Page 1 of 5

Nifty Random Features

Posted: 05 Apr 2011 03:41
by roninbodhisattva
I thought it might be fun to have a thread where people can post nifty random features/parts of languages that they're reading about. Here's one:

...

Actually I don't have one right now. Shit. But anyway, might serve as a good place for inspiration.

Re: Nifty Random Features

Posted: 23 Apr 2011 19:32
by Sodomor
Muskogean language have interesting ways of deriving verbs. The coolest is the use of disfixes to derive pluractionality.
noktiłifka "choke", noktiłiika "choke repeatedly"
batatli "hits", batli "hits repeatedly"

Re: Nifty Random Features

Posted: 23 Apr 2011 21:43
by eldin raigmore
switch-reference markers
same-subject and different-subject markers
clause-chaining

Re: Nifty Random Features

Posted: 24 Apr 2011 03:54
by Micamo
eldin raigmore wrote:switch-reference markers
same-subject and different-subject markers
Aren't these the same thing?

Anyway... everything about Navajo. I took a look at the wikipedia page on its grammar, and holy *fuck.* I was totally hit by the Anadew stick. And I thought my Midhera aspect system was crazy... I think I'm in love.

Re: Nifty Random Features

Posted: 24 Apr 2011 04:07
by roninbodhisattva
Micamo wrote:Anyway... everything about Navajo. I took a look at the wikipedia page on its grammar, and holy *fuck.* I was totally hit by the Anadew stick. And I thought my Midhera aspect system was crazy... I think I'm in love.
I have basically the same reaction to every Athabaskan language. They're amazing.

Re: Nifty Random Features

Posted: 29 Apr 2011 23:26
by eldin raigmore
Micamo wrote:
eldin raigmore wrote:switch-reference markers
same-subject and different-subject markers
Aren't these the same thing?
Not quite.

If you get a report and a collection of papers presented from a conference on switch-reference, you'll see a lot of talk about same-subject markers and different-subject markers.

But there are markers called "switch-reference markers" that;
* have different values than "same" and "different", (for instance, "properly contains", or, "is properly contained in") (or, for places and times, "is contiguous to")
and/or
* are about other referents than "subject"s, (for instance, "same or different objects", "subject of marked clause same as or different than object of referred-to clause", "object of marked clause same as or different from subject of referred-to clause") (or, "same or different place", or "same or different time").

For instance, my Adpihi switch-reference system is not far-fetched.
There are two switch-reference markers for each Consecutive (that is, non-Initial) verb.
Each can take any one of five values.

(For the following discussions, think "intransitive subject" for "Subject"; think "agent" for "Actor"; think "direct object" for "Undergoer"; and think "indirect object" for "Extended-core term".)

The "same or different" "subject" marker's values have the following meanings.
(1). The marked clause's Subject or Actor is identically the same as the referred-to clause's Subject or Actor.
(2). Not (1), but the marked clause's Subject or Actor is identically the same as the referred-to clause's Undergoer or Extended-core term.
(3). Neither (1) nor (2), but the marked clause's Subject or Actor either properly contains, or is properly contained in, the referred-to clause's Subject or Actor.
(4). Not (1) and not (2) and not (3), but the marked clause's Subject or Actor either properly contains, or is properly contained in, the referred-to clause's Undergoer or Extended-core term.
(5). None of the above.

The "same or different" "object" marker's values have the following meanings.
(1). The marked clause's Subject or Undergoer is identically the same as the referred-to clause's Subject or Undergoer.
(2). Not (1), but the marked clause's Subject or Undergoer is identically the same as the referred-to clause's Actor or Extended-core term.
(3). Neither (1) nor (2), but the marked clause's Subject or Undergoer either properly contains, or is properly contained in, the referred-to clause's Subject or Undergoer.
(4). Not (1) and not (2) and not (3), but the marked clause's Subject or Undergoer either properly contains, or is properly contained in, the referred-to clause's Actor or Extended-core term.
(5). None of the above.

What exactly a given value of one of these switch-reference markers means in any given use, depends on the types of the marked clause and the referred-to clause.
A clause may have just a Subject; or it may have just an Actor and an Undergoer; or it may have a Subject and an Extended-core term (basically a "bivalent intransitive clause", sort of a clause with an indirect object but no direct object); or it may have an Actor and an Undergoer and an Extended-core term (the clause may be a ditransitive clause).

