Ronald W. Langacker (UCSD)
Friday, May 18th
Buchanan A201
The meaning and grammatical status of the indefinite article in core English quantifiers—notably a few, a little, a lot of—are examined from the standpoint of Cognitive Grammar. Arguments are given that these expressions are analyzable, that the initial element grounds the quantifying noun, and that it represents a semantic variant of the indefinite article appropriate for the meaning of that noun. Supporting its article status are the myriad quantifying expressions, exhibiting various degrees of grammaticization and grammatical flexibility, where a alternates with one, with other numerals, and with unquantified plurality, as exemplified in (1).
The a appearing in grammaticized quantifiers grounds the quantifying noun rather than the quantified noun: a few beers (but *a beers), a lot of verbiage (*a verbiage). Like possessives (that boy’s shoes), which provide an indirect means of definite grounding, complex quantifiers provide an indirect means of indefinite grounding. The distinction between grounding and quantification is not a sharp one. Whereas “relative” quantifiers (e.g. all, most, some, every, each) are grounding elements, “absolute” quantifiers (many, few, several, three, etc.) would seem not to be because they co-occur with definite grounding (e.g. those three women, the many houses I have owned). However, they also occur separately (many houses, three women), in which case they are comparable to relative quantifiers in the degree of “epistemic control” afforded; (2)(a) and (2)(b) are comparable in this respect. The functions of identification and quantification figure in both grounding elements and absolute quantifiers, but with different degrees of prominence. For instance, a and one alternate because both are indefinite and specify the same quantity; the choice depends on discourse factors, as in (3).
In its canonical use, the indefinite article singles out an instance of some type but specifies its non-identifiability given the prior discourse context. With complex quantifiers, identification of a particular instance is generally not at issue: when buying a bag of mulch we are usually not concerned with the identity of the bag. Nouns recruited for quantifying use come to designate units of measurement, invoking their original physical referents only to indicate magnitude. Quantity thus prevails over identification for the a in grammaticized quantifiers. In expressions like a cup of flour, the measurement noun designates a single positive step along a scale of measurement; it represents a type with multiple instances only in that the scale comprises multiple, interchangeable increments of this sort. Identification is thus irrelevant.
In the most highly grammaticized cases, quantification is non-metric, so a merely indicates positive value. Lot specifies that the quantity is greater than some norm. For little and few the quantity is less than a norm, so they function as negative polarity items, e.g. they sanction any, as in (4)(a). The complex forms a little and a few do not, as seen in (4)(b). The reason is that a imposes a positive construal, viewing the same quantities as upward departures from zero.
Examples
(1)(a) {a / one} {handful / gob / cup / ton / box} of N
(b) two {handfuls / gobs / cups / tons / boxes} of N
(c) {lots / handfuls / gobs / cups / tons / boxes} of N
(2)(a) Most students did well, but some had problems.
(b) Many students did well, but a few had problems.
(3)(a) A student complained.
(b) One student complained—the others didn’t.
(4)(a) Few linguists have any common sense.
(b) *A few linguists have any common sense.
References
Langacker, Ronald W. 1991. Foundations of Cognitive Grammar, vol. 2, Descriptive Application. Stanford: Stanford University Press.
Langacker, Ronald W. 2009. Investigations in Cognitive Grammar. Berlin and New York: Mouton de Gruyter. Cognitive Linguistics Research 42.
Langacker, Ronald W. 2010. A Lot of Quantifiers. In Sally Rice and John Newman (eds.), Empirical and Experimental Methods in Cognitive/Functional Research, 41-57. Stanford: CSLI.
Talmy, Leonard. 1988. The Relation of Grammar to Cognition. In Brygida Rudzka-Ostyn (ed.), Topics in Cognitive Linguistics, 165-205. Amsterdam and Philadelphia: John Benjamins. Current Issues in Linguistic Theory 50.
