The project investigates the relationship between negation and cancellation by (i) taking a close and fresh look at the logical foundations of analysing negation as a complement mapping operator as opposed to negation as a content cancelling operator, and (ii) investigating linguistic cancellation phenomena. The main aim is to develop a theory of cancellation that is based on modern logical approaches and is capable of handling linguistically relevant empirical data. We argue that cancellation is a negation-like phenomenon, however not proposition-based, but rather discourse-related. Our project thus investigates negation phenomena at the discourse level and addresses the CRC’s general research question “Which aspects of negation are universal and why are they universal?”.

The envisaged theory will unfold in two dimensions which mutually inform each other. On the one hand, there is an investigation of the general (‘logical’) properties of the common ground (CG) and its dynamics, given the identified parameters and certain formal properties of them. On the other hand, the required parameters need to be identified in empirical language data so as to enable a meaningful application of the formal framework to concrete pieces of conversation.

We thus aim to collect and analyse linguistic data exemplifying cancellation that have so far not been treated in a unitary manner comprising data that have been discussed in the literature as well as experimentally elicited new data points. We furthermore extend the empirical data base to phenomena of multi-modal communication such as (non)manual co-speech gestures used in contexts of cancellation. The overarching aim of our project is to clarify the relationship of negation to cancellation and its relation to well-known discourse moves such as denial, refusal, and correction, as well as to develop a concise theory of the CG dynamics that can handle the cases of cancellation that we observe in human discourse.

While we will take visual and multi-modal data into account in the first funding period, the CG model we develop will not be one that can do full justice to visual data and the multi-modality of language. We plan to focus on multi-modality in the second and third funding period, collect more gestural and other visual data of cancellation and extend the CG model we develop in this first phase so as to appropriately handle also multi-modal input. Furthermore, we envisage to investigate the relationship of cancellation and abstention in future funding periods. While certain discourse configurations, where conflicting information and belief states are involved, give rise to CG updates that involve revisions (which is the topic of the project proposal at hand), speakers can also refuse to commit to one or the other stance. At this moment it is unclear how such cases of abstention should be modelled. This and the relationship of abstention to cancellation (that is, two belief states being in conflict vs. refusing to make a judgement about a certain proposition) will be the topic of future work in upcoming funding periods.