Loading ...
Sorry, an error occurred while loading the content.

1228Re: Relationship from concept to non G2 entities

Expand Messages
  • Jo
    Sep 13, 2011
    • 0 Attachment
      Hi Misha,

      In line below.

      --- In newsml-g2@yahoogroups.com, misha.wolf@... wrote:
      > Hi Jo,
      > Long time no see
      We should rectify that off list!

      > Consider Yugoslavia. Constituent republics peeled away, one after the other, but the Alpha-2 code "YU" was retained till (alomost) the end. Was each instance of "Yugoslavia" the same as the one before? Was "Yugoslavia" without Croatia the same country as "Yugoslavia" with Croatia?

      In my opinion it's important to understand what the states themselves think. Whether you or I think it's the same country is probably not as important as whether the country considers itself the continuation of the earlier entity or not. There are numerous examples such as the YU example you quote where despite the loss or gain of territory, population etc. the country considers itself to be the same country. [On the other hand, when Czechoslovakia split into the Czech Republic and Slovakia neither state continued to use the old alpha or numeric codes - presumably indicating that neither country considered itself to be the successor]

      We may possibly agree that whether it is the same or not depends on your point of view and what attributes are the defining characteristics of sameness or difference. Either way, it would appear that the alpha code system in ISO-3166 has different criteria to the numeric codes for judging sameness.

      > Any analysis of indicators, eg territory, population, GDP, breakdown of state expenditure, etc, must take into account that a country before such a merger/de-merger is not the same country as the one(s) before/after the merger/de-merger.

      I agree, but is only one set of characteristics that you might consider. Does it have the same executive, the same flag, the same diplomatic representatives overseas would be another - it depends on the application, doesn't it?

      > The fact that many characteristics survive the merger/de-merger does not mean that an entity after the merger/de-merger is the same entity as a similarly-named one before the merger/de-merger.

      Nor is it necessarily the case that it is not the same.

      > In Thomson Reuters we assign Permanent Identifiers to all concepts (including entities) we consider useful, and link these using typed relationships. In our system, the "old" Sudan and the "new" Sudan have different PermIDs.

      That's interesting, but is presumably based on the fact that a lot of your customers track indicators of that kind, and makes their analysis easier in some respect or another. It presumably makes it harder to track, say, voting patterns at the UN General Assembly.

      I sometimes quote a remark that I attribute to you, from many moons ago, which is that you can't talk about metadata, you have to talk about metadata for some purpose. Which I have interpreted as meaning that there is no single objective stance from which you can make a judgment.

      According to that, Sudan is both the same country as it was before and a different country, depending on the purpose for which you are making the assessment.

      Coming back to your original statement: "In case someone thinks that the continuity of the country's name indicates a continuity of identity, they are mistaken. " For reasons that I hope are clear and make some kind of sense my point is that this might be better put as "In case someone thinks that the continuity of the country's name indicates a continuity of identity, whether they are mistaken or not depends on whether the maintenance policy of the naming scheme is consistent with their intended purpose. "

      If you're interested in doing assessments that involve territorial extent it would be better to use the numeric scheme, if you're interested in doing assessments that are primarily political then the alpha schemes may be better. Either way, be careful.

      That's all

      > Regards,
      > Misha
      > -----Original Message-----
      > From: newsml-g2@yahoogroups.com [mailto:newsml-g2@yahoogroups.com] On Behalf Of Jo
      > Sent: 12 September 2011 16:49
      > To: newsml-g2@yahoogroups.com
      > Subject: [newsml-g2] Re: Relationship from concept to non G2 entities
      > Hi Misha
      > re "In case someone thinks that the continuity
      > of the country's name indicates a continuity of identity, they are mistaken. " - I'm intrigued.
      > Though the territorial extent of the old Sudan has changed, most other aspects of it have not, for example afaik it is still the same geopolitical actor, continues to have representation abroad in the same form etc. I think the numeric code has changed because the territorial limits have changed and that the ISO alpha code has remained the same because the (continuing) government of Sudan did not request a change to its code.
      > A similar thing happened for Germany, I believe, upon the integration of the former DDR. That doesn't mean, afaik, that DE represents a different country (qua geopolitical actor) before and after the change of numeric code, though likewise its territorial extend did change.
      > Contrariwise, a while ago, Zaire became the Democratic Republic of the Congo and changed its code from ZR to CD but kept its numeric code of 180.
      > A good illustration of the need to understand the policies behind code maintenance, and in this case possibly a good illustration of the need for clarity in distinguishing a geopolitical actor and its associated territory. You can take a skiing holiday in France (location) and make a treaty with France (geopolitical actor) but you can't make a treaty with a location nor take a skiing holiday in a geopolitical actor.
      > For some applications it may be useful to distinguish SDN pre-9th July 2011 from SDN post-9th July - perhaps one should use a different alias from within NewsML?
      > Likewise, by analogy, for Philippe's question. Ideally he'd have a chronology of the introduction of and deprecation of codes, of course.
      > Best
      > Jo
    • Show all 12 messages in this topic