Well, as usual we disagree. Etymologiacally, there may only be two origins
for the set set. But synchronically, I am set on there being quite a larger
and more scattered set set before us. Now I am set to hear reasons why they
should all still be concidered just two groupings, but that will have to
wait till after the Jell-o sets and I have set my hair.
Put, determine, specify, confine, establish, prepare, move, locate, go down,
staging, arrange, adapt, plant, jell, a grouping, put back together, start
-- and that's bound to be an incomplete list, may all be traceable to only
two sources, but I just don't agree that hey are, at this point in time, in
real usage, reducable to only two lumps without recourse to etymologies that
are meaningless to the man in the street. They do not "feel" like the same
words any more.
On Tue, Mar 8, 2011 at 11:04 AM, Alex Fink <[log in to unmask]> wrote:
> On Mon, 7 Mar 2011 16:00:36 -0600, Adam Walker <[log in to unmask]> wrote:
> >Grab an English dictionary and look up the word *set*. That will give you
> >some idea of how far homophony can go for a single word.
> Both you and Pete have offered "set" now, and I don't think it's a good
> example of _homonymy_. Of _polysemy_, of having a very large semantic
> spread, it certainly is. But if homonymy is to mean something distinct
> polysemy, I'd not attribute it to "set". To my mind the vast majority of
> the senses of "set" are specialisations of maybe two main senses, a verb
> 'make to sit, make to stay in some condition' (and its formally identical
> participle and zero-derived nominalisations) and a noun 'collection of
> similar things'; since those two clusters are still discernable and the
> central senses still usable nonspecifically, the polysemous analysis
> me as a better one inside them.
> Admittedly I have a bit of an etymological bias here; there are about two
> etymologies of "set", one as the causative of "sit" with various lexical
> category conversions, and one < French < Latin _secta_ which fell together
> and was confounded with it.
> There are objective tests for polysemy vs. homonymy that one might deploy
> here, sentences that use a word two ways that are germane in the former
> but odd in the latter, but unfortunately I don't remember them.
> IMO a better example of a family of English homonyms (and not necessarily
> the best one but just the first one to come to mind) would be something
> /bej/, whose senses 'laurel', 'mostly-surrounded body of water; recess',
> 'bark', 'reddish-brown', and 'Turkish governor' if that counts as
> are all clearly separated each from the others.
> Anyway, to the original question, I agree that a language could probably
> by without severe difficulty with an average of three words per form. If
> you're after naturalism, though, they shouldn't be distributed too evenly
> some forms should have just one word, if not zero, while others have four
> five or more.