently meaningless symbols, yet they were attempting to characterize meaningful thought and language. I accepted the usual assumptions of the day and went along with them as long as I could.

By 1975, I had heard four lectures by Berkeley colleagues that showed that what I was doing was impossible. Eleanor Rosch (e.g., Rosch & Mervis, 1975) has done her great research on basic level concepts which are defined by motor programs, mental imagery, and visual perception, all of which are embodied. That meant our conceptual concepts are embodied.

Paul Kay and Chad McDaniel (1978) grounded their work on basic color terms on the neurophysiology of color vision. As you probably all know, if you’ve studied any neuroscience, there is no color in the world—no red in blood, no green in grass, no blue in the sky. It’s not out there. Color makes use of the wavelengths of light reflected by objects, but color is not wavelengths. Color categories are whole ranges of colors with best examples and with lesser examples fading off to the edges. They are created by color cones and neural circuitry. Color is embodied. Color language is embodied. And colors are somewhat different for men and women. So, if you’ve ever had an argument about whether something is blue or green or brown or orange with someone of the opposite gender, you are both right. You have different color cones, and 2 to 3 percent of women have an extra set of color cones, which allows them to see many more variations than men. My wife is an artist and seems to be one of them.

The third lecture was by Leonard Talmy (2000), pointing out that spatial relations terms are radically different across languages. That is, the particular words and morphemes are different. But however different the words and morphemes are, they use the same primitive concepts in different combinations. Primitives such as Above–Below, In Front–Behind, Near–Far, and so on. These primitives are not out there in the universe independent of our bodies. They only make sense given our bodies. In physics, there’s no above and below, no in front and behind. You have to have a body to understand these concepts.

The fourth lecture was by Chuck Fillmore (1977). He introduced frame semantics, which is based on fundamental human experience, both physical and social. The whole study of linguistic frames is based on embodied experience. Fillmore then argued that every word in every language is defined relative to such conceptual frames. Think of a restaurant frame, where you have waiters, and eaters, and chefs, and so on, who play roles. They are in the restaurant frame, but there’s no herd of elephants in the restaurant frame.

In short, in all sorts of ways, thought and language are embodied.

By 1978, Mike Reddy (1979) and I had independently shown that metaphor is not just a matter of language. Metaphor is conceptual. We think metaphorically. Conceptual metaphors, we discovered, are frame-to-frame mappings. Then in 1979, Mark Johnson and I showed that many conceptual metaphors are embodied for emotions; examples like Happy Is Up, Sad Is Down, as in I’m feeling up today, My spirits rose, I’m feeling depressed, I’m down in the dumps, and so on. Or take Affection Is Warmth. Someone can be a warm person, a cold person and so on. We realized that these cases have to do with correspondences in your embodied experiences. When you’re held affectionately by your parents, you feel their body warmth. When you’re happy, you smile, and your facial muscles go up. Metaphors show linguistic correlates of embodiment.

In 1983, Zoltán Kövecses arrived in Berkeley. He was doing an idiom dictionary of English, and when he got to A, he found over 400 idioms for anger. He showed up in my office with examples like boiling mad, blow your top, incensed, burn up, flip your lid, consumed with anger, burst a blood vessel, foaming at the mouth, wrestling with your anger, on a short fuse, letting off steam, fly off the handle, appease your anger, doing a slow burn, simmer down, hopping mad aligned with rage, and so on.

During 9 months of work we figured out the system. The clue came from Paul Ekman. Paul pointed that the physiology of anger includes a rise in skin temperature, a rise in blood pressure and heart beat rate, interference with active reception, and interference to motor control.

Kövecses and I discovered that the conceptual metaphors for anger were based on the physiology of anger. There are conceptual metaphors like Anger Is Heat, Anger Is Pressure, and Anger Is a Wild Animal. It’s something that’s inside of you that you can’t control. Anger Is a Forceful Opponent, something that interferes with how you can move your body. Anger is madness, which interferes with accurate perception.

