Many thanks to everyone who replied on- or off-list with suggestions and
reports on metrical tagging in TEI and its exploitation.
A little more background on what my colleague is thinking about: a group
of English scholars are considering creating a teaching tool that allows
for visual representations and other manipulation of metrically tagged
verse, including a programme to automatically indentify and therefore
tag metre and scansion to assist students in the interpretation of poetry.
This seems to be both a very promising and an insanely ambitious
project. My advice (which they've not really asked for) will be to focus
on the teaching tool, which could be very useful. As we have seen from
some of the responses on this list, visual representations and
searching/processing of metrical schemes can help castlight on texts in
all sorts of ways. If this tool also allowed live manipulation of the
tagged metre, for example, and a library of common metrical schemes and
patterns (perhaps even connected to a dictionary with word-stress
tagged), it could make for some very interesting class demonstrations.
Hand-tag a few syllables, and see which of the patterns in the library
fit the rest of the line.
With this as the core of the project, one could then gradually improve
the suggestions feature of this tool by the addition of a more
sophisticated library of metres and word or sentence stresses; of
algorithms for syllable detection; of weighting and recommending
multiple options rather than claiming to be able to automatically detect
the "correct" answer; of text mining and machine learning from a
hand-tagged corpus; etc. This is a serious artificial intelligence and
natural language processing task, so I think it should be seen as (a) an
e-Science project, and (b) a plug-in to the teaching tool rather than
the core upon which all else stands or falls.
Thoughts? Is this even relevant to TEI?
Stine Brenna Taugbøl a écrit :
> At Henrik Ibsen's Writings we have marked up the metrical structures of
> all Ibsen's verse texts (verse dramas and poems) manually, as we have a
> lot of irregular metres in our material.
> The metrical encoding will be used in textual presentations with
> information of metre, rhyme scheme, scansions and so on. We will also
> present statistics showing for instance use of metres in Ibsen's
> authorship and we will give opportunity to search for metres, rhyme
> schemes etc. directly in the texts.
> So far we have encoded the texts and made presentations of them (not
> released yet). We have not come so far as to develop search engines. We
> would appreciate hearing how similar projects are doing this.
> All the best,
> Stine Brenna Taugbøl.
> Henrik Ibsen's Writings
>> Dear all,
>> On behalf of a colleague who is exploring a new project possibility, I'd
>> like to conduct a quick survey on the recording and exploitation of
>> metrical tagging in TEI. Two questions therefore:
>> 1) What sorts of implementations have you or projects you know about
>> developed to make use of the metrical data encoded in your TEI texts?
>> Can you perform metrical searching? Perform stress-aware text-to-speech?
>> Etc.? What do you use your tagged metre for?
>> 2) Has anyone developed a tool or a script to interpret verse texts and
>> automagically tag the metre (with any degree of probability)? I expect
>> this would be relatively easy for Ancient Greek, a lot harder for
>> English. Does anyone have experience with work of this kind?
>> Many thanks in advance,
>> Dr Gabriel BODARD
>> (Epigrapher & Digital Classicist)
>> Centre for Computing in the Humanities
>> King's College London
>> 26-29 Drury Lane
>> London WC2B 5RL
>> Email: [log in to unmask]
>> Tel: +44 (0)20 7848 1388
>> Fax: +44 (0)20 7848 2980
> Henrik Ibsens skrifter
Dr Gabriel BODARD
(Epigrapher & Digital Classicist)
Centre for Computing in the Humanities
King's College London
26-29 Drury Lane
London WC2B 5RL
Email: [log in to unmask]
Tel: +44 (0)20 7848 1388
Fax: +44 (0)20 7848 2980