Abstract
This survey paper proposes a clearer view of natural language reasoning inthe field of Natural Language Processing (NLP), both conceptually andpractically. Conceptually, we provide a distinct definition for naturallanguage reasoning in NLP, based on both philosophy and NLP scenarios, discusswhat types of tasks require reasoning, and introduce a taxonomy of reasoning.Practically, we conduct a comprehensive literature review on natural languagereasoning in NLP, mainly covering classical logical reasoning, natural languageinference, multi-hop question answering, and commonsense reasoning. The paperalso identifies and views backward reasoning, a powerful paradigm formulti-step reasoning, and introduces defeasible reasoning as one of the mostimportant future directions in natural language reasoning research. We focus onsingle-modality unstructured natural language text, excluding neuro-symbolictechniques and mathematical reasoning.