Latest comment: 13 years ago5 comments5 people in discussion
Is it just me or does this template need a lot of work? I'm not an expert in Japanese, but it seems to be lacking (for example the verb 食べる) the form 食べません or tabemasen- and should it perhaps show "translations", like under "tabemasu" there would be written "I/you/he/we ----" so people can infer that "tabemasu" can mean "I eat" or "we eat"? --BiT22:14, 30 October 2007 (UTC)Reply
While this reply is eternally late, yes, it needs working. I'll be putting it in my userspace for some work and table-isation VNNS16:39, 20 March 2010 (UTC)Reply
Rather than deixis along the lines of the verb actor -- who's doing what -- Japanese has deixis along the lines of audience -- who's being spoken to, and the social relationship with the speaker. Consequently, Japanese has no concept of grammatical person, so there's no need for or utility in adding "I / we / etc". Also consequently, Japanese does have an extensive system of honorifics, which is where the difference between plain form and formal desu/masu speech comes into play. This is a bit like the tu - vous distinction in French, tu - usted in Spanish, or du - Sie in German. English used to have this too, and we can still find vestiges of this in Shakespeare's use of thou versus you.
Then again, you probably already know this. :) Ultimately, though, if someone is going to the trouble of learning how to conjugate Japanese verbs, I think it's safe to assume they've already learned that there's no grammatical person involved.
@Everyone --
I'm making a minor alteration to the beginning of the template to match Template:es-conj-er etc., as it seems significantly clearer to me to have the link to the Appendix more explicitly indicated.
Incidentally, what do you all think of reworking this table a bit? It seems odd to have the "Formal" row and then not have "Perfective Formal" and "Negative Formal". For that matter, ます is just a plain old straightforwardly conjugating 助動詞 -- so shouldn't we just point folks to the Appendix and explain the formal constructions there?
@Everyone --
I started this template many years ago when Wiktionary was just a wee experiment ... nice to see that it's survived this long. Just to address a couple of points that have been brought up: my assumption is that anyone who is learning Japanese presumably has learned how to conjugate "-masu", and if not, should probably go somewhere else to learn that so as not to make this template unnecessarily long. The "stem forms" section was meant to show all conjugations as formally taught in Japanese schools, the "classical" view. Since these forms are not adequate to express the variety of ways in which verb stems change in actual modern usage, however, I included a "key constructions" section for the purpose of ensuring that every possible conjugation a Japanese student would be likely to encounter in actual usage was represented. I included only constructions for which irregularities existed that needed to be explicity called out: for example, I included "formal" because "-masu" doesn't always take the continuative stem, as in "irassharu" becoming "irasshaimasu" instead of "irassharimasu". I omitted the "non-hypothetical conditional" form "-tara" because in every case, "-tara" conjugates the same as the perfective "-ta".
The reason I kept this base template separate from the sub-templates {{ja-ichi}}, {{ja-go}}, etc. was because I wanted it to be easy to modify the template to add forms or verb subtypes without having to redraw the entire template. I hope it's been useful, and of course, I'm sure it can still be improved.
Of course, one needs to copy the accent type from the pronunciation section to the conjugation template, which is huge work for the existing entries.
(Actually, I came up with the idea when thinking about applying the current Chinese entry layout to Japanese, grouping Modern and Classical Japanese together and having a pronunciation template like this (using あじわう as an example):
{{ja-pron
|m=あじわう // modern
|macc=3,0,ref1=DJR,ref2=DJR
|mcat=v5
|c=あぢはふ // classical, can also be used to infer the historical spelling of modern form (maybe)
|ccat=v4,vl2
}}
Then it occurred to me that this was enough information to generate the conjugation box from the ja-pron template. I'm not sure if it is feasible, though.)
Interesting idea. I'm wondering if the accents of conjugated forms can be accurately predicted from the lemma forms in all cases...? Wyang (talk) 22:53, 4 September 2018 (UTC)Reply
I like the general idea. Some of the pitch patterns above look a little odd, however. For instance, I'm more used to hearing the imperative with the final mora at a higher pitch, more like [hànáséꜜ]. The hypothetical conditional might also have the downstep at the end. For the plain negative, I feel like the downstep comes in the middle of the negation suffix, as in [hànásánáꜜì]; and I recall reading somewhere of an analysis of the ～た・～て forms pointing to the downstep technically coming after the し, but since that vowel is often unvoiced, it might sound like it comes before.
