ontolog-forum
[Top] [All Lists]

Re: [ontolog-forum] What goes into a Lexicon?

To: ontolog-forum@xxxxxxxxxxxxxxxx
From: "John F. Sowa" <sowa@xxxxxxxxxxx>
Date: Sun, 19 Feb 2012 13:09:50 -0500
Message-id: <4F413AEE.5060605@xxxxxxxxxxx>
Dear Matthew, John, and David,    (01)

In this note, I talk about a variety of issues.  For anybody who wants
to skip the details, please go to the paragraph at the end of this
note that begins "In conclusion, ..."    (02)

MW
> A lexicon is a list of words and phrases for a domain. There is
> no reason why there should not be more than one word or phrase
> with the same meaning. The word term is sometimes used to refer
> to the entries in the lexicon.    (03)

That is a good summary of the issues, but there is more to say.    (04)

JB
> The publisher, and maybe the author, says that the lexicon contains
> concepts. I had expected to see terms and maybe proper nouns. I don't
> understand how you can have a concept in isolation in a lexicon.
> It seems that if it does not have relations and constrains specified
> then it is not really a concept.    (05)

DP
> SKOS Concepts can have multiple labels and can be shared between/grouped
> into Concept Schemes (aka 'Lexicon' or 'Vocabulary').    (06)

Unfortunately, many people freely mix terminology from different fields
-- linguistics, logic, philosophy, psychology, computer science, and
whatever programming language they happen to use.    (07)

Fortunately, there are professionals who have done the "heavy lifting"
to define words.  They're called lexicographers.  And they publish
their findings about "word senses" in things called dictionaries.    (08)

I recommend that anybody who has a question about word usage should
"look it up!"  Check how the word is defined in a good dictionary.
For the word 'concept', we can find two word senses in the online
Merriam-Webster dictionary:    (09)

    1. something conceived in the mind:  thought, notion    (010)

    2. an abstract or generic idea generalized from particular instances    (011)

The M-W dictionary also gives the etymology:    (012)

    Latin /conceptum,/ neuter of /conceptus,/ past participle of
    /concipere/ to conceive — more at conceive.    (013)

    First Known Use: 1556    (014)

This entry also highlights the words 'thought', 'notion', and 'conceive'
in blue.  You can click on them for more information.    (015)

But it's always a good idea to get a second opinion from other
dictionaries and encyclopedias.  Since ontology is a branch of
philosophy, look at what the philosophers say.  Following is the
article about concepts from the Stanford Encyclopedia of Philosophy:    (016)

    http://plato.stanford.edu/entries/concepts/    (017)

Another important word is 'meaning':    (018)

    http://plato.stanford.edu/entries/meaning/    (019)

And don't forget 'ambiguity':    (020)

    http://plato.stanford.edu/entries/ambiguity/    (021)

Reading all of that will take a bit of time and effort.  For anybody
who doesn't have that much time, I'll state my suggestions:    (022)

  1. The best dictionaries and the consensus of the philosophers is
     that the word 'concept' is related to 'conceive' and to the way
     people think.  Since that topic gets into a huge amount of debate
     in philosophy, psychology, and neuroscience, it's better to chose
     a less controversial term for what words mean.    (023)

  2. The lexicographers, who are experts in defining words, use the
     term 'word sense' instead of 'concept' or 'meaning'.  It is also
     the term used in computational linguistics.  I recommend it when
     we're talking about word meanings.    (024)

  3. For formal ontology, definitions are stated in some version of
     logic.  Since there are many different notations for logic and
     different terms for describing them, the terminology will vary
     from one implementation to another.    (025)

  4. For a common terminology that is independent of any particular
     implementation, I recommend the usage by logicians and philosophers
     who are experts in both logic and ontology.  The consensus in both
     traditional logic up to the 19th century and in symbolic logic
     from the late 19th century to the present is to use the words
     'predicate', 'property', and 'relation':    (026)

     The M-W online says that 'predicate' is "a term designating a
     property or relation".  It also says that the word is derived
     from the Latin 'praedicatum', which refers to what is said about
     the subjectum or subject of a sentence.    (027)

In conclusion, the word 'predicate', as in predicate calculus, is
the consensus term for what is said about something.  As the M-W says,
that word designates "a property or relation".  Therefore, we can use
the word 'property' for what a monadic predicate designates, and use 
'relation' for what a predicate with two or more arguments designates.    (028)

When we want to talk about ontology independently of any particular
implementation, I suggest that we use the following terminology:    (029)

  1. A class is a set whose membership is specified by a monadic
     predicate that designates a property shared by every member
     of the set.    (030)

  2. Each word in a natural language has one or more word senses.
     Each word sense can be informally defined by a phrase in a
     natural language or formally defined by a predicate in some
     version of logic.    (031)

  3. A lexicon is a collection of words whose word senses are
     defined informally in a natural language or formally
     by predicates in some version of logic.    (032)

John    (033)

_________________________________________________________________
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    (034)

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