Why no formalized collaboration between Wiktionaries?

[\Mike]

Hi,

Lately, I have started wondering why we don't
have a set of multilingual discussion pages - one
to deal with English, say, which could be used
for people interested in working on English words
in any wiktionary. Another for Swedish, and so
on... Then we could, I believe, reduce the amount
of *repeatedly* produced "hot air".... ;)


I mean, I see right now a discussion concerning
Romance languages' participles underway in
en:wikt. I see no reason, however, why a such
discussion won't arise again, in the Greek, or
the Polish, or even the Italian wiktionary.
 
I see the same discussions concerning various
details on certain Swedish words being held on
both sv: and en: - and in how many more places
are these words discussed without me noticing
because I'm not active in those wiktionaries, and
perhaps unable to understand the language in
which it is held? Perhaps I (or someone else)
familiar with other wiktionaries could point out
how the same problem may have been solved already
in this second wiktionary, would I only know
about the discussion...

Someone found out a while ago that several
wiktionaries had made mistakes in their treatment
of Irish nation names - and had to rise the same
issue over and over and over again, once in each
wiktionary where this user found this particular
error.



Though I know meta - in theory at least - has
been multilingual for quite some time, I'm not
very active there and hence don't really know
about how successful (or not) their attempts to
deal with large numbers of extensive multilingual
discussions  have been. (Perhaps someone could
enlighten me?)

Of course I understand that there are some
serious  complications with any attempt of a
"multilingual discussion" - maybe most
importantly the continuous need to translate
things, but I guess there also will be issues
with various wiktionaries wanting to arrange
things in very different manners.

Now the question is: would anyone be interested
in trying to follow a multilingual discussion of
their favourite language if it took place in meta
or on another site than they ordinarily work on?
Or would such an attempt be considered as an
attempt of *someone* (=outsiders) to decide how
"my" wiktonary is run?

Comments?


Regards,
\Mike

(p.s. This actually makes me regret that we
basically decided to split the wiktionaries
according to the user interface language and not
according to "content language", way back in
2003/2004 or whenever the first two non-English
wiktionaries were created.... :/ Well, no point
crying over spilled milk.)

I'm \Mike.

You'll find me at [[wikt:sv:Användare:Mike]], [[wikt:en:User:Mike]] and elsewhere.


      __________________________________________________________
Sent from Yahoo! Mail.
A Smarter Inbox. http://uk.docs.yahoo.com/nowyoucan.html

_______________________________________________
Wiktionary-l mailing list
Wiktionary-l@...
https://lists.wikimedia.org/mailman/listinfo/wiktionary-l

[Ray Saintonge]

\Mike wrote:
> I mean, I see right now a discussion concerning
> Romance languages' participles underway in
> en:wikt. I see no reason, however, why a such
> discussion won't arise again, in the Greek, or
> the Polish, or even the Italian wiktionary.
>  
I'm sure it will.  But with an entirely new cast of characters, and
perhaps they will find different byt equally valid solutions

> I see the same discussions concerning various
> details on certain Swedish words being held on
> both sv: and en: - and in how many more places
> are these words discussed without me noticing
> because I'm not active in those wiktionaries, and
> perhaps unable to understand the language in
> which it is held? Perhaps I (or someone else)
> familiar with other wiktionaries could point out
> how the same problem may have been solved already
> in this second wiktionary, would I only know
> about the discussion...
>  

