I just finished reading this paper, which describes the situations in which Japanese allows a dative subject and a nominative object. For example, the verb 分かる can be used like this:


"彼" is the subject but is marked with "に" instead of "が", and similarly "英語", the object, is marked irregularly with the nominative particle "が" instead of the usual "を". According to the paper, this happens in Japanese when a verb 1) licenses this case marking and 2) the verb is transitive (there is an object). There are two parts to my questions:

1)The 可能形 licenses dative subjects:


but not


This is still ruled out for intransitives, so 5a in the paper shows that you cannot say 「*赤ちゃんにもう歩ける」. Does this change if we add another argument such as a location? Can I say 「人間にはその道が歩けない」 (let's say it's covered in lava or something).

  1. The paper also mentions that there are exceptions to this rule, one of them being certain kinds of questions. Can anyone think of a Japanese question with a transitive verb or a verb that doesn't normally allow a dative subject, but which has a dative subject anyway?
  • I cannot think of examples of the other exception either "not in embedded clauses(Shibatani 1977: 807, Dubinsky 1992)". I don't feel like paying 38$ for the referred paper, but it would be interesting to hear any listed examples.
    – dainichi
    Mar 14, 2012 at 8:28
  • I've requested it through interlibrary loan, so if I get a copy I'll post them here.
    – Nate Glenn
    Mar 14, 2012 at 16:46
  • 1
    Btw why do you say that in the sentence "彼に英語が分かる", "彼" is the subject? If the translated sentence is "English is understood, as for him.", shouldn't "英語" be the subject?
    – Pacerier
    Mar 30, 2012 at 21:41
  • I would not translate it that way, and that is not how it is treated in the literature.
    – Nate Glenn
    Mar 30, 2012 at 23:37
  • 1
    By way of counterpoint, translating is entirely the problem -- after translating, you're looking at it in terms of the target language (English), not the source language (Japanese). This is deeply problematic -- especially for languages as grammatically different as English and Japanese. Analyzing Japanese after translation into English is going to inevitably introduce all kinds of opportunities for misapprehending how Japanese works. I think @Pacerier has the right of it here. Mar 14, 2023 at 17:28

2 Answers 2

  1. >Can I say 「人間にはその道が歩けない」
    Yes I think you can say that. Maybe you can also say 人間にはその道は歩けない/人間にその道は歩けない.
  2. Hmm... Would it be something like... you can say 君に(この車が)運転できるかい? but not 僕に運転できます。?? 私には耐えられない/私に耐えられるだろうか but not 私に耐えられます。?? or maybe 私にやって行けるだろうか/私にはやって行けない but not 私にやって行けます。??
  • 1
    Are "私には耐えられない" and "私にはやって行けない" meant as questions or statements?
    – Nate Glenn
    Mar 14, 2012 at 1:25
  • @NateGlenn They're negative statements... That's why I wasn't sure.
    – user1016
    Mar 14, 2012 at 16:29

Looking at the examples in the paper linked above, of Japanese sentences purportedly showing "dative subjects" (page 8 in the PDF), I note that the nouns marked by に in the Japanese are only subjects after translating into English.

For verbs in the potential, much as in passive constructions, the に in these sentences marks the agent, not the subject.


  • Passive: 「彼に英語が話される」 → "English is spoken by him."
  • Potential: 「彼に英語が話せる」 → "English is speakable by him."

In both sentences, the verbs are intransitive, 英語 is the subject of those intransitive verbs, and 彼 is marked as the agent of the actions.

Notably, intransitive verbs of potential that are derived from transitive verbs, like 話【はな】せる above, still semantically describe an action happening upon something, where that something is not the agent -- this is different from a "regular" intransitive verb, which semantically does not happen upon something, but instead describes an action occurring by means of that something itself.

More here in an older answer post, regarding the development and semantics of Japanese potential verbs.

In like fashion, the verbs わかる and できる mark the subjects with が and the agents with に. While not derived from transitive verbs, they function grammatically and semantically in a similar way -- they describe a quality of the subject, as something that is "do-able by" an agent marked with に.

The author of the linked paper Case Patterns is one Ellen Woolford, who appears to be this person. She specializes in syntax, but doesn't seem to have any particular familiarity with Japanese itself.

In the linked paper, she pulls the Japanese examples of purported "dative subjects" from Masayoshi Shibatani's 1977 work Grammatical Relations and Surface Cases, available here via JSTOR (free registration required). Shibatani goes in quite deeply, and his work is rife with abbreviations that make it quite difficult to read for non-specialists. That said, his analysis from page 799 (referenced by Woolford's paper) consistently describes subjects marked by に and objects marked by が, for sentences using verbs that Japanese-language references consistently describe as 自動詞 or "intransitive verbs" that take no objects -- verbs that only become "transitive" after translating into the common English glosses, verbs like intransitive ある translated as the transitive "to have".

Reading a bit further in Shibatani on pages 800-801, he makes a complicated argument about "subject"-ness based on how reflexive (using 「自分」) and honorific constructions focus on the noun marked with に. I posit that construing this as "subject-ness" is a mistake -- this indicates not that the に-marked nouns are subjects, but rather that the agents in such constructions have a higher primacy of focus for reflexives or honorifics than do the grammatical が-marked subjects of the intransitive verbs.

Re-casting this as "dative subject" and "nominative object" for verb constructions like ある is a strange stretch. Many languages use intransitive constructions not dissimilar to Japanese ある ("there exists") to express that someone "has" something. Compare:

  • JA: 彼には家がある → "by him, there is a house" → "he has a house"
  • HU: [nek]{に}[i]{彼} [ház]{家}[a]{POS} [van]{ある} → "by him/her, a [third-person singular possessor] house there is" → "s/he has a house"
  • NV: [kin]{家} [bee]{彼に} [hólǫ́]{ある} → "a house by him/her there is" → "s/he has a house"

In each case, the verb is simply a marker that something exists -- this is not transitive in any of these languages, and the statement only manifests any transitivity after translation into idiomatic English using the transitive verb "have". Claiming the existence of a "dative subject" and "nominative object" violates the principle of Occam's razor, and does much to muddy the waters.

  • 1
    The paper itself, as many other papers on this matter make compelling arguments as to why dative subjects, and nominative objects, behave more like subjects and objects respectively. It doesn't violate Ockam's razor because this complexity is needed to explain how Japanese grammatically functions. A common example is subject-honorrification. Honorrific forms of verbs can be used to raise the status of the subject, but this applies to the dative subject, not the nominative object in these kinds of verbs. There are many other examples such as how they interact with “自分” or “こと”.
    – Zorf
    Mar 13, 2023 at 4:57
  • @Zorf, it's like you completely missed my comment about agents. Mar 13, 2023 at 16:58
  • What language does NV stand for?
    – aguijonazo
    Apr 18 at 15:03
  • @aguijonazo: Navajo. Apologies for the aggressive abbreviation. See also the list of two-letter language codes at en.wikipedia.org/wiki/List_of_ISO_639_language_codes. Apr 18 at 15:42

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .