I've dealt with some issues surrounding aesthetics here in the past, and my approach has at least partially been biological and socio-cultural: asking what external reasons there are for the granting of aesthetic merit to certain artworks as opposed to others.
However another approach is also valuable: asking what internal reasons there might be for artistic approval and disapproval. One central, internalist topic has to do with emotional response. Speaking very roughly, an artwork that elicits a strong emotional reaction, or perhaps a strong emotional reaction with certain characteristics, is seen as superior to one that, as the saying goes, "leaves us cold". Now, of course one worthwhile way to go here is to begin to ask, "What sorts of emotional responses are the right ones?" And perhaps we can come up with some initial, rough criteria: in order to be a particularly good work of art, the emotion shouldn't include strong antipathy towards the work. Now, a sly, budding art critic may disagree with such a claim: perhaps reactions of distaste, disgust, dislike, and so on are examples that "épater la bourgeoisie", and as such, perhaps we should be suspicious of such strong antipathic emotional responses. But nevertheless I think most of us would agree that there is some scope of strong dislike that includes artworks that really are worthless by any reasonable aesthetic merit. That said, it doesn't really matter for what follows whether we make such a move or not.
Second, we may say that in order to be the right sort of emotional response, the emotion must be somehow complex. After all, a simple horror, humor, or romance film, a propaganda piece, or even pornography can elicit very strong emotions without having any real aesthetic merit. One typical response of the self-styled sophisticate is to claim to be unmoved in the presence of such works, of course. Although I am highly dubious of such poses, nevertheless it also doesn't really matter for what follows whether we accept them or not.
The question I prefer to get at is more basic: why should it matter that a piece elicit strong emotions? What's so important about emotions? Why should a piece that makes us cry, laugh, or feel terribly emotionally confused be somehow superior to one that does not, or that, say, makes us think of the color blue or feel particularly hungry?
I don't claim to have anything like a full answer to this question. I think it's enough to raise it as an object of genuine concern. However I do have something at least approaching one plausible explanation why emotion might be important in art criticism.
Emotion and Recall
When we think back on the artworks we have seen in our lives, which ones tend to stick out? A simple answer is that it's the ones that caused the deepest emotional impact in us. I'd argue this goes much more for art critics, whose job involves interacting with large numbers of similar pieces on a regular basis. In the fullness of time, they all tend to blend together; the ones that stick out are the ones that caused the deepest emotional impact. (And that may have had other, associated characteristics, like originality, and so on).
It's often said that time is the greatest art critic of all: often the pieces that were beloved by this generation of critics prove forgettable to the next, when seen in a new light. Once again, the operant cause here may at least partially have to do with emotional impact over time. That is particularly so if the emotion of the work became confused with disgust in the first critics who saw it.
At any rate, over time memory comes to the fore in internal determinations of artistic merit: the best pieces are said to be "memorable". Less worthy pieces are, in a word, "forgettable".
Now I think we have the framework to bring the two together: emotion and memory. To put it simply, the artworks that are "forgettable" just tend to be those that "leave us cold". Scientific studies have shown conclusively that we have better and longer recall of emotional events than we do of ones that are not emotion-laden. This, of course, makes evolutionary sense: our emotions are tailored to arise during times that tend to be evolutionarily critical. We are frightened or hate filled when we feel threatened. We are aroused when we sense a potential mate. We feel love surrounding mating, child rearing, and in-group bonding. Humor is a more complex case, however it is also one not typically associated with great artwork. Nevertheless times of emotional arousal are also times it would be good to underline in our long-term memory traces: they likely involve processes it would be useful to come back to at a later date. They may, for the same reason, be seen and encoded as particularly valid or truthful, whether in fact they are or not.
If this is so, then the evolutionarily useful link between emotion and memory plays a crucial role in our aesthetic evaluation. It may not be that there is any particular reason why emotional artworks are aesthetically better than ones that are not, apart from the mere fact that because they elicit strong emotions they are easier to recall at a later date.
I believe similar processes may occur as well with narrative events: they are more easily encoded and hence recalled at a later date. Emotional narratives, therefore, may be particularly well suited to be taken as artistically valuable: thus our Homers, Sophocleses, Shakespeares, Dantes, Cervanteses, Lady Murasakis, and so on.
Although the creation of aesthetic merit out of evolutionarily attuned memory and other allied causes may be a relatively benign phenomenon, there is still room for concern. Insofar as the mind tends to recall such emotion-laden narratives more easily, and imbue them with a validity they may lack, we may be prone to learning and recalling false lessons. It's all too easy for any of us, from the most humble up to the most powerful, to put ourselves into easily recalled, emotion-laden narratives from history, believing they provide lessons which they may not.
If in fact emotion is an enemy of clear seeing, then relying on emotion to provide us with our most salient data is a recipe for promoting only confusion, falsehood, and ignorance. No doubt this overstates the case to some degree, but once again, perhaps it's enough to raise it as an object of genuine concern.