The discussion between those who are already fluent in Swedish will be
far different from corresponding discussions on other projects.  The
things about the language and its culture which one takes for granted
when raised in that language are not at all obvious to those who learned
about the language later in life.
> Someone found out a while ago that several
> wiktionaries had made mistakes in their treatment
> of Irish nation names - and had to rise the same
> issue over and over and over again, once in each
> wiktionary where this user found this particular
> error.
>  
That sort of thing is inevitable.  Would you have understood what they
were talking about in Gaelic language comments?  It does not get any
easier when the language is non-ino-European.
> Of course I understand that there are some
> serious  complications with any attempt of a
> "multilingual discussion" - maybe most
> importantly the continuous need to translate
> things, but I guess there also will be issues
> with various wiktionaries wanting to arrange
> things in very different manners.
>  
Being able to arrange things differently will be important to the
longevity of the wiktionaries.  If different approaches are allowed to
flourish, they will be a constant source of fresh ideas for the others.
> Now the question is: would anyone be interested
> in trying to follow a multilingual discussion of
> their favourite language if it took place in meta
> or on another site than they ordinarily work on?
> Or would such an attempt be considered as an
> attempt of *someone* (=outsiders) to decide how
> "my" wiktonary is run?
>  
I don't think it would even get as far as complaints about outsider
actions.  People would just not bother to read comments in other
languages.  Given the long-windedness of many Wikimedians providing
translation for those who might be interested would not be practical.  
Already there are frequent and continuing requests for translating
documents that are important to Wikimedia in general.
> (p.s. This actually makes me regret that we
> basically decided to split the wiktionaries
> according to the user interface language and not
> according to "content language", way back in
> 2003/2004 or whenever the first two non-English
> wiktionaries were created.... :/ Well, no point
> crying over spilled milk.)
The underlying rationale was that each Wiktionary was intended to
address the needs of speakers of that language.  One could argue that
user interface language and content language are the same thing.  The
content for this purpose is not so much the words that are defined, but
the way they are defined and otherwise written up.  Part of the vision
is to have tools for identifying material when you don't even know what
language it's in.

Ec

_______________________________________________
Wiktionary-l mailing list
Wiktionary-l@...
https://lists.wikimedia.org/mailman/listinfo/wiktionary-l

[Oldak Quill]

On 04/03/2008, Mike <mike_wikipedia@...> wrote:

> Hi,
>
>  Lately, I have started wondering why we don't
>  have a set of multilingual discussion pages - one
>  to deal with English, say, which could be used
>  for people interested in working on English words
>  in any wiktionary. Another for Swedish, and so
>  on... Then we could, I believe, reduce the amount
>  of *repeatedly* produced "hot air".... ;)
>
>
>  I mean, I see right now a discussion concerning
>  Romance languages' participles underway in
>  en:wikt. I see no reason, however, why a such
>  discussion won't arise again, in the Greek, or
>  the Polish, or even the Italian wiktionary.
>
>  I see the same discussions concerning various
>  details on certain Swedish words being held on
>  both sv: and en: - and in how many more places
>  are these words discussed without me noticing
>  because I'm not active in those wiktionaries, and
>  perhaps unable to understand the language in
>  which it is held? Perhaps I (or someone else)
>  familiar with other wiktionaries could point out
>  how the same problem may have been solved already
>  in this second wiktionary, would I only know
>  about the discussion...
>
>  Someone found out a while ago that several
>  wiktionaries had made mistakes in their treatment
>  of Irish nation names - and had to rise the same
>  issue over and over and over again, once in each
>  wiktionary where this user found this particular
>  error.
>
>
>
>  Though I know meta - in theory at least - has
>  been multilingual for quite some time, I'm not
>  very active there and hence don't really know
>  about how successful (or not) their attempts to
>  deal with large numbers of extensive multilingual
>  discussions  have been. (Perhaps someone could
>  enlighten me?)
>
>  Of course I understand that there are some
>  serious  complications with any attempt of a
>  "multilingual discussion" - maybe most
>  importantly the continuous need to translate
>  things, but I guess there also will be issues
>  with various wiktionaries wanting to arrange
>  things in very different manners.
>
>  Now the question is: would anyone be interested
>  in trying to follow a multilingual discussion of
>  their favourite language if it took place in meta
>  or on another site than they ordinarily work on?
>  Or would such an attempt be considered as an
>  attempt of *someone* (=outsiders) to decide how
>  "my" wiktonary is run?
>
>  Comments?
>
>
>  Regards,
>  \Mike
>
>  (p.s. This actually makes me regret that we
>  basically decided to split the wiktionaries
>  according to the user interface language and not
>  according to "content language", way back in
>  2003/2004 or whenever the first two non-English
>  wiktionaries were created.... :/ Well, no point
>  crying over spilled milk.)
>
>  I'm \Mike.
>
>  You'll find me at [[wikt:sv:Användare:Mike]], [[wikt:en:User:Mike]] and elsewhere.

Since each Wiktionary project aims to define every word in every
language, a single multilingual project would be more optimal than the
current division per-interface language.

Interestingly, such a project has already started (albeit, not under
the auspices of Wikimedia Foundation): http://www.omegawiki.org/

This project was started by Wikimedians and is almost identical to the
aims of Wiktionary. The main difference is that omegawiki is based on
a more dictionary-friendly version of MediaWiki, allowing definitions
in different languages to be attached to an expression. The user
selects their preferred language and definitions in this language are
presented first.

