When virtue ethicists invoke “happiness”, they are inevitably going to bump into the objection that “happiness” is incredibly vague and that it is subjective. And, given the common understanding of what “happiness” means, this objection is perfectly understandable. This forces the virtue ethicist to clarify that that is not what they mean by the term “happiness” in this context. What they tend to actually mean is something more along the lines of “well-being” and “flourishing”, irrespective of mere feelings. In this context, “happiness” does not refer to a feeling of pleasure.
This dual meaning of “happiness” is at the heart of some of the confusion that arises in ethical discourse with virtue ethicists. For the opponent of the virtue ethicist can always argue something like this: “A murderer can concievably obtain happiness by murdering, especially in the case of the pathological. Thus, if you are using happiness as the criteria for ethics, I don’t see why plenty of actions that would normally be considered immoral could be justified based on such a criteria. There are plenty of people in the world who obtain happiness via immoral means”. At this point, the virtue ethicist has to either clarify a different meaning of “happiness” or invoke something else.
Alternatively, the virtue ethicist can argue from an absurd psychological premise: that there is an absolute correlation between morality and happiness, to the point of claiming that a murderer or thief is actually inherently unhappy. This premise is demonstrably false based on common experience and known facts: there are people who are “moral” who are unhappy and there are people who are “immoral” who seem to be quite content. Especially in the case of the pathological, it is possible to engage in “immoral” behavior without experiencing any negative psychological fallout. It simply would be absurd to propose that “nature” inherently punishes all “immoral” behavior with unhappiness and unsuccess (otherwise it would be impossible for “immoral” people to survive and succeed in life – and they clearly do).
Fortunately, I haven’t ever seen any virtue ethicist argue from such a premise (however, I have seen various Objectivists do so). To avoid this trap, it would probably be wise for the virtue ethicist to say that morality is necessary but not entirely sufficient for happiness, I.E. it does not necessarily absolutely gaurantee happiness. Likewise, one could propose that liberty is necessary but not sufficient for survival and flourishing; liberty does not necessarily gaurantee that one will survive or even have a good quality of life (one could concievably be free yet miserable). Liberty doesn’t mean “freedom from the constraints of reality itself”. So while one may propose that there is some correlation between happiness and moral behavior (and between survival and moral behavior, to use Rand’s criteria instead of that of the virtue ethicist), such a correlation is not absolute or purely causal.
A similar confusion plagues the concept of “selfishness”. While I actually agree with Rand that the cliche concept of selfishness as being synonomous with immorality is nonsensical (I.E. the tendency in moral philosophy to completely divorce ethics from self-interest is a grave error), I think that her own pet definition of “selfishness” merely makes the opposite error by completely and absolutely conflating selfishness with morality. Rand essentially redefines “selfishness” to be the very essence of morality. In her view, to be selfish is to be moral, to be moral is to be selfish, to be rational is to be moral, to be moral is to be rational, to be selfish is to be rational, to be rational is to be selfish.
In other words, selfishness, rationality and morality are all absolutely conflated into the same thing by Rand. To be immoral is inherenty to not be selfish, using her definitions. This is highly problematic in light of the more psychological understanding of selfishness. It seems absurd to say that a thief is not being selfish; the very act of stealing is presumably motivated by the desire of the thief to obtain the thing that they are stealing. It is highly counter-intuitive to claim that people who forcibly dominate others are not being selfish, or even to imply that they are inherently altruistic. Is a man who rapes a woman because he wants to appease his sexual desires “not being truly selfish”? This seems incredibly silly. The only way that Rand can get away with this is by essentially divorcing “selfishness” from motivation and emotion altogether. This is an incredibly rare and narrow definition of “selfishness”.
Hence, one of the most common objections that Objectivists are going to run into, besides the altruist definition of selfishness, is the psychological egoist or subjectivist definition of selfishness (I.E. that, in some sense, everyone already inherently is selfish). This definition of selfishness is psychological: it defines selfishness as being inherent to motivation itself, and in this sense it is impossible to not be selfish, since all motiviations derive from the subject. From a broad psychological egoist perspective, behavior that we would normally consider both “rational” and “irrational”, and both “moral” and “immoral”, are all “selfish” in some sense. There is nothing about “selfishness” that inherently means “moral” or “rational” – it could go either way depending on the context.
So we see that in the case of both “selfishness” and “happiness”, there is an inevitable conflict between the psychological definitions of these terms and esoteric definitions in particular moral philosophies. Since the common understanding of the meaning of these terms is not likely to change any time soon, it seems counterproductive for proponents of these moral philosophies to continue insisting on using such terms in this way. It will always lead to a need for clarification. Either one must establish that there is a dual definition of such terms, concede to the psychological definition of the terms, or continue to brashly insist that these esoteric definitions are the “true” meanings (to the likely result of continueing the confusion). It probably would be wise to at least make one’s context clear before even using such terms in such ways; otherwise, I’m tempted to suggest dropping them altogether.