Chapter 10: Position of indefinite article in the noun phrase
1. Indefinite articles ⇫ ¶
This chapter deals with the position of the indefinite article with respect to the noun in noun phrases. While there are more APiCS languages that have an indefinite article than a definite article (cf. Chapter 9), there is less variation: Almost all indefinite articles precede the noun:
| excl | shrd | all | ||
|---|---|---|---|---|
 | Indefinite article is preposed | 63 | 1 | 64 |
 | Indefinite article is postposed | 2 | 1 | 3 |
 | The language has no indefinite article | 10 | 0 | 10 |
| Representation: | 76 | |||
As in Chapter 29, we define an indefinite article as a morpheme that frequently occurs in noun phrases and signals that the referent is not uniquely identifiable by the hearer, as in We have a cat. In almost all languages where we have information about their origin, indefinite articles go back to the numeral ‘one’, and as a result, they are usually not used with plural noun phrases. As we see in Chapter 29, in many languages they are still synchronically identical to the numeral ‘one’.
2. Preposed indefinite articles ⇫ ¶
Indefinite articles are preposed (value 1) in the great majority of APiCS languages. Some examples are given in (1)-(4).
3. Postposed indefinite articles⇫¶
Postposed indefinite articles (value 2) are found only in two APiCS languages in Africa, Kinubi and Principense, and in Sri Lankan Malay.
Since word order tends to be very strongly influenced by the lexifier (e.g. Chapters 1-3), it is not surprising to find a postposed indefinite article in Arabic-lexified Kinubi, because in Arabic, the numeral ‘one’ is always postposed, like adjectives (e.g. ṭiflun waaħidun ‘one child’). Kinubi wáy derives from Arabic waaħid ‘one’.
The other language where the indefinite article always follows the noun is Principense:
Principense and Kinubi are two of those few APiCS languages where the numeral may (or must) follow the noun (see Chapter 6). Since the indefinite article derives from the numeral ‘one’, it is precisely in these languages that we expect postposed indefinite articles. (Most of the other languages with postposed numerals lack indefinite articles.)
As we saw in Chapter 6, the postposed order of numerals in the African APiCS languages goes back to the indigenous African languages, which almost always have postposed numerals. In fact, even the postposed ordering of the numeral ‘one’ and the indefinite article in Kinubi may have to do with the indigenous languages of South Sudan and Uganda (rather than with the Arabic lexifier, as suggested earlier), because Kinubi has all numerals following the noun (e.g. yal-á tísa [child-PL nine] ‘nine children’), not just the numerals for ‘one’ and ‘two’, as in Arabic.
The third language that has a postposed indefinite article is Sri Lankan Malay, but here both orders of the indefinite article attu (from Malay satu) are allowed and about equally common:
In Sri Lankan Malay, the order of the indefinite article is actually quite free, as described in detail by Nordhoff (2009: 440-444). We see in (7a) that the indefinite article follows the adjective, for instance, but the reverse order would also be possible. As argued by Nordhoff (2012: 28-31), the widespread (and sometimes obligatory) use of the indefinite article is due to Sinhala influence. Nordhoff cites the Sri Lankan Malay example in (8) and the parallel Sinhala example in (9). So it seems clear that the postposed ordering is also due to Sinhala influence.
4. Languages lacking indefinite articles ⇫ ¶
Ten languages lack in definite articles (value 3). These are mostly the languages that have lexifiers which are not Germanic or Romance: Eskimo Pidgin, Chinuk Wawa, and Chinese Pidgin Russian in the north, Bantu-based Lingala, Kikongo-Kituba, and Mixed Ma’a/Mbugu (as well as Juba Arabic) in Africa, and Ambon Malay, Yimas-Arafundi Pidgin, and Pidgin Hindustani in the Pacific region. Since indefinite articles are actually not all that easy to distinguish from the numeral ‘one’ (see Chapter 29), it could be that the difference is primarily due to the descriptive tradition: Germanic and Romance languages are usually described as having indefinite articles, while in other languages indefinite articles are much less expected.