The Role of Explanatory Value in Natural Language Processing

TL;DR

This paper advocates for prioritizing explanatory value in NLP, emphasizing that explanation of linguistic phenomena should be central, distinct from model interpretability, and explores implications for research and policy.

Contribution

It introduces the importance of explanatory value in NLP, differentiates it from explainability, and discusses its potential impact on research directions and institutional policies.

Findings

Comparison of recent models of human language production

Discussion on implications of emphasizing explanatory value

Analysis of potential pitfalls in adopting explanatory focus

Abstract

A key aim of science is explanation, yet the idea of explaining language phenomena has taken a backseat in mainstream Natural Language Processing (NLP) and many other areas of Artificial Intelligence. I argue that explanation of linguistic behaviour should be a main goal of NLP, and that this is not the same as making NLP models explainable. To illustrate these ideas, some recent models of human language production are compared with each other. I conclude by asking what it would mean for NLP research and institutional policies if our community took explanatory value seriously, while heeding some possible pitfalls.