The Corner

The one and only.

Marriage Revisited


Congratulations to Sherif Girgis, Robert P. George, and Ryan T. Anderson on the impending publication of What Is Marriage? I read a galley copy over the weekend. It is a formidable piece of thought expressed with precision and elegance. That authors of such thoughtfulness and goodwill are routinely dismissed as bigots is a gross unfairness, and one that angers me.

I am not persuaded by the argument. I think it remains vulnerable to the objections I raised in “Two Views of Marriage, and the Falsity of the Choice between Them” and “Two Views of Marriage: Appendix.” I won’t recapitulate all the details, but I’d like to restate the central weakness I perceive in the position of Girgis, George, and Anderson (henceforward “Girgis”), as well as the central elements of my alternative position. I will then make a few general comments about the marriage debate.


The weakness is that Girgis does not convincingly explain why marriage as he defines it is a distinct good for all the couples who fall under the definition, or why they all should abide by the norms of permanence, monogamy, and sexual exclusivity.

What distinguishes marriage and its intrinsic good from other types of relationship, says Girgis, is that a married man and woman engage in the behavioral part of the body’s reproductive function. “They are united . . . much as one’s heart, lungs, and other organs are united: by coordinating toward a biological good of the whole that they form together. Here the whole is the couple; the single biological good, their reproduction.” The reproductive function is also said to ground the norms of marriage, since these norms establish the conditions best suited to child-rearing. All of this is said to hold regardless of whether the couple expect to have children or are able to have them.

The problem is in that regardless. Let’s allow, as Girgis says, that reproduction is an intrinsic good. This does not establish that simply taking a step in the reproductive process is good. Imagine a married man and woman who know to a certainty that they cannot reproduce. Is it plausible that they will find the sexual dimension of their relationship valuable because it is the behavioral component of the reproductive process? Or will they value it for some other reason? Will they abide by the norms of marriage because these are the norms suited to reproduction? Or will they do it for some other reason?

To put the point more formally: Girgis claims that a type (kind) of act—coitus—makes possible a distinct value, and is governed by distinct norms, for all individuals who engage in tokens (instances) of it. But not every token is capable of achieving the typical end with reference to which the norms and the value are defined. And to bridge the normative and valuational gap between type and token, Girgis offers nothing but his say-so.

By contrast, I present, in the “Appendix,” the following definition of “procreative orientation,” and I apply it as well to the “toward” in Girgis’s “coordinating toward a biological good”: A token of coition is procreatively oriented if it is possible that it will result in reproduction, or if those who engage in it believe that reproduction is possible. (The extension of the type “procreatively oriented sex” is thus a subset of the extension of the type “coital sex.”) It is the possibility of reproduction that makes bodily union valuable, and it is the belief in this possibility that leads married men and women to abide by whatever norms its realization implies.

If I were inclined toward a traditional natural-law ethics, the sheer ungroundedness of Girgis’s claims would lead me to reflect on what a shriveled, crumbling thing the natural law can become when shorn away from a metaphysical foundation (for example, a belief in the metaphysical reality and teleological orientation of types1). “New natural law? I’ll keep the old one, thanks!” On the other hand, “The New Natural Lawyers” strikes me as very cool name for a band—maybe a bluegrass band. (Just kidding.)


As it happens, I’m not inclined toward natural-law ethics. I think any claim that something is valuable—as distinct from the most precise possible description of what is valued—is nothing more than a say-so.2 So I don’t begrudge Girgis his say-so, at least on meta-ethical grounds. But I do think mine is more plausible. It is as follows.

Romantic attraction is a unique type of desire in which a person is wanted in his or her unity and totality, and sexual activity is the unique expression and bodily dimension of such desire. The desire is thus unique in both its “inner” (“subjective,” “mental”) and its “outer” (“objective,” “bodily”) dimensions, and its fulfillment is intrinsically good.

