Chris said to me this morning that he thinks linguists are fighting a losing battle.
I asked what he meant.
It turns out he was referring to a few recent posts on http://www.languagelog.com
dedicated to discussing the minutiae of some overheard/seen-in-print construction, such as "I'm feeling all Olympic-y." His observation was that language users innovate faster than linguists' ability to track and classify the innovations. And you know, that's true. For every construction that someone happens to observe, how many more are there that people fail to notice? Amd multiply that over the world's languages and the problem gets very big very fast.
So this started me asking him what battle he thinks linguists are fighting. Because in fact linguistics isn't much different from physics, or chemistry, or any other science in this regard: the world of observable phenomena is bigger than scientists have the resources to observe. But that makes you wonder: is the point of science to observe everything that can be observed? I really don't think that it is.
I think the point of science is to take a consistent data set and explain it. You get new data, you change your explanation. But so as long as established principles are respected in the way some newly observed phenomenon behaves, that that newly observed piece of data really doesn't make a difference. Once it gets to a certain point, science looks for counterexamples, not for examples.
So if the battle of linguists is to explain the principles governing language structure and use, then it's perfectly okay that language changes faster than we can keep up with the changes. But so long as the principles guiding change consistently apply -- and unless we're talking on an evolutionary timescale here, I think the field has more or less decided that they do -- then no individual change matters.