Semantic transduction

Semantic transduction names the reconstitution of a distinction or significance as it moves across settings or operational media. A tentative comment can become a recorded category; a reported experience can become a complaint; an institutional finding can become input to an AI service. The wording may remain similar while the contribution’s evidential status, authority and consequences change. (Watson and Brezovec, 2026, AE2.6)

What changes in a passage?

Translation between languages is one possible case, but the concept is broader. An expression can change its practical role while remaining in the same language. “The student may need help with fractions” might begin as a provisional prompt for discussion and later appear in a profile as an established learning difficulty. The transformation concerns what the contribution counts as and what it permits others to do.

AE asks researchers to follow these changes rather than assume that meaning travels unchanged. Relevant evidence includes who produced the expression, what uncertainty accompanied it, how it was interpreted, which procedure recognised it and what later activity relied on it. A preserved string of words does not establish continuity of these relations. (Watson, 2026, Operability)

Origins and neighbouring ideas

Simondon’s work makes the formation of individuals, rather than already completed individuals, a central philosophical problem. His treatment of transduction belongs to that account of individuation. AE’s semantic usage is a particular development in dialogue with this wider tradition; a shared term does not make the two accounts identical. (Simondon, 2020)

The concept also relates to scholarship on communication and infrastructure. Esposito studies the social significance of algorithmically generated contributions. Star and Ruhleder show why technical provision and situated use cannot be understood independently of organisational and communicative relations. These are resources for asking how a contribution becomes effective in a setting; neither source should be presented as having already formulated the whole AE account. (Esposito, 2022) (Star and Ruhleder, 1996)

A worked example: the correction that does not travel

In the guide’s constructed educational case, a student gives a valid solution using an unfamiliar method. An AI service generates a tentative diagnosis of misunderstanding. The teacher invites an explanation and concludes that the diagnosis was wrong. Meanwhile, a summary process has entered the original diagnosis in a learner profile without its qualification. (Watson, 2026, Operability)

Several transductions can now be distinguished. The student’s response became computational input. Generated text became a prompt for pedagogical interpretation. A tentative diagnosis became a recorded category. The category then became apparently established context for a later recommendation. These passages differ in their warrants and effects, even if parts of the original wording recur.

The teacher’s correction succeeds within the immediate dialogue. Whether it also succeeds institutionally depends on the record that governs the next relevant action. If the visible feedback changes but the profile does not, a subsequent service can retrieve the stored information accurately while perpetuating the original error. Explaining this outcome requires following the change in status and the pathway of correction, alongside the quality of the initial generation.

Transduction can also create possibilities

The concept is not restricted to distortion or loss. A student may turn an inarticulate concern into a question that others can address. A community’s account of an inaccessible service may become evidence that changes a procedure. An AI-generated comparison may help someone recognise a distinction they can subsequently explain and challenge. These are possible transformations whose benefits need their own evidence.

The question is what is preserved, changed or made possible at each passage. It is equally important to ask what becomes unavailable. A useful simplification for one purpose can remove a qualification needed for another. Those outcomes cannot be judged without specifying the purpose, audience and consequences.

How to investigate it

• Identify a contribution and the settings through which it passes.
• Record its assigned status, its actual uptake and the use its evidence warrants.
• Follow transformations of wording, uncertainty, provenance and authority.
• Examine whether corrections reach the next consequential use.
• Compare the explanation with alternatives such as source error, misleading presentation or a knowingly unjustified decision.

Bender and Koller’s distinction between form and meaning helps identify one limit: continuity or fluency of linguistic form cannot establish continuity of understanding. AE’s additional question concerns the practical organisation through which different contributions acquire significance and consequences. (Bender and Koller, 2020) (Watson and Brezovec, 2026, AE2.6)

Limits and responsibilities

Researchers should demonstrate a relevant transformation rather than label every movement of data semantic transduction. Technical conversion may be important without changing the contribution’s evidential role. Conversely, a consequential shift can occur without an observable wording change. The account requires evidence appropriate to both possibilities.

Following a chain of mediation also does not excuse the people or organisations controlling it. Responsibility depends on authority, knowledge, obligations and capacity to intervene. Making a transformation visible is useful when it helps locate an effective correction or a justified change in practice; visibility alone does not guarantee either.

Related entries

Meaning mediation introduces significance and uptake. Rough closure examines what remains consequentially unsettled. Ecological hyperknots relates these passages to different continuations.

Sources support the particular claims discussed; their inclusion does not imply endorsement of AE. Publication and manuscript status are identified below.

Watson and Brezovec, 2026, AE2.6

Watson, S., and Brezovec, E. (2026). Autopoietic Ecology: Rethinking Systems, Meaning, and Matter. AE2.6, revised manuscript dated 15 September 2026. Unpublished manuscript. Relevant locators: preface; chapters 3–6, 10–12 and 15–18; appendices A–C.

Watson, 2026, Operability

Watson, S. (2026). Generative AI and the recursive constitution of operability: Rough closure, semantic transduction and ecological hyperknots. Cambridge EdgeLab working paper, version 1, 18 September. Manuscript consulted; not peer reviewed. Constructed case and prospective research propositions.

Simondon, 2020

Simondon, G. (2020). Individuation in Light of Notions of Form and Information. Translated by T. Adkins. University of Minnesota Press. Source.

Esposito, 2022

Esposito, E. (2022). Artificial Communication: How Algorithms Produce Social Intelligence. MIT Press. Source.

Star and Ruhleder, 1996

Star, S. L., and Ruhleder, K. (1996). Steps toward an ecology of infrastructure: Design and access for large information spaces. Information Systems Research, 7(1), 111–134. Source.

Bender and Koller, 2020

Bender, E. M., and Koller, A. (2020). Climbing towards NLU: On meaning, form, and understanding in the age of data. Proceedings of ACL 2020, 5185–5198. Source.