Chapter 10

HOLISM

“The totality of our so-called knowledge or beliefs, from the most casual matters of geography and history to the profoundest laws of atomic physics or even of pure mathematics and logic, is a man-made fabric which impinges on experience only along the edges. Or, to change the figure, total science is like a field of force whose boundary conditions are experience. A conflict with experience at the periphery occasions readjustments in the interior of the field. . . . But the total field is so underdetermined by its boundary conditions, experience, that there is much latitude of choice as to what statements to re-evaluate in the light of any single contrary experience. No particular experiences are linked with any particular statements in the interior of the field, except indirectly through considerations of equilibrium affecting the field as a whole.”

- Willard Van Orman Quine (1908-2000) “Two Dogmas of Empiricism”, §6, in From a Logical Point of View (Quine 1953)

            Our emotional judgments give us a guide to what truly matters. They provide a guide, however, that is diverse, plural, and particular.

            What matters is diverse because we have seen that no one sort of thing is all that matters. Neither having blissful experiences, fulfilling our human potential, nor satisfying our desires is all that matters to us. Emotional judgments, on the other hand, capture the full diversity of what is potentially meaningful.

            What matters is plural because things do not matter to us in just one way. Some things we marvel at, while others disgust us. Some people we love, while others anger us. Emotional judgments capture this plurality of meaning. Emotional judgments require no single, overall judgment of valuableness or meaningfulness. Different people, things, and events matter in a plurality of different ways.

            What matters is particular to each of us. We often find different activities enjoyable-to-each-of-us or find different people deserving of each of our anger. Sometimes we need a coordinated response to an event or person. Then we need to share emotional judgments: What he did deserves the anger of everyone. In questions of interpersonal morality, we need to make emotional judgments that are universal. To find what matters to each of us, however, we need to make only particular emotional judgments.

            Our search is not just for what feels as though it matters, but for what truly matters. Can our emotional judgments be true? The question of truth will occupy us for the next several chapters. Do we have good reason to think that an emotional judgment is the sort of thing that can be true? We expect truth of our beliefs but not of our desires, emotions, and other mental states. The question of whether emotional judgment can be true depends on the prior question of whether they are the sort of things, beliefs, that can be true or false or whether they are something else, such as expressions of feeling, that cannot. Are emotional judgments beliefs?

The Problem

            Emotional judgments are hypotheses about the emotions that we would have under distortion-free conditions. Emotional judgments predict our emotions in possible but non-actual circumstances.

            On the one hand, emotional judgments seem like beliefs. The most natural attitude to take toward a hypothesis or prediction is a cognitive one. We postulate, assume, consider, are certain about, or are convinced of a hypothesis. When we think a hypothesis to be true, then we believe it. Predictions are cognitive, not emotive. Predictions are beliefs about what we would feel in more ideal circumstances. Emotional judgments have a cognitive role.

            On the other hand, emotional judgments seem like emotions. Emotional judgments are evaluative. When we talk of someone being admirable, or deserving of anger, we have crossed the line from factual judgments to evaluative judgment. Evaluative judgments have something that factual ones do not: They move us in a certain way; they engage us; they give us reasons for caring or acting. The philosopher Peter Railton puts the point this way, “It would be an intolerably alienated conception of someone’s good to imagine that it might fail in any way to engage him.” (Railton 1986:9) Any answer to the question of what is meaningful must be an answer that matters to us; we want an answer that engages our emotions. Emotional judgments have emotive force.

            Now we have a problem. (Smith 1994:4-13) According to the common conception of belief, beliefs are dispassionate and inert. They do not move us, or engage us, or give us reasons for caring. They have no emotive force. If emotional judgments have emotive force and beliefs do not, then emotional judgments apparently cannot be beliefs.

            Our dilemma is this: If emotional judgments are beliefs, then they appear to lose touch with what matters. Yet if emotional judgments are not beliefs, then they appear to lose touch with truth.

