Encyclopedia of Educational Philosophy and Theory

2017 Edition
| Editors: Michael A. Peters

Wittgenstein and the Learning of Emotions

  • Steinar Bøyum
Reference work entry
DOI: https://doi.org/10.1007/978-981-287-588-4_401


It is a methodological precept of Wittgenstein’s later philosophy that greater clarity may be obtained by looking at how words are taught and learned: “One thing we always do when discussing a word is to ask how we were taught it. Doing this … destroys a variety of misconceptions” (Wittgenstein 1966, I, §5). Looking at how we learn psychological concepts can therefore be a way to break the hold that certain pictures of psychological states have on us. It may also be a way to break the hold that certain pictures of learning have on us.

Yet learning is not a uniform phenomenon. In his work on the foundations of mathematics, Wittgenstein talks about the motley of mathematics (Wittgenstein 1956, III, §48). Likewise, the concept of learning covers a motley of processes, and we should resist the impulse to force them all into a single mold or two. The aim of this chapter is accordingly to dip into the great variety of things that learning can be, concentrating mainly on some of the characteristic ways in which we learn emotional concepts. In doing so, light will also be thrown on what Wittgenstein may have meant when he compared the learning of emotions to the learning of musical styles (Wittgenstein 1992, pp. 42–43).

Asymmetry and Expression

A good place to begin is the so-called asymmetry between first- and third-person statements about mental states. This asymmetry has two aspects. First, I am the authority on my thoughts, feelings, and emotions. If you want to know what I think or feel, then, normally, I am the one to ask (Finkelstein 2003, p. 9). You may also ask someone who knows me well, but the validity of that person’s judgments will ordinarily be conditional on my acceptance of them. Second, I do not usually need to back up my claims about what I think and feel with behavioral evidence (Wittgenstein 1967, §472). When someone says that she is angry, we might very well ask why, but not how she knows. In special cases, we may think she is wrong, but if so we have to adduce quite strong evidence, and even then her sincere avowal may cancel out the evidence.

When we are in a certain philosophical frame of mind, this asymmetry might seem puzzling. How can we be said to know something simply on the basis of our saying so, without citing any evidence and sometimes even against the evidence? I do not know anything about my brain, but even when I am in an fMRI-machine I am the one to ask about what I feel and think, not the neuroscientist. And even though my wife seems to attend more carefully to my behavior than what my distracted self does, I still get to have the final say on what I feel and think. Why?

According to what Finkelstein (2003) calls detectivism, I have direct knowledge of my thoughts and feelings because I, unlike others, am able to detect them. Analogous to our perception of the outer world, we perceive (or “scan” or “monitor”) our own inner world and report on what we “see” there. Other people do not enjoy this special access to my inner states, and therefore, they have to deduce them from more or less unreliable behavioral clues. This kind of view is a prominent target of Wittgenstein’s criticism in his remarks on psychology, including the private language argument. A common thread in these remarks is Wittgenstein’s juxtaposition of detectivism with expressivism: we dissolve many of the philosophical problems about first-person ascriptions of mental states by seeing them as expressions rather than reports (Wittgenstein 1967, §472).

Wittgenstein’s target here is just as much a view of the body as a view of the mind. It is not only the picture of the inner world that creates problems but the corresponding picture of the body as a mere body, according to which bodily expressions are really only physical movements, devoid of significance, and in need of interpretation to invest them with meaning. Against this, Wittgenstein maintains that there is a literal sense in which we make emotions visible (or audible) by expressing them. Crying makes sadness manifest, and saying that you are sad can do the same (Finkelstein 2003, p. 93). When we talk about seeing what other people feel, this should not be dismissed as merely a metaphor for interpretation. Indeed, a child that had to learn how to interpret these expressions with the help of rules like “If people cry, they are sad” would be missing something.

“We see emotion.” – As opposed to what? – We do not see facial contortions and make inferences from them (like a doctor framing a diagnosis) to joy, grief, boredom. We describe a face immediately as sad, radiant, bored, even when we are unable to give any other description of the features. – Grief, one would like to say, is personified in the face. This belongs to the concept of emotion. (Wittgenstein 1967, §225)

This “expressivism” should not be understood as a general theory, but as an illuminating analogy. If we are puzzled about first-person authority, bodily expressions can function as helpful objects of comparison (Wittgenstein 1953, §130; Kuusela 2013). Seen in that light, it should be no more puzzling that I am the best one to ask than that “my face is the best one to look at” if you want to know how I feel (Finkelstein 2003, p. 101). Likewise, if we are mystified by the fact that we do not normally need evidence for our claims about how we feel, it may help to note that neither do we need evidence for our smiling or crying. Yet since Wittgenstein’s view is that psychological self-ascriptions are akin to expressions, there are differences, too. Unlike a smile, for instance, an avowal can be said to be true or false, as Finkelstein (2003) emphasizes. Moreover, psychological self-ascriptions can be used like reports in some situations, for instance, in the therapist’s office.

