[A2k] Re: [A2k] Re: [A2k] Uwe Müller: Academic Research o n OAJ

[Claude Almansi]

Hi, Arif, Manon and All

Between your lines, Arif

On Wed, Apr 1, 2009 at 12:18 AM, Arif Jinha <arif@stratongina.net> wrote:
> (...)
>
> this looks incredibly interesing and important to my own thesis.   M=
y congratulations to the author on their dissertation, and for investigating such an important subject at the root of what we do.  Unfortunately, I am a hopeless (only partly bilingual in French) anglophone and do not speak German.

It does indeed look very important. While I have good passive
knowledge of German, the whole dissertation is 269 p. long, though, so
I'll need some time to read it ;-)

>
> So.  I would like to try and experiment.  I want to discover, a=
s an issue of A2K, whether it is possible for someone, or a group, with a good knowledge of the SUBJECT MATTER would be able to interpret the automated computer translation of a service available free to everyone with an internet connection - such as Google Translate, and produce a VERSION of the article, retaining the meaning and the data but with some flexibility of wording to provide a natural language style in English.  I think it is possible, with some help, to do so and am interested in how quickly it can be done this way. But of course, that proceess needs review before a version could be published.
>

I'd be game to take part in your experiment. As a human translator, my
take on computer translation is that it is - so far - more
time-consuming to edit one into something readable than to do it
"humanly" from scratch. Better to make people aware of what they will
and won't get from a text when it is computer-translated (1).

However computer translation is progressing. So your proposal is
interesting, especially in its collaborative aspect. With some
colleagues, we created a "collaborative translation" online social
network where we also discuss computer translation issues,
occasionally. If you wish, I could ask them if they'd be interested.

> (...)
>
> So as an A2K research question - will it be possible to re-mix research a=
rticles by versionizing translations in this fashion?  can we have Peer-Review of such versions - clearly with needed input from professional translators and people who have knowledge and know either language, people with a senstivity to the idioms and idiosyncracies of language, and the appropriateness of interpretations in terms of language style and culture?  The experiment I suggest to start asks whether it is possible for some without knowledge of the originating language to create the version from the automated.  I think it will depend on the interpretative, intuitive, creative skills of the person doing the version, AND the progress of the semantic intelligence of the automated translator.

As a translator, I first balked at the word "re-mix" . But if you mean
working together on a translation and keeping all stages, then all
stages of further adaptations, I think it is feasible - as with any
other collectively elaborated text (2). And then there are tools for
collectively translating the interface of a computer application that
might perhaps be adapted for "normal" translation (3).

>
> Would such a process accelerate the access to knowledge across any sets o=
f languages? What is required to create high-quality versions, and how much latitude should there be in interpretation for best quality? Is a presence of the voice of the interpreter acceptable if a work is considered a version, rather than  translation?

"Provided the difference between version and translation is clearly
stated" is the official answer. But in practice, the difference is not
as clear cut. There is an element of interpretation in all
translations, because of the different ways in which different
languages represent reality and ideas, and the different ways in which
each individual uses each language

One more argument in favor of collaborative translation, keeping
different stages accessible (wiki-like).


> What are the implications for IP, copyright and attributions in the dimen=
sion of versions vs. translations?

IP, copyright and attributions are issues already involved in "pure"
translation (which is a derivative work anyway), where you must get
the right holder's authorization, have a proper contract as to sharing
possible revenue etc. In the case of texts under an open license, cc
for instance, things are a bit simpler.

> How much can be learned from the debates in poetry and literature, and in=
deed religious texts, regarding versions/translations?

Heaps from the  translators themselves  - much less from pure
translation scholars, in my experience. I.e., for instance After Babel
by George Steiner (see <http://en.wikipedia.org/wiki/After_Babel>) is
a very seminal book because he is also a very good translator (and he
always collaborates with his translators).

>
> The possibilities seem exciting.  I would like to see if Journals in=
 the future can use web 2.0 to create 'Translation-Version Queues' whenever an article is published, where communities of people with knowledge of the subject matter work collaboratively on versions starting with computer translation, and where the 'queue' ends with a type of peer-review process, and then the version is published.  Rules could be created as to how much could be introduced in the version, or how close to the original it should follow, depending on the perspective taken and permissions.  In the process, technology could be used in such a way to provide human feedback to web 3.0 semantic intelligence of computer translators.  It would be exciting to see where such things as medical textbooks could be versionized, so that not only language style and idiom is appropriate, but new content is introduced that is relevant to local culture, geography, society, health systems and resources.  Where the publisher uses a Creative Commons Attribution license and original authors are credited, the creator of the version also should be attributed, but most importantly, we move beyond Access to Knowledge to Transformation of Knowledge.

Fascinating - re rules, though, I think they should be agreed on case
by case between the original author and the derivative authors.

>
> The A2K/IP questions that arise are whether Creative Commons licensing ca=
n allow for this kind of re-mix culture in academic research across languages, through this use of technology and creativity, and whether we can have a process where unlike re-mix culture in music, there is a peer-review of mulit-lingual versions.


