MindForm — Social Expression: Manner as a Learned Layer Shaped by How Utterances Land

The Cognitive Patterns node of this architecture is stateless: it transforms an incoming appraisal and stores nothing. This paper documents its output-side twin, Social Expression, which holds state — and argues the asymmetry is forced rather than chosen. Perception is private and generates no feedback, so a lens has nothing to learn from; expression is emitted into a world that answers, and a substrate that can be reinforced must have somewhere to keep the reinforcement. The node maintains a communication manner on four signed dimensions in two layers: a target derived fresh each turn from the character's self-image, self-regard and accumulated need pressure, and a persisted formed style that relaxes toward that target while being operantly shaped between times by reception — the manner actually spoken with is entrenched when answered warmly and extinguished when rebuffed. The gap between the two layers is a learned mannerism. Three commitments define the node. It performs from the self-image rather than the traits, following Goffman, so a character's misapprehension of itself becomes audible rather than merely internal, and composes with the self-image's own lag to produce a character heard as its former self long after it has changed. It separates managed manner from leaked affect, following Ekman and Friesen's two accounts of expressive behaviour: assertion, warmth and energy are learned and reinforced, while strain is tension escaping the manner adopted, fed by chronic need pressure and bounded one-sidedly because leakage is a presence rather than a polarity. And its deterministic path produces an actual reply rather than an approximation of one, so two characters with different histories sound different with no language model in the loop. The node closes the architecture's first exogenous feedback cycle, whose sign is not a property of the system: the same character under a warm and a cold interlocutor runs the same mechanism to different attractors. Every stability question this series has raised assumed the character was the system under test; this node makes that assumption insufficient.