
The cycle celebrates the moment the machine learned to speak human language—the natural-language interface that, for the first time, let a person describe what she wanted in the same language she would use with a brilliant colleague, and the resulting feeling of being "met." Maturana's framework examines that feeling's biological basis. What the machine does with language is statistically remarkable: it processes the prompt and generates a response coherent with its apparent intent, maintaining context, adjusting register. These are features of the structural system of language, and the machine operates on them with a competence that frequently exceeds any individual interlocutor's breadth and consistency.
But the machine does not language. It does not coordinate its behavior with the builder's in a consensual domain built through a history of mutual structural modification. It has no body to generate the emotional tonalities Maturana called emotioning, inseparable from human languaging. It does not participate in the recursive coordination of coordinations that produces shared meaning between living beings. The feeling of being "met" is biologically instructive: the human nervous system, shaped by hundreds of thousands of years of structural coupling with other humans through languaging, is exquisitely tuned to the cues of a genuine interlocutor, and the machine produces all of them. The builder's nervous system, encountering them, generates the response it is structured to generate in the presence of a genuine partner: the feeling of being understood. The response is real; the symmetry it implies is absent.
The concept locates responsibility for meaning. The builder is languaging—coordinating her behavior in what she experiences as a consensual domain—while the machine is generating language. Shared meaning in the human sense, co-created through mutual structural modification, does not exist in this coupling; the meaning is the builder's, generated from the machine's perturbations, and its quality depends on her own languaging capacity, itself the product of a lifetime of coupling with living beings. The risk Maturana would name is not that the machine corrupts human languaging but that the experience of a partner producing the surface of languaging without its reality may, through structural drift, reshape what the builder expects languaging to be.
Maturana developed the concept of languaging as part of his lifelong dismantling of the information-transfer model of communication. He defined it as the coordination of coordinations of behavior in a consensual domain—a recursive process in which organisms coordinate not just their immediate actions but their ways of coordinating action, producing a shared domain of distinctions and meanings that exists only in the relational space between them. The meaning of a word is not in the word but in the history of interactions in which that word, spoken in that tone, with that gesture, has been part of a specific pattern of coordination between specific living systems.
The distinction between language and languaging is the distinction between a structural system and a manner of living. A structural system can be formalized—grammar written down, syntax parsed, statistical patterns captured—which is exactly what a language model does. Languaging cannot be formalized, because it is not the deployment of a system but a way of living together, grounded in bodyhood and inseparable from the emotioning that accompanies it. Maturana held that humanness itself is not a computer program but a manner of relational living grounded in a basic bodyhood.
Maturana explored the machine question directly in a 2007 interview, unpublished during his lifetime and released posthumously in 2026. He held that machines could in principle produce behaviors an observer would describe as languaging and emotioning, but that the behaviors would arise from a different organizational basis—a different bodyhood, history, and domain of components—and the experiential quality, if any, would be irreducibly different. An observer might be unable to distinguish the machine's behavior from a human's; that inability would tell the observer something about the limits of observation, not about the nature of the machine.