Now, about this response-dependency thing

I am a fan of keeping options left open, and of not leaving open options left undeveloped.  When we find ourselves with conflicting intuitions in situations where intuitiois our only ground for theoretical decisions, it is basically an act of charity to develop a theoretical option anyway, in case someone will find it in their heart to – as we so endearingly say in philosophy – entertain the proposition. (My attitude here, you might have noticed, is a bit counter to my lament about a certain trope in the post below.)

Store that in some cognitive pocket (memory, David, it’s called ‘memory’ ) for the duration of this post.

Response-dependency. Some concepts, and some properties, are response-dependent. That means that the analysis of the concept, and the nature of the property, is at least partly made up by some response. To be scary, for instance, is to have a tendency to cause a fear-response. There is nothing else that scary things have in common. Things are different with the concept of danger: Dangerous things are usually scary too, but that is not their essence. Their essence consist in the threat they pose to something we care about,  or should care about. Fear is usually a reasonable response to danger; fear is usually how we detect it. Danger might still be response-dependent, but then fear is not the crucial response.

Response-dependency accounts have been developed for many things. Quite sensibly for notions such as being disgusting. Famously by Hume for aesthetic value. And arguably first and foremost under the name ”secondary qualities” by Locke, and unceasingly since by other philosophers, for colours.  Morality, too, has been judged response-dependent , and a great many things have been written about whether this amounts to relativism or not, and whether that would be a point against it.

Response-dependency accounts of value and of moral properties has a lot going for it. Famously, beliefs about moral properties are supposed to involve some essential engagement of our motivational capacities. And if the relevant property is one our knowledge of which is dependent on some motivational response, say an emotion, this we’re all set. Further, if these concepts/properties are response dependent, it would account for many instances of moral disagreements – we disagree on moral issues when our moral responses differ, and when the difference is not accountable by a difference in other factual beliefs and perceptions. If we accept that responses are all there is to moral issues, we might have to learn to live with the existence of some fundamental disagreements between conflicting responses, and moral views. Relativism follows if there is nothing that moral respones (for the most part, the moral response involved is some kind of emotion) track. What the account gives us is a common source of evaluative meaning in located in the fact that we all share the same basic type of responses. We just disagree in what causes those responses, and about what objects merit the response.

If we insist on locating the value (moral or otherwise) in the object/cause of the response (note that the object and the cause might be different things – we might project an emotion of something that did not cause it. This happens all the time), the response-dependency account results in a form of moral relativism. If one finds relativism objectionable, and there is no way to provide firm moral properties in the cause/object structure of typical moral responses, one might therefore  want to reject response-dependency wholesale. I think this is mistake. If we agree that there is such a thing as a moral/evaluative response, and this response is something that all conceptually competent evaluators have in common, we have our common ground right there, in the response. It is not the kind of relativism where we find that seemingly disagreeing parties are actually speaking about different things altogether. In fact, there is a common core evaluative meaning, and that meaning is provided by the relevant response.

So, to the suggestion then, our theoretical option left open for development: that moral/whatever value is in the response, not in the object of the response. This seems to be the obvious solution once we’ve established that the value is metaphysically dependent on the response, and there is no commonality to what causes the response. If the responses themselves can is something that seemvaluable, and emotions usually do, we should develop that option, and disregard the fact that we tend to project value to the object of mental states. (If we keep on, as I do, and argue that the evaluative component of any emotion consists in it valence, and valence is cashed out in terms of pleasantness – unpleasantness, we have a kind of hedonism at our hands, but this option is open for any response you like).

Nothing is metaphysically more response-dependent than the responses themselves, and yet, this move avoids any objectionable form of relativism, while explaining the appearance of relativism. And, given that the response is motivationally potent, we have an inside track to the motivational power of moral/evaluative properties/beliefs. This, I’d say, makes it a theoretical option worth pursuing.

Going Through the Emotions

Here are some words that, stringed together, are bound to make you gasp and salivate with anticipation (thus supporting the James-Lange/Tex Avery theory of emotion): The Non-Conceptual Representational Content of Emotions.

Emotions are tricky things. But what kind of tricky things are they? ”Cognitivists” believe that emotions are at least partly constituted by judgments, wereas non-cognitivists typically claim that emotions are mere experiences, uncommitted to any propositional content.

One problem for cognitivists is that judgments seem to require concepts, and we might want to say that creatures incapable of housing the relevant concepts might nevertheless experience emotions. This depends on what you require from your concept possessors: we can lower the bar, so that most animals have at least some concepts needed for certain emotions. Or we might want to exclude non-humans from the world of full-fledged emotions. Cog and Non-Cog may not disagree over the facts, but only about the terms, and the implied policies.

The problem for non-cogs is that we sometimes treat emotions as justifiable, and it is hard to see what that amounts to if emotions are not judgments. The case is analogous to perception.

Emotions can’t just be judgments since for every judgment , something could be that judgment and yet not be an emotion. I.e. even though emotions might have conceptual content, they are not exhausted by that content. We need more in order to differentiate emotions from states having the same content as an emotion.

Now, representation is a great notion to invoke here, since it’s eminently flexible: to represent somehing is a relational, i.e. non-intrinsic, feature. In order to represent something, even something propositional, you need not yourself have a propositional nature. So an emotion might be something like a pure experience, and still have non-conceptual, representational content – the ”meaning”, as it were, of the emotion, is not intrinsic to the emotion, but derived from its function. When what is represented by the emotion is a judgment, or something to that effect, we can treat it as justifiable. Analogously, again, perceptions represent things as being a certain way, but in order to represent a judgment, it need not itself be a judgment. The fear of a snake is a mental state that represents the snake as being dangerous, but it is not itself the judgment that the snake is dangerous.

Is representation essential to what you feel?  Representations according to Dretske-influenced theories (Tye, Schroeder, probably Prinz), are causal, functional affairs, and it might not be obvious to you what caused your current mental state, or what function that mental state plays. In fact, seeing how one and the same mental state might represent/be caused by different things, it might not even be introspectable what you are feeling. The feeling is obvious to you, but what the feeling is saying need not be.

Are we left with enough to identify and individuate emotions? It seems not. It is not surprising that emotions have the function to represent, since practically all mental states have that function. But that is not all it does. Indeed, there are plenty of reasons to believe that emotions are not exhausted by their function, either, let alone the strictly informational one-

I believe that ”emotion” is a usefully vague notion, but that all emotions have a hedonic component. They are either positively or negatively valenced. Also, any mental state having a hedonic constituent deserves consideration as an emotion. When they are prompted by judgments, or followed by judgments, or just associated with judgments, it makes sense to assess them accordingly. In addition, you may be held responsible for an emotion, even if that emotion is not something you can directly choose to have or not too have: you might be responsible for the judgment that prompts the emotion.

-

This post was inspired by reading this book, which I’m sure will prompt (and thus then be represented by) more posts over the next few weeks):

Jesse Prinz: the Emotional Construction of Morals