Why do people act morally? Why do people act altruistically, keep their promises, or act fairly? It is sometimes held that a part of the answer is that people have "moral emotions", and these emotions play a key role in the creation of moral actions.
What is a moral emotion? I'm sure that there are specialists who would offer different definitions of this concept; but I suggest that a moral emotion is a feeling or affect that is responsive to the situation of other living beings. Sympathy, compassion, humor, affection, and respect are all examples of moral emotions; but so are antipathy, rivalry, envy, and racial animosity. This inventory shows that what I'm calling "moral emotions" are not necessarily "moral" -- taking pleasure in the suffering of others is morally unattractive, but falls in the category of a feeling that is responsive to the situation of the other.
There is a related category of emotion that philosophers sometimes refer to as "cognitive emotions." These are feelings that are dependent on possessing certain kinds of beliefs. Feeling grateful is a cognitive emotion; it doesn't make sense to attribute this mental state to someone without also attributing to the person some set of factual beliefs about what has occurred in light of which being grateful makes sense. (Andrew Ortony, Gerald Clore, and Allan Collins provide some theoretical discussion of this topic in The Cognitive Structure of Emotions.)
These two categories do not fully overlap. There are moral emotions that have a cognitive basis. But there are also moral emotions that do not have a cognitive foundation -- for example, the emotional response most people have to a smiling infant. And there are cognitive emotions that do not have a social component -- for example, fear of illness.
It is clear that normal human beings experience these kinds of emotions and feelings. How should we factor them into our theory of action? How do emotions affect behavior? Some emotions seem to have an immediate causal power to create dispositions to specific kinds of action (dispositions that can nonetheless be overridden by higher functions of self-control). An angry person is disposed to lashing out at others. A person experiencing sympathy is disposed to providing aid to people in immediate need. A frightened person is disposed to retreat from the frightening situation. A person experiencing sadness may be inhibited from any kind of action. So emotions have a fairly direct relationship to action.
It is a short step from recognizing the fact of these kinds of emotions, to asking whether there is an evolutionary basis for them. Were social emotions like sympathy psychological capacities that conferred reproductive advantage on early primates? Did these emotions create the possibility of forms of cooperation that permitted primates and early humans to achieve greater success in their environments than would otherwise have been possible? This is a topic that Allan Gibbard explores in Wise Choices, Apt Feelings: A Theory of Normative Judgment.
Some philosophers of action (e.g. David Hume) have wanted to understand the emotions as the primary or even sole motivating forces that drive actions, and reason serves only to permit the actor to tailor action to circumstance. Others (e.g. Aristotle) have looked at reason as the master of the emotions: the rational being decides what to do no matter what emotions he or she is experiencing. The "passions" are not trustworthy guides to action, on this philosophy. Kant represented the extreme version of the reason-centric theory of action: only actions motivated by the recognition of duty are morally worthy.
Some philosophers have held that the moral emotions are a necessary ingredient of morally generated behavior. These emotions take the individual out of his/her particular interests and provide a basis for other-regarding action. The moral emotions are thought to provide the motive power underlying altruism, benevolence, and sympathy. Lacking these emotions, it is thought that the actor would be unmoved by the needs and cares of others. The most noteworthy exception to this line of thought is that provided by Tom Nagel in The Possibility of Altruism, where he argues that the cognitive act of recognizing the reality of other persons is sufficient to generate altruistic behavior.
These ideas highlight once again the point made in earlier posts: that a theory of action needs to be complex and needs to take into account the several ways in which consciousness drives behavior. It seems apparent that we do not yet have a theory of action that does justice to the nuance of thought and behavior. Habit, emotion, character, rules, and deliberation all play roles in the creation of actions, but we do not yet have good models for how they work together.
Here is a diagram offered by Ortony, Clore and Collins to help to classify emotions (19).
(Quite a few earlier posts are relevant to this topic. Searching for relevant keywords including altruism, sentiment, reciprocity, and cooperation will lead to some of these discussions.)