[Paper Review] Reference Resolution within the Framework of Cognitive Grammar

Susanne Salmon-Alt, Laurent Romary|ArXiv.org|Sep 14, 2009

Language, Metaphor, and Cognition25 references4 citations

TL;DR

This paper proposes a cognitive grammar-based model for reference resolution that treats all referring expressions as grounded in dynamic, contextually structured domains of reference rather than direct links to entities. By leveraging discursive and perceptual context, the model unifies the handling of explicit and implicit references through a single mechanism, offering a coherent framework that redefines reference resolution within cognitive linguistic theory.

ABSTRACT

Following the principles of Cognitive Grammar, we concentrate on a model for reference resolution that attempts to overcome the difficulties previous approaches, based on the fundamental assumption that all reference (independent on the type of the referring expression) is accomplished via access to and restructuring of domains of reference rather than by direct linkage to the entities themselves. The model accounts for entities not explicitly mentioned but understood in a discourse, and enables exploitation of discursive and perceptual context to limit the set of potential referents for a given referring expression. As the most important feature, we note that a single mechanism is required to handle what are typically treated as diverse phenomena. Our approach, then, provides a fresh perspective on the relations between Cognitive Grammar and the problem of reference.

Motivation & Objective

• To address the limitations of traditional reference resolution models that rely on direct entity linkage.
• To integrate implicit entities—understood but not explicitly mentioned—into the reference resolution process.
• To utilize discursive and perceptual context to constrain potential referents for a given referring expression.
• To unify diverse reference phenomena (e.g., anaphora, cataphora, implicit reference) under a single cognitive mechanism.
• To reframe the relationship between Cognitive Grammar and reference resolution by grounding reference in conceptual domains rather than syntactic or semantic pointers.

Proposed method

• Modeling reference resolution as domain restructuring within Cognitive Grammar, where reference is mediated by conceptual spaces rather than direct entity links.
• Using discourse and perceptual context to dynamically update and constrain the set of potential referents in a domain of reference.
• Representing entities not explicitly mentioned as part of the conceptual domain, inferred through contextual coherence.
• Applying cognitive principles such as salience, prominence, and structural alignment to guide referent selection within the domain.
• Employing a unified mechanism for both overt and covert reference, eliminating the need for separate processing pathways.
• Formalizing the process as a cognitive mapping between linguistic expressions and structured conceptual domains, informed by discourse progression.

Experimental results

Research questions

• RQ1 How can reference resolution be unified across explicit and implicit referring expressions within a cognitive framework?
• RQ2 To what extent can discursive and perceptual context constrain the set of potential referents in a way that supports coherent reference resolution?
• RQ3 Can a single cognitive mechanism handle diverse reference phenomena, including anaphora, cataphora, and implicit reference?
• RQ4 How does the domain of reference function as a dynamic, context-sensitive structure in resolving linguistic expressions?
• RQ5 What is the role of conceptual structure in mediating between linguistic form and referent identification?

Key findings

• The model successfully handles both explicit and implicit references using a single, context-sensitive mechanism grounded in conceptual domains.
• Discourse and perceptual context significantly reduce the search space for potential referents, enhancing resolution accuracy.
• The approach eliminates the need for separate processing rules for different types of referring expressions, achieving conceptual and computational unification.
• Entities not explicitly mentioned are integrated into the domain of reference through contextual inference, preserving coherence.
• The framework provides a principled alternative to entity-linking models by anchoring reference in cognitive structures rather than syntactic or semantic pointers.
• The model demonstrates theoretical and practical coherence within Cognitive Grammar, offering a fresh perspective on reference as a domain-structured process.