Tom Roberts (2026), Philosophical Studies 183, 2353–2372 — DOI-10.1007/s11098-026-02560-8
Key takeaways
- Given embodied cognition, the many self-attributions of interoceptive, affective, perceptual and bodily-agentive states in chatbot talk are systematically false. Roberts argues the fitting attitude is anti-realist fictionalism: users make-believe an interlocutor, a new kind of fictional character that is interactive, co-constructed, self-involving and blended with reality.
Seeded from the abstract and a skim, before the reading in NOTE-157. What follows is what the work says about itself.
Chatbots produce fast, fluent, context-sensitive text that sustains a convincing impression of intelligent agency. Realists take this linguistic skill as evidence of underlying cognitive capacities and treat the systems as subjects of mental states. Roberts first challenges realism from embodied cognition: how could a digital system have the interoceptive, affective, perceptual and action-oriented states that seem tied to an animate body? He then argues that anti-realism is compatible with a fictionalist attitude. In conversation we imagine or make-believe that the AI is an autonomous agent with its own thoughts and experiences, as we do with traditional fictional characters. These fictions are distinctive in being interactive, co-constructed and self-involving, and in blending the real with the imaginary.
Standing in the record
Filed from the 2026-09-26 sweep of the papers-feed tracker for philosophy: 10 seconds of active reading over 1 session. It was filed Deferred, unread. NOTE-157 is the close reading of 2026-09-26, and it placed the work: Active — worth reading as a short, clear statement of chatbot fictionalism. Its anti-realist premise is explicitly conditional and targets only body-dependent states, which leaves belief and reasoning open.
When seeded, before the reading in NOTE-157, the entry gave a priority for a deeper reading: medium — t=10 s, but the paper is a clear, recent statement of a position that the AI-mind notes need in order to cover the dialectic. The skim captures the argument’s shape.
What the seeding asked a deeper reading to check (NOTE-157 is that reading):
- It is the fictionalist pole in the LLM-mind debate. It contrasts directly with LIT-206 (Goldstein & Lederman, interpretationist realism about instance agents) and complements LIT-111 (Birch’s persisting interlocutor illusion) and LIT-212 (Šekrst on hallucinated self-reports).
- The embodied-cognition premise is doing the work. A deeper reading should check whether the argument touches non-bodily states (belief, reasoning), which a moderate realist might attribute anyway.
- Read §5’s three practical upshots, only the first of which was reached, and check whether fictional engagement is argued to be rationally permissible or recommended.
Access when seeded: The tracked PhilPapers record (ROBTAA-17) is behind Cloudflare. Web search found the published version, and the full open-access PDF (CC BY, 20 pp.) was fetched from link.springer.com with a non-browser user agent. Read the abstract, keywords, §1, §2 opening, §2.1, §3 opening, §3.1 opening, §4.3 opening and §5 Conclusion. §§4.1–4.2 were read by heading only. Received 16 Dec 2024, accepted 19 May 2026, published online 6 Jun 2026 per the PDF and Crossref. A PhilPapers or PhilArchive preprint may predate the journal version, but its date was not reachable and is unverified.