Re: Splitting of a word or morpheme into several
- On Feb 2, 2011, at 9:38 AM, Alex Fink wrote:
> First off, in view of the amount you're flinging around morphemes, have youOh! Sadly, no -- I had bookmarked it a long time ago, but never gotten around to watching it. I must needs do that.
> seen this?
> On Wed, 2 Feb 2011 00:44:15 -0600, Eric Christopherson <rakko@...>
>> E.g.:True... I think my thinking on these is sort of muddy. It might be that they are originally predicate pronominals, e.g. "it is I", but evolve into plain pronoun forms.
>> original conjugated pronoun form:
>> /?atam/ "I, I am"
>> /ka:n/ "You, you are"
>> leading to:
>> /?at/ pronominal/copular form: "I, I am"
>> /am/ 1st-person singular agreement marker
>> /k/ pron./cop. form: "You, you are"
>> /a:n/ 2nd-p. sg. agreement marker
>> So, what does everyone think?
> I more or less agree with JV. I wouldn't expect this to happen straight
> off. From a language-internal perspective there's just zero reason to take
> a construction where pronouns can serve as copulae and break that down into
> two elements. Copulae aren't some sort of natural feature of semantics
> which languages yearn to have, they're just one way which a language might
> fall into expressing such notions.
(Actually one thing I'd like to do is make them specifically *ergative* pronouns, but I'm not sure about that yet. I may use an ergative preposition and conjugate it with the new personal endings, the way Semitic and Celtic languages do with various prepositions.)
>True. I was just skimming that Nuuchahnulth paper, and noticed suffixes that drop their initial consonants when they come after consonants. I'm unsure of how to formulate sound changes that would drop those affix-initial consonants across the board, though.
> I could see it happening if, again as JV has it, /-am/ and /-a:n/ occurred
> as suffixes with 1sg and 2sg value elsewhere. And that's not a hard thing
> to achieve. Grammaticalisation often drops a lot of material; it could
> easily (as in this example) drop all but the last rhyme of original pronouns
> when suffixal across the board.
(I generally have the problem that I'm not happy with just positing synchronic alternation without also somehow explaining how it got that way diachronically. But eventually it's an issue of having turtles all the way down...)
> If you like multipart verbs, you'll love Ket. Ket was once a suffixingYes, please; you may email it to this address.
> language, but thanks to areal influence has become largely prefixing. The
> way it went about this was by taking an old multipart verb construction,
> with usually a nominal incorporate (which is allowed to be an infinitive)
> sitting in the far left position, and reinterpreting that far left position
> as the root position and using only semantically bleached roots in the old
> root position on the right. I have a paper of Vajda's discussing this I can
> send, if you like.
(I'm trying to track down the Dene-Yeniseian issue of _Anthropological Papers of the University of Alaska_ for interlibrary loan right now.)
>John Vertical wrote:
> (But there is no splitting of unitary forms here.)
>> - Maybe if my system developed infixation, the infixes could come to be
> seen as non-infixal affixes, meaning the stuff on each side of them would
> cease to be seen as one morpheme.
> This also seems unlikely to me, just on the same grounds that there probably
> won't be any language-internal motivation to split the semantics in two so
> that one can be assigned to each piece of the form.
> Still, just /?atam/ being 1PS, etc. could probably motivate an /-am/ verbalYeah. With /?atam/ it might be pretty easy - I could just have the suffixal variant of it erode to /aDm/ or /a?m/ and thence to /am/.
> ending, etc, anyway - and from there you might first get a situation where
> the same ending is analyzed as a subject marker on the pronouns and stripped
> off of pronominal objects.
> This might imply extension of the personalYes! That is exactly one of the things that I have planned for this system.
> suffixes to nouns too: if /wip/ is "king", and /jatpu/ is "to command", then
> by the example of /?at-am jatpu-am/ "I command...", also /wip-am jatpu-am/
> "I, the king, command..." (Was it Sumerian that also did this?)