Allomorphy conditioned by syntactic adjacency

[Enrique L. Palancar Vizcaya]

Dear typologists,

I'm writing a paper on what I think it is an uncommon grammatical phenomenon
in languages. This phenomenon involves morphophonological alternations
triggered by the adjacency of a given syntactic constituent in a clause.
This is shown in a dialect of Otomi (Otomangue):

San Ildefonso Otomi verbs have two inflectional forms (Form I & Form II).
The conditioning is the following:

- Form I is used when the verb occurs at a clausal boundary or when it is
followed by a non-spatial PP adjunct. This form can be uttered in isolation
and is employed as the citation form.
- Form II is used when the verb does not occur at the end of a clause, i.e.
when it is followed by an NP or by a spatial PP adjunct. This form
constitutes a phonological unit with the following word and it cannot be
uttered in isolation.

Examples  (“7” represents a glottal stop, “ë” an epsilon):

Form I   (1)  dá 7wëti
                  1.PST sew(tr)
                  “I sewed it”
Form II   (2) dá 7wëta=ma duhtu
                  1.PST sew(tr)=1.POSS clothes
                  “I sewed my clothes”

In (1), the verb “7wëhti” “sew” is used in Form I, as it occurs at a clausal
boundary. Form II of the verb is used in (2) given that an NP follows the
verb; it is a bound morph that cliticizes to the word “ma”, the possessive
for first person.
   We see here an alternation targeting verbs triggered by the adjacency of
a
following syntactic constituent, in this case an NP. Even though it may
remind some of sandhi
phenomena, the conditioning is not purely phonological since the verb occurs
for example in Form I before non-spatial adverbs, as in (3):

Form I   (1)  dá 7wëti xiñho
                  1.PST sew(tr) good
                  “I sewed it well”

As far as my knowledge goes, a similar phenomenon has been observed in some
French Creoles in the Indian Ocean, such as Isle-de-France and Reunion
Creole with a rule known as “Final Vowel Truncation” (Corne 1982). A
some-how related process is found in Welsh with the so-called
“soft-mutation”. Apart from other contexts, in Welsh, the mutation targets
the first unit of a NP or a non-finite VP when a syntactic phrase (an NP, a
non-finite VP, an AP, or a PP) precedes the targeted constituent (Borsley &
Tallerman 1996).

I wonder whether you know of a similar or related phenomenon in other
languages.

Thank you very much for your help,


Enrique Palancar


PD. References cited: - Corne, Chris. 1982. Final Vowel Truncation in Indian
Ocean Creole French, in Philip Baker & Chris Corne (eds.), Isle de France
Creole: Affinities and origins. USA: Karoma Publishers, Inc.; - Borsley
Robert, D. & Maggie Tallerman. 1996. Phrases and Soft Mutation in Welsh.
Journal of Celtic Linguistics, 5 (1-49).

_________________________________________________________________
Add photos to your e-mail with MSN 8. Get 2 months FREE*.
http://join.msn.com/?page=features/featuredemail