More on the Mean

To recap from yesterday...

The Doctrine of the Mean
Actions (and emotions) have 'parameters' (what objects, when, how, how much, for the sake of what), and, when an action (or emotion) goes wrong, it does so because it is too much or too little with respect to at least one parameter.

Hursthouse
Sometimes an action (or emotion) goes wrong because of an error in quality, not quantity. Consider the emotion of fear, and the parameter of 'what objects'. A virtuous person fears (let us suppose) only death, trauma, and severe pain. Consider now a 'fearless phobic' who fears only mice, the dark, and open spaces. His error is not one of quantity, because he too fears precisely three sorts of things. His error, then, is one of quality.

Curzer
We should probably say that a 'fearless phobic' is mentally ill, not vicious. But putting this point aside, we can say that such a person does indeed go wrong in quantity: he fears mice (and the others) too much, and he fears death (and the others) too little.

Pearson
It is odd to say that a person fears mice 'too much', when what we mean is that he should not fear them at all. The reason is that a statement of the form, "He F's X too much", carries with it the implicature that some degree of F toward X is appropriate. The oddness of "He fears mice too much" indicates that the difference between fearing mice and not fear them is not a difference in degree.

But now to carry on:

(1) I wonder if Pearson's objection doesn't depend on an artefact of language. 'Too much' means 'more than one should'. Thus, "He fears mice too much" means "He fears mice more than he should (fear mice)." It smuggles in the object, when we wanted to say that the right amount of fear toward the object is none. But that phrasing is inessential. We could easily say, rather: "He shows too much fear in fearing mice", which carries with it no unwanted implicature.

(2) Surely it is right to say that someone "shows too much fear in fearing mice", and also surely the fearless phobic example is contrived and not a serious example in ethics. This needs to be accounted for. I propose the the following.

As regards the parameter of "what objects"-- we do not evaluate actions (or emotions) in isolation, but we interpret each as an expression of a 'distribution'. For instance, as regards fear, we think that a person's fear of something should be proportionate to the actual threat or danger it poses (thus Aristotle, 1115b7-14); furthermore, we presume that the usual principles of rational preference should apply to this (e.g. transitivity--if someone fears a snarling dog more than a mild dog, and a wolf more than a snarling dog, he is rationally bound to fear a wolf more than a mild dog.) So we view his reactions of fear as reasonably and typically distributed in a series: more fear for objectively more threatening objects, less fear for objectively less threatening objects. And that is why the 'fearless phobic' is an unreal example: someone who feared a mouse should reasonably fear even more a heavily armed enemy coming at you in battle.

Now it seems natural to suppose that there will be a relationship between (i) degree of fear as shown towards an object that should be feared, and (ii) extent of distribution of one's fear. E.g. someone who fears a snarling dog too much might fear a mild dog when he shouldn't; someone who fears a wolf too little might fail to fear a snarling dog when he should. Strong emotion makes the distribution broad; weak emotion makes it narrow. We correctly interpret fear toward an inappropriate object, then, in terms of a distribution we take that fear to display. (This is one way in which we view actions as expressions of character.) That is why it is right to say such things as "He shows too much fear in fearing mice", that is, the excess fear he shows toward appropriate objects of fear spills over, even, to things that are not fearful at all.

This notion of a 'extent of distribution' of fear provides a reply, too, to the objection that Ben Miller raised in a comment: If not fearing a mouse is the mean--he asked--and fearing a mouse is excessive, then what counts as the contrary extreme? What counts as deficiency of fear in this case--a negative fear of mice??

In reply to this one might deny that, for Aristotle, all particular actions (and emotions) have erroneous extremes on either side of them (why couldn't Milo's diet ever be to eat all 10 minae?), even if habits leading to extremes (i.e. in some parameter or other) are matched by habits leading to deficiencies.

But perhaps a better reply is available, using the notion of a distribution. Someone who fears too little as regards the 'object' parameter will typically distribute his fear too narrowly. Thus objects that should be on the border are more remote from being feared than they should be. For instance, if it's appropriate to have slight fear for a snarling dog, but not a mild dog, then, as a dog changes from mild to snarling, one should appropriately change from not fearing to fearing it. But if a person generally fears things too little, he won't begin to fear a mild dog that begins to snarl--and from this we might say that, even though he did not fear the mild dog at all, still he feared it less than he should, in the precise sense that he did not place it close enough in the scale to objects that should appropriately be feared.

(Granted, some objects shouldn't be conceived of in relation to the scale at all, e.g. my easy chair, or an earthquake. But that's just to say that these are not 'human' objects of fear, and inappropriate fear as regards such things constitutes mental illness, not moral badness. Again, see 1115b7-9.)