David Beck dbeck at UALBERTA.CA
Fri Mar 9 17:10:47 UTC 2001

I have not-very-good examples of both types of language (languages
where only subject can be relativized and languages where the object
but not the subject can be relativized. Both of these examples are
from Salishan languages, but neither of these show any signs of
ergativity (as do some of their relatives).

The first language is Lushootseed, where if both subject and object
are third person only the subject can be relativized (there is data
given in a paper that just came out in Studies in Language (24-2) if
people want the details). However, this seems to be a pragmatic
rather than a grammatical restriction -- I have found a few examples
in texts where an animate object is relativized when the subject is
inanimate and the verb denotes a volitional act, so I suspect the
restriction is based on the lack of case-marking in the language that
would disambiguate the grammatical roles of two third persons under

The second example is from Bella Coola (Nuxalk), where (again, only
if both subject and object of the relative are third person),
subject-centred RCs take a special set of pronominal endings (I'll
send this data, though I'm sure it will come out a mess):

	Bella Coola
(i)	(a)	ti+nus?uulX	ti+k'x+t		ti+?aq'wlikw+tx
		D+thief		D+see+3SG:3	D+policeman+D
		'the thief who saw the policeman'

	(b)	ti+nus?uulX	ti+k'x+tan	wa+?aq'wlikw+c
		D+thief		D+see+3PL:3	D+policeman+D
		'the thief who saw the policemen'

	(c)	wa+nus?uulX	wa+k'x+t	ti+?aq'wlikw+tx
		D+thief		D+see+3SG:3	D+policeman+D
		'the thieves who saw the policeman'

	(d)	wa+nus?uulX	wa+k'x+tan	wa+?aq'wlikw+c
		D+thief		D+see+3PL:3	D+policeman+D
		'the thieves who saw the policemen'
			(Davis & Saunders 1997: 99)

(cf.	(e)	k'x+is+kw		ta+Ò©msta+tX
		see+3SG:3SG+QTV	D+person+D
		'the person saw it'
			(Davis & Saunders 1980: 13, line 89)

	(f)	k'x+tis+kw+tuu		ta+sissi+c+tX
		see+3PL:3SG+QTV+and	D+uncle+D
		'and my uncle saw them'
			(Davis & Saunders 1980: 59, line 5))

The RCs in (ia) - (d) make use of a different set of pronominal
suffixes than are used in  (e) and (f) and in  the object-centred
RCs) in (ii)

	Bella Coola
(ii)	(a)	k'x+it		ti+?imlk	ti+qup'+cs+tx
		see+3SG:3PL	D+man	D+punch+1SG:3SG+D
		'they see the man who is punching me'
			(Davis & Saunders 1978: 46)

	(b)	?alhk'yuk+ilh	ti+?imlk	ti+qup'+¬+is
		know+3SG:1PL	D+man	D+punch+PERF+3SG:3SG	D+woman+D
		'we know the man who the woman punched'
			(Davis & Saunders 1978: 49)

Unlike the ordinary pronominal suffixes, these affixes show number
agreement only for the direct object and not the syntactic subject.
Nater (1984: 38) refers to them as participial endings, though to be
honest it is not clear that the "participles" are non-finite (if that
term has any meaning in Salishan languages). Kroeber (1999: 288ff)
has a good discussion of the historical provenance of these

For those interested, the works cited for Bella Coola are:

Davis, Philip W., & Saunders, Ross. (1998). The place of Bella Coola
(Nuxalk) in a typology of the relative clause. In Ewa
Czaykowska-Higgins & M. Dale Kinkade (Eds.), Salish languages and
linguistics, 219 - 34. Berlin: Mouton de Gruyter.
Kroeber, Paul D. (19991). The Salish language family. University of
Nebraska Press.
Nater, Hank F. (1984). The Bella Coola language. Ottawa: National
Museum of Man.

David Beck
Assistant Professor
Department of Linguistics
University of Alberta
4-45 Assiniboia Hall
Edmonton, AB T6G 2E7
office: (780) 492-0807
FAX:    (780) 492-0806
e-mail: dbeck at

       Man has such a predilection for systems and abstract deductions that he
       is ready to distort the truth intentionally, he is ready to deny the
       evidence of his senses in order to justify his logic.

                                     -- Dostoyevsky, Notes from the Underground
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the Lingtyp mailing list