Abstract

In a classic paper, Brennan and Clark argued that when interlocutors agree on a name for an object, they are forming a temporary agreement on how to conceptualize that object; that is, they are forming a conceptual pact. The literature on conceptual pacts has largely focused on the costs and benefits of breaking and maintaining lexical precedents, and the degree to which they might be partner-specific. The research presented here focuses on a question about conceptual pacts that has been largely neglected in the literature: To what extent are conceptual pacts specific to the local context of the interaction? If conceptual pacts are indeed temporary, then when the local context changes in ways that are accessible to participants, we would expect participants to seamlessly shift to referential expressions that reflect novel conceptualizations. Two experiments examined how referential forms change across context in collaborative, task-oriented dialog between naïve participants. In Experiment 1, names for parts of an unknown object were established in an “item” identification stage (e.g., a shape that looked like a wrench was called “the wrench”). In a second “build” stage, that name was often supplanted by an object-oriented name, e.g., the “leg.” These changes happened abruptly and without negotiation. In Experiment 2, interlocutors manipulated clip art and more abstract tangram pictures in a “slider” puzzle to arrange the objects into a target configuration. On some trials moving an object revealed a picture that could be construed as a contrast competitor, e.g., a clip art picture of a camel after “the camel” had been negotiated as a name for a tangram shape, or vice versa. As would be expected, modification rates increased when a potential contrast was revealed. More strikingly, the degree to which a name had been negotiated or the frequency with which it had been used did not affect the likelihood that the revealed shape would be considered as a potential contrast. We find little evidence that names that are introduced as part of a conceptual pact persist when either the task goals or informational needs change. Rather, conceptual pacts are fluid temporary agreements.

Highlights

  • There is a many-to-many mapping between names and potential referents

  • The answer was clearly “No,” highlighting that conceptual pacts are strongly dependent on task goal

  • A conceptual pact has historically been framed as a temporary agreement between interlocutors to refer to objects by a particular name and to adopt a particular conceptualization of the referent (Brennan and Clark, 1996)

Read more

Summary

Introduction

There is a many-to-many mapping between names and potential referents. A picture of a Bernese Mountain Dog could be referred to as “the dog”, “the Bernese Mountain Dog,” “the Berner,” etc. Building on the foundation established by Clark and Wilkes-Gibbs (1986) and Isaacs and Clark (1987) in developing the collaborative model of conversation, Brennan and Clark argued against what they termed “a historical” accounts of the generation of referring expressions. They proposed that when participants agree upon a name, that is when the name has been grounded, they are making a temporary agreement about how to conceptualize that object. There is an ongoing debate about how and when the use of a particular name for an object will affect processing, when an established name (a maintained precedent) or a different name (a broken precedent) is used by the same partner or by a different partner (see Kronmüller and Barr, 2015 for a recent review and meta-analysis)

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call