[Lingtyp] incipient ergativity conditioned by number of A

[Sebastian Nordhoff]

Dear David,
in both Sinhala and Sri Lanka Malay, institutional actors (government, 
board, committee, police) are marked with the instrumental. I suppose 
that your Likud examples would get the instrumental as well in those 
languages.

You write that [number] seems to be the relevant factor. But if you have 
"one baker" and "thirteen bakers", you would probably not get the 
difference. So it might be more the feature [+institutional] or 
[+collective], as you say.

When looking into the instrumental in the Sri Lankan languages, I was 
wondering whether British English agreement ("The committee have 
discussed ... ") and Dutch feminine institutional reference ("het 
kabinet en haar beleid" 'the cabinet[N] and her[F] policies') are 
actually triggered by the same semantics. I found that interesting since 
this is a grammatical fact that relies on the society having some kind 
of bureaucracy, which poses interesting questions with regard to innateness.
Best wishes
Sebastian




On 6/1/23 06:43, David Gil wrote:
> Hi all,
> 
> Is anybody familiar with a case of split ergativity in which the 
> conditioning factor is the number of the Agent NP?
> 
> My reason for asking:in Hebrew, especially in a journalese register, in 
> a transitive A V P construction, when the A is semantically plural, 
> typically denoting a collective entity, it is often marked with the 
> locative proclitic /b-/ while the verb takes plural subject agreement in 
> an apparent impersonal construction.For example, in a sentence about the 
> Likud political party:
> 
> balikud muxanim lidħot et hamahapexa hamišpatit ...
> 
> LOC-Likud prepare:3.PLM INF-postpone ACC DEF-revolution 
> DEF-legislative.F ...
> 
> idiomatically: 'The Likud is willing to postpone the legislative 
> revolution ...'
> 
> literally: 'In the Likud they're willing to postpone the legislative 
> revolution ...'
> 
> Such constructions are extremely widespread in journalistic writing.The 
> above example, part of a newspaper headline, is followed by a string of 
> several clauses all exhibiting the same construction, each beginning 
> with a semantically plural agent marked with locative /b-:/ 'in the 
> ruling party', 'in closed rooms', 'in the other side', etc. 
> [https://www.ynet.co.il/news/article/bk5kubsin#autoplay]
> 
> In the above construction, the locative proclitic /b-/ seems to be 
> approaching the function of an ergative marker, albeit a rather atypical 
> one: in particular, when the P is definite, as in the above example, it 
> is marked with the definite direct object, thereby retaining accusative 
> alignment.
> 
> I wonder whether anybody has come across similar constructions, in which 
> an incipient or apparent ergative case marking system is licensed by 
> number (rather than by more commonly-cited features such as aspect or 
> person).
> 
> Thanks,
> 
> David
> 
> 
> _______________________________________________
> Lingtyp mailing list
> Lingtyp at listserv.linguistlist.org
> https://listserv.linguistlist.org/cgi-bin/mailman/listinfo/lingtyp