Topics Logic

Question During a 'debate' with a friend about same sex marriage, he raised the issue of marriage being 'by definition union between a man and a woman', and appeared to hold that this was grounds for rejecting same sex marriage. My question does not relate to the ethics surrounding the issue, but rather to the fallacy I thought he had commited in saying this. It seemed to me as if he was stating the conclusion of an argument that had not been argued (at least, not by us either at or prior to that time) namely whether marriage is, in fact, the union as mentioned - is this what is known as 'begging the question' (i.e., stating a point that remains to be proven as foundation for another conclusion)? If not, then what is the formal term for this fallacy (if it is, indeed, fallacious)?

Accepted:
June 8, 2006

Comments

There certainly is a fallacy here, but I don't know what it should be called. In the end, perhaps it is a simple fallacy of equivocation: an equivocation between two senses of the word "definition".

Philosophers since Aristotle (at least) have distinguished two types of definitions: definitions of words and definitions of things. Personally, I find the latter notion hard to understand, but the idea is that a definition of a thing tells you what it really is. So gold, for example, might be defined as that element that has atomic number 79. This is very different from saying that the word "gold" is defined this way. The word "gold" could not have been defined that way before the establishment of modern chemistry, but, nonetheless, the true definition of gold—what gold really is—is: the element with atomic number 79. You cannot, therefore, find out what the real definition of a thing is by consulting a dictionary: That will tell you only how the word is defined. What the "real definition" of a thing is may be unknown to us, and the "verbal definition" we employ of the word that denotes that thing may therefore be incorrect. Even the venerable Kant seems not to have understood this point. He claims that it is analytic that gold is a yellow metal: It follows from the definition. But, as Saul Kripke pointed out, we can well imagine finding out that gold was not yellow or even that it was not, in fact, a metal. It is a substantive empirical question what the real definition of gold is.

Now: Is your friend's claim that marriage is "by definition a union between a man and a woman" a claim about how the word is defined or is it a claim about how the thing is defined? My sense is that the argument your friend is making waffles between these two. One often hears people say: Just look it up in the dictionary! Well, if you do, then perhaps you find something about a man and a woman. But that only tells us how the word is defined, and, as just said, the definition of a word can just be wrong in the sense that it conflicts with the correct real definition. And that, it seems to me, is what many proponents of gay marriage (including the justices on the Supreme Judicial Court of the Commonwealth of Massachusetts) believe: It turns out, despite what has long been thought, that marriage isn't necessarily a union between a man and a woman. This view cannot be refuted by a dictionary or even a history book: If, as has sometimes been claimed, every recorded civilization has regarded marriage as a relationship between a man and a woman, well, then every recorded civilization has been wrong about what marriage really is. Maybe that's surprising, but it's hardly inconceivable, especially when one thinks about how asymmetric marriage traditionally was and, indeed, continues to be in many parts of the world.

The SJC was, of course, concerned with a question about the laws that govern marriage and so were concerned with a question about a certain legal institution. And what they asked themselves was this: When one looks at the laws that actually govern marriage, are there any such laws that actually require one of the parties to be male and one of them to be female? The question is not simply whether words like "man" and "woman" get used in formulating the laws, but whether those words are doing any real work, whether there is anything in the law that would be lost if the term "person (of consenting age)" were substituted for both. And their answer was that, no, there are not. That was not always so. Not very long ago marriage was essentially a form of ownership: The husband (all but?) owned the wife. At that time, then, certainly there were aspects of the law that would have made no sense had we made the replacement just mentioned: How would you know who owned whom? But, over the last fifty years or so, marriage has come to be regarded, by many people, anyway, as an essentially symmetric relationship, and that is reflected in the laws that govern it. In so far as it is not so, that would be widely regarded, and rightly so in my view, as unjustifiable discrimination on the basis of gender. Hence, at the very least, swapping the terms "man" and "woman", throughout the marriage laws, should make no difference.

Of course, it does not follow that replacing "man" and "woman" by "person" should make no difference: Perhaps there are laws that, while not being asymmetric as regards male and female, nonetheless require the parties to the marriage to be of different genders. But the Court held that there are no reasonable such laws. Of course one could simply add one, as many states have: Simply make it a law that the parties must be of different genders. Here, however, the Court held, such a law serves no reasonable purpose, any more than would a law that required the man to be taller than the woman.

If we want to understand what marriage really is, then we need to ask ourselves similar questions: When we think about what really makes, or should make, a relationship a marriage, or if we think about what makes for a good, thriving, and healthy marriage, is there anything there that involves the parties being of different genders? Perhaps many people do instinctively regard the parties as of different genders, but of course that is not the question. The question is whether there is anything in the very nature of such relationships that require the parties to be of different genders.

It is, of course, there that the debate threatens to become intractable. The only argument against gay marriage that I know that is not based upon simple prejudice is this one: There is a reasonable purpose to laws requiring the parties to be of different genders, one that is connected with child rearing. To my mind, this argument assumes a connection between marriage and child rearing that is not in evidence. I do not regard childless marriages as any less marriages and, more importantly, gay couples have been raising children for a long time, will continue to do so whatever the laws governing marriage, and gay couples are, at least in Massachusetts, legally permitted to adopt. It can be argued, indeed, that if we want to support all children, including children of gay parents, then we should allow gay parents to marry. But the question is, as I said, a difficult one to resolve.

One of the more amusing things about this debate is that opponents of gay marriage are often heard to make the argument you mentioned and then say, in the next breath, that, if gay marriage is permitted, what is to stop the courts from permitting polygamy? And I have even heard it said that marriage is, by definition, a union between one man and one woman. Such people have obviously not read their Bibles. In any event, the question here is whether, as a matter of real definition, marriage is a union of two people or whether a marriage might be a relationship among more than two people or, again, whether one can enter into such a relationship with more than one person. Certainly the reasoning the SJC used in Massachusetts can not be used to argue that prohibitions against polygamy are discriminatory. It seems obvious that plenty of the laws that govern marriage—for example, laws concerning divorce—do essentially assume that a marriage is a relationship between two people or assume—consider laws governing marital property—that one has such a relationship with only one other person.

The argument from definition was actually put forward in the Canadian courts a few years ago, when they were debating the question of whether gay marriage should legalised. It was argued (roughly speaking) that it makes no sense to try to legalise gay marriage, because, by definition, marriage is a union between people of different sexes. So - it was argued - trying to lagalise gay marriage would be like trying to legalise married bachelors or male vixens. The argument was rejected. One unsubtle problem with the argument is that the premise seems to be false. Here is a quote from the Random House Dictionary (1983 edition):

"Marriage: a relationship in which two people have pledged themselves to each other in the manner of a husband and wife, without legal sanction: trial marriage, homosexual marriage"

It seems to me that this is a common occurrence: people fairly often try to argue for some conclusion or other by appealing to the definition of some key term, when they actually don't know the definition (or definitions). And they get it wrong. Maybe we should call this 'the fallacy of thinking you know what's in the dictionary'.

Of course, one might try to get around the problem in the 'marriage' case by appealing to something like the 'normal' or 'standard' definition, which does indeed require that people married to each other be of different sexes. But then the argument commits something in the ballpark of the fallacy of trading on ambiguity, because people arguing in favour of gay marriage obviously use 'marriage' to mean something like 'a relationship in which two people have pledged themselves to each other in the manner of a husband and wife'. They obviously are not using the term in a way which rules out the possibility of gay marriage by definition!