            We can state the problem in another way. Emotional judgments have a certain functional role in our psychology. What type of psychological state they are is whatever type of psychological state fits this functional role. (Gibbard 1990:71-75) The functional role of emotional judgments has two parts:

(1) We are disposed to involve them in reasoning, in making predictions, in constructing arguments, and to assert them in conversation.

(2) We are disposed to be moved by them, to be emotionally engaged by them, and to have them matter to us.

            We can think of three possible types of occupants for this role. The occupants could be emotions, they could be beliefs, or they could be another, special type of psychological state for which we do not yet have a name. (Perhaps, on this third option, judgments are combination states that we could call ‘bemotions’, or ‘emoliefs.’) In the next section, we will see that they are not emotions. In the final section, we will see that, contrary to common philosophical opinion, beliefs about emotions under mistake-free conditions are connected to emotive force. If the conception of all beliefs as disconnected from emotion is wrong, then we have no need to postulate a special, sui generis, type of psychological state to fill the role of emotional judgments. If such beliefs have emotive force, then we can see emotional judgments as beliefs and thus as assessable for truth or falsity.

Judgment and Emotion

            Emotions are involved in highly specific sorts of value judgments, as when we judge someone admirable or despicable or judge an event enjoyable or boring. Either emotions have a cognate value judgments (admiration/admirable) or we can easily construct an associated value judgment (anger/deserving of anger). The simplest view of evaluative judgments would be that an emotional judgment expresses the cognate emotion. When we say that so-and-so is boring, we are expressing our actual boredom with him. This view automatically makes a connection between an emotional judgment and its cognate emotion. Yet, this view implies that judgments, as expressions of emotion, are not the sort of thing that can be true or false.

            The view that emotional judgments are expressions of their cognate emotion makes the connection between judgment and emotion too tight. Someone who is grieving the recent loss of a loved one may be unable to enjoy life, yet still sincerely judge that life is, in fact, enjoyable. We should not take this person’s judgment that life is enjoyable as an expression of his enjoyment of life. In his grief, he is not enjoying life at all. When we make emotional judgments, we try to compensate for any unusual factors that we think may be affecting our emotions and leading us astray in our judgments.

            Emotional judgments are predictions regarding the emotions that we would have in possible, but seldom actual, circumstances where we are free of the distortions to which emotions are prone. Sometimes our actual emotion is the same as our predicted emotion, and sometimes it is not. Nonetheless, predicted emotions are not actual emotions.

            To see this, imagine someone who does not now admire a certain public figure, but who predicts that she would become an admirer if she knew more about the public figure. Her prediction is not the actual emotion of admiration.

            Similarly, someone can predict, now, that when an older friend of hers dies in the future, she will feel sadness and grief. She judges, correctly, that her friend’s death will be a sad event to her. Judging this future event to be sad and predicting her future grief is not actually to grieve in the present. The judgment and the emotion are different.

            Because of the many ways that emotion can go wrong, the simple view that an emotional value judgment expresses the cognate emotion cannot be correct. One person’s judgment that a second person deserves her anger does more than just express her anger at him. On a rational-emotion account of emotional judgments, she judges that he is deserves her anger just in case she would be angry with him if she were free of the sorts of distortions to which emotions are prone, or at least were able to discount for them. When she judges that he deserves her anger, she also implies that she is not making any of these sorts of mistakes. If this account of emotional judgments is roughly right, then emotional judgments express conditional emotions, emotions that we may not actually have, but which we would have if we were in the right, distortion-free, circumstances.

            Rational emotions can differ from actual emotions in two ways. We can illustrate this by focusing on one way that emotions can be mistaken, that of being dependent on false beliefs.

            First, suppose that one person is angry with another. Either finding out more information or finding that some of her beliefs about him are false – finding out, for example, that he did not do some of the deeds for which she is angry with him – may bring her to lose her anger at him. Although she is angry with him, more information would eliminate her anger, and so he is not deserving of her anger.

