Re: Artificial Bureaucracy - Language Codes

Dear Gannon,

we have not gone far enough into the weeks of HXL to reach a point
where we see the need for data about languages spoken by persons or in
organizations. We are incrementally going through existing datasets
that we see should be representable in HXL in the future, and we have
not come across this case yet. Having that said, this may well be the
case, and the two types of URIs you mention support an important
distinction, IMO. Once we reach that point, the portal you have made
will be really useful for us, thanks for this effort!

- Carsten

2012/2/28 Gannon Dick <gannon_dick@yahoo.com>:
> All,
>
> "Artificial Bureaucracy" is like Artificial Intelligence (AI) for Civil
> Servants.  A very important tool for a  bureaucrat are codes and standards -
> a language only they speak.  The codes used in the standards function as
> encryption to keep out "the enemy" both foreign and domestic, and BTW, that
> includes citizens.  Since every e-Gov uses only a small subset of Country
> (ISO 3166), Language (ISO 639) and Currency (ISO 4217) - one size does fit
> all - it is practical to make up a Repository Profile from a single DCMI
> subject list, and without believing that everybody speaks English because
> that's the default language of the national website.  But I'm getting a bit
> ahead ... The intent of "Artificial Bureaucracy" is to do away with the
> Codes in favor of Names (from an IT perspective, Name Tokens which are
> themselves language neutral).
>
> The ISO 639 Language Codes present a special concern.  There are two
> different sets of Name Tokens needed, and to mix them up is to invite false
> inferences about the meta data:
> 1. Naming the Website Display Page Language or the language of the text in a
> data set; and
> 2. Specifying a Property of a Person or Organization - a population or
> person speaks, reads writes, etc. a particular language.
>
> Neither the The Core Vocabularies Working Group [1] nor the Humanitarian
> eXchange Language (HXL) [2] address this, AFAICT.  This means to me that
> LOD'ers need it spelled out.  The Specification makes the distinction, but
> not right out loud.  The three letter codes have an extra
> {bibliographic|terminology} attribute.  The two letter codes have no such
> attribute.  So, for eGov work, the two letter codes refer to a display and
> the three letter bibliographic codes refer to a Person, with the three
> letter 'terminology' codes acting as alternates to the two letter codes.
>
> It's fine to be friendly (and in the meantime promote tourism) on websites,
> but there are other circumstances, humanitarian causes for example, where
> more accuracy is necessary.  Any thoughts (while I go home and eat dinner,
> and leave details until tomorrow) ?
>
> --Gannon
>
> [1]
> http://joinup.ec.europa.eu/asset/core_business/document/core-vocabularies-working-group-members
> [2] http://carsten.io/hxl/ns-2012-02-22/index.html
>

Received on Thursday, 1 March 2012 09:07:52 UTC