Retour vers liste

Détail de la contribution

Auteur: Ashley ATKINS

Things to Come: An Examination of Ceteris Paribus Outcomes

Abstract/Résumé: For forty years, there has been debate over whether the progressive has a modal component of meaning. Some theorists, including David Dowty (1977), (1979), Fred Landman (1992), and Paul Portner (1998), have argued for the view that it does. In favor of this position, one might claim that a sentence like 'Mary is leaping to the moon' is false and that it is false (roughly) because Mary cannot leap to the moon; the truth value of this claim seems to be partly determined by whether the outcome associated with it is possible. Other theorists, including Alice ter Meulen (1985), Terence Parsons (1990), and Zoltán Szabó (2004), (2008), have argued for the view that the progressive does not have a modal component of meaning. In favor of this position, one might observe that a sentence like 'Mary is leaping' does not invite a modal analysis; its truth value appears to be exclusively determined by what Mary is actually doing. Although the dominant view today is that the progressive has a modal meaning, the opposition to this view is entrenched. My aim is to settle this debate and to reconcile these parties by disagreeing with each in equal measure. Any account of the progressive must, I claim, account for the following pairs of asymmetrical entailments: (1) Mary was walking. --> Mary walked. (2) Mary was crossing the Atlantic. -/->Mary crossed the Atlantic. (3) Mary walked. --> Mary was walking. (4) Mary crossed the Atlantic. -/-> Mary was crossing the Atlantic. Against the modal theorists, I will argue that modal theories of the progressive cannot account for these asymmetries. Against the extensionalists, I will argue that these asymmetries can only be explained if many of the insights of modal theories are incorporated---not at the level of progressive meaning but at the level of outcome-associated event descriptions (which figure in (2) and (4)). I will offer an extensional account of the progressive along these lines, which accounts for these two pairs of asymmetries. I will claim that the progressive is simply a relation between an event and its proper parts and that outcome-associated event descriptions are associated with a ceteris paribus condition, which requires that an event develop into an event of the relevant type, other things equal. Alongside my proposed semantics for the progressive is a plea for what I call 'ceteris paribus events.' I demonstrate how positing these events in a semantic theory proves useful not only in accounting for the meaning of the progressive but also in providing a unified account of puzzling phenomena involving ability attributions and allowance claims, and in accounting for cross-linguistic patterns that are unattested in English (e.g. non-culminating accomplishments).