A press release came over my transom today that troubled me, about a novel test of the "snowball Earth" hypothesis. The research in question helps clarify an outstanding puzzle about how snowball episodes end, showing that one of the guesses is probably incorrect. According to this paper, based on carbon-13 abundances, snowball episodes could not have ended with a large belch of methane that made the planet quickly warm up. The signs point instead toward a hydrothermal mechanism. It seems that there were no methane-oxidizing bacteria around at the time to make the scenario work.
That's all good; that's how science works. We put out ideas, and most of them are wrong, even those that manage to get published. The result is an overarching story about the world that becomes ever more true. But I really don't like the headline of the press release: "Caltech-led team debunks theory on end of 'Snowball Earth' ice age."
Scientists don't "debunk" each other. Debunking presupposes that something is bunk to begin with. It's a prejudicial word for an ordinary development in the (usually) polite debate that constitutes science.
Very few developments in science can be construed as "debunking." There's the notorious case of Piltdown Man, where outright fraud was committed. And that's about it. You might consider the acrimonious case of the small-comets hypothesis in the 1990s, which involved no fraud, merely an annoyingly persistent advocate of a provocative idea. A more recent example might be the Younger Dryas impact hypothesis, a dramatic scenario that has met a storm of opposition with a strong hint of exasperation. A press release last year used the D-word in that case, too. If I had noticed at the time, I would have protested, because the people involved are all legitimate scientists who happen to be in the grip of a compelling idea.
There is not a thing wrong with that. Press officers, please watch your language.