LINGUIST List 6.270

Wed 22 Feb 1995

Sum: Verbal/pronominal feature differences

Editor for this issue: <>

Directory

1. Maik Gibson, Sum: verbal/pronominal feature differences

Earlier this month I posted the following query:

======================
What languages (if any) do people know about, where there are
distinctions carried in the verbal morphology which cannot be indicated
in the pronominal system? (for example, if a language expressed gender
differences in the verbs, but not in pronouns).
======================

The following people very kindly replied with various pieces of very
useful information:

Antton Elosegi Aldasoro (fvpelalasd.ehu.es)
ariel mira (marielccsg.tau.ac.il)
"Ellen L. Contini-Morava" (elc9jfaraday.clas.virginia.edu)
Simon Corston (corstonhumanitas.ucsb.edu)
Brian D Joseph (bjosephmagnus.acs.ohio-state.edu)
June Wickboldt: jwickbolucs.indiana.edu

Here are the replies:

The two most promising examples of what I was asking about come
from Basque and Hebrew, where the pronominal and verbal systems
do not match for gender:
-----------------------

Antton Elosegi Aldasoro (fvpelalasd.ehu.es)

In Basque there is not gender at all in nouns or pronouns, but some
verbal forms carry the distinction of the gender of the 2nd person
 nik ekarri diat nik ekarri dinat
 I(erg) give it to you-masc I(erg) give it to you-fem
Antton Elosegi (University of the Basque Country)

ariel mira (marielccsg.tau.ac.il)

In my own Hebrew "normal" verbal inflections (past and
future) distinguish person gender and number in 2nd and 3rd persons, but
only number in 1st person. So far so good, because so does the free
pronominal system. However, our now present tense is morphologically a
nominal form (for eg. holexet='walk (fem)' as well as 'walker (fem)'.
Now, our nominal forms, having nothing to do with pronominal forms
inflects for number and gender, but does not distinguish persons. The
result is that now 1st person present inflection distinguishes between
fem and masc though the independent pronoun does not.

Modern Hebrew is losing some gender distinctions too!
1. Future tense 3rd pers plural: fem form is hardly ever used. Even the
purist language academy has "abolished" the form. (We don't seem to ever
have had it in past tense).
2. In colloquial speech, the same is happening to 2nd person, no doubt
because the 2nd and 3rd person plural fem forms are identical (though not
the masc ones they have converged with).
3. This needs to be checked! the free plural feminine pronoun, as well as
past inflections of same seem to me to be shaking. But real data has to
be recorded for this, so don't make much of this. I wonder whether mixed
forms are possible, namely, you-fem go+masc, etc.

*and on a more theoretical note*:

Hopefully you will find very few examples for what you are looking,
because I believe (what others have been saying for about a hundred
years) that inflections tend to develop out of free pronouns. Hence, the
dependence between the meanings encoded, though there is no principled
reason for the inflection to change later, I guess. It is really
unlikely, though, because once the form is inflected, fused with the
verb, chances are it won't develop its separate semantics.

If you're interested in theories about the development of inflection out
of pronouns, I have my own, which accounts for the well-known fact that
inflections for 1st and 2nd person are much more prevalent than 3rd
person. I claim this is so not because 3rd person is unmarked, but
rather, since referents of 3rd person are usually much less accessible
than referents of 1st and 2nd person (the speaker and the addressee).
Minimal forms are reserved for more accessible referents (in general), hence
inflections are the natural development out of free pronouns for highly
accessible antecedents.

You can have a look at my
book 'Accessing NP antecedents', Routledge, 1990, chapter 6.

-----------------------

The following two respondents quote cases where logophoric and
obviative marking occur attached to the verb rather than to the pronoun
to which they refer. To me, it seems that these particles are not inflections
as such, though I am not sure exactly how I would classify them.

-----------------------
June Wickboldt: jwickbolucs.indiana.edu

Some languages having logophoric reference mark the reference with
verbal affixes, not pronouns or pronominals. Two are Newari, see
Karen Ebert. 1986. Reported speech in some languages of Nepal. In
F. Coulmas (ed.) Direct and Indirect Speech, Berlin: Mouton de Gruyter,
and Gokana, see Hyman, Larry M. and Bernard Comrie. 1981. Logophoric
Reference
in Gokana. Journal of African Languages and Linguistics. 3:19-37.

Brian D Joseph (bjosephmagnus.acs.ohio-state.edu)

The Algonquian language Cree, spoken in Canada, has a category known
in the literature as "obviative", which provides a way of distinguishing
between different third-persons in a discourse (the first one mentioned is
"proximate", and the next one mentioned is "obviative", so Cree can
distinguish unambiguously between "John met Bill as he was walking down
the street" where Rhe" in English can refer to either John or Bill -- in
Cree it would be unambiguously one or the other).

For the most part, and this is where it is relevant to you, this marking
shows up on the verb (thus the verb form in the above sentence would be
different if it was John walking or Bill walking); nouns can show
obviative/proximate marking (though one class, the so-called "inanimate"
nouns do not), and there are proximate and obviative forms of demonstrative
pronouns, but not of the personal pronouns. Thus "wi:ya" is 'he/Proximate'
as well as 'he/Obviative'.

Pronouns are not usually expressed in Cree, but if you just look at the
personal pronouns, then Cree would be a language of the sort you were looking
for. For that matter, these facts are similar in virtually all the
Algonquian languages, so it isn't just Cree.
-----------------------------

Finally, two respondents noted that independent pronouns are not
necessarily marked for case in the same way as pronominal affixes.
Such phenomena do not seem rare to me (e.g. Arabic does the same).
I am assuming then that such cases are not marked.

-----------------------
"Ellen L. Contini-Morava" (elc9jfaraday.clas.virginia.edu)

In Swahili part of the verb morphology is subject and object
prefixes, that signal info. about participant role with respect
to the action of the verb. This distinction is not made among
independent pronouns, which distinguish only person and number.
But since the subject and object prefixes in Swahili are often
called "pronominal", I don't know if this counts as a distinction
that can't be made by pronouns.

Simon Corston (corstonhumanitas.ucsb.edu)

In my MA, in press in the Pacific Linguistics Series from ANU, I discuss
'Ergativity in Roviana'. Roviana has special pronominal forms used for
absolutive (S ('subject of intr') or O ('object')), and different forms
for A ('subject of tr'). The pronouns make distinctions in person,
number, and incl/ecl for 1PL. There are pronominal verbal affixes on the
verb which are always only O. I.e. whereas the independent prons don't
distinguish S/O, the pronominal affixes do. Somewhere around here I have
a brief sketch of Roviana which I have been sending to people.
-------------------

I found everything very interesting, and am happy to receive any more
info on my query above. Thanks again to those who responded!

Maik Gibson
University of Reading