I think that our commonsense

Richmond Journal of Philosophy 1 (June 2002) I think that our commonsense intuitions tend to draw us in two apparently opposing directions about this question. On the one hand, we are inclined to say that we can certainly trust our emotions. Indeed, our emotions can sometimes tell us things about the world that reason alone will miss, as all the recent books about ‘emotional intelligence’ attest. In this respect, our emotions serve us very well; that, one might think, is why we evolved as creatures with emotions. Yet, on the other hand, we are inclined to say that our emotions can and do profoundly distort our view of things: in anger or jealousy, for example, when the red mist comes down over the eyes, and we can feel the blood pulsing in the temples, things look other than the way they are, and, accordingly, our emotions can mislead us profoundly; literature is replete with examples. A cheap resolution of these competing intuitions would be to say that there are cases and cases: sometimes our emotions help us to gain knowledge of the world around us, and sometimes they hinder us. No doubt this is true so far as it goes, but I think there is more to be said than just that. So what I want to do here is to say what lies behind these competing intuitions, and the sense in which both are right (without simply appealing to the idea that there are cases and cases). But in doing this I want to raise what I see as two quite deep epistemological worries about the emotions as a source of empirical knowledge—that is to say, they are worries about whether the emotions can help us to know certain things— empirical facts—about the world around us. I think that there is something especially troubling about the emotions here. The value of emotions: instrumental and non-instrumental To the extent that the emotions do serve the purpose of being a source of knowledge about the world, then they are of instrumental value. (A knife, for example, is of instrumental value in so far as it is useful for cutting things, and if it ceases to be able to do that, then it is no longer of instrumental value in that respect.) Although my focus here will be only on the potential instrumental value of emotions—on their value in enabling us to see things as they really are—I also think that they can be valuable non-instrumentally; that is to say, they can be valuable even if they serve no further purpose or create no other sort of valuable thing. This is in contrast to a popular view in ethics, which I think should be resisted, which goes only part of the way in accepting the value of emotions. This is done by, first, accepting that emotions matter, and then, secondly, going on to insist that they only matter because of the way that they can affect other things— their effect on how you act, for example. In other words, emotions are only valuable instrumentally. I will not argue the point here, partly because it is not the central topic of the paper, and partly because I find it very difficult to think of an argument that can persuade someone who disagrees. So, rather than putting up an unpersuasive argument (one that will only persuade the already converted, and what is the use of that?), I will give Richmond Journal of Philosophy 1 (June 2002) Peter Goldie an example; examples can sometimes persuade where arguments cannot. It comes from a wonderful novel, Disgrace, by the South African writer, J. M. Coetzee. The central character, David Lurie, is a university lecturer who is forced to resign in disgrace from his post for having an affair with a student, which involves him treating her very badly. He tries to make a life of sorts with his daughter, who has a hopeless smallholding deep in the country. He starts, as part of his new life, to help a friend of his daughter, Bev Shaw, at the animal welfare clinic that she runs. Many sorry animals come in, and many have to be put down, either because they are sick, or because no one wants them. David asks Bev, when he first visits the clinic, whether she minds putting down the animals. She replies ‘I do mind. I mind deeply. I wouldn’t want someone doing it for me who didn’t mind.’ Much later, right at the end of the novel (after some awful things have happened to David and his daughter), there is a wonderful redemptive moment, almost unbearably poignant. One by one dogs and cats are brought in to be put down, now by David as well as Bev. Coetzee writes: ‘One by one Bev touches them, speaks to them, comforts them, and puts them away, then stands back and watches while he seals up the remains in a black plastic shroud. He and Bev do not speak. He has learned by now, from her, to concentrate all his attention on the animal they are killing, giving it what he no longer has difficulty in calling by its proper name: love.’ At last, in this final scene of the novel, David has the right emotion; and, I want to add, it matters non-instrumentally that he should have the right emotion (and that he should be able to name it). Even if his treatment of the soon-to-be-dead animals were no different because of his emotion, and even if David’s feelings of love were to have no other consequences elsewhere (even for him), it still matters that he was able to give love to the dying animals. But now I must get down to the task at hand: Can we trust our emotions to give us knowledge of the empirical world? Having an emotion I will first consider what it is to have an emotion, and then go on to consider what it is to have an emotion that reveals things as they really are— that is, as I will put it, to have the right emotion. Let me begin with an example. You are trying to cross a field during your afternoon walk, and you see a bull in the field. You feel afraid of it. Your fear, being an emotion, is what is called intentional, in the sense that it is directed in thought and feelings towards an object, in this case the bull. You think that your fear is justified (although in your fear you might hardly dwell on the point): you think it is justified not only because the bull seems to you to be dangerous, but also because you think it really is dangerous (because it might harm you, with its long horns and menacing look in your direction). In your fear you are frozen to the spot, and feel adrenaline coursing through your veins. Then suddenly you run! Only when you get over the stile into the next field do you feel safe; your heart is still racing, but the fear is now nearly over. As this example brings out, an emotion (in this case your fear) is complex, episodic, dynamic, and structured. It is complex in that it involves many different elements: thoughts and feelings directed towards the object of the emotion (the bull), bodily Richmond Journal of Philosophy 1 (June 2002) Peter Goldie changes (increased adrenaline), and so forth. It is episodic and dynamic in that, over time, these elements wax and wane (your fear ceases after you have got into the next field). And it is structured in that the emotion constitutes part of a larger unfolding sequence of actions and events (you still feel nervous many hours later). Having the right emotion Now, what is it to have the right emotion? Well, staying with this example, it would not be the right emotion if it were not really a bull, but one of those Milton Keynes stone cows that we see through the window of the train, and you had mistaken this for a real live bull. Also, it would not be the right emotion if it was really a bull, but the bull was not really dangerous because it was very firmly tethered. In both these cases, it would not be right to feel fear because the object of your emotion is not really a source of danger; and thus the fear would not be justified. But even if it was really a bull and it really was dangerous, so it is right to feel fear, you could still feel fear in the wrong way: perhaps being too afraid (all it wants to do is follow you out of curiosity and it would only be dangerous if you made it panic); or perhaps being afraid for too long (even after you are safely in the next field). The idea, then, is that the right emotion is the one that is not only appropriate to its object, but is also proportionate and of the right duration. In short, the right emotion is the emotion that is reasonable or justified. In the bull example, and the example is typical of emotional experience in this respect (there are other non-typical cases that I will turn to next), the emotional response involves the experience of the emotion as being reasonable or justified. In other words you fear the bull, and at the same time think that the bull that seems to you to be dangerous really is dangerous, and that it really is dangerous because of its having other features (long horns that might harm you, a menacing look etc.), and that its having these other features justifies your fear. In the non-typical case, one does not see one’s emotion as reasonable or justified in this way. For example, you are afraid of the mouse in the corner of the room, and yet at the same time you know that the mouse is not really dangerous. So, in these circumstances, you would not try to justify your fear by appealing to the mouse’s features which you think make it dangerous; rather, you might give an explanation of your fear which does not seek at the same time to justify it: you might, for example, say ‘I have always been afraid of mice ever since I woke up and found one in my bed when I was six; but I know they are perfectly harmless.’ Nevertheless, even in these nontypical cases, the experience is still of the mouse as seeming to be dangerous. There is, thus, the possibility of acknowledging, in one’s own case, and at the same time as the emotional experience takes place, that things are not really as they seem: the mouse seems to you to be dangerous; but you know that it is not. And this is why you give an explaining reason why you are afraid (your childhood experience), without holding that this explaining reason for your fear also justifies it. So far, then, we have the following picture. Having the right emotion is having the emotion that can be justified by features of the object of the emotion. In the bull example, you are right that your fear is reasonable or justified, and that things really are as they seem: the bull seems to be dangerous and it really is. You think you are Richmond Journal of Philosophy 1 (June 2002) Peter Goldie having the right emotion and you are. Where the bull is firmly tethered but you have not seen the tether, you think the bull really is dangerous, but it is not, and so you are wrong that your fear is justified or reasonable. You think you are having the right emotion and you are not. In the mouse example, you think the mouse is not really dangerous (even though it seems to be), and it is not, and so you are right that your fear is not reasonable or justified. You think you are not having the right emotion and you are not. (To fill in the rest of the picture, try to think of an