This paper aims to provide a clearer understanding of the widely discussed topic of grammaticalization of constructions. One unresolved problem concerns the identification of the relevant constructional level of language change, i.e., either the substantial level of specific occurrences or the schematic level of generalizations. Another issue addressed in many current analyses is the degree of schematicity of analogical forces.The present case study addresses these questions in the context of Spanish binominal quantifiers, basing its claims on synchronic and diachronic corpus analysis. The synchronic layering of uses in Present-Day Spanish—which encompasses quantifying and premodifying uses in addition to literal head uses—will be argued to reflect the outcome of grammaticalization. The synchronic as well as diachronic semantic clustering of binominal quantifiers and the prototype function of high-frequency montón de ‘heap of’ are hypothesized to result from the close interaction between analogical thinking and conceptual persistence. In line with the usage-based approach to language change, the dramatic increase in frequency observed for quantifying montón de will lead me to distinguish several constructional levels of change and to establish a schematic ‘N1 de N2’ pattern linked to quantity as a driving force. The strong collocational preferences characterizing the individual binominal constructions, the systematicity of the premodifying use and the tendency to reinforce conceptual elements of the original head use will be accounted for by considering the quantity assessment of the (schematic) binominal string as just one possible realization of co-extensiveness, in addition to quality assessment.

In the long-running and framework-independent discussion among historical linguists on the dominant mechanism in grammaticalization, viz. "reanalysis" or "analogy", recent case studies (

No abstract

In this paper we present a corpus-based, cognitive-linguistic account of the singular/plural agreement alternation with binominal constructions belonging to the [N1 singular + de + N2 plural ]-type. From a syntagmatic point of view, plural agreement reflects a shift from N1 as the head of the Noun Phrase towards N1 being a quantifier to N2. This view is in line with grammaticalization theory and holds for many examples, yet there are important counterexamples where we find a singular verb despite N1 being used as a quantifier. The Hispanic linguistic literature mentions several factors that would favor plural agreement with N2 such as an indefinite article, the distance between the verb and the subject, the position of the subject and pseudopartitive rather than partitive use. The corpus analysis will show that these factors are not conclusive. Our alternative account has recourse to the concept of lexical persistence and shows to what extent the quantifier use of N1 retains semantic features associated with the head use of N1. We argue that verbal agreement phenomena can be motivated by the interaction between lexical persistence and the morpho-syntactic and semantic criteria listed in literature. Lexical persistence is thus not a side effect of the grammaticalization process, but a crucial dimension of binominal constructions in that it allows us to explain quantifier uses of N1 with a singular verb.

No abstract