No Center Holds

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.

The Persistence of the Center

Information systems have long been organized around a center. In libraries, this center takes the form of the catalog or the authority file. In archives, it appears as the repository and the hierarchical structure that governs it. In digital environments, it is reproduced through databases, platforms, and coordinating standards. Even systems that describe themselves as distributed tend to retain central points of reference at the level of indexing, schema design, or governance.

This persistence reflects a deeper assumption. Coherence is expected to depend on a stable point of coordination. Information must be anchored, validated, and interpreted in relation to something that does not move.

That assumption has structured the field for decades. It is now increasingly difficult to maintain.

As information environments expand, records are no longer created, described, and preserved within a single institutional frame. They circulate across systems, accumulate layers of intervention, and remain subject to ongoing transformation. Under these conditions, centralization does not simply organize information. It constrains the ways in which coherence can be produced.

Coherence Without Anchor

In decentralized environments, no single node contains a complete representation of the system. There is no master record that stabilizes meaning, no definitive version against which all others can be measured. Records persist across multiple sites, each shaped by different descriptive practices, technical conditions, and forms of use.

Despite this, such systems do not dissolve into incoherence. They remain navigable. They continue to produce meaning.

What changes is the mechanism by which coherence is sustained. It is no longer imposed from a central point. It emerges from the persistence of relations across distributed elements. Links hold where authority does not. Identifiers continue to connect instances that are no longer identical. Interpretation becomes a function of position within a network rather than proximity to an origin.

Under these conditions, informational integrity can no longer be located in a single object. It must be understood as a property of the system as a whole, expressed through the consistency and durability of its internal relations.

Redundancy and Transformation

Conventional models of preservation depend on duplication. Multiple copies of a record are maintained in separate locations in order to ensure survival. The premise is that integrity requires sameness. If the copies remain identical, the record persists.

In distributed environments, this premise does not hold. Records do not remain unchanged as they move. They are reformatted, re-described, enriched, and reinterpreted. Each instance diverges from the others, sometimes subtly, sometimes significantly.

What persists across these transformations is not identity, but relation.

Redundancy, in this context, cannot be understood as the maintenance of identical copies. It must be understood as the distribution of related forms. Each instance carries a partial continuity of the record. Together, they sustain it. Instead of preserving a fixed state, they maintain a field of transformations that remain intelligibly connected.

Loss, under these conditions, does not erase a record. It alters the configuration through which that record continues to exist.

Authority as a Distributed Condition

In centralized systems, authority is tied to institutional position. A record is legitimate insofar as it remains within the structures that produce and recognize it. Authority does not emerge from use. It is imposed and maintained.

Distributed environments complicate this model. Records circulate beyond the boundaries of their originating institutions. They circulate across multiple contexts and are altered as they move. Origin remains relevant, but it no longer stabilizes meaning.

In such contexts, authority is distributed. It cannot be located in a single source. It takes form across repeated use and through its integration into multiple systems of interpretation. Institutional authority persists, but no longer as an exclusive source of legitimacy. It operates within a broader field in which legitimacy is continuously formed through interaction.

Provenance and the Problem of Linearity

Archival theory has long relied on the concept of provenance as a means of securing contextual integrity. The record is anchored to its origin, and that origin provides the framework for its interpretation.

This model presupposes stability. It assumes that records remain sufficiently bounded for their origin to function as a primary reference.

In distributed environments, that condition is increasingly rare. Records emerge from multiple contexts, move across systems, and accumulate layers of intervention that cannot be reduced to a single lineage. Their histories are not sequential but overlapping.

To describe such records through a linear model of provenance is to impose an order that no longer corresponds to their mode of existence. Provenance must instead be understood as distributed, constituted by the set of relations through which a record has been formed, modified, and sustained.

The problem is not simply descriptive. It is structural. Existing frameworks are not designed to represent this condition adequately.

Failure and Reconfiguration

Centralized systems are vulnerable in specific ways. When the structures that coordinate them fail, the consequences propagate quickly. A compromised repository, a lost catalog, or a broken infrastructure can disrupt access to entire bodies of information.

Distributed systems exhibit a different form of vulnerability. Failures occur, but they tend to remain localized. Nodes disappear, connections weaken, pathways are interrupted. The system does not remain unchanged, but neither does it collapse entirely.

It reorganizes.

Continuity, in such systems, depends less on the preservation of particular components than on the persistence of relational density. As long as sufficient connections remain, the system retains the capacity to reconfigure itself.

The Limits of Centralization

Centralized architectures offer clear advantages. They facilitate control, enforce consistency, and provide identifiable points of responsibility. These qualities have been essential to the historical development of memory institutions.

At the same time, they impose constraints. They privilege uniformity over variation, stability over transformation, and control over adaptability. As information environments become more complex and more distributed, these constraints become increasingly visible.

The issue is not whether centralization should be abandoned. It is whether it can continue to function as the primary organizing principle under conditions that no longer support its assumptions.

Memory Without a Fixed Site

If coherence can be sustained without a center, then memory need not be anchored in a single location. It can persist across distributed systems, maintained through the continuity of relations rather than the stability of objects.

This does not eliminate structure. It changes its locus.

Structure is no longer imposed from above through centralized coordination. It is maintained through interaction, through the ongoing alignment and realignment of distributed elements.

Under such conditions, memory is not something that resides in a place. It is something that persists in a system.

Conclusion

The centrality of the archive, the catalog, and the repository has long defined how information systems are imagined. These structures have provided coherence by anchoring knowledge in stable, authoritative forms.

But they are not the only way coherence can be achieved.

Distributed environments demonstrate that continuity can emerge without central coordination, that integrity can be maintained across transformation, and that memory can persist without being fixed in a single site.

The question is not whether such systems can exist. They already do.

The question is whether the conceptual frameworks of Library and Information Science are prepared to account for them.

  This entry mirrors the chronicle "There is No Central Server," a narrative reflection on the same theme.