[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message] Re: Noise-free Complex Systems
I have recently returned from the annual American Philological Association meeting where I said to all who might listen (not many): Eighty years after Milman Parry we have not begun serious work on an aesthetics of oral poetry--that is, dealing with oral poetry as what it is, formulae which exhibit properties of metrics, morphology and vocabulary. Instead we continue to predicate our aesthetics on comfortable abstractions which might describe 'poetry', but certainly nothing 'oral'. Despite claiming to embrace Parry's discovery, philology seems to have backslidden into the comfortable habits of aesthetics of the 18th and, especially, 19th centuries though these are abstractions elaborated from entirely different underlying stuff. Yet aesthetics is an activity; it is procedural; and it is performed in the case of text against concrete syntax. In 19 years of practice I have developed a practical aesthetics of certain commercial and financial documents: bond indentures, insurance policies, and some other specialized stuff. Using such aesthetics I can arbitrage bonds against insurance policies, but not in the general case. The aesthetic activity requires a process--in these examples, it is a parsing--of particular instances on a particular occasion and from a very particular point of view in order to establish the equivalence of 'chunks' from different environments, at least so far as is required to base an arbitrage upon that equivalence of those chunks in those particular circumstances. The output of such a parsing is never portable precisely because its nature is determined by the circumstances in which it is parsed, and portability is the attempt to move it out of those very circumstances. When XML came along it was easy to shift the accumulated experience of aesthetics in financial documents into XML 1.0 syntax and thereby to gain the considerable advantage of standard XML parsers and other tools. But that shift was possible at all because XML begins from the premises of the instance document and the instance parse. XML 1.0's innovation of well-formedness as opposed to validity emphasized this: that the syntactic integrity of the document was distinct from, and in XML 1.0's case took precedence over, validity as conformance to a content model, schematic, or other a priori abstraction. To speak of moving an infoset or of determining equivalence in the terms of infosets is nonsense because these 'infosets' are divorced from any concrete instance which might be transported or measured. Portable XML, on the other hand, and the determination of equivalence in concrete instances of XML are feasible and, indeed, are the functional raison d'etre for an appropriate aesthetics of XML documents. Arbitrage based on financial documents which schematically are radically non-congruent is both possible and profitable. I am convinced that in philology the resolution of all sorts of apparently anomalous problems awaits only an aesthetics appropriate for oral poetry. But I am convinced (not least by success in the financial application) that interoperability--of which arbitrage is an exquisitely refined form--requires an appropriate aesthetics for the documents which are to be processed, and any such aesthetics will necessarily be grounded in the concrete form of those document instances. Respectfully, Walter Perry "Bullard, Claude L (Len)" wrote: > It seems to circle back to the "If you depart from syntax as the means to > ensure portability of data, XML makes no promises about its > interoperability" position. A conservative and sane position, but it > seems to get in the way of relentless innovation because it comes down to > "If it hurts, don't do that!"
|
PURCHASE STYLUS STUDIO ONLINE TODAY!Purchasing Stylus Studio from our online shop is Easy, Secure and Value Priced! Download The World's Best XML IDE!Accelerate XML development with our award-winning XML IDE - Download a free trial today! Subscribe in XML format
|