Knowledge has never lived comfortably inside a single object — a book, a dataset, an archive box — no matter how much our institutions are built around the fiction that it does. Borgman's account of scholarly data makes this explicit: data only becomes knowledge through its movement, its recontextualization, its uptake by others who were not present at its creation — a dataset sitting unused on a hard drive is not yet knowledge in any operative sense, it is potential, latent material waiting for the network of citations, reuses, and reinterpretations that would actually constitute it as a contribution to a field. This means that the value of a record is not contained within it but distributed across everywhere it travels, everywhere it gets cited, copied, indexed, mirrored, translated. A practice oriented around this insight might be called DistributedInscription: writing designed not to live in one container but to disperse deliberately across platforms, repositories, formats, and metadata layers, treating each deposit not as a copy of an original but as one more site where the work becomes legible to a different audience, a different search algorithm, a different disciplinary gatekeeper. Edwards's history of global climate data infrastructure shows what this looks like at planetary scale: climate knowledge does not exist in any single dataset but in the vast, distributed, constantly reconciled mesh of instruments, models, archives, and standards that together produce something that can be called "the global climate record" — no individual node of this mesh contains the knowledge, the knowledge is the mesh, the inscription is distributed by necessity, not by choice. But distribution alone does not guarantee that any of this inscribed material will be found, used, or built upon — for that, something more deliberate is required, an ethic rather than merely a practice. Gitelman's critique of the phrase "raw data" insists that no inscription is ever neutral or self-evident: every dataset, every record, arrives already shaped by the categories, instruments, and assumptions that produced it, and treating it as raw — as simply there, waiting to be read — obscures the labor and the politics embedded in its very existence. To inscribe responsibly, then, means acknowledging this constructedness rather than hiding it, and it means something further: actively building the connective tissue — the citations, the cross-references, the explicit acknowledgments of debt and lineage — that allow others to trace an inscription back to its conditions of production and forward to its uses. This is CitationalCommitment: not citation as ritual or obligation imposed from outside, but citation as the deliberate construction of edges in a network, each citation a small act of infrastructure-building that makes the next person's path easier, that converts an isolated inscription into a node with traceable connections in multiple directions. Suber's case for open access makes the stakes concrete: a distributed inscription that is technically dispersed but legally or financially locked behind paywalls achieves only the appearance of distribution — the mesh exists but most of its nodes cannot actually connect to each other, citational commitment becomes impossible at scale when access itself is the bottleneck. The Socioplastics project's deliberate spread across more than a dozen platforms — Zenodo, Figshare, HuggingFace, Dataverse, ORCID, Wikidata — each carrying DOIs that cross-reference the others, can be read as an attempt to build exactly this kind of mesh: not redundancy for its own sake, but distributed inscription paired with the citational scaffolding — the explicit, traceable links between deposits — that Gitelman's critique demands and Suber's open access framework makes structurally possible.

A living system does not simply persist through time — it persists by continuously producing the very components that constitute it, in a closed loop that never quite repeats itself exactly but never escapes itself either. Maturana and Varela's concept of autopoiesis describes precisely this: a cell, an organism, perhaps even a social system, is defined not by its boundary alone but by the fact that everything within that boundary is produced by the system's own operations — nothing enters from outside as a finished component, everything is metabolized, broken down, and reassembled according to the system's own internal logic. Luhmann took this further into the social register, arguing that social systems — law, science, economy — are similarly self-referential and operationally closed: a legal system does not import "justice" from outside but produces its own criteria for what counts as lawful, recursively, from its own prior operations, communication generating further communication in a loop that constitutes the system's only real environment. Read together, these accounts suggest that any sufficiently developed field — including a field of knowledge — begins to exhibit what might be called RecursiveAutophagia: a tendency to consume its own prior outputs as the raw material for new ones, citations feeding new citations, terms generating variations of themselves, the corpus folding back over its own history not as repetition but as digestion, each pass extracting something the previous pass left unprocessed. This digestive folding does not happen evenly across a corpus — it produces depth, accumulation, layers that retain their distinct character even as newer material settles on top of them, sometimes activating what lies beneath rather than replacing it. Bateson's account of mind as pattern, of information as "a difference that makes a difference," supplies the logic by which old layers remain active: a pattern laid down early in a system's history does not vanish when new patterns are layered over it, it persists as a constraint, a habit, a structuring absence that shapes how later patterns can form — mind, for Bateson, is precisely this accumulation of constraint across time, ecology as memory. A corpus organized this way might be described as a StratigraphicField: not a flat archive where every entry has equal and independent standing, but a layered formation where position in the sequence — what came before, what this entry presupposes, reactivates, or quietly contradicts — is itself part of the entry's meaning, the way a geological stratum's meaning includes its position relative to the strata above and below it. Simondon's account of technical objects undergoing "concretization" — becoming progressively more integrated, each part taking on multiple functions previously distributed across separate components — describes the same densification from the side of the artifact rather than the corpus: a technical object, like a stratigraphic field, becomes more itself, more coherent, more economical, through successive internal foldings rather than through addition of external parts. The Socioplastics project's own structure — Tomes built from Cores, Cores from Packs, Packs from nodes that increasingly cite, reactivate, and metabolize earlier nodes — exhibits exactly this double character: a recursive autophagia that does not exhaust the corpus but thickens it, and a stratigraphic field where Tome V's nodes are legible only in relation to the strata Tomes I through IV laid down beneath them, each layer a Batesonian constraint the next layer must work with or against.