Modal indefinites

Natural Language Semantics, Mar 2010

Across languages, we find indefinites that trigger modal inferences. This article contributes to a semantic typology of these items by contrasting Spanish algún with indefinites like German irgendein or Italian uno qualsiasi. While irgendein-type indefinites trigger a Free Choice effect (Kratzer and Shimoyama 2002; Chierchia 2006), algún simply signals that at least two individuals in its domain are possibilities. Additionally, algún, but not irgendein, can convey that the speaker does not know how many individuals satisfy the existential claim in the world of evaluation. We contend that the two types of indefinites impose different constraints on their domain of quantification: irgendein and its kin are domain wideners (Kratzer and Shimoyama 2002), whereas algún is an ‘anti-singleton’ indefinite (its domain cannot be restricted to a singleton). This, together with the fact that algún does not require uniqueness, allows us to derive the contrast between irgendein and algún by using the pragmatic reasoning presented by Kratzer and Shimoyama.

Article PDF cannot be displayed. You can download it here:

https://link.springer.com/content/pdf/10.1007%2Fs11050-009-9048-4.pdf

Modal indefinites

Luis Alonso-Ovalle 0 1 Paula Menendez-Benito 0 1 0 P. Menendez-Benito Seminar fur Englische Philologie, University of Gottingen , Kate-Hamburger-Weg 3, 37073 Gottingen, Germany 1 L. Alonso-Ovalle (&) Department of Hispanic Studies , McCormack 4-609, University of Massachusetts Boston , 100 Morrissey Blvd., Boston, MA 02125, USA Across languages, we find indefinites that trigger modal inferences. This article contributes to a semantic typology of these items by contrasting Spanish algun with indefinites like German irgendein or Italian uno qualsiasi. While irgendein-type indefinites trigger a Free Choice effect (Kratzer and Shimoyama 2002; Chierchia 2006), algun simply signals that at least two individuals in its domain are possibilities. Additionally, algun, but not irgendein, can convey that the speaker does not know how many individuals satisfy the existential claim in the world of evaluation. We contend that the two types of indefinites impose different constraints on their domain of quantification: irgendein and its kin are domain wideners (Kratzer and Shimoyama 2002), whereas algun is an 'anti-singleton' indefinite (its domain cannot be restricted to a singleton). This, together with the fact that algun does not require uniqueness, allows us to derive the contrast between irgendein and algun by using the pragmatic reasoning presented by Kratzer and Shimoyama. 1 Introduction Across languages, we find indefinites that trigger modal inferences. One such indefinite is Spanish algun. Consider, as an illustration, (1) below. This sentence makes an existential claim (that there is a student that Mara married), and additionally conveys that the speaker does not know who the witness of this claim is (i.e., the speaker doesnt know which student Mara married). Hence, adding the continuation namely, Pedro, which explicitly identifies the witness, results in oddity, as (2) illustrates. In contrast, the plain indefinite un allows for this type of continuation, as in (3). ] Mara se caso con algu n estudiante Mara SE married with ALGU N student departamento de lingu stica: en concreto con department of linguistics: namely with Mara married a linguistics student, namely Pedro. Mara se caso con un estudiante Mara SE married with UN student ling ustica: en concreto con Pedro. linguistics: namely with Pedro Mara married a linguistics student, namely Pedro. In a possible world semantics, the ignorance component of sentences like (1) can be modeled by saying that algun imposes a constraint on the speakers epistemic alternatives (the set of worlds compatible with what the speaker believes), namely that Mara didnt marry the same linguistics student in all those worlds. When algun is in the scope of an intensional operator, it imposes the same type of constraint on the worlds that the operator quantifies over. This is illustrated by (4) below, where algun is in the complement clause of a propositional attitude verb, pensar (to think): Pedro piensa que Mara se caso con Pedro thinks that Mara SE married with del departamento de lingu stica. of the department of linguistics Pedro thinks that Mara married a linguistics student. In (4), algun can have scope over or under the attitude verb. On the wide scope reading of algun, (4) conveys that there is a particular student that Pedro thinks Mara married, but the speaker does not know who. This is the speakers ignorance reading that we saw above. When algun has narrow scope, (4) says that Pedro is uncertain about the identity of the student that Mara married. In other words, Pedros epistemic alternatives vary with respect to the identity of the student that Mara married. In cases like (1) and (4) above, we are likely to make a uniqueness assumption: in each accessible world, Mara married only one student. When uniqueness cannot be taken for granted, algun can convey ignorance with respect to the total number of individuals that satisfy the existential claim. The example in (5), for instance, strongly suggests that the speaker does not know how many dents her car has.1 A number of recent works focus on indefinites that convey a modal component, henceforth modal indefinites.2 These studies differ widely with respect to the description and analysis of the modal component. Since no systematic crosslinguistic investigation of this class of indefinites has been undertaken, it is not clear whether these divergences correspond to typological differences. This sets the stage for a research program which aims to understand along which lines modal indefinites can vary, and to seek a unifying core underlying the observed diversity. This paper contributes to this enterprise by describing the modal component of algun and contrasting it with that of modal indefinites like German irgendein or Italian uno qualsiasi. These indefinites have been characterized in the literature as Existential Free Choice Items because they convey, roughly, that each of the individuals in the domain of quantification can satisfy the existential claimthe Free Choice component: see Kratzer and Shimoyama (2002), Kratzer (2005), Chierchia (2006). The sentence in (6), for instance, claims that Mary had to marry a doctor, and, additionally, that any doctor in the domain of quantification was a permitted option. (Kratzer 2005, p. 129) Kratzer and Shimoyama (2002), who analyze the Free Choice component associated with irgendein-type indefinites, put forward an account for this component that crucially relies on the assumption that these indefinites are domain wideners, i.e., they cannot be contextually restricted. For instance, irgendein Arzt picks out the 1 A note about the translations of our example sentences: we use English a in the translations of the examples in which algun conveys ignorance regarding the identity of the witness, even though a lacks the modal component of algun. In cases like (5), where algun conveys ignorance with respect to the number of witnesses, we will only provide a gloss. The use of a or (singular) some in those cases would convey that there is a unique individual satisfying the existential claim. 2 Some examples of modal indefinites are: English singular some (Strawson 1974; Becker 1999; Farkas 2002), German irgendein (Kratzer and Shimoyama 2002; Aloni and van Rooij 2004; Aloni 2007), the -to series in Russian (Yanovich 2005; Kagan 2007), the -kin series in Finnish (Kagan 2007), Romanian vreun and un NP oarecare (Farkas 2006; Ciucivara 2007), French quelque, un NP quelconque, and nimporte quoi (Zabbal 2004; Tovena and Jayez 2006), and Italian (un) qualche and uno qualsiasi (Aloni and van Rooij 2004; Chierchia 2006; Zamparelli 2007). set of all doctors in the world of evaluation, rather than a contextually salient subset of doctors. In this paper, we show that the modal effect induced by algun is weaker than Free Choice: algun only requires that there be at least two individuals in the domain of quantification that satisfy the existential clai (...truncated)


This is a preview of a remote PDF: https://link.springer.com/content/pdf/10.1007%2Fs11050-009-9048-4.pdf
Article home page: https://link.springer.com/article/10.1007/s11050-009-9048-4

Luis Alonso-Ovalle, Paula Menéndez-Benito. Modal indefinites, Natural Language Semantics, 2010, pp. 1-31, Volume 18, Issue 1, DOI: 10.1007/s11050-009-9048-4