The Austin Institute’s recently released Relationships in America survey is a must-read for anyone interested in surveying American attitudes about family, marriage, religion, and sexual ethics. Containing good and bad news both—some of which Mark Regnerus, senior fellow at the Austin Institute, documented last week at Public Discourse—the report is also structured in a maximally digestible manner; one can simply click through the table of contents to particular questions that interest him.

It’s an excellent study and highly commendable. One of its questions, or rather statements, is badly flawed, however: the statement about same-sex marriage. The figures and description of the respondents’ views on same-sex marriage are gathered on a page titled “Should same-sex marriage be legal?” But the statement with which respondents are instructed to agree, disagree, or indicate neutral feelings toward, is worded thus: “It should be legal for gays and lesbians to marry in America.”

This statement is problematic for two reasons. First, it obscures a crucial dimension of present debates over the meaning of marriage by locating the discussion in an alleged correspondence between the civil right to marry and one’s ability to enjoy that right, or not, due to one’s sexual attractions. But understanding the marriage question this way hinders one’s thinking clearly about it. It should be manifest to all Americans that no adult’s sexual inclinations ought to disqualify him or her from being able to civilly marry in America. The conflation of the question of “whether gays and lesbians should be able [it be legal for them] to marry in America” with the question of what marriage is, is deadly for careful thought.

Rather, the appropriate question is: “Should same-sex unions be legally recognized as marriages?” As Aaron Taylor has pointed out, drawing such linguistic and conceptual distinctions doesn’t get to the bottom of what explains the impetus behind the same-sex marriage movement. Such distinctions must be drawn nevertheless for the sake of sound thinking, which will readily recognize that while a person’s sexual inclinations may render marriage an unattractive—though still intelligible and choiceworthy—option for full-being, they do not come to bear upon one’s ability to enter a civil marriage.

Second, the same-sex marriage statement is problematic because its use of “gays and lesbians” suggests what Michael Hannon has called “orientation essentialism,” even as the survey elsewhere corroborates the suggestion that sexual attraction and disposition are fluid and mutable rather than fixed, and gradated rather than discrete, especially for females (Regnerus too cautions against “biological essentialism” in his Public Discourse essay).

One cannot speak unqualifiedly of a person’s being “gay or lesbian” in the same way that one would speak of one’s being Italian-American, or female. Say that a woman, Jane, identifies as lesbian during her college and immediate post-collegiate years, but in her late 20s identifies as heterosexual—say, to a degree such that she checks the “mostly heterosexual” box in response to a survey statement like the one posed in Relationships in America (such a transition was not uncommon among female respondents). Do 31 percent of American adults believe that Jane should be denied civil marriage to a man should she seek it as a 22-year-old, but permitted civil marriage to the same man as a 30-year-old, simply because she is “lesbian” at 22 and (mostly) “straight” at 30?

Almost certainly not. Most people would sense how preposterous the suggestion is, and would reply that in responding to the question they really had same-sex marriage in mind. Yet for many, the response to the survey statement as stands would be "yes," but the response to a "same-sex unions as marriages" statement would be "no." And this reveals why the survey’s wording here is unfortunate: Not necessarily because wording it soundly would have altered these responses much (if at all), but because when the rhetorical edifice of conflation becomes the linguistic paradigm for a complex and controversial discussion, that paradigm metastasizes over time and obscures concepts that are distinct, and are even understood rightly to be distinct by some (or many) who see no harm in indulging the paradigm for precisely that reason.

One might object that I wouldn’t make the same point if the survey statement were, “it should be legal for young children to marry,” even though “child-ness,” as well, is neither discrete nor immutable. But being a child is relevant to one’s ability to enter into marriage, which is a consenting (not to mention sexual) union—a conviction shared by all in the marriage debate.

This is exactly the point: Homosexual persons, unlike children, are capable of and therefore eligible for civil marriage because one’s being homosexual, unlike one’s being a child, doesn’t preclude one from being able to marry (though it may render marriage unattractive in the aforementioned sense). So in wording the statement thus, the survey already communicates the suggestion that being “gay or lesbian” effects one’s capacity to marry, and that there is some legally significant correspondence between sexual attraction and ability to civilly marry that is different in kind from both 1) an analogous potential correspondence between, say, being Italian-American and being able to civilly marry, and 2) other mutable and gradated characteristics, such as one’s being a child, that do bear upon one’s capacity to civilly marry.

Of course, most people understand by the term “gay marriage” same-sex marriage, but the concepts are distinct. What could a “gay marriage” be except a marriage in which one or both spouses self-identify as gay? But such a marriage needn’t be a same-sex marriage, and nobody opposed to same-sex marriage should be opposed to allowing “gays and lesbians” to marry, and even to allowing them to marry each other, for even these marriages needn’t be same-sex marriages.

The coupling of the question of same-sex marriage with the sexual identity of a small swath of Americans has proven unhelpful for understanding marriage as the comprehensive union of man and woman. Where marriage is concerned, we should avail ourselves of every opportunity to speak clearly and to correct unhelpful speech, just as the pro-life movement should take care (and largely has taken care) to speak of the child in the womb as a real child—for example, by men speaking of themselves as fathers during their wives’ first pregnancy, not simply upon the birth of their firstborn.

The effort to uncouple these distinct concepts when discussing marriage is rendered less effective by an accompanying disregard for the concrete situations of homosexual persons, but that only indicates that speaking precisely about same-sex marriage isn’t sufficient to making a convincing case for marriage as comprehensive union. Dislodging faulty linguistic paradigms and their vague rhetoric, however, may well be necessary in a discussion being largely dominated by memetic thought.