Is Consciousness a Story We Tell Ourselves?

Internal language and consciousness after chain of thought reasoning.

Where and when? A Symposium at the AISB Convention 2026 aisb.org.uk/aisb-convention-2026 1-2 July 2026 University of Sussex, Brighton, UK Hosted by The Society for the study of Artificial Intelligence and Simulation of Behaviour Organizers Conor Houghton conorhoughton.github.io / conor.houghton@bristol.ac.uk Jonathon Shock shocklab.net Seth Bullock homepage

Key dates

There will be a session for 15 minute contributed talks. If you would like to take part submit an abstract of less than 400 words by 15 April to:

• easychair.org/conferences/?conf=astory2026

Plan

• Session 1: panel discussion (1 hour). Is consciousness a story we tell ourselves?
• Session 2: contributed talks (1 hour). A set of four or five short talks.
• Session 3: panel discussion (1 hour). Is consciousness an experimental question?

Description

The aim of this symposium is to consider to sharpen our consideration of consciousness by looking at it through the lens of language; this may be a useful approach or one that distracts us from important issues of embodiment or emotion. This discussion aims to explore this! This symposium will consider

• The relationship between language and consciousness
• Language and inference in humans and machines
• The influence the role of language has on its structure.

Consciousness may arise from, or be structured by, our ability to access and reason about internal representations through the internal use of language or it may be completely unrelated to language. This half-day symposium will address the question: "Is consciousness a story we tell ourselves?" by examining the relationship between consciousness, inner speech in humans and chain-of-thought reasoning in machines.

Many theories of consciousness assume some role for internal representations. However, language has often been treated as secondary. This is surprising. although language is most obviously a tool for communication, it provides the structured, sequential space in which humans build multi-step inferences.

Conversely language use may be unrelated to consciousness; some people report thinking in pictures rather than words and a language-based definition of consciousness appears to restrict consciousness to humans and deny that other animals are conscious. Many regard this restriction as absurd.

Questions of consciousness and language have a new urgency because of the recent machine learning revolution. Chain-of-thought reasoning in large language models provides a striking parallel: models become more capable when they are allowed to "talk to themselves", producing extended internal sequences that guide inference.

This field lacks an agreed experimental programme but is rich in experimental questions. To what extent can we disrupt inner speech without fragmenting conscious experience? Can non-human animals or artificial systems support forms of consciousness? What are the crucial questions that are capable of empirical resolution?