Whether integrativity bears on agreement categories only

[René-Joseph Lavie]

Hello everybody!

In the Latin word _dôna_ (gifts, neuter, accusative, plural), the
inflectional morpheme _-a_ marks at once the neuter gender, the
accusative case and the plural number; for this, it is said to be
'integrative'. Gender, case and number are agreement categories. I have
the following conjecture. 

Conjecture C: the categories that can be integrated into one
(integrative) inflectional morpheme necessarily are agreement
categories. 

Conjecture C is one way only, the reverse implication does not hold: the
marks of agreement categories do not necessarily get integrated together
in one morpheme. A counter-example can be the French sentence elles sont
jolies (they are nice, feminine, plural), in which -e- marks the
feminine and -s is the mark of the plural. This French morphology bears
indeed on agreement categories (gender and number) but it is
agglutinative, not integrative. 

I have not been able to falsify C with the (limited number of) languages
I know or for which I have documentation. I have not seen C stated in
the literature I read on agreement or on inflectional morphology. So I
stay with these questions: 

Is conjecture C verified?
If so, why should C be necessary? 
Any publication some of you could be aware of?

Thanks in advance for any help.

-- 
René-Joseph Lavie 
MoDyCo (Université Paris Ouest Nanterre la Défense et CNRS) 
rjl at ehop.com 
http://rjl.ehop.com [1] 
33 (0)9 8065 6722 ---- 33 (0)6 0818 6973

 

Links:
------
[1] http://rjl.ehop.com