Pre-linguistic Cognitive Substrate (PCS): A Semantic Decomposition Operational Framework as Executable Specification

Cognitive science has long faced a tension at the terminal level of semantic decomposition:static primitive sets cannot grow from experience, connectionist continuous representationslack discrete interpretability, and existing theories mostly remain at the descriptivelevel, lacking formalized operational specifications available for independent verification.This paper proposes that the underlying representation of language understanding canbe traced back to a set of Pre-linguistic Cognitive Substrate (PCS). This substrate is not afixed primitive set, but a set of fixed casting rules; primitives themselves are discrete hypotheticalhierarchies actively generated from continuous input through differentiation operationsunder the constraint of rules. Specifically, three functionally irreducible pre-linguisticoperations—Difference, Identity, Integration—operate under the partial-order skeleton andpolarity vectors provided by two structural conditions—Order and Direction—constitutingthe L0 core; the dynamic primitive casting protocol, driven by definition-reference frequencyand compressibility constraints, enables the system to continuously expand its expressivepower across historical stages.This paper provides the complete formalized kernel of this framework, a stage-parameterizedderivation system, and a hierarchical ontology, and constructs a semantic decompositionprotocol covering 14 classes of content-word Chinese characters including nouns, verbs, adjectives,quantifiers, and pronouns.To examine the operability of this protocol, .this paper uses a large language model(LLM) as an execution interface, selecting representative cases from the content-word classes,including boundary cases at the system’s expressive limit; the outputs are verified by thethree metrics of structural closure, stage compatibility, and polarity consistency derivedfrom the formalized kernel (§3.6). The demonstration confirms the protocol’s compatibilitydetermination and boundary-rejection behavior, providing preliminary evidence for theframework’s operability. Thereby taking a step toward transforming theoretical cognitivescience into executable specifications.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-09-10
DOI
https://doi.org/10.5281/zenodo.22688578
Primary Topic
Neurobiology of Language and Bilingualism
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
preprint

Pre-linguistic Cognitive Substrate (PCS): A Semantic Decomposition Operational Framework as Executable Specification

Tong Liu, Zhen Cao
Zenodo (CERN European Organization for Nuclear Research)
Neurobiology of Language and Bilingualism
preprint

Pre-linguistic Cognitive Substrate (PCS): A Semantic Decomposition Operational Framework as Executable Specification

Tong Liu, Zhen Cao
preprint en

Abstract

Cognitive science has long faced a tension at the terminal level of semantic decomposition:static primitive sets cannot grow from experience, connectionist continuous representationslack discrete interpretability, and existing theories mostly remain at the descriptivelevel, lacking formalized operational specifications available for independent verification.This paper proposes that the underlying representation of language understanding canbe traced back to a set of Pre-linguistic Cognitive Substrate (PCS). This substrate is not afixed primitive set, but a set of fixed casting rules; primitives themselves are discrete hypotheticalhierarchies actively generated from continuous input through differentiation operationsunder the constraint of rules. Specifically, three functionally irreducible pre-linguisticoperations—Difference, Identity, Integration—operate under the partial-order skeleton andpolarity vectors provided by two structural conditions—Order and Direction—constitutingthe L0 core; the dynamic primitive casting protocol, driven by definition-reference frequencyand compressibility constraints, enables the system to continuously expand its expressivepower across historical stages.This paper provides the complete formalized kernel of this framework, a stage-parameterizedderivation system, and a hierarchical ontology, and constructs a semantic decompositionprotocol covering 14 classes of content-word Chinese characters including nouns, verbs, adjectives,quantifiers, and pronouns.To examine the operability of this protocol, .this paper uses a large language model(LLM) as an execution interface, selecting representative cases from the content-word classes,including boundary cases at the system’s expressive limit; the outputs are verified by thethree metrics of structural closure, stage compatibility, and polarity consistency derivedfrom the formalized kernel (§3.6). The demonstration confirms the protocol’s compatibilitydetermination and boundary-rejection behavior, providing preliminary evidence for theframework’s operability. Thereby taking a step toward transforming theoretical cognitivescience into executable specifications.

Zenodo (CERN European Organization for Nuclear Research)
Neurobiology of Language and Bilingualism
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.