Natural Expressions

The most famous and controversial of Wittgenstein’s ideas on the learning of psychological words is his suggestion that they are learned as replacements for natural expressions:

Here is one possibility: words are connected with the primitive, the natural, expressions of the sensation and used in their place. A child has hurt himself and he cries; and then adults talk to him and teach him exclamations and, later, sentences. They teach the child new pain-behaviour. “So you are saying that the word ‘pain’ really means crying?” – On the contrary: the verbal expression of pain replaces crying and does not describe it. (Wittgenstein 1953, §244)

Wittgenstein’s remark is on pain, but it seems reasonable to extend it to emotions like anger or sadness. Certainly, learning emotions involves being taught to use words instead of reacting “primitively”: to say that one is angry instead of raging, or to go from screaming to crying to speaking. There is here both a gradual calming (or disciplining) of bodily expressions and a partial replacement of those expressions with words. This partial replacement allows for both reflection on and refinement of emotions: reflection, in the sense that language enables the child to think and talk about emotions, and refinement, in the sense that the child’s space of expression becomes infinitely more nuanced and complex with language. Hence, learning emotions also involves coming to have new emotions – hope is often cited as an emotion that can only be had by those who possess a language.

Many have been critical to this picture of learning. To note just one problem, it does not seem to be valid for all mental states, not even all emotions. More complex emotions, like guilt, do not even have natural expressions in the same sense as more elementary emotions like anger or sadness. They may have characteristic expressions, like a “guilty look,” but these are more symbols of guilt than they are natural expressions of it. Hence, the replacement thesis looks incomplete and without empirical support. The question is what kind of thesis it is and whether it is a thesis at all. After all, Wittgenstein introduces it by calling it a mere possibility. In a similar discussion of language learning, Wittgenstein asks, “Am I doing child psychology?”, and he answers, “I am making a connexion between the concept of teaching and the concept of meaning” (Wittgenstein 1967, §412). Hence, it is a logical point rather than an empirical one about the actual genesis of concepts.

That emotions like fear and joy have natural expressions is one of those general facts of nature without which “our normal language-games lose their point” (Wittgenstein 1953, §142). Children’s learning of the language of emotion thus rests on these natural expressions: without them, the grammar of emotion would be very different; indeed, there is a sense in which there would be no such thing as fear and joy in that event. There is, that is to say, an internal connection between these natural expressions and the emotions that they express: knowing what sadness is involves knowing that crying is typically a reliable basis for ascribing sadness to others (and if they fake, then it is sadness they fake). The logical point can also be formulated temporally: we do not first know what sadness is and then make the empirical discovery that there is a connection between crying and sadness. In the words of Lars Hertzberg (2014), the account in Wittgenstein 1953, §244 is “an account of what might be termed a logical order: an indication of the circumstances in which we would be prepared to say that someone has learnt verbal expressions of pain” (p. 368).

Patterns of Life

One simple way to understand the philosophy of the later Wittgenstein is to say that it involves a widening of the contextual principle of Frege and the Tractatus (Finkelstein 2003, pp. 107–109). The meaning of words must be seen not only in light of the sentences they form part of but also in light of the particular situation and narrative to which they belong (Wittgenstein 1953, §525; Wittgenstein 1953, II, §ix). The same applies to all emotional expressions: only in the context of a “far-reaching particular manifestation of life” is there such a thing as the expression of, say, sorrow (Wittgenstein 1967, §534).

