Noun incorporation and applicative voice question

[nmnmv123]

Vowel harmony can be phonemic.

Do you know of any languages that have phonemic harmony?
I've never heard of it, but I can see how some triggering morpheme in a binary system (ex, marking past or present) would simply get deleted, since the harmony alone is enough to get the information (though this is kinda cutting the line between harmony and ablaut)
ex: (since my description is kind of confusing)
rounding harmony: i and y, i and u
-i past tense
-u present tense
to run = iuk

ran: iuki -> iik
running: iuku -> yuk
In either case, the tense marker just drops after applying harmony.

I have no idea why I just wrote up that example, but thats actually pretty interesting. Thanks for pointing that out! (bonus points if you post a natlang with phonemic harmony)

[Zaarin]

Vowel harmony can be phonemic.

Do you know of any languages that have phonemic harmony?

In Nez Perce, harmony is based on "dominance," with (I believe--this is from memory) /a i e/ being dominant over /e u o/. /e/ is not neutral in this system; the dominant /e/ comes from historical /ə/, which means you simply have to know whether a given /e/ is dominant or non-dominant for purpose of harmony. The harmony is both progressive and regressive: if there is a dominant vowel anywhere in the word then all vowels will be dominant.

[KathTheDragon]

NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

NO! NO! NO! NO! NO! NO! NO! NO! NO! NO! NO! NO!

[nmnmv123]

@soap & @nmnmv123: of course sound changes! I thought that went without saying. When I said reduced/abbreviated, I meant phonologically reduced. And if one looks at an athabaskan grammar, they sometimes show a grand chart of all the verbal slots, showing all sorts of interdependencies and variations, showing the results of diachronic changes (and of course other developments).

An important point that Methun makes about INs is this: You can tell that the noun is not simply juxtaposed on the verb complex if it has been subjected to phonological developments that can occur only within a morpheme; if internal changes are not reflected in the compound, then the noun is really just juxtaposed, and may show sound changes that are germane only to word margins. And of course this applies to all the components of an utterance, so one can see (if not erased by too many changes) that various components have undergone word-initial, word-medial, or word-final changes (and the whole gamut of other phonological developments), and where two components abutt, and there are word-medial changes, then you know that the speakers were treating the two as a single unit.

Are there a seperate set of sound changes that apply only to nouns before they become incorporated? I'm thinking of something where each noun has a corresponding incorporated form, created by a set of systemic changes to the original stem, where the changed form is then inserted into the corresponding noun slot, but from what little I know of realistic sound changes, they're always consistent. Is this an "exception" of sorts, where they only apply to create the incorporated stem?

Also just to address the second part of your comment, I am planning to have changes that occur word-initially that affect the incorporated stem (and all affixes) equally.

[Frislander]

However, I agree that in a stereotypical polysynthetic language, "toy" is not likely to be a single indivisible moirpheme, unless recent sound changes have compressed something that used to be composed of morphemes meaning something like "kids play with (it)" into a shorter word.

I think it's also likely that "determination" won't be a single morpheme either, and that you could solve your problem by taking just one piece of that word and using it as a verb instead of a noun. In English, I would analyze "You are filled with determination" as simply an emphatic form of "You are determined", and in my language I simply wouldnt bother to nominalize the word for "determination". Even the English word "determined" has one more morpheme than I would like to use, since there is no reason why a verb emphasizing a person's strong will should need to be marked with a passive participle marker like English -ed.

Good to know. So polysynthetics tend to function somewhat like oligiosynthetics in a way? As for why I treat it as a single word, primarily because my lexicon has literally 30 words, so its just kind of a place-holder for now so I dont need to make 5 new morphemes every time i want to translate something, though I guess it'll just end up impeding me later on in terms of realism, so I'll go through and re-define things when I have some time.

W

Well oligosynthetic is really just polysynthesis done to such a degree that it becomes ridiculous and nearly unworkable. You may get some parts of your lexicon which are full of single-morpheme words (like in Nuu-Chah-Nulth there are separate words for all of the different types of salmon that are found in their territory), but mostly you are talking sizeable amounts of derivation. There's also the issue of the Western perception of a "toy" as "something kids have fun with" as opposed to the more historical "miniature versions of invaluable tools and skills they will need as grown-ups."

I'd probably express them more like so: 1.SUB-teach-HAB, 1.OBJ-teach-NOM, 1.SUB-teach-ONG, and I'd not have the last one because you don't actually need action nominals.

Action nominals are deriving a verb from a noun, like the Inuit languages, right? If you don't have them, then how would you handle something like the last one - 1.SUB-teach, and just have context differentiate it from the verb? Also, a little side question - what is ONG in your gloss?

It's to other way round - noun from verb. I linked to the WALS chapter which talks about it, and it suggests one of the ways you can overcome this is to use a subordinate clause (e.g. something like "when I teach it is good") or equivalent construction. If you played around with this you'd probably find other interesting ways of expressing similar ideas.

ONG stands for Ongoing, which is kind of a synonym for imperfective aspect, though generally with a more present-tense interpretation, I find.

2s-toy-buy-PST-3s 1s-cat

Note that the noun "cat" has lost its case marking; that it has "gone".

So, the benefactive (or whatever adposition/other case it could be) just kind of gets implied by the verb already having an incorporated noun, and inferred by context?


getting into the polysynthetic minset is hard (>_<), but thanks for all the help guys!

[/quote]

You got it!

Slots are really the key to any inflecting verbal system.

True for PS and agglutinating languages, only marginally and inconsistently true for fusional languages. That's why they're fusional.

The closest thing to a counter-example in agglutinating languages is Eskimo-Aleut, where the post-bases and most of the inflection can occur in any order to get the meaning you want. However, even there the person inflection always occurs at the end of the verb complex, so there's your answer.

There must be some language somewhere which only uses stem changes for its inflection (it's probably Nilo-Saharan).

[Soap]

As for the sound changes, thanks to everyone that brought it up - another point I'll be adding. For the reduction, I'll be using a searchable version of the index diachronica, along with some syllable merging and (TBD) vowel changes (if any future readers need this for reference)

Thank you, I wasnt aware that that website even existed. I recognize some of the entries as having been put in by ZBB board members, and one of them is even from me. (/p/ > /w/ between vowels in Marathi).

I think this thread could be a good candidate for moving to C&C so that it will survive longer. I dont know what the pruning time is in this forum, but it looks like it's about 45 days judging by the timestamp of the oldest thread in the forum.

edit: either I misread the original forum or a moderator has done the move now.

[nmnmv123]

I think this thread could be a good candidate for moving to C&C so that it will survive longer. I dont know what the pruning time is in this forum, but it looks like it's about 45 days judging by the timestamp of the oldest thread in the forum.

edit: either I misread the original forum or a moderator has done the move now.

Yep, its been moved.

It's to other way round - noun from verb. I linked to the WALS chapter which talks about it, and it suggests one of the ways you can overcome this is to use a subordinate clause (e.g. something like "when I teach it is good") or equivalent construction. If you played around with this you'd probably find other interesting ways of expressing similar ideas.

ONG stands for Ongoing, which is kind of a synonym for imperfective aspect, though generally with a more present-tense interpretation, I find.

Thanks for the clarification - will do!