Language X within scope of language Y

L.Gillam L.Gillam at
Fri Jan 21 12:02:47 CET 2005


Though we would all like to travel to the stars and the far reaches of the galaxy, we first have to get into space.

I can conceive of an application context where I would like to retrieve text written by person A, and be able to read it in my "own language", however the necessary efforts required to undertake such an endeavour (6 billion people?) suggests a very long time for realisation - and if you were to follow the arguments on national identify card schemes in the UK, you'd discover why. Of course, this does not preclude experimental work that uses any or all of ISO 639/3166 and RFC3066/3066bis to explore this. Indeed, this would seem to be another application of these items, and tends to validate their existence.

Discussions here have considered, broadly, what "tagging" entails for content - indeed, what can be tagged (written language predominantly - how you tag fragments of speech in "binary" data using xml:lang I've yet to fully comprehend - BinX perhaps?). As we get towards finer grained aspects - when can a word/phrase be said to belong to a language? - we start to discover more "dark matter" in the linguistic universe. Such issues are questions also for identifying sublanguages - is that a term from "Theoretical Physics" or "Nuclear Physics"? Where does Biology end and Computing begin in Bioinformatics? Is "kaput" English or German? When you begin asking whether person X has adopted "deja vu" as part of their personal English dictionary or whether it is in their dictionary of phrases borrowed from French, you start exploring the boundaries of your own sanity.

In order to get to the stars, you need substances like teflon, and from this, sometimes you can create non-stick frying pans. If "fr-Latn-US-ietf_language_style-ietf_language at" is the goal, stabilising "fr-Latn-US" would seem to be the teflon you can start with.

Now, the question I don't think has been asked (and here a additional searching mechanisms may or may not have helped) is: do generative mechanisms allow for both "fr-Latn-US" and "fr-US-Latn" - are the semantics of these constructions considered different? 

The other question I have is, could the "algorithm" and the assumptions be simplified, for example using wildcards (e.g. zh-**-Hans) as, for example, in font descriptions in Unix? Rushed programmers sometimes seek shortcuts!

Best Regards.

> -----Original Message-----
> From: JFC (Jefsey) Morfin [mailto:jefsey at]
> Sent: 20 January 2005 16:02
> To: Gillam L Dr (Computing); petercon
> Cc: ietf-languages
> Subject: RE: Language X within scope of language Y
> At 13:36 20/01/2005, L.Gillam wrote:
> >[With tongue planted somewhere in cheek]
> don't gulp at the response or you may be hurt !
> >How does one register, get consensual agreement on, and 
> become allowed to 
> >use Peter Constable?
> >Can Peter Constable only be used for tagging content, or 
> does he have a 
> >broader scope, and does he get a say in this?
> Never thought that THIS list is authoritative on nothing else 
> than its own 
> decisions?
> This is what I concede leaving you the choice of the default 
> value. So, 
> "fr-Latn-US-ietf_language_style-ietf_language at" 
> is something 
> you say you want to default to fr-Latn-US (if I read through 
> the off topic 
> responses). If it pleases you so much to be the lingustic 
> center of the 
> world, there is no problem with that. But your delivery (the attached 
> content) will have to be very appealing if you want it to be 
> widely used.
> But
> > > "fr-Latn-US-ietf_language_style-Peter_Constable"
> is something you need the OK of Peter to register. If he 
> wants to give you 
> authority on him. But frankly why would he want to register 
> it with you ?
> HA's Troll

More information about the Ietf-languages mailing list