Quantifier, Numeral or Determiner? The so-called indefinite article in French
Leszek Barczak
University of British Columbia
NOTE: This abstract contains special symbols coded using the Lucida Sans Unicode font. You may download the font from here or, alternatively, download the whole abstract in the pdf format.
For decades, the French words un/une/des (M.SG/F.SG/PL) were described as forms of the indefinite article. The apparent function of such indefinite article was (following Grevisse, 1964) to introduce a common noun phrase (NP), the referent of which was either not previously mentioned in the discourse or an element of a set of (at least two) individuals that could be described by such a common NP. In more recent analyses of nominal expressions, the traditional term article has been replaced with and included (together with demonstratives and possessives) into that of determiner (D). Thus, one may conclude that the French forms un/une/des are instantiations of D. Unfortunately, such conclusion is far from the truth when the semantics and syntactic behaviour the so-called indefinite article are taken into consideration.
Taking into account the meaning of un/une/des, as in (1), it seems that it has the interpretation of the existential quantifier (∃) similar to this of the English some (type: <<e,t>,<<e,t>,t>>. Another possibility is to analyse the French un/une/des as a determiner of type <<e,t>,e> as shown in (1b). In addition, if asked as to what is the French word for the numeral one as in one dog, two dogs, a French native speaker will cite either un or une depending on the gender of the noun s/he happens to think about first. Therefore, one may be prompted to analyse the singular forms of so-called indefinite article as the numeral one.
|
Un chien aboie dans la cour. |
|
un.M.SG dog.M.SG bark.PRES.3SG in the.F.SG courtyard.F.SG |
|
‘A dog barks/is barking in the courtyard.’ |
|
a. = ∃x [dog(x), x barks/is barking in the courtyard] |
|
[[un]] = λf ∈ D<e,t> . [λg ∈ D<e,t> . there exist some x ∈ De such that f(x) = 1 and g(x) = 1] |
|
b. = ∃x [dog(x) = 1] |
|
[[un]] = λf ∈ D<e,t> . there exist some x ∈ De such that f(x) = 1 |
The question that I address in my paper refers to these three potential analyses of un/une/des. I.e. in the subsequent sections of my work, I look at the (language internal) semantic and syntactic evidence in order to determine what the nature of un/une/des is – i.e. is it (a) a quantifier; (b) a determiner and/or (c) a (cardinal) numeral?
Based on: (i) the scope properties that such items exhibit; (ii) the influence of negation on their scope and form; (iii) restrictions on their co-occurrence with other quantifiers, determiners and numerals and (iv) the comparison of their meaning to the latter; as well as (v) the process of pronominalisation of un-Ps (known as en-replacement) which involves quantifier stranding, amongst others; I come to the conclusion that the primary (but not the unique) property of the French un/une/des is that of a quantifier. Finally, I postulate that un/une are either (a) used as numeral one due to the lack of such numeral in French cardinal numerals system or (b) real numeral homophonous with the singular forms of the quantifier un/une/des. Given the fact that French numerals are unspecified for gender and un/une is, the former analysis should be preferred.