Syllabicity

[CONNOLLY at LATTE.MEMPHIS.EDU]

Pat promised:

>henceforward, unless I
>inadvertently regress, I will refer to lexical and grammatical differences.

That will help clarify things greatly.  A question on the use of the word
morpheme: Some of use it for the smallest element of a word (i.e., either root
or affix), but in some of our discussion we have tended to restrict it to
grammatical morphemes (essentially, to affixes).  I think all of us (including
me) have to be more careful here than we have been.

> <snip>

>Pat answers:

>I do not think 'withdrawing' is the correct characterization. I do not
>subscribe to the current theory of Lallwo{"}rter as I believe you do, and
>here we must simply agree to disagree since neither of us will be persuaded
>by the other.

I agree that there's little to be gained by pursuing the matter.  I would only
caution that one must be as careful of possible Lallwoerter as of onomatopoetic
expressions, since either could be created anew at any time.  I mean, how old
is the word _cuckoo_?  Since that's how the European cuckoo really sounds (even
outside clocks_, it is futile to look for etymological connections, since
anyone hearing it at any time could create the word anew.  And that's the
problem with Lallwoerter: babies do babble the same everywhere, and parents
love to think that their offspring are talking to them.  So etymologies of
words auch as 'father' and 'mother' are riskier than normal as far as the root
goes.  The suffix, of course, if that's what it is, is another matter, since
that's not babbling.

[stuff deleted]

Pat claimed:

>I do believe there is a strong possibility that, in the very earliest stages
>of language, phonemes did have actual meanings but even by the time of CV
>roots, this association (if it ever existed a la sound symbolism) had been
>lost in terms of the basic meaning of these monosyllables (though it might
>linger on almost as a grace-note to the meaning in the form of nuances or
>emotional interpretations).

This can be interpreted several ways.  The most literal reading would be that
any [s], anywhere, anytime, meant the same thing.  If that's what you mean, all
I could say is that it's pure conjecture, and naturally there's no way to prove
or disprove it.  It would be a non-scientific statement and thus have no place
in a serious discussion on the origins of language.

There's also a phonetic problem: while vowels, nasals, [r] and [l], and
fricatives could plausibly exist by themselves, stops really can't (and
couldn't back then either, unless our vocal apparatus functioned very
differently).  Yet stops dominate in PIE roots, in the sense that there are
many more of them than of the other consonant phonemes.  Why would that be?

Pat again:

>Now, evidently, my early training in linguistics differed from your own
>since, as another wrote recently, this definition of phoneme has to be with
>the (once fashionable?) idea of minimal pairs.

I don't quite understand this sentence, but I don't that that's the explanation
for our differences.  My original training was structuralist (phoneme = a class
of sounds, out there in the real world); I now regard it as a psychological
entity, and not necessarily the best analysis for all situations (remember what
I wrote privately about the vowels of Turkish grammatical morphemes, where
distinctive feature analysis works well but phonemes are clumsy).  Minimal
pairs, closely associated with structuralism, are still a very useful
diagnostic even for the psychological analyses.  The fact that [phit] and [bit]
are different still tells us that their mental phonological representation must
be different.

[stuff omotted]

>Pat, puzzled:

>Then how did "original" e: creep into the discussion? Do you believe that
>there could be two morphemes in IE, *CeC and *Ce:C, that differed
>**lexically** when *Ce:C is not the result of earlier *CeHC?

I don't know of any for *early* PIE, and there shouldn't have been any.  I'll
need help here from someone who knows more, but don't some scholars posit
original long vowels for *late* PIE?  We would be talking about either newly
created roots or about borroiwings from non-IE languages.  Can anyone
contribute something?

>Pat, more or less agreeing [that nouns and pronouns may have different syntax]:

>But do you not think that where their employments differ, one of the major
>reasons is the typical brevity of many pronominal elements, and their
>encliticity?

This is probably true for word order -- at least, if there's any truth to
Behaghel's "Gesetz der wachsenden Glieder", by which words and phrases tend to
get longer as the clause goes on.  This "law", of course, is based on empirical
observation but seems to work in a good many languages.  But that won't explain
why they can't be used with determiners, or with adjectives and may behave
oddly when used with prepositions.  (All of this, by the way, is
language-specific; noun and pronoun direct objects differ much more in Spanish
than in English.)

Leo

Leo A. Connolly                         Foreign Languages & Literatures
connolly at latte.memphis.edu              University of Memphis