When you deposit a paper on Figshare with sixty DOIs embedded in it — linking to the sixty Core objects of the Socioplastics corpus — and that paper surfaces quickly and visibly in Google searches, while the Zenodo records of those same sixty objects remain essentially invisible, you are not seeing a quality difference between platforms. You are seeing two completely different indexing logics at work. Figshare is designed to be crawled. Google Scholar has a direct harvesting relationship with it. The moment a Figshare record is published, it gets a landing page with clean HTML metadata, an open-access PDF, and a DOI that resolves to a publicly readable object. Google's academic crawlers recognise this pattern immediately because Figshare pages are structurally identical to what Google expects a citable scholarly object to look like: title, author, abstract, DOI, download link, all in one clean page. It also helps that Figshare has very high domain authority — it is trusted infrastructure in Google's model.
But here is the critical mechanism that explains the specific technique: the Figshare paper contains sixty DOIs in its body text, Google does not just index the paper. It follows those DOIs as links. Each one becomes a crawlable signal pointing outward from a high-authority, well-structured document. The paper is functioning as a citational hub — a single indexed object that radiates sixty connections simultaneously. Google reads this as a document of significant relational weight. It is not just a text; it is a node in a larger network, and the network is visible in the document itself. This is structurally identical to what CamelTags do at the lexical level — but here it operates at the citation level. The DOIs are hyperlinks with scholarly authority baked in. Sixty of them in one document is not clutter; it is a density signal.
Why Zenodo stays invisible
Zenodo is CERN infrastructure. It is built for long-term preservation, not for fast surface-level discoverability. Its landing pages are valid and structured, but they are less aggressively optimised for Google's crawlers than Figshare's. More importantly, Zenodo records are often deposited without accompanying full-text PDFs that Google can read — or with PDFs that are uploaded but not indexed as readable documents. Google Scholar can struggle to distinguish a Zenodo PDF deposit from a data file. The metadata is there, but the signal it sends to Google's crawlers is quieter. There is also a community effect. Figshare has been actively courting academic search engine indexing for years. Zenodo's priority is institutional reliability and open science compliance — it is the preservation layer, not the discovery layer. Both are important, but they are doing different jobs.
The technique in plain terms
What has been built, then, is a two-layer architecture: The sixty Core objects live on Zenodo as the preservation layer — sealed, versioned, DOI-anchored, stable. They are the hardened nucleus. They do not need to be Googleable because their job is not to be found by casual search. Their job is to exist permanently, to be citable, and to give every paper that references them a stable address to point at. The Figshare paper is the surface layer — the single, highly legible, Google-crawlable document that makes the whole network visible at once. It pulls the sixty Zenodo DOIs into its own body, and in doing so it creates a high-authority document that Google finds quickly and that radiates sixty stable connections outward. It is simultaneously an introduction to the corpus and a navigation device for machines. The Zenodo records benefit from this indirectly. When the Figshare paper surfaces in Google and readers follow the embedded DOIs, each Zenodo landing page receives traffic from a trusted referrer. Google updates its understanding of those pages accordingly — slowly, but it does.
Why this matters structurally for Socioplastics
This is not just a platform trick. It is a precise enactment of the ThresholdClosure and DistributedInscription logic that Tome III theorises. The Zenodo records are the closed, stable, hardened layer — the part of the corpus that does not move. The Figshare paper is the open, legible, public-facing layer — the part that makes contact with the external detection systems (Google, Google Scholar, institutional crawlers) that decide what gets found. The technique works because it keeps those two functions separate and gives each the right tool. Preservation on Zenodo. Discovery on Figshare. The sixty DOIs in the Figshare body are the connective tissue — the physical links that make the two layers one field rather than two separate deposits.
What we have essentially built is a citational infrastructure that mimics the structure of the corpus itself: a hardened nucleus (Zenodo, invisible but permanent) surrounded by a legible, navigable surface (Figshare paper, fast and visible) that makes the nucleus accessible without destabilising it.
The Soft Ontology Papers do the same thing at the essay level. The Figshare paper does it at the indexing level. The technique is the same technique, applied twice, at different scales.