            Second, suppose that she is not actually angry with him, but instead is neutral about him. Finding out more information about him or that some of her beliefs about him are false – finding out, for example, that he did do some deeds that she had previously believed he had not – may bring her to gain anger at him. Although she is not angry with him, more information would make her angry with him and so he deserves her anger.

            Notice something important. To deal with the second sort of case, the rational-emotions account employs a hypothetical emotion that she does not really have. In the second case, she is not actually angry with him, though if she knew more, then she would be angry with him. Her rational emotions are hypothetical emotions that she would have only if she were fully and correctly informed. What are these rational emotions?

            The most natural interpretation presents a problem for the view that her judgment that he deserves her anger merely expresses anger for him. On this interpretation, rational emotions are conditional, hypothetical emotions, not actual ones. So, they are not real emotions, the ones that make things matter to her. If, as avoiding the first problem required, her rational emotions were just an edited set of her actual emotions, from which inappropriate emotions were subtracted, then the connection to mattering would remain. Her edited set of actual emotions would still be a set of actual emotions. However, to avoid the second problem, the rational-emotion theory has to add emotions that she does not actually have. These conditional emotions are not real emotions. In some hypothetical situations, new information would result in her gaining an emotion that she does not in fact have. New information about him would lead her, in hypothetical circumstances, to be angry with him. Anger in hypothetical circumstances, however, is not real anger. So her judgment that he deserves her anger does not express actual anger for him.

            We should not think of emotional judgments as emotions at all. Emotional judgments are hypotheses about the emotions that we would have under mistake-free conditions. Hypotheses are not emotions. Hypotheses can be accurate or inaccurate, correct or mistaken, and true or false. Hypotheses can be confirmed or disconfirmed by new evidence. Despite their name, emotional judgments are more like beliefs. They are fallible beliefs about what people would conclude under hypothetical circumstances. Now, however, if we take emotional judgments to be expressions of belief, then we must face the problem of connecting emotional judgments to their cognate emotions.

Hypothesis Testing

            The conclusion of the previous section was that our emotional judgments are hypotheses that predict the emotions that we would have under distortion-free conditions. Hypotheses can be confirmed or disconfirmed by the evidence. Hypothesis testing is holistic.

            A too simple picture of scientific reasoning goes like this. A scientist formulates a general hypothesis, performs an experiment, and conclusively confirms or disproves the hypothesis. This simple picture is wrong because whenever a scientist tests a hypothesis, she has to make assumptions about the environment of the experiment, about the equipment that she uses, and about other aspects of scientific theory. If her experiment fails to produce the result predicted by the hypothesis, then all she can logically conclude is either that her hypothesis is wrong or that one or more of her assumptions are wrong. She cannot conclusively infer that her hypothesis is wrong.

            Even in unsophisticated experiments, evidence is never logically conclusive. Suppose that her hypothesis implies that a voltmeter will measure a certain voltage. If the voltage is different from what she predicts, then it could be that her hypothesis is wrong, or it could be that no one has calibrated the voltmeter correctly, or that its needle is bent, etc. The situation is far more complicated in a high energy physics experiment performed to detect a previously unknown particle. Physicists must presuppose enormous amounts of physical theory, and perform complex calculations, to get a prediction. If the prediction is wrong, it might be because the hypothesized particle does not exist, or it might be some problem either in the rest of the theory or in the experimental design.

            Only if the assumptions are better confirmed than the hypothesis can the scientist conclude anything. On the balance of probabilities, it is the hypothesis that the evidence disconfirms and not the assumptions. The assumptions are better confirmed than the hypothesis if they also have been previously subject to testing and if they fit with other assumptions of the scientist’s overall theory.

