What Meaning-Form Correlation Has to Compose With
This work addresses the challenge of accurately measuring compositionality in natural languages, which is a foundational problem for researchers in computational linguistics and natural language processing.
This paper investigates the assessment of compositionality in natural languages by measuring meaning-form correlation (MFC). It analyzes MFC across artificial toy languages, English dictionary definitions, and English sentences from literature, finding that linguistic phenomena like synonymy and ungrounded stop-words significantly impact MFC measurements.
Compositionality is a widely discussed property of natural languages, although its exact definition has been elusive. We focus on the proposal that compositionality can be assessed by measuring meaning-form correlation. We analyze meaning-form correlation on three sets of languages: (i) artificial toy languages tailored to be compositional, (ii) a set of English dictionary definitions, and (iii) a set of English sentences drawn from literature. We find that linguistic phenomena such as synonymy and ungrounded stop-words weigh on MFC measurements, and that straightforward methods to mitigate their effects have widely varying results depending on the dataset they are applied to. Data and code are made publicly available.