Values 2 and 4 could be ambiguous if the referred-to clause is ditransitive, because, for instance, the switch-subject marker is assigned the 2 value whether the marked clause's Subject or Actor is the same as the referred-to clause's Undergoer or to its Extended-core term. I think it will usually be clear to the addressee which is meant; if not, s/he can ask the speaker to disambiguate.

Values 3 and 4 are ambiguous as well. For instance, the switch-subject marker is assigned the value 3 whether the marked clause's Subject or Actor properly contains the referred-to clause's Subject or Actor, or the other way around. If one of them has a smaller grammatical number than the other -- for instance, one is singular and the other is plural -- the smaller one is properly contained in the larger one. If one of them has the Common, Epicene, or Mixed gender and the other doesn't, then the Common one contains the other one. Otherwise, the addressee can probably disambiguate; or can ask the speaker to.


Suppose for a moment that the Initial, "anchor" clause, and the Consecutive clause, both are two-participant transitive clauses with an Agent and a Patient.
Suppose Consecutive clause is switch-reference-marked referring to the Initial, "anchor" clause.

Let's call the referred-to clause's agent "rA" and its patient "rP"; let's call the marked clause's agent "mA" and its patient "mP".
Spoiler:
If mA=rA and mP=rP, then the values will be 1,1.
If mA=rA and mP<rP, then the values will be 1,3.
If mA=rA and mP>rP, then the values will be 1,3.
If mA=rA and mP is disjoint from rP, then the values will be 1,5.
If mA=rA and mP overlaps rP but neither contains the other, then the values will be 1,5.

If mA=rA and mP=rP, then the values will be 1,1.
If mA<rA and mP=rP, then the values will be 3,1.
If mA>rA and mP=rP, then the values will be 3,1.
If mA is disjoint from rA and mP=rP, then the values will be 5,1.
If mA overlaps rA but neither contains the other, and mP=rP, then the values will be 5,1.

If mA=rP and mP=rA, then the values will be 2,2.
If mA=rP and mP<rA, then the values will be 2,4.
If mA=rP and mP>rA, then the values will be 2,4.
If mA=rP and mP is disjoint from rA, then the values will be 2,5.
If mA=rP and mP overlaps rA but neither contains the other, then the values will be 2,5.

If mA=rP and mP=rA, then the valuses will be 2,2.
If mA<rP and mP<rA, then the valuses will be 4,2.
If mA>rP and mP>rA, then the valuses will be 4,2.
If mA is disjoint from rP and mP is disjoint from rA, then the values will be 5,2.
If mA overlaps rP but neither contains the other, and mP=rA, then the values will be 5,2.

If mA<rA and mP=rP, then the values will be 3,1.
If mA>rA and mP=rP, then the values will be 3,1.
If mA<rA and mP<rP, then the values will be 3,3.
If mA<rA and mP>rP, then the values will be 3,3.
If mA>rA and mP<rP, then the values will be 3,3.
If mA>rA and mP>rP, then the values will be 3,3.
If mA<rA and mP is disjoint from rP, then the values will be 3,5.
If mA<rA and mP overlaps rP but neither contains the other, then the values will be 3,5.
If mA>rA and mP is disjoint from rP, then the values will be 3,5.
If mA>rA and mP overlaps rP but neither contains the other, then the values will be 3,5.
And so on.

Re: Nifty Random Features

Posted: 30 Apr 2011 05:06
by Micamo
o.o

Eldin, I love you.

But one question about all of this: You say switch-reference works with non-initial verbs. How plausible would it be, natlang wise, for switch-reference to occur in a rigidly verb-initial language, where it's not always syntactically possible for verbs sharing referents to appear next to each other?

Re: Nifty Random Features

Posted: 30 Apr 2011 05:39
by cybrxkhan
Apparently some language around Nepal called Limbu has a case called the "Intrative Case", which means "amidst" or "in between" the noun.


There was also some Dravidian language (I think?) that grammaticalized quoting someone. And then I also remember reading on Wikipedia a long, long time ago that there was a grammatical feature in another Dravidian language that functioned exactly like the em dash, although I really don't know if my memory just botched that up.

Re: Nifty Random Features

Posted: 30 Apr 2011 05:40
by roninbodhisattva
cybrxkhan wrote:There was also some Dravidian language (I think?) that grammaticalized quoting someone.
What do you mean by this?

