[UPDATE: I'm afraid I misentered the data in the table when I first posted this; please see the revised information. The analysis remains correct — but the data now matches it.]

AEI has a useful compilation of poll data on attitudes about homosexuality and related topics. Here's one particular interesting item, from p. 3, reporting on what percentage of respondents to a National Opinion Research Center survey said that homosexual sexual relations are "always wrong" (as opposed to almost always wrong, only sometimes wrong, or not wrong at all):

Age	1973	2002
Total	73	55
18-29	56	48
30-44	74	48
45-59	75	55
60 and over	89	68

There are three obviously striking items here: First, 55% of Americans still think homosexual sexual relations are always wrong. Second, public attitudes have shifted considerably (by 18%) on this in the last thirty years. Third, younger people have always been less likely than older people to say that homosexual sexual relations are always wrong.

But the fourth thing may be less obvious, and yet I think just as important: If you look at the 18-29 age range in 1973 and the 45-59 range in 2002, which represent pretty much the same people (18-29-year-olds in 1973 would be 47-58 in 2002), the percentages are statistically identical, 56% and 55%. If you look at 30-59-year-olds in 1973 and 60-and-over in 2002, which should also be pretty much the same people (since only a small fraction of the 60-plus in 1973 survive in 2002), the change is from 74-75% to 68%, a significant change but a relatively small one.

So the primary reason for the 18% change does not seem to be that adults are hearing more about gay rights claims, seeing more out-of-the-closet gays at work or in social circles, and thus changing their views. There seems to be a modest such effect among those who were over 30 in 1973, but only a modest one.

Rather, the main change is in the views of the new generations (the ones who are now 18-44). And this change started with those who came of age in the 1960s and early 1970s (note that the "always wrong" figure has declined only from 56% to 48% from 1973 to 2002), and therefore seems likely to have been caused by the Sexual Revolution, which predated 1973, more than by the gay rights movement.

UPDATE: Many thanks to reader Marco Parillo, who gave the quote that this reminded me of, but the details and the author of which I couldn't remember: It's known as Planck's Principle (after it's author, physicist Max Planck), and it is that

A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die and a new generation grows up that is familiar with it.

Naturally, the analogy is not intended to be perfect, and those who prefer to see tolerance of homosexuality as being more akin to new error than to new truth may feel free to replace "truth" with "error." The important point here isn't about truth as such, but rather about how public opinion changes.