Communicating in Emergent Language with an Induced Morphological Phrasebook

ACL ARR 2026 January Submission454 Authors

Keywords: emergent language, morphology, emergent communication, compositionality

Abstract: We build rule-based emergent language (EL) agents using induced morphological phrasebooks and test their communicative performance in the EL environment with its neural network agents. This contributes three things: First, it assesses the quality of the morphemes discovered by the induction algorithm in situ, which we find to be effective for communicating in the EL. Second, it allows us to uncover morphosyntactic properties of EL through ablating the morpheme induction and the phrasebook algorithms, showing that the ELs rely on repetition as well as morpheme ordering to convey meaning. Third, we find that the bijectivity of morphemes (measured via normalized pointwise mutual information), serves as a metric of compositionality that is more closely correlated with the ability of the phrasebook-agents to "speak" and "hear" an EL than existing metrics such as topographic similarity or bag-of-symbols disentanglement.

Paper Type: Long

Research Area: Phonology, Morphology and Word Segmentation

Research Area Keywords: morphological segmentation, morphological analysis

Contribution Types: Model analysis & interpretability, Publicly available software and/or pre-trained models

Languages Studied: emergent language

Submission Number: 454