@Frank Apisa,
Without getting into too much math:
Subjectivity (by definition) is the perspective of an observer.
Objectivity (by definition) is what is real absent an observer.
Our (human) condition is that we are observers, our experiences are by definition subjective ones.
Objective reality is only implied by a "consensus" of observations, objective reality "makes sense" as a model because it conforms to our observation. Things seem to exist whether you observe them or not [cue the QM discussion].
We have no reason to believe that there exist observers without interests. Every "observer" we observe has them. We imply the existence of these other observers by the way they interact with the objective model of reality. In short they behave as though they can also observe objective reality. They also behave as though they have interests. The skill of observing other observers and attempting to understand their perspective is called empathy. Empathy is amoral. It is a skill.
As an observer
you feel preferences (interests) these may be anything, but they have to be something. Show me a person without interests if you think this untrue.
Compassion is using the skill of empathy for the interests of another observer. Empirical evidence shows that the more interests that individual human beings place in another observer the happier and healthier the are. They shows of signs "happiness" and "health". This is still not a moral position. This is a labeling of certain states. These states are more stable, they confer on the observer less risk of death (they are healthier). Being "happy" makes you more inclined toward more compassion (feel good, do good phenomenon). This is a constructive circularity or a vicious loop (depending on whether "happy" is good or bad).
Anti-compassion is the non-use of empathy for the interests of another. This might be by not having empathy as a skill, or by not having interests outside of self, or both.
For the case of not having empathy. Not having the skill to know what the interests of another observer are. In this case there could be no intentional help provided, since there is no knowledge of what the other wants. There is also no possibility for intentional harm, since there is no knowledge of what the other wants. This is ignorance of others' interests. There can be no moral actions in this case. Amorality.
For the case of having empathy, but no interests outside of self. This case is one in which all interest is self-interest. If this self-interested person is rational, they will examine which behaviors are going to lead to fulfilling their own self-interests (Machiavellian). If the Machiavellian acts in ways which appear self-interested they will of course have some success. Success as measured by health which is simply a measure of stability. If the Machiavellian chooses to place interest outside of self, they will have more success (by psychological empirical evidence), they will "convert" to compassion.
Assume the self-interested person is not rational, they will be incapable of judging what is or is not in their own best interest. They will be less successful on average than the either Machiavellian, due to their inability to even predict their own interests. This is ignorance of self interest. Moral behavior is impossible, because it cannot be intentional. Amorality.
The only situations which allow for moral behaviors are the compassionate and the Machiavellian. The difference being in whether the totality of interest is in the self (M) or any interest outside (C).
C's are healthier than M's. They are healthier, they live longer. (as supported by psychological evidence)
Now look at the situation as an optimization problem, looking at the stable state. The C strategy is superior on a "sociological" level. It is a more stable solution. The M strategy is volatile (less stable).
Societies favoring the C strategy will tend toward existing longer than societies favoring the M strategy. Because collections of C-ish people create a more stable state, than collections of M-ish people.
Dynamics favors C to M as a strategy of interacting agents.
This would be true no matter what the arbitrary interests of the individuals happen to be. You could assume that anything is the "base preference" of individuals (
the best thing in the world is watching reruns of Lassie for example, or running around in circles) the process of optimization will favor the steady states that employ C-ish strategies in the individuals.