A special case of Anger Is Heat is Anger Is Fire. Expressions of the conceptual metaphor include, inflammatory remarks, all burned up, consumed with anger, add fuel to the fire. Another special case of Anger is Heat is Anger is a Hot Fluid in a Container. Examples include, boiling mad, all steamed up, being a hothead, seeing red, fuming, letting off steam, he exploded, he blew up, there are outbursts, she erupted, exploded, hit the ceiling, flip your lid, go through the roof. With Anger Is a Wild Animal, you have examples like a ferocious temper and unleash your anger, getting a grip on yourself, insatiable anger, reawaken his anger. When you conceptualize Anger as an Opponent, you can be seized by anger, you can wrestle with it. You can fight back your anger, hold back your anger, surrender to your anger, overcome your anger and it can be appeased. When you think with that conceptual the Anger is Madness, you can throw a fit, start foaming at the mouth, go crazy, go bananas, have an insane rage—that drives me nuts—and, of course, there is the word mad. The anger metaphors are clearly based on the physiology of anger.

We also found that there is a prototypical anger scenario in which you start off with an offending offense: you get angry, your attempt to control it, you lose your control. And then there is some active retribution that you perform, and the anger goes away. What Ray Gibbs at UC Santa Cruz showed, in experiment, is that the linguistic metaphors fit the various phases of the anger. In short, the phase analysis and the metaphor analysis fit together (Gibbs & Matlock, 1999).

Kövecses and I also found idioms that don’t fit the prototypical scenario. Expressions like wrath or rage or slow burn, cool anger, cold anger, redirected anger, etcetera. These work by the general account of variations discussed earlier in Women, Fire and Dangerous Things (Lakoff, 1987).

The neural theory of how embodied metaphor works dates back to 1997.

Srini Narayanan, in his 1997 thesis, pointed out that the circuitry for moving your body—moving circuitry—also characterizes the structure of events and actions in the world. In linguistics, this is called Aspect, and the same structure appears in every language in the world.

Think about it! You understand the world in terms of how you can move your body. That includes abstract events and causal events of all kinds, including emotion. From the perspective of the brain, the circuitry is the same. It’s the same, if not necessarily the same circuitry, it’s the same type of circuitry. It’s the same structure that appears over and over in motor control, but also in thought and language.

This gets us into the study of what’s called primary metaphor, which Johnson and I noticed (1999).

Consider the conceptual metaphor Love Is a Journey. It is a general way of thinking metaphorically about intended long-term love relationships and what difficulties arise in those relationships. This way of conceptualizing and reasoning about love in terms of journeys and what goes wrong with journeys leads to linguistic metaphorical expressions like, the marriage is on the rocks, we’re going in different directions, we’re spinning our wheels, etcetera.

In the Love Is a Journey metaphor, Lovers Are Travelers, Relationships Are Vehicles, Common Life Goals are Common Destinations, and Difficulties in the Relationship are Problems in Getting There.

That general metaphor is decomposable into more basic embodied metaphors. Why are love relations vehicles? Because A Vehicle Is a Container: in a relationship you enter a relationship, you can leave a relationship, you can be deep in it, on the edge of it, etcetera. Why is a Relationship a Container? Because as children, we grow up in the same home (contained space) as our relatives—our “relations.” Intimacy is seen as Closeness: We’re very close, we’re drifting apart, and so on. Because Achieving a Purpose is Reaching a Destination, and because we are expected to have life goals, a Relationship is seen as a “vehicle” for achieving our life goals in this culture. A vehicle is a container in which people are “close” to one another. In an expected long-term love relationship, you are expected to have common life goals. That is often difficult to achieve, and this metaphor is about the problems encountered along the way.

Love Is a Journey is a complex conceptual metaphor composed of other more basic metaphors, called “primary” metaphors, like Intimacy is Closeness and Anger Is Heat. These primary metaphors arise from embodiment. The same ones occur in many unrelated languages around the world. They have to do with commonplace experiences that regularly occur together, like being angry and having a rise in skin temperature, or feeling your parents’ affection when held and simultaneously feeling their bodily warmth.

This gave rise to a neural theory of metaphor again in 1997 by Srini Narayanan, who constructed the neural theory of how primary metaphors work. The basic idea is this: Suppose you have two basic human experiences that regularly occur together and that activate different brain regions. For example, when you are held affectionately, you feel the bodily of warmth of your parents. Different brain regions are active, the circuitry in both is activated, the more the circuits are activated the stronger their synapses get. The activation from each spreads along existing pathways by Hebbian learning. As the synapses get stronger, the activation spreads from both neural centers, until the shortest pathway is found, and a single circuit is formed. At this point, First-Spike Time Dependent Plasticity takes over at the neural level: the neurons that regularly spike first get strengthened in their direction and those aimed in the opposite direction get weakened. The result is an asymmetric circuit that constitutes the metaphor. The metaphor is physically there in your brain, linking two distinct brain regions.