In Wittgenstein’s later writings, the contextuality of emotions comes out most sharply in his idea of patterns in the weave of life. In this perspective, emotions are seen as constituted by complex and dynamic configurations of words and gestures, actions and reactions, appearing within the stream of life. Emotions like love and grief are more like such patterns than they are like feelings in the narrower sense:

“Grief” describes a pattern which recurs, with different variations, in the weave of our life. If a man’s bodily expression of sorrow and of joy alternated, say with the ticking of a clock, here we should not have the characteristic formation of the pattern of sorrow or of the pattern of joy. “For a second he felt violent pain.” – Why does it sound queer to say: “For a second he felt deep grief”? Only because it so seldom happens? (Wittgenstein 1953, II, p. 174)

The importance of context is not at odds with the importance of immediacy mentioned above. Dialectically put, an expression is only immediate when mediated by context. A smile is only a smile in a face, which again is only the kind of smile it is as part of a situation (Wittgenstein 1953, §583). Wittgenstein remarks that if we are to imagine a kind smile or a malicious smile, we typically imagine the face or, rather, the human being as a whole, within different contexts, smiling at playing children or at suffering enemies (Wittgenstein 1953, §539). Still, within those contexts we can see the kindness in the kind smile and the maliciousness in the malicious smile.

Acquiring concepts of emotions can therefore be understood as learning how to recognize certain patterns in the weave of life. These can be of an almost endless variety. No two patterns of grief are exactly alike – they may even be completely unlike each other. Moreover, “one pattern in the weave is interwoven with many others” (Wittgenstein 1967, §569). The plasticity of emotional concepts, the variety of patterns they cover, as well as their being intricately intertwined with each other make it hard to understand how we can learn to recognize them. Wittgenstein suggests that we tend to learn the simple figures first and then proceed to the more complicated, “the way I learn to distinguish the styles of two composers” (Wittgenstein 1992, pp. 42–43).

Early on we learn what typically makes people happy or sad, whereas later in life we come to understand that one can become sad by happy events. Still, the concept of sadness would be very different from ours if we learned the concept of sadness in these latter circumstances. Hence, the intricate variety of emotional patterns is rooted in simpler connections. Analogously, even if we are inclined to call the letter “e” yellow, as Wittgenstein notes in his discussion of secondary sense (Wittgenstein 1953, II, p. 216), our color concepts would not be what they are if we learned them in connection with letters. As a matter of logical grammar, children will have to learn the concept of sadness in the context of sad things and may later learn to transfer it to very different cases (Wittgenstein 1982, pp. 966–967).

Imponderable Evidence

Agreement is part of the grammar of mathematics in the following sense: if we arrive at different results, then at least one of us is doing something wrong, and usually we will find out who (Wittgenstein 1976, p. 107). Now it would be an exaggeration to say that disagreement constitutes psychological concepts, since conflicting judgments are quite exceptional in the face of raw expressions of elementary feelings. In modern adult life, however, such raw expressions are almost the exception, and concerning judgments of subtler, verbal expressions, disagreement is nearly the norm (Wittgenstein 1953, II, p. 227).

Wittgenstein (1953, II, p. 228) remarks that judgments about sincerity and pretense are often backed up by imponderable evidence, which includes “subtleties of glance, of gesture, of tone.” These nuances may be difficult to describe except in the vaguest of terms: “something about his smile,” “the way he looked at her,” or “as though a shadow came over his face.” This is a kind of evidence, but not the kind that any rational being has to see or accept (indeed, in many such cases seeing is accepting). True, we do not judge the sincerity of an expression merely on the basis of what is given to us in the moment but also by the surroundings, in particular our past dealings with the person in question. Moreover, pretense will have different consequences than sincerity, and these may decide the issue, for instance, if we hear her laughing after she thinks we have left. Then again, the question may arise again: wasn’t there something hollow about her laughter? The consequences may be just as hard to agree on as the original expression and the imponderable evidence just as ineradicable.

The pervasiveness of imponderable evidence means that learning emotions involves learning to appreciate this kind of evidence, although the skill with which we do so varies widely. Wittgenstein suggests that there is a connection between the judging of imponderable evidence and being a “Menschenkenner,” an expert judge of character (Wittgenstein 1953, II, p. 227). There is a striking illustration of this kind of knowledge in a novel by his Austrian contemporary Robert Musil, where “Menschenkenntnis” is translated as “knowledge of human nature”:

Keeping company with the prince thus became a source of refined psychological pleasure for Törless. Dawning within him was the kind of knowledge of human nature that teaches us to know and appreciate another person by the fall of his voice, the way he picks something up, even the timbre of his silence and the expression of the physical posture with which he occupies a space; in short, by that agile way, barely tangible and yet the only truly complete way, of being something spiritual and human, which is layered around the tangible, effable core as around a bare skeleton, and by means of that appreciation to anticipate his mental personality. (Musil 2001, p. 8)

