Terminology of convenience: group and situation


#1

I’ve encountered a short section on wiki introducing terms situation and group (http://moin.delph-in.net/ErgSemantics/Basics in the Terminology of convenience section). I am working with similar concepts and was curious if these definitions are elaborated on anywhere else?


#2

Hi @EwaM!

As far as I can reconstruct, those definitions are not elaborated on anywhere else. I think I can say they are largely my fault: we put them in because I had the sense in writing some of the pages for specific phenomena in the ESD wiki space that I wanted to be able to refer to those things — and didn’t want to have to resort to calling in syntactic concepts like “(rest of the) NP”, as those pages are meant to describe the ERSs themselves and not focus on the process of composition.

On searching, it appears that we don’t actually use the word group in any of the phenomenon pages yet, and only use situation on the vocatives page. I’d be really interested to hear more about what leads you to want terms for similar concepts! If you develop something more well-founded than what we have, I expect it will make sense for us to adopt your terms.


Side note on searching the wiki: the wiki software itself doesn’t provide facilities for searching subspaces within the wiki, but Google does. In case it’s useful for future reference, these are the searches I used:

site:moin.delph-in.net/ErgSemantics group
site:moin.delph-in.net/ErgSemantics situation


#3

Thanks for the quick reply, @ebender!
I am working on how to semantically chunk large DMRS into smaller parts that are easier to process and for which the results of processing can be put back together to yield a valid full-sentence analysis. As part of that, I’m identifying subgraphs that are semantically self-contained to some extent, trying to tie my working definitions to existing theoretical concepts such as propositions, situations, event, etc.

I also know that Matic Horvat used semantic subgraphs in his hierarchical realization/translation thesis/paper, and they are basically the groups and situations from the wiki page. As far as I know, he used them as practical objects and didn’t develop any specific terminology or theory for them.


#4

Thanks for the further info :slight_smile:

I wouldn’t want to claim that our notions of situation and group are theoretical concepts … we just felt the need (or at least I did) to have a way to refer to at least those two types of subgraphs, and didn’t want to have to keep saying things like “subgraph centered on an EP with an ARG0 of type event”. I’ll be very interested to see whatever you learn about possible theoretical grounding!


#5

I put a lot of thought into these classifications of semantic fragments, but in my thesis I never committed to a specific terminology. Like Matic, I also used “semantic subgraph” since we often refer to the representations (particularly dependency ones) as graphs, but I was not quite satisfied with that term because it is not precise enough. I ended up using phrases like “connected” or “rooted” subgraphs to describe well-formed fragments. I was steered away from terms that might conflate syntactic notions with semantic ones, so “semantic phrase”, “semantic constituent”, “semantic head”, “verbal subgraph”, etc. were out.

Regarding the terms used in the wiki, “situation” seems appropriate, although I don’t find “group” very intuitive. Emmon Bach (e.g., “On time, tense, and aspect: An essay in English metaphysics”, 1981) introduced “eventuality” and “individual” which we already use (though, as the wiki says, we often use “instance” instead of “individual”). Hana Filip (“Nominal and Verbal Semantic Structure: Analogies and Interactions”, 2001) uses “eventuality descriptions” for subgraphs centered on an eventuality, with “basic” and “non-basic” eventuality descriptions for those with all arguments satisfied, or not, respectively. Others have called these “atomic” and “non-atomic”, respectively, but I think that would be confusing when talking about semantic composition (i.e., that an “atomic eventuality” could be composed of smaller parts). Hana Filip also uses “eventuality description modifiers” as the semantic analogy of adverbs, prepositions, etc.

Maybe we could talk about “simple”, “complex”, and “complete” eventualities as those with none, some, or all of their arguments satisfied, and another term (“situation”?) when also considering modifiers of that eventuality. I’ve also heard Stephan refers to semantic subgraphs as “constellations”, and I think the term has precedence in the literature, and it has potential for further analogies to astronomy (e.g., orbit, satellite). There’s also precedence for using physics terminology, e.g., valence and atomic, but maybe that’s already been colonized by syntax.

In any case, I’m curious to see what you end up with (terminology and theory). Keep us updated :slight_smile: