Re: [edict-jmdict] 愛する issues

Not wishing at all to be disruptive, I must say that I have about as little understanding of your verb classifications as a man on the street in Kyoto or a kid in the Japanese school system. I learned Japanese in total immersion way up in the country near kashihara in Kansai, a place where I absolutely had no recourse to English. In fact, that is the very reason I moved into this small country village - to get away from the foreigners and the "eigo-ya". We had no books, relevant schools or internet in those days - only Japanese friends and of course the television.

So when you talk about v5s and GG5 verbs, etc., I must confess that I do not know what the hell you are talking about. I suspect that most people who make use of Edict, Yahoo, Goo and ALC, the last three of which are considerably easier to navigate, are just as lost and confused as I am. For what it's worth, my Japanese is adequate to the point that the person on the telephone is not aware that I am not Japanese. It's always great fun to meet them in person after first meeting them on the telephone.

So I would be one to vote strongly that we shove the pseuod-intellectual analytics and over-classifications to the back, dusty shelves of the library and go about the business of simplifying this wretchedly over-complicated dictionary which is usable only by those in a special fraternity and not by those of us who only speak the language.

If I might be a bit sarcastic for a moment, all this Battleship Yamato needs to sink itself is more labels, buttons to select and pretentious nomenclature. While we're at it, I would suggest all the fonts be pink.

Example of the "Logical Mess" as follows:

I'm very tempted to:

(a) for the 愛す, etc. verbs simply call them "v5s". In what way will they
not conjugate correctly (if at all)?

(b) leave the verb tables as they are for the "vs-s" verbs, but add a footnote
saying that the "さない", etc. inflections really come from the す version of the
verb, but thse days many Japanese regard them as part of the する version. I note
that big JEs such as GG5 simply don't even list 愛す.

2008/9/11 René Malenfant <rene_malenfant@***********>

1) If you remove the (vs-x) from 愛す(v5s,vs-x,vt), it will
cause no problems.

As I understand it, the conjugation chart for 愛する(vs-s)
contains all possible conjugations for 愛す(vs-x) (and then
some) save for one: 愛す itself (which is used as a 終止
形). If you removed the (vs-x) tag, 愛す would still
exist in the database as (v5s,vt), so there would never be any lookup
problems.

The problem with removing the (vs-x) tags comes from the other words
like 閲す(vs-x) and 謁す(vs-x) that would cease to exist
as entries if you removed the tag. Of course, that could be
okay. す is a predictable classical form of する, just
like 四段 is a predictable classical form of 五段. I
think the only reasons Daijirin makes an effort to include 閲
す, 謁す, etc. are: 1) to draw attention to which of
the ～す forms can be used as (v5s) and which ones can't; and 2)
for completeness/consistency.

You could go the same route with 閲す, 謁す, etc. that you
did with 得る[うる]. Retain them, and label them as
(exp), letting the conjugation work be done entirely by 閲す
る, 謁する, etc. Of course, that doesn't make as much
sense here. 得る[うる] is a single throw-away irregular
"verb", while 閲す, 謁す, etc. are members of a large
regular verb class.

2) I don't like the idea of leaving ～さない as the negative
conjugation in the (vs-s) chart. As I pointed out, there are lots of
other conjugations for which the 愛す(v5s) form is prevalent
over the equivalent 愛する(vs-s) form. I don't see any reason
to single out ～さない and leave it there. Plus, doing so
leaves 愛しない (which is 100% grammatically correct, and not
uncommon--20k Yahoo hits) without an entry.

Besides. If you look at the (vs-s) entries for which I *didn't* add
a (v5s) equivalent, I'll think you'll find that ～さない is far
less common than ～しない. For example (going by Yahoo here
because I don't trust Google's results quantitatively, although they
always agree that ～しない is greater):

"関しない" = 71000 hits; "関さない" = 934 hits
"対しない" = 18000 hits; "対さない" = 565 hits
"了しない" = 2210 hits; "了さない" = 9 hits

However, the less-common ～さない form is still the one that's
showing up for each of these verbs on their conjugation charts.

I say just x-ref all of the ～する(vs-s) forms to their
equivalent ～す(v5s) forms and vice versa. People will figure
out what's going on.

