Sorting it all Out Michael Kaplan's random stuff of dubious value Be sure to read the disclaimer here first!
I have mentioned Portuguese and some of the many issues that come up with the fact that Microsoft does not usually localize int two different dialects of the same language before in this very blog (whether that language is English, Spanish, Arabic, or Italian).
Anyway, I was installing the Server 2003 MUI language packs the other day, and I noticed in the list of languages two varieties of Portuguese were present:
Ok, no worries -- many people feel that the difference between the two is much greater than the difference between US English and UK English, or between Spanish in Spain and Mexico. In other words, it reportedly goes above the level of being annoying/distracting and into the realm where mutual intelligibility is in question.
Now one could argue the same issue exists in Arabic as spoken in Morocco and Saudi Arabia. And one would even be correct, except for the fact that this ignores a more 'formal' mode of Arabic that is apparently accepted across many Arabic locales, one that does not delve too deeply into the dialectical differences.
Certainly their are standardization bodies in both Canada and France that each have strong (sometimes conflicting) opinions about the way French should be spoken. I am fairly convinced (after having had translations from both locations mistaken for their source by some of the 'experts' in the respective dialects) that the differences in French have a lot less to do with the language than those who are passionate on the topic will readily admit.
But Portuguese really is different -- historically, it is obvious that Brazilian Portuguese is a child of Iberian (European) Portuguese, but just as obviously it has gone through many subtle and not-so-subtle changes. And that Brazilian market is much bigger, too (a factor that a software company cannot really ignore!).
Anyway, when I went to install those language packs, I was somewhat bemused to see the dialog come up when the Brazilian Portuguese language pack was being installed:
compared to when the Iberian Portuguese one was being installed:
Now these two names are definitely not coming from the NLS data -- it would have been interesting to look at the localized strings for this piece of setup, had this UI been localized (unfortunately, it is not). This is all in the ProductName field anyway -- someone just decided to put them out there this way, even though other places in Windows don't seem to do it this way.
And I wonder in any case whether this was someone being clever on the English version or if it was with feedback from the language people?
Anyway, once it is installed let's look at how they are listed in the MUI language list:
Wow, at least they are using NLS again. They are picking up the LOCALE_SNATIVELANGNAME, with Brazilian Portuguese (0x0416) calling it Português and Iberian Portuguese (0x0816) calling it português.
Of course, this seems like a poor distinguishing feature to me, even if it does truly represent the preference in each locale. If you are a native speaker, doesn't this seem like a little bit of a mind game to guess which one to pick? And if that preference changes, what happens to the items in this list? Could you end up with two items that are identical? And could that lead to a bug where the first one is always selected and the Continental variety gets the shaft?
Maybe I should try this out in Vista at some point with a replacement locale that makes one's SNATIVELANGNAME look like other's and see what happens....
(don't worry, some day soon I'll be posting a whole bunch about replacement locales and some of the more compelling scenarios!)
Anyway, in speaking to people from both Mexico and Spain (as well as conversations with people from other Spanish-speaking countries), the differences between the various Spanish dialects can sometimes be quite marked. It has been a source of some bitterness among the more pure Castellano speakers that so much of the Spanish product appears to be produced in somewhere other than Spain.
So what is the actual difference, the one that makes it compelling to ship two versions of Portuguese but only one version of some of these other languages?
From those conversations I will hazard a semi-educated guess. :-)
I think it has more to do with what markets will tolerate -- the differences in English or Spanish or Italian are certainly noticable, but the dialects are mutually intelligible. And in the case of Arabic the attempt to try to stay as neutral of dialect as possible may make a product feel less "local" to the inhabitants any one country but more likely to be understood across a wider area.
So a real study of this would have to be focused on the actual differences between the various languages. According to Wikipedia, Phrase- and sentence-level stress and tone patterns and differ significantly between dialects: European Portuguese is often described as a stress-timed language (consistent with the its loss of pre-stress vowels), while Brazilian Portuguese is syllable-timed.
If true, then this is obviously a much more significant type of difference than those in for example English (which seem to be more focused on word choices and word spellings), or in French (which as I said I have been able to unintentionally fool people in Quebec about translations done in France and vice versa). And in situations like those with the Arabic language (where the choice seems to be more focused on literacy vs. illteracy than on capturing the dialectical variations between populations), it seems much more natural to simply go with what one might consider to be the Arabic language equivalent of BBC Standard Received (you know, pure, cultured, etc.).
Or perhaps we have a new standard of degree of language difference that spans linguistic, geopolitical, and other concerns. And we can alter Max Weinrich's famous quote (A shprakh iz a dialekt mit an armey un a flot) and say that a language is a dialect with its own localized version of Windows.
Boy, if ever were the first paragraph of my disclaimer to apply, it would be just about now, huh?
But in my opinion, if Microsoft as a company really is trying to give people local experiences with Windows, we need to work harder to target localized versions to a finer level than we do currently.
An aussie and a yank and a brit may be able to understand each other, but that does not mean it is a comfortable local experience to force that difference on them on a full-time basis....
This post brought to you by "ڦ" (U+06a6, a.k.a. ARABIC LETTER PEHEH)
Ah yes, it has happened again. As we all knew, it could not end with اردو , Inuktitut , മലയാളം , Qhichwa
A question came up the other day that some regular readers might find vaguely familiar: We have been