Is it still possible, then, to maintain the union theory of romantic love? If we do wish to defend a union theory of love, it looks as if the moderate ontological model would represent the only promising candidate as far as a philosophically reliable position is concerned. Nonetheless, and here we come to the decisive point, it turns out that “romantic love” can neither be defined by the presence of a union in the moderate ontological sense, nor can the presence of such a union be understood merely as a sufficient condition of such love. One could, as we have just seen, respond to objection (e) by regarding the presence of particularly deep or intensive forms of union as sufficient for the presence of romantic love. But it is by no means obvious why we should not also be prepared to recognize the presence of a similarly deep union in the moderate ontological sense in the early love between mother and child, for instance. We might point out, with Fisher, that the mother also experiences the infant – at least in some cases – as part of herself, while the infant is not yet in a position to distinguish between the mother, the world, and itself; the life of the mother may completely revolve around the needs of the newborn, while the child may exist by completely depending on the mother.Footnote 28 Another example of a deep union that is not romantic love is often claimed to be found in the mystical love to God; and maybe the deep, but not romantic, love between twins (siblings) is another example. But if deep unions are also possible in other cases, the presence of such a union cannot possibly be a sufficient and distinctive feature of romantic love. And that can only mean that the essence of romantic love does not consist in the presence of a deep union after all. Rather, the presence of such a deep union is only (if at all) one feature amongst several such features of (fully realized) romantic love.
Here it would still be possible for the union theorist to develop a concept of a deep and a specifically romantic form of union that would include, for example, the aspects of sexuality, concern, trust, and the creation of a shared life. On this account, the (deep) union would basically be the genus proximum for a definition of romantic love; the differentia specifica of romantic union could then be located in the particular aspects we have just mentioned. Yet the ascription of additional characteristics, such as sexuality or concern (i.e., the ascription of the differentia specifica), is highly problematic: There are such things as asexual love, or jealous love marked by distrust, or loving relationships of people who do not, for whatever reason, wish to share their life in the long term, and so on. The notion of “concern” (or “care”) may indeed look like a promising candidate to specify the definition, but there is no doubt that this is just as much a central feature of friendship and of the love between parents and children. It cannot therefore be adduced (even in combination with the presence of union) as an unambiguous distinguishing characteristic here.Footnote 29 And it turns out to be the same in every case: Whenever specific and apparently necessary elements of a romantic notion of union are proposed, it is always possible to think of counter examples that challenge the necessity of these elements. Moreover, it is clearly not the actual presence of particular elements (desire, concern, etc.) which is required, rather it is the concept of disposition that must be brought into play here in one way or another (for those who love each other are not always concerned about one another). And we should also take account of the temporal aspects involved here, such as the question as to how we would think about a couple who once enjoyed sexual intimacy but from a certain age or point of time have ceased to do so.
We must acknowledge, therefore, that the presence of union in the moderate ontological sense is not a sufficient criterion for the presence of romantic love. At best, the presence of such union could be understood as a necessary element of love in general, including the love between parents and children and other forms of love (insofar as we are ignoring problematic cases here).Footnote 30 Hence, however much may be said to make the union theory of romantic love appear plausible (and it is certainly not absurd), it cannot, in the final analysis, be regarded as satisfactory.
In the light of the difficulties we have mentioned, however, the attempt to discover any definition of romantic love seems equally hopeless. If we look for several necessary features (such as concern, sexuality, specific emotional-phenomenal qualities, etc.), which, taken together, could seem sufficient to identify cases of romantic love (whether union is understood as one of these necessary features or not), we can always find counter examples that challenge the necessity of these features. Now one might object that such problematic examples represent extreme rather than paradigmatic cases and that we could at least still manage to define an ideal or typical love relation (especially since definitions are always somehow normative in character). Yet, once again, this seems hardly promising or even desirable because why should we think it justifiable, for example, to rank asexual love as non-ideal and thus as inferior? The fact that such love is not typical – i.e., that it does not correspond to the average case – does not show that it is in any normative sense deficient.
This presents us with a fundamental problem: A definition furnishes the essence of something and answers the question “what is x?” Should the fact that no convincing definition of love can be found lead us to conclude that love has no essential character at all? Must we perhaps content ourselves simply by identifying certain family resemblances here? But this does not seem very plausible either. For we are all familiar with love; we usually recognize love when we encounter it and sometimes we think that someone is merely imagining it. We believe that we know what love is and in any case that it certainly does have an essential character. And this familiarity with the phenomenon suggests one final possibility: Perhaps romantic love could be defined through a (partially) variable cluster of properties (although this would naturally mean abandoning the classical theory of definition).Footnote 31 According to this model, we could draw from a pool of, let us say, 15 conditions – none of which is either necessary or sufficient – and say, for example, that seven of these conditions must be present in any case of romantic love. Precisely which seven out of the 15 are relevant in a particular case would not be given in advance, but the presence of a certain union could be an element in this pool. We could thus list the characteristic 15 features a to o (union, concern, trust, affection, shared feelings, desire, jealousy, attractiveness, the wish to spend time together, permanence, etc.), and if at least seven elements from this set, whatever they may be, perhaps a to g or i to o, are present, one would be justified in speaking of romantic love. In this sense, there could be cases of love which exhibit no common properties, apart from precisely possessing seven (different) features from the given pool. But then there could also be love without union. This seems to me to be the most promising model, although it certainly remains unclear how large the pool of potential characteristic features of love would really need to be, precisely how many characteristic features must be present in order to be able to speak of romantic love, and which characteristic features belong to this pool at all. Perhaps one could postulate some necessary core elements in this pool (such as union, for example) that must be present in every case. Still, a particular number of further variable characteristic features from the pool must gather around those core elements if we are to speak of romantic love in a concrete case. And then we could say: It cannot be that two people love one another if they have never created a certain union, although it alone is not sufficient. To explore all this would be part of a larger project.