ICANN/GNSO GNSO Email List Archives

[ga]


<<< Chronological Index >>>    <<< Thread Index >>>

Re: [OT] RE: [ga] Aviation languages


Dear all,
This post uses the GA debate example to the advantage of the MLTF mailing list. However, some on the GA mailing list may be interested in commenting on it, since it is a key issue for naming as well as local Internet usage. I will forward their comments on the MLTF and will be used to formalise this post into an IETF Draft for information.
jfc


---

Subject: "Aviation Languages"

The subject of this ICANN/GNSO/GA thread seems to have a particularly well defined meaning in English...since the GA Members are extensively discussing it. In French, it is only a non-translatable confusion. This is important for understanding as to just why Debbie and her WLDC (<http://thewldc.org/>http://thewldc.org) and our (<http://mltf.org/>http://mltf.org) MLTF tend to have various converging and diverging points of view and motivations, that I propose to make into coopeting (<http://mltf.org/070622-prop.pdf>http://mltf.org/070622-prop.pdf). The same remark applies to ISO TC 37 and ISO TC 46 and current NWIP.


1. The confusion

The confusion comes into play when the English "language" term is used to support two very different concepts that French differentiates:
* the language of a linguistic community (English, Italian, French) - in French: "langue"
* the language of a specialized community (Aviation, Internet [for ALAC], programming, e-commerce, standards) down to even a single individual in a given situation - in French: "langage".
To clarify this confusion, one must take into account that:
* most do not realize that to technically support billions of "langages" and a hundred "langues" is quite similar, in which each linguistic unit is supported via an ad hoc corpus enacted through the same architecture. The real issue is the architecture, and not the number of languages.
* a human person has his/her own "langages" in his/her own single or few "langues" for education and memory limitation reasons.
* the Internet community largely has more difficulties due to the translation costs and administation.
* Internet technology has a near imossibility due to the lack of a presentation layer in the TCP/IP pile.
* the Unicode community has fewer due to the UCS and CLDR (locale files).
* the DNS has no real problem except for the confusion over what it is to do, how, and where.
* the computer has no difficulty, provided that it has the proper architecture and documentation.
* imposing a standardized "langage" (for example: administrative language as documented by ISO 3166-1) is not a technical barrier to Trade.


2. The current situation

Atop that situation, a cultural technical, political, and commercial status quo has developed:

* English represents most of the presentation-session-less missing Internet Interoperating System
It is like aviation: only one single global English "presentation" (general usage) is permitted, and possible locally limited Italian sessions (on a per flight basis). This is not too dangerous, because if you hear an aviation "langage" dialog, whatever the "langue", you can still understand most of its aviation specific "langage". The same with ICANN: you can understand most of what an ICANN paper is about (not what it says) in many languages.
* It builds language support as an option atop that of the environment.
This is the consistent IETF approach for the last 10 years. This results in the "internationalization" of the mediums, "localization" of the ends, and description (tagging) of the content. This is documented by the BCP 47. We all know as IDNs, as an application additional option.
* It calls for language filtering instead of support
When a language is requested, and is not supported, the dialog is redirected towards a less acceptable but possible language, or filtered out altogether (RFC 4647). I do not speak Finnish, but you can try to understand my Estonian or my Hungarian pages. At the end of the day everyone will e-commerce and work together using the default English language.


3. The poor consequences

This may seem a good thing to English speakers. However...
* it is likely that human beings, being human beings, that there will be no e-commerce by that time.
* the resulting impact on cultures would be a tremendous catastrophy.
* the lack of resultant technical development would delay the world by decades.
* etc.
This debate belongs to that cultural status quo among technical people, which denies people the Human Right to use their own language to live, manage, and innovate. The problem (well documented as a basic Internet problem by the IAB in RFC 3869) is that there is no commercial incentive today to support more "langues". From this comes, the difficulties that Debbie documents and you (GA) meet. This is obviously highly detrimental to culture and humanity's legacy. And this matches two of the five international legal definitions of what a genocide is (to force people to live in using another language, to educate kids in a culture not desired by their parents).


