Re: The monumental stupidity of PIE theorists further illustrated


"Nathan Sanders" <nsanders@xxxxxxxxxxxx> ...
"John Atkinson" <johnacko@xxxxxxxxxxx> wrote:
Nathan Sanders wrote:
> "John Atkinson" <johnacko@xxxxxxxxxxx> wrote:
>> Nathan Sanders wrote:
>>> Jack Campin - bogus address <bogus@xxxxxxxxxxxxxxxx> wrote:
>>>
>>>>>> Is there evidence of sound change be conditioned by
>>>>>> part-of-speech?
>>>>> I'm not quite sure what you mean, but the answer is probably >>>>> no.
>>>>> If two words are homophonous, they will undergo the same sound
>>>>> changes with the same results, regardless of their parts of
>>>>> speech. (Keep in mind that I'm talking about systematic change
>>>>> here. Sporadic change is always possible, and it's one of the >>>>> few
>>>>> ways homophones can diverge.)
>>>>
>>>> Can't word order affect that? If some particular part of speech
>>>> is always sentence-final, won't it be subject to different sound
>>>> changes than a homophone which is aways initial or medial? (I'm
>>>> not pretending to have an example).
>>>
>>> Note that occurring phrase-finally or phrase-medially is still
>>> technically a phonetic environment, since you have the
>>> absence/presence of some sound following the word.
>>>
>>> But yes, this sort of thing could be a way of distinguishing two
>>> homophones (an analogous example is "to" versus "two"; the former >>> is
>>> nearly always unstressed, leading to reduction of the vowel, >>> while
>>> the latter is nearly always stressed, preserving the vowel >>> quality).
>>
>> A more convincing example (at least in my dialect) is the >> difference
>> in the final consonant of <of> and <off>, which were originally >> the
>> same word -- when used adverbly <off> it was typically >> phrase-final,
>> thus had the voiceless allophone of the OE fricative, while
>> preposition <of> was frequently followed by a vowel or voiced
>> consonant, which induced the voiced allophone.
>
> Right, this is the sort of thing I'm talking about. It's possible > for
> homophones, by the nature of their meanings or lexical categories, > to
> appear predominately in different phonetic environments, and
> eventually cease being homophonous via leveling of the altered
> pronunciation of one of them across all environments (one factor in
> language change is allomorphy avoidance, which can sometimes cause > one
> of many allomorphs to be lexicalized, wiping out many/all of the > other
> allomorphs).
>
> But it is still the phonetic environment that triggers (or blocks) > the
> relevant sound change. "Of" didn't undergo voicing because it was > a
> preposition---it underwent voicing because it was in a phonetic
> environment conducive to voicing, and instances of "of" not in that
> environment were changed to match the most commonly used form > (again,
> not because they are prepositions, but because of another > functional
> factor).
>
>> ISTM that sound changes may occasionally be blocked (at least
>> temporarily) when they would otherwise lead to loss of an >> "important"
>> distinction. I'm thinking of things like the loss of Latin final
>> /s/ in Spanish, which appears to have been blocked only when it
>> denoted the plural of nouns (there, it was later extended even to
>> nouns that didn't have it in the first place). Today, even that
>> final /s/ is beginning to be lost in some varieties -- over a
>> millenium after it was dropped elsewhere.
>>
>> Without having thought too hard about it, isn't this sort of thing
>> pretty close to Analyst's "sound change conditioned by
>> part-of-speech"?
>
> Whatever blocked s-deletion was most likely a phonetic environment
> that the relevant words typically occurred in. I don't know the
> relevant facts here however, so this is just a guess, but I'd > expect
> the non-deleting /s/s to have typically occurred pre-vocalically
> (causing them to be resyllabified as onsets, and thus, blocking
> deletion).

