Term Request: Act of Refrain

[PhiBabs935]

I would like to request / propose the term 'Act of Refrain' be added to the Event Ontology.

I am currently curating the term in an ontology whose scope is deterrence and compellence, broadly construed (with classes such as Act of Deterrence). I have defined Act of Refrain as follows:

              Act of Refrain =def A Planned Act in which some Agent avoids or stops themselves from participating in some Planned Act.

               Source > https://dictionary.cambridge.org/us/dictionary/english/refrain

I am currently using this class in the modelling of uses cases for acts of deterrence and their effects. The idea is that the point of deterrence is to convince another agent to refrain from some Planned Act. For example,

The implementation of sanctions against country X (an act of deterrence) is cause of gov't of country x refraining from developing WMDs.

Obviously, 'Act of Refrain' has broader applicability than deterrence, e.g. Shane refraining from sweets for health reasons. For that reason, I am suggesting inclusion in CCO.

[APCox]

This term turns a negative into a positive - that is, it turns a lack of doing X into an Act of Not-Xing. As the band Rush said, "If you choose not to decide, you still have made a choice."

Act of Refrain is designed to be generic, which is good. If Act of Refrain gets added to the Event Ontology, it would be a good idea to add a scope note stating that it should not be further subtyped or it could easily result in duplicating the entire Planned Act hierarchy.

With that in mind, I'm interested in knowing how you use this class to model the Planned Act that the agent refrained from doing. Do you use the Modal Relations Ontology or another approach? I anticipate that some users will get confused about how to use this term unless clear guidance is provided for it.

[PhiBabs935]

Hi Alex,

I agree on having a scope that it is not meant to be subtyped. Rather I only envisioned this being used to populate instances of act of refrain in a knowledge graph.

I had not fully ironed out the modeling, but I was using classes from the Cognitive Process Ontology, and the Modal Relations Ontology. Specifically, 'Decision Alternative'
"a decision alternative is a classification that indicates some entity is to be considered as a possible choice in a decision described in a decision process plan​" --which in CPO is a kind of Nominal ICE, though I wonder if it should be a directive ICE
and 'Selection Decision Specification'
"A decision specification that specifies that a decision process has identified one alternative to be pursued from a set of possible alternatives.​" Which descends from Directive ICE
And 'Decision', which is a mental process in CPO

The idea was that the agent is the carrier of instances of Decision Alternative, one which is a nominal measure of a Planned Act A, the other a nominal measure of an Act of Refraining from Planned Act A (using modal versions of the relation).
Furthermore, the agent participates in an instance of Decision D that has output a Selection Decision Specification that (mro) prescribes the Act of Refraining from A. Then that decision would in turn be a cause of the Act of Refraining from Planned Act A.

I wonder if there is a simpler way to do it though? (Note that the above involves the instance Act of Refraining from Planned Act A being linked to a directive ICE via mro prescribed and also linked to an instance of decision via non-mro 'is cause of'. So, there is an instance that is both a possible instance and an actual instance in the graph, and not sure if that really works.)

[PhiBabs935]

What is under 'discussion'? How the class is being used to model the refrained from act (per Alex's comment)? Or whether this will be added?

[PhiBabs935]

Is this term still in consideration for a future release? I assume this was moved to discussion because of the question of how it would be used in modeling efforts (see above). But is that separate from the question of inclusion? Is there anything else to be done on that front besides the scopenote? @APCox @johnbeve

[PhiBabs935]

Ah, I see now that in the original issue this was updated to 'For 2.2. release'. Sorry, I had not seen that until now.

[PhiBabs935]

Here is a (WIP) use case illustrating how to model deterrence with an act of refrain alongside the corresponding merely possible course of action that is refrained from:

[PhiBabs935]

@APCox

One thing I wanted to note r.e. subclasses of Act of Refrain. I can see a case for limited subclasses of Act of Refrain in certain domains. For example,

"Act of Suspending Hostilities" is a useful class for the domain of military conflict and diplomacy. For instance, where after the declaring a ceasefire, the parties to the conflict refrain from further aggressions toward each other [in line with a ceasefire agreement]. One can then model an Act of Declaring a Ceasefire [a class I am currently curating as a subclass of Act of Commissive Communication] as having output a Process Requirement that requires an Act of Suspending Hostilities. E.g., the declaration of a ceasefire between Israel/Hezbollah has output a Process Requirement that requires both to participate in an Act of Suspending Hostilities.

[alanruttenberg]

It's difficult to understand the diagram because there is no legend. Which things are classes, which are instances? What does the dotted box mean? What defines the anonymous class (and why are you using an anonymous class?).
Also note, as a matter of graphic design, it is difficult to read white lettering on a light color.

As an example of my difficulty understanding , you use instance-of links to relate to orange ovals on the lower end suggesting maybe you mean to have purple mean class and yellow instance, but then you have something as subclass of act of buying which is yellow, so that must be a class. Or the realization link from disposition to buy, if the relata are instances it would be wrong, since the disposition isn't realized, since the title says there was no buying.

I suggest a legend that explains: What shapes mean, what the colors mean, what the link colors mean. and what the links mean, e.g.are the links intended to mean instance-instance relation or do they mean something like a class axiom such as "every act of buying realizes some disposition to buy".

[PhiBabs935]

Yeah, I need to re-upload the diagram. I have already adjusted it since the original posting. Will respond more fully tomorrow.

The diagram is meant to reflect the strategy for representing merely possible entities defended by John Beverley and Barry here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5370912. It's an alternative to the modal relations ontology strategy for representing the merely possible.

I am piggybacking off of John's visualization (of a different kind of case) here: https://www.youtube.com/watch?v=rXEsc0dDdsA&t=518s
(The motivation there being that an artifact spec that doesn't have any existing instances conforming to it isn't really about any instance of artifact.)

[PhiBabs935]

Thanks for the suggestions