(unknown project)

Proposal for a TEI Module for Correspondence

Table of contents

About this document

This document proposes the addition of a TEI module for correspondence. It is the product of a Correspondence SIG task force (Marcel Illetschko, Sabine Seifert Peter Stadler) with valuable input from various people.

The aim of the proposal is to define a common set of correspondence-specific meta data, brought together at one place in the TEI header. The lack of special guidelines for encoding letters, postcards, etc. with TEI P5 has been a recurring topic since well before the formation of the Correspondence SIG in 2008. Additionally, there is a growing demand of correspondence projects for linked data facilities. While some of this correspondence-specific information can already be encoded in various sections of the TEI header we believe it to be of greater value for projects dealing with correspondence material to have this information in one place. For the encoding, we suggest two approaches within this set:

A guiding idea has been to only tentatively modify the TEI standard by merely adding a few elements, wrapping them in a special purpose element <ct:correspDesc> and injecting the latter into <profileDesc>. Please note that <ct:correspDesc> is not intended to and can not replace the whole TEI header.

These efforts are supposed to be just a first step in the Correspondence SIG's strive for a complete model of encoding correspondence material.

Conventions used

As required for TEI conformance, non-TEI elements are defined in a distinct non-TEI namespace. In the usage examples and throughout this document that namespace is mapped to the prefix ct: (for ‘correspondence task force’), while TEI elements are not marked by any namespace prefix.

1. Preceding work

The development of <ct:correspDesc> was inspired mostly but not exclusively by the work of DALF as well as the WeGA.

The DALF project, the Digital Archive of Letters by Flemish Authors and Composers from the 19th & 20th century, was launched by the Belgian Centre for Scholarly Editing and Document Studies (Centrum voor Teksteditie en Bronnenstudie) and is one of the best documented projects in the area of correspondence encoding. In 2003, specific DTDs and guidelines for the description and encoding of modern correspondence material, both intended for use with TEI P4, were published.[1] The practical realization as well as the main features for encoding meta data are discussed in [2]. In August 2013 the DALF guidelines were reworked for the new requirements in TEI P5 as a ‘Preliminary P5 Proposal’; that version served as an important inspiration for this proposal.

Before DALF updated its customization from P4 to P5, the Carl-Maria-von-Weber-Gesamtausgabe (WeGA, Carl Maria von Weber – Collected Works) had developed project-specific guidelines for encoding correspondence material in TEI P5[3] and published recommendations in June 2011.

Both projects define a special wrapper element for correspondence meta data while the depth of its structure and its child elements differ slightly. In each case this wrapper element (<dalf:letDesc> and <wega:correspDesc>, respectively) is part of the <sourceDesc> element. DALF introduces the child element <dalf:letHeading> in which they encode the author and addressee of the letter as well as place and date when it was written, and next to that some other child elements that contain additional information, e.g. the formal classification of the letter, or the presence of an envelope or illustrations. The WEGA, however, groups this information directly within the wrapper element: sender1 and addressee, the place where and the date when the letter was not only written but also received, incipit and position within the thread of correspondence (korrespondenzstelle). Both projects decided to leave all other information on providence, physical appearance, history, etc. in <msDesc> where there are already sufficient TEI elements with which to encode them.

In this proposal, we mainly considered the overlapping of these two projects’ encoding guidelines and merged them into one.

2. Theoretical Considerations

Usually, letters (or any other pieces of correspondence) are considered as having a double nature as object and event.

The aspect of objecthood predominantly, but not exclusively, is tied to materiality. The facets of this materiality are manifold, but they do not differ much between letters and other manuscripts and therfore need not to be discussed here because the TEI Guidelines have elaborated instruments for describing them, especially within the manuscript description <msDesc>.

More important than questions of objecthood and materiality is the aspect of eventness (in a wide ontological sense): Correspondence does not mean merely written messages. Very often the surrounding circumstances - sent works, presents, works of art, other oral messages - show that the letter is embedded within an ensemble of communication media, which enriches the importance of the letter in itself. Sometimes there may be no actual letter at all - still ‘correspondence’ happens.

