Since this question relates to reflexives, I thought I'd just add it to this thread.
I have seen an example, I think from Hindi, where using an ergative subject vs. nom/abs gives a reflexive interpretation. It's something like:
man.ABS die.PERFECTIVE = the man died.
man.ERG die.PERFECTIVE = the man died (purposefully, as in commited suicide/killed himself).
After some searching, I haven't been able to relocate a source that is complete w/ Hindi words. I think the above will suffice.
Is this reflexive interpretation common w/ ergative subjects in intransitive sentences, or is this specific to the verb "to die"? Hindi also has reflexive pronouns. If the reflexive interpretation of the ergative subject is common, what's the difference between that vs. using a transitive verb and a reflexive pronoun object? Is it just that an ergative subject w/ an intransitive verb can for a few (many?) verbs give a reflexive meaning in certain contexts, and a construction with an reflexive pronoun can be used to disambiguate?
reflexive verb marking (current: ergative subjects in Hindi)
Re: reflexive verb marking origins
Tibetan Dwarvish - My own ergative "dwarf-lang"
Quasi-Khuzdul - An expansion of J.R.R. Tolkien's Dwarvish language from The Lord of the Rings
Quasi-Khuzdul - An expansion of J.R.R. Tolkien's Dwarvish language from The Lord of the Rings
Re: reflexive verb marking origins
It's actually a little more complicated than this. There are three main (fossilized) markers of transitivity in Japanese verbs.Terra wrote:Reflexives are a kind of intranstive verb. Japanese has many fossilized patterns for making transitive and intransitive pairs. There's 2 main kinds: vowel change and adding -su.
Examples:
http://en.wikibooks.org/wiki/Japanese/G ... ansitivity
http://www.csse.monash.edu.au/~jwb/ti_list.html
Maybe the vowel change comes from some kind of suffix that umlauted the preceding vowel, and maybe the -su comes from 'suru' ('do'). I'm not sure about either of these though.
First, the two most obvious:
-su: transitive marker: as you say, it is just the verb su "do" (modern suru). It is generally always added to a stem ending in -a.
-aru: intransitive marker: this is the verb "be" or "exist"
-e: This suffix swaps the transitivity.
Now, what you said was a vowel change that indicates a change in transitivity is actually two separate functions that look similar on the surface but are actually unrelated. All Old Japanese shimo-nidan verbs have a fossilized suffix -e(r) in their modern forms, while kami-nidan verbs have -i(r) (e.g. OJ 寝 nu vs. MJ 寝る neru "sleep", OJ 食ぶ tabu vs. MJ 食べる taberu "eat" OJ 生く iku vs. MJ 生きる ikiru "live"). Adding any of the derivational suffixes above ensures that the resulting verb will not be shimo- or kami-nidan, which means that the addition of -e(r)/-i(r) can only happen on the form without the suffix.
Let's look at some examples to clarify:
付く tsuku "stick; attach" (intransitive)
付ける tsukeru "stick; attach" (transitive)
This one takes the -e suffix to reverse transitivity and make a transitive verb. The unmodified stem is yodan class, so it does not end up with an -e(r) or -i(r) suffix.
焼く yaku "cook; burn" (transitive)
焼ける yakeru "cook; burn" (transitive)
The same thing happens here, except the root is transitive and the derived form with -e is intransitive. Yaku is also yodan, like tsuku.
Here's where it gets more confusing:
上げる ageru "raise" (transitive)
上がる agaru "rise" (intransitive)
Old Japanese 上ぐ agu belonged to the shimo-nidan class, so its modern Japanese form is ageru.
Intransitive 上がる agaru is formed with the addition of the -aru intransitive suffix
満ちる michiru "become full" (intransitive)
満たす mitasu "fill up" (transitive)
Old Japanese 満つ mitsu belonged to the kami-nidan class, so its modern Japanese form is michiru
Transitive 満たす is formed with the transitive suffix -su.
There are also allophonic changes happening with the stem-final consonant /t/.
Typically, stems do not take both -aru and -su. I can't think of any pair that does, so I cannot provide any examples. I guess there are none, or possibly one or two weird exceptions that I can't think of.
There are some pairs that end in -r, but drop this before the suffix -su.
壊れる kowareru "break" (intransitive)
壊す kowasu "break" (transitive)
隠れる kakureru "hide" (intransitive)
隠す kakusu "hide" (transitive)
The sound change that dropped /w/ (former /h/) before every vowel but /a/ causes this process to be a little obscured:
終える oeru "finish" (transitive)
終わる owaru "finish" (intransitive)
(original stem was OH, so ohu -> oheru -> oeru vs. oharu)
There is also a small subset of verbs that have -i in the intransitive and -o in the transitive. I haven't figured out the cause of this yet.
落ちる ochiru "fall" (intransitive)
落とす otosu "drop" (transitive)
降りる oriru "go down; get off" (intransitive)
下ろす orosu "take down; unload; withdraw" (transitive)
起きる okiru "get up" (intransitive)
起こす okosu "awaken; cause" (transitive)
A little off topic, but I hope this clears it up.
EDIT: The page you linked to seems to support what I said about -aru and -su never appearing on the same root.
Also, it pointed out the weird set that have -seru. Maybe -su somehow becomes kami-nidan only for these few verbs. Weird. I forgot about these ones.
Re: reflexive verb marking (current: ergative subjects in Hi
That's elucidating. Thank you.Now, what you said was a vowel change that indicates a change in transitivity is actually two separate functions that look similar on the surface but are actually unrelated. All Old Japanese shimo-nidan verbs have a fossilized suffix -e(r) in their modern forms, while kami-nidan verbs have -i(r) (e.g. OJ 寝 nu vs. MJ 寝る neru "sleep", OJ 食ぶ tabu vs. MJ 食べる taberu "eat" OJ 生く iku vs. MJ 生きる ikiru "live"). Adding any of the derivational suffixes above ensures that the resulting verb will not be shimo- or kami-nidan, which means that the addition of -e(r)/-i(r) can only happen on the form without the suffix.