Willard McCarty <MCCARTY@vm.epas.utoronto.ca> (02/23/90)
Humanist Discussion Group, Vol. 3, No. 1086. Thursday, 22 Feb 1990. (1) Date: Tuesday, 20 February 1990 2338-EST (18 lines) From: KRAFT@PENNDRLS Subject: Warning: ChiWriter - Megawriter Split (2) Date: Tue, 20 Feb 90 22:46:26 EST (17 lines) From: "Steven J. DeRose" <IR400011@BROWNVM> Subject: SGML and hypertext (3) Date: Tue, 20 Feb 90 22:50:11 EST (31 lines) From: "Steven J. DeRose" <IR400011@BROWNVM> Subject: Annotated e-texts, retrieval (1) -------------------------------------------------------------------- Date: Tuesday, 20 February 1990 2338-EST From: KRAFT@PENNDRLS Subject: Warning: ChiWriter - Megawriter Split Although I am unable this moment to put my hands on the actual pieces of paper somewhere in this room, a communication arrived just this week to the effect that ChiWriter, on which MegaWriter is/was based has recinded (or words to that effect) its agreement with Paraclete Software for the latter to develop and market MegaWriter. Various reasons are given. In any event, if anyone out there wants details, I'm sure the letter will resurface at some point. Or simply contact ChiWriter, which seems committed to developing and supporting the multilingual capabilities along the lines of MegaWriter's designs. I don't know where this leaves MegaWriter. Bob Kraft (U. Penn) (2) --------------------------------------------------------------25---- Date: Tue, 20 Feb 90 22:46:26 EST From: "Steven J. DeRose" <IR400011@BROWNVM> Subject: SGML and hypertext Much as I appreciate Michael's posting, I would make 2 adjustments: First, I think "great sage" rather overstates my stature in the hypertext community (though I'm working to get there!). Second, regarding the quote: "SGML and hypertext: two great tastes that taste great together." This should be attributed to another Humanist, namely David G. Durand. I fully support his sentiment, and quoted it in a paper at Hypertext '89 (which may be the source of the association). David's sagacity in SGML and Hypertext has helped many, myself included, to see the issues much more clearly and deeply. Steve DeRose (3) --------------------------------------------------------------39---- Date: Tue, 20 Feb 90 22:50:11 EST From: "Steven J. DeRose" <IR400011@BROWNVM> Subject: Annotated e-texts, retrieval In response to Pieter C. Masereeuw: Substantial work has recently been done with both tagged corpora, and the automatic tagging of corpora. (Also, the Brown Corpus is still much in use, e.g., for my dissertation last year) But I don't know of any full-featured programs for the kind of search and retrieval sought, except perhaps GramCord, which has (so far as I know) been applied only to Biblical texts. The features you described are basically the extensions of everyday search tools to ***hierarchical*** documents. For example, in most texts sentences and words are demarcated, but not discourse units above the sentence, nor elements smaller than words, such as morphemes. Any scheme which represents these levels should allow annotations at all levels. One program I know supports word- and morpheme-level annotations ("IT", for PC and MAc, from the Summer Institue of LInguistics, Dallas TX); no program I know supports annotations at an unbounded set of levels, though SGML provides the necessary syntax. Committees of TEI are working on some of these issues, and would no doubt welcome your input. In conclusion: no, there's no magic bullet, but it's a very saleable tool if someone builds a good retrieval engine for annotated, hierarchical texts, based on SGML. Steve DeRose