Re: [edict-jmdict] Conjugations and PoS tags for だ, くれる

[René Malenfant <rene_malenfant@***********>]

I think the point is that いい does not inflect.  Historically, it’s just a shorthand way of saying よい that’s come to dominate in 終止形 and 連体形 forms.  So よく, よくない, etc. all belong to よい and いい doesn’t really have any inflections of its own.  So does it really need a PoS all to itself if it arguably doesn’t even inflect at all?  [exp], etc. should do.



Rene



On Jul 24, 2014, at 8:48 PM, Stuart McGraw smcg4191@frii.com [edict-jmdict] <edict-jmdict@yahoogroups.com> wrote:

> On 07/22/2014 10:08 PM, Jim Breen jimbreen@gmail.com [edict-jmdict] wrote:
> > On 23 July 2014 08:16, René Malenfant rene_malenfant@hotmail.com
> > [edict-jmdict] <edict-jmdict@yahoogroups.com> wrote:
> > 
> >> Those merged いい/よい expressions are cumbersome and ugly, and I’m fine
> >> with a complete いい vs. よい split, as long as there’s a note in the main entry
> >> for 良い[よい] that says that it can also be read as “いい” and x-refs between
> >> the main いい and よい entries.
> > 
> > OK, let's go with a complete split.
> > 
> > There is a question as to whether the 良い entry should have the reading
> > いい or not.
> > - Koj doesn't - it has よい and the old えい. In fact it doesn't have いい
> > at all (5th ed.)
> > - Daijr does, as do GG5 and 中辞典. GG5 has 11 senses for the 良い/いい entry and
> > 9 for the 良い/よい entry. Daijr says at the front: "形容詞「よい」の終止形".
> > - ルミナス doesn't, having kana only for its いい entry.
> > 
> > My preference is to leave 良い off the いい entry, and to have an "also read いい"
> > comment on 良い/よい.
> 
> In that case 良い would conjugate like all other 'adj-i' words
> and would keep its 'adj-i' PoS I presume.
> 
> However, いい still conjugates differently and thus I would like
> いい to have a PoS that identifies this difference ('adj-ii'?)
> 
> >> A [cop] PoS might work for だ, but I don’t think it helps the conjugation problem
> >> any. だ isn’t the only copula, so I think [cop] would be applied to many entries.
> >> In the end, it would just be another grab-bag category, and I think that role is
> >> already filled by [aux].
> > 
> > One of the troubles we have is I am not sure we can really regard "copula" as a
> > part-of-speech. It's more a role played by a word (usually a verb, but
> > it depends
> > on the language.) Of course the concept of "part-of-speech" is rather
> > woolly. The
> > OED has the POS of "the" as "demonstrative adjective", then glosses it
> > as "definite
> > article". Role vs POS.
> > 
> > Then, as René points out, there are copula other than だ in Japanese. Strictly
> > speaking we should be having です/ある/いる/おる/そろ/(で)ござる/おじゃる/et al. labelled
> > too if we are going to tag だ, and of course they inflect in different ways.
> > 
> > I guess I have misgivings about assigning "cop" to だ alone, and implying it
> > indicates a class of inflections.
> 
> Then the problem here is not an objection to assigning a 
> unique PoS to だ (because it has a unique conjugation pattern)
> but that the suggested PoS was spelled 'cop' (copula)?
> 
> Then how about spelling it differently: 'da', 'cop-da', 'da-cop'
> (ok, that last one sounds like a line from a B-grade movie.)
> 
> To summarize (since some of the discussion is becoming a little 
> tangential to my request), there are close to 24000 common verbs, 
> adjectives and 'vs' entries (and close to 12000 excluding the 'vs' 
> entries) that are conjugatable based solely on their PoS.
> (For comparison all the other v2*, v4*, aux, aux-v, vr, vn, vz
> entries total 234.)
> 
> There are 3 common and important entries that are not conjugatable 
> based on their PoS: だ, くれる and いい. I would like those entries
> (and preferably related entries like 頭がいい) to get a PoS that 
> allows them to be conjugated based on PoS the same way the other
> 24000 entries can be.
> 
>