However, there are technical, political, and further on enormous commercial incentives to support "langages", and an awful lot of money that is poorly spent because the proper model is confused by this English language confusion (II am referring to the metadata domain and semantic web).


4. The correct approach

The other approach we support is architectural, and is motivated by network, security, etc. development, which we call "multilingualization". It consists of being able to support every language the way English is supported today.
* either people do not understand what it means and consider us as dumb fools and boring trolls;
* either they understand us too well and they are afraid of the impact on their technology, policy, e-commerce, cultures, etc. and oppose or disregard us.
* or they understand the importance of this impact and support us, but we are at the beginning of a long march for research, devlopment, and deployment. It takes a very long time to understand and accept the obvious: this is no exception.
This only delays development and innovation for eveyone, and endangers a correct approach to semantic processing and networking.



5. The stakes

Languages are human brain to brain interintelligibility protocols, which use different syntaxes and terminologies with an impact on logical exchanges, supporting semantic (meaning's level) relations, in order to exchange information (content) in a manner that is itself (pragmatics) a compression of knowledge [in a relation wherein you assume that much information is already known from your interlocutor]. Syntax, terminology [in one or several languages], semantic approaches, content thematics, pragmatic usages together form a pattern that could identify originating individual brains (human or computer). There are therefore enormous incentives for engaging in "langage" related developments and protection research.

Languages are also brain programming languages. Working on languages permits one to better educate, inform, etc. and analyze knowledge, in which working on a "polynymy" (invariance in describing a concept in different languages) one greatly simplifies the conceptual analysis and the interest of the semantic web. There are also, therefore, huge incentives for engaging in "langue" related developments.


6. The presentation layer

However, if all this is confused by the "language/langue-langage" lack of IETF/ISO TC37 differentiation, this is further complexified (as in IDNs) by the fact that Internet technology has no "presentation" layer.

This is perceived as an advantage for e-commerce and e-political English lobbies, because it leads everything to be specified, supported, and operated in ASCII and the entire network to be possibly unified under IANA and ICANN. This is in fact highly detrimental because it prevents the Internet from supporting our (each human being and organisation) lingual, cultural, local, family, private, corporate, etc. visions ("presentations"). This is partly, resulting in most of the architectural problems that we now face. Not only with IDNs, but also with IPv6, routing, governance, etc.

This is in fact a "layer violation" (you do something at the wrong place, which then quickly prevents scalability). Languages, as in IDNs, like in the RFC 4647 filtering approach, etc. are considered as options that are being supported atop the English background. The result is that roughly 150 locale files for 130 languages (some have two scripts) are supported today, while the ISO 639-3 that IETF wants to support documents 7,500 languages. Debbie's ISO 639-6 project targets 20/30,000 language "archeological" entities (they will be long dead when it is accurate!), while new languages appear every day.


7. The correct approach

Whatever the solution, it can only be based on sound premises. This demands the acceptance that there is a layer violation on the Internet side and confusion on the ISO side.

The necessary technical culture is progressing. For example, in chronological order:

* I tricked the IETF in producing an interoperable RFC 4646 text (to support its language doctrine) that is clear enough to be interoperable.
* John Klensin proposed to remove the linguistic aspects from the IDNA and keep them purely script oriented.
* the metacoms concept (transporting information about the content and meaning of information) advances slowly through:
* the Semantic Web concepts,
* the possible IETF OPES,
* the JTC1/SC32/WG2 work on metadata registries, etc.
* my pending proposal
However, we are so embedded in the IETF "English", "decentralized", and "end to end" paradigm that there is some difficulty in understanding that many of the Internet poblems result from there specifically: the differences between the way we perceive reality and the model of virtuality that a limited Internet imposes on us.


