My point of tradition isn't that it is bad, but rather that it is an invalid justification.
The notion that someone in the past would be regarded as an Emperor rather than as a King does not dictate how we must use words today. A senator in Ancient Rome is very different from a senator in the US Congress today, but we call both senators. Marriage functions in the same way, as it began as a property arrangement (with women as property) and has evolved (in some countries) to entail a consensual relationship that affirms both men and women as free people. Given how much marriage has changed and how much it varies between countries (some of which already call same-sex marriage "marriage"), it is a simple step for a country whose forms asks for "Wife" and "Husband" to instead say "Partner A" and "Partner B."
My Christian denomination examples weren't about religion -- they were merely to illustrate that we don't all have to recognize someone's marriage. If you don't think "gay marriage" is really "marriage," no one will force you to call it that.
The olympics example has a bona fide justification: there are tendencies concerning the birth sexes that in many sports would effectively exclude one birth sex from competition at the highest level. It is not the sort of thing I'm thrilled about, but there's a point to it, and the olympics gives each division an equally wordy title (as men compete in Men's, and women compete in Women's -- this is contrast to calling one group's marriage "gay marriage" but letting the other get by with just "marriage"). There is no substantial bona fide justification for a separate marriage name.
As for something that concerns definitions, sexual orientation (the attraction) is different from sexual relationships. I should have clarified my previous argument more to that effect, but basically "gay" in more accepting circles is reserved for sex-based attraction -- and labeling the sex-pairing of a relationship is basically pointless (if someone is acting on their attractions in a responsible way, that's good, and if they're acting on attractions irresponsibly or are seeking out sexual relationships without attraction, that's bad -- it doesn't need to be made more complicated than that!) As far as gay people are concerned, they have reproduced quite a bit! Historically, gay people often had other-sex spouses/partners and still had children with them; nowadays, in vitro fertilization allows one parent of a same-sex couple to pass on genetic material.
Hopes of finding "the gay gene" are misplaced, because there are probably many genes that play a role, and there can be random factors at work, too (identical twins are more likely to have the same sexual orientation, but it's basically a 50/50 proposition that a gay man's twin is also gay). Even if we find gene(s) responsible for determining sexual orientation, it doesn't offer much practical value except to people who want to "cure" gay, which is a nuclear ethical issue.
As to the idea of judging minority groups until their difference is scientifically proven to be okay, that's not science works. Statistical hypothesis testing works from the direction of assuming equality (that men and women are equal on a certain dimension, that the weight of people on a diet is the same before compared to after), and a study bears the burden of proof to show some claim of difference is true. Even when a study shows a difference (with a certain level of confidence), it must be replicated by peers in the field (to make sure there wasn't a false positive or a fabricated study). Usually no difference will emerge, although it is possible for the majority or the minority to score higher on some dimension. But even then, science is limited with respect to many dimensions, including sexual orientation, because it can't perform random assignment. A correlational study can only tease out tendencies, rather than prove a cause-and-effect relationship. If one aspires to come up with as good of a comparison as possible, though, we have to remove confounding variables: to optimize our measurements, we need to get rid of any factors possible that would skew results, and that means doing away with inequality and discrimination *before* conducting the studies (rather than afterward). In fact, I do care quite dearly about science, and I am quite eager to level the playing field so that we might actually be able to observe and learn about variation as it occurs naturally (not that we get to see causality, but we get the best data possible) -- without equal treatment, the effects of inequality will tip the scales against the minority and provide the false justification some people seek to discriminate and hold people down. Yet even when there are differences, each person remains their own, and each individual shall be dealt with according to their character or ability -- after all, there may come a day when you find yourself in an "inferior" minority or "inferior" majority, and you will want to be judged on your own merits.
< Message edited by Kaelin -- 8/2/2012 13:48:53 >