Such desire and activity need not (but may) result in bodily union as Girgis defines this. He will therefore object to my “unity and totality,” since the bodily dimension of the individual as defined with reference to reproductive function may be omitted. This would be a persuasive objection if romance worked like this: First you think, “I desire a mind, a reproductive function, etc.”—what Girgis calls the “basic dimensions” of a person—and then you seek some suitable instantiation of them. But that doesn’t seem to be how it is. It’s not like wanting some good dinner, some pleasant vacation, etc., where all instances meeting certain criteria are fungible. The desire is for that individual; this is the most fundamental level of analysis.

From there, one seeks the most total possible union with the beloved (consistent with the hope that the union will be sustainable; for example, you don’t expect to spend every waking minute together, because that would be destabilizing). Sexual activity is the characteristic expression of such desire, because it is the most complete possible integration, in a single act, of the individuals’ bodies, minds, and wills in their directedness toward each other, as well as the most vulnerable kind of reciprocal self-surrender available to them. The surrender would not be absolute if it were not exclusive; and when the individuals wish to extend the union it expresses into shared domestic life and a public vow of permanence, it becomes a marriage.

In the absence of children, there is nothing external to the relationship compelling the couple to live up to their vow; but this is also true, on my view, of certain relationships that Girgis is willing to recognize as marriages: namely, those of couples who know they will not reproduce. Similarly, childless couples have no motivation for monogamy and fidelity other than their belief that the absoluteness of their reciprocal self-surrender entails these commitments, but that doesn’t mean the traditional norms are implausible as a common-sense default. (It also doesn’t mean we should condemn a childless couple who, owing to idiosyncrasies of their dispositions, prefer to have an open relationship. It is possible for something to be aberrational without being blameworthy.) Again, if our hypothetical couple are a man and a woman, Girgis will say that the mere fact of their bodily union (as he defines this) is sufficient to ground the norms; but that is precisely what is in dispute.

I see some hands going up to object that the “subjective” or “inner” dimension must be excluded from discussion, since it isn’t susceptible of public verification. I will return to this thought in discussing civil marriage, but to grant the objection here—while we ask what marriage is as a pre-legal matter—would amount to ruling out discussion of the “inner” altogether. It’s true that we communicate about the “inner” only insofar as it can be correlated with public signs, but this hardly prevents meaningful discussion of the entire mental dimension of our existence.

Indeed, Girgis’s claim that, on my view, marriage is distinguished mainly as offering more of what is valuable (emotional intimacy) rather than something qualitatively unique depends on treating emotional intimacy as a kind of free-floating abstraction rather than attending to the publicly observable circumstances in which it is found. He could claim just as easily, and just as implausibly, that there is no appreciable difference in kind between what one feels for a grandparent and what one feels for a best friend. Besides, I dispute his claim that the non-bodily dimension of marriage (or any other relationship) is merely emotional, merely a private good concurrently experienced by two people. For love is not just a feeling. It is also a group of mutual expectations and commitments, and the willingness to live up to them; and this aspect—what I would call its “intersubjective significance”—is no less a shared good than is our mutual understanding of each other’s words if we converse. (For more on this, see my “Appendix.”)


Onward to civil marriage. I think traditionalists are correct that the law has no special reason to concern itself with the emotional and intersubjective dimensions of relationships. (I do think it ought to ensure that any combination of individuals who, for whatever reason, wish to ensure for one another such things as hospital-visitation rights, property inheritance, and so on are able easily to do so.) On the other hand, if my claims are correct, there is a clear sense in which same-sex couples make the same kind of choice as men and women in dyadic relationships, and in which this choice is different from the choice of individuals in polygamous or polyamorous relationships. The sense is this: Whom one is romantically attracted to is not a choice, but whether to pursue an attraction is. Someone who pursues only one relationship at a time (whether or not it involves a vow of permanence) therefore makes a different kind of choice from someone who pursues simultaneous relationships (whether or not they involve a vow of permanence).

