<<I think you are wrong, Steve. A language often borrows a redundant word,
and then employs it for a particular stylistic effect or creates a new
semantic niche for it, relocating its resources, as it were. You project the
new distinction on the pre-borrowing stage of the language, suggesting a
pre-existing gap waiting for a loan to come along and fill it. >>
Try to look at it the other way around. From the time the first word was
spoken there's been an infinite number of pre-existing gaps. When another
language has attempted to fill one of those gaps, and my language hasn't, I
have three choices. And one of them is borrowing that word from the other
Now, I may not have preceived a difference between crimsom and red before you
and your language brought it up. Was there a gap there in my language before
you brought it up? The answer is I guess only if the difference is of
consequence for me. Otherwise, there was really no gap to fill. And I might
as well just continue using red for all reds.
When I go around calling red, crimsom, to my fellow speakers to improve my
status, that's a prestige borrowing. I do run the severe risk however of
them not knowing what I am talking about. If a borrowing is only for
prestige and has no other function in the language - doesn't really also fill
a another gap - it refers to nothing and is gibberish.
And if I want to share in the prestige of your language among your speakers,
using the word "crimsom" alone will do me very little good. Here you are not
talking about borrowing, but learning a new language - a very different way
of acquiring prestige through language. (As opposed to acquiring it by
buying a Rolex watch - which I'm told are a lot more effective at getting
quick status than walking around saying "chateaubriand" to the guys down at
the body shop.)
<<However, a new word often ousts an older (near-)synonym entirely -- why, if
_both_ were allegedly needed in the first place, if I understand you
Why? Because things change. The old need for two different words may have
disappeared. (And BTW "near-synonym" may be presumptive. The near
difference lexically may be very consequential in the real world. Like the
difference between a snake and a poisonous snake.)
In perceptual psychology, there are two very basic and discrete actions or
behaviours. One is discrimination. The other is generalization. If there
is a difference between just being an eme and an uncle, we might begin to use
two different words to describe that difference. If that difference over
time disappears, the two different words MAY no longer refer to any
difference we can think of. Logically, one or the other word may fall out of
use. Or go on to some new use (e.g, a quaint expression.) The other word
may be generalized to remove any trace of the old difference - uncle becomes
any uncle, not just a mother's brother.
<<You claim, for example, that the meaning of <eme> (OE e:am) was different
from that of <uncle>, the latter having special legal connotations, so both
were needed to express different semantic shades. So far so good. But the
speakers of Middle English can't have been too pedantic about the
distinction: <uncle> was used in the sense 'one's father brother' or 'an
aunt's husband' already in Middle English times, replacing <eme> completely
as a kinship term.>>
Canon law came to England with Christianity. My Black's Law Dictionary
tells me that "avunculus" was not just a description, but a term of specific
legal status. I have every reason to believe that is how it was "borrowed"
into English. How it generalized beyond that meaning I can't say, but it's
VERY important to remember that that legal status was ALSO changing over that
time, as canon law receded in importance.
Saying that the English could not have been too "pedantic" about the word's
use is to miss the real world consequence of that legal term. The courts
weren't simply being pedantic, they had the power to act on it. Whether they
decided you were an uncle or not could be of significant importance.
Apparently, <eme> had no such legal consequence. That alone could explain
why uncle came into use and eme fell out of use.
This is a functionalist approach. What were the consequences of using the
word uncle versus eme. How did it change? You probably have to know more
about the old inheritance laws than about the prestige of Latin to account
for those words. And of course every word has its own history.
<<Why didn't it happen the other way? Why didn't <eme> add a legal touch to
its meaning and kick the newcomer out?>>
Well, you know the answer. Words are tools. <Eme> couldn't do anything on
its own. And claiming you were an <eme> got you nowhere in court and so
perhaps also it got you no free beer in anticipation of an inheritance at the
local pub. The law made <uncle> a term of consequence. What happened to
<eme>? It seems to have lingered on as a term of endearment dialectially or
as a restoration word in titles. Apparently endearment is about as
functional and enduring as prestige when it comes to lexical matters. And so
it passed on.
I'll try to finish this reply soon.