Well, I think that linguistics as you have described it here is a poor analogy. Natural selection does not have a goal(supposedly). Language has a goal, namely communication.
I think you are missing the gist of the analogy. Natural selection is a mechanism that affects biological organisms. In the language analogy the analogues are biological organisms and languages not languages and natural selection.
If language has the goal of communication then biological organisms have the goal of reproduction. Language does not change through selection as much as drift but since drift also affects evolution then the analogy is accurate.
AndrewC, an interesting and stimulating point, for which I thank you. Now that we are in the 'Smoky Backroom,' perhaps I may be permitted to expand a bit on my original--and necessarily truncated--post.
I think it an error to assume 'communication' as some sort of teleological 'goal' of language, for several reasons. Of course, 'communication' is its primary function, and no language that did not fulfill the basic function of communication could arise, or survive if it could (indeed, can scarcely be imagined). But that basic functional requirement can be and is performed by a number of other channels in nature, from pheromones and plummage to gestures and auditory signals. A pack of wild dogs can communicate perfectly effectively while conducting a co-ordinated hunt (but no one supposes they discuss Heidegger and Kant in their lair). In other words, communication may be a functional requirement of a number of organisms, and it could be stated language is one of the many channels that have evolved that meet this requirement, but that does not explain why language should continue to evolve once that 'goal' of functional fulfillment had been achieved. I think that is Doctor Stochastic's point in post #1117:
The goal of language isn't to change things like Latin into French.
Precisely. If one states "Language has a goal, namely communication," then we have failed to explain the observed phenomenon of the development of new languages from previous ones: Latin is not one jot less 'communicative' than Italian, French, Spanish, Portuguese, Romanian etc.
But I think your point well illustrates, by analogy, the teleological fallacies I have often seen applied as crticism to biological ToE. There are many variants, e.g. "what good would 'half-an-eye' be" (assuming a "goal" of eyes as we know them), or "the mathematical odds are astronomical against homo sapiens evolving by gradual stages of variations in allele frequency" (assuming a "goal" of mankind). Such objections can and are easily and decisively answered--but they persist, I suspect, because of the teleogical fallacy.
[AndrewC] Well, I think that linguistics as you have described it here is a poor analogy.
I'm not endeavouring to argue one comprehensive analogy, simply to point out several interesting (to me, at least) points of similarity--with my previously stated caveat firmly in place (to wit, no analogy here is intended to argue for biological ToE, but rather to illustrate examples of similar processes elsewhere to better illustrate some of the weaknesses in some of the objections to biological ToE):
[1] One possible analogy here is between speciation (which was, after all, the phenomenon Darwin set out to understand) and the emergence of new languages: both entail the issue of 'transitional forms.' In the language analogy, the question might be stated, at what precise date did English arise? Or, who was the very first person to speak English? Nonsense questions, of course. Parts of the house in which I currently dwell are a little over 400 years old; were it possible (let's dream up a convenient time machine here) to travel back and meet the builders of this home, we could speak with them without too much difficulty, their English sounding somewhat provincial and 'Shakesperian' to us, ours sounding a tad peculiar to them in about the same measure. Just down the road is The Lamb and Flag, which dates from circa 1320; it would be more of a stretch to understand the Chaucerian Middle English of its builders, though we'd eventually get our ear in and manage ok. And the local parish church was founded before the Conquest, in 948: understanding the Old English of its founders would be very difficult indeed. And so on, back to the first Anglo-Saxon settlers--some of whose relatives who did not migrate, but spoke the same language, are the ancestors of today's speakers of German.
This process seems to me very closely analogous to the biological concept of speciation. There are a number of other analogies that can be drawn from linguistics, and I'd be happy to kick this or others around if anyone here is interested--but at the moment, my lunch hour is drawing to a close and, as one of the Roman inhabitants of this village might have said, labor me vocat.