There is accordingly no reason in principle not to extend civil marriage to committed same-sex couples: It introduces no novel confusion about the norms and good of marriage, since Girgis’s concept of bodily union already fails to ground the norms and explain the good for certain included couples, and since these couples and same-sex couples have identical grounds to abide by the norms. The reason to include the latter isn’t that we’ve suddenly embarked on “the legal regulation of tenderness,” as Girgis pithily puts it, but rather that an asymmetry has come to our attention and, as a matter of fairness, we want to even it out. After all, we don’t have to regulate procreation indirectly via regulation of relationships. We could regulate it directly instead: by making parents strongly accountable for the material support of their children, by giving them incentives to merge their domestic affairs, etc. These methods might be less effective or efficient than the traditional approach. But the price we pay for concerning our law directly with relationships rather than procreation is that we must attend to the structure and nature of the relationships themselves, being careful to treat like cases alike.3


My argument is theoretical, and considerations of prudence may lead even those who agree with it to oppose same-sex marriage. Such considerations might include a fear that recognizing same-sex unions would launch a campaign of legal persecution against conservative churches and businesses, and a suspicion that such recognition would in practice distract the public from the importance of child-rearing by increasing the attention paid to the intersubjective and emotional dimensions of marriage.

Traditionalists should nonetheless think through the theory carefully as they decide how they will oppose same-sex marriage. Girgis’s position offers little room to maneuver. It entails that we might as well start marrying groups of three and twenty-three if we start marrying same-sex couples. It is a last stand. If it’s what you really believe, there’s no helping it, and there can even be something beautiful about jumping off a cliff.

My position offers a principled basis to resist further reforms. Note as well the deep consistency between the two-tiered marriage law I have proposed—in which couples are discouraged from divorcing only after they have children—and my claim that reproduction becomes normatively relevant when reproduction occurs. It’s not a reversal of the sexual revolution, as some traditionalists hope for, but it is a way of reconnecting marriage law with the interests of children in the world we actually live in.

Girgis has tended to write as if no alternative to his position were even coherent, and maybe I have as well. But that just isn’t true. Each of our positions is internally consistent, and each of us is perfectly able to meet the other’s objections—just not on the other’s conceptual turf. This is in the nature of philosophical conversation: One rarely wins one’s premises, but disagreements can still be a useful mirror—if sometimes an inverting mirror—in which to recognize one’s thoughts.


Some quibbles with the references to me in the book.

First, Girgis thinks that in the the final sentence of “Two Views” I “scoffed at the implication that ‘the value of a relationship between two persons in love [would depend] on the structure of their genitals.’” Last time he quoted me I “sneer[ed].” To me the sentence is affectless, but if I must have one attitude or another projected into it, I would rather scoff than sneer—so thanks for the upgrade. Substantively, I grant that the sentence obscures what Girgis wishes to emphasize. Its purpose, though, was not to “ridicule[]” his view but to illuminate my own: specifically, my objection to his claim that coital sex has a distinct value apart from whether it might lead to reproduction. His preferred term, “life-giving act,” obscures that. (For my riposte to his remark about Romeo and Juliet, see again my “Appendix.”)

Second, in the book’s penultimate paragraph, Girgis writes that “it would be a mistake to infer, from these fine distinctions about what counts as real bodily union, any abandonment (in principle or practice) of the integrated view of human beings as body-mind wholes.” Here he gives an endnote reference to “Two Views.” I’ll concede that if his presuppositions are granted, his argument involves no such abandonment. But there can be other presuppositions, and the disagreement turns on a deeper meta-ethical disagreement between us.

Girgis reflects upon the mind and body as generalities prior to imagining the situations in which a body-mind unity could find itself. He arrives at a general idea of what bodily union is and a general idea of what mental or spiritual union is, and then he makes a claim about why both types of union, as he has defined them, are valuable to a human being. By contrast, I think we must concretely imagine situations in which a body-mind unity could find itself, then reflect upon the choices available to it in such a situation.

