3

What is the current accepted origin of the Japanese particles no/na/ni/ga?

One account I heard was that all were descended from a common root: an existential verb nu or ni, where ni was the infinitive/conjunctive form, and no/na/ga are various evolutions of the non-finite/adjectival or irrealis forms. Another account I heard was that ga and no were true particles (ga being a more concrete genitive relation and no being a more abstract genitive relation), while na is a shortened form of naru/nari (which might itself be a contraction of ni aru/ari). I wouldn't be surprised if there were other explanations.

Is there an accepted etymology for some or all of these particles?

|improve this question
5

Very few of the case markers you are asking about have any kind of an accepted etymology. Most of the time, it's because we just don't have any evidence to go on. Generally, they all share their functions and forms throughout the Japonic languages, so we have no real evidence to call them anything other than what they were, even going back to Proto-Japonic times.

But first, a detour. The "existential verb" you're talking about is not an existential verb at all, but a copula. In Vovin (2005, 2009)'s analysis of Old Japanese, the copula n- is a defective verb, only having three forms: an infinitive n-i |COP-INF|, an attributive (used to form relative clauses) n-ö |COP-ATTR|, and a subordinating converb form (used to link sentences) n-i-te |COP-INF-SUB| (Vovin 2009: 511-37). You can find reflexes of this not only in modern Japanese, but also in the related Ryukyuan languages, so the copula n- certainly goes back to Proto-Japonic.

As far as I know, =ga (PJ *ŋga) has no known etymology other than just a possessive case marker with a subject-marking usage.

It's conceivable that the genitive case marker =no (PJ *nə) and the attributive form of the copula no (also PJ *nə) are related (and even today most people continually confuse them--traditional Japanese approaches to grammar don't even recognize the copula n-).

na, which is used in modern Japanese to use certain nouns (also called adjectival nouns or na-adjectives) as attributives of some other noun is clearly a special attributive form of the copula. As far as I know, this is an innovation only found in Late Middle Japanese (so at the earliest around 1200 CE) (Frellesvig 2010: 341-2). Frellesvig seems to think LMJ na is from ealier naru, but he doesn't say whether this is from even earlier n(-i) ar-u |COP(-INF) exist-FIN| or from nar-u |become-FIN|.

The dative-locative case marker =ni is also of unclear etymological origin beyond being the Proto-Japonic dative-locative case marker *nəi (Vovin 2005: 148, reconstruction Whitman 1985: 36-8).

|improve this answer
  • 1
    Bonus question: is there a well-established reason to consider nari (copula) and naru (become) as having different origins? It seems to me that both meanings could easily be accounted for by a "ni [dative] ari/aru" etymology using different shades of the dative. – Justin Olbrantz Oct 10 '13 at 16:05
  • 2
    There is at least one very good reason. In Old Japanese, nar- 'to become' is a regular consonant stem verb. It's final (or 終止形 shūshikei, if you prefer) is nar-u |become-FIN|. The copula nar- from n-i ar- is an r-irregular verb, and its final form is nar-i |COP-FIN|. Only a limited group of verbs belong to this r-irregular class. The only non-derived members of this class, to the best of my knowledge, are ar- 'to exist', wor- 'to exist, stay', and por- 'to want'. – limetom Oct 10 '13 at 19:45
  • Vovin (2009) also gives some distributional evidence for nar- < n-i ar- |COP-INF exist-|, nar- 'become', and nar- < =ni ar- |=LOC exist-| being distinct. Following the attributive form (連体形) of verbs in OJ, the DAT-LOC case marker =ni takes on the meanings 'because' or 'when'. It doesn't follow other verb forms, and many of the uses we find of the various nar-*s just don't make sense if it were the locative. The only potential ambiguity is locative arguments, but it's certainly the copula there, as the DAT-LOC case marker can't take the converb *-te (cf. modern =de < n-i-te). – limetom Oct 10 '13 at 19:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.