Stratigraphic Time and the Architecture of Information Systems

This post is part of a series that explores how metadata can be used as a site of resistance, refusal, and poetic subversion. From classification to linked data, the series investigates how cataloging practices can encode oppression, and how they can be reimagined to challenge dominant systems and speak from the margins. Check all the posts in this section's index.

Linear Time as Infrastructure

Library and Information Science has historically operationalized time as sequence. Bibliographic description privileges publication date; archival arrangement often follows chronological order within fonds; digital preservation frameworks record events in ordered logs; version control systems increment states numerically. These conventions are not merely practical choices. They encode an ontological commitment: that temporality is external to the record and can be mapped as a linear progression of discrete moments.

This commitment is embedded in standards and protocols. MARC fields such as 260 and 264 situate works within temporal coordinates that support sorting and authority control. Dublin Core's date element functions as a scalar marker, reducible to machine-readable formats that enable chronological filtering. In preservation environments, PREMIS events document actions performed on digital objects in timestamped sequences. Even in semantic web architectures, temporal modeling frequently relies on ordered intervals, instants, and successor relations.

Such infrastructures assume that a record can be separated from its context, stabilized, and placed along a timeline without ontological distortion. Time is treated as metadata — an attribute appended to an otherwise self-contained entity. The persistence of this model reflects the administrative origins of archival practice and the bibliographic tradition's reliance on publication chronology as an organizing principle.

Versioning and the Logic of Succession

Digital environments have intensified this linear model through formalized versioning systems. In repository architectures, objects acquire identifiers that persist across updates while new versions are enumerated. Software-based systems such as Git articulate branching and merging, but even these more complex structures retain an underlying assumption of traceable succession. Each state derives from a previous one; the history of the object is represented as a path through a graph whose edges imply temporal order.

In archival metadata, similar logic governs the recording of migrations, format shifts, and fixity checks. A digital object exists in state A, undergoes an event at time t1, transitions to state B, and so on. Integrity is demonstrated through the ability to reconstruct this sequence. The epistemic weight lies in the chain: authenticity is preserved by proving continuity across ordered transformations.

This procedural understanding of time privileges traceability over stratification. Earlier states are preserved as recoverable instances, but they are conceptually superseded. Version 3 replaces version 2 in active use; prior states become archival residues accessible on demand. Temporality is conceived as forward movement through discrete iterations.

The stratigraphic model suggested by subsoil systems complicates this assumption. If temporal experience is embodied as spatial modification rather than as sequential replacement, then versioning might be reconsidered not as succession but as compression. A document's history of transformation could be understood as sedimented within its current configuration rather than externalized as a list of prior states. In such a model, earlier forms are not merely retrievable; they are constitutive of the present.

Provenance Beyond Singular Lineage

Traditional archival theory has treated provenance as the cornerstone of contextual integrity. Records are linked to a creator, and that linkage establishes both authority and interpretive frame. While post-custodial paradigms have expanded this view to acknowledge distributed custody and networked environments, provenance is still frequently represented as a lineage, even when multiple agents are involved.

Digital documentation complicates this model. Records circulate across platforms, are enriched by successive layers of metadata, and are transformed by aggregation, digitization, and reuse. The resulting objects are polyprovenancial in practice, even if descriptive frameworks struggle to articulate that condition. Temporal modeling remains largely additive: new contributors are appended as additional agents in a sequence of events.

A stratigraphic approach to provenance would emphasize overlap rather than succession. Instead of conceiving of creators and custodians as occupying discrete positions in time, it would recognize that their contributions coexist within the present form of the record. Metadata created during digitization does not merely follow the original inscription; it reshapes how that inscription functions. User annotations, algorithmic enrichments, and subsequent curatorial interventions are not external commentaries but structural modifications.

Representing such layered authorship requires moving beyond linear chains toward topological mappings of relation. Temporal depth would be expressed not only through dates but through relational density: the extent to which a record has been recontextualized, translated, or integrated into new systems. Provenance becomes less a question of origin and more a question of entanglement.

Temporal Ontologies and Semantic Modeling

Semantic web technologies have introduced more flexible temporal representations through ontologies capable of expressing intervals, durations, and event relations. OWL-Time and related vocabularies allow the modeling of instants, recurring events, and temporal containment. Yet even these frameworks often retain a fundamentally chronological orientation. They describe when something occurred and how events relate in ordered fashion, but they rarely capture how prior states persist structurally within subsequent ones.

To incorporate stratigraphic temporality, semantic models would need to express transformation as spatial reconfiguration rather than mere succession. Predicates such as "was transformed into" or "incorporates trace of" could complement standard temporal relations. The goal would not be to abandon chronological description but to enrich it with a vocabulary capable of representing compression, recurrence, and layered persistence.

Such modeling has implications for preservation metadata. Instead of documenting each migration solely as a timestamped event, systems could encode the ways in which prior formats continue to shape current accessibility, interpretation, or functionality. Deprecated standards and obsolete schemas would not appear merely as historical notes but as active substrata influencing present configurations.

This shift would align temporal modeling more closely with ecological notions of continuity. In living systems, earlier conditions do not vanish when new ones emerge; they modify the substrate within which change occurs. Translating this insight into LIS infrastructure requires acknowledging that information systems, too, are environments shaped by accumulated interventions.

Designing for Temporal Depth

Reconceptualizing time as depth rather than sequence challenges user interface design as well. Most discovery systems present records along timelines or sort results by date. While effective for certain queries, such interfaces reinforce the notion that relevance is tied to recency or historical placement along a line.

Interfaces that visualize relational density, layers of annotation, or degrees of contextual embedding could offer alternative ways of navigating temporal complexity. Rather than presenting a document as a point on a timeline, systems might render it as a node within a stratified network, where earlier and later contributions coexist visibly.

This approach does not reject chronological information. It situates it within a broader temporal ecology. Date remains important, but it ceases to monopolize the representation of time. The emphasis shifts toward understanding how records accumulate modifications and how those modifications condition present meaning.

Moving Toward a Stratigraphic Information Science

Adopting a stratigraphic model of time within LIS does not entail abandoning existing standards. Chronology remains indispensable for legal, administrative, and historical purposes. However, treating it as the exclusive or primary mode of temporal representation limits the field's capacity to model complex transformations.

The subsoil metaphor of roots storing time spatially invites a reconsideration of foundational assumptions. Records may be better understood as configurations shaped by layered interventions rather than as sequences of replaceable states. Versioning becomes an expression of accumulated change; provenance becomes a topology of relations; preservation becomes the management of temporal depth.

Such a reorientation would not romanticize ecological processes nor impose natural analogies uncritically. It would, instead, recognize that information systems, like ecological systems, are structured by the ways they incorporate their own histories. To design infrastructures attentive to temporal depth is to acknowledge that memory persists not only by moving forward along a line, but by folding past transformations into the architecture of the present.

  This entry mirrors the chronicle "Where the Roots Keep Time," a narrative reflection on the same theme.