One can learn to master the art of judging character, Wittgenstein adds, but only through experience, perhaps accompanied by an expert judge who can teach us by hints and tips. Here, we do not first learn a method and then arrive at whatever results or judgments that the method leads to. We learn, Wittgenstein says (Wittgenstein 1953, II, p. 227), correct judgments (“Mummy looks a bit worried today”), perhaps helped by a few hints (“Isn’t there something distant in her eyes, as though she’s not really listening?”). There is no method or technique involved apart from seeing for yourself, guided by the verdicts of the more experienced. Gradually, one comes to formulate judgments of this sort oneself, autonomously, as it were, judgments that may serve as guides for others: the novice has then become a teacher. “This is what “learning” and “teaching” are like here” (Wittgenstein 1953, II, p. 227).


A nice way to summarize is by returning to Wittgenstein’s comparison between learning emotional concepts and learning styles of music (Wittgenstein 1992, pp. 42–43). The analogy highlights several interconnected similarities. First, primitive reactions and natural expressions play a corresponding role in the two fields, as has been well explored by Simo Säätelä (2002). Second, recognizing emotions and distinguishing musical styles involve perceptual discernment rather than rule-based reasoning. One may see the difference between real and feigned sorrow, and one may hear the difference between Beethoven and Brahms. Third, the analogy suggests that an emotion is more like a piece of music as a whole than a single note within this piece. And the character of a single note depends on its place in the piece, like the character of a single expression depends on its place in a pattern of life. Fourth, in both fields we learn by moving from simpler to more complex examples. We are shown the simplest and most typical cases first, and then, when recognizing these has become a matter of course, we proceed to more complex and less typical cases. Fifth, imponderable evidence is vital in both psychology and music: quite often we are not able to justify our judgment in any other way than by gesturing to the most delicate of nuances.

In this way, emotional concepts and musical styles function as objects of comparison for each other, highlighting similarities (and differences) in order to clarify their grammars and destroy misconceptions of them by attending to how they are learned.



  1. Finkelstein, D. H. (2003). Expression and the inner. Cambridge, MA: Harvard University Press.Google Scholar
  2. Hertzberg, L. (2014). Very general facts of nature. In M. McGinn & O. Kuusela (Eds.), The Oxford handbook of Wittgenstein (pp. 351–374). Oxford, UK: Oxford University Press.Google Scholar
  3. Kuusela, O. (2013). Wittgenstein’s method of conceptual investigation and concept formation in psychology. In T. P. Racine & K. L. Slaney (Eds.), A Wittgensteinian perspective on the use of conceptual analysis in psychology (pp. 51–71). London: Palgrave Macmillan.CrossRefGoogle Scholar
  4. Musil, R. (2001). The confusions of young Törless. New York: Penguin.Google Scholar
  5. Säätelä, S. (2002). “Perhaps the most important thing in connection with aesthetics” Wittgenstein on “aesthetic reactions”. Revue Internationale De Philosophie, 219(1), 49–72.Google Scholar
  6. Wittgenstein, L. (1953). Philosophical investigations (trans: Anscombe, G. E. M.). Oxford, UK: Blackwell.Google Scholar
  7. Wittgenstein, L. (1956). Remarks on the foundations of mathematics (trans: Anscombe, G. E. M.). Oxford, UK: Blackwell.Google Scholar
  8. Wittgenstein, L. (1966). In C. Barrett (Ed.), Lectures and conversations on aesthetics, psychology and religious belief. Berkeley, CA: University of California Press.Google Scholar
  9. Wittgenstein, L. (1967). Zettel. (trans: Anscombe, G. E. M.). Oxford, UK: Blackwell.Google Scholar
  10. Wittgenstein, L. (1976). In C. Diamond (Ed.), Lectures on the foundations of mathematics. Ithaca, NY: Cornell University Press.Google Scholar
  11. Wittgenstein, L. (1982). Last writings on the philosophy of psychology (Vol. 1). (trans: Luckhardt, C. G., & Aue, M. A. E.). Oxford, UK: Blackwell.Google Scholar
  12. Wittgenstein, L. (1992). Last writings on the philosophy of psychology (Vol. II). (trans: Luckhardt, C. G., & Aue, M. A. E.). Oxford, UK: Blackwell.Google Scholar

Copyright information

© Springer Science+Business Media Singapore 2017

Authors and Affiliations

  1. 1.Department of Education, Faculty of PsychologyUniversity of BergenBergenNorway