ontolog-forum
[Top] [All Lists]

Re: [ontolog-forum] The concepts can change?

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: Ed Barkmeyer <edbark@xxxxxxxx>
Date: Mon, 16 Jul 2012 12:02:00 -0400
Message-id: <50043AF8.6000203@xxxxxxxx>


Ali Hashemi wrote:
> Dear Matthew and Marcelino,
>
> In socially constructed domains, it may be the case that it is useful 
> to model a concept with changing definitions through time. 
>
> Take for example the medical domain and the definition of a disorder 
> X. In some authoritative medical journal or text at year YYYY, the 
> disorder may be defined as containing certain characteristics. In some 
> future edition of a similar authoritative text, what exactly counts as 
> an X may evolve. It would then be useful / valuable / instructive to 
> scope the definitions of X through time, and in so doing to capture 
> how the different interpretations of X have been applied. 
>
> Such a scenario is quite common in regulatory and legal domains. A 
> legal concept may be defined (and created) in some article of some 
> legislation and it may be amended at a later date. Arguably, you could 
> model this evolution as two distinct concepts, related by an amending 
> action, though colloquially, the notion of say, "Tax Payer" would 
> remain constant. Modeling the changing definitions of "Tax Payer" is 
> definitely of use to the intended consumer of said legal concepts and 
> from their point of view, there is only one concept under consideration.
>
> In such a scenario, the concept of "Tax Payer" was created for a given 
> jurisdiction due to a particular article / section of law, and the 
> meaning of this term may change according to other pieces of 
> legislation (or court decisions) as issued by the relevant legislative 
> authorities in said jurisdiction. Should this jurisdiction (or 
> geo-political entity) cease to exist, the concept of "Tax Payer" may 
> consequently also cease to exist.
>
> Best,
> Ali    (01)

Ali makes a point.  It is important to distinguish the idea that a 
concept may have different referents (a different extension) over time 
from the idea that the 'concept' itself may evolve over time.  Matthew 
addressed the 4-D solution to the "changing referents" problem -- make 
all referents 'states of things' rather than "endurant" things.    (02)

We also need to distinguish the evolution of a concept from the 
relationship between the term and the concept.  One can argue that the 
evolved concept is a new concept and the term has simply been "moved".  
This is not uncommon in some business and social practices, and is a 
critical feature of "buzzwords" -- the term is applied to a 
generalization that includes several vaguely related concepts that were 
formerly distinguished.    (03)

One can argue about what happens in the legal domain, but in the science 
and engineering domains one really does see evolution of a concept.  The 
problem is almost always that the scientific 'concept' is not exactly 
its definition.  Scientists sometimes refer to a 'working definition'.  
What is meant is that there is a set of phenomena that have something in 
common, and the concept is intended to refer to whatever that is.  
Initial observations lead to the working definition; further research 
refines that notion.  In a similar way, engineering folk distinguish 
between what they call 'concept' and 'design'.  The concept is the set 
of properties the new thing will have, more or less, and the design is 
the detailed specifications for its makeup.  In an engineering 
'concept', some properties are critical -- they are what creates its 
value -- but other properties may be important in creating market and 
distinction from competitors.  Those properties are 'softer' -- they can 
be traded off in the actual design, as long as the design meets the twin 
purposes of appeal and distinction.    (04)

The whole idea of concept evolution is that the original intent -- the 
characterizing properties that motivated the creation of the concept -- 
does not change but the formal intension -- the formal definition of the 
concept -- does.     (05)

A clear case from the NIST realm:  The international standard 'metre' 
was originally defined to be a particular fraction of the best 
measurement of the circumference of the earth at the equator in 1875.  
In practice, it was defined by a metal alloy bar that was kept in Paris 
and carefully cloned for each of the participating national measurement 
standards organizations.  In 1972, the nominal geographical definition 
was replaced by the wavelength of a readily reproduceable radiation, 
which happened to be the length of the metre bar but measured to many 
orders of magnitude greater refinement, and now independent of the 
temperature in the room and accidental deposition of airborne 
molecules.  The intent never changed.  Nothing that was 2 metres long 
became something different, but some scientific measurements of silicon 
deposition and crystal structure, for example, might actually change in 
the last decimal place or two, if the instruments were re-calibrated.  
Theretofore, however, those decimal places were a matter of disagreement 
among conforming measurement instruments.  The new definition enabled 
better measurement, by establishing a better reference.  The important 
idea here is that the concept -- the intent -- of the metre did not 
change, but its definition improved.    (06)

