Syntax 1: Clausal Structure
So, before I explain it, let me put in a diagram that shows the whole clausal structure:
First, let me explain what each phrase and syntactic head are, and then describe them in Činlimli Nujmil:
CP: complementizer phrase
CPs are clauses, which can be sentences ("I went to the store"), or subordinate clauses ("for him to be happy"). In most syntactic analyses of natlangs, it is assumed that there are different C heads for declarative sentences vs. yes-no questions vs. wh-questions, root/matrix clauses (the main sentence) vs. embedded/subordinate clauses, and sometimes even more. Some are phonologically null, such as the complementizer in English declarative (root) sentences, such as "
Ø I went to the store", and sometimes a phonologically pronounced word, such as 'that' is "I know [
that he will be late today]".
Činlimli Nujmil has phonologically null matrix/root clause complementizers, except for yes/no questions, which acquired a new overt C after the last set of sound changes some of the polar question cues that marked the tense + agreement were obscured.
Code: Select all
/Ø/ root clause declarative
/Ø/ root clause wh-question [moves wh-items]
/luuš/ root/embed. clause yes/no question
/kal/ embed. clause declarative
/kaf/ embed. clause wh-question [moves wh-items]
Here are some examples:
1) [
Ø mąą neŋįį pus’nęę.]
"I am eating bread."
2) Loo [
kal mąą neŋįį pus’nęę] nuu’nas.
"They know
that I am eating bread."
(1) & (2) show the declarative complementizer in root and embedded clauses. Now for demonstration of wh-movement
3) lenįį leekųų
Ø lels’naa?
what-acc who-nom
C do.impfv=npast.3s
“Who is doing what?”
4) Loo [lenįį leekųų
kaf lels’naa] kuulečif’šis
3p.nom [what-acc who-nom
C do.impfv=npast.3s] ask.pfv=past.3p
“They asked
who was doing what.”
PolP: polarity phrase
This is where some languages such as Irish and Russian house negation. It also sometimes is realized as
contrastive positives (like "I
DID do it, contrary to your portrayal...")
In Činlimli Nujmil, Pol plays an important part. Negation is housed and realized only in Pol, along with irrealis mood (the subjunctive, and conditional). It's specifier position (one slot to the left of the Pol word) is reserved for a
contrastive item. For example:
5)
Neŋįį Ø mąą pus’nęę.
bread-acc Pol 3s.nom eat.impfv=npast.1s
"I am eating
some bread. (not lobster or anything)"
Here is a list of the Pol heads:
Code: Select all
/Ø/ positive
/Ø/ positive contrastive [moves contrastive item]
/nas/ negative
/nas/ negative contrastive [moves contrastive item]
/nužla/ positive irrealis [allows subjunctive or conditional]
/nazla/ negative irrealis [allows subjunctive or conditional]
Here are some positive and negative Pol examples:
6)
a) Luuš
Ø puvlįį čenf’šitųų?
Q Pol face-def.sg.acc see.pfv=past.2s.Q
"Did you see his face?"
b)
Nas čis čenf’šįį.
neg 3s.acc see.pfv=past.1s
"No. I didn’t."
(6a) has asks a positive-valency question via the positive Pol /Ø/. The answer in (6b) is a negative-valency statement, with the negative Pol /nas/. Below are examples of the irrealis Pol with a subjunctive in (7), and a conditional in (8).
7)
Nužla tąą molnootulii’na kuŋf'
šįį.
irrealis 3s.nom home-def.sg.obl=dat go.pfv=
subj.3s
“S/He should go home.”
8) [
Nužla nal kutas’
tuu] mąą kutas’nęę
[
irrealis 2s.nom happy=
cond.2s] 1s.nom happy=npast.1s
“If you are happy, I’ll be happy”
TP: tense phrase
TPs, or IPs (inflectional phrase), have been a big turn in syntax. It is a strong case of assuming the surface structure of a sentence can be different from the underlying framework that builds it up. What does that mean? Well, it is a claim that
Tense, although mostly marked on the verb (in English and some other languages), is not something that is part of the verb. The discussion for it is long and different for different languages, but it seems that there are languages where tense (T) is realized independent from the verb, and weird behaviors (like English
do-support under negation) that suggest tense is syntactically higher than the verb, and just affects the marking of the verb. Long story short, tense is separate from the verb.
Činlimli Nujmil doesn't mark tense on the verb. Instead, tense is realized as a separate word, with subject agreement on it. At least, that used to be the case in Middle Činlimli Nujmil (the one before the last batch of sound changes). Now, Tense + agreement is phonologically clitic to the verb, showing
Obstruent Lenition as a sign of phonological cliticization (The initial obstruent lenides when cliticizing to the left, ex: noun=ta --> noun=sa, ex: verb=čįį --> verb=šįį)
Tense shows agreement with the subject in person and number. The tenses are
past, non-past, subjunctive, & conditional. A polar question morpheme used to be attached into this complex in Old Činlimli Nujmil, but sound change has conglomerated all
tense + person + number + polar question together. Here are the forms of tense:
Code: Select all
PAST Y/N Q
*ti- SG PL SG PL
1 čįį činis činįį čiitǫǫ
2 čis čitus čitųų čistǫǫ
3 čii čis čįį čitǫǫ
Code: Select all
PRES Y/N Q
*na- SG PL SG PL
1 nęę nenis nenįį neetǫǫ
2 nos notus notųų nostǫǫ
3 naa nas nǫǫ natǫǫ
Code: Select all
SUBJ Y/N Q
*tinu- SG PL SG PL
1 činųų čiinis čiinįį čiinitǫǫ
2 činus čiitus čiitųų čiitutǫǫ
3 čįį činus činųų čiitǫǫ
Code: Select all
COND
*Ø- SG PL
1 nii nis
2 tuu tus
3 Ø taa
The left two columns are the declarative tense forms. In polar questions, the right two columns starting with "Y/N Q" are used. These forms are different, but very similar, because the only difference was the incorporation in Old Č.N. of /*-ŋu/.
