Re: [edict-jmdict] Entries such as ～朝時代

["Francis Bond" <bond@********>]

G'day,

how fantastic to find a fellow Francis in the fray.

2008/10/22 Jim Breen <jimbreen@gmail.com>:
> People who have been following the discussion going on in the
> WWWJDIC feedback system will have seen Francis Miller's
> arguments for entries such as 年生まれ and 朝時代. I have
> been resisting as to me they are text fragments rather than
> expressions in their own right.
>
> Having said that, I can see some sense in having
> something like:
> ～朝時代 the era of the ～ dynasty
> or
> ～年生まれ born in ～
> available as some sort of guide to people who don't
> recognize such constructions when they encounter them.

I think they would be a great idea.   In a perfect world, replacing ～
by the part-of-speech tag of the missing thing would be useful, but I
suspect deciding on the the POS inventory we allow and then entering
it consistently would be too difficult.

> Adding such entries brings other problems, e.g.
> - what is the "reading" of ～? I guess it would have to be
> null, and ～朝時代 would simply be ちょうじだい.
> - people encountering 朝時代 wouldn't know that the
> "entry" started with a ～, and would miss it unless they
> did as Francis did and did a search for the string anywhere
> in an entry.

I think we should also have the ～ in the reading, as it make it more
clear. People normally read it as ニョロ when they read dictionary
examples, but I would expect the kanji to be easier to follow.  Would
it complicate indexing a lot to index from position two iff position
one is ～ (and I guess there is more than one character, as we would
want to index ～  itself)?

I notice that the Lawdic already has several such entries:
〜講義 【〜こうぎ】 Lectures on -; LW1 [G][GI][S][A]

> Anyway, I'm looking for some feedback as to the usefulness
> of these sorts of possible entries.

I consider them very useful.   If we could have a policy that they
should be accompanied by an actual example, it would make their use
even clearer.

-- 
Francis Bond <http://www2.nict.go.jp/x/x161/en/member/bond/>
NICT Language Infrastructure Group