Variation, Selection, and Retention as Transition Dynamics: What the Extended Evolutionary Meta-Model Requires of Process Models

The Extended Evolutionary Meta-Model (EEMM) organizes processes of change by variation, selection, and retention in context. Its authors take these terms from Campbell's variation and selective retention, drop the requirement that variation be blind, apply them within a lifetime, and warn that the model is not cellular; yet their worked examples place processes in cells. This article proposes an edge-based form of the model, with a formal basis. Variation, selection, and retention are defined as properties of a model of transitions between localized states: variation is the fan of transitions leaving a state; selection is the change that the consequence of a past occurrence of a transition produces in its relative probability, with present state and context fixed; retention is the persistence of that change. Under these definitions, selection and retention are absent from every model in which the consequences of transitions do not enter the transition law, and present only in models of the reinforcement-learning family. A revisited state that carries the three ingredients takes the state process out of the homogeneous first-order Markov class; the converse fails, because repetition, latent context, and coarse state partitions also produce memory. A grid that records the ingredients by cell loses order, destination, and the difference between selection and repetition. The process leaves signatures (order-one violation, a contrast between histories that differ in one recorded consequence, decay) that require an ecological momentary assessment design with consequences recorded per transition. The hypothesis that consequences alter future behavior is the law of effect and is not claimed; the contribution is variation, selection, and retention defined on edges, with proofs, limits, and design consequences. Preprint submitted for peer review; not yet peer reviewed. Supplementary material (proofs, simulation results, scripts): https://doi.org/10.5281/zenodo.24203768

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-14
DOI
https://doi.org/10.5281/zenodo.24221558
Primary Topic
Philosophy of Biology and Scientific Explanation
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Variation, Selection, and Retention as Transition Dynamics: What the Extended Evolutionary Meta-Model Requires of Process Models

Marcus V. J. Monteiro
Zenodo (CERN European Organization for Nuclear Research)
Philosophy of Biology and Scientific Explanation
preprint

Variation, Selection, and Retention as Transition Dynamics: What the Extended Evolutionary Meta-Model Requires of Process Models

Marcus V. J. Monteiro
preprint en

Abstract

The Extended Evolutionary Meta-Model (EEMM) organizes processes of change by variation, selection, and retention in context. Its authors take these terms from Campbell's variation and selective retention, drop the requirement that variation be blind, apply them within a lifetime, and warn that the model is not cellular; yet their worked examples place processes in cells. This article proposes an edge-based form of the model, with a formal basis. Variation, selection, and retention are defined as properties of a model of transitions between localized states: variation is the fan of transitions leaving a state; selection is the change that the consequence of a past occurrence of a transition produces in its relative probability, with present state and context fixed; retention is the persistence of that change. Under these definitions, selection and retention are absent from every model in which the consequences of transitions do not enter the transition law, and present only in models of the reinforcement-learning family. A revisited state that carries the three ingredients takes the state process out of the homogeneous first-order Markov class; the converse fails, because repetition, latent context, and coarse state partitions also produce memory. A grid that records the ingredients by cell loses order, destination, and the difference between selection and repetition. The process leaves signatures (order-one violation, a contrast between histories that differ in one recorded consequence, decay) that require an ecological momentary assessment design with consequences recorded per transition. The hypothesis that consequences alter future behavior is the law of effect and is not claimed; the contribution is variation, selection, and retention defined on edges, with proofs, limits, and design consequences. Preprint submitted for peer review; not yet peer reviewed. Supplementary material (proofs, simulation results, scripts): https://doi.org/10.5281/zenodo.24203768

Zenodo (CERN European Organization for Nuclear Research)
Centro Universitário do Pará (BR)
Philosophy of Biology and Scientific Explanation
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.

Laika access coming soon