W hen I came out with a book making the case for same-sex marriage a few years ago, I expected to spend time selling gay marriage to straight people and marriage to gay people. The surprise was how much time I spent selling marriage to straight people.

By marriage, I mean not just a commitment that two people make to each other. Marriage is a commitment that the two spouses also make to their community. They promise to look after each other and their children so society won’t have to; in exchange, society deems them a family and provides an assortment of privileges, obligations, and caregiving tools. (Not, mostly, "benefits.") Marriage does much more than ratify relationships, I would tell audiences; it fortifies relationships by embedding them in a dense web of social expectations. That is why marriage, with or without children, is a win-win deal, strengthening individuals, families, and communities all at the same time. Gay marriage, I said, would be the same positive-sum transaction. The example gay couples set by marrying instead of shacking up might even strengthen marriage itself.

Audiences received my gay-marriage pitch in predictably varied ways. What consistently surprised me, however, was how few people thought of marriage as anything more than a private contract. Particularly among groups of younger people, the standard view was that marriage is just an individual lifestyle choice. If chosen, great. If not chosen, great. I would leave such encounters with a troubling thought: Perhaps straights were becoming receptive to gay marriage partly because they had devalued marriage itself.

In his new book, The Future of Marriage, David Blankenhorn begins where my doubts left off. Blankenhorn is the founder and president of his own think tank, the Institute for American Values, and has built his career on the restoration of fatherhood to the center of American family life. In The Future of Marriage, he emerges as an articulate, humane, and fair-minded opponent of same-sex marriage, which he regards as nothing less than part of an effort to steal children’s patrimony. "It would require us, legally and formally, to withdraw marriage’s greatest promise to the child–the promise that, insofar as society can make it possible, I will be loved and raised by the mother and father who made me." He takes jabs at me, among other gay-marriage advocates, but in my case he plays fair. And Blankenhorn is ambitious. He wants to lift the gay-marriage debate from its isolation in the mud-pit of the partisan culture wars and place it within a larger theory of marriage. He also wants to put an end to the days when gay-marriage advocates can say that there is no serious case against gay marriage. In both respects, he succeeds.

As I read, I made note of points on which he and I agree. I soon found myself running out of paper. Marriage, we both believe, is a vital institution, not just equal to competing family arrangements from society’s point of view but preferable; it is an institution embedded in society, not a mere contract between individuals; it is social, not just legal, and so cannot be twisted like a pretzel by court order; it has (almost) everywhere and always been heterosexual and entwined with procreation, and should be. Gay marriage, we both believe, is a significant change that entails risk (though we assess the risks very differently); but gay marriage, we also believe, is a supporting character in the much larger drama of shifting social values. We agree that heterosexuals, not homosexuals, will determine marriage’s fate and have handled matrimony pretty poorly without any gay help. And we agree that children, on average (please note the qualifier), do best when raised by their biological mother and father, though he makes more sweeping claims on that score than I would. That is a great deal of common ground, which makes it all the more interesting that we come out in utterly different places and that gay marriage, in some ways, turns out to be the least of our disagreements.

For Blankenhorn, "the most important trend affecting marriage in America" is not same-sex marriage. It is the "deinstitutionalization" of marriage–that is, "the belief that marriage is exclusively a private relationship"–of which gay marriage is merely a prominent offshoot. To his credit, he understands and forthrightly acknowledges that the individualistic view of marriage "has deep roots in our society and has been growing for decades, propagated overwhelmingly by heterosexuals."

Marriage creates kin. In society’s eyes, it distinguishes a relationship from a family. The trouble, for Blankenhorn, with declaring any old kind of relationship a family–with turning marriage into "a pretty label for a private relationship"–is that marriage evolved and exists for a specific social reason, which is to bind both parents, especially fathers, to their biological children. Same-sex marriage, he argues, denies this principle, because its "deep logic" is that a family is whatever we say it is, and it changes the meaning of marriage "for everyone" (his italics). For support, he draws on the writings of left-wing activists and academics who favor same-sex marriage precisely because, they hope, it would knock mom-dad-child marriage off its pedestal. Granting marriage rights to gay couples, who even in principle cannot unite biological fathers and mothers with their children, would "require us in both law and culture to deny the double origin of the child." Once that happens, we "transform marriage once and for all from a pro-child social institution into a post-institutional private relationship."

In plainer English, Blankenhorn is saying that marriage is designed to discriminate in favor of conjugal families and must continue to do so. Egalitarians may hate that idea, but it isn’t stupid or bigoted. Blankenhorn is correct to think society has a strong interest in keeping fathers, mothers, and children together; many of today’s problems of crime, poverty, and inequality flow directly from the breakdown of families. But there Blankenhorn and I part ways. He says he is all for maintaining the dignity and equality of gay people, but he believes that changing marriage’s most venerable boundary is the wrong way to do so. I am all for maintaining the strength of marriage and family, but I think that telling homosexuals (and their kids) they can’t form legal families is the wrong way to do so.

Having written a whole book on the subject, I won’t rehearse here why I think gay marriage is good family policy. Suffice it to say that, in a society riddled with divorce and fatherlessness, family policy’s essential task is to shore up marriage’s status as a norm. In a world where gay couples look married, act married, talk married, raise kids together, and are increasingly accepted as married, the best way to preserve marriage’s normative status is to bring gay couples inside the tent. Failing to do so, over time, will tar marriage as discriminatory, legitimize co-habitation and other kinds of non-marriage, and turn every successful gay couple into a cultural advertisement for the expendability of matrimony.

Blankenhorn clearly disagrees. Our disagreement over how gay marriage will affect marriage’s normative status, however, is well-plowed ground. So I’ll move along to what Blankenhorn rightly considers his deeper and more important arguments, which are about the nature of marriage itself. Near the beginning of his book, Blankenhorn calls childrearing (by which he means the rearing of children by their biological parents) "probably the single most important social need that marriage is designed to meet, but there are numerous others as well." Two pages later, however, he makes a more unequivocal statement: "Without children, marriage as an institution makes little sense." Though he regularly uses qualifiers, it quickly becomes clear that, in practice, the unqualified statement is much closer to his view.

Blankenhorn succeeds in showing that binding fathers and mothers to their biological children is a core purpose of marriage, and more power to him for that. But the logic of his argument is that binding fathers and mothers to their biological children is the only purpose that has any compelling claim on society, and that allowing marriage to serve any other purpose hurts children by pushing them to the sidelines. From a purpose to the purpose is a long leap, and one that leaves the public far behind. Blankenhorn himself cites a poll showing that 13 percent of Americans say "promot[ing] the happiness and wellbeing of the married individuals" is the "more important characteristic of a good marriage," and 10 percent choose "produces children who are well-adjusted and who will become good citizens," but three-quarters say: "The two are about equally important." In other words, the public believes that a good marriage is good for adults and good for children, and that there is no conflict. And the public is obviously right. Marriage has more than one essential public purpose. Providing a healthy and secure environment for the rearing of children (biological or adoptive) is certainly one of them (and, of course, many gay couples are raising children), but at least three others, in my view, compel respect: providing a transition to stable domestic life for young adults (especially men), providing a safe harbor for sex, and providing lifelong caregivers.

Still others can be found in a 2000 document called The Marriage Movement: A Statement of Principles. In a section headed "What Is Marriage?" the manifesto declares that "marriage has at least six important dimensions": it is a legal contract, a financial partnership, a sacred promise, a sexual union, a personal bond, and a family-making bond. "In all these ways," the statement continues, "marriage is a productive institution, not a consumer good."