[WikiEN-l] Re: transliteration is stupider

[Ray Saintonge]

Having succombed to the urgings of both Mav and Toby to join the 
WikiEN-l list, despite my protestations about having yet more e-mails to 
thwart my constructive participation in Wikipedia, and having dutifully 
ploughed through the mass of messages that have accumulated therein in a 
very short period of time, I am now prepared to move back in the stack 
to those substantive matters that are a part of the mailing list.  I 
dream that some day I may be able to escape the stack and rejoin 
Wikipedia's Mainframe.

Toby Bartels wrote:

>Mav wrote:
>
>>Toby Bartels wrote:
>>
>>>NPOV, of course, has nothing to do with this (despite what Lir says).
>>>As you know, NPOV doesn't mean the majority point of view.
>>>It means presenting every point of view in a manner fair to it.
>>>
>>Key words "in a manner fair to it". In my world view what is most "fair" in
>>terms of NPOV is to express majority opinions as majority opinions and
>>minority ones as minority.
>>
I find it hard to see how NPOV is a serious part of this debate.  Sure 
there are POV titles; I've raised that point in connection with 
"Pseudo-science".  This debate is about a convention.  Establishing a 
convention is about choosing between equally functional positions. 
 Accident rates in Britain and Japan have not been significantly 
affected by their choice to drive on the wrong side of the road in 
defiance of Pope Boniface VIII's edict for the year 1300.

>>This also affects the amount of text we give to
>>any particular opinion in an article. So majority opinions get majority time.
>>Therefore if something is known by a particular name by the great majority of
>>English speakers we should reflect this fact in our choice of what to name
>>the article. So yes, NPOV does most certainly apply.
>>
Greater amount of text?  With the title "München" the article adds "It 
is called Munich in English."  With the title "Munich" the article adds 
"It is called München in German."  Where the length difference.

>...we must also indicate minority opinions where applicable.
>It is applicable in article titles -- but we can't do so there,
>because (as you say, and this is *central* to my argument,
>not contradictory to it), such discussion goes into the article itself,
>not into the title, which must present only one thing.
>We must make a decision in the title, we can't be NPOV there,
>(and we don't have to in order to be a neutral encyclopædia
>since everything is explained immediately beneath said title).
>
It's not just that we "must" present only one thing in the title; it's 
that we "can" present only one thing.  There are still some instances 
where NPOV becomes a problem in selecting article titles.  One example 
is in choosing between Oder River (from the German name) or Odra River 
(from the Polish name).  Wikipedia should not be put in the position of 
having to choose sides in that dispute.

>>>Under the current plan we also have to delve into linguistic usage
>>>(not *wars* that I can see, but I don't anticipate those in any case),
>>>to decide which usage is most common.  That's an issue of linguistic usage.
>>>
>>Eh? Finding out what most English speakers actually use is much easier than
>>having to research what the residents of where the term is derived use it.
>>
This could work either way, and although what most English speakers use 
is easy to determine for some of the better known entities, there is a 
lot more uncertainty as the subject becomes more obscure.  Disputes have 
already arisen over what name to use for Charlemagne, or, more broadly 
various European monarchs named Charles, Henry, William or Peter.

>Not at all.  There are a lot fewer residents of München
>than English speakers.  No, Google does *not* do the job;
>Google isn't anywhere *close* to a reprsentative sample of English speakers.
>The fact that a Google measurement is *easy* to do
>doesn't change the fact that it's highly biased and thus almost meaningless.
>
I would avoid an unhealthy addiction to Google.  It is clearly the most 
popular search engine now, but Yahoo and Lycos have been there in the 
past, A new engine could appear at any time and change that fact.  I do 
look at Google statistics, and take that into account as only one of 
several possible influences.

>>Their language evolves too, just like ours. Should we use what they call the
>>term now, back when the term was coined or some arbitrary date inbetween?
>>
>
>Well, if our language evolves, then what *English* term should we use,
>what English said back then, or what English says now?
>Actually, I know the answer to my question: what English says now.
>Similarly, we can come up with an answer to your question.
>I suggest referring to a place by the name used by the current residents
>(or the residents at the time for articles focussing on a historic period)
>and referring to a person by the name that they used for themself.
>That seems pretty natural to me.
>
That makes sense.

>>There will also be different spellings and different use of diacriticals to
>>deal with, not to mention competing terms. How in the world can an
>>English-only speaker sort this out?
>>
>The same way that such a speaker currently
>sorts out what's most commonly used in English:
>by making the best guess that they can
>using the resources easily available to them
>and then being corrected by later editors.
>Such corrections happen now under the current policy, too.
>
Yes, even under an original language regime the first author of an 
article can use the English name to get the matter going.  Perhaps his 
only information source is something like the 1911 Encyclopedia.  If 
that does not fit with our naming policy somebody else can and probably 
will change it..  We'll appreciate both efforts.

