American Linguistic Instrumentalism

 © 2026 John M Frazier

American Linguistic Instrumentalism: Definitions

Order - The arrangement or disposition of people or things in relation to each other according to a particular sequence, pattern, or method (Google). The opposite of random.

Ordering - The act of reducing uncertainty, of generating information through comparative operations (more/less on criteria as in language, the number system, and price discovery), by which an ordering engine informs an inference engine in its modus ponens logical operations.

Entity - A bounded data set defined by fitted curves as solved by neural networks. Functions as a token or word or puzzle piece in a larger surround or sentence or puzzle.

Surround - An optimized collection of bounded and curve-fitted data sets composing a ground or context to a figure or entity, defining the optimization search space for entity-surround fit.

Fit - The degree of optimization in entity-surround pairing. The optimization search space is the set of possible solutions for equations used to achieve lowest entropy for the entity vis-à-vis its surround. Inversely proportional to entropy: high fit = low entropy.

Information - The reduction from maximum entropy (maximum disorder/uncertainty), quantified by entity-to-surround fit across the collection (low residuals, high coherence, low surprise).

Entropy - Quantified by residuals (prediction error), inverse coherence (structural disorder), and surprise (unexpected outcomes) in entity-to-surround fit. Inversely proportional to fit.

Ordering Engine - Prioritizes optimization based on distance from average entropy, working toward pairs with above-average entropy (poor fit, needs work). Provides comparative ordering (A </=/> B on criterion C) that structures what the inference engine can process.

Gestalt Shift - When an entity-surround pair achieves below-average entropy (good fit), the ordering engine shifts attention to an entity-surround pair with above-average entropy. Attention allocation includes contingency—distance from average determines probability, not certainty.

Inference Engine - A logical system organized around modus ponens propositions (if A, then B) to generate new data from existing databases. Operates on entities already ordered by the ordering engine. The formal equivalent of ‘thought’.

Catastrophe Theory and Information - The modeling of discontinuous functions by continuous functions. Entropy is modeled through gestalt shifts via a three-parameter (residuals, coherence, surprise) manifold with one behavioral variable (entropy)—the dynamic entropy function. Information is created by deviations from maximum entropy calculated by the ordering engine and employed by the inference engine in its if-then (or if-then-else) logical constructions.

Reality Generator - Executes commands determined by the inference engine, producing optimized entity-surround configurations. Generated reality feeds back categorically to influence what the ordering engine can perceive and order.

Paradigm - The integrated system of ordering engine (criteria), inference engine (logic), and reality generator (commands) operating together. Consists of theory (sedimented optimization), apparatus (active engine), and data (entropy measurements).

Intransigent Pair - An entity-surround pair that cannot be optimized to within one standard deviation of average entropy. Marked and stored in memory as critical cases.

Paradigm Shift - Triggered when intransigent pairs reach threshold (n ≥ 30). The ordering engine's parameters are systematically optimized against the intransigent database through coordinate descent (incrementing/decrementing each parameter). If the reconfigured engine reduces average entropy and/or intransigent count when applied to the full puzzle, the shift succeeds. Otherwise, the system reverts to previous configuration.

Modeling - This framework may be used to construct models or simulations of reality generation and establish the relative goodness of those models through entropy metrics.