Hence, the basic assumption in this document is the understanding of correspondence in general as an event - namely an of communication - connected to a document and/or an (electronical) text. The topos of correspondence being a ‘half dialogue’, a ‘conversation amongst absentees’ is very often found in theoretical considerations about epistolography, also ‘the letter’s function as a connector between two distant points, as a bridge between sender and receiver’ (Janet Gurkin Altman). This concept of communication (sender, transmission, receiver) can not yet be expressed with existing TEI elements though. To answer the questions WHO WRITES? - TO WHOM? - WHERE FROM? - WHERE TO? - ANSWERING WHAT? - AT WHAT DATE? - and (very often unknown) AT WHAT DATE IS THE LETTER BEING RECEIVED? a correspondence module needs to provide information about persons (or organisations) as sender, receiver or messenger as well as about the respective dates and places. Additionally, one needs mechanisms for pointing at (or: referencing) preceding and subsequent messages.

2.1. What is correspondence-specific information (in terms of the TEI)?

With this concept of correspondence = communication in mind a letter (or any other piece of correspondence) has different participants:

Very often these bits of information are evident in the text of the message, in address lines, postmarks, electronically generated sending meta data, etc.

Furthermore a particular message usually is sent in order to trigger an answer (in form of one or more pieces of correspondence). A single piece of correspondence therefore is normally not a secluded entity but a (written) act of communication in a communication continuum, in which it is defined by its relative position between messages sent ‘before’ and ‘after’. Dates can be helpful – but finding out the correct chronology is part of the editing process, and should be given in a specific element.

2.2. And what is left out?

Information concerning the manuscript description, e.g. (postage) stamps, seals, water marks, form fields, letter heads, and incipits, are not considered part of the proposed correspondence description. These aspects are not necessarily correspondence-specific but rather characteristic for manuscripts in general and are thus best encoded within <msDesc>.

As this proposal deals with the meta data of correspondence, aspects relating to correspondence-specific transcriptional encoding within <text>, e.g. with <postscript>, <opener>, <closer>, or <signed>, are not considered.

3. Implementation of the correspondence description

The <ct:correspDesc> element provides a detailed description of correspondence-specific metadata. The <ct:correspDesc> element focuses on the communicative function of correspondence, i.e. sender, addressee, and transmission. It is provided as a supplement to the physical description of a correspondence object which is to be encoded within the <sourceDesc>. The description is organized by the following elements

These elements are members of a model class called model.correspDescPart; additional features may be defined by defining new elements and adding them to that class.

A simple example might be encoded as follows:2
<ct:correspDesc    xmlns:ct="http://wiki.tei-c.org/index.php/SIG:Correspondence/task-force-correspDesc">  <ct:correspAction type="sending">   <persName>Adelbert von Chamisso</persName>   <settlement>Vertus</settlement>   <date when="1807-01-29">29 January 1807</date></ct:correspAction>  <ct:correspAction type="receiving">   <persName>Louis de La Foye</persName>   <settlement>Caen</settlement></ct:correspAction>  <ct:correspContext>   <ref type="prev"    target="http://tei.ibi.hu-berlin.de/berliner-intellektuelle/manuscript?Brief023ChamissoandeLaFoye#1">Previous letter of Adelbert von Chamisso to Louis de La Foye:      16 January 1807</ref>   <ref type="next"    target="http://tei.ibi.hu-berlin.de/berliner-intellektuelle/manuscript?Brief025ChamissoandeLaFoye#1">Next letter of Adelbert von Chamisso to Louis de La Foye: 07      May 1810</ref></ct:correspContext></ct:correspDesc>

3.1. Actors, places and dates of correspondence

The very basic information about correspondence consists of the participating actors, their respective locations and dates (of writing/sending and reception). While there is a semantic difference between the sender of a message and the author of the corresponding text, in most cases those two will be identical. To reduce the redundancy of markup and error-proneness, standard linking mechanisms can be applied to connect the <author> element within <titleStmt> to the appropriate name within the <ct:correspAction> element.
<fileDesc    xmlns:ct="http://wiki.tei-c.org/index.php/SIG:Correspondence/task-force-correspDesc">  <titleStmt>   <title>Letter of Weber to his wife. Dresden, January 1,      1817</title>   <author xml:id="author1">Carl Maria von Weber</author>  </titleStmt> <!-- … -->  <profileDesc>   <ct:correspAction type="sending">    <name role="sendersameAs="#author1"/> <!-- … --></ct:correspAction>  </profileDesc> </fileDesc>
In other cases, there might as well be a clear distinction of the intellectual author of the text and the sender of the message, be it in disguise or frankly. Examples include pre-formulated greeting cards, forced dictated messages, or copied love letters.
The detailed information about names, places and dates of each participant are to be encoded with the standard TEI elements provided by the namesdates module described in 13 Names, Dates, People, and Places under the respective <correspAction> element. It is possible to differentiate those declarations further with the @type attribute, e.g. to distinguish the intended place of the addressee (as noted on the address) from the actual place of reception; or to capture several places of writing when the letter was produced over several days at several places. If an (external) placeography or personography is present or authority files are available it is also highly recommended to identify those names via unique URIs.
<ct:correspDesc    xmlns:ct="http://wiki.tei-c.org/index.php/SIG:Correspondence/task-force-correspDesc">  <ct:correspAction type="sending">   <persName ref="http://d-nb.info/gnd/118520040">Adelbert von Chamisso</persName>   <placeName ref="http://d-nb.info/gnd/4835139-8">Vertus</placeName>   <date when="1807-01-29">29 January 1807</date></ct:correspAction>  <ct:correspAction type="receiving">   <persName ref="http://d-nb.info/gnd/116057203">Louis de La Foye</persName>   <placeName ref="http://www.geonames.org/3029241">Caen</placeName></ct:correspAction> <!-- … --></ct:correspDesc>
As with other information compiled within <ct:correspDesc>, the spatial information is supposed to be editorial, normalised data, not necessarily the exact reproduction of the original address which should rather be encoded within <text>. However, it is possible to provide more address-like detail:
<ct:correspAction    xmlns:ct="http://wiki.tei-c.org/index.php/SIG:Correspondence/task-force-correspDesc">  <address>   <orgName type="Company">SubMarine Telegraph Company</orgName>   <street>58 Threadneedle Street</street>   <settlement>London</settlement>   <postCode>EC</postCode>  </address></ct:correspAction>

4. Future Work

After dealing with correspondence-specific meta data, the next steps are:

Appendix A Bibliography

  1. Edward Vanhoutte, Ron Van den Branden: Describing, transcribing, encoding and editing modern correspondence material: a textbase approach. In: Literary and Linguistic Computing 24,1 (2009), pp. 77-98.
  2. Anne Bohnenkamp, Elke Richter (Ed): Brief-Edition im digitalen Zeitalter. Berlin:DeGruyter 2009
  3. Klaus Hurlebusch: Divergenzen des Schreibens vom Lesen. Besonderheiten der Tagebuch- und Briefedition. In: editio 9 (1995), pp.18-36.
  4. Hans-Gert Roloff (Ed.): Wissenschaftliche Briefeditionen und ihre Probleme. Berlin: Weidler 1998
  5. Winfried Woesler: Vorschläge für eine Normierung von Briefeditionen. In: editio 2 (1988), pp. 8-18.
Notes
1
The CIF format is not part of the official feature request but will be maintained by the SIG. As of November 2014 the work on CIF has been postponed until the official request is in a more stable state.
1
In regard to the person who wrote the letter or other piece of correspondence, there is a conceptual difference. Thus, it is author in DALF and sender in WeGA.
2
The namesdates module needs to be included to support this encoding.
TEI Correspondence SIG. Date: 2014-11-19