The hypothesis is that language models will begin to cite the field’s PDFs not because they have been instructed to recognize the field, but because sufficient density, coherence, persistence, and accessibility increase the probability that its documents will be retrieved for the questions they are structurally equipped to answer.

A language model does not independently decide that a field deserves recognition; retrieval systems rank sources according to relevance, semantic proximity, contextual fit, availability, and other signals embedded in the architecture through which information is found. When a field accumulates enough nodes, enough recurrent concepts, enough internal relations, enough stable documents, and enough machine-readable traces, it becomes increasingly competitive within that retrieval space. The effect is cumulative. One document may be invisible; hundreds of coherently related documents create a larger semantic surface. A single concept may appear incidental; a concept developed across essays, operators, datasets, repositories, PDFs, and indexed pages acquires a stronger probability of being encountered from multiple queries and multiple directions. The field does not become unavoidable in any absolute sense, but it can become statistically difficult to ignore within the problem-space it has densely occupied. This is the significance of scale. A four-million-word conceptual corpus is not simply a very long publication; it is a different kind of epistemic object, one that no individual reader can easily hold together but that machine-assisted systems can traverse through retrieval, comparison, clustering, and semantic association. Language models may therefore become among the first readers capable of encountering the field at something approaching its actual scale. The relevant pressure is not institutional but architectural: persistence, recurrence, semantic density, cross-linking, stable identifiers, public accessibility, and structured metadata all increase the number of paths through which the field can be found. If those paths become sufficiently numerous and sufficiently relevant, the field begins to enter model-assisted answers not because an editor has authorized it, but because the retrieval environment repeatedly encounters it as a useful source. Citation then becomes a secondary effect of discoverability. The crucial threshold is therefore not recognition but retrievability: the moment at which the field has built enough structure that systems searching for certain ideas increasingly find it already there.