Minutes London 03-2009

From TEIWiki

Jump to: navigation, search

SEE DISCUSSION PAGE: Needs to indicate whose minutes, what sig, what subgroup, etc. preferrably in page-title/url

Present: Katrin Dennerlein, Paolo D'Iorio, Fotis Jannidis, Gregor Middell, Elena Pierazzo, Malte Rehbein, Moritz Wissenbach.

Discussion led by Elena Pierazzo, minutes by Gregor Middell.



Technical infrastructure/ Workflow

  • coordination takes place in this wiki
  • project has to be announced on the news bar at http://tei-c.org/
  • wiki will be reorganzied in 2 distinct parts: a public one (with announcements, draft specs …) and an internal one with information, that is less interesting to the general public
  • list of sponsors has to go on the wiki as well
  • Google Docs used for the (internal) specification process
  • draft versions of the specification are checked out as a PDF document and put on the wiki every week, so we can gather feedback

Conference in Paris

  • ITEM provides up to 3 rooms for the conference; definitely 1 on both days, possibly 3 on the second
  • Paolo D'Iorio coordinates the room allocation with ITEM
  • we want to offer lunch at the conference (attendants shall confirm their participation upon registration)
  • SIG will invite speakers and organizers to a dinner as well
  • Paolo D'Iorio organizes hotel, dinner bookings, internet access, beamer
  • Malte Rehbein manages registration
  • registration deadline is 17.04. (first come, first serve)
  • conference program will be split up into 2 sections:
    • a common program with presentations and public discussions
    • a workgroup-oriented program with talks, which are oriented toward certain aspects of the specification

Spending institution

  • TU Darmstadt/ Universität Würzburg (chair of Fotis Jannidis) will be spending institution
  • in case of any restrictions NUI Galway can serve as a workaround
  • funding will be provided by ALLC, ACH, TEI and Galway
  • Fotis Jannidis contacts sponsors as soon as the ultimate recipient of reimbursements is determined

Program: 1. day

  • start at 02:00pm
  • one coffee break
  • 2 sessions, 4 talks each, 15 min/talk + 5 min/discussion
  • (edition) project presentations

Program: 2. day

  • start at 09:00am
  • present the SIG work: 45 min
  • 10:00am: discussion
  • 11:00am: coffee
  • 11:30am: discussion
  • 01:00pm: lunch
  • 02:00pm:discussion, wrap up, roadmap, outlook
  • finish at 03:30pm

Technical support

  • we want to offer an optional web publication service, so attendants can refer online to all the examples during discussion [not necessarily online: important is that the examples, i.e. images etc. are easily accessible during the discussion, so one powerpoint presentation including all images would do the job --Malte 13:33, 25 March 2009 (EDT)]



  1. a presentation of the draft specification at the conference
  2. the XML schema
  3. a contribution to the TEI guidelines

Theoretical framework

  • the specification shall be independent of presuppositions made by a particular theoretical framework
  • therefore a couple of typical dichotomies in editorial theory have to be recognized
  • first there is the notion of fact (or representation) vs. interpretation
  • maybe we should think of differing levels of interpretation instead
  • reason: we might not be able to cleary differentiate between „what's there“ (document/fact) and „how does it relate“ (text/interpretation)
  • this leads to the second opposition, that is central to editorial theory: document vs. text
  • Paolo D'Iorio explains the HyperNietzsche approach to handle this opposition
  • HyperNietzsche bases its functionality on a thorough manuscript description, that happens on the documentary level
  • the interpretative acts (constituting a text) build upon the manuscript description
  • to summarize the HyperNietzsche experience while trying to adopt TEI guidelines: the Text Encoding Initiative does not handle "the document level" very well thus far
  • for genetic editions though, this level is crucial


  • we want to propose a standard, we do not want (even better: we cannot) prescribe one, given the complex landscape of editorial practices
  • in the beginning we will work on a pseudo-encoding, which is tightly bound to our own terminology
  • in a second step, we will try to map this pseudo-encoding to the actual TEI encoding framework, probably in cooperation with someone from the TEI Council
  • the same step-by-step approach shall be taken for the standardization process
  • first we develop an application profile aka. TEI customization
  • when that proves to be of general interest, we rework it into a separate chapter for the TEI guidelines

Aspects of Genetic Editions

Topological description on the document level

  • the description functions hierarchically on 3 levels:
  • highest level: the page
  • pages contain zones
  • zones are are nestable, groupable, can overlap and have a depth level (aka. belong to a layer)
  • zones are arbitrary in as much as they can be defined by layout and/or semantic aspects
  • on a third level, zones normally contain (among other entities) lines


  • on the one hand the dating of a witness may not be directly related to its encoded content, it therefore has to go into the header
  • on the other hand, dates added by the original author as metadata to the text on the witness should be encoded without necessarily going into the header [can be embedded within a Functional Mark]
  • dating can be justified by prose and/or by reference to a characteristic of a manuscript (e.g. hand)
    <dating ....>

