In the final chapter of The Given, I describe the rich complexity involved in experiencing emotions. I restrict my attention to occurrent emotional episodes that are not only conscious, but are also intentional. I consider two questions: [i] What kind of property attributions do we typically make in having emotions? [ii] What kind of phenomenology do conscious emotions have?
I start with the idea that in having various emotions, the world appears disgusting, maddening, scary, hateful, wonderful, sad, contemptible, amazing and so on. This description reveals two important intentional facts about emotions, thus providing a partial answer to [i]:
(1) In having emotions, the properties of being sad, joyous, disgusting and so on are properties we attribute to objects and states of affairs and thus are experienced as objective properties of objects and states of affairs
and
(2) The property attributions of sad, joyous, disgusting and so on are essentially value property attributions, and so I call them ‘emotion-value properties’.
One may feel some discomfort with (1). It’s true that we say people feel sad, happy, joyous, and so on, but do we really attribute such properties to states of affairs in the world? It seems completely natural for a subject to attribute the property of sadness, for example, to the state of affairs of a friend’s death: ‘my friend’s death is sad’. This kind of attribution of an emotion-value property to an object or state of affairs is common in our language: we feel the behaviour of irresponsible bankers is contemptible; we find our boss annoying; we feel the wedding of a best friend joyous, and so on. Moreover, these emotion-value properties seem to be properties these states of affairs possess quite independently of us. When I claim that my best friend’s death is sad, I take myself to be claiming something about the state of affairs itself—that it is itself objectively sad.
One may continue to object: isn’t my friend’s death something that causes me to feel sad rather than having the property of sadness itself? If emotion-value properties were only a matter of what state of affairs can cause us to feel, emotions would be merely subjective in a way that they typically (or at least ideally) are not. Emotions seem to connect us to people and the world in a way that takes us entirely beyond ourselves or outside of ourselves. When I feel sad about my friend’s death, my sadness essentially involves the value or disvalue of something beyond me and completely independent of me— my friend’s death. My emotional feelings extend beyond me to connect to and correctly register this disvalue, which is (or appears to be) independent of me. If emotions were only a matter of how things affected us or caused us to feel, it would be difficult to characterize the sense in which they allow us to reach beyond ourselves.
My answer to [ii] involves the claim that
(3) In addition to whatever cognitive and sensory phenomenology they may have, emotions also and always have a distinctive sui generis kind of phenomenology, which I call ‘evaluative phenomenology’.
To begin, evaluative phenomenology cannot be reduced to some variety of cognitive phenomenology that is (someone might argue) distinctive of conscious evaluative judgment. Although a conscious evaluative judgment may be said to be an experiential representation of value, because all conscious episodes are experiential or phenomenological in nature, conscious emotion, with its evaluative phenomenology, is also and essentially a distinctive kind of experience of value. This difference should be evident when we reflect on the phenomenological difference between judging (without feeling) that someone’s death is sad and actually feeling the sadness of a friend’s death (putting aside the claim that bodily sensations can account for this felt difference). I argue that this phenomenological difference resides in the fact that the way in which evaluative phenomenology represents value properties is very different from the way in which a conscious thought or judgment represents value properties.
What about the standard kinds of sensory phenomenology associated with emotion? Let’s suppose it’s true that emotions are generally and even necessarily accompanied by bodily sensations. My heart may race when I’m angry, my throat may constrict when I’m afraid, and my emotional experience of sadness may include certain ‘fatigue’-like bodily sensations. Even so, I propose, the ‘psychic deflation’ distinctive of the phenomenology of sadness, for example, is not reducible to such bodily fatigue or any other combination of bodily sensations. The first reason is purely an appeal to how things seem to me while I’m experiencing a vivid emotion. When I feel anguish about the displacement of my sister with Down syndrome from our family, and her struggle to find her own way in the world, my phenomenology is not reducible to a set of bodily sensations, although it may be true that this emotion is necessarily accompanied by, and underlain by, bodily sensations. The second reason is that evaluative phenomenology is (essentially) representational in a way that bodily sensations like pain, fatigue, and the feeling of a rapid heartbeat are not, and the kind of representation distinctive of evaluative phenomenology is necessary to account for the nature of conscious emotion.
I conclude with two final claims:
(4) Emotions afford a distinctive kind of evaluative awareness of the world;
and
(5) It is partly in virtue of an emotion’s evaluative phenomenology that value properties are attributed to objects and states of affairs in the having of an emotion.
The evaluative phenomenology of experiencing sadness about a friend’s death, the negative feeling one experiences, in the specific mode of sadness, accounts for one’s attribution of the property of sadness (and so disvalue) to the friend’s death. So although one may be able to know intellectually (in some sense) that the death of a friend is of disvalue, and indeed that such a death is sad, without feeling an emotion, one can experience the disvalue of the friend’s death in this distinctive way only if one has this emotional experience. Commander Data in Star Trek, who doesn’t have any emotions, may be capable of saying what is of value and disvalue, and, in turn, of saying what is sad and what is happy, but he cannot experience sadness or happiness and thus he cannot experience value or disvalue (in the special way I am indicating). A stronger claim about Data is that although he can say what is of value and disvalue, and in that sense know what is of value and disvalue, he cannot really know value and disvalue at all: he cannot really know what value and disvalue are. So too, someone who is blind from birth can say that something is red, when told that it is a ripe tomato, and indeed know that it is red, but not know what red is.
