There is a part of every Socioplastics node that most readers skip. It sits below the conceptual body, after the protocol sequence, before the citation line. It contains a slug, a version number, a date, a licence string, a file format designation, a keyword list, and a set of author identifiers. It takes perhaps thirty seconds to read and looks, at first glance, like administrative residue — the bureaucratic wrapper around the real content. This essay argues the opposite: that metadata is not the wrapper around the corpus but one of its primary load-bearing structures, and that understanding why changes how the entire project should be read.

Metadata is data about data. In the context of a textual corpus, it is the structured description of a document that allows systems — human and machine — to locate, identify, classify, retrieve, relate, and cite it. A title is metadata. A date is metadata. A DOI is metadata. A keyword list is metadata. An author identifier is metadata. A version number is metadata. A licence string is metadata. A canonical filename is metadata. None of these elements is the argument. None of them is the theory. None of them is the prose. And yet without them, the argument cannot be found, the theory cannot be cited, and the prose cannot be distinguished from the thousands of other texts deposited in the same repositories on the same day. Metadata is the condition of discoverability, and discoverability is the condition of existence in any practical sense. A text without metadata is a text that exists only for whoever already has it. A text with precise, stable, machine-readable metadata exists for anyone who might ever look.