W3C home > Mailing lists > Public > public-vocabs@w3.org > September 2014

Re: itemprop="translator"

From: Mats Blakstad <mats.gbproject@gmail.com>
Date: Sun, 7 Sep 2014 05:39:53 +0200
Message-ID: <CAP=1PAUkX=c_pKgRS6qn4DipSxmDXM4tqbT9xihUoJRU2sLOLQ@mail.gmail.com>
To: "Wallis,Richard" <Richard.Wallis@oclc.org>, "<public-vocabs@w3.org>" <public-vocabs@w3.org>
Thanks for sharing this!
If it is a "translation of translation", should there not be any way to
link the content both to the translation source, and to the original work?
And what if a translation used two other translations, can it be mapped
with two source translations?

The way you write this it sounds like it can work good for mapping e.g.
books in a library, but what about websites and translations of
website pages. Will they be able to use the same system?

I ask because websites are usually not static work. What happens when
somebody update the source text? Then the other translations becomes
out-dated. Maybe it should be a way to mark a translation as out-dated too?
If not it is important that the translations are mapped with the correct
revision of the original work.

Just some thoughts, will be interesting to see the outcome of this


2014-09-03 14:57 GMT+02:00 Wallis,Richard <Richard.Wallis@oclc.org>:

>  The pattern that seems be emerging from the work we are in the middle of
> at the moment goes something like this (assuming translator,
> translationOfWork, and workTranslation are in place):
>
>
>    1. First describe the [translated] work as if it wasn’t a translation
>    2. add as alternateName the the title of the original work (language
>    tags are important)
>    3. author set to the author of original work (preferably URI in an
>    authority e.g.VIAF)
>    4. Translator set to the translator’s identity (preferably URI in an
>    authority e.g.VIAF)
>    5. translationOfWork set to URI of the work this was translated from -
>    care needs to be taken as many translations are actually translations of
>    translations.
>    6. If access is available to the description of a work that has been
>    translated, add a workTranslation URI to link to its translation(s)
>    7. Go drink some more coffee!
>
>
>  As I said I would love to share more details, but the project is not
> completed, therefore not quite ready to share conclusions yet.
>
>  ~Richard
>
>  On 3 Sep 2014, at 12:15, Mats Blakstad <mats.gbproject@gmail.com> wrote:
>
>  On Mon, 01 Sep 2014 10:01, Jean-Christophe Lavocat wrote:
>  > Regarding an author/translator, one might want to know which "content"
> has been written by the person. Is there any official way to do that ?
>
>  I guess it will only make sense to add a translator tag (or "translator
> role") to a text first defined as a translation. Author tag could then be
> added to any language version of the text and will then be the one that has
> written the original source text of that translation.
>
>  I was not aware about the roles pattern used. To simply add "translator"
> as a contribution role to the work sounds very logical and would make it
> easier to manage for webmasters and to keep overview. However, the
> contribution from a translator to the work is in fact really significant
> for the person that needs the translation to understand the text. The issue
> is not about simply having a list of translators to give them attributions
> to a work, but to be able to display who translated a text together with
> the specific translation done. It serves an important function: Then if
> there are confusions in the text, you can actually contact the translator
> to clear up what you don't understand. The translator understand a language
> you don't, and you're in fact helpless to understand the original work
> without the translator.
>
>  Maybe knowing the name of the translator would not make you buy a book,
> but would you buy the book at all if it was only written in a language you
> didn't understand? I think we can hardly underestimate the importance of
> the translator.
>
>  A good translator understand a lot about the cultural context both of
> the source language/culture and the target language/culture, so there are
> many interpretations made for you by the translators that will be
> transmitted directly to your understanding of the work. To know who made
> the translation can say a lot about the quality of the translation. Culture
> and language is really not a 1-1 issue that can be translated word by word
> and different translators will often give different translations. Then of
> course the creative contribution from a translator to a translation can
> differ a lot; in a poem the translator needs to be more poetic and the
> translator will probably have a stronger voice in the text, in a legal text
> of course the translator will try to avoid having any voice themselves at
> all.
>
>  On Tue, 02 Sep 2014 01:49, chaals@yandex-team.ru wrote:
> > I think the sticking point on translation is about the difference
> between identifying the thing translated and the translation, or between
> noting that there *is* a translation without specifying which is the
> original...
>
>  Good point! But should this not be integrated with the way different
> language versions of a website is declared today? Usually a html document
> will declare all language versions of itself in the header with a link like
> this to each language version:
> <link rel="alternate" hreflang="LANGUAGE-CODE" href="LINK">
> As far as I understand, there is nothing that indicate what is the
> original content (rel="canonical" only indicate that two different html
> pages have same content/translation). I guess adding something like
> *itemprop="hasTranslation"* or *itemprop="translationOf"* inside these
> link tags would help clarify what is the original version, but they would
> not be very human readable. What will *translationOf* mean inside the
> link tag in the header? That the html document is a translation of the
> link? Or that the link is a translation of the html document?
>
>  *rel="alternate"* in fact already means an alternative version of the
> document, it can be different language version, but also a version with
> different style sheet (e.g. for people that need high contrast because of
> reduced sight). Maybe what we're actually missing is *rel="source"* or
> something similar, to indicate what (language) version is the original one?
>
>
> 2014-09-01 8:44 GMT+02:00 Wallis,Richard <Richard.Wallis@oclc.org>:
>
>> There is not a translator property at the moment, I agree that one would
>> be useful along with translationOf/hasTranslation.
>>
>> These are on my list for potential proposals from the SchemaBibEx group.
>>
>> ~Richard
>>
>> On 31 Aug 2014, at 14:44, Mats Blakstad <mats.gbproject@gmail.com> wrote:
>>
>> > Does there exist in attribute for translator, like the author attribute?
>> > If not, would be great to add that!
>>
>>
>>
>
>
Received on Sunday, 7 September 2014 03:40:21 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:29:44 UTC