            In the chapter motto, Quine offered, as a metaphor for the holism of science, the idea of a “fabric” or “force field” of scientific belief tacked down only at the edges by the evidence. Perhaps more familiar would be Susan Haack’s metaphor of a crossword puzzle. (Haack 1993:84-86) Consider the following very simple little example:

■   ■ ¹B  ■   ■   ■

■² F   L   0 ³W  S

■   ■  U   ■   I   ■

⁴W H  E   E  L  S

■    ■  ■   ■  D   ■

Across: 

2. A river does this. 

4. A bicycle has two of these.                        

Down:

1. A color

3. Untamed

The clues provide inconclusive evidence for the correct entries. In one-down, for example, the clue is a four-letter word meaning a color. Both “blue” and “pink” fit the evidence of the clue, but only “blue” fits with two-across, “flows,” and with four-across, “wheel.” These latter entries are auxiliary hypotheses that lead us to prefer “blue” to “pink.” These entries, in turn, fit with the evidence in their clues and with the reasoning in other entries. On the evidence of its clue, two-across could be “flood.” However, that would not fit with three-down, “wild.” And so on. In this analogy, the clues provide evidence for an entry, and its fit with other entries provides auxiliary reasons for an entry.

            Our actual entry is the one that we will make based on the clues and on the other entries that we have already made. The true entry, however, is the one that we would make if all our intersecting entries were true ones. Intersecting entries are true, in turn, if they fit both with their clues and with the entries that they intersect. The truth of any entry depends on getting the whole puzzle to fit together. Whatever truth might be, it requires a holistic fit with all possible reasons and evidence.

            At this point the analogy breaks down. For a crossword puzzle, we might think that the correct answer is the one that corresponds to what the puzzle’s creator had in mind. The clues, and the ways words fit together, are merely means to guessing what the puzzle creator was thinking. For the Godless, however, the real world has no creator. No divine mind exists for us to guess.

            The structure of evaluative reasoning is similar. On the one hand, our judgments try to fit with our actual emotions. On the other hand, our judgments must fit with auxiliary judgments that we assume based on past evidence and reasoning.

Distortion-free Emotions

            Our actual emotions are evidence for our emotional judgments, but not the sole evidence. Our primary hypothesis is that we would feel a certain emotion does not face the evidence of our actual emotions alone. It does so only in conjunction with the auxiliary hypothesis that we are avoiding the sorts of distortion to which emotions are prone.

            What are the distortion-free conditions for making an emotional judgment? Distortion-free conditions are an ideal that we will seldom attain. Even if we do, on occasion attain them, we will be unable to know that we have done so. In distortion-free circumstances, we would know everything that is relevant. For any particular emotional judgments this may be possible, since not all that much may be relevant. Yet, to know that we know everything relevant, we would have to know everything there is to know. Examining each piece of knowledge is the only way to decide whether it is relevant. Given our finitude, we can never fully specify what distortion-free conditions are.

            Nevertheless, we can still make emotional judgments. We can do so because we have some knowledge of the distortions to which emotions are prone. When she judges that he deserves her anger, her main hypothesis is that she would feel anger toward him. However, she understands that the evidence of her actual emotions will only confirm her prediction if various auxiliary hypotheses hold. Her auxiliary hypotheses are that she is not making any of the mistakes to which anger is prone, mistakes about direction, cause, information, evaluation, focus, intensity, and physiology. Her auxiliary hypotheses are not ad hoc because they are based on experience with emotions. Not just any hypothesis can be added here, only ones based in the nature of emotions.

            Now compare emotional judgments with a simple perceptual judgment. Someone’s judgment that something is green expresses the belief that, if his visual system were free of the distortions to which vision is prone, then he would see it as green. His evidence for this primary hypothesis is his experience of the object as green. Evidence against his auxiliary hypotheses could change his belief. Such evidence would be information about the ambient lighting, information about faults in his perceptual system, information about the effects of background on color experience, and so on. The acquisition of this auxiliary evidence, however, will not usually effect how he experiences the object. Finding out that the background light is yellow, not white, may change his belief in the color of the object, but he will mostly still see it as green.

