Syncretism by analogy

[Matthew Baerman]

----------------------------Original message----------------------------
> ----------------------------Original message----------------------------
> Does the original syncretism have to be due to regular SC? As far as
> principles of morphological change are concerned, one might think it made
> no difference how the original syncretism came about, provided one could
> show the analogical extension was subsequent to the original syncretism.

The nice thing about sound change is that it can produce bizarre patterns,
which are not reduceable to any obvious morphosyntactic or semantic natural
class. To the extent that these are then extended analogically, it's evidence
of the existence of purely morphological classes of values, something argued
against by plenty of folk.

In the Catalan example, is the case that the indicative form ousts the
subjunctive?

Matthew

Matthew Baerman
Surrey Morphology Group
University of Surrey
Guildford, UK




> In many dialects of Catalan, syncretism 1pl.prs.subj=1pl.prs.ind and
> 2pl.prs.subj=2pl.prs.ind got established in the default Conjugation 1
> before spreading 'by analogy' to the endings of the other conjugations
> (though in some verbs of Conj II, ind and subj retain distinct stem
> variants). But the original syncretism doesn't follow from regular SC (but
> rather, I have argued, from an inadequate and confusing system of vowel
> contrasts as exponents of the categories ind, subj, and imp in 1pl and
> 2pl). Of course, the extension here is across arbitrary inflectional
> classes, but perhaps that is not so different from the Russian case Matthew
> mentions.
>
> Max Wheeler
>
> --On Friday, June 27, 2003 10:43 -0400 Matthew Baerman
> <M.Baerman at SURREY.AC.UK> wrote:
>
> > ----------------------------Original message----------------------------
> > Dear list members
> >
> > A historiographical-cum-factual query. In Livonian, in place of the
> > expected 1sg present tense ending (zero), you get the 3sg ending -b.
> > Kettunen (1938) says that this was on analogy with the simple past tense
> > paradigm, where 1sg and 3sg fell together as the result of regular sound
> > change. Merits of this particular account aside, what interests me is the
> > nature of the claim, namely, that you can have a pattern of syncretism
> > (in the sense of homophony between cells of an inflectional paradigm)
> > which arose by regular sound change in one paradigm, which is then
> > extended by analogy to other paradigms which had not been affected by
> > that sound change. I get a real sense of dij` vu when I read accounts
> > like the above, one which quickly evaporates when I try to recall
> > specific examples. So my question is:
> >
> > Can anyone think of other examples that have been claimed to work like the
> > above (even unlikely ones)? Or ones that seem like they work that way,
> > even if nobody s ever claimed it?
> >
> > Ones I can think of are:
> >
> > --Old Icelandic: 2sg=3sg in the present extended from consonant-stem
> > verbs to all verbs (Kurylowicz 1949).
> > --Eastern Finnmark Sami dialects: comitative sg=inessive/elative pl
> > extended from nouns to pronouns(Hansson 1996).
> > --Russian: genitive sg=dative/locative sg extended from i-stem nouns to
> > a-stems (Sologub 1983).
> > --Dhasanaac/Dasenech: 1pl=2sg/2pl/3SG fem extended from resonant-final
> > stems to all stems (Sasse 1976, Tosco 2001).
> >
> >
> > many thanks in advance
> > Matthew
> >
> >
> > Matthew Baerman
> > Surrey Morphology Group
> > University of Surrey
> > Guildford, UK
> >
> >
> > ---------------------------------------------
> > This message was sent using UNIS MailSystem.
>
>
>
> ____________________________________________________________
> Max W. Wheeler
> School of Cognitive & Computing Sciences
> University of Sussex
> Falmer
> BRIGHTON BN1 9QH, G.B.
>
> Tel: +44 (0)1273 678975 Fax: +44 (0)1273 671320 Email:
> maxw at cogs.susx.ac.uk
> ____________________________________________________________
>


---------------------------------------------
This message was sent using UNIS MailSystem.