On the impossibility of defining anything whatsoever (as Rob might have you believe) (Was: Re: New issue?: vocabulary maintenance))
thomas at astro.umd.edu
Wed Feb 6 10:35:36 PST 2008
As usual..no end.
This will be my last email on this matter.
On Wednesday 06 February 2008 12:52:18 pm you wrote:
> Yo, Bri,
> > I knew as I was writting my email that you would respond and
> > be positioned at the other side of the fence.
> I thought we were on the same side of the fence. You said, "better
> yet, just drop any definition at all and save the argument for a rainy
Only in matters of controversy or time needed. Getting a draft
of IVOAT would be paramount. Yes, in this regard we agree although
you would argue that all definitions are controversial and we
will never have enough time.
> > Let me ask this simple question : if the definition of astronomy
> > concepts (or the concepts of any field) are in general impossible
> > as you seem to indicate, then how do any text books get written?
> Not impossible. Expensive. Like I said, it took the OED 71 years, at
> which point they started all over again.
IMO, they are being overly pedantic. We are not the OED. We should not
try to emulate them. We are not trying for all encompassing definitions,
only short ones which capture the meaning of a concept for Astronomy.
The OED is trying (overreaching?) to capture the meaning of a word for
all possible domains. Our problem is quite a bit narrower (and simpler).
> Also, textbooks are written by individual members of the community,
> who are free to assert any definition they like. The IVOA must be
> responsive to the IAU (through commission 5).
Responsive only means that if there is a problem with a
definition, we listen to the IAU, and we fix it. It does not mean
that we don't try.
> Please advise, would the appended picture be considered a
> "Frankenstein rake"?
No no. Its a wooden thing. Think 18th century field rake (couldn't
find a picture). I don't think they had the modern item you attached
in Mary Shelly's time.
> > We are not looking for "complete" definitions, but rather, simple,
> > workaday ones which will serve the purpose of general identification
> > of a subject/concept.
> Like I said, what is the purpose of this exercise? What does "general
> identification" mean?
The purpose, primarily, is to sort out the meaning of a concept/token.
In my perusal of 800 definitions assigned by WordNet to the IVOAT
(an old version, admittedly) I found several degeneracies in meaning.
By assigning a short descriptive text we catch these problems.
> > I'll try a simple definition of gamma ray burst from the wikipedia:
> Would referencing external sources, i.e., via a URL to wikiville, be
> sufficient - whatever the purpose(s)? A collection of hyperlinks is
> equivalent to the OED editor's compendium of usage.
I don't know. Its for the group to decide. Reliance on transient
links seems to be a bad idea. Maybe one day the IVOA will
sponsor an "astro-wiki" and we could reference that. Until then
short descriptions in text which is part of the SKOS document
> > I would like to further add that if some people argue with this
> > definition..
> > so what?
> The definition of "planet" started a near uprising at the IAU.
Alright! Our work providing definitions should then set the planet
on fire then,. first time thats happened with anything I've done.
> > They may start their own vocabulary, -or-, they may engage
> > us to either fix the definition or to evolve the vocabulary.
> ...or they may regard the IVOA as arrogant and reject participation in
> virtual endeavors.
I doubt it as long as we are responsive to requests for changes.
> Could someone clarify whether each vocabulary is intended to have its
> own definitions, or rather whether all synonyms in the greater
> thesauri are intended to map to a single centralized unitary definition?
The definition of a concept is only in the scope of the given vocabulary.
Another vocabulary may choose to adopt that definition, or ignore it, as it
desires (any mapping between the 2 vocabularies would indicate this)
More information about the semantics