            The situation is different for emotional judgments. Someone’s judgment that someone else deserves her anger expresses the belief that, if her emotions were free of distortion, then she would be angry with him. Her primary evidence for this belief is that she finds that she is angry with him. Additionally, she believes auxiliary hypotheses that she is not making the sorts of mistakes mentioned above as those to avoid in getting anger right. For example, information he did not do what she believes he did can change her judgment that he deserves her anger. In contrast to the perceptual case, however, the acquisition of this auxiliary evidence will, over time, also change her emotional response to him. Finding out that he did not do it will change both her belief that he deserves her anger and her feelings of anger toward him. A mechanism connects beliefs and emotions. Emotions change, albeit slowly and with help, to converge with their corresponding emotional beliefs. Feedback mechanisms generate a slow convergence between emotional beliefs and their cognate emotions.

Judgment, Belief, and Emotion

            We are now able to explain the nature of the internal connection between an emotional judgment and its corresponding emotion. First, though, seeing what we are not trying to explain is important.

            Sometimes people think that there has to be a necessary or conceptual connection between valuing and mattering or between an emotional judgment and its cognate emotion. If the connection were necessary, then we would invariably find the emotional judgment and the cognate emotion together. However, this overstates the problem. Reflection shows that a necessary connection is not what we should be looking for. From common experience, we can see that having an emotional belief without the cognate emotion and vice versa is always possible.          A grieving person lives in a mood of prevailing sadness, unable to enjoy the world around him. Nevertheless, she may still believe that the world is an enjoyable place. When asked, she may sincerely avow that the world is, in fact, enjoyable. It is just that, in her melancholy, she does not feel any enjoyment. Here we have an emotional belief without the cognate emotion.

            A person with a phobia about mice may not believe that mice are fearsome animals. Nevertheless, he still fears them. No matter how strong the evidence that mice are not dangerous, he cannot shake his fear. Here we have a strong emotion without the cognate emotional belief.

            Thus, we are not searching for a necessary connection between belief and emotion. We are searching for something weaker. This internal connection is the same as the rational connection between belief and evidence. We can think of someone’s emotional judgment, for example that he deserves her anger, as a hypothesis that is subject to confirming or disconfirming evidence. A primary piece of evidence is her emotional response to him. Like all hypotheses, her belief is never tested alone. She can test it only in conjunction with the auxiliary hypothesis that she is free of the distortions to which emotions are prone. In a test situation, her hypothesis that he deserves her anger will imply that she will be angry with him only in conjunction with the auxiliary hypotheses that he did do the things that she believes he did, and that she is not subject to any undermining psychological conditions. If she finds that she is angry with him, then this still may not confirm the hypothesis that he deserves her anger. One or more auxiliary hypotheses may be false. Maybe he did not do the deed, or her psychological conditions are inappropriate. Normally, however, whenever she judges someone to deserve her anger she also is angry with him, and whenever she is angry with someone she also believes him to deserve her anger. “Normally” here catches the rational relationship between hypothesis and evidence, and between emotional judgment and actual emotion.

            Thus emotional judgments can be beliefs because, as the special sort of belief that they are, normally they will have emotive force. Emotions and emotional judgments will gradually converge. When we make emotional judgments, we make hypotheses. We test these hypotheses in much the same way that we test factual hypotheses. Both judgment and emotion can change in the testing process, and the result, over time, is convergence.

            The insight that emotional judgments are beliefs not emotions, is tremendously important. For beliefs can be true or false while emotions cannot. Emotions can be appropriate, justified, or fitting, but not true. Yet, when we are searching for an answer to the question of whether our lives can be meaningful, we want an answer that not just feels right but that is right. The search for meaning is the search for truth, and beliefs are the proper candidates for truth.