A thought provoking post.
‘When I claim that my best friend’s death is sad, I take myself to be claiming something about the state of affairs itself—that it is itself objectively sad.’
I wonder; Say, my dear friend x dies, and her death represents a sad state of affairs for me. But for y, who hated x, her death is a cause of joy and celebration. I recognise this to be the case for y, and so I accept that x’s death is not an objectively sad state of affairs per se, but only for me and her other friends. So you might conclude that x’s death only has a particular emotional value for the subject.
I also wonder if all emotional values are necessarily intentional; as in the case of moods and dispositions, where you may feel in a sad (or happy) mood with no identifiable intentional object- as might be the case for someone with clinical depression, or high on recreational drugs. The only intentional content of the mood is oneself- ‘I feel terrible!’ or ‘Wow, I feel great!’ So once again, the emotional (dis)value is limited to the subject.
‘…one may be able to know intellectually (in some sense) that the death of a friend is of disvalue, and indeed that such a death is sad, without feeling an emotion..’
Again, I wonder if this is really the case. As perhaps shown here by your use of parenthesis, it’s hard to imagine having an awareness of a sad event that is purely neutral and objective; that is, without feeling the associated subjective disvalue at least to some extent. Whether Data, as portrayed in the series, really does feel no emotional values is I think is a moot point. But assuming that to be the case, yet he can still intellectually predict the correct emotional value of any event, then we assume he does this through simple data processing, without any of the accompany evaluative phenomenology. If we accept that any conscious experience is homogeneous, it’s hard to see how this could be the case with us. You could go further and say that consciousness is necessary for evaluative phenomenology, but not not so for objective evaluation of events in the world, which might be achieved with or without any kind of phenomenal experience. Indeed, it might be conceivably be achieved more efficiently by a non-conscious machine.
I can’t speak for Michelle — but I’d reply to the first objection by saying that this is parallel to a case where I judge that my perception may be misleading. I can experience something as sad without thinking, at the level of reflective judgment, that it really is, or is that way objectively or for everyone. Nevertheless it can still appear (objectively) sad from the point of view of my emotional experience.
I like John’s comparison to the case of perception. I’ll just add that disagreement in general doesn’t immediately lead one to a relativistic conclusion. You could be right and your friend could be wrong.
It’s a difficult question whether all emotions are intentional. One move is to distinguish moods from emotions and allow that moods are not intentional. However, I’m inclined to think that moods are intentional. For a depressed person or a person high on drugs, for example, everything is presented as bad or good, respectively. One might describe one’s general mood without appealing to an intentional object, “I feel good”, but that doesn’t mean that the mood occurs without being intentionally directed.
There’s a lot to say about your last point, and we may have deep disagreements about the nature of consciousness and its relation to intentionality. I don’t think a non-conscious machine has any thought content at all. I use the case of Data to focus on the particular case of emotions, but if Data lacked all phenomenology, it would only look as if he were a thinking machine.
Thanks for your interesting reply, Michelle. I’m by no means an expert in this field (as you probably guessed) but still find myself drawn to play devil’s advocate here.
‘You could be right and your friend could be wrong’.
In the case of a purely epistemic judgement about the world, this statement would be correct, as we could appeal to the authority of a suitably qualified third-person judge. But in the case of evaluative phenomenology, where is the source of authority behind this assertion? If I am sad about the death of x, this has necessary first-person authority, in as much as I cannot be mistaken about my own evaluative feelings v, at any given moment of time, t. My feelings may be different, v1, at another time, t1, but again I cannot be said to be mistaken about my own evaluative phenomenology v1, at t1. It is open to another subject to experience an opposing evaluative phenomenology, say -v, about the same event at the same time t, but this is purely relativistic as there can be no third-person authority qualified to judge which of us is ‘right’. For me, this implies that our evaluative phenomenology is not straightforwardly intentional in the same way as our perceptual or empirical judgements about the world may be. So John’s use of a perceptual analogy is perhaps misleading.
‘..evaluative phenomenology is (essentially) representational in a way that bodily sensations like pain, fatigue, and the feeling of a rapid heartbeat are not, and the kind of representation distinctive of evaluative phenomenology is necessary to account for the nature of conscious emotion.’
I quite agree, and I think this goes along with what I’m trying to say, here. If evaluative phenomenology, defined as the subjective good/bad feeling at the heart of any emotion, is essentially representational, couldn’t we go further and say that for this reason it is not – necessarily – intentional? That it is somehow a brute and essential- that is, an irreducible – quality of a conscious emotion?
I realise that this raises all the usual metaphysical and ontological problems about the phenomenal. A bit of a leap from your general argument perhaps, but for me, the whole question of the primary role of the subjective within evaluative phenomenology shows just how hard the hard problem remains . So now I’ll go out and buy your book and read it.
What are your thoughts on A.I. and emotionality?
What are emotions for an A.I. (such as ‘Ava’ from the film ‘Ex Machina’)?