>>>I don't know about Lir, but I don't propose such a thing,
>>>because "mammal" is a common noun, not a proper noun.
>>>I certainly don't want to change all of our article titles
>>>back to Proto-Indo-European ^_^!
>>>
>The new policy is proposed for *proper*nouns* only,
>and refers to historical forms *only* in historical contexts.
>It's not about retracing the origins of English etymology *at*all*.
>
It is primarily about proper nouns, though I still support the principle 
that the primary entry for life forms should be by the unambiguous Latin 
binomial.  On the other hand, I still very much disagree with Tarquin's 
idiosyncratic predilection for using the french form "département" (with 
the extra "e") instead of "department" when referring the the political 
sub-divisions of France.  

>>Already there is a
>>continuum of opinion on how such a convention would work; Lir on the extreme
>>"all anglicization is bad" end, you in the middle and Ec on the more liberal
>>and IMO sane side.
>>
This is frightening!  I'm being accused of sanity.

>Yes, there's a continuum, going from Lir to me to Eclecticology to you.
>Why does it follow that you must then be right?  Why not Lir?
>>From her perspective, there's a continuum of opinion about
>how words derived from proper names should sometimes be anglicised;
>you on the extreme "any anglicization done by most anglophones is good" end,
>Ec in the middle and me on the more liberal and IHO sane side.
>It's not like there's some wide gulf between you and Eclecticology
>that doesn't exist between me and Lir.
>
I don't know how I feel about having the centre-right position in such a 
political spectrum, even as Lir's uncertain credibility makes me more of 
a centrist.  (Ed Poor eat your heart out!)

>>There is also the fact that the proposed change absolutely requires the use
>>of a technological fix to work (namely redirects). How is it less complicated
>>when redirects are absolutely required? Not to mention the fact that the
>>current display of redirects is rather ugly after following them. To fix this
>>would require yet another technological fix.
>>
Redirects are one of the earliest techniques that new Wikipedians learn. 
 They are so common that I hardly consider them as technological fixes. 
 IIRC when I suggested deleting some such entries a few months back 
because they were useless, it was Mav who objected because Google could 
use them for a search, or somebody might have them bookmarked; this 
could mean the loss of potential new Wikipedians.  He further argued 
that redirects didn't take up a lot of space in the system anyway.  Now 
he has elevated redirects to the status of technological fixes.

>>Google's language tools can be used as an objective measure of widest usage
>>whereas the proposed plan depends on subjective choices between different
>>more native transliterations and language sets.
>>
It is only one measure.  Why wouldn't it also work for original language 
frequencies?

>>4) What about the audience: How is it more useful for them to use words they
>>can't pronounce, spell or are familiar with?
>>
>It's more useful for them to have the name used by the person in question
>or in the place in question up front and direct in their minds.
>We use pronunciation guides for weird words like "Kong-fu-zi" (/koN 'fu dz/)
>and "Confucius" (/k at n 'fju S at s/), which as you can see in this example,
>include both native and common English forms in most cases.
>
Using SAMPA makes it even less pronounceable.  I find "Kong-fu-zi" more 
pronounceable than its SAMPA counterpart, but then I would also advocate 
IPA for pronunciation guides.  (But that's an entirely different issue 
that I'll save for another day.)  Strangely enough, the present issue is 
not about pronunciation, but about the visual representation of names. 
 A Chinese ideograph is more closely connected to its meaning than its 
pronunciation.  At the risk of seeming to go out on a limb, I believe 
that there is a linkage between a culture and the way it writes things. 
 Köln does not communicate the same picture as Cologne even though they 
refer to the same city.  This is perhaps the saddest result of 
anglocentric expressions of these names:  that somehow a cultural 
subtlety is not communicated, and the Dresden before and after the 
bombing remains the same pile of bricks and mortar.

>>What about the writers: How is it more useful for them to have articles
>>that they can't link to directly (or at all when redirects are not made)?
>>
If all the redirects are not made, we need to do more housework.  Direct 
linkage is not a necessity as long as they can get there.

>>Also, most English speakers do not know how to make the more elaborate
>>diacrtic marks with their keyboard. So they will have to copy and paste to
>>make direct links. How is that at all user-friendly?
>>
Present ignorance is no excuse for future ignorance.  Making the 
characters in ISO 8859-1 is not difficult, just a little inconvenient. 
 The evedent difficulty will be with languages using the roman alphabet 
with characters outside if 8859-1, which even I will admit may need to 
wait for a better solution.  In languages requiring transliteration or 
romanization there may not even be a problem since the transliterations 
mostly use 8859-1.

Eclecticology