[topicmapmail] Variants

[Motomu Naito]

Dear Heuer-san,

> Hi Naito-san,
> 
> Thanks for your mail. The problem is not somewhat clearer to me.
> 
> > I will show you the example again.
> [...]
> > XXjkSATO_HACHIRO
> [...]
> >  - alias: "XXkkMUTSU_HAYAO"
> @Kanji                                   # 7
> >  - alias: "XXjhMutsu Hayao" @ Kanji, Hiragana, Reading               #
> 8
> >  - alias: "Mutsu Hayao" @ Kanji, Romaji, Reading                     #
> 9
> >  - alias: "XXkkYAMANO SABURO" @
> Kanji                                # 10
> >  - alias: "XXjhYamano Saburou" @ Kanji, Hiragana, Reading            #
> 11
> >  - alias: "Yamano Saburou" @ Kanji, Romaji, Reading.                 #
> 12
> >
> > But I can't specify # 8 is not Reading of # 10 but Reading of # 7.
> > I can't specify # 11 is not Reading of # 7 but Reading of # 10.
> 
> That's true. The version which uses names only does not imply any kind
> of hierarchy. Without modifying the example further (i.e. introducing
> another theme or another name type, or using reification) it is not
> possible to specify that #11 is meant as reading of #10.
> 
> I'd tend to use reification to create such a hierarchy (if the
> hierarchy is necessary at all).
>

Thank you for your understanding of the problem. I feel relieved a little.

Yes, I think we need some kind of hierarchy.
I agree I can use another name type or reification technically.

Concerning another name type, in the example I can introduce the new name types such as "penName" or "stageName".
But some people have two or more "penName" or "stageName". So those name types may have the same problem. The new name types such as "penName_1", "penName_2", ... seem to be nonsense. Because those types are the same level and have the same meaning.

Concerning the "reification", I think it's not intuitive and a little bit hard to use and understand especially for users i.e. not Topic Maps experts. So I think the "variant" is superior to the "reification" in this case.

> But what does it mean practically if I use variants? Does it mean,
> that #11 only exists if #10 exists? Enables the existence of #10 the
> existence of #11? What happens if I know #11 but not #10: I've to add
> #11 as name do to the fact that I don't about the existence of #10.
>

It is a delicate and interesting matter. It is hard to explain in English for me though. Let me try.

Kanji character is the main character for expressing things (subjects) including person names in Japanese.
In Japanese Kanji character has two way of reading Chinese-style reading and Japanese reading. Because we imported Kanji characters from China and apply them to Japanese speaking words in ancient times I think. Each way has one or more reading for each Kanji character. So we have at least two or more reading for the characters. The combination of Kanji characters become more complicated. In some degree we can guess the reading logically but those guesses are not always right. There are a lot of irregulars. Sometimes we can read them rightly only when we have already known the reading. Especially person names and place names are the worst cases. So we need #11 to show the right reading of #10.

In contrast, Hiragana and Katakana characters are phonogram. Those are used for the reading in this example. We can assign many different Kanji characters to the same Hiragana and Katakana characters. And we can't decide corresponding right Kanji only from Hiragana and Katakana. If there was only #11 we could know the reading of the person name at least but there was no way to know the right Kanji i.e. right name of the person.

I hope the explanation is understandable for you.

Best regards,

Motomu Naito

> 
> Best regards,
> Lars
> --
> Semagia
> <http://www.semagia.com/>