On the Optimality of Discrete Object Naming: a Kinship Case Study

TL;DR

This paper develops an information-theoretic framework for discrete object naming systems, demonstrating that optimal trade-offs between informativeness and complexity are achievable and empirically observed in kinship communication models.

Contribution

It introduces a new framework addressing previous simplifications and proves the conditions for optimality in kinship naming systems, supported by empirical evidence.

Findings

Optimal trade-off is achievable if listener's decoder matches Bayesian decoding.

Empirical emergence of optimality in learned kinship communication systems.

Theoretical proof linking decoder type to naming system optimality.

Abstract

The structure of naming systems in natural languages hinges on a trade-off between high informativeness and low complexity. Prior work capitalizes on information theory to formalize these notions; however, these studies generally rely on two simplifications: (i) optimal listeners, and (ii) universal communicative need across languages. Here, we address these limitations by introducing an information-theoretic framework for discrete object naming systems, and we use it to prove that an optimal trade-off is achievable if and only if the listener's decoder is equivalent to the Bayesian decoder of the speaker. Adopting a referential game setup from emergent communication, and focusing on the semantic domain of kinship, we show that our notion of optimality is not only theoretically achievable but also emerges empirically in learned communication systems.