The response of the GA of the Internet DNS users should be to complain about the fact that the ISO, IETF, UNICODE, etc. have not yet made it possible to have universal pervasive multilingal Internet support, in turn making the problem obsolete. Amazingly, that complain was absent from the debate. Instead they discussed costs, treeshold computation for HR enforcement, marketing.

This only shows that their assembly do not represent the users and explain why China adopted a grassroots independent strategy, development, and deployment.

jfc


At 14:01 17/07/2007, Debbie Garside wrote:
Language is important to the people that speak it.  It is part of their
cultural heritage.  Whilst English is recognised as a lingua franca around
the world it is important that policies are put into place to protect
"small" languages.  According to a well known linguist, 3000 of the recorded
6-7,000 languages (Ethnologue/ISO 639-3 lists 7,200) currently alive will
die within 100 years.
http://www.crystalreference.com/DC_articles/Langdeath19.pdf

Who cares?  Not governments that's for sure because there are very little or
no funds for language documentation or language revival.  Not industry
because it doesn't make economic sense -
http://www.native-languages.org/linguistics.htm#tree. I have tried for 6
years to get people interested enough to fund language documentation -it is
a thankless task (one of my many). The only people interested are linguists
and the speakers themselves.

Bringing the topic of conversation back in line with the objectives of this
forum, should we translate the NARALO MoU into Inuit/Inuktitut* and Cree**
and other native languages?  Answer, yes we should.  Does it make economic
sense?  Answer, sadly no.  Very often, and certainly in this case, the
translation of documents is about getting information to the most people
possible and this does not mean translating into languages that represent
the mother tongue of less than 1% of the population albeit 100% of the
indigenous population; especially where these people have knowledge of a
second language - in this case English and French.  Sad but true.

Best

Debbie
PS Anyone interested in language documentation or language
protection/revival may make large donations to The World Language
Documentation Centre www.thewldc.org I happen to be CEO :-)

*http://en.wikipedia.org/wiki/Inuit_language
**http://en.wikipedia.org/wiki/Cree_language


> -----Original Message----- > From: owner-ga@xxxxxxxxxxxxxx > [mailto:owner-ga@xxxxxxxxxxxxxx] On Behalf Of Roberto Gaetano > Sent: 16 July 2007 22:37 > To: 'Jaap Akkerhuis'; ga@xxxxxxxxxxxxxx; martin.boyle@xxxxxxxxxx > Subject: RE: [ga] Aviation languages > > Jaap, > > > > > As far as I know English is mandatory. > > > > It is. And there is a protocol to switch to another > languages in case > > the parties desire. > > > > For more details, google for "aviation languages" and similar terms. > > > <maybe off topic, or just borderline...> > > With all due respect, I made the statement about languages > not because I *thought*, but because I *know* from personal > experience that Italian can rightfully be used under certain > circumstances by pilots while flying over the italian air > space. Specifically, the use of italian is lawful for all VFR > flights, that are still the vast majority of the flights > where amateur pilots engage. And since it is the pilot who > initiates the conversation, for instance with the control > tower when still in parking position, there is no protocol > needed: he/she chooses the language. > Please have a look at the document of the National (Italian) > Authority for Civil Aviation: > http://www.enac-italia.it/documents/download/nor/Reg-Regole_Ar > ia.pdf. This is updated as of 2006-10-03, and has clear text > on paragraph 3.10 (in italian). > Incidentally, this text is similar to the one used in > different countries, simply because it is taken from the > Annex 10 to the Convention on International Civil Aviation, > issued by ICAO. > > I do acknowledge that English is, and will remain for the > foreseeable future, *the* language used in international > situations (like this list, for instance). But also that > there are local situations (that also are here to stay for a > while) where communication is done using local languages. I > know that opinions vary about to which extent an > international (common) language should be used vs. a national > (local) language. Dutch and French, to quote an example.... ;>) > > Best regards, > Roberto > > > >


<<< Chronological Index >>>    <<< Thread Index >>>