This allows us to predict the directionality of the metaphors, and therefore, which domain is source and which is target. So, for example, why is it that affection is warmth and not warmth is affection? You don’t say the temperature became more affectionate, meaning that it gets hotter. The reason is that the brain is always computing temperature not always computing affection, and so you’re going to get more first spikes from temperature to affection than in the opposite direction and that will result in the asymmetry of the metaphor. This works for all of the hundreds of cases we’ve looked at so far.

I believe that we can explain why we have embodied metaphors like: Anger Is Heat and Happy Is Up and Affection Is Warmth. We know why they occur.

Moreover, the brain has circuitry that is embodied, that is, directly linked bidirectionally to the body. Neural circuitry in the body and embodied circuitry in the brain form a single integrated system. The brain’s associative circuitry links the various regions of embodied circuitry, and it is in that associative linking circuitry that the metaphorical circuitry exists.

So far I haven’t said much about grammar. I’ll turn to that now. There is a set of related grammatical constructions governed by predicates—adjectives and verbs—that linguists call emotive factives. Examples are amuse, amaze, bewilder, surprise, disgust, please, and more than 100 more. Examples of the four construction types are:

Each has (a) an emotion, (b) an experiencer of the emotion, (c) a stimulus resulting in the experience of the emotion, (d) the stimulus is presupposed and the emotional experience is asserted. Here are some more: baffled, confused, amazed, thrilled, delighted/delightful, enheartened, interested, bored, disappointed, sickened, disturbed, angered, upset, shocked, embarrassed, embittered, aggravated, depressed, exasperated, irritated, frustrated, overwhelmed, pained, discouraged, humiliated, scared.

I bring this up because there are constant discussions of how to “define” emotions. In the case of this set of constructions, it is the grammar of English that does the defining. They form a natural category right there in the grammar of the language every English speaker uses every day.

Now I want to turn to the concept of being overwhelmed by emotion. When can you be overwhelmed? All the states referred to by emotion factors can overwhelm you emotionally. You can be overwhelmed by anger, joy, sadness, disgust, confusion, bother. Are you bored right now? Shame, horror, embarrassment, surprise, and amusement for all. Any of those can overwhelm you. What does that mean? Each of these has a physical quarrel in the face of body or both. Each is understood metaphorically as a force that stops you in your tracks. There is apparently an empirical universal metaphor that achieving a goal is reaching your destination. And in trying to “reach” your goal, an emotion is interfering, stopping you metaphorically in your tracks. Overwhelming you.

In short, extreme emotions have a force. What this suggests is a disruptive theory of emotions: Which emotions disrupt you and how? Have the members of this society studied this and figured out how it works? And if not, why not?

Finally I want to talk about affect and what is called the brain’s “reward system.”

There is brain circuitry that releases hormones in certain brain regions that we experience as “affect,” either positive or negative. This neural system has to do with achieving goals and maintaining normal homeostasis. It determines when you feel good or bad—when you have a sense of well-being or ill-being.

Morality is about well-being—our own or that of others. As discussed in detail in Chapter 14 of Philosophy in the Flesh (Lakoff & Johnson, 1999) and Chapter 4 of The Political Mind, (Lakoff, 2008) our concepts of morality arise via conceptual metaphor—in particular, primary metaphor. The metaphorical concepts that define the various forms of morality arise from recurrent correlations between well- and ill-being on the one hand, and everyday other experiences on the other. For example, you eat pure food as a child; you feel good. You eat rotten food, you feel bad. The metaphor: morality is purity; immorality is rottenness. That was a rotten thing to do. Something is rotten in the state of Denmark.

Walking upright versus crawling. Any 1 year old would tell you that you’re better off if you can walk upright than if you can crawl. Therefore, Morality Is Uprightness. Being immoral is being a low-down snake, being underhanded and so on. Then you have Morality as Light; Immorality is Darkness. The Prince of Darkness. The white hats and the black hats.

