Determiners and Demonstratives
Determiners are post-nominal and appear as clitics or words rather than as suffixes. A reason for not treating them as suffixes is that when nouns are modified, the modifying constituent intervenes between the noun and the determiner. Futehrmore, the determiner agrees with the noun class. It also has word properties and is able to be positioned phrase-initially as well as undergoes morphophonological agreement changes, and encodes proximal-distal distinctions.
Determiners and Proximity
Determiners undergo a four-way proximity distinction, as the following table summarizes:
(I understand that final -e is optional for all proximal and medial ones. -a is not optional but mandatory for the distals).
The first column consists of a consonant (C), which is defined according to the noun class of the verb it is combining with, but the vowel and coda of that syllable expone the unique semantic features of the determiner, that is, whether it is a proximal, medial or distal, or not visible distal. The coda consonants of these clitics expone proximity as follows: -k- is proximal, -n- is distal and medial, and -g- is super-distal. In terms of vowel distinction, -ee- is associated with proximal and medial, and -aa- with distal in general.
The second column details the deictic distinctions for a class of demonstrative pronouns roughly translatable as ‘the ~ one(s)’. The singular is oxe and the plural owe (where -x- and -w- encode the number distinction throughout the pronominal and determiner paradigms). Some examples of this include the following, all of which translate as ‘the one who is bored’ or ‘the one who is not happy’:
oxe tayl na lit. one who is lazy
oxe sawareer na lit. one who is sad
oxe bugateer na lit. one who is not liking
oxe ɗayeer na lit. one who is not happy
Interestingly, an adjective can intervene between o and xe/we:
oxe jeg na --> o cegu xe ‘the rich one’
owe tayl na --> o taylu we ‘the lazy one’
This suggests that o- is really a pronominal form and can be modified by an adjective, while -xe and -we are its number-marking clitics for singular and plural, respectively. In the relativizing construciton with na, each of these includes an unsplit determiner pronoun oxe plus a bare form verb/adjective. The determiner pronoun can become demonstrative by attaching the appropriate deictic ending with -k, e.g., o cegu xeek.
Noun-Class-less Determiners
As suggested above, xe and we are default demonstratives when the noun is omitted and there is deictic reference. See the section on Noun Modification with na for more discussion on xe/we.
The following table details, for a few noun classes, how oxe/we work, and also summarized modification for those nouns with -u and na:
In the first column, it looks like oxe/we are pronominals that are modified by the adjectives faax 'good' and ranig 'white'. As I suggested in the discussion in the section above, oxe can possible be broken up into o-, a pronoun, and -xe, a determiner. Similarly for we, there is a null pronoun and an overt determiner. However, to remain agnostic as to the nominal status of oxe/we, I have suggested that the phrases may in fact be headless.
Demonstratives preserve the noun class of the noun they modify, even when the demonstrative is a fronted bare demonstrative. Take an example with pis ke ‘horses’:
Oana 21:34, 7 December 2012 (UTC)
Determiners and Wh-interrogatives
See Questions for the section on "which" interrogatives (used in WH-constituent questions), which agree for noun class with the head noun in their phrase.
