[Lingtyp] differential argument marking conditioned by modification

Fri May 6 07:47:20 UTC 2022

Hi Dejan,

Sidaama (Highland East Cushitic, Ethiopia) is a language you could have a
look at. In the description by Kazuhiro Kawachi (2007), available here:
https://www.acsu.buffalo.edu/~dryer/KawachiSidaama.pdf, it is shown that
the language has different case markers for modified and unmodified nouns -
cf. for instance the list of abbreviations where we find dedicated glosses
for nominative, genitive, dative and locative of unmodified vs. modified
common nouns. Relevant examples that demonstrate the case marking on
modified nouns can be found if you search for ".MOD" in the glosses.

Here is a quote from the thesis:

"The nominative, genitive, and dative-locative case suffixes on nouns can
have different forms depending on three factors: (i) whether the noun is
common or proper, (ii) the gender of the noun, and (iii) (in the case of a
common noun) whether or not the noun is Modified by another element." (p.
353)

Example (p. 355): unmodified vs. modified nominative masculine common noun
(I highlight the relevant case markers by ** in the following)

Unmodified:

beett-**u** kaw-í-ra da-∅-i

child-**NOM.M** here-GEN.PRON.M-ALL come-3SG.M-S.PRF.3SG.M

'The boy came here.'

If preceded by a demonstrative, a genitive modifier or an adjective the
nominative is marked by -i (for modified nominative masculine common
nouns), e.g.

Modified:

hakku beett-**i** kaw-íra da-∅-i

that.M.NOM child-**NOM.MOD.M** here-GEN.PRON.M-ALL come-3SG.M-S.PRF.3SG.M

'That boy came here.'

Other examples in the grammar show that modification by a relative clause
also triggers the use of the "modified" case (-i in the nominative):

p. 472

y-oo-tt-o  r-**i**  k’arr-i-s-am-ø-ino-’e.

say-P.PRF.2SG-2SG-M things-**NOM.MOD.M**
be.in.trouble-EP-CAUS-PASS-3SG.M-P.PRF.3-1SG

‘I got confused with what you (SG.M) said.’ (literal translation seems to
be: 'The things that you said troubled me.', YT)

The actual position of the modifier is not relevant for case-marking.
Modifiers usually precede the head noun in Sidaama, but there are also
examples in the grammar where postposed modifiers (such as the universal
quantifier ‘all’) trigger the use of the "modified" case on a noun.

p. 306

duuču-nk-u mann-**i** ha'r-∅-i

all-EMP-NOM.M people-**NOM.MOD.M** leave-3SG.M-S.PRF.3SG.M

'All the people left.'

or:

mann-**i** duuču-nk-u ha'r-∅-i

people-**NOM.MOD.M** all-EMP-NOM.M leave-3SG.M-S.PRF.3SG.M

'All the people left.'

It might be interesting to compare languages with respect to what qualifies
as "modification" (or said differently: what triggers the use of the
dedicated "modified" case). Kawachi writes on this subject (p. 353): "Note
that 'modification', which is normally a syntactic notion, has to be used
here in a sense specific to Sidaama common nouns, and it is capitalized as
'Modification' (and its related forms are also capitalized whenever the
distinction is relevant: Modified, Unmodified, Modify, Modifier) [...].
Though affixation is not 'modification' and nouns are not described as
'modified' (in the ordinary sense) by an affix, Sidaama common nouns
accompanied by the possessive pronominal suffix behave the same way in the
selection of the allomorphs of the nominative, genitive, and
dative-locative case suffixes as those that have dependents (in other
words, those 'modified' (in the ordinary sense) by genitive NPs or
adnominals), and these Modified common nouns behave differently from
Unmodified common nouns, namely those neither with a dependent nor the
possessive pronominal suffix. Thus, the use of the possessive pronominal
suffix on a common noun counts as Modification of the noun."

A final note: Although related Highland East Cushitic languages also mark
case (some of which even have more elaborate case systems than Sidaama), I
am not aware that the distinction between unmodified vs. modified is
relevant for case marking in these languages.

Best,

Yvonne

*___________ Yvonne TREIS *• Chargée de recherche
Directrice Adjointe • *LLACAN (UMR 8135 du CNRS) <https://llacan.cnrs.fr/>*
Rédactrice en chef • *Linguistique et Langues Africaines (LLA)
<https://llacan.cnrs.fr/lla/index.html>*
Web : *Page personnelle <https://llacan.cnrs.fr/p_treis.php>* • *Publications
sur HAL
<https://hal.archives-ouvertes.fr/search/index/?q=yvonne+treis&sort=producedDate_tdate+desc&rows=100>*
Tél. : +33-1-4958-3710

On Thu, 5 May 2022 at 17:31, Daniel Ross <djross3 at gmail.com> wrote:

> Dear Dejan,
>
> In the broadest sense of differential morphosyntax for objects, one thing
> that comes to mind is object incorporation in polysynthetic languages,
> where I believe this is typical only of bare noun forms without
> modification (and especially where those verb+noun combinations are
> culturally typical). That isn't differential object marking in a
> traditional sense (and I think typically optional and stylistic, even for
> unmodified nouns), but it is a case of objects being expressed differently.
> I don't believe that incorporation is considered a typical feature of
> Yukaghir languages, but it is a feature of some neighbors, so one possible
> explanation for this feature might be contact? Could one the non-modified
> form of this be considered less typologically marked, almost as some kind
> of promotion? Or does the modified form seem like some kind of non-default
> as an exception due to being modified?
>
> Differential object marking isn't my area of research, but my impression
> is that what you describe is not typical (for example, in the Romance
> languages).
>
> Daniel
>
> On Thu, May 5, 2022 at 3:02 AM <matic at uni-muenster.de> wrote:
>
>> Dear typologists,
>>
>> Has any of you ever seen a case of differential argument marking
>> conditioned by whether the argument is modified (i.e. has an attributive
>> modifier or some other adnominal dependent) or not?
>>
>> I have this in Tundra Yukaghir. The story goes roughly as follows: If I
>> want to say "I see a reindeer", "reindeer" will take the case suffix A; if
>> I say "I see a good reindeer", "reindeer" takes the case suffix B. In both
>> instances, the argument behaves syntactically the same, the only difference
>> between the two types of case forms being the presence vs. absence of an
>> attributive modifier.
>>
>> I have been wondering (a) if any of you know of similar cases in any
>> other language, and (b) how to name and gloss the case form that is used
>> only with modified arguments.
>>
>> (I am aware that there is a similar situation in the only relative of
>> Tundra Yukaghir, Kolyma Yukaghir, and that there have been some analyses
>> based on these data, but that's not what I am asking about.)
>>
>> Thank you for your answers!
>>
>> Best,
>> Dejan
>>
>> --
>> Prof. Dr. Dejan Matic
>> Institut für Sprachwissenschaft
>> Westfälische Wilhelms-Universität Münster
>> Aegidiistr. 5
>> 48143 Münster
>> Germany
>> tel. +49-251-8324494
>> _______________________________________________
>> Lingtyp mailing list
>> Lingtyp at listserv.linguistlist.org
>> http://listserv.linguistlist.org/mailman/listinfo/lingtyp
>>
> _______________________________________________
> Lingtyp mailing list
> Lingtyp at listserv.linguistlist.org
> http://listserv.linguistlist.org/mailman/listinfo/lingtyp
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20220506/2d2df48d/attachment.htm>