The two faces of emotion
In practice, psychological research on emotions has focused on two quite separate phenomena. Many ethologists and social psychologists, following Charles Darwin, have focused on the expression of emotions, particularly on the face. These investigators have sometimes argued that emotional expressions are special forms of communication that are universal, and therefore probably inherited.17 Other researchers, particularly clinical psychologists, have been more concerned with subjectively experienced emotions as revealed by first-person reports such as ‘I feel depressed’, ‘I feel anxious’, or ‘I feel ashamed.’ It is easy to assume that there is a special relationship between these two types of phenomena, that each facial expression is associated with a particular subjective feeling. Consistent with this assumption, there is obviously some degree of correspondence between feelings and expressions – when we feel depressed we are more likely to frown than to smile. However, we should not therefore leap to the further assumption that expressions and subjective experiences are different manifestations of a single underlying process, that there is a discrete number of basic emotions that are reflected on our faces and in our minds.
The idea of basic emotions seems so natural and self-evident that it has dominated research on normal emotions, and also on the abnormal emotions encountered in psychiatric practice. It has particularly influenced research on facial expressions. In an important series of studies, Paul Ekman and his colleagues at the University of California in San Francisco have collected evidence that common expressions of anger, fear, enjoyment, sadness, disgust and possibly surprise are recognized in diverse societies of the world.18 However, not all psychologists have accepted this evidence uncritically. Canadian researcher James Russell has carefully scrutinized the work of Ekman and others, and has noted that most studies have employed forced-choice procedures in which informants are shown photographs of faces and are then asked to select from a list of possible emotions. Russell has argued that these procedures minimize disagreements and that there is less agreement when more open-ended procedures are used.19 Even when forced-choice procedures have been employed, differences between cultures are greater than Ekman has supposed. (For example, Japanese and American informants generally agree that certain expressions are best described as happiness, sadness and surprise, but show less agreement when asked to describe expressions corresponding to North American notions of anger, disgust or fear.)
Subjective emotions
In the case of subjective experience, the idea of basic emotions is closely related to the idea that we discover our emotions by a process of introspection. According to this theory (and, no matter how natural and self-evident it may appear, it is a theory), when I say that I feel depressed something happens along the following lines. First, I notice some kind of internal state or sensation; second, I recognize that inner state or sensation as the feeling of ‘depression’; and, third, I report it to anyone who is likely to be interested.
Some elements of this theory are uncontentious. For example, there is nothing wrong with the idea that we are influenced by events occurring inside our bodies. Writing in 1948, the influential American behaviourist B. F. Skinner observed that:
The response ‘My tooth aches’ is partly under the control of a state of affairs to which the speaker alone is able to react, since no one else can establish the required connection with the tooth in question. There is nothing mysterious or metaphysical about this; the simple fact is that each speaker possesses a small but important world of private stimuli.20
Some psychologists have suggested that this private world is the source of our emotional experiences. For example, in his celebrated two-volume Principles of Psychology published in 1890, the American psychologist and philosopher William James argued that:
Bodily changes follow directly the perception of the exciting fact, and our feeling of the same changes as they occur is the emotion. Common sense says… we meet a bear are frightened and run… The hypothesis here to be defended says that the order of the sequence is incorrect, that the one mental state is not immediately induced by the other [the perception of the bear], that the bodily manifestations must first be interposed between them, and that the more rational statement is we feel… afraid because we tremble.21
Recent versions of this theory have implicated various types of internal stimuli in subjective emotions, including arousal of the autonomic nervous system,22 feedback from motor responses involved in facial and bodily expressions,23 and even changes in brain blood temperature resulting from the disruption of vascular flow by facial muscle movements.24 Overall, there is quite convincing evidence that each of these processes does have some influence.25 For example, if people are persuaded to adopt particular expressions, their feelings tend to be affected in predictable ways. However, as long ago as 1927, the American physiologist Walter Cannon pointed out that bodily sensations are too diffuse to account for the rich variety of emotional experiences reported by ordinary people. More importantly, even if it was to transpire, following further physiological studies, that the range of internal stimuli is sufficiently varied, the problem of how people can learn to notice and report these stimuli as emotional states represents an insuperable obstacle to James’s theory.
In order to see the power of this criticism, we must consider how individuals come to possess a vocabulary of feelings and internal states. This question was considered by the Austrian-British philosopher Ludwig Wittgenstein, and was famously debated in his posthumously published book Philosophical Investigations, which first appeared in 1953.26 However, Wittgenstein’s analysis was anticipated in a less well-known paper published by B. F. Skinner in 1945.27 Although written in the terse jargon beloved by American behaviourists of the period, Skinner’s account is much simpler than Wittgenstein’s, and so it is his version of the argument that I will outline here.
Having established that we each possess an internal world of stimuli to which we have exclusive access (no one else can feel our toothache), Skinner points out that the way we learn to label those stimuli is far from obvious. The problem is that speaking is a social process in which the appropriate use of words depends on rules and conventions that we learn from other people.* These conventions can only be discovered in the context of the right kind of relationship between the person who is learning to speak and what Skinner describes as the wider ‘verbal community’. To take a simple example, learning object names like ‘cow’ or ‘dog’ presents no particular problem as both the learner and the verbal community have access to the stimuli that prompt these names (cows and dogs). I can correct you if you call a dog a cow, and my failure to respond appropriately to your request to ‘pat the cow’ may alert you to the fact that you have used the wrong name. However, these possibilities are absent in the case of internal stimuli. I cannot point to a particular profile of physiological states inside you, no matter how complex, and teach you to call it ‘depression’. Nor am I in a position to disagree with you should you label a physiological state as ‘depression’ on the grounds that I would call the very same state ‘anxiety’.
Nonetheless, we plainly do talk about internal states and Skinner discusses a number of indirect ways by which we can acquire the necessary vocabulary. Any publicly observable event that regularly accompanies a private stimulus may be exploited by the verbal community in much the same way that a sighted person might use visual cues when teaching a blind person to name objects. For example, a child’s protest, ‘That hurts’, may be responded to appropriately if it occurs following a fall resulting in a grazed knee. Similarly, the verbal community may encourage statements that are made together with particular types of non-verbal behaviour (groaning and clutching the jaw in the case of ‘toothache’, smiling and related gestures in the case of ‘happiness’). Nor must we assume that some kind of overt reinforcement is required for these kinds of learning to occur – children have many opportunities to observe in others associations between particular types of
emotional statements and the relevant public stimuli.
Some times we are able to learn terms that appear to describe internal states because those states occasionally occur in a public form. For example, the word ‘murderous’ may be used to describe an emotional state that typically accompanies murderous behaviour. Other emotion words, for example ‘agitated’ or ‘ebullient’, began their life referring to public events but have been extended to the psychological domain by a process of analogy. Recently, the British psychologist Graham Richards devoted an entire book to this phenomenon, and the following is just one of the many examples that he cites:
Fire-control terms have proved especially useful for the encoding of what we now term arousal states, and constitute an extension of the temperature-based PL [psychological language] (hot, cold, tepid, frigid, lukewarm, icy). They include the following:
to blaze up, to burn, to be burnt up/out/into one’s memory, to burn one’s fingers, to dampen down, to douse, dying embers (e.g. ‘her smile rekindled the dying embers of passion’), fiery, flaming, to fan the flames, to flare up, to fuel, to ignite (e.g. ‘his words ignited the passions of the crowd’), inflammatory, to kindle, to light (e.g. ‘I just want to light a flame in your heart’), to rake over the ashes, red hot, to rekindle, scorched, searing (e.g. especially in the phrase heart-searing), singed, to smoulder/smouldering, to spark off, to stoke up.28
The word ‘depression’ began its life in just this way. Until the early years of the nineteenth century, doctors labelled negative emotional states as ‘melancholia’, ‘neurasthenia’ or ‘moppishness’ (the last term being reserved for the uneducated classes). ‘Depression’ was first used in physical medicine to describe a reduction in cadiovascular function and was afterwards adopted by early psychiatrists to indicate emotional states that were considered to be the opposite of excitation.29 During the early years of the twentieth century the term intruded into the language of ordinary people, so that we now think of ‘depression’ as a natural label for how we feel during times of loss.
This account suggests the possibility that different cultures will describe subjective emotions differently. Collating the evidence that addresses this question, James Russell has demonstrated that societies vary dramatically in the range of words they allocate to emotional states.30 For example, the English language contains over 2000 emotion words, whereas most languages contain fewer than 200. Emotional states that appear quite fundamental from the perspective of an English speaker are not always mirrored in the lives and languages of other cultures. In some African languages, the same word covers what in English would be described as anger and sadness whereas the Gidjingali aborigines of Australia do not discriminate between fear and shame. Prototypical emotions that have a central place in Western concepts of psychopathology appear to be entirely absent in some cultures. For example, there is no word equivalent to depression in many non-Western cultures and no word equivalent to anxiety among Eskimos or Yorubas.
Even when emotion concepts do appear to translate from one language to another the relevant words in the two cultures may not be precisely equivalent in meaning. A number of studies have been carried out in which the Japanese equivalents of the English words ‘anxiety’ and ‘depression’ have been identified by the process of back-translation (translating the English words into Japanese and then checking that independent translators translate them back to their original English expressions). However, further studies of the meanings of these words (for example by asking people for other words they associate with them) have revealed substantial differences in the way that they are used in the two cultures.
Cultures even appear to differ in where they draw the line between emotional and non-emotional internal states. For example, the Japanese word ‘jodo’ has been translated as the equivalent of the English word emotion, but includes states that translate as considerate, motivated, lucky and calculating. To take another example cited by Russell, the Chewong of Malaysia, who are apparently limited to only seven words that clearly translate as emotion states, group together both thoughts and feelings, locating them in the liver. Of course, one possible way of interpreting these findings would be to assume that Japanese and Chewong informants are simply wrong about their emotional states. However, in pursuing this argument, we would be making the chauvinistic assumption that the structure of the human mind is uniquely and accurately reflected in the vocabulary of English.
The dimensional structure of subjective emotion
Despite the different ways in which cultures describe internal states, it is possible that a considerable degree of cross-cultural consistency exists when emotions are considered within a dimensional framework. The idea here is that, underlying the diverse ways in which different cultures describe their emotions, there may be a relatively small number of emotional dimensions that are universal. Various statistical methods have been used in an attempt to discover such dimensions, the most common being multidimensional scaling in which informants are asked to rate the similarity of various emotion words. The informants’ ratings are then analysed using a complex mathematical procedure, which tries to account for the ratings in terms of a minimum number of dimensions. (This approach is similar to the technique of factor analysis, which I briefly described in Chapter 4.)
This type of research has consistently shown that ratings of emotions can be plotted as a circular pattern or circumplex in a two-dimensional space. This structure has been found not only using words but also using photographs of facial expressions, and with people from such varying places in the world as North America, Greece, Spain, Vietnam, Hong Kong and Haiti. The two dimensions that define this space have in turn have been interpreted in different ways by different researchers. Russell has taken what might seem to be the commonsense view, and has suggested that the two dimensions are, first, pleasantness versus unpleasantness and, second, the degree to which the individual is physiologically aroused. On this view, pleasure is the opposite of dysphoria (negative emotion).31 In contrast, American psychologists David Watson and Auke Tellegren have argued that the same structure is better described in terms of two independent dimensions of positive and negative affect, which are revealed by rotating Russell’s model through 45 degrees32 (see Figure 9.1).
Figure 9.1 The two emotional circumplexes (adapted from B. Parkinson, P. Totterdell, R. B. Briner and S. Reynolds (1996) Changing Moods: The Psychology of Mood and Mood Regulation. Harlow: Longman). Russell’s model, on the left, proposes two dimensions of affect – positive vs negative and arousal vs sleepiness. Rotating this model through 45°, we arrive at Watson and Tellegren’s model of independent positive (PA) and negative affects (NA), which is on the right.
These systems are clearly inter-translatable (for example, a state of high positive affect according to Watson and Tellegren corresponds with a state of high arousal and high pleasure according to Russell). The debate about which system is preferable is there fore quite technical and, to some commentators, the differences between them are largely arbitrary. However, to my mind at least, Watson and Tellegren’s system has several compelling advantages.
Least importantly, Watson and Tellegren have argued that the majority of emotion words in most languages describe the regions of the circumplex that they have characterized as high positive affect and high negative affect. Therefore, they argue, these two types of subjective emotion are the most salient, and the most psychologically real, to most people in the world.
More importantly, Watson and Tellegren’s model of positive and negative affect clearly maps on to two fundamental psychological processes that are common to all animal species, and which govern much of our behaviour. Reinforcement (or reward in ordinary language) refers to the process by which behaviours tend to be repeated if they lead to stimuli that have positive consequences for the individual. Stimuli of this sort are known as reinforcers. The most basic reinforcers satisfy biological drives such as hunger and the need for sex. Not surprisingly, organisms (whether human or
non-human) tend to approach reinforcing stimuli, and the accompanying emotion experienced by human beings is positive affect. Punishment, in contrast, refers to the process by which organisms (whether human or non-human) tend to desist from actions that lead to negative consequences, such as painful stimuli. The emotion we experience when encountering such stimuli is negative affect. The neurophysiological systems responsible for these two processes are beyond the scope of this book; suffice it to say that there has been an impressive accumulation of evidence indicating that distinctive brain mechanisms are responsible for these processes and hence, presumably, for the two types of emotion.33
An important implication of Watson and Tellegren’s model is that, contrary to folk wisdom, positive and negative emotions can occur independently of each other. This claim has sometimes been regarded as controversial, partly because, in most circumstances, positive and negative affect are negatively correlated.34 (Stimuli that evoke positive affect tend not to evoke negative affect and vice versa, so that they often appear to lie at opposite ends of a single dimension of emotion.) However, even in ordinary life the two types of emotion sometimes occur together – for example during periods of transition between one phase of life and another, such as when starting at college or when moving to a new job.35 In a later chapter we will see that this model helps to make sense of the experiences of manic patients, who often experience positive and negative affect simultaneously.
Madness Explained Page 26