Markos wrote:That was my original point. The whole thing is mere speculation. There is simply too much we don't know. We don't know anything, really, about the ancestors of PIE. We don't know how old such languges are. There is speculation that there was a common ancestor of PIE and Proto-Semitic. You
probably know about Hebrew KEREN, "horn" which is presumed to be a PIE loanword, but we don't know for sure.
The question is not only the speculative basis of this "science" but its value. Does it help you understand either word if you discover whether they are related or not? I would say etymology almost always leads to less understanding, not more. Does it help you REMEMBER words to discover cognates? Absolutely. Then it matters not whether the etymology be true or false. If the mixing in horns image helps you remember the words, go for it. Argue thus. No one will be able to refute you. Since we know absolutely nothing about the people who may have spoken the language ten thousand, fifty thouand years ago, you will be on safe ground asserting that these words were at one time related.
You make some excellent points, some of which I very much agree with. But I think several different issues have gotten muddled up in this discussion, so let me try to separate them:
First, the question of whether historical linguistics is a science: it definitely is a science, as practiced by real linguists
(as opposed to people speculating on message boards). But the key is that you can't push the etymologies farther back than the evidence can support. Beyond that point, it really is
just idle speculation, impossible to prove or disprove. This is true of any inquiry of a historical nature. Whether you're investigating the Vietnam War or ancient Greek mystery religions or Indo-European migration patterns, you can be scientific up to a point; but the farther you go beyond that point, the less scientific it becomes. Incidentally, historical linguistics seems particularly speculative if you only read about Indo-European stuff; but there's been a lot of excellent work done on, say, the spread of Bantu languages in Africa, or various language families of Southeast Asia; if you read a little bit of that, you see that it's hardly speculative or controversial at all.
But I completely agree with you in questioning whether etymologies help you understand Greek - in fact, that has long been one of my own pet peeves. People don't realize that when you delve into that stuff, you're now studying Proto-Indo-European, not
Greek. It's a perfectly valuable area of inquiry, but it's something different
from studying how the Greek language worked in the historical period. (And the same goes for Latin, obviously.) The question I like to ask is, "Does Proto-Indo-European help you understand Shakespeare?" (No.) Then why would it help you understand Sophocles? In both cases, PIE is more than 5,000 years removed from the period you're actually interested in. The main instance where PIE is really helpful is as a last resort, when we can't figure out the meaning of a word from any other evidence.
So in summary, I think the most helpful works on Greek linguistics are those that take a synchronic perspective; for instance, Rijksbaron, The Syntax and Semantics of the Verb in Classical Greek.
EDIT: Rijksbaron deals with syntax, not lexical meaning, but I mention him as an example of the synchronic approach. Dickey, Greek Forms of Address
, is an example of how one can assess shades of meaning based on how the words were actually used in extant texts.