This is what Ali means when he talks about the 'social' aspect.  What is 
important is the intent and the expectations for the use of the concept, 
and the formal definition can change over time to support those purposes 
better.  Statistically, the change in definition may come to clearly 
include or exclude 'outlying cases', like the deposition measurements, 
but that almost always means that their former inclusion or exclusion 
was debatable.    (07)

We don't want to excuse "term migration" as "concept evolution", but we 
do have to recognize the existence of true concept evolution in science 
and engineering.  (And my concerns have nothing to do with 'social 
domains'.)    (08)

-Ed    (09)

P.S.  I have to say that I owe much of this insight to Eswaran Subrahmanian.    (010)

>
> On Sat, Jul 14, 2012 at 10:47 AM, Matthew West 
> <dr.matthew.west@xxxxxxxxx <mailto:dr.matthew.west@xxxxxxxxx>> wrote:
>
>     Dear Marcelino,
>
>      
>
>     What does it mean to say that concepts can change?
>
>      
>
>     MW: For a concept to be able to change would suggest a very
>     unusual usage of the term. Usually the whole idea of a concept is
>     that it refers to the same thing always.
>
>      
>
>     It is the case that the meaning of the concepts can change over
>     time? In this case, what is the meaning of meaning? How one can to
>     trace the identity of concepts over time, in order to judge that a
>     (same) concept was changed? What remains the same when occur a
>     change in a concept?
>
>      
>
>     MW: These would be excellent questions to ask anyone who suggests
>     that concepts can change.
>
>      
>
>     Or it is the case that each meaning is related to a single
>     concept? In this perspective, seems that when we say that a
>     concept was changed, in truth, we have two concepts: the previous
>     concept and a new concept (with a new meaning). 
>
>      
>
>     MW: This would be the usual usage.
>
>      
>
>     Does it make sense to say that a concept can cease to exists?
>
>      
>
>     MW: That depends on how you are using the word “concept”. If you
>     are using it as a synonym for class then it probably does not make
>     sense. But one of the usages of concept has it as the
>     representation of some thing in some particular human brain, in
>     which case presumably they cease to exist when the person dies or
>     forgets them.
>
>      
>
>     How these questions are related to the practice of ontology
>     engineering?
>
>      
>
>     MW: They questions you need to get out of the way.
>
>      
>
>      
>
>     Regards
>
>      
>
>     Matthew West                            
>
>     Information  Junction
>
>     Tel: +44 1489 880185 <tel:%2B44%201489%20880185>
>
>     Mobile: +44 750 3385279 <tel:%2B44%20750%203385279>
>
>     Skype: dr.matthew.west
>
>     matthew.west@xxxxxxxxxxxxxxxxxxxxxxxxx
>     <mailto:matthew.west@xxxxxxxxxxxxxxxxxxxxxxxxx>
>
>     http://www.informationjunction.co.uk/
>
>     http://www.matthew-west.org.uk/
>
>      
>
>     This email originates from Information Junction Ltd. Registered in
>     England and Wales No. 6632177.
>
>     Registered office: 2 Brookside, Meadow Way, Letchworth Garden
>     City, Hertfordshire, SG6 3JE.
>
>      
>
>      
>
>     Best regards.
>
>
>
>     _________________________________________________________________
>     Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
>     Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
>     Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
>     <mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx>
>     Shared Files: http://ontolog.cim3.net/file/
>     Community Wiki: http://ontolog.cim3.net/wiki/
>     To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
>      
>
>
>
>
> -- 
>
>
> (•`'·.¸(`'·.¸(•)¸.·'´)¸.·'´•) .,.,    (011)

-- 
Edward J. Barkmeyer                        Email: edbark@xxxxxxxx
National Institute of Standards & Technology
Manufacturing Systems Integration Division
100 Bureau Drive, Stop 8263                Tel: +1 301-975-3528
Gaithersburg, MD 20899-8263                Cel: +1 240-672-5800    (012)

"The opinions expressed above do not reflect consensus of NIST, 
 and have not been reviewed by any Government authority."    (013)


_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (014)

<Prev in Thread] Current Thread [Next in Thread>