LINGUIST List 15.1941

Mon Jun 28 2004

Disc: Re: Comments on things no language does

Editor for this issue: Sarah Murray <sarahlinguistlist.org>

Directory

1. Koontz John E, Re: 15.1816, Disc: Re: Comments on things no language does

Message 1: Re: 15.1816, Disc: Re: Comments on things no language does

On Tue, 15 Jun 2004, William J. Poser wrote:
> Another problematic claim [as reformalized by Poser] is:
>
> There exists no language in which there exists an inflectional
> affix I and a derivational affix D such that I is closer to the
> root than D.

> If the former is intended, on most people's notions of "inflectional"
> vs. "derivational", it is falsified by all of the Athabaskan
> languages.

A paper which addresses this very point is:

Speas, Margaret. 1987. Position classes and morphological universals.
pp. 199-214 in Native American Languages and Grammatical Typology:
Papers from a Conference at the U of Chicago, April 22, 1987, ed. by
Paul D. Kroeber and Robert E. Moore. IULC 1987.

She shows that the claim cannot be maintained as is, but suggests that
it might be maintainable because "In an abstract sense, Inflectional
morphology is on the outside of Derivational morphology, in that
Inflectional morphemes are added after [D]erivational ones." In other
words, the claim may fail with respect to morpheme order, but be
salvaged with respect to order of application of rules. She rejects
the constraint in terms of morpheme order specifically on a basis of
Navajo.

However, a few years back I presented a description of the Omaha-Ponca
(Dhegiha Siouan) dative construction that might even present problems
for the rule ordering version.

Koontz, J. E. (1989). Ordering of morphological rules in
Omaha-Ponca. Read at the 88th Annual Meeting of the AAA (CAIL 28),
Washington, DC.

It appears to me that the formation of the OP dative paradigm can only
be accounted for sensibly by starting with the inflected form of the
non-dative verb and then applying various vowel changes or dative
morpheme insertions. The rules governing the insertion process depend
on the person of the verb and what locative prefix, if any, it has.

The scheme that exists develops, probably, from an earlier one in
which the second, vowel-only allomorph of the dative prefix gi- ~ i-
fused with the preceding morpheme, changing its vowel. Today,
however, the changed vowel may be rather far from the reconstructe
original slot of the dative following the pronominal and preceding the
underlying stem.

For the uninitiated, which, based on the attendence at the Siouan and
Caddoan Conference must be nearly everyone, a dative prefix converts a
verb from an unmarked transitive or in transitive to one that agrees
with a more remote transitive object or intransitive subject, e.g.,
naN?aN 'to listen to, hear' becomes ginaN?aN 'to listen to for
someone', or t?e 'to die' becomes git?e 'one's own to die'. A
locative prefix adds a central or peripheral locative argument, e.g.,
gase 'to cut' becomes i...gase 'to cut with' or gdhiN 'to sit' becomes
a...gdhiN 'to sit on'. Datives and locatives in Omaha-Ponca are
fairly productive, but both are usually taken to be derivational in
Siouan grammars. (In these stems ... represents the principle locus
of pronominalization.)

This is were Poser's following comment applies:

> One problem of course is that "inflectional" and "derivational" have
> different meanings for different people.

No doubt an extended debate of the issue of the
inflectional/derivation status of the Omaha-Ponca dative would be
possible. It seems to me, however, that such a debate might be at
least partly misguided, as probably most interested parties would
agree that personal inflection of verbs is more inflectional and less
derivational than case marking schemes that use verb affixes. So, the
overall situation is relatively contrary to the generalization even if
it isn't absolutely contrary.

In fact, in some of the better behaved personal forms the dative marking
remains a prefix "inside" of the pronominal prefix. The difficulty is
that in other cases it leaps over pronominals or even parts of pronominals
and fuses with a vowel outside them or even with several vowels in a row.

LOC PRO i ROOT > LOC-i PRO ROOT
 LOC PR-i-O ROOT

Examples of these would be:

eagdhaN (a-a-gdhaN LOC-I-put 'I put it on' + DATIVE) 'I put it on for
him'

a + i => e

iNgadhiN (aNg-a-dhiN we-LOC-have 'we have it' + DATIVE) 'we have for
him'

aNg + i => iNg

The Siouan languages in general provide many trivially contrary
situations to the original morpheme-based ordering generalization.
There are many indubitably derivational affixes that preceed all
inflectional and derivational prefixes. Stems that have this pattern
are called infixing and the initial sequences is called a preverb.
The preverb + rest of stem structure is fairly common in North
America.

Speas' sharpening of the inflectional-derivational constraint would
make short work of infixing verbs and of the class of verbs that place
some pronouns before locatives and some after them, not to mention
those verbs that take duplicate sets of pronominals in several
different slots, sometimes two adjacent slots. In all of these cases
it can be taken that pronominalization follows application of the
other processes, which form the stem that is pronominalized and are
therefore derivational with respect to pronominalization.


- John E Koontz