[Lingtyp] realis: definition

[Juergen Bohnemeyer]

Dear Vladimir — I hesitate to offer a definition that’s explicitly typological (for reasons I’ll mention below). But in unpublished work presented in various places (and versions), I’ve worked with a definition of realis mood for Yucatec according to which it locates the topic world inside (i.e., as part of) the utterance world. This assumes an ontology in which worlds are maximal spacetime entities that “grow” into the future, meaning their own past, but not their future, is part of them. 

This effectively makes realis mood very similar to non-future tense, with the crucial difference that, unlike a non-future tense, it cannot be used in past counterfactual contexts. I show that this prediction is borne out for Yucatec.

Here is a version of the talk that specifically focuses on the distinction between realis mood and nonfuture tense:

http://www.acsu.buffalo.edu/~jb77/Bohnemeyer_2019_tenselessness_Lisbon.pdf

And here is an earlier talk that focuses on the conflation of mood and viewpoint aspect in inflectional paradigms in Mayan languages:

http://www.acsu.buffalo.edu/~jb77/Handouts/TLS_JB_v1.pdf

Now, the reason I hesitate when it comes to endorsing the above definition for typological purposes is that the “growing worlds” model will cause the entire past of the utterance world to be realis and none of the future. But I’m pretty sure there are languages in which the remote past is treated as irrealis, and others in which parts of the future (or perhaps we should say certain types of future time reference) are treated as realis.

For typological purposes, I would suggest that a language has a realis-irrealis contrast if it classifies utterances in terms of whether they are (purported to be) about the real/factual world or not. What counts as “real” varies somewhat from language to language, but from my understanding of the typological literature, the crosslinguistic prototype seems to be the speech situation and at least part of its past, with the likelihood of inclusion perhaps decreasing with distance from the speech situation. 

Now, when it comes to distinguishing between ‘realis’, ‘declarative’, and ‘assertive’, in practice, there is of course a great deal of overlap in how these are used. That said, I would reserve the terms ‘declarative’ and ‘assertive’ for mood categories that classify utterances exclusively by speech act, i.e., all assertions/‘representational’ speech acts (in Searle’s terms) occur with the declarative, regardless of whether they concerns the real world, a future world, or some (present, past, or future) counterfactual scenario. 

Lastly, I should mention that Manfred Krifka recently developed a very different approach to the semantics of the realis/irrealis contrast in Daakie or Port Vato (Oceanic, Vanuatu):

https://journals.linguisticsociety.org/proceedings/index.php/SALT/article/view/26.566

As I recall, Krifka’s proposal is based on presuppositions.

Best — Juergen 



> On Jul 20, 2021, at 4:04 AM, Vladimir Panov <panovmeister at gmail.com> wrote:
> 
> Dear colleagues,
> 
> I would be grateful you could recommend me works which provide typological definitions of the realis. Crucially, I wonder how it is distinguished from the assertive speech act or the declarative.
> 
> Best,
> Vladimir
> _______________________________________________
> Lingtyp mailing list
> Lingtyp at listserv.linguistlist.org
> http://listserv.linguistlist.org/mailman/listinfo/lingtyp