‘Act of Directive Communication’ definition mistakenly requires hearer

[gregfowlerphd]

One can directively communicate (e.g., advise or order something, to take just a few of the examples given in the rdfs:comment) in a non-auditory manner, so ‘hearer’ seems mistaken. Better to use ‘audience’, I suspect.

[gregfowlerphd]

I think there are other issues with this class--in particular, with the relation between it and its subclasses--that might necessitate other revisions to the definition. So revising now would be premature.

(@cameronmore: Tagging you here since you gave the original post a thumbs up, suggesting you agree about the problem stated and perhaps about its solution.)

[swartik]

Recognizing what @gregfowlerphd just said about delaying revision, I still want to mention that the Oxford Language Dictionary defines "audience" as follows:

The assembled spectators or listeners at a public event, such as a play, movie, concert, or meeting.

That interpretation of "audience" isn't much better than "hearer". The following definition for Act of Directive Communication, although a little obscure, possibly is closer to the intended meaning:

An Act of Communication that is intended to cause one or more Agents to receive a Directive Information Content Entity prescribing some Planned Act.

"Prescribing some Planned Act" covers "take a particular action" in the current definition. However, from the examples in the rdfs:comment, I'm not clear whether an Act of Directive Communication always intends to take an action. For example, is "instructing" a synonym for "ordering", or is it what a teacher does to students?

[gregfowlerphd]

I'm not sure what to think @swartik's point concerning 'audience'. I agree that the sense in which I was using it is a semi-technical one that doesn't match the ordinary sense captured by the OLD definition he quotes. However, 'audience' is already used in that semi-technical sense elsewhere in the Event Ontology. See the definitions for Act of Personal Communication, Act of Deceptive Communication, and Act of Inviting. (Moreover, its other uses in the Event Ontology also don't seem to conform to the OLD definition.)

In my previous comment, I noted that there are other issues with the Act of Directive Communication class that might necessitate further revisions to its definition:

An Act of Communication that is intended to cause the hearer to take a particular action.

I'll discuss these below.

First, not all Acts of Advising--which, by definition, are 'performed by providing advice or counsel to another agent'--seem to be intended to cause the "hearer" to take a particular action. A student's faculty advisor might lay out for the student multiple paths for the student to choose between based on their interests, availability, etc.

Second, Acts of Requesting (which are 'performed by asking for something') are subject to similar concerns. Suppose I ask you to accompany me to the movies or to coffee. In that case, there's no particular action I intend to cause you to take, but it would seem I've still performed an Act of Requesting.

Third, a related problem arises for Acts of Warning (which are 'performed by indicating a possible or impending danger, problem, or other unpleasant situation'). Indeed, in the typical case, an act of warning is not be intended to cause the "hearer" to take a particular action. Instead, it's intended to cause them not to take a particular action.

Thus, given the current definition for Act of Directive Communication, it would appear that Act of Advising, Act of Requesting, and Act of Warning aren't subclasses of ADC, contra the asserted subclass relations. (Is the same true of the fourth asserted subclass, Act of Commanding? Perhaps, we accept the possibility of disjunctive commands.)

(@cameronmore: Tagging you here for the same reason as before.)

[swartik]

Well, this issue of what "audience" means points to a larger problem. If an ontology exists to give standard meaning to a set of concepts, and those concepts are supplemented by natural-language explanations, then do we also agree there should be standards on how to write the explanations? CCO already has some syntactic standards (e.g., "A class that elaboration"). Should CCO also state a definitive source to use for the meanings of words in explanations?

I don't want to fight battles over whether "audience" is acceptable. When I think of "audience", the first thing I think of is people in an auditorium, receiving audio. You can retort with any number of counterexamples, from silent movies to John Cage's 4'33" for piano. And those aren't even figurative interpretations. I'm only hoping we can come up with a way to adjudicate whether a word is acceptable, along with helping CCO users interpret definitions.

[neilotte]

@gregfowlerphd The issues raised in this thread require a more comprehensive review. For this reason, I am converting this into a discussion where we can work to identify a project plan to implement decisions made here. Thanks!