Rene

On 11-Sep-08, at 1:54 AM, Jim Breen wrote:

> We've been having a discussion on the EDICT/JMdict feedback system
> about the conjugation of Xする verbs. I want to move it here, as
> that system is a bit clunky for proper dialogues.
>
> The issue raise was whether WWWJDIC's tables were correct about the
> -さない inflections, which is what I'd been taught (by NSoJs).
> That generated page, BTB, had been vetted and cleared by some NSoJ
> language teachers too.
>
> Rene commented:
>
> =========================================================
> Yes. I'm close to 100% certain that they're wrong.
>
> They aren't even self-consistent. Like the page assumes that
> "愛さない"
> uses さ as its -nai stem, but "愛せず" uses せ as
> its -nai stem for some reason.
>
> The former is properly attributed to 愛す (v5s), and the
> latter to 愛する (vs-s)
> or 愛す (vs-s). The ず form of 愛す is 愛さ
> ず, which isn't accounted for on the
> conjugation page at all. (Even though 愛さず is more common
> than 愛せず based on
> web hits.)
>
> Compare (the ones with asterisks aren't accounted for on your
> conjugation page):
> 愛せず (from 愛する/愛す vs-s) = 23,500 googits
> 愛さず (from 愛す v5s) = 141,000 googits*****
>
> 愛しよう (from 愛する/愛す vs-s) = 7,520 googits
> 愛そう (from 愛す v5s) = 141,000 googits (although some
> are probably false
> positives)*****
>
> 愛できる (from 愛する/愛す vs-s) = 2,120 googits
> 愛せる (from 愛す v5s) = 458,000 googits*****
>
> 愛すれば (from 愛する/愛す vs-s) = 82,900 googits
> 愛せば (from 愛す v5s) = 66,100 googits*****
>
> 愛せ brings up too many false positives to make a comparison
> with
> 愛しろ/愛せよ.
>
> Now, Wikipedia and kokugo dictionaries DO list さ as one
> possible 未然形
> ending for する in modern Japanese:
> し（せ・さ）・し・する・する・すれ・しろ（せよ）
>
> So it seems *possible* that you could attribute さない
> and さず to する.
> However, the さ here seems to be limited to use when creating
> the passive
> さ-れる and the causative さ-せる.
> The ない is always supposed used as し-ない.
>
> Here's the smoking gun, from Meikyo's entry for ない:
> 【語法】(1)動詞、動詞型活用の助動詞の未然形に付く。サ変
> 動詞は「し」に付いて
> 「しない」となる。*****なお、「愛しない」は「愛する
> (サ変)」に、「愛さない」は
> 「愛す(五段)」に付いたもの。*****
> ======================================================================
> ==========
>
> and later:
>
> ======================================================================
> ==========
> One more thing.
>
> I assume that the reason that する is labeled as (vs-s) is
> because it was being
> considered as having that extra "special" conjugation さない
> in the case of things
> like 愛する and 介する, but not in the case of regular
> (vs) verbs like 勉強する and
> 表現する. I think it would actually make a lot more sense, and
> be more consistent
> to flip things around.
>
> If the カ変 verb 来る is tagged (vk), not (vk-s), then
> the サ変 verb する should
> be (vs), not (vs-s). The (vs-s) tag would actually be better on
> things like 勉強
> and 表現, since the additional "s" could imply that you have
> to "suffix" the
> "suru" yourself.
>
> Of course the easier path to consistency would just be to change (vk)
> -> (vk-s),
> but that seems a little less sensible.
> ======================================================================
> ==========
>
> I'm very tempted to:
>
> (a) for the 愛す, etc. verbs simply call them "v5s". In what
> way will they
> not conjugate correctly (if at all)?
>
> (b) leave the verb tables as they are for the "vs-s" verbs, but add
> a footnote
> saying that the "さない", etc. inflections really come from
> the す version of the
> verb, but thse days many Japanese regard them as part of the
> する version. I note
> that big JEs such as GG5 simply don't even list 愛す.
>
> Jim
>
> --
> Jim Breen
> Honorary Senior Research Fellow
> Clayton School of Information Technology,
> Monash University, VIC 3800, Australia
> http://www.csse.monash.edu.au/~jwb/
>
>