Well, Andrew Woode completely shot down that particular example, at
least as far as Spanish is concerned (should have checked my facts
before posting!). Maybe it (blocked s-deletion in plurals) still works
in other Western Romance languages (Catalan, Occitan, earlier
French?) -- or even Middle English -- but I don't know enough about them
to say.

I guess the question is whether phonetic environment is the _only_
factor governing sound changes, or whether preservation of an important
(for the language involved) distinction is _ever_ a factor in blocking a
merger or deletion.

>> Today, even that final s is beginning to
>> be lost in some varieties -- over a millenium after it was dropped
>> elsewhere.

[Andrew:]
> Though in many the phonological distinction in the relevant > syllables
> is not quite gone (often being expressed in new vowel > distinctions).

Like this:
Before s-deletion, <la mula>, [la mula], /la mula / has plural <las
mulas>, [l&s mul&s], /las mulas /.
After s-deletion, <la mula>, [la mula], /la mula / has plural <las
mulas>, [l& mul&], /l& mul& /.
Similarly, /e / > /e, E / and /o / > /o, O /.

If the singular/plural distinction in nouns and adjectives and the
2nd/3rd person distinction in verbs weren't "important" ones, and
therefore worth maintaining, presumably the loss of final -s wouldn't
have induced the splits in the vowels in these Eastern Andalusian
varieties? That is:

After s-deletion, <la mula>, [la mula], /la mula/ would have plural <las
mulas>, [la mula], /la mula/.

Here, [a] and [&] have remained allophones of /a /, and <las mulas> is
pronounced [la mula], because the allophony is governed by whether the
syllables are open or closed. The singular/plural distinction would be
completely lost in nouns ending in -a and -o, though it would still be
there in the masculine definite article (/el / vs /lo /), and in nouns
ending in a consonant (<la paz> /la pas / would have plural <las paces>
/la pase /).

In French, final s-deletion _did_ lead to loss of the singular/plural
distinction in most nouns with no compensating split elsewhere.
(Admittedly, French maintains the distinction in both masculine and
feminine definite articles). Does this imply that the sing/pl distinction
is "less important" in French than in Spanish?

The sound changes in question were triggered by phonetic factors, and
they happened to occur in such a way as to create a phonemic split.
Eastern Andalusians didn't predict the forthcoming s-deletion and plan
for it by changing their vowels in advance. The vowel changes just
happened on their, and sometime later, s-deletion happened (both are
independently attested changes). It's nothing more complex than that,
with the "importance" of number-marking playing no role.

Maintenance of contrast is definitely a functional factor in sound
change, but it doesn't quite work the way you're describing. Anna
Lubowicz has offered up a formalization of the kind of thing you're
talking about with her Contrast Preservation Theory, but there are
some serious problems with her approach (which I discussed in my
dissertation).

I must say I agree with you. The trouble with the "preservation of important contrasts" argument is that word "important". If half the world's languages get on just fine with no obligatory sing/pl contrast expressed on nouns, why should speakers of language X go out of their way not to lose it? It has the flavour of being a Just So story -- if it was lost, it wasn't important, if it wasn't, it must have been. Not to say that in occasional particular cases one might not be able to find additional (non-phonological) supporting evidence for why something happened -- but simply to say that it happened in order to preserve a contrast really says nothing at all.

The best way to think of how contrast preservation affects sound
change is to look at what happens when a contrast is relatively
difficult to perceive/make. If a sound change is going to occur, one
of two things will happen: the contrast will be given up on resulting
in a merger, or the contrast will be enhanced by making one or both of
the contrasting units more distinct. But either possibility can and
does occur, and across languages, even within languages at different
points in time, we see the same contrast being handled differently,
sometimes merging, sometimes being enhanced (and often, just being
left alone).

Yes. But the question remains unanswered -- why did they go one way rather than the other? Which is a bit unsatisfactory. Though of course unsatisfactorinesses like this abound in historical linguistics. Are languages chaotic systems? (In the sense that tiny differences typically get amplified -- the amazonian butterfly effect.)

John.

.