A recipe for annotating grounded clarifications

Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 4065–4077 June 6–11, 2021. ©2021 Association for Computational Linguistics 4065 A recipe for annotating grounded clarifications Luciana Benotti Universidad Nacional de Córdoba CONICET, Argentina luciana.benotti@unc.edu.ar Patrick Blackburn Philosophy and Science Studies IKH, Roskilde University, Denmark patrickb@ruc.dk Abstract In order to interpret the communicative intents of an utterance, it needs to be grounded in something that is outside of language; that is, grounded in world modalities. In this paper we argue that dialogue clarification mechanisms make explicit the process of interpreting the communicative intents of the speaker’s utter- ances by grounding them in the various modal- ities in which the dialogue is situated. This pa- per frames dialogue clarification mechanisms as an understudied research problem and a key missing piece in the giant jigsaw puzzle of natural language understanding. We discuss both the theoretical background and practical challenges posed by this problem, and propose a recipe for obtaining grounding annotations. We conclude by highlighting ethical issues that need to be addressed in future work. 1 Introduction Clarifications are crucial to robust dialogues, and pragmatic factors — notably those shaped by the world modalities situating the conversation — have a key role to play. Referring expressions have in vision a modality in which to ground clarifications concerning objects in the world (de Vries et al., 2017); navigation instructions have in movement a modality in which to ground clarifications con- cerning collaborative wayfinding (Thomason et al., 2019). Clarifications grounded in situationally rel- evant modalities boost the redundancy required to learn to use language without explicit supervision, as they make explicit the process of negotiating the communicative intent. But despite its importance, work on clarification remains scattered. Humans switch between clarifications grounded in different modalities seamlessly but (we shall ar- gue) systematically. Our discussion is based around a general recipe for detecting grounded clarifica- tions; we work towards this in Section 2 by first reviewing the distinction between perceptual and collaborative grounding, and then discussing clar- ification mechanisms, Clark (1996)’s action lad- der of communication, and Ginzburg, Purver and colleagues (2012)’s classification of clarification phenomena. In Section 3 we draw these threads together and present the central idea: Given an utterance U, a subsequent turn is its clarification grounded in modality m if it cannot be preceded by positive evidence of understanding of U in m. This provides a unified way to frame clarification mechanisms and their interactions across various modalities; a graphical specification of the recipe it gives rise to can be found in Figure 2 of the supple- mentary material. It covers clarifications grounded in moving, grabbing and changing the physical world: these have traditionally been considered plain-old-questions (Purver et al., 2018), but we view them as useful clarification ingredients.1 In Sections 4 and A we test the practical implications of our recipe by identifying and characterizing (ac- cording to their modalities) the clarifications in a corpus of long dialogues in English. …

Content truncated to 500 words