Most or the Art of Compositionality: Dutch de/het meeste at the Syntax-Semantics Interface
Koen Roelandt · Lirias · 2016
In this dissertation, I propose an analysis for the two interpretations of Dutch sentences with de meeste: (1) Jan heeft de meeste bergen beklommen The sentence in (1) can have the proportional interpretation where the number of mountains climbed by John is greater than the number of mountains he didn’t climb. The relative interpretation states that John climbed more mountains than other climbers (e.g. Bill, Mary or Paul). In English, these two interpretations are realised differently: the sentence with the proportional interpretation does not contain a definite determiner, but the sentence with the relative interpretation does (2). (2a) John climbed most mountains. (proportional) (2b) John climbed the most mountains. (relative) The central research question is how the two different interpretations can arise in Dutch if the definite determiner is present in both cases. I will claim that de meeste is very similar to regular superlative constructions, since these have similar interpretations. (3) Jan heeft de hoogste berg beklommen. Note that the sentence in (3) does not have a proportional interpretation: it can either mean that John climbed the highest mountain in the set (absolute interpretation) or that he climbed a higher mountain than Bill, Mary and Paul (relative interpretation). The difference between the absolute and the proportional interpretation is explained by the observation that de meeste ranges over individual sums of entities and de hoogste ranges over entities. The second research question pertains to the expression het meeste in Flemish Dutch. (4) Jan heeft het meeste bergen beklommen. This expression is puzzling for several reasons. Firstly, there is no agreement between the definite determiner het and the plural noun bergen; normally, het only appears with neuter mass and singular nouns. Secondly, sentences with het meeste can only have the relative interpretation. Using diagnostic tests for indefiniteness and specificity, I will demonstrate that het meeste behaves radically different than de meeste and regular superlatives such as de hoogste. Against the backdrop of de/het meeste I will also ask three questions. Firstly, I will consider how a semantic analysis of superlatives can be developed that follows the Containment Hypothesis in Bobaljik (2012): the syntax of the superlative contains the comparative and the adjective. In other words, if meeste ‘most’ contains the comparative meer ‘more’ and the adjective veel ‘many’, then what does its semantics look like? Secondly, I investigate to what extent this semantic account can be strictly compositional: can one assign a meaning to the different atoms without assuming null morphemes (semantic content that is not realised syntactically), semantically vacuous morphemes (syntactic elements that have no semantic content) or unmotivated movement? As it turns out, this is possible, but there are some problematic points in the analysis. The third question pertains to adjectives: to what extent can Aristotelian relations (contradictoriness, contrariety, subalternation and subcontrariety) be used to describe patterns in adjectival lexical fields? Using the algorithm of universe restriction described in Seuren and Jaspers (2014), I demonstrate that these relations are elementary to these lexical fields, but that a recursive application of the algorithm raises interesting questions.