Time/ Chronology

  • we want to express time as absolute and relative measures
  • express absolute time by adapting the existing date element/attributes; see http://www.tei-c.org/release/doc/tei-p5-doc/en/html/ND.html#NDATTSda
  • express relative time as predecessor/successor-relationships
    • might be a direct relationship: next, previous
    • might be a „sorting“ relationship: before, after
    • relationships may also be expressed via a numbering scheme
    • relationships may have 1:n cardinality, e.g. one witness is dated before a set of others

Grouping changes (“changesets”)

  • forming changesets is necessary to group text passages, which all form one revision
  • implemented either via a wrapping element (e.g. <revision/>) or via a stand-off mechanism for spatially disparate changes


  • alterations are practically related to the collation of texts
  • we need to deal with collation as a separate aspect of genetic editions
  • for example we need an expression for ommissions as a textual feature, that results from relating texts via collation
  • another set of alterations to be expressed is the one, that results from authors fixating a text passage by overwriting it (e.g. a penciled passage fixated with ink)
  • marginal notes have to be differentiated from functional marks commanding an alteration (e.g. „move this passage over there”)


“marked as used”

  • passages happen to be striked through, which marks the passage to be used at another location
  • two-step markup
  • first: markup the block on the document level
  • secondly: describe the function (“mark as used”)


  • overwriting is a special case of a deletion

Undoing alterations

  • how to express on the document level, that an alteration has been taken back, e.g. a striked through passage being reinstantiated via a dotted underlining
  • <restore/> already exists for undoing an deletion, but we might need a more general approach
  • possible solutions via an element or via an attribute “undo”
  • proposal: add a generic element <undo/> with an attribute referring to the action being undone
  • can be further differentiated by a type attribute, e.g. immediate correction vs. revising
  • this is preliminary: we have to open up this solution to discussion


  • current approach: express transpositions as pairs of additions and deletions
  • we also want to express transpositions as an atomic operation, because it might be expressed as such on the document as well
  • first: make the segment to be transposed adressable
  • either via an identifiable wrapping element or via a “cross-cutting” <seg/>
  • secondly: markup “functional mark”, which might be in the text (e.g. a superlinear “add ‘is’ here”), with an element like <fm/>
  • relate the functional mark to the identified passages
<add><fm type="addition">add</fm> ‘is’ here</add>
  • either refer to the passage added via attributes
  • or it might be implicit, which passage is transposed, because the <fm/> is contained in an <add/> or <delete/>
  • maybe we can generalize the idea of a functional mark (<fm/>)
  • we markup passages on the representational level, but also on the semantic level, on which they command alterations


  • think of clarification as a repetition of a text
  • this idea can also offer a solution for encoding fixations
<repetition type="clarification|fixation|..." position="superlinear" hand=""/>


  • supply an additional attribute to <add/> and <delete/> that typifies the alteration (e. g. “instant correction”)

“Coordinate system”

  • how to express genetic relationships on an inter- and intra-document level
  • we need a scheme for addressing the linked passages and a means for describing such relationships/ links
  • adopt the HyperNietzsche concept: relational description via paths
  • paths are typed, depending on the relation expressed (e.g. a timeline, a conceptual link …)
  • paths: define complete paths via a set of vertices (steps)
  • <linkGrp/> is a candidate for grouping a set of steps forming a path
  • steps of a path can be ordered or unordered, but that might belong to the aspect of uncertainty)
  • the steps of a path should be describable, so the expressed relationships can be justified or further characterized
  • we should import concepts from the critical apparatus here: http://www.tei-c.org/release/doc/tei-p5-doc/en/html/TC.html
  • relationships might be transitive in nature, should we express that as well
  • then: is there a potential scope conflict between the TEI and other graph-oriented markup languages like RDF/XML
  • maybe we can differentiate between the description of the relationships itself and their characterization/justification
  • the relationships might be described as an RDF graph, which is argumented with TEI-based prose


Organizational tasks

  • conference announcement: Malte Rehbein
  • call for more examples on the TEI list: Elena Pierazzo
  • apply for a panel on the next TEI Member’s meeting

Specification tasks

  • as we were not able to deal with all aspects, the further work on the remaining ones is distributed among the SIG members
Aspect responsible
Alternatives Elena Pierazzo
Substitution Malte Rehbein
Gaps Fotis Jannidis
Editorial Decisions Paolo D'Iorio
Uncertainty Malte Rehbein
Critical Apparatus Fotis Jannidis
Textual Constitution Paolo D'Iorio

Upcoming events

Date/Time Event
21.04. 09:30am CET conference talk: content-related issues
05.05. 09:30am CET conference talk: organization-related issues
11.05. deadline for publication of draft specification to the conf. participants
14./15.05. Conference in Paris
16.05. post-conference meeting
16.06. 09:30 CET conference talk: specification
07.08. 09:30 CET conference talk: specification
16./17.09. Wuerzburg: finalize specification (place might change; depends on funding)
Personal tools