My own position is the same as what I understand the Vatican's official one to be, at least if its most extensive document on the subject be taken at face value. Given the relative frequency of condom misuse and/or failure, and condom use's counterproductive effect of encouraging sexual excess, even strictly prophylactic condom use is manifestly imprudent as a general course of action. On the personal level, the only prudent way to handle the sort of situation in question is abstinence; on the policy level, strengthening marriage, the family, and sexual morality would be far more effective than promoting the illusion of "safe sex" by distributing condoms. There is, though, a possible exception: the case when somebody aims to protect herself from an infected husband who demands his "conjugal right" regardless of both the situation and her inclination. That indeed is the crucial test case. Like Cardinals Barragan and Martini, who cite it as one in which "the lesser evil" could be justified, I am not willing to say that every act of condomistic sex is an inherently immoral sort of act aside from any considerations of prudence, intention, or anything else. I am unwilling to say that because it would rule out even the self-defense intended in the test case. But even that little bit of "give" gets me into trouble with the other end of the spectrum.
Privately in 2006, I was told by a prominent, scholarly absolutist at that end that the Barragan-Martini view ('BMV' for short), and thus my own, is flatly incompatible with the entire logical structure of the Church's teaching about sex and marriage. In the combox of my most recent post pertinent to this topic, it has even been suggested that my defense of said view is incompatible with the very idea of certain sorts of act being intrinsically evil, and thus with a pivotal, irreformable principle of Catholic moral teaching: evil may never be done so that good might come of it. Both charges are very serious, indeed decisive rebuttals if true. Such reactions are not of course as emotionally extreme as the homicide accusation from the other end; but conceptually, they go almost as astray. I shall now show how. At the end, I shall restate my position to accommodate the critics' concerns.
In general it is a necessary condition, for any action or omission A to be subject to moral evaluation at all, that A be an intentional action or omission. That means the agent must be able, at least in principle, to answer the question "Why did you do [not do] that?", and the answer must constitute a reason for what was done or not done. Many acts are of course voluntary without being intentional. E.g., I voluntarily if marginally contribute to air pollution by operating a car; but that is not my reason for operating a car and therefore not my intention in doing so, so that ceteris paribus I can't be held blameworthy for driving a vehicle that contributes to air pollution. In general, to be held blameworthy for an undesirable side effect of one's action, it is not enough that one foresee it and voluntarily cause it. One must intend it, either as an end in itself or as a means to the end in view. Of course omissions can be intentional. E.g., if I own a factory that fouls the river-water some people need, foreseeing but not positively "intending" the pollution, I can still be held to intend the pollution by omission if it is practically possible for me either to deal with the waste by-products by a safer means or, if not, to earn a living by some other means. From the perspective of just-war theory, considerations of such a kind are especially important for evaluating actions both ad bellum and in bello. Yet given what's at stake in the dispute over BMV, I shall now confine myself to discussing positive actions as opposed to omissions.
In order for positive action A to count as intentional, it must "embody" an intention, to use a term from the late Elizabeth Anscombe's classic book Intention. To specify the intention embodied, the question "What are you doing?" has to be answerable by a description of the person's physical action that makes the action intelligible precisely as intentional. For that purpose, however, it is not enough to cite some further or ultimate intention of the agent. E.g., if I go to the grocery store, pick some healthy foods off the shelves, pay for them and leave, it is not enough, for purposes of answering the question "What are you doing?", to say "maintaining my health," even though that would often be true. That might indeed be my ultimate intention; but since there are still indefinitely many other actions that could just as readily be described as "maintaining my health," I do not describe my action in particular by calling it health maintenance. No, what I am doing is "getting groceries," which is the intention embodied by all the little sub-actions I perform in the store. What I'm doing in the store embodies the intention of getting groceries, so that the complex, intentional action I'm performing is "getting groceries." Typically, getting groceries can also be cited as a means to the further end of maintaining health. But what characterizes the intention embodied by my action is not my further end of maintaining health as a reason for action; rather, the intention embodied is described by that answer to the question "What are you doing?" which makes my complex physical action itself intelligible as intentional.
Now in Veritatis Splendor, John Paul II spoke of the "object" of an action thus: "The primary and decisive element for moral judgment is the object of the human act, which establishes whether it is capable of being ordered to the good and to the ultimate end, which is God." (§80). By the "object" of an action I took and take him to mean the intention embodied in the action, where embodiment is to be understood as I have just explained it. So if the object, so understood, is of such a sort that it is a disorder of the will to intend it, then the intentional action itself is of an intrinsically evil sort, irrespective of the further intention with which somebody might do it. That principle is essential for upholding the still more basic principle that evil may never be done so that good might come of it, which rules out "proportionalism," "utilitarianism," or indeed any variety of what is now called (to use another term Anscombe coined) "consequentialism." I am as opposed to consequentialism as any of my conservative critics. So if anybody wants to say that BMV is consequentialist, they would have to show that it either assumes or entails consequentialism even if its proponents do not intend as much. That is not easy to do. And that's mainly because the conservative critics of BMV do not have a cogent way of specifying just what object of action in prophylactic condom use is of a sort that it is a disorder of the will to intend, thus making a sexual act involving such use an intrinsically evil sort of act irrespective of the further end of defending the health of the uninfected partner.
