Thematic Session Proposal for a Workshop on the Concept of Possibility and its Morphological, Syntactic and Pragmatic Realizations in Natural Language

Date: 21-Aug-2024 - 24-Aug-2024
Location: Helsinki, Finland
Contact: Patrick Duffley
Contact Email: [email protected]

Linguistic Field(s): General Linguistics

Meeting Description:

Within linguistics, possibility has been approached in different ways, but generally with circumspection. Palmer (1979) treats it as constituting, along with necessity, the core of the category of modality. In Palmer (2001), modality is defined as depicting the reality-status of the proposition in terms of some form of irrealis. This implies that the relation between possibility and irrealis needs to be explored. Mithun (1999) draws a distinction between realis, which presents events as real, actualized or occurring, and irrealis, which presents them as not actualized or “purely within the realm of thought.” Bybee (1998) has criticized the notion of irrealis as being too inconsistent crosslinguistically to constitute a useful analytical category and very rarely attested in binary opposition to realis in the languages of the world. On the other hand, von Prince, Krajinović and Krifka (2022) have argued recently that “irrealis is real” and represents a crosslinguistically meaningful notion that can be properly understood if split into two domains – the possible and the counterfactual. They characterize these domains in terms of a branching time framework in which the possible is defined as the successor of the actual and the counterfactual as being neither actual nor possible. But isn’t the successor of the actual the future and not the merely possible? And if possibility is a species belonging to the genus of irrealis, what is its relation to the other species belonging to this genus?

In the verbal domain, the possible worlds model has been used by Kripke (1959, 1972) to distinguish the notions of possibility and necessity by mobilizing the distinction between existential and universal quantification. Possibility is defined via existential quantification, as the case where a proposition is true in at least one but not all possible worlds; necessity is defined by universal quantification, as representing a proposition as true in all possible worlds. One may wonder however whether the distinction between She may be at work and She must be at work is adequately described by saying that the first is true in at least one and perhaps more but not all possible worlds, while the second is true in all such worlds. Kripke’s definition of possibility could also be accused of being circular, as it employs the notion of ‘possible worlds’ in order to define the notion of ‘possible’. In addition, it seems incapable of distinguishing between the two forms of possibility expressed by the modal auxiliaries can and may in English. The peculiar meaning of epistemic may in a use such as She may be at work has been characterized by van der Auwera (2001: 28-31) as ‘POSS (NOT P)’. In contrast, following Aristotle, he defines the standard notion of possibility (‘POSS P’) in terms of necessity, as ‘NOT NEC (NOT P)’. None of these glosses corresponds however to the meaning expressed by may in uses such as You may go to the washroom now or Enrollment information may be found on the Faculty website, nor to the use of can in She can swim. This casts doubt on their adequacy as tools for describing the meanings of these modals in a manner that respects their natural semantic unity. But what tools do we need to achieve this type of description and to handle differences such as that between can and may?

Kratzer (1991) has added to the modal toolkit the notions of ordering source and modal base. The ordering source is the set of propositions that represents the normal course of events in a given possible world. The modal base is the set of propositions which forms the basis of the evaluation that a given modality obtains, the result of this evaluation being termed the modal force (e.g. ‘possible’/‘necessary’). Thus John can open a beer bottle with his teeth is analyzed as ‘Given his abilities, the strength of his teeth, etc., it is possible for John to open a beer bottle with his teeth.’ While this framework specifies the setting in which possibility is embedded, one may wonder what it says about the definition of possibility itself. Moreover, it seems to offer no means for distinguishing between the uses of may vs. can to express permission (You may/can go to the washroom): both would be paraphrased as ‘Given the fact that the relevant authority-figure has granted permission, it is possible for you to go to the washroom.’ Are other models able to capture this difference?