example where you think that you are not having the right emotion but you are.) It seems true to say that we want to have the right emotions: this would mean, so far as fear is concerned, being the sort of person who is afraid when and only when fear is reasonable or justified. This sort of person, Aristotle would say, has the virtue of courage, as to have this virtue just is to be disposed to be afraid when and only when it is reasonable or justified. We want our emotional dispositions, so to speak, to attune us to the world around us, enabling us to see things as they really are and to respond as we should—in short, enabling us to get it right. Emotion and Virtue In Book II of his Nicomachean Ethics, Aristotle discusses the virtues, of which courage is, of course, an example. Courage is also (and not all the virtues are like this; justice is not) an example of an emotional mean-disposition: to be courageous is to have the virtue of courage, and to have this virtue just is to be disposed to feel fear when and only when fear is justified—that is, disposed to get it right (in thought, feeling and action) so far as fear is concerned. (Thus, to be a just person is to have a virtue that is not also an emotional mean-disposition (although it is a mean-disposition), because being a just person does not involve being disposed to feel any particular sort of emotion; rather, it involves being disposed to have just thoughts, to decide on what is truly just, and to act justly.) Sometimes Aristotle is read as if he is saying (and there are things that he says that encourage this reading) that to be virtuous in some respect, courageous for example, is to be somewhere in between two extremes: moderation in one’s fear at all times, so to speak. But this is not what I think he is really getting at; it would be an absurd view if it were. So when I call it an emotional mean-disposition, I do not intend to suggest that I endorse the so-called doctrine of the mean; for I reject it. Rather, as I think Aristotle makes clear, the virtuous person will feel (that is, have emotions) and act ‘at the right times, about the right things, towards the right people, and in the right way; … this is the intermediate and best condition, and this is proper to virtue’ (Nicomachean Ethics 1106b20). And, with courage specifically in mind, he says, ‘Hence whoever stands firm against the right things, and fears the right things, for the right end, in the right way, at the right time … is the brave person; for the brave person’s actions and feelings reflect what something is worth and what reason prescribes’ (1115b17). Not moderation in all things, then, but just getting it right. In each case, on each occasion, there will be one way of getting it right, and many ways of getting it wrong; it is, as Aristotle says, like hitting a target (1106b16). Having the emotional mean-disposition, the deployment of which will enable one to get things right, is a profoundly normative notion, governed by the norms of reason, and not by what is merely typical or normal. It would be absurd to suggest that one can check to see whether or not one is getting it right by comparing one’s emotional Richmond Journal of Philosophy 1 (June 2002) Peter Goldie responses with those of other humans or with those of others within one’s community, and concluding that if they match up with what is typical or normal, then they are fine, and if not, then they are ‘wrong’. For example, it is in an important sense normal for humans to feel envy and sexual jealousy, but in both cases it is at least questionable whether envy and sexual jealousy are ever justified. Moreover, in respect of an emotion towards a particular object or type of object, an entire community (or near enough entire) can be wrong, as for example, were English people at the beginning of the First World War, who almost universally felt profound anger and disgust at all things German: Dachshunds, Wagner, and so on. We now see that as silly (or worse), and surely it is we who are right. And it was certainly ‘normal’ to experience what was called an ‘outpouring of grief’ at the death of Princess Diana, but this too has been argued to be wrong; and here again the contrary view is not wrong simply in virtue of not being normal. And, so far as fear is concerned, it might be normal not to fear the microwaves from mobile ‘phones, and those who feel no fear may think that fear is not justified, for they think that there are no good reasons to consider them dangerous. But perhaps we will find out at some future date that most of us are wrong about this: we should be afraid of mobile ‘phones. In this respect, our emotional dispositions are different from our perceptual mechanisms. We need only check that our perceptual mechanisms (sight, hearing, smell etc.) are normal, and the idea of ‘normal’ here is not a normative idea. Getting it right and getting it wrong If we are not properly attuned to the world around us, then we will be disposed to get it wrong. If, for example, you are, by disposition, a timorous person, then you will not have the emotional mean-disposition for fear, and accordingly you will respond with fear to all sorts of things (such as mice) that are not really dangerous—or at least to things that are not as dangerous as you take them to be. And if you are disposed to be unduly indifferent to fear, then again you will not have the emotional mean-disposition for fear, and here you will fail to fear things when you ought to. It