The same conceptual text can exist as a printed page, PDF, HTML document, repository record, plain-text file, database entry, video transcript or fragment retrieved by an artificial intelligence system, and while these manifestations may be nearly identical at the level of words, they are not identical at the level of operation. A PDF stabilizes pagination, typography and visual composition; HTML enables linking, indexing and continual revision; plain text strips away most visual hierarchy while increasing portability and computational accessibility; structured data turns selected elements into explicit fields and relations; a repository adds metadata, persistence and citation; a social platform places the material within streams governed by recency, recommendation and attention. Medium is therefore not a neutral envelope surrounding semantic content. It determines what can be found, quoted, parsed, linked, ranked, recombined and remembered. This becomes especially visible when concepts circulate independently from their original formats. A heading becomes a search result, a paragraph becomes a retrieval fragment, a keyword becomes a graph node, a quotation circulates without the argumentative sequence that once framed it, and a machine-readable representation may expose relationships that were implicit in prose. None of these transformations necessarily corrupts the concept, but each produces another operational condition of it. FormatAdmission begins at this threshold, where entry into a new medium also means entry into a new regime of possible actions and forms of legibility. The distinction between textual identity and medial identity therefore becomes useful. Two occurrences can reproduce exactly the same sentence while performing very different intellectual functions because one acts as evidence within an argument and another as metadata within an index. This also complicates the idea of duplication. A repository copy may contribute no new argument while significantly increasing durability; an HTML version may improve discoverability; a TXT version may facilitate machine processing; a translated transcript may create access for another linguistic community. Redundancy can therefore be functional rather than merely repetitive. The relevant question is what each manifestation contributes to the conceptual object’s continued life. A distributed scholarly architecture should preserve enough common identity for manifestations to remain connected while allowing their different media to perform different tasks. Knowledge does not simply survive despite transformations of format; often it survives through them. The concept that moves between media acquires new capacities each time it enters another technical environment, making circulation itself part of the architecture of meaning.
Briet, S. (1951) Qu’est-ce que la documentation?
Buckland, M. K. (1991) ‘Information as Thing’, JASIS, 42(5), 351–360.
Drucker, J. (2011) ‘Humanities Approaches to Graphical Display’, Digital Humanities Quarterly, 5(1).
Manovich, L. (2001) The Language of New Media. MIT Press.
McLuhan, M. (1964) Understanding Media. McGraw-Hill.