AspP: aspect phrase
This is just a syntactic position for aspect (perfective vs. imperfective) which is realized on the verb.
For example, the verb for 'eat' has two forms; imperfective /pus/, and perfective /putaf/. In this case Asp either houses the impfv /-Ø/ or the pfv /-f/.
vP: voice phrase
I will not go much into what
v is, because it is quite complicated and I don't quite understand all the reasons we want or don't want a separate functional head
v in syntactic theory. I think the take-away is that the verb itself is housed in VP, while the argument structure (whether there is a subject and an indirect object, whether it is passive or active) is housed in
v. I will just use it in diagrams, and refer to vP instead of VP, but as of yet,
v does not have any special thing in Činlimli Nujmil. I justed wanted to say that I am going to use it, but in the end, it is not important, pay it no more heed than this paragraph does.
VP & arguments
VP is the good old verb phrase, which consists of the
verb, and the
object if the verb is transitive. As you can see from the diagram, the order of the arguments, as well the hierarchical positions is: [ indirect-object [ subject [ object verb ] ] ]
This is the canonical order. But, as you have read a little bit above in the PolP section, one of these arguments (or even an adverb or something) can move to the left of Pol to get contrastive focus. It doesn't matter which, any of these arguments, and also adjuncts, adverbs, etc can get contrastive focus and move. But
only one element. There is only one position for contrastive items, so only one can move to the left of Pol.
Adjuncts: adverbs, time, manner, "because ...", "when ..."
Adjuncts are optional things that loosely do not change the core proposition of something, but add more meaning, such as adverbs. Syntactically parenthetical clauses such as "
If he would open the door, he would see." or "
When I try to turn it on, I first look at the safety switch." are also adjuncts. An adjunct in syntax is loosely something you can attach to certain points in the syntactic tree, without affecting all the stuff in the tree. They are optional, both "Yesterday I ate pizza" and "I ate pizza" essentially express that the consumption of pizza is complete, and the agent of the consumption was me.
So, in Č.N. manner adverbs such as "quickly", "fast", "slow", & "well" attach at the AspP level, with the linear order as:
[ wh-item C contr-item Pol
manner-adv [ indir-obj subj obj verb-Asp ]
AspP T=agreement ]
Temporal adverbs, more complex adjuncts such as "when ..." or "in order to ..." etc, conditionals, all attach at TP level (which is indistinguishable from AspP-adjunction for now, maybe I'll add some other stuff that will make is possible to tell apart in some situations), or at CP level.
[
CP-adjunct [ wh-item C contr-item Pol
TP-adjunct [ indir-obj subj obj verb-Asp T=agreement ]
TP ]
CP ]
Scope
This language has surface scope. So, quantifiers, indefinites, negation, and all that jazz scope based of surface positions. So, if a quantifier such as 'every' scopes lover than negation as long as the DP it doesn't move to the left of (spec PolP to be exact) the negation at Pol.
Here are some examples with
negation and
everybody. Think about the English equivalent with this sentence, "Everybody didn't go". For the English sentence, there are two interpretations: the first is "It is not the case that everyone went", just a logical negation of
all persons involved going, the second one is "For each person in question, that person did not go." This is because of English allowing different
scopes for quantifiers like
every, some, etc.
In Č.N. the surface position determines the scope of quantifiers, negation, and other semantic functions. The first example is the canonical word-order, without movement to the left of Pol:
9) Nas
moskųų činaa’na kuŋf’naa.
neg everyone road.indef.obl=dat go.pfv=npast.3s
¬ [ ∀x [ go(x) ] ]
“It is not the case that everybody went.” [neg >> every]
The negation in (9) scopes over everything else, i.e. negates the truth value of the statement "Everyone went." So, half of the people might have left, or everyone could have stayed, or everybody except Jeff could have went along. This is the only interpretation for this sentence.
Now, if
everyone had moved to the specifier of Pol, then the interpretation would be different:
10)
Moskųų nas činaa’na kuŋf’naa.
everyone neg road.indef.obl=dat go.pfv=npast.3s
∀x [ ¬ [ go(x) ] ]
“Nobody left.” [every >> neg]
In (10), the interpretation is that "Nobody left." Because
everyone is higher up on the syntactic tree, it serves as a "for all people, the following is true" kind of meaning.
So, scoping interacts in a really cool way with the clausal structure. I have to say, I'm feeling really proud of this scope interaction, so thank you Irish for having scope interactions with negation on Pol. You're the language!
Coming soon is probably a lexicon, followed soon by the currently non-existent derivational morphology.