[Lingtyp] non-agentive transitives

[Joshua Birchall]

Dear Ferenc,

In a classic paper, Tsunoda (1985) proposes a semantic hierarchy for case
splits with bivalent verbs:

EFFECTIVE (break, kick) > PERCEPTION (see, listen) > PURSUIT (search,
await) > KNOWLEDGE (know, forget) > FEELING (love, fear) >  RELATIONSHIP
(have, be similar to) > ABILITY (be good at, be capable of)

The further down the scale, the less likely a verb is to show a case
marking pattern typically associated with a transitive construction, such
as NOM-ACC or ERG-ABS. He downplays the role that agentivity has in splits
in case marking, stating that "it is always the affectedness of the
patient, rather than the volitionality/agency/agentivity of the agent, that
is crucial [in manifesting a transitive case frame]" (p. 393). He
illustrates this with an example from Nepali where 'hear' (non-agentive)
shows the same case marking pattern as a prototypical transitive
construction, whereas 'listen' (agentive) shows a different case marking
pattern. However, more recent works such as Witzlack-Makarevich & Seržant
(2017: 14-16), and references therein, show that agentivity can indeed be a
factor in some instances of differential argument marking, among others.

All the best,
Joshua



Tsunoda, T. (1985). Remarks on transitivity. *Journal of Linguistics*,
21(2), 385-396. doi:10.1017/S0022226700010318

Witzlack-Makarevich, A. & I. A. Seržant. (2017). Differential argument
marking: Patterns of variation. In Ilja A. Seržant & Alena
Witzlack-Makarevich (eds.),* Diachrony of differential argument marking*,
1–40. Berlin: Language Science Press. doi:
10.5281/zenodo.1228243

On Thu, Jan 30, 2020 at 6:38 AM Havas Ferenc <hfz at iif.hu> wrote:

> Dear Colleagues,
>
> My question is about the differential marking of agentive and non-agentive
> subjects in transitive sentences. It is well known that in some languages,
> called active or agentive, the marking of the subject of intransitive verbs
> (whether by a case or by agreement) differs depending on whether the
> subject is agentive or patient-like. As Marianne Mithun puts it: “ln all of
> these languages one case is used for semantic agents of most transitive
> verbs and the single argument of some intransitives while a different case
> is used for the semantic patients of most transitive verbs and the single
> argument of other intransitives. The sets of verbs occurring with each case
> are largely the same from one language to the next. Most verbs in the first
> set denote events performed, effected, instigated and controlled by their
> participants (’jump', 'go', 'catch'). Most verbs in the second set denote
> state significantly affecting their participants ('be sick', 'be tired',
> 'be caught’)”. (Language 1991, 67/3, 523)
>
> So much about subjects of intransitive predicates. Less light seems to
> have been cast on transitive subjects in the dedicated languages, though
> the pattern exists. Consider e.g. these Kaddoan sentences (selected from
> the same paper, 525–528):
>
>    1. *ci-hahyúnčah*  'I'm going to go home.’
>    2. *ku-táyʡayah* 'I'm tired, disgusted, fed up.'
>
> (3)  *ci-kíʡčah* 'I'm going to kill him.'
>
>    1. *kú-ʡnutah* ’I like it.’
>
>
>
> (1) and (2) show that agentive and non-agentive subjects of intransitive
> sentences have distinct verbal prefixes: *ci-* versus *ku-.*  (3) and (4)
> in turn illustrate the differential marking of agentive and non-agentive
> subjects in transitive sentences with the very same prefixes as
> in intransitive sentences.
>
> So my questions are
>
> a) WHICH SUBJECTS?
> Which subjects are non-agentively marked in transitive sentences? I would
> expect them to be passive experiencers (of verbs like ’see’, ’hear’ as
> opposed to ‘look at’, ‘listen to’), recipients (’get’, ’inherit’),
> possessors (’have’), undergoers of unintentional mental processes
> (’remember’, ’forget’), emotions (’like’, ’dislike’, ’hate’).
>
> b) UNIFORM MARKING?
> If a language has splits in both transitive and intransitive sentences,
> are the agentive and non-agentive markers of the same form in the two types
> of sentences?
>
> At the end of the day, the central issue is whether the agentive –
> non-agentive split does or does not work the same way in intransitive and
> transitive sentences. If it did, the mere “split intransivity” concept of
> agentive languages would be worth reconsidering.
>
> Thank you for your assistance.
>
> Ferenc Havas
>
> Professor Emeritus of Linguistics
>
> ELTE University, Budapest
> _______________________________________________
> 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/20200130/9138b5e6/attachment.htm>