Introduction

Traditional design methodologies frequently begin with the question: “What should be communicated?”

Under this approach, words, images, colors, composition, sound, and motion are selected as tools for delivering a predefined meaning to an audience.

However, creative practice suggests that viewers do not receive meaning passively. They observe fragments, detect boundaries, recognize objects, complete missing information, and construct their own narratives.

Creators may not design outcomes directly.
They may design the state from which outcomes emerge.

This paper introduces State as an intermediate layer connecting expression and meaning.

Conventional Communication Model

A simplified conventional model treats expression as the direct carrier of meaning.

Let E denote an observable expression and M denote the meaning interpreted by the viewer.

This model is useful for functional communication, but it does not fully explain why identical expressions can generate different meanings while producing a similar sense of stillness, tension, curiosity, or attention among different viewers.

Proposed State-Mediated Model

The proposed model introduces an intermediate cognitive layer, defined as State.

Here, S does not represent semantic meaning or a specific emotion. It represents a temporary cognitive condition that precedes meaning generation.

Examples of State include:

• stopping
• observing
• becoming quiet
• focusing
• imagining
• questioning

Bidirectional Transformation

Creator and Viewer are modeled as inverse transformation processes.

Creator Transformation

The Creator begins with an intention, concept, or narrative and identifies the State through which it may be experienced. Scene, composition, light, color, sound, motion, distance, and information density are then selected to form the expression.

Viewer Transformation

The Viewer encounters the expression, enters a cognitive State, and constructs a personal meaning from perception, memory, culture, experience, and imagination.

The resulting meaning M′ does not necessarily equal the creator’s original meaning M.

Reversibility Hypothesis

If Creator and Viewer operate as inverse processes, the combined relationship may be expressed conceptually as:

Exact semantic reconstruction is neither assumed nor required. Instead, successful expression may depend on preserving a target State through both transformations.

In this conceptual expression, d represents a future measurement function for the difference between the State intended by the Creator and the State experienced by the Viewer.

Perception and Expression Sequences

Preliminary observation suggests that visual perception may develop through a staged sequence.

Viewer Sequence

Creator Sequence

The Creator may proceed in the inverse direction:

The Creator decomposes an internal narrative into increasingly perceptible elements. The Viewer reconstructs a narrative from those elements.

State Design

The model changes the primary objective of creative design.

Instead of directly targeting outcomes such as surprise, understanding, emotion, or action, the Creator targets a State from which those outcomes may emerge.

Hierarchical Expression Parameters

Expression parameters are assumed to form a hierarchical rather than a flat structure.

1. Concept Layer
   What should be explored or communicated?
2. State Layer
   What cognitive condition should be created?
3. Scene Layer
   In what world, location, time, or situation should it appear?
4. Composition Layer
   How should objects, distances, boundaries, and visual attention be arranged?
5. Rendering Layer
   How should color, light, contrast, texture, sound, typography, and margin be expressed?

This equation is conceptual notation rather than a validated mathematical function. It is introduced to clarify the layered dependency of expression generation.

Implications for Artificial Intelligence

Many current generative systems are modeled conceptually as a direct transformation from prompt or meaning to output expression.

The proposed model suggests an alternative architecture:

Here, P represents hierarchical expression parameters. AI functions as an Expression Transformer that renders expressions while preserving the intended State.

Viewer feedback may then be used to estimate the difference between the target and experienced States.

This equation is presented as a conceptual learning model and does not yet define a concrete optimization algorithm.

Future Research

Future research should investigate:

• the taxonomy and measurement of cognitive States;
• the distinction between State, emotion, and meaning;
• the relationship between scenes and State formation;
• the effects of margin, light, time, sound, and information density;
• the degree to which State can be shared across viewers;
• the empirical validity of Creator–Viewer reversibility;
• AI-assisted optimization of expression parameters;
• applications to film, photography, Web design, architecture, UI, branding, presentation, and spatial experience.

Conclusion

This paper proposes a reversible model connecting creation and perception through an intermediate cognitive State.

The Creator transforms Meaning into Expression by designing State. The Viewer transforms Expression into personal Meaning by entering a State.

Although the present model remains a conceptual hypothesis, it provides a possible common framework linking structural science, cognitive perception, creative expression, and artificial intelligence.