LINGUIST List 3.34

Wed 15 Jan 1992

Disc: Diachronic lengthening, inflection development

Editor for this issue: <>

Directory

1. bert peeters, 3.26 Diachronic Lengthening
2. "Randy J. LaPolla", Diachronic Development of Inflectional Endings
3. John E. Koontz, Re: 3.24 Queries: ...; Diachrony; ...

> Date: Thu, 9 Jan 1992 9:25:17 GMT-10:00
> From: Fran Karttunen <kartunenuhccux.uhcc.Hawaii.Edu>
>
> In Nantucket, where I grew up, monosyllabic words ending in r were
> pronounced as two syllables:
>
> door > dowa (can't do schwas on this keyboard either)
> flour > flauwa (homophonous with flower)
> beer > biya
> there > theiya
>
> As a kid I thought it rather strange but chalked it up to the irrationality
> of
> English spelling. Later on Broadcast English and travel away from peripheral
> New England showed me that it was a rather localized dialect feature.

Ruth Blair (p.c.) tells me that in Australian English the word _law_ now
seems to consist of two syllables: [lowa]. [a] endings in Australian English
are rampant in replacement of schwas (e.g. moda for mother, fada for father,
etc.)
See you lada. (no publicity intended)

David Eddington asks "Are there languages
 thathave developed inflectional systems where there once were none?"
It seems from the work I have been doing on Tibeto-Burman morphology
that most if not all the inflectional morphology (not the derivational
morphology) developed sometime after the breakup of Proto-Tibeto-Burman.
That is, none (or almost none--and there are people who will disagree
with me) is reconstructable to Proto-Tibeto-Burman, though many of the
languages now have complex inflectional systems (both nominal and verbal).
I have a paper coming out in Feb.'s issue of BSOAS on one type of
verbal inflection, and another paper coming out in Linguistics of the
Tibeto-Burman Area (14.2 or 15.1) on one type of nominal marking.
--Randy LaPOlla

Message 3: Re: 3.24 Queries: ...; Diachrony; ...

David Eddington asks:
> ... The drift of the Indo-european languages has been toward the loss of >
inflectional endings with the result that word order becomes crucial (as >
well as prepositions) in determining case. Are there languages that have >
developed inflectional systems where there once were none?

I think I can safely say that the inflectional systems of the Siouan
languages are at least observationally stable. The time depth of Siouan (as
opposed to Siouan + Catawban) must be comparable to the time depth of the
branches of Indo-European, but the inflectional systems are quite similar in
detail, and, when different in detail, still similar in character. One has
to assume that details have been variously retained or innovated in
parallel.

Inflection consists primarily of personal inflection in the verb: agent
and/or patient, roughly, as appropriate to the sense. Secondarily, there are
(a) a system of deriving verbs stems with dative, reflexive, reflexive
possessive, reciprocal, etc., reference, which has syntactic implications,
and might be regarded as inflectional; and (b) a separate, but comparable
system of deriving stems with various applicative/locative senses which
might also be regarded as inflectional.

At least the formation of the first and second person agents, and some sort
of proto-dative/reflexive possessive and must be of Proto-Siouan age, since
these involve complex and irregular phonological patterns repeated across the
family. Other systems, e.g., other pronominals or the applicatives, are
universal, but so transparent in formation that they are open to
interpretation as fairly recently absorbed proclitics.

Pronominal markers like the inclusive or the third person indefinite, which
are further from the stem than the applicatives may be similarly recent,
though there are attested mechanisms in Siouan and elsewhere for
metasthetizing elements in ways which make it impossible to considently date
morphology by the "tree-ring" approach.

In some cases inflectional markers in the Siouan languages are clearly
recent innovations, e.g., the animate third person plural patient marker
-wic^ha- in Dakotan, which seems to be an incorporated form of the free noun
wic^ha's^a ~ wic^hasta `person; man'.

It is also worth noting that the Siouan languages are extremely prone to
supplementing relatively obscure markers or marker systems with more
transparent ones, without replacing the original system, producing extensive
pleonasm in the relevant paradigms.

To summarize, the inflectional systems of the attested Siouan languages are
composed of both inherited and innovated inflectional material, with the
difference being sometimes difficult to determine, i.e., when inherited
material may have been independent material at earlier stages. If the only
inflections in Proto-Siouan are those that we can find attested today,
whether clearly or ambiguously, then the attested Siouan languages have at
least as much inflection as Proto-Siouan, and probably rather more.

Returning to the question, I'm not sure if it is fair to say that
Indo-European as a whole exhibits any overall trend to absolute loss of
inflection, and the whole issue of trends in any sense save as ex post facto
discoveries is fraught with pitfalls. There are IE languages with much less
inflection than PIE (e.g., English), and some with about the same amount
(e.g., Lithuanian or Hindi). In some cases the existing inflection is to a
fair extent "original" (e.g., Lithuanian); in others it is to a greater
extent innovated (e.g., Hindi). I am not going to fight to the death over
these examples, if someone has other views of the matter. In my view all
languages are simultaneously engaged in the twin (related) processes of
losing and gaining inflectional systems, but sometimes one process or the
other leads.