The Theory

2.2 Communication Between Sender and Receiver

Before presenting a theory, we need to consider what happens when we try to communicate it. A text does not directly transport the writer’s understanding into the reader’s mind. It offers words, examples, and relationships from which the reader constructs an interpretation.

The inquiry into communication presented here emerges from ORT, but does not constitute the exposition of the theory. Its role at this stage is to prepare the ground for reading: to examine how perception, interpretation, and representation participate in constructing an explanation and in another person’s understanding of it.

The starting point was a practical difficulty: producing a message whose meaning would remain invariant between sender and receiver, initially between a human being and an artificial intelligence. Adding definitions, examples, and constraints seemed a way to eliminate ambiguity. Yet each clarification relied on new terms that also needed interpretation. Precision could reduce some possible readings without eliminating all the others.

This difficulty shifted the question. Rather than first seeking a message immune to every transformation of meaning, it became necessary to investigate the process through which something is perceived, interpreted, represented, and communicated. A message reaches someone who already has experiences, references, and conditions of their own; it does not arrive at an empty starting point.

The technical project proposes examining these conditions together. An individual’s trajectory participates in how they respond, while circumstance includes the present conditions of their relationship with themselves and their environment. Perception, interpretation, and representation are separated in the explanation so that we can investigate them, without assuming that they occur as isolated stages or in a rigid sequence. The model under development must make their recursive relationships explicit.

In communicating, the sender selects something of what they intend to express and organizes it through a code. The receiver encounters the resulting message, not the totality of the relationships that participated in its production. From it, the receiver develops their own interpretation. The project proposes the hypothesis that representations produced by distinct individuals do not possess absolute identity; this hypothesis must be examined, rather than treated as demonstrated merely because it has been stated or written mathematically.

This does not require treating all communication as failure. Two interpretations may correspond sufficiently for a particular purpose without being identical. For this reason, the project does not use correspondence as an absolute magnitude: we must specify the criterion by which two representations are being compared. An approximation sufficient for one purpose may be insufficient for another.

Dialogue allows this process to be revisited. A response offers new elements for interpretation, may reveal differences, and makes it possible to reformulate the message. But repeating the exchange does not, by itself, guarantee closer correspondence. Feedback may increase, reduce, or reorganize correspondence, and the model must investigate the conditions under which this happens.

For reading ORT, the consequence is methodological: before comparing a concept with another already familiar to us, we need to understand how it is being used in the text and which relationships it seeks to express. An explanation can guide that understanding, but cannot replace the reader’s interpretive work. The following premises return to these considerations before the exposition of the theory.

The research project linked to this section preserves the technical development of this inquiry: its modules, provisional definitions, hypotheses, proposed formalizations, and criteria for future verification. It is a work plan under development. Formalization does not constitute evidence, and hypotheses must be modified or rejected if the investigation does not support them.

View technical development — 5.1