[
Lists Home |
Date Index |
Thread Index
]
"Manos Batsis" <m.batsis@bsnet.gr> writes:
> Take RDF for example, (AFAIK and even worse, last
> time I checked) where literals cannot be subjects in
> a sentence (triple). This essentially has to do with
> the RDF model; a literal there is just that; what if
> the system was able to uniquely identify this literal
> and refer to it with a *unique* URI; all implemented
> features applicable to resources would be available
> for literals as well.
Sounds like you're ready to be converted to the Topic
Maps way of seeing things. In Topic Maps, every, uh,
piece of addressable data can be seen from two
different perspectives:
(1) as a *subject-indicating* piece of addressable
data, or
(2) as a *subject-constituting* piece of addressable
data.
The address of every piece of addressable data can
thus be used as a binding point for two different, uh,
"subjects" (as in "subjects of conversation"):
(1) the subject that the piece of data compellingly
"indicates" to a human being when the piece of data
is rendered (together with its context) for human
perception, or
(2) the subject that *is* the piece of data itself,
considered as a piece of data, regardless of its
meaning to machines or human beings.
The context in which the address of the piece of data
is uttered establishes which of the two subjects is
being addressed. In the XTM syntax for Topic Maps,
(1) the address contained in a <subjectIndicatorRef>
element refers to the subject that is *indicated*
by the piece of data that's being addressed, and
(2) the address contained in a <resourceRef> element
refers to the subject that *is* the piece of data
that's being addressed.
Note: In Topic Maps land, the word "resource"
means "piece of addressable data". "Subject"
means "subject of conversation" -- anything that
anyone might ever want to talk about. This
contrasts strongly with the Web world's usage,
in which "resource" sometimes confusingly means
"subject of conversation".
In Topic Maps land, a subject that *is* a piece
of addressable data is called an "addressable
subject", while all other subjects are called
"non-addressable subjects". However, in Topic
Maps, even non-addressable subjects can, in
effect, have addresses; the addresses are the
addresses of the pieces of data used to
"indicate" the non-addressable subjects.
Now to your point. Having explained all that, what
does it have to do with literals?
In Topic Maps, every literal is, by definition, a piece
of addressable information. Like every other piece of
addressable information, it can be used as a binding
point for two different subjects:
(1) the subject that the literal compellingly
"indicates" to a human being when the piece of data
is rendered (together with its context) for human
perception, or
(2) the subject that *is* the literal itself,
considered as a piece of data, regardless of its
meaning to machines or human beings.
If the literal is a name, its context indicates that it
is a name. When the literal is addressed as a subject
indicator, the subject being addressed is the name. In
Topic Maps, a name is an abstraction that can be
equally compellingly indicated by any number of copies
of same literal, and/or by other literals that may use
different character sets, etc.
If, on the other hand, the literal is addressed as a
subject constituter, then the subject being addressed
is NOT the abstract name that the literal indicates.
Instead, the subject is those particular literal data,
right where they are being addressed, irrespective of
the fact that they can be interpreted as a indicating a
name, and irrespective of their context (except in the
sense that its context makes that particular copy of
that piece of data different from all other copies of
that piece of data).
Now, you may ask, how do you address the subject of
conversation that is being named by a literal that
happens to be a name? Well, that's a completely
different thing! To do that, you utter the name
literal itself in the context of the appropriate
reference-resolving system. Depending on the notation
in which the name is expressed, and the intent of the
utterer of the name, such a reference-resolving system
could be the Web, or a human librarian, or the
pushbuttons on his telephone, etc. If and only if the
name is the name of an addressable piece of data (and
if, for example, you happen to be using the XTM
syntax), then you can utter the name in the context of
either a <subjectIndicatorRef> or a <resourceRef>,
depending on what you're really trying to talk about:
the data, or what the data indicate.
-- Steve
Steven R. Newcomb, Consultant
srn@coolheads.com
voice: +1 972 359 8160
fax: +1 972 359 0270
1527 Northaven Drive
Allen, Texas 75002-1648 USA
|
