There are two fairly standard claims about love:
- benevolence--willing the good--towards the beloved is an essential component of love, and
- love gives one the reason par excellence to benefit the beloved.
As usual in the case of such a tension, there are four options available. Discarding both (1) and (2) is implausible.
Frankfurt in effect discards (1)--it is care, but not benevolence, that is an essential part of love on his view, and this care justifies the benevolence. But this is implausible, as it seems to imply that one could love someone and at the same time will the worst evils for one's beloved. Even if one adds a claim that justice towards the beloved is a component of love, still the idea that one could fail to will anything for one's beloved, despite the opportunity to do so, and still count as loving is implausible. Also, such a view is incompatible with the Christian idea that love fulfills the law--a love that does not imply some benevolence does not fulfill the moral law.
One could discard (2) but keep (1). On this view, I think one will end up saying that it is not love that justifies benevolence, but whatever it is that justifies the love likewise justifies the benevolence that is a part of this love. Another attractive feature of this view is that if one explains why one made some sacrifice for Bobby, and one says: "Bobby is my son and I love him", the "and I love him" is one thought too much--it is the sonship that justifies both the love and the sacrifice that is a part of the love. If one hates Bobby, one still has reason to make that sacrifice. I like this view a lot.
One could just be careful and hold on to both (1) and (2). Thus, some benevolence will be an essential part of love, and the love will then justify further benevolence. But what justifies the bit of benevolence that forms an essential part of love, and why does it not also justify all the benevolence by itself? Suppose our proposal is that the initial bit of benevolence is justified by the fact that we're all children of God, or that if something is good for x then it is good and hence it is worth promoting, or some other general claim like that. But such a fact seems to equally give one a reason for unlimited benevolence. Moreover, let "love*" be the aspects of love beyond the bit of benevolence that is an essential part of love (maybe love* is appreciation and/or a desire for union). Then, ex hypothesi, love* plus the little bit of benevolence justifies more benevolence. But it is not clear how a little bit of benevolence can help justify more benevolence. So, then, it seems it is love* that justifies the extra benevolence. But if so, then it is not love, properly speaking, that justifies the extra benevolence, but love*.
Perhaps one can combine the best of the last two views. Whatever it is that justifies the love justifies all the benevolence involved in the love. However, the love creates an additional reason for continued benevolence, and after the initial moment of love, the benevolence is rationally overdetermined. Thus, "Bobby is my son and I love him" correctly states two distinct reasons for making a sacrifice for Bobby, but is felt to be one thought too many because of an implicature that if one did not love Bobby, one wouldn't have the reason. This seems to be the best view.