Re: Nifty Random Features

Posted: 30 Apr 2011 05:42
by cybrxkhan
roninbodhisattva wrote:
cybrxkhan wrote:There was also some Dravidian language (I think?) that grammaticalized quoting someone.
What do you mean by this?
I just found it on Wikipedia, although it doesn't cite any references (dammit), although personally I think it sounds plausible: http://en.wikipedia.org/wiki/Quotative

Re: Nifty Random Features

Posted: 30 Apr 2011 05:42
by Micamo
You mean quotative vs. reportative distinction?

Re: Nifty Random Features

Posted: 30 Apr 2011 05:43
by cybrxkhan
Micamo wrote:You mean quotative vs. reportative distinction?
No, it doesn't seem to indicate evidentiality per se, although I dunno. Can't trust Wikipedia on everything, I guess. :roll:

Re: Nifty Random Features

Posted: 30 Apr 2011 23:12
by eldin raigmore
Micamo wrote:o.o

Eldin, I love you.
Aww, shucks, ma'am! :oops: :-) <bashful grin>
Micamo wrote:But one question about all of this: You say switch-reference works with non-initial verbs. How plausible would it be, natlang wise, for switch-reference to occur in a rigidly verb-initial language, where it's not always syntactically possible for verbs sharing referents to appear next to each other?
That's not what those terms mean in this use.

In many languages with lots of clause-chaining, there's no such thing as a subordinate clause, and so no such mood as a subjunctive mood.

But in many of them there is a "conjunctive mood" (I don't think I made up that term, but I can't find who did in the time I have now). All of the verbs of clauses in any clause-chain except the anchor clause, are in this "conjunctive mood".

If the anchor clause of each chain is the final clause in the chain, the anchor is said to be a "Final Clause" and all the other clauses in the chain are said to be "Medial Clauses". In such languages, each marked clause's referred-to clause comes after it, sometimes being the next clause in the chain and sometimes being the anchor (Final) clause. (Some languages make it the next clause, some make it the Final clause, some allow either, and some allow any later clause in the chain.)

If the anchor clause of each chain is the first clause in the chain, the anchor is said to be an "Initial Clause" and all the other clauses in the chain are said to be "Consecutive Clauses". In such languages, each marked clause's referred-to clause comes before it, sometimes being the preceding clause in the chain and sometimes being the anchor (Initial) clause. (Some languages make it the preceding clause, some make it the Initial clause, some allow either, and some allow any earlier clause in the chain.)

In any cause, clauses in this "conjunctive" mood are dependent. For instance, their verbs have relative tense (aka "anteriority/simultaneity/posteriority") rather than absolute tense (aka "past/present/future").

___________________________

(These languages' modality/mode/mood systems might not just be "Initial vs Consecutive" or "Final vs Medial"; there may be, for instance, several "conjunctive moods" instead of just one. Also, aspect and/or m/m/m of a Consecutive or Medial clause may depend on that of the Anchor clause -- to some degree. Or, FAIK, not.)

______________________________________________________________________________

EDIT: that's not what I meant by those terms in this use.

Re: Nifty Random Features

Posted: 01 May 2011 12:26
by teh_Foxx0rz
eldin raigmore wrote:But in many of them there is a "conjunctive mood" (I don't think I made up that term, but I can't find who did in the time I have now). All of the verbs of clauses in any clause-chain except the anchor clause, are in this "conjunctive mood".
Japanese is like this with it's "-te" form. I'm not sure it's called a "mood", but it seems to work like how you're describing it here.

Re: Nifty Random Features

Posted: 01 May 2011 17:22
by Micamo

Re: Nifty Random Features

Posted: 01 May 2011 17:26
by Testyal
Ittamyminecraftwoodenhouse!

Re: Nifty Random Features

Posted: 01 May 2011 19:14
by cybrxkhan
lolwut

Another Native American weirdness...

Re: Nifty Random Features

Posted: 01 May 2011 20:31
by roninbodhisattva
What do you find so surprising / exciting about this?

Re: Nifty Random Features

Posted: 01 May 2011 20:40
by Micamo
roninbodhisattva wrote:What do you find so surprising / exciting about this?
Mostly because it's ANADEW. I thought the idea of locative marking on the verb was original...

Re: Nifty Random Features

Posted: 01 May 2011 20:50
by roninbodhisattva
Micamo wrote:
roninbodhisattva wrote:What do you find so surprising / exciting about this?
Mostly because it's ANADEW. I thought the idea of locative marking on the verb was original...
Not at all.