>
> I would like to try and versionize this dissertation in English, for my o=
wn edification in terms of the content and to see how this experiment would address those questions, and to make that work available in English as a version.
>
> Any suggestions?

I'd start with a collective translation rather, then go on to
versionizing. If you use a wiki or wikioid app, or even with a decent
desktop text editor, you will then be able to keep track of all
departures from the as-close-as-possible translation

>
> You can also see the very interesting 'self-referential' aspects of this =
kind of research, where a kind of meta-scholarship arises.  I am very interested in this, and some compelling dimensions that arise in the philosophy of mind - books like Emperor's New Mind and Godel, Escher, Bach.  One can also apply theoretical views communications mediums and refer to works of Marshal and MacLuhan.  These all need attention in the contemporary digital age, but for me I am referring to English works mainly and so we are back to the frontiers of translation.
>

English-speaking scholars tend to keep closer to the reality of the
things or processes they write about than French - or Italian - ones,
for instance. And to use simpler language ;-)
But yes, the approaches of the Russian formalism, the  "Prague
school", and of the German "esthetics of reception" school might
perhaps be relevant too.

> The future of philosophy should be thoroughly shaken up by the communicat=
ions age, but as yet I see some disconnect between our technocratic, evidence-based and policy-oriented approach and seeing this opportunity for discussion of the rontological and epistemological underpinnings of what we discuss - knowledge, and the implications of far greater diversity and far greater cross-language, cross-cultural interaction.  Those philosophical questions ought to be at the root of the question of quality, access and diversity in the accreditation of knowledge from a global perspective - in addressing peer-review in this century.

Perhaps not "at the root" . It's a dialectic process between practice
and theorizing about the goals of the practice. See Wikipedia, its
goals and its rules: you had an initial bunch of people who shared the
goal of making knowledge accessible, a very good knowledge of the tool
(wiki) they would use for it, and enough systematic imagination to lay
down rules that would make using that tool for that purpose work
efficiently. . Same with the Web Content Accessibility Guidelines, in
a way. I.e. in both cases, the "meta" reflection that lead to
re-elaboration was based on active experience.


>
>  (for humour and for critique, I hope that we can avoid using the te=
rms post-modern or any other post-dash term to do so - our grandchildren should not suffer having to live in a post-post-post-modern world!!!)

Oyez!!!!  The 70 (actually 72, according to the legend: 6 per each
tribe of Israel) blokes who are meant to have produced the
"Septuagint" Greek translation of the Bible would have loved a wiki.
And if remix culture is "post-modern", then Homer - if s/he existed -
was a post-modern DJ remixing the works of preceding indy folk singers
("aedes"), whose stuff got "remixed" in turn by Virgil, Sophocles,
Euripides,  Dante, Chaucer, Shakespeare, Joyce, Walcott and heaps of
others in between. As the Greek poet Seferis wrote in one of the
prefaces to his translation of Eliot's "The Waste Land": "There is no
parthenogenesis in art" (4).

Best

Claude


(1) I made a half jocular attempt at this in a "Babelfish 101" blog
post at the Digital Divide Network
<http://www.digitaldivide.net/blog/Claude/view?PostID=2277>. Then with
the students of an intensive French workshop, part of a bilingual
(French/English) Master's course in intercultural study, we compared
Babelfish and Google language tools results for a text on their
reading list (see <http://micusif.wikispaces.com/Mardi+2> and
<http://micusif.wikispaces.com/Priska+Troxler>).

(2) For instance last Summer I was asked to translate into French a
text that was  half in Italian and half in German, with the German
part being about using econometrics to assess development projects in
Alpine valleys. As I only knew that econometrics existed, I accepted
on condition that I could ask questions whenever I didn't understand
something. The authors of that part accepted and it worked well: we
used an online glossary made with a Google spreadsheet
<http://spreadsheets.google.com/ccc?key=p6XS5oqmWKcj81HBoaiLeCw&hl=en>
 for recurring terms, and the versioning and commenting features of
.doc files. We didn't use a wiki or a google doc for the text itself
because there were heaps of tables that don't transfer well, but I've
used wikis and google docs on other occasions.

(3) wikispaces.com set up a Pootle tool for the collaborative
translation of their interface in other languages. And Miro TV
getmiro.com  uses a similar one, Launchpad. I've only done some
translations in them, so I don't know if it is possible to  them for
other texts - and with longer "strings".  And maybe it would be
possible to hijack DotSUB.com - a collaborative multilingual
captioning tool for normal texts. Normally - see Lee LeFever's
Computer Software in Plain English
<http://dotsub.com/view/253fc0e4-32fe-4c36-80e4-5dfb692dbb95> - you
get a full plain text in each language under "Video transcription".
But when I tried myself in
<http://dotsub.com/view/1bd49084-e9de-4f33-9259-fd9c71519546>, I
didn't get the "Video transcription" feature.

(4) Btw, you could add Seferis to your list of "meta" authors - as
well as other poet-translators who wrote about translation and why
they translated: Lowell, Bonnefoy and many others.