David Durand writes:
> It's just seems that the flexibility the TEI requires for describing
>existing documents is compatible with the kinds of prescriptive
>one might want for data creation.
True enough, and perhaps that is "good enough." It seems to me if that
there are going to be "new dictionaries designed for use with automatic
parsing systems," that a standard such as am I proposing is even more
necessary. If this is not the case, it is likely that the vendor producing
the new dictionary will likely be selling the only parsing system that it
is compatible with (ala spelling checkers). It is exactly this situation
that I am looking to avoid.
Keep in mind that I am not looking to propose a _prescriptive_ system --
even if a dictionary vendor has a more descriptive set of values for part
of speech (for example), that vendor could still provide the "standard" POS
value in addition to their value in exactly the manner you are describing.
Also, as it became obvious that many vendors were providing more specific
information, the standard could be revised to include them as optional
Keep in mind that we are looking at this from the point of view of a
dictionary "consumer" (rather than a dictionary vendor) -- we want our
customers to be able to choose from a variety of dictionaries. If there is
no standard, we cannot do this as it is unlikely that we can convince
dictionary vendors to adopt our proprietary standard; particularly if
another company creates a conflicting standard. Our customers will only be
able to choose from dictionaries that we license and modify (which is not
likely to be a large number).
Is this clear? We (Origami, Ltd.) would gain very little from such a
standard, but our customers gain a lot.