Languages without pronouns?

[Michael Cysouw]

[disclaimer: The genetic and areal indications in this mail are only for
convenience of identifying the languages named!]

Elizabeth Ritter asked:
> Also, does anyone know of a language that has a single (true) pronoun
> for either
> a.    1st OR 2nd person
> b.    2nd OR 3rd person

A few definitions have to be more precise to be able to give an answer:

1) if by 'true' you mean something like 'free' or 'morphological
independent', then:

- in the singular, it is almost unsttested as far as I know. The best case
seems to be the SOuth American language Qawesqar (or 'Kaweskar') which has
a pronoun _caw_ for second and third person. Another possible case is cited
by Noyer (1992:151, citing Lipkind 1951; I have been unable to check this
claim). He claims a single independent pronouns _nee_ for first and second
person singular in the Siouan language Winnebago. As far as I have seen,
other Siouan languages do not do this. It might be an misunderstanding.

- in the non-singular, it is not unusual (if there is no
inclusive-exclusive distinction). Except the two kinds of combinations you
ask for, there is of course also a third possibility: a combination of 1st
and 3rd person.
	1/2: Athabscan object pronouns;
	     Awa and Usarufa pronouns (East New Guinea Highlands)
	2/3: Nez Perce (Sahaptin); Warekena (Arawakan)
	1/3: Bagirmi (Nilo-Saharan)

- in the non-singular with an inclusive-exclusive distinctions it seems to
be much rarer. The only clear example is the South American language
Sanuma. Depending on your interpretation, you might add the 'pronominal
root' from Nama as well. Maybe also Nehan (Austronesian form New Ireland)
and Puragi and Kasuweri from the Bird's Head of New Guinea.

- if you also allow a combination of singular and non-singular, you might
want to check the 'Papuan' language Suki, which has a pronoun _e_ for
second person singular and first person plural.

- if you are also intersted in such combinations in the dual, trial or
paucal, then you might want to check the combinations of 2nd and 3rd person
in Amele (Madang), Baruya (Angan), Kalam, Fore and Wiru, (New Guinea
Highlands) Yareba, Korafe (Central and Southeastern New Guinea), Wantoat,
Kewieng (Finisterre) and in Slovene and Upper Sorbian (Slavonic)

2) if by 'true' you mean something like 'pronominal argument' or
'inflectional pronoun' then all kind of combinations are much more common.
I have no time to list all cases that I have found, but there are many. The
problem though is the definition of an 'inflectional pronominal argument'.
In fact, I still do not know precisely how to deal with this notion when
dealing with a large sample of languages.

bye
michael cysouw
ZAS, Berlin


PS. Most of these data are from my thesis on precisely this topic, entitled
"The Paradigmatic Structure of Person Marking". The thesis might not be
easily accessible outside the Netherlands, and it will probably be taking
some time before it will be published internationally. I can send a .pdf
file of the thesis to the ones interested.