One issue is the project works in a different way, it is not possible
to simply transfer Wiktionary contents into it. To embrace omegawiki,
we'd have to phase out Wiktionary in its current state. I think
omegawiki as Wiktionary 2.0 is a controversial idea among many
Wiktionarians, though I can't give you any more details on that.

A unified multilingual project could make progress far more quickly.
That omegawiki is based on WikiData also makes it more flexible and
useful (as part of the semantic web, for researchers, for end-users,
for contributors, &c.) Also, because users from different languages
aggregate together, correctness of spelling and definition is more
likely.

Are there long-term plans to bring omegawiki under Wikimedia
Foundation? Do Wiktionarians oppose this? Do omegawiki-people oppose
this? Does the Foundation oppose this?

--
Oldak Quill (oldakquill@...)
_______________________________________________
Wiktionary-l mailing list
Wiktionary-l@...
https://lists.wikimedia.org/mailman/listinfo/wiktionary-l

[Ray Saintonge]

Oldak Quill wrote:

> One issue is the project works in a different way, it is not possible
> to simply transfer Wiktionary contents into it. To embrace omegawiki,
> we'd have to phase out Wiktionary in its current state. I think
> omegawiki as Wiktionary 2.0 is a controversial idea among many
> Wiktionarians, though I can't give you any more details on that.
>
> A unified multilingual project could make progress far more quickly.
> That omegawiki is based on WikiData also makes it more flexible and
> useful (as part of the semantic web, for researchers, for end-users,
> for contributors, &c.) Also, because users from different languages
> aggregate together, correctness of spelling and definition is more
> likely.

A lot of this depends on your philosophy of dictionaries.  Are they
descriptive or prescriptive?  Translating dictionaries tend very heavily
toward the prescriptive, directed toward efficiently solving the very
real problem of rendering a text in one language into a readable text in
another language.  But translations taken from translating dictionaries
will at best seem stillted, and often misleading unless you have fully
explored the subtleties of both languages.  That is not particularly
efficient.

A top level single-language dictionary is built on descriptive
historical principles in that it is evidence based.  And, as much as in
Wikipedia should be citing sources for usage. To the extent that a
dictionary describes words in its own language it should always strive
for a depth that is unachievable in a translation dictionary.  A wise
literary translator must be aware of the subtleties, preferably of both
languages, but especially of the target language.

The problem for Wiktionary is how best to merge these conflicting
tendencies.

Ec

_______________________________________________
Wiktionary-l mailing list
Wiktionary-l@...
https://lists.wikimedia.org/mailman/listinfo/wiktionary-l

[Gerard Meijssen-3]

Hoi,
It was for the Wikimedia Foundation to adopt what was then called "Ultimate
Wiktionary". The WMF decided at the time that it did not want to adopt our
project and as a consequence Stichting Open Progress was founded and the
name of the project was changed to "OmegaWiki" in order not to confuse with
the Wiktionary projects.

The content of OmegaWiki is available under a different licensing scheme
from the Wiktionary data. This has everything to do with the fact that you
cannot license facts in the first place and because in our opinion there are
many competing effectively incompatible Open/Free licenses that deal with
lexical data. By providing our data both as GFDLD or CC-by, it is at least
possible for everyone to collaborate with *us *and *our *data.

As to bringing OmegaWiki under the WMF, there was a time when this was the
obvious solution. At this stage it is not so obvious nor necessary any more,
What is needed is a wish to collaborate. Collaboration is something that is
obvious for us. Contrary to some other projects we cherish a good relation
with the Wikimedia Foundation.. There is plenty of room for people of good
will to accomplish shared objectives.
Thanks,
     GerardM

On Wed, Mar 5, 2008 at 12:37 PM, Oldak Quill <oldakquill@...> wrote:

> On 04/03/2008, Mike <mike_wikipedia@...> wrote:
> > Hi,
> >
> >  Lately, I have started wondering why we don't
> >  have a set of multilingual discussion pages - one
> >  to deal with English, say, which could be used
> >  for people interested in working on English words
> >  in any wiktionary. Another for Swedish, and so
> >  on... Then we could, I believe, reduce the amount
> >  of *repeatedly* produced "hot air".... ;)
> >
> >
> >  I mean, I see right now a discussion concerning
> >  Romance languages' participles underway in
> >  en:wikt. I see no reason, however, why a such
> >  discussion won't arise again, in the Greek, or
> >  the Polish, or even the Italian wiktionary.
> >
> >  I see the same discussions concerning various
> >  details on certain Swedish words being held on
> >  both sv: and en: - and in how many more places
> >  are these words discussed without me noticing
> >  because I'm not active in those wiktionaries, and
> >  perhaps unable to understand the language in
> >  which it is held? Perhaps I (or someone else)
> >  familiar with other wiktionaries could point out
> >  how the same problem may have been solved already
> >  in this second wiktionary, would I only know
> >  about the discussion...
> >
> >  Someone found out a while ago that several
> >  wiktionaries had made mistakes in their treatment
> >  of Irish nation names - and had to rise the same
> >  issue over and over and over again, once in each
> >  wiktionary where this user found this particular
> >  error.
> >
> >
> >
> >  Though I know meta - in theory at least - has
> >  been multilingual for quite some time, I'm not
> >  very active there and hence don't really know
> >  about how successful (or not) their attempts to
> >  deal with large numbers of extensive multilingual
> >  discussions  have been. (Perhaps someone could
> >  enlighten me?)
> >
> >  Of course I understand that there are some
> >  serious  complications with any attempt of a
> >  "multilingual discussion" - maybe most
> >  importantly the continuous need to translate
> >  things, but I guess there also will be issues
> >  with various wiktionaries wanting to arrange
> >  things in very different manners.
> >
> >  Now the question is: would anyone be interested
> >  in trying to follow a multilingual discussion of
> >  their favourite language if it took place in meta
> >  or on another site than they ordinarily work on?
> >  Or would such an attempt be considered as an
> >  attempt of *someone* (=outsiders) to decide how
> >  "my" wiktonary is run?
> >
> >  Comments?
> >
> >
> >  Regards,
> >  \Mike
> >
> >  (p.s. This actually makes me regret that we
> >  basically decided to split the wiktionaries
> >  according to the user interface language and not
> >  according to "content language", way back in
> >  2003/2004 or whenever the first two non-English
> >  wiktionaries were created.... :/ Well, no point
> >  crying over spilled milk.)
> >
> >  I'm \Mike.
> >
> >  You'll find me at [[wikt:sv:Användare:Mike]], [[wikt:en:User:Mike]] and
> elsewhere.
>
> Since each Wiktionary project aims to define every word in every
> language, a single multilingual project would be more optimal than the
> current division per-interface language.
>
> Interestingly, such a project has already started (albeit, not under
> the auspices of Wikimedia Foundation): http://www.omegawiki.org/
>
> This project was started by Wikimedians and is almost identical to the
> aims of Wiktionary. The main difference is that omegawiki is based on
> a more dictionary-friendly version of MediaWiki, allowing definitions
> in different languages to be attached to an expression. The user
> selects their preferred language and definitions in this language are
> presented first.
>
> One issue is the project works in a different way, it is not possible
> to simply transfer Wiktionary contents into it. To embrace omegawiki,
> we'd have to phase out Wiktionary in its current state. I think
> omegawiki as Wiktionary 2.0 is a controversial idea among many
> Wiktionarians, though I can't give you any more details on that.
>
> A unified multilingual project could make progress far more quickly.
> That omegawiki is based on WikiData also makes it more flexible and
> useful (as part of the semantic web, for researchers, for end-users,
> for contributors, &c.) Also, because users from different languages
> aggregate together, correctness of spelling and definition is more
> likely.
>
> Are there long-term plans to bring omegawiki under Wikimedia
> Foundation? Do Wiktionarians oppose this? Do omegawiki-people oppose
> this? Does the Foundation oppose this?
>
> --
> Oldak Quill (oldakquill@...)
> _______________________________________________
> Wiktionary-l mailing list
> Wiktionary-l@...
> https://lists.wikimedia.org/mailman/listinfo/wiktionary-l
>

_______________________________________________
Wiktionary-l mailing list
Wiktionary-l@...
https://lists.wikimedia.org/mailman/listinfo/wiktionary-l

[Gerard Meijssen-3]

Hoi,
When you intend to make a Wiktionary and have it include all words of all
languages and, have wiktionaries in all the languages that ask for one, you
have a situation where every project aims to do exactly the same thing and
without collaboration between the projects you have extreme inefficiency. It
is also not feasible to have all these projects be of the same standard and
only a few projects are usable.

