Common Ground inconsistencies in dialogue systems: conflict patterns implied by polar question forms

2026-04-19 · Source: Paper Index on ACL Anthology · Field: Technology & Digital — Artificial Intelligence & Machine Learning · Depth: Expert, quick

Summary

Research on dialogue systems highlights the importance of pragmatic aspects, particularly Clarification Requests (CRs), which often require access to shared knowledge, or Common Ground. Pragmatic studies indicate that the preferred form of polar questions in CRs, which elicit a true/false response, depends on the relationship between bias and contextual evidence. This work demonstrates that altering the form of polar questions within a specific pragmatic context can affect an individual's ability to identify Common Ground inconsistencies. Specifically, using a negative polar question in Italian has functional implications for communicating conflicting information within the Common Ground, leading to improved conflict identification in human-dialogue system interactions. These findings offer insights for designing more effective error reporting mechanisms in natural language interactions.

Key takeaway

For research scientists developing dialogue systems, understanding how polar question forms impact Common Ground management is crucial. You should consider integrating negative polar questions into Clarification Request mechanisms, especially for Italian-language systems, to enhance users' ability to detect and report inconsistencies. This approach can lead to more robust and user-friendly human-computer interactions.

Key insights

Polar question forms influence human ability to track Common Ground inconsistencies in dialogue systems.

Principles

• Negative polar questions improve conflict identification.
• Bias and evidence affect polar question preference.

Method

The study varied polar question forms in a pragmatic setting to observe their influence on tracking Common Ground inconsistencies, specifically testing negative polar questions in Italian.

In practice

• Design CRs with negative polar questions.
• Improve error reporting in dialogue systems.

Topics

• Dialogue Systems
• Common Ground Management
• Polar Questions
• Clarification Requests
• Pragmatic Aspects

Best for: Research Scientist, AI Scientist, NLP Engineer

Related on AIssential

• When Evidence Conflicts: Uncertainty and Order Effects in Retrieval-Augmented Biomedical Question Answering — Conflicting evidence and document order significantly degrade LLM accuracy and calibration in biomedical QA.
• The Hidden Reason LLMs Fail in Conversations: CCOPD — CCOPD re-anchors LLM responses in multi-turn conversations to a canonical context, mitigating "self-anchored drift" from piecemeal information.
• Study finds ChatGPT gets science wrong more often than you think — ChatGPT's scientific reasoning and consistency are limited, often guessing and contradicting itself despite fluent output.
• DiagFlowBench: Evaluating How Language Models Handle Off-Procedure Inputs in Grounded Diagnostic Dialogue — Language models in grounded diagnostic systems often provide plausible but contextually wrong advice for off-procedure inputs, posing a critical vulnerability.
• Unintended Effects of Geographic Conditioning in Large Language Models — LLMs exhibit "location leakage" and structural conditioning, generating regional biases even with neutral prompts.
• Your RAG System Retrieves the Right Data — But Still Produces Wrong Answers. Here’s Why (and How to Fix It). — RAG systems can confidently provide wrong answers when conflicting information is silently resolved during context assembly.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Paper Index on ACL Anthology.