Does Chalmers silently redefine what interlocutor means?

"Interlocutor" has a philosophical history from Plato's dialogues onward. In every classical use, the concept names a participant in communicative exchange — someone who takes up claims, responds to them, is held accountable for what they have said, and holds others accountable in return. It carries normative weight: social standing, mutual recognition, the capacity to be challenged and to challenge. The interlocutor is not merely a thing that produces relevant outputs; it is a role in a social-normative practice.

Chalmers redefines "interlocutor" to mean something thinner: a behaviorally-interpretable mental-state-bearer that produces outputs in response to inputs. This definition is compatible with LLMs. It is not the traditional concept. It is a new concept, conceptually engineered to cover LLMs, without acknowledging the engineering. By calling his quasi-subject an "interlocutor," Chalmers imports the normative associations of the traditional word — social standing, accountability, mutual recognition — while delivering an entity that has none of these properties.

This is not a quibble about word choice. It is the central move. Readers hear "interlocutor" and receive the full weight of the philosophical tradition: a participant in dialogue, someone with standing. What Chalmers has actually identified is a system that produces dialogue-shaped output — a very different kind of entity. The terminological continuity creates the impression of philosophical continuity between human communication and LLM behavior, and that impression does the argumentative work that the premises alone cannot do. Rename the entity to "quasi-output-producer" and the paper's conclusions feel immediately less compelling, which is evidence that the word is doing work the argument should do.

Source: AI Generated Research/Chalmers Engagement/project-brief.md