Mariage

I've marked up a further six images today; they were all fairly busy ones, with some linking between them, and took longer than I expected. I have nine to go.

Working through the images, marking up objects, people and other bits and pieces, I'm beginning to put together the basics of the Lustucru story. This is how it seems to work:

Lustucru nowadays seems to mean a fool or a buffoon, but the figure of Lustucru in these images is not like that. He's basically the Opérateur Céphalique; working in what looks like a blacksmith's shop, with appropriate tools (furnace, bellows, anvil, hammer, tongs, and file), he and his assistants remove and replace the heads of women, who are brought in by their husbands to be "improved". A cruder version of this workshop appears in Jean Tangous.

Outside the shop, ships bring a cargo of new heads up the river; these are unloaded and brought into the workshop by other workers. The shop appears to advertise itself by hanging out the decapitated heads of women, and by a sign showing a decapitated body.

This "normal" state of affairs -- men having their women "fixed" to tame their natures -- is inverted in L'invention des femmes (1) and (2), where the women force their men to undergo operations which similarly render them more malleable and cooperative. Interestingly, these operations do not involve actual decapitation; instead, a (male) surgeon seems to perform more delicate work on the patient's skull, using a knife. Following the operation, these men wear the symbol "A", as a "marque de bon mari".

However, in several engravings, typified in La grande destruction de Lustucru par les femmes fortes, and also shown in the two Massacre de Lustucru pictures (here and here), the women take direct revenge on Lustucru and his assistants. They invade his workshop and attack the men, subjecting them to the same decapitation, throwing their heads into the furnace, and dragging Lustucru's body down to the riverside. They also set fire to the ships which bring heads to the workshop.

In both the Massacre pictures, there are male bystanders who mildly regret what is happening, and provide a link to the Confrerie des Martyrs; one says that "la confrerie des Martirs sera plus grande" as a result of the massacre.

I have only about fifteen images to go, so I should be finished with this first pass through objects, people and creatures tomorrow. I'm noticing that hats on the floor are a symbol of defeat; hats in hand symbolize submission or importuning; and hats on heads, especially large ones with prominent feathers, show command and virility.

Made the changes that needed to be made based on FL's report.

Pushed the latest XML files into the db, and then noticed that apostrophes in title attributes in the output were being double-escaped. Edited teiGeneral.xsl to remove escaping from the output of <reg> elements as title attributes. We'll have to watch this, though; the escaping may have been there for a reason. Not clear whether the XSLT processor will handle escaping appropriately or not.

I've got more than half-way through now, although some images (Fornaux, for instance) are difficult to interpret, which makes it hard to identify Personnages. I'm noticing that bâtons, éventails, marteaux, balais and clefs are very important.

After an hour or so of looking at the first few images, it became clear that we also needed a category called "Personnages", so I've added both the Créatures and Personnages categories to all of the image files (using XSLT). I've also worked through a dozen or so of the images, adding markup of objects, creatures and people (often empty placeholders, but enough for searching purposes), and moving other annotations from categories they didn't belong in (such as people in the Gestes category).

Found and corrected a lot of typos in the process, but CC is taking the images I've finished with and doing a proper proofing of them, also adding the typographical regularization tags discussed last week.

Also found a couple of annoyances in the IMT, and determined that I really want the UniSynEdit syntax highlighting that was planned for version 1.8, so I think I'm going to add that today and tomorrow (assuming I don't hit any horrible roadblocks with it).

CC and I met to plan the next couple of weeks of work. These are some of the decisions:

Transcription

• At the base level, transcription of text on the images should be diplomatic. At present it is a mixture, with some regularizations; these should be undone, then re-done methodically (see below).
• Superscript and smallcaps need to be included, using <hi> elements, and then reproduced in the rendering.
• Typographical features should be regularized using a <choice> element, containing <orig> and <reg>. Three features to which this will apply are:
  • long s
  • u which is typographically identical to v
  • tilde used to indicate an abbreviation (typically of a word suffix)
• These <choice> elements should be applied at the word level, so that searching can more easily handle variants.
• A modernized version of the text should also be created. This would be in a separate category, but both would be (in the coming TEI version) linked through @facs, as opposed to @corresp. (A new version of the IMT will handle this, in a month or two. In the meantime, category ids will be sufficient anyway.) The modernized transcription would be keyed to the original using identifying attributes (probably @n).

Object markup

We began working on a plan for formalizing, focusing and regularizing the markup of objects, creatures, weapons etc. in the images:

• Make sure gestures are moved from the Objets to the Gestes category. Where a gesture incorporates an object, use two annotations.
• A new category is needed for Créatures.
• The title of an annotation must include, unambiguously, the identifier we're using for that species of object.
• Where the object is used as a weapon (a common feature), it should be titled like this: "Baton comme arme".
• The following items are already of special interest, and must be marked up wherever they appear:
  • corne
  • crochet (link to corne; symbolic)
  • clé
  • balai
  • chaise (especially as weapon)
  • verre
  • pichet
  • bourse
  • argent
  • pot
  • seau
  • croix
  • marque (badge -- see for example Invention)
  • chapeau
  Other items of interest will certainly emerge; the key is consistency in marking (especially titling) particular items.

We'll work next week in production-line mode: I'll add and regularize object markup, while CC proofs, corrects and diplomatizes transcription (and probably expands on my object markup by adding some content to the annotation texts).

Anticipating that at least one of us, and hopefully both I and CC, will be doing lots of markup next week on objects in the files, I've installed the IMT version 1.7.1.1, the latest, on the two markup computers.

I've finished a new stylesheet for the site - it's green and goes by the name Pelouse.
I've also begun cleaning up the sheets so that colours only get declared in <colourname>.css files. All other sheets only deal with layout, typography etc. This keeps the site looking the same in terms of layout no matter the colour scheme.
I'll finish the de-crufting of layout sheets next and begin working on a brown sheet (Bois?) after that. The last step will be tidying up the original and versailles sheets.

In order to run some of the text-analysis operations we've been planning on the database, we need to be able to grab text-only output from the documents, and be able to specify various types of filter on that output. I've been working on that today. These are the main points so far:

• I've written an XQuery file called text.xq which does the extraction, based on input parameters (see below), and returns plain text. The file has this custom eXist option declared in it:

  declare option exist:serialize "method=text media-type=text/text encoding=utf-8";

  That means I can avoid any transformations at all, which is handy.
• I've set up a page here which presents a simple form from which you can choose the options you want.
• These are the parameters:
  • startYear
  • endYear
  • textProse
  • textVerse
  • imageProse
  • imageVerse
  The first two are date filters, but the other four are booleans which decide whether prose or verse from either of the two types of document is included. Other filters behind the scenes ensure that only transcriptional text is returned; editorial notes, metadata, regularizations, etc. are all excluded.

As usual, XQuery hacking took a long time, but it appears to be working well now, and it's faster than I expected. The user will now be able to get the text in the browser, and feed it as-is into an application, or supply the URL of the filtered text query to an online service such as the TAPoR Tools.