LINGUIST List 14.960

Tue Apr 1 2003

Sum: Cognate Objects/Unaccusatives

Editor for this issue: Steve Moran <>


  1. Andrew McIntyre, summary: cognate objects/unaccusatives

Message 1: summary: cognate objects/unaccusatives

Date: Mon, 31 Mar 2003 17:25:42 +0200
From: Andrew McIntyre <>
Subject: summary: cognate objects/unaccusatives

Dear linguists, 

this summarises the responses to a query (Linguist 14.697) I sent a
while ago about the occurrence of cognate objects with unaccusative

The following people are thanked for answering: Werner Abraham, Nino
Amiridze, Martin Haiden, Jaume Mateu, Ora Matushanksy, Thomas
McFadden, Asja Pereltswaig, Natalya Serdobolskaya, Anja Wanner

A couple of apologies: the summary will be long because I found it
hard to condense some of the replies in a fair way. I did not
summarise every response because some points are duplicated in other
replies or in the cited literature. 


Asya Pereltswaig, who has worked on matters directly bearing on the
query, made the following observations: "Essentially, the normal
assumption that cognate objects are impossible with unaccusatives is
correct with two notable exceptions (which you've predicted in your

1) Some languages (e.g., Russian, Hebrew) allow a different type of
"cognate objects" (which, as I have argued, are neither really
objects, nor cognate in any syntactically interesting way). These
"quasi-cognate objects" exhibit a number of non-object-like properties
(catalogued in my WCCFL and CJL papers [see below, A.M.]). Where
morphological case is available (as in Russian) they are
non-accusative (e.g., instrumental in Russian).

2) Even languages like English and French, which normally do not allow
cognate objects with unaccusatives (perhaps, German is like that too),
allow cognate objects with some "unaccusative" verbs, DIE being one of
the most common ones cross-linguistically. GO, which you have in the
second example, may also turn out to be really an unergative
verb. Talke Macfarland has a whole chapter in her dissertation on why
DIE allows cognate objects and whether it is really unaccusative"

The literature cited by Asya was:

-Macfarland, Talke (1995) Cognate Objects and the Argument/Adjunct
Distinction in English. Ph.D. dissertation, Northwestern University,
Evanston, Illinois.

Finally, I've found examples of cognate objects with causativized
verbs in Hebrew and Russian (see my papers)."

-Pereltsvaig, Asya (1999) Two Classes of Cognate Objects. In Kimary
Shahin, Susan Blake, and Eun-Sook Kim (eds.) The Proceedings of the
WCCFL XVII. Stanford, CA: CSLI Publications. Pp. 537-551.20

-Pereltsvaig, Asya (1999) Cognate Objects in Russian: Is the Notion
"Cognate" Relevant for Syntax? Canadian Journal of Linguistics

-Pereltsvaig, Asya (2002) Cognate objects in Modern and Biblical
Hebrew. In Jamal Ouhalla and Ur Shlonsky (eds.) Themes in Arabic and
Hebrew Syntax. Dordrecht: Kluwer. Pp. 107-136.


Martin Haiden, who is writing an article dealing with German cognate
objects, makes the following points:

-Tanya Reinhart has work suggesting that some unaccusativity
diagnostics are sensitive to argument reduction rather than to
unaccusativity; if this applies to the verbs at hand, the paradox with
cognate objects is only apparent. This work can be downloaded:

-The verbs taking cognate objects (and path and vehicle objects) tend
to have volitional subjects. This suggests that they are agents,
giving us grounds for querying whether we are dealing with
unaccusative (variants of) verbs. 


Thomas McFadden notes that (older) German also exhibits
genitive-marked cognate objects with "die" like (1), attestable by web
search. This could be part of a more general diachronic tendency in
which the formerly more common verb-selected genitives been dying a
slow death, dwindling down to about half a dozen memorised
relics. (The claim is not that this predicts the constructions to
exist, since with the demise of the genitive, we might a priori have
expected a blanket ban on cognate objects with this verb, rather than
an accusative object.)

(1) sie sind eines elenden Todes gestorben
 they are a miserable death(GENITIVE) died


Nino Amiridze noted that the constructions I alluded to in the
original query are possible in Georgian. For example:

(2) igi sashinel-i sikvdil-it mo-kvd-a 
 (s)he.NOM horrible-NOM death-INSTR PREVERB-die-S3.SG(3D(s)he.died))
 "(S)he died a horrible death"

(4) man igi sashinel-i sikvdil-it mo-0-kl-a
 (s)he.ERG (s)he.NOM horrible-NOM
 "(S)he killed him/her a horrible death"


Jaume Mateu wrote:

I remember that for example "morire una bella morte" (to die a
wonderful death) is an example attested in Italian, but you can't say
"*Gianni � morto una bella morte" (Gianni IS died a wonderful death)
nor "*Gianni ha morto una bella morte" (Gianni HAS died a wonderful
death). I think that those defectivity effects are described in the
Grande Grammatica Italiana di Consultazione by Renzi et al. Take a
look at this Italian grammar.


Werner Abraham wrote: "There is no good reason against, and there are
only good reasons for, the occurrence of cognate objects with
unaccusative verbs, at least in Ls where the uV can reliably be
identified. In English, this is not the case. In German they are
always perfectives, and with perfectives (irrespective of whether they
are intransitives or transitives), the end of the resultative may very
well be expressed by a nominal or PP (if the semantics allows
that). Please see my paper:

Abraham, W. 1999. "The aspect-case typology correlation: perfectivity
triggering (split) ergativity. Burzio's generalization explained." In
E. Reuland (hg.) Arguments and case: explaining Burzio's
Generalization. Amsterdam: J. Benjamins, 129-190.

All writings on unaccusativity in English blur the issue a lot because
there is nothing that makes unaccusativity clearly and intuitively
identifiable, while it is in German. I argued in this paper that this
is due to the lack of perfective morphology in English."

Dr. Andrew McIntyre
Universitaet Leipzig
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue