W3C home > Mailing lists > Public > public-prov-wg@w3.org > October 2011

Re: vocabulary simplification: two proposals to vote on [deadline, Oct 26 midnight, GMT]

From: Reza B'Far (Oracle) <reza.bfar@oracle.com>
Date: Wed, 26 Oct 2011 11:25:03 -0700
Message-ID: <4EA8507F.5030109@oracle.com>
To: public-prov-wg@w3.org
Jim -

I'm not saying that earth's whether system is an agent.  To your examples -

 1. If there is a compound that is changing because there is a chemical
    reaction, then the forces that are causing the chemical reaction are
    the agent
 2. If there is a weather system that is changing, then the forces that
    change the whether system are the agents.

I'm not sure where you get the _*/fact/*_ that "it's simply not 
correct".  Can you give me a reference please?  Just because you say 
it's not correct, it doesn't mean it's not correct.  I can give you a 
dozen references where this is defined as a "System Actor" or "System 
Agent".  Your representing a subjective argument an objective one.

This is a problem that OMG struggled with and UML, IMHO successfully has 
an extension for.  I've sent out the references for it before.  There 
are non-human agents that can do stuff.  They don't have to be 
intelligent.  There is nothing that says an Actor has to be intelligent 
or that there needs to be intention.  Being an Actor is more about 
causality from a control systems perspective.

So, I reiterate:  There are Agents that cause weather system changes.  
Chemical forces as agents (and at the micro-level one of the four basic 
forces) cause the spilled code on the desk to eat through the desk.  The 
actual word Agent is used in pharmaceuticals, physics, and a bunch of 
other fields in this same exact way: indicating causality, whether 
intentional or not, whether intelligent or not.

On 10/26/11 11:11 AM, Jim McCusker wrote:
> You're talking about documenting the agent of an observation of a
> natural event, I'm talking about the documenting the actual natural
> event itself. To call earth's weather system an agent is really
> stretching the definition, though.
>
> If we want to restrict PROV to just things that have agency, then
> that's fine, but that misses out on lots of things that don't, and
> it's simply incorrect to say that every event has an actor.
>
> Jim
>
> On Wed, Oct 26, 2011 at 12:49 PM, Reza B'Far (Oracle)
> <reza.bfar@oracle.com>  wrote:
>> Jim -
>>
>> I disagree.  The fact that there is a process whose agent is not well known,
>> or that that agent's behavior is not deterministic (random, chaotic,
>> probabilistic, or some other type) doesn't mean that there is no agent.  I
>> would argue that in a weather modeling system, there are 1 or more agents
>> that encapsulate the algorithms that model the whether behavior given a set
>> of stimuli.
>>
>> I would claim that there is ALWAYS an agent and that this can be resolved
>> via Agent "typing" as requested on a different thread versus saying that
>> there is no agent.
>>
>> Best.
>>
>> On 10/26/11 9:42 AM, Jim McCusker wrote:
>>
>> No, there isn't. Stellar formation doesn't happen because of specific
>> agency, it just happens as an effect of gravity and having the right
>> mass in the right place at the right time. Things happen all the time
>> that have no agency - weather is a perfect example.
>>
>> Jim
>>
>> On Wed, Oct 26, 2011 at 11:27 AM, Luc Moreau<L.Moreau@ecs.soton.ac.uk>
>> wrote:
>>
>> Hi Jim,
>>
>> That's what my OED says:
>>
>> A condition in which things are happening or being done.
>>
>> Is there a notion of agency when we say "things are happening"?
>>
>> Luc
>>
>> On 10/26/2011 03:56 PM, Jim McCusker wrote:
>>
>> Then not Event. But I think a key goal of our work is to find terms
>> that align best with the intended usage. This makes it much easier for
>> people who are coming to the model for the first time. It's perfect
>> that we've started with concepts, but these concepts are being
>> grounded in terminology, and that should align with the chosen,
>> default language.
>>
>> If someone can give me a counterexample where an act or activity
>> doesn't have an implied actor, I'll withdraw my negative vote.
>>
>> Jim
>>
>> On Wed, Oct 26, 2011 at 10:40 AM, Paolo Ncl
>> <paolo.missier@newcastle.ac.uk>    wrote:
>>
>> Hi,
>>
>> I think we are in fact reading a bit too much into this. The intent was
>> to simplify and harmonize the key terms used in the model.  an agent may
>> play a part in the activity, and we do have a way to express that, but that
>> doesn't have to be (does that mean we cater to eastern cultures as well? :-)
>> )
>> But I strongly advise against using the term "event" to refer to
>> activities that have a time duration. Events already have a clear role to
>> play in the model, and have no duration.
>>
>> Thanks, Paolo
>>
>> Sent from my iPad
>>
>> On 26 Oct 2011, at 15:05, Jim McCusker<mccusj@rpi.edu>    wrote:
>>
>>
>> On Wed, Oct 26, 2011 at 7:33 AM, Luc Moreau<L.Moreau@ecs.soton.ac.uk>
>>   wrote:
>>
>> JimMcC indicated that activity implied a notion of agency. I am not
>> familiar
>> with this
>> interpretation. Where does it come from? He suggests 'event', but this
>> term
>> is already in
>> the document (and will be the subject of a future clarification
>> proposal).
>>
>> Activity (http://www.merriam-webster.com/dictionary/activity) is
>> defined as a quality or state of being active. If you look at the
>> examples at MW, all of them have some sort of agent or actor. There is
>> one natural process example, which is that a volcano is active. Even
>> in that case, the volcano is being considered an actor (which is fine
>> in discourse, but isn't technically correct). The root word, "act",
>> when used, requires an actor. An act can happen with an unknown actor,
>> but there is always an entity that is behind an act.
>>
>> Using this word to describe all events (including natural events),
>> especially formally in a standard, gives the model a pre-scientific
>> bais (the idea that a prime mover is needed, because all events are
>> acts). Note that this is actually a western bais too, as many eastern
>> traditions do not require a prime mover.
>>
>> Maybe I'm reading far too much into this, but if we're looking to
>> simplify, I would far prefer Event or Process (but with a clear
>> explanation that it is a occurrent, not a specification of an
>> occurrent) to Activity.
>>
>> Jim
>> --
>> Jim McCusker
>> Programmer Analyst
>> Krauthammer Lab, Pathology Informatics
>> Yale School of Medicine
>> james.mccusker@yale.edu | (203) 785-6330
>> http://krauthammerlab.med.yale.edu
>>
>> PhD Student
>> Tetherless World Constellation
>> Rensselaer Polytechnic Institute
>> mccusj@cs.rpi.edu
>> http://tw.rpi.edu
>>
>>
>>
>>
>>
>> --
>> Professor Luc Moreau
>> Electronics and Computer Science   tel:   +44 23 8059 4487
>> University of Southampton          fax:   +44 23 8059 2865
>> Southampton SO17 1BJ               email: l.moreau@ecs.soton.ac.uk
>> United Kingdom                     http://www.ecs.soton.ac.uk/~lavm
>>
>>
>>
>>
>>
>>
>
>
Received on Wednesday, 26 October 2011 18:25:50 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 13:06:46 GMT