If the situation is that of someone who is in love with a person of the same sex, I think Girgis’s approach will give a distorted view of his or her interests. It will ignore important aspects of the relation between mind and body as they are actually integrated in such an individual, and will present as a universal good something that such a person may have reason not to seek. (In passing, though, let me say that the sentence in “Two Views” about lexical ordering now strikes me as presumptuous. A decision to forgo procreation for the sake of romantic love is momentous, as is its reverse. Both seem understandable and defensible as long as no deception is involved, and both can fit under the concept of marriage I have presented.4) I hardly blame Girgis for not spelling out the finer points of my view, but I would be grateful if he supplemented his endnote with references to the “Appendix” and to this article.

I end by again offering Girgis and his co-authors my sincere congratulations.



1. But note that some further step would be needed to establish that the metaphysically and ethically relevant type must be defined solely in terms of the behavioral portion of the reproductive process, and exclude broader consideration of types of individuals who might engage in it: for example, whether they are naturally capable of reproducing together, and whether they are naturally attracted to persons of the opposite sex. Note further that the answer to the latter question may depend on physiological, which is to say bodily, factors. Compare remark IV:4 of the “Appendix,” and the comment about “myopic biology” and related discussion in “Two Views.” (11/8/12)

2. I emphatically do not mean anything nihilistic by this. See section II of the “Appendix.” (11/8/12)

3. A traditionalist could still reply as follows: “I grant that a relationship between two persons of the same sex can instantiate the same kind of moral reality as a relationship between a man and a woman who know they will not reproduce. But this doesn’t mean the cases are alike with respect to the concerns of public policy. Civil marriage came about only to protect children, and the legal category is drawn as narrowly as possible around those couples who may reproduce. The inclusion of men and women who know they will not have children is due purely to the impracticality and intrusiveness of administering fertility tests prior to granting marriage licenses; it is a kind of accident. Same-sex couples, then, should understand that the law’s purpose was not to exclude them.” As far as it goes, this reply seems sound to me, and it points to the way in which efforts to recognize same-sex unions are different from the civil-rights movement, which sought to overturn a legal regime whose specific purpose was to discriminate against a racial minority. But note two things. First, the reply surrenders the standard argument as to why marrying same-sex couples would be harmful—viz., that doing so would undermine the norms of marriage—since it concedes that the norms make sense for such couples. At least in principle, the traditionalist will be left without a compelling answer to the question “Why not?” Second, our purpose in retaining a law need not be the same as our purpose in creating it. Indeed, while there is a fact-of-the-matter, discoverable in the historical record, about why a law came to exist, there is no fact-of-the-matter about why we retain it. And it is simply true that much of the public sees civil marriage not only as a mechanism to protect children, but also as a form of social inclusion and moral approbation. (Such a view becomes more plausible when we recognize — as I mention above — that the law might regulate procreation directly. Instead, it recognizes and confers benefits on a type of relationship whether or not children have been born into it. This reinforces the belief that civil marriage achieves something more than the protection of children.) To the extent that this understanding of marriage prevails, and to the extent that my objections to Girgis’s analysis of the pre-legal moral reality are sound, it is unfair not to marry same-sex couples. This unfairness would not be a reason for courts to redefine civil marriage, since courts should concern themselves only with the documented purposes for which laws were passed. But it would be a reason for legislatures and voters to redefine civil marriage. (11/21/12)

4. I am presenting a unitary concept of marriage as a publicly identifiable pattern of life-sharing that is subject to a single set of norms. The norms are to be regarded as a common-sense default that holds up under reflection. I am presenting a disjunctive concept of what may justify these norms in a given case: either reproductive partnership (where “procreatively oriented” is defined as I’ve defined it) or romantic commitment. (Note: Even the first should be seen as giving rise only to a common-sense default, since it’s conceivable that some particular set of parents could raise children successfully without committing to sexual exclusivity.) In most cases of marriage, these disjunctive justifications both apply; we might say the norms are overjustified. At the margins, the justifications can come apart—either because two people are in love but can’t reproduce, or because two people want to reproduce but aren’t in love—but the norms still apply, and observers would still detect the same pattern of life-sharing. (11/8/12)


Sign up for free NRO e-mails today:

Subscribe to National Review