Dividing dictionaries in either mono-lingual or translation dictionaries is
historically correct. However with the Internet and with the massive
collaboration and the mashing of resources that is now possible, it is a bit
of a simplification. Also the sad reality of Wiktionary is that it is
inherently useful for looking things up by people. Thanks to the massive
effort to standardise data it becomes possible to extract data from
Wiktionary. This is what enhances the potential of Wiktionary enormously.

OmegaWiki is inherently able to reuse data because of the way the data is
stored. OmegaWiki is able to provide the same information in many languages.


Thanks,
    GerardM

On Wed, Mar 5, 2008 at 10:00 PM, Ray Saintonge <saintonge@...> wrote:

> Oldak Quill wrote:
> > One issue is the project works in a different way, it is not possible
> > to simply transfer Wiktionary contents into it. To embrace omegawiki,
> > we'd have to phase out Wiktionary in its current state. I think
> > omegawiki as Wiktionary 2.0 is a controversial idea among many
> > Wiktionarians, though I can't give you any more details on that.
> >
> > A unified multilingual project could make progress far more quickly.
> > That omegawiki is based on WikiData also makes it more flexible and
> > useful (as part of the semantic web, for researchers, for end-users,
> > for contributors, &c.) Also, because users from different languages
> > aggregate together, correctness of spelling and definition is more
> > likely.
> A lot of this depends on your philosophy of dictionaries.  Are they
> descriptive or prescriptive?  Translating dictionaries tend very heavily
> toward the prescriptive, directed toward efficiently solving the very
> real problem of rendering a text in one language into a readable text in
> another language.  But translations taken from translating dictionaries
> will at best seem stillted, and often misleading unless you have fully
> explored the subtleties of both languages.  That is not particularly
> efficient.
>
> A top level single-language dictionary is built on descriptive
> historical principles in that it is evidence based.  And, as much as in
> Wikipedia should be citing sources for usage. To the extent that a
> dictionary describes words in its own language it should always strive
> for a depth that is unachievable in a translation dictionary.  A wise
> literary translator must be aware of the subtleties, preferably of both
> languages, but especially of the target language.
>
> The problem for Wiktionary is how best to merge these conflicting
> tendencies.
>
> Ec
>
> _______________________________________________
> Wiktionary-l mailing list
> Wiktionary-l@...
> https://lists.wikimedia.org/mailman/listinfo/wiktionary-l
>

_______________________________________________
Wiktionary-l mailing list
Wiktionary-l@...
https://lists.wikimedia.org/mailman/listinfo/wiktionary-l

[Lars Aronsson]

\Mike wrote:

> Lately, I have started wondering why we don't
> have a set of multilingual discussion pages - one
> to deal with English, say, which could be used
> for people interested in working on English words
> in any wiktionary. Another for Swedish, and so
> on... Then we could, I believe, reduce the amount
> of *repeatedly* produced "hot air".... ;)

This sounds like a bad idea.  If you mean that we should start to
write in Swedish on the talk pages of the English and French
Wiktionary, what good would that be?  Who would understand those
discussions?  But Wiktionary is already ruined, so why don't you
go ahead.

Earlier this year, I called a meeting in Stockholm about free
dictionaries for the Swedish language.  Ten people showed up,
representing as many projects.  But nobody from the Swedish
Wiktionary bothered to attend.  All who attended considered
Wiktionary to be broken by design and not worth spending time on.  
And hey, I'm representing the Swedish chapter of the Wikimedia
Foundation.  I have fifteen years of experience of free
dictionaries and I was among the younger at this meeting.

So the first issue should be: Why are there no discussions? I.e.
skip the word multilingual.  Why do the important and competent
people, who do contribute to free dictionaries, shun Wiktionary?
Shouldn't we just scrap Wiktionary and start from zero?  If it is
good and worth keeping, where is the good in it?


--
  Lars Aronsson (lars@...)
  Aronsson Datateknik - http://aronsson.se

_______________________________________________
Wiktionary-l mailing list
Wiktionary-l@...
https://lists.wikimedia.org/mailman/listinfo/wiktionary-l