Consciousness as Process, Not Agent¶

Consciousness is not an entity that causes things -- it is a process performed by the substrate, and its functional role is explained by the dual evaluation architecture without requiring independent causal power.

A persistent source of confusion in consciousness studies is the treatment of consciousness as an "it" -- an agent that either does or does not cause things. This framing generates false dilemmas: either consciousness causes behavior (which seems to require mysterious downward causation) or it does not (which makes it epiphenomenal and apparently pointless). The Four-Model Theory dissolves this dilemma by rejecting its premise.

The Self-Reading Thermostat¶

Consciousness is not a thing; it is a process performed by the substrate. Asking whether consciousness "causes" anything is a category error -- analogous to asking whether the display of a self-reading thermostat causes the heating.

The furnace generates heat; the thermostat's sensor reads the temperature; but the display is also read by the system itself to calibrate further action. The display does not cause the heating in the way the furnace does, yet it is constitutive of the feedback loop that makes the system a thermostat rather than an unregulated heater.

Consciousness occupies the same structural position: not a separate cause but a constitutive element of the feedback architecture. The implicit models generate the virtual simulation for concrete adaptive reasons. The EWM integrates multimodal sensory data into a unified scene. The ESM provides a self-model against which consequences can be evaluated. Qualia, as constitutive elements of that simulation, lack independent causal power over the substrate, yet they are not causally inert: as the self-model's evaluative representations they do genuine causal work at the computational level, in the sense in which a spreadsheet's "sum" does. Remove the simulation and the substrate still processes, but it loses its cheap and general route to consequence-evaluation.

The Dual Evaluation Architecture¶

The relationship between substrate and simulation is not epiphenomenal accompaniment. It is best understood as a dual evaluation architecture with two-way traffic.

Primary direction: The implicit system actively deploys the virtual simulation as its evaluation mechanism. It presents decisions, actions, and their consequences to the simulation so that it can assess outcomes, run scenarios, and register hedonic valence. This is the substrate's mechanism for consequence-observation -- the very thing natural selection shaped the architecture to do.

Secondary direction: The explicit models evaluate independently, albeit with significantly less computational bandwidth. The conscious simulation operates at a limited frame rate -- on the order of 20 Hz at its upper, trained end, with the perceptual sampling literature placing typical rates nearer 7–13 Hz -- while the substrate processes at vastly higher throughput. Over time, these conscious evaluations -- the explicit system's independent assessments of situations, actions, and outcomes -- feed back to reshape the implicit models through learning.

The result is continuous feedback: the implicit system uses the explicit as an evaluation tool (primary), and the explicit system contributes its own assessments back (secondary), shaping the substrate that generates it.

Not Epiphenomenalism¶

This makes the theory's position distinct from classical epiphenomenalism, in which consciousness is a causally inert by-product with no functional role. In the Four-Model Theory, the virtual models are in continuous feedback with the implicit models: the simulation's outputs feed back to update implicit processing, shaping future behavior. The virtual simulation has a genuine functional role -- consequence-observation and future-oriented adaptation -- without requiring independent causal power.

Figure¶

graph LR
    subgraph IMPL["Substrate (Implicit Models)"]
        IWM["IWM"]
        ISM["ISM"]
    end

    subgraph EXPL["Simulation (Explicit Models)"]
        EWM["EWM<br/>Scene integration"]
        ESM["ESM<br/>Consequence evaluation"]
    end

    IMPL -->|"deploys for evaluation<br/>(primary)"| EXPL
    EXPL -->|"reshapes through learning<br/>(secondary)"| IMPL

    style IMPL fill:#1a1a2e,stroke:#333,color:#aaa
    style EXPL fill:#2d1b69,stroke:#9b59b6,color:#fff
    style IWM fill:#2c3e50,stroke:#34495e,color:#bbb
    style ISM fill:#2c3e50,stroke:#34495e,color:#bbb
    style EWM fill:#8e44ad,stroke:#9b59b6,color:#fff
    style ESM fill:#8e44ad,stroke:#9b59b6,color:#fff

The dual evaluation architecture. The substrate deploys the simulation for consequence-evaluation (primary direction, left to right). The simulation's assessments feed back to reshape the substrate through learning (secondary direction, right to left). Two-way traffic, not epiphenomenal accompaniment.

Key Takeaway¶

Consciousness is a process, not an agent. It does not "cause" behavior the way a furnace causes heat -- but a thermostat is not a thermostat without the display it reads, and the brain's adaptive architecture is not adaptive without the simulation it runs.