@Eirikr: The accents listed above are from the Online Japanese Accent Dictionary (OJAD). Is there great variation in practice? Is there any standard accent of conjugated forms we can present to learners without being biased? As for the difference between the classical form and the modern form in historical spelling, the only ones I know is due to change in conjugational class: （二段）たづぬ→（一段）たづねる，（ラ変）をり→（四段）をる， or due to rentaikei replacing the old shuushikei く→くる，す→する --Dine2016 (talk) 03:37, 5 September 2018 (UTC)Reply
In Tokyo accent [ha˨.na˦ꜜ.se˨] and [ha˨.na˦.sa˦ꜜ.na˨.i˨] are correct. Verbs are either accented or non-accented, and their accent positions are predictable (i.e. there is no accent position difference in verbs). Minor “exceptions” are long vowels and diphthongs such as 通る(tōru) and 入る(hairu), which are not [to˨.o˦ꜜ.ɾɯ˨] and [ha˨.i˦ꜜ.ɾɯ˨] but [toː˦˨ꜜ.ɾɯ˨] and [hai˦˨ꜜ.ɾɯ˨]. (Here a downstep of a long vowel or a diphthong is realized as falling tone, and therefore they are not really exeptions.) — TAKASUGI Shinji (talk) 04:33, 5 September 2018 (UTC)Reply
@Dine2016: Ah, we were talking past each other: you're describing a difference in grammatical forms (which did change, as you note), whereas I was describing a difference in kana spelling of a single form (which didn't change, I think, until the spelling reforms). I've got you now.
@Shinji: My living experience started in the Tōhoku, in Morioka. There's definitely a difference in pronunciation there, especially among older speakers -- らりるれろ sounds much more like /la li lu le lo/, for example, with a real [l] sound rather than the apical tap [ɾ] typical in Tokyo. Any chance there's also a different pitch pattern up there? For instance, [ha˨.na˦ꜜ.se˨] and [ha˨.na˦.sa˦ꜜ.na˨.i˨] sound super unnatural to me, and I can't think why that would be unless I learned a different pattern. When I later moved further south, I was consciously aware of other clear pitch differences, such as for the term 居酒屋(izakaya, “pub”). ‑‑ Eiríkr Útlendi │Tala við mig16:06, 5 September 2018 (UTC)Reply
Very interesting, thank you! I was taught about the Kansai - Kantō difference in pitch accent quite early on, but I was not aware of other finer distinctions. 東京式アクセント has been a most interesting read. Much appreciated! ‑‑ Eiríkr Útlendi │Tala við mig23:42, 5 September 2018 (UTC)Reply
I think this is a good idea. It is especially useful since entries for conjugated forms often don't exist. I would also suggest something similar be considered for -い adjectives as well since (at least in Tokyo dialect) they seem to also be predictable given an accented/unaccented status. If this were to happen, it would be nice to add an extra label to give the template other than just the current system of identifying an accented mora number - something semantically equivalent to "unaccented verb" or "accented adjective". At least in Tokyo dialect, this would match the actual degrees of freedom of the verb or adjectives's accent pattern. If entries can be identified as -い adjectives or of one of the verb classes then conversion to an "accented"/"unaccented" label should be quite easy to automate by checking if the listed pitch is 0 or n-1, or both if some entries have references to both types. (Aogaeru4 (talk) 18:45, 6 September 2018 (UTC))Reply
Just adding on to myself, we should be a bit careful since the accent interaction between the base of the verb and the various suffixes that can be placed on them is predictable but not obvious. This is why (OJAD) mentioned above has such a long list of forms for each verb. I have seen some descriptive rules that describe the patterns but they end up being tough for the typical person to use vs. a full pre-made conjugation table. Thus having such a brief table as shown above might be misleading to a typical user unless there is a link somewhere to describe the other forms. Aiming for something similar to what OJAD would be useful but will take lots of space and might not combine well with the size people want for the typical conjugation box. (Aogaeru4 (talk) 19:01, 6 September 2018 (UTC))Reply
Yes, I think we should include this in a collapsed section. Probably this would be more relevant in the pronunciation section than in the verb conjugation box though. Some thought definitely needs to go into the formatting of the section due to its potential size. I'm not quite sure how the formatting from the Chinese pronunciation section would carry over since I think the intent and categorization is different there. (Aogaeru4 (talk) 21:32, 17 September 2018 (UTC))Reply
Interestingly, OJAD says that 帰ります has accent 4, while Web茶まめ analyses it as having accent 1. (Type 帰ります to the text box, check アクセント型 in 出力項目, and you get 帰り [accent: 1] + ます [accent: empty].) --Dine2016 (talk) 15:32, 10 September 2018 (UTC)Reply
I think this is a mistake on the part of Web茶まめ. It looks like Web茶まめ is analyzing 帰ります as two separate words, which is not valid for pitch accent. In verbs "ます" is a pitch-dominating suffix, so the accent pattern of ます, 1, overrides the typical pattern of かえる which is type "accented" => (n - 1) => (via sometimes ignored sequential vowel rule) (n - 2) = 1. So 帰ります should be analyzed as (n - 1). Earlier I commented that the system of pitch accents when combined with suffixes can be complicated. This is just one of those instances. Only some of the possible suffixes change the accented status of the verb and only some of them move the location accent (ます yields n - 1). (Aogaeru4 (talk) 21:19, 17 September 2018 (UTC))Reply