To appreciate the difficulty, consider first that it is agreed on all sides that, if and when a given sexual act A intentionally culminates in male ejaculation, A is morally acceptable only if A is a "conjugal act" between spouses. Thus, if A is not a conjugal act, then A is of a intrinsically evil sort, and thus morally unacceptable regardless of its further or ultimate intention. So, the question under dispute may be framed thus: Can an otherwise "conjugal" act of sexual intercourse still be conjugal if a couple, one of whose members has AIDS, use a condom for the sole purpose of preventing the virus' transmission to the uninfected partner? The critics of course say no. Obviously, the cogency of their position depends on that of their definition of a conjugal act. It is on this point that the discussion needs to focus.
The critics take their cue from magisterial documents. Thus Andrew McCarthy:
Church teaching on this subject appears to be settled by Pope Paul VI’s official confirmation on 28th January 1978 of a CDF declaration dated 13th May 1977 which dealt with the question of what is required in order for a sexual act to be capable of consummating a marriage. See Sacra Congregatio pro Doctrina Fidei, Decretum circa impotentiam quae matrimonium dirimet, Acta Apostolicae Sedis 69 (1977), p426, Pope Paul VI, Allocutio ad Tribunalis Sacrae Romanae Rotae Decanum, Praelatos Auditores Officiales et Advocates, ineunte anno iudiciali, die 28 ianuarii 1978, Acta Apostolicae Sedis 70 (1977), pp183-184. For more on this whole debate see McGrath Aidan O.F.M, "A Controversy Concerning Male Impotence," Editrice Pontifica Universita Gregorina Roma (1988). The basis for understanding what is and is not correct behaviour in sexual ethics is a correct understanding of the conjugal act. The conjugal act must be, minimally, an act capable of consummating a marriage. Deliberately ejaculating outside of a woman's vagina cannot constitute a consummatory act of a marriage and therefore cannot be a conjugal act. An ejaculatory act that is not a conjugal act is grave matter. The traditional teaching of the Church has always condemned ejaculatory acts deliberately aimed at places other than a spouse's vagina. The 1978 confirmation of Pope Paul VI clears up this matter and shows that a condom affects the act per se, not per accidens.
But as it stands, that does not suffice to rebut BMV. For in the sort of case we're considering, the man does indeed ejaculate within the woman's vagina. Of course it is replied that that doesn't count, because the man does not actually deposit seminal fluid in the woman's vagina. So the key question then becomes: is it a necessary condition for a sexual act to count as a conjugal act that seminal fluid actually and intentionally be deposited in the vagina, not merely ejaculated within it? If the answer is yes, then no act of condomistic sex can count as a conjugal act, regardless of the purpose of the condom's use. If so, and given premises already shared, then any act of condomistic sex embodies an intention to do something that, objectively speaking, it is a disorder of the will to intend. Therefore, any such act would be of an intrinsically evil sort, and thus evil regardless of how laudable the ultimate intention might be.
Even so, to the question why deposition of semen is thus necessary, the answer is by no means obvious. Following McCarthy, one might think that such deposition is a necessary condition of a sexual act A's being procreative per se, and that A's being procreative per se is itself a necessary condition for A's being a conjugal act. But it is simply not the teaching of the Church that, for a sexual act A to count as a "conjugal act," A must be procreative per se. In what is perhaps its most widely quoted sentence, Paul VI's Humanae Vitae (§11) asserted that "...each and every marital act must of necessity retain its intrinsic relationship to the procreation of human life." (The original Latin of HV's assertion reads quilibet matrimonii usus ad vitam humanam procreandam per se destinatus permaneat; the Vatican translation I've used, though not literal, is superior to the better-known one using the phrase "open to the transmission of life," but I mention that only to minimize confusion about which passage I'm discussing.) That means that the act must be procreative per se even if it happens to be non-procreative per accidens (which, incidentally, is how it is possible for NFP to be sometimes right even though contraception is always wrong). Now if a given sexual act is a "marital act," a matrimonii usus, then it is a conjugal act, i.e., an act of a kind that would count as consummating a marriage. That can and often does hold even if the couple are contracepting, which is why it is possible to consummate a marriage even if, say, the wife is taking birth-control pills. Yet according to HV, contraception is wrong because it severs the conjugal act's "intrinsic relationship to the procreation of life." If contraception does that, then contraceptive intercourse between spouses of a kind that nonetheless involves deposition of semen in the vagina fails to qualify as procreative per se, but is still a conjugal act. Therefore, being procreative per se is not a necessary condition of a sexual act's being a conjugal act. Being procreative per se is a necessary condition only for the conjugal act's being morally licit. And so, if condomistic sex fails to qualify as a conjugal act inasmuch as it embodies the intention to avoid depositing semen in the vagina, that cannot be because its embodying such an intention prevents the sex from being procreative per se.
Alexander Pruss, a philosopher at Baylor, has suggested what one might view as a way round that objection. According to him, a "positive intention to transmit genetic material" from husband to wife is a necessary condition for an act to be "marital" and thus conjugal. If that were so, then no act of condomistic sex could be a conjugal act. But Pruss' answer can't be right either. Suppose a widower who had undergone a vasectomy during his marriage goes on to marry a woman who is well past her childbearing years. As a Catholic of newly invigorated orthodoxy, he now regrets his vasectomy and would have it reversed if he were marrying somebody capable of bearing children. But given his wife's perfectly natural infertility, he sees no point in going to the trouble and expense of reversing his vasectomy. So he doesn't. Accordingly, no sexual act of his could embody the positive intention to transmit his genetic material. For he knows he cannot transmit the sperm that carry such material, and nobody can be said to positively intend to do something he knows he cannot do. Does that mean that this man and his new wife are incapable of performing a conjugal act and thus of consummating their marriage? Of course not. But if a "positive intention to transmit genetic material" were a necessary condition for a conjugal act, that obviously true answer would be false.
One might still maintain, however, that even if condomistic sex can sometimes qualify as a conjugal act, it is intrinsically evil inasmuch as condom use that is purely prophylactic and without contraceptive intent still precludes the sexual act's being procreative per se. From this point of view, condomistic sex is morally no different from ejaculating within a human orifice other than the vagina: even if there is no intent to block procreation, neither is there an action bearing any "intrinsic relationship" to procreation. But I don't think that will do either. Vaginal intercourse just is the sort of act that is procreative per se, and condomistic sex is vaginal intercourse; that is precisely why conception can occur when condoms fail, even when one's intent is contraceptive, whereas ejaculating within some human orifice other than the vagina cannot lead to conception even if one wishes, perversely, that it could. It would be more sensible to say that condomistic sex fails to qualify as procreative per se, and is thus immoral, when it embodies the intent to contracept. But ex hypothesi, condom use for purely prophylactic purposes embodies no such intent; so if such use is intrinsically wrong, it must be so for some other reason.
The only specific such reason I've seen suggested is that of Luke Gormally and Hugh Henry: deposition of semen in the vagina is a necessary condition for the conjugal act to retain its unitive significance, even aside from the question whether condomistic sex is procreative per se or not. Relying on JP2's theology of the body, they argue that each and every conjugal act must embody the intention of complete "self-donation," and that no sexual act which is incomplete, in virtue of withholding from one's partner something natural to the act, could embody the intention of complete self-donation. Of course it is important to remember here that condom use for purely prophylactic purposes does not prevent semen deposition as a mere sexual preference or as a way of withholding love. Semen retention by the condom is intended only a means to the end of prophylaxis. But the point being made by LG and HH is that regardless of the further end in view, the intended means deprives the conjugal act of the unitive significance it must retain in order to be morally licit.
Now one must certainly concede that condomistic sex is truncated sex. I've never met anybody who claims to prefer it to, or even to like it as much as, purely natural sex. Something is indeed missing. But provided that the intent is purely prophylactic, I confess being unable to follow the argument that what's missing in condomistic sex is something essential for complete spiritual self-donation. Perhaps others can find something in the argument that I've missed. But I find myself unpersuaded that condom use in the sort of case under consideration is intrinsically evil qua anti-unitive, as distinct from a lesser but still tolerable evil, physically truncated and thus less satisfying than purely natural sex.
In the sort of case under consideration, though, most of this debate may turn out to be inconsequential. One could plausibly argue that an HIV-infected person who insists on sex with their spouse, despite the spouse's wishes, intends an act that really does lack unitive significance apart from the question of condoms. And so the sort of self-defense intended by condom use in such a case would be a defense against an act of rape not a conjugal act. Of course there are always those cases in which both spouses are willing to take the risk of sex. For such couples, I can only cite the same considerations of prudence that Cardinal Trujillo did in the Vatican document I've already cited. In either case, I do not believe it can be accounted an act of love, as distinct from an act of mere physical gratification, to have sex under such conditions. And maybe that's all one needs to specify what it's a disorder of the will to intend in such a case.