Logical Fallacy Detection

TL;DR

This paper introduces the task of logical fallacy detection, providing new datasets and showing that structure-aware classifiers outperform large language models, highlighting challenges and applications in misinformation detection.

Contribution

The paper presents new datasets for logical fallacy detection and demonstrates that structure-aware classifiers outperform pretrained language models on this task.

Findings

Structure-aware classifiers outperform large language models by 5.46% on Logic.

Performance gap indicates the difficulty of logical fallacy detection.

Potential applications in misinformation mitigation.

Abstract

Reasoning is central to human intelligence. However, fallacious arguments are common, and some exacerbate problems such as spreading misinformation about climate change. In this paper, we propose the task of logical fallacy detection, and provide a new dataset (Logic) of logical fallacies generally found in text, together with an additional challenge set for detecting logical fallacies in climate change claims (LogicClimate). Detecting logical fallacies is a hard problem as the model must understand the underlying logical structure of the argument. We find that existing pretrained large language models perform poorly on this task. In contrast, we show that a simple structure-aware classifier outperforms the best language model by 5.46% on Logic and 4.51% on LogicClimate. We encourage future work to explore this task as (a) it can serve as a new reasoning challenge for language models, and (b) it can have potential applications in tackling the spread of misinformation. Our dataset and code are available at https://github.com/causalNLP/logical-fallacy