Since you’re better off if you have the things you need than if you don’t, Well-being is Wealth and Morality is Accounting. So if I do you a favor, it’s metaphorically like I gave you money: I owe you one, I’m in your debt, How can I repay you, etcetera.

If I harm you, you can balance the books by harming me back (that’s “retribution”) or by taking something good away from me (that’s revenge). Karma is moral accounting with the universe.

There are also family-based metaphorical concepts of morality. So, (a) If you’re a young child, you’re better off in general if you listen to your parents than if you don’t. (b) You’re better off in general if your parents nurture you, than if they don’t. So, morality is obedience to legitimate authority (a), and morality is the nurturance (b).

And then there are about two dozen others. I won’t go through all of them here. They are spelled out in the books.

The point is that the conceptions of morality widespread around the world arise via primary metaphor!

This brings me to the appraisal theory of emotions. Well-being is what appraisal is about.

Appraisals centrally involve scaling. They’re linear scales. And linear scales make use of primary metaphors. The basic metaphors are these:

Examples: Henry is far smarter than Bill. Bill’s intelligence goes way beyond Henry’s.

Since Achieving purposes are normally seen as good things that give you a sense of well-being, there is a derived metaphor:

These metaphors are neurally bound together, and the concept of appraisal is defined relative to that complex metaphor. “Higher” appraisal is more “positive.” Negative appraisals are on a linear scale going in the opposite direction—downhill.

What this suggests is that metaphor is significantly constitutive of emotion, if appraisal is essential to emotion. So our metaphor systems are entering into our emotional systems. Not surprising, if they’re all there in the brain, and they’re all part of the same integrated system of circuitry.

That, I think, is crucial, since the concept of morality is based on metaphors for well-being, moral appraisal is based on primary metaphors for morality. And where emotion uses appraisal, emotion makes use of metaphorical scales.

What is the point of all this? The conceptual is inseparable from the emotional, and vice versa.

Now, a final point. Important words like freedom, empathy, and emotion do not have objective definitions. They are contested concepts.

The idea came from Walter Bryce Gallie in an article in 1956 called “Essentially Contested Concepts” in the Aristotelian Society proceedings. The most detailed study of a contested concept so far is my book (Lakoff, 2006), Whose Freedom? It is a 250-page book on how conservatives and progressives differ on the concept of freedom systematically in every domain of life. It follows Gallie’s theory. Gallie got it right and the idea is this: Suppose there is some central underspecified concept that we generally agree upon, like a simple physical freedom. Imagine you’re walking down the street, and somebody jumps on you and ties you up. You are not physically free.

If you apply the moral systems of conservatives and liberals to this underspecified notion of freedom, you get two very different extended concepts of freedom—opposite concepts in many domains.

So far as I can tell, “emotion” is a contested concept. It depends on the interests, skills, and academic ideology of the emotion researcher. There is no one “correct definition” of “emotion” or “basic emotion.” But there are real phenomena that it is reasonable to call “emotional” and they are probably of many types. I think it is important for each emotion researcher to get an understanding of exactly what values and frames lie behind her or his choice of phenomena to study, as well as the kinds of technology and analytic tools used, their intellectual background, their personal goals, and the academic politics involved. Personally, I think an open discussion of this would enlighten us all.

I am fortunate to have friends who are exemplary in this respect. First, Andrew Ortony, who states straightforwardly that emotions for him have to be either positive or negative, which would rule out surprise and interest from the start. Another example is Antonio Damasio, for whom emotions are body states that are “somatically marked” in the brain in certain regions (insula, amygdala, etc.). And of course, Paul Ekman, who has been following Darwin’s work on animal emotions by studying facial muscles and other aspects of physiology. I have learned from all of them, and from many others in this audience.

Words on the whole do not have unique, objective meanings. Cognitive mechanisms of various sorts—prototypes (of half a dozen kinds), radial categories, conceptual frames, metonymy, and conceptual metaphor tend to yield multiple meanings of words. If you are not convinced, take a look at Case Study 2 in my Women, Fire, and Dangerous Things (1990). It is an analysis based on Claudia Brugman’s book Story of Over (1988), which is an account of the word over and its over 100 related meanings.