xml-dev - Re: Begging the Question

["W. E. Perry" <wperry@fiduciary.com>]

Since David Megginson has introduce the quite valid analogy between our
fundamental specifications and constitutional law, I will extend it a bit in the
hope of casting more light than heat. As general practice, the federal courts
(and ultimately the Supreme Court) build upon earlier interpretations when
deciding constitutional questions. They are not, however, obliged to do so, and
the adaptability of the Constitution is based upon this fundamental mechanism of
extensibility:  the court may always return to the literal syntax of the
Constitution and, in the context of the case at hand, elaborate an entirely
different semantic outcome than achieved by the body of previous decisions on the
same constitutional questions. All of the great inflection points in the history
of U.S. law rest on such reinterpretations.

As I return once again to insist:  a node in a Semantic Web *may* rely upon
previously-agreed semantics when processing a new question, but
    1) that process is 'legally' constrained only by the syntactic rules of the
fundamental specification itself; and
    2) the business of elaborating the particular semantics of an instance is the
proper role of the processing node alone; and
    3) the salient semantic constraints on the outcome of that process are those
expressed, through that process, by the specific data or environmental
circumstances of the instance, just as the constitutional nexus in a judicial
interpretation must be solidly grounded in the facts of the specific case at
hand.

Respectfully,

Walter Perry


David Megginson wrote:

>   The U.S. president is head of state because the U.S. constitution
>   says so.
>
> In each case, while they're often ambiguous, self-contradictory, and
> underspecified, the XML Namespaces spec and the U.S. Constitution are
> normative documents, not simply unproven assumptions.  That doesn't
> establish that the it is natural or morally right that the
> U.S. president be (or not be) head of state, but it does correctly
> state that the president's position conforms to the specification that
> defines the United States in the first place.