Persistent identifiers—DOIs registered on Zenodo (for core monographs) and Figshare (for preprints)—function as DOI anchors, temporal locks that fix specific segments (books, conceptual cores, working papers) into citable, non-mutable objects. These are not decorative badges; they are the project’s “conceptual anchors,” “DOI infrastructure anchors,” and “field nodes,” cross-linked via ORCID (0009-0009-9820-3319) into a machine-readable JSON-LD graph that binds Person → Organization (LAPIEZA-LAB) → ResearchProject → Dataset → Book → ScholarlyArticle. The Decalogue of Knowledge Formation formalizes this as ten interlocking components: Glossary, Dataset, DOI, Preprint, Book, Blog, Software, ORCID, CSV, and Links. Independence here is achieved through deliberate non-dependence: no single repository holds the totality, and no external gatekeeper controls the citation graph.
The current distribution enacts this logic with surgical precision. The primary living corpus resides on a decentralized network of Blogspot domains (antolloveras.blogspot.com and satellite blogs), declared explicitly as “the primary distribution node for a sovereign epistemic infrastructure” rather than a personal site—over 2,000 numbered working papers, century packs, and real-time nodes. Hugging Face hosts the live, machine-actionable datasets the structural index of 2,000+ papers), optimized for computational traversal, vector search, and RAG pipelines. Zenodo and Figshare supply the fixed deposits with DOIs, enabling scholarly citation and archival stability. GitHub anchors the software layer (MUSE system). ORCID serves as the persistent authorial spine, linking every layer without requiring institutional affiliation beyond LAPIEZA itself. This is not platform pluralism for its own sake; it is stratified legibility: blogs for human interface and serial continuity, HF for metabolic flow and machine legibility, Zenodo/Figshare for contractive fixation and citability. The result is corpus sovereignty over platform dependency, achieved through “persistent identifiers as anchors” and “metadata tail as sovereign.”
At 2,000 nodes exceeding two million words, the corpus has outgrown any single container. Further independence requires extending the same stratigraphic principle rather than mere replication. The obvious next layers are: (1) IPFS pinning or a self-hosted static mirror (via GitHub Pages or a personal domain) to escape Blogger’s Google dependency entirely while preserving the numerical spine and slugs; (2) Internet Archive as a redundant, crawlable backup for long-term web preservation; (3) HAL or SSRN for additional scholarly circulation in European or social-science contexts, maintaining the same DOI mesh; and (4) a compressed “summary dataset” on Dataverse that indexes only the lexical gravity metrics, recurrence indices, and cross-references—reducing the two-million-word mass to a navigable, FAIR-compliant meta-layer without flattening the living organism. These additions do not dilute the project; they harden its sovereignty. In Socioplastics, independence is not the fantasy of a platformless field but the engineered capacity of a field to survive, remain legible, and continue metabolizing across any combination of repositories—precisely because its anchors (lexical, numerical, DOI, ORCID) are internal, not borrowed. At this scale, the decisive question is no longer “where to store the text” but “how to ensure the structure itself remains the sovereign carrier of its intelligence.”
