Floating Quantifiers, Partitives and Distributivity [chapter]

Jacob Hoeksema, Jacob Hoeksema
1996 Partitives  
Floating Quantifiers Commonly, the quantifier expressions of natural language are arguments of the verb, such as subjects, direct objects or objects of prepositions. In this respect, they differ crucially from the free-standing quantifier prefixes of first-order logic, as has often been observed by the proponents of generalized quantifier theory ( Barwise and Cooper (1981) ). This theory treats quantifiers such as every student or most teachers on a par with prototypical argument expressions
more » ... ment expressions such as proper names and pronouns (following Montague's (1973) innovation to employ 'lifted' types <,t> for proper names and other referring terms). At the same time, it must also be acknowledged that nominal arguments are not the only elements with quantificational force and in some languages not even the most frequent or central ones. Thus, there are adverbs of quantification, such as e.g. often in (1a) below (see Lewis (1975) and De Swart (1991) for discussion). Quantifiers may also be (pro-) nominal expressions used as modifiers, such as all in sentence (1b) and none of them in sentence (1c). (1) a. Australian terriers are often good mouse-catchers. b. We were all wounded at Wounded Knee. c. The deans were none of them fond of jeans. In sentence (1b), the quantifier all is neither in an argument position, nor is it selected by any expression within the sentence. Consequently it is entirely optional. Compare this with the use of all in (2) , where it is neither optional nor in a nonargument position. (2) a. All were wounded at Wounded Knee. b. He destroyed all to save himself. c. One for all and all for one! A quantifier expression used in the way all is used in (1b) is said to be a floating quantifier. The adjective floating is meant to indicate that its position is not necessarily fixed, but variable. Compare the three positional variants in (3). (3) a. We all should have been drinking tea. b. We should all have been drinking tea. c. We should have all been drinking tea. This variability has been the main topic of discussion in the smallish literature on floating quantifiers. How is it to be accounted for? Must we assume a movement transformation (and if so, of what kind?), or is it perhaps preferable to base-generate floating quantifiers? However, there are more questions that demand an answer, before a reasonable understanding is gained of the phenomenon of quantifier flotation. For example, what is the semantic status of floating quantifiers? What kinds of quantifiers float? And what are the syntactic category and the internal structure of a floating quantifier? It is a basic assumption of this paper that all these questions are related and ought to be answered together, in an integrated theory of the syntax and semantics of floating quantifiers. It is further assumed that it will be useful to
doi:10.1515/9783110908985.57 fatcat:amc3dv43uvdfff6yo5sap4vnle