Solving the Partition Problem: Signaling Games and the Cost-Based Perception Model.

Signaling games have been widely used as a powerful theoretical tool for investigating the propositional content of signals in the evolution of language and communication. Although significant progress has been made, information-theoretic approaches to content within the sender-receiver framework fa...

Full description

Bibliographic Details
Published in:Erkenntnis Vol. 91; no. 4; pp. 2035 - 2063
Main Authors: Wang, Jidong, Zhang, Mingjun
Format: Article
Published: Springer Nature Apr2026
Subjects:
Online Access:View this record in EBSCOhost
Description
Summary:Signaling games have been widely used as a powerful theoretical tool for investigating the propositional content of signals in the evolution of language and communication. Although significant progress has been made, information-theoretic approaches to content within the sender-receiver framework face what Jonathan Birch calls the "partition problem." In this article, we address this problem by integrating signaling games with cost-based perception to model the internal state of an organism as it perceives objects in its environment. We propose a simple model that shows how an organism can spontaneously learn to partition external states in different ways through reinforcement learning, thereby offering a promising solution to the partition problem.