with the "fittest" making it to the next shakeout.
Whilst I agree with the overall tenor of your post, wseiber, I don't think language really has "shakeouts."
Surely it is a continuous process, not necessarily a smooth one but having peaks and troughs. Certain events will promote a good deal of linguistic activity which may well be followed by a "quiet" period, but entry and exit doesn't ever completly stop.
But I do agree that there is a place in the language for archaisms, and tend to disagree with Wiliam's proposition that words "die." Like King Arthur, they lie beneath the linguistic hills, to rise and save us in our hour of need.
I suppose it might be the case in a language that has no, or only a recent, written tradition. Perhaps someone with more knowledge than I (not difficult to find) can inform us on this point.
One of my own favourite archaisms is "yclept" - not that I very often find an opportunity for its use.
