No abstract

Much literature has explored the interpretation of the bare singular (BS) in Brazilian Portuguese. Pires de Oliveira & Rothstein (2011) claim that BS nouns are mass because they denote kinds and argue that this explains why only the BS in Brazilian Portuguese can have a non-cardinal interpretation. In this paper, based on an experimental task with Brazilian Portuguese adult speakers, we explore one of their predictions, namely that the 'volume interpretation' of the BS cannot be explained as a case of Grinding. Our results show that Grinding and Volume readings of a BS noun are not equivalent (in favor of their hypothesis). We also show that a volume interpretation of a noun is never preferred when a cardinal interpretation is available, but that this can be explained by other lexical and pragmatic factors. We conclude by suggesting that Rothstein's (in press) distinction between counting and measuring accounts for the fact that non-cardinal readings are not grinding.

The chapter reports the results of applying quantity judgment tests to bare nouns in Brazilian Portuguese. The results show that the bare singular and sentences with no nouns (no cues to atomicity, as argued in Scontras et al. 2017) behave alike, in contrast with plural nouns. Brazilian bare singulars allow for cardinal and volume readings. We show that this is explained by the fact that they carry no linguistic cues to atomicity. Finally, we argue that the difference between English and Brazilian Portuguese may be explained by Chierchia’s (2010, 2014) nominal parameter: English is a predicate language, whereas Brazilian Portuguese is a kind language. The conclusion explores some consequences of this semantic parameter.

No abstract

Muito, no Português Brasileiro (PB), é um quantificador massivo: *muito livro-S(PL), muita água. Contudo, ele se combina com o singular nu (SNu): muito livro. Pires de Oliveira & Rothstein (2011) argumentam que o muito SNu gera leituras massivas; uma evidência de que o SNu é mass, já que nomes contáveis geram apenas leituras cardinais. Beviláqua (2015) verificou experimentalmente se muito SNu tem uma leitura massiva. Os resultados não falsearam a predição de Pires de Oliveira & Rothstein (2011), mas os dados não são conclusivos, pois não exclui a possibilidade de que “grinding” esteja licenciando essa leitura. Nós trazemos argumentos teóricos contra essa possiblidade de “grinding”, para, depois, explorar a semântico do muito SNu. Além do mais, Rothstein & Pires de Oliveira (in press) propõem que a leitura contável do SNu é resultado de “measure” (não “contagem”). Neste artigo, exploramos algumas das consequências disso, em especial um melhor entendimento de massa.