is, of course, an over-simplification (one to which Aristotle was perhaps prone) to think in terms of us having (or lacking) a single emotional mean-disposition for fear of all sorts of thing. Matters are less simple than that: some people are brave about one sort of thing and timid or overly fearful about others. A particular person could, for example, be both unduly indifferent in respect of the risk of being attacked by bulls in fields, and unduly timorous in respect of the risk of being attacked by muggers in dark alleyways. There may be explanations of these dispositions that will appeal to his past experiences: perhaps he was brought up in the company of an unusually friendly bull, and was also brought up by parents who were terrified of street crime, so that their timorousness in this respect rubbed off on him. To have the emotional mean-disposition is not, however, sufficient for getting it right. Other temporary factors can also unduly interfere with one’s emotional response on an occasion, leading one to fail to get things right. I will mention two notable ones. First, one’s mood can affect one’s emotional response: for example, if one is in a jittery mood (perhaps through drinking too much coffee), then one is more likely to be frightened by a strange noise as you walk through an alleyway. Secondly, a recent Richmond Journal of Philosophy 1 (June 2002) Peter Goldie emotional experience in relation to one thing can resonate across to some other, unrelated thing: for example, if one has just had the terrifying experience of being mugged in the alleyway, then one may be especially likely to be jumpy every time the doorbell rings; your emotional disposition gets temporarily put ‘out of tune’. So let us see where we have got to so far. If one is of the right disposition, that is, if one has the emotional mean-disposition, and if there are no other undue influences on one’s thinking, then one will see things as they really are, and one will respond emotionally in the right way, in thought feeling and action. But if one is not properly disposed, or if there is some temporary undue interference on one’s emotional response, then there is a significant risk of getting things wrong; one’s emotions can distort how things are, and one will fail to respond emotionally in the right way, or as one ought. The first epistemological worry: reason-skewing This is just where my epistemological worries being to arise. As I have already said, it is typical of emotional experience to consider one’s emotion to be justified—that is, to see the object of one’s emotion as having the features that justify your feeling. So far so good. But what if, without one’s knowing it, one’s emotional response is wrong or unjustified, and the object of your emotion does not have the features that it seems to have? (Perhaps you think you have the emotional mean-disposition but you do not; or perhaps there are other temporary undue influences on your thinking that you are not aware of.) In such cases (and here is the worry), one’s emotional responses tend to skew one’s reasons to make them cohere with the emotional experience. To be clear, I am not here concerned with those non-typical occasions, like the mouse example, when one knows at the time that one’s emotional response is not justified, but the emotion remains, for on those occasions one’s reasons stand opposed to one’s emotional response, and one recognizes that it is one’s emotional response that is in error. I am, rather, concerned with those more typical cases where, in the here and now of emotional experience, one does not know that one has reason to doubt one’s own emotional response, so one sees no reason to question one’s experience of the object of the emotion as having the features that it seems to have. In such cases, I think we tend to look for and find ‘reasons’ where there are none—‘reasons’ which are supposed to justify what is really an unjustified emotional response. The emotion becomes a sort of idée fixe to which other thoughts have to conform. This is the first epistemological worry; let us call it the worry about reason-skewing. This skewing process can be continuous whilst the emotional response is in place, operating on new information as it comes in. One’s emotions and emotionally-held judgements ought to be open to be shown to be wrong by new evidence, but when new evidence does emerge, one tends not only to be insensitive to that evidence, but also, for the sake of internal coherence, to doubt the reliability of the source of that new evidence. An extreme case is Leontes in Shakespeare’s A Winter’s Tale. Once he becomes convinced that he has been cuckolded by his boyhood friend Polixenes, he refuses to listen to what previously trustworthy Camillo tells him to the contrary; he even refuses to accept the words of the oracle of Apollo. Anyone with a contrary view to his own must be either mistaken or an enemy, acting on ulterior motives. Only when Leontes’ child and wife die at the hand of Apollo does he recognize that he has ‘too much believ’d his own suspicion’. Richmond Journal of Philosophy 1 (June 2002) Peter Goldie So what’s special about the emotions here? A possible objection to my position is that there is nothing special about the emotional case: people are generally subject to all sorts of well-documented cognitive deficiencies, and the emotional case is just an instance of this. One response to this objection, which I find independently attractive but will not pursue here (although it is not unrelated to the second worry), is that perhaps more of these cognitive deficiencies can be traced back to the emotions than might at first be thought. The other response, which I will put forward here, is that there is something special about the emotional case: emotionally-held judgements, about things as having emotion-proper properties, are more intransigent than are their non-emotional counterparts, and thus the skewing (for the sake of internal coherence) tends to be towards the preservation of the emotionally-held idée fixe at the cost of the unemotional beliefs. Now, it is surely a reasonable epistemic requirement that one be willing and able to ‘stand back’ to reflect on, criticize, and if necessary change our way of thinking of things. And this requirement surely rightly extends to critical reflection on the way that one’s emotions can have this skewing effect. This is obviously the case when one knows that one’s emotional dispositions are not as they should be (as in the mouse example). But it is also the case when one has no particular reason to doubt one’s emotional dispositions: even then one should try to be especially watchful and reflect dispassionately on the evidential support for one’s emotionally-held judgements. The contrast is stark here between the epistemic requirement to check up on our perceptual mechanisms and the epistemic requirement to check up on our emotional dispositions. The contrast lies not only in the fact, which I have already discussed, that we need only be sure that our perceptual mechanisms are normal (within a certain tolerance), whereas we need to be sure that our emotional dispositions enable us to get it right, which is a normative notion. This alone makes the epistemic requirement harder to satisfy in the emotional case (comparing one’s emotional reaction with that of others may not be the right check). But it is more problematic than that. The contrast with ordinary perception lies also in the fact that we can readily observe that our perceptual mechanisms have fallen away from what they should be: car number-plates become illegible; you now cannot hear a noise which others can hear; and so on. Whereas we typically cannot observe that our emotional dispositions have fallen away from the norm: as Simon Blackburn puts it, there is no ‘loss of immediately felt phenomenal quality … when we become, say, corrupt’. The problem is a very familiar one to everyday experience: how one is to satisfy this epistemic requirement when one is in the swim of emotional experience. Consider this example. You are in despair about your job. The job seems hopeless, and it seems to be hopeless for all sorts of reasons which seem to justify your despair: there are no decent prospects for promotion; most of your colleagues are people with whom you really have very little in common; you do not seem to be able to get the work done properly; the journey to and from home is a nightmare; and so on. Your friends, not in the here and now of this emotional experience, assure you that things only seem this black because you are feeling so despairing (you used not to be like this; perhaps some Prozac might help?). You try to stand back and see things as others do (maybe things will look a bit brighter in the morning). And you might succeed in doing this to Richmond Journal of Philosophy 1 (June 2002) Peter Goldie some extent. But you could still think that it is your friends who are wrong: they believe these things because they do not see that things really are hopeless and how right you are to be in despair (Prozac might lift the despair, but the job will still be hopeless). The second epistemological worry: water lilies This leads me directly to the second, deeper epistemological worry. Emotions continue to resonate in one’s mind long after they are, as it seems, ‘over’. It is a fundamental error to think of emotions as being just about mental turbulence, of the sort that one can immediately recognize from introspection, so that one cannot be in error as to whether or not one is being emotional at any given time. Rather, we can be emotional without knowing it: for example, one may think that one has ‘got over’ some emotional experience or other, and that its potentially distorting effects are no longer at work, whilst the emotion still, at a deeper level, continues to resonate in the psyche. One can therefore be inclined to think that one is being ‘dispassionate’ when one is not, and thus one has no way of knowing that special watchfulness is required. On such an occasion, then, one might ask oneself ‘Am I emotionally involved here? Because if I am, I should be especially watchful.’ Yet the answer comes back ‘No, I am not emotionally involved here’; moreover, one might sense a certain puzzlement as to what sort of emotion might be relevant here. I call this the water lilies worry after this marvellous passage from Robert Musil’s The Man Without Qualities: ‘We . imagine that the world is unambiguous, whatever the relationship between the things out there and the inner processes may be; and what we call an emotion is a personal matter that is added to our own pleasure or uneasiness but does not otherwise change anything in the world. Not just the way we see red when we get angry – that too, moreover; it is only erroneously that one considers it something that is an occasional exception, without suspecting what deep and general law one has touched upon! - but rather like this: things swim in emotions the way water lilies consist not only of leaves and flowers and white and green but also of “gently lying there”’. If Musil is right, then, that we are always in the swim of emotion (‘no emotion … ever comes to an entirely specifiable end’), and that we will often not know what emotions are at work in our minds at any given time, then we could be in this worrying position. Our reasons continue to be skewed as they are in emotional experience (the reason-skewing worry), and yet we have no way of knowing in what way they are being skewed, because we do not know what emotion is at work (the water lilies worry). Let me give an example. A long time ago you were very angry with a colleague at work because he failed to turn up to a meeting that you were chairing where his presence was essential. And he promised to be there. You thought your anger to be thoroughly justified—how could he not have told you in advance! The following day, though, he gave you a full explanation, and was extremely apologetic. You put your anger behind you, as you should do, realising that he really had a good reason not to be there, and a good reason why he could not give you advance warning. Later still—much later—you are asked to give your colleague a reference. Without your realising it, what you say is influenced by your emotional experience, which still has its residue deep in the recesses of your mind: you do not say that he is unreliable (for the long-past incident is no longer in the forefront of your mind), but your reference is not as Richmond Journal of Philosophy 1 (June 2002) Peter Goldie favourable as it would have been if the incident had never taken place. You are, in a subtle way, and without knowing that you are doing it, getting your revenge. So where does thie leave us? The two epistemological worries, then, are as follows. First, whilst one is in the swim of life, emotionally engaged with what is going on, one’s reasons are liable to be skewed by one’s emotions, which become sort of idées fixes. This is the reason-skewing worry. To avoid this as much as possible, one should see oneself as subject to the epistemic requirement to reflect on one’s reasons, and to correct them where necessary, and to be aware that one should be especially watchful when one is emotionally engaged. The second water lilies worry is that one can be emotionally engaged without knowing it, so one has no way of knowing which of one’s reasons are being skewed, and in what ways. Even if one were to embrace Musil’s thought (or my interpretation of it) that we are always in the swim of emotion, and thus to accept that a special watchfulness is always required, one will still be no wiser as to how to apply this epistemic requirement at any particular moment. This seems to me to be especially troubling. But then perhaps I am just being unduly emotional. Peter Goldie King’s College London 1 Intentionality is a notion that I cannot hope adequately to explain here. It is sometimes explained as the property of aboutness that is one aspect of the mental; for example, if you think about your kitchen at home, it is your kitchen at home that your thought is about. There is a very good, but not easy, paper by Tim Crane, ‘Intentionality as the mark of the mental’, in A. O’Hear ed., Current Issues in the Philosophy of Mind (Oxford: Oxford University Press, 1998), pp229–51, where Crane argues that intentionality is better understood as directedness towards an object. This is the view I adopt here. 2 I argue for this in The Emotions: A Philosophical Exploration (Oxford: Clarendon Press, 2000). 3 Jonathan Dancy discusses the distinction between explaining and justifying reasons in his Practical Reality (Oxford: Oxford University Press, 2002). 4 Aristotle discusses these ideas in Chapters 6 and 7 of Book II of his Nicomachean Ethics. There is an excellent discussion of the doctrine of the mean (and a rejection of it as false), in a recent and very readable book by Rosalind Hursthouse, On Virtue Ethics, Oxford: Oxford University Press, 1999; I strongly recommend this book to anyone beginning to find his or her way into Aristotle’s ethics. Also, for anyone thinking of buying a copy of the Nicomachean Ethics, the edition that I recommend is translated by Terence Irwin, Indianapolis: Hackett. It has especially useful notes and a glossary. 5 See also, with courage in mind, 1115b35 and 1116a4. 6 See Anthony O’Hear’s ‘Diana, queen of hearts: sentimentality personified and canonised’ in D. Anderson and P. Mullen (eds.), Faking It: The Sentimentalisation of Modern Society (Social Affairs Unit, 1998), pp181–190. 7 For a fascinating survey of these deficiencies, see R. E. Nisbett and L. Ross, Human Inference: Strategies and Shortcomings of Social Judgement (Englewood Cliffs, NJ: Prentice-Hall, 1980). 8 See his ‘Errors and the Phenomenology of Value’, in his Essays in Quasi-Realism (New York: Oxford University Press, 1993), pp149–65, at page 160. 9 The Man Without Qualities, tr. S. Wilkins and B. Pike (New York: Alfred A. Knopf, 1995), page 1561. 10 Ibid. page 1307. 11 This example is very loosely based on some empirical research, which seems to lend some support to my case here. See Dolf Zillman and Joanne Cantor, ‘Effect of timing of information about mitigating circumstances on emotional responses to provocation and retaliatory behaviour’, Journal of Experimental Social Psychology 12, 1976, pp. 38–55.

Source: http://www.richmond-philosophy.net/rjp/back_issues/rjp1_goldie.pdf