The land, energy, water, and labor behind retained data

A retained file needs a place to persist and people to keep it usable. Its storage can feel remote from the desk where a reader opens it, but the ongoing work remains situated in equipment, power, buildings, and institutional decisions.

Mél Hogan’s Data Flows and Water Woes examines that material relation through the Utah Data Center. Published in 2015, the article joins surveillance, resource dependence, and the politics of location.

For librarians, the question concerns how preservation is described and governed. What resources and responsibilities sustain a collection’s continuing availability, and which parts of that account can the institution actually verify?

Retention takes place somewhere

Hogan’s study focuses on the data center’s architecture, location, cooling needs, and relation to surveillance. Water is central to her analysis because the facility’s operation depends on material conditions that an account of data as abstract information can leave out.

Her argument links privacy and surveillance to the infrastructure that makes collection and retention possible. It does not provide unrestricted access to intelligence operations or a current inventory of everything held at the site. Public evidence and its limits remain part of the study.

The Atlas interprets the case as a relation between persistent records and persistent material support. Data can remain operative beyond the encounter that produced it, while the infrastructure holding it continues to draw on situated resources.

I find the difference between visibility and dependence useful. A person may be unable to inspect a stored record or determine its future use while the facility supporting that use has consequences for a particular landscape. The missing public account does not make the material operation disappear.

That distinction also qualifies a tempting calculation. A historical site’s reported consumption cannot simply be divided by an assumed number of files to produce the environmental cost of an arbitrary library object. The allocation would require evidence about the system, workload, time, and resources being compared.

Hogan’s case therefore supports an inquiry into infrastructure, because a rigorous discussion needs to establish what a claim about storage includes, whose resource use it describes, and which public records support it before deciding what the number means.

The location also matters politically. Storage is governed through land, facilities, resource arrangements, and institutions whose decisions may be separated from the people represented in the data. Treating the infrastructure as an unlimited background would remove those relations from the account.

Preservation is a continuing service

A library’s preservation system has a different purpose and governance. Its copies may protect against loss, enable recovery, or support access in different places. Those functions deserve an explicit account before any recommendation about retention or reduction.

The NDSA Levels of Digital Preservation provide a professional framework for examining preservation practice. Attention to integrity, storage, control, and metadata supports the collection’s continuing use. The framework does not make every additional copy necessary or supply an environmental figure for a particular service.

I would begin with one bounded collection and document the role of each controlled copy, including which copy supports recovery, which supports access, and who is responsible for checking that those functions remain available when another part of the system fails.

That inventory should distinguish material the institution intends to preserve from disposable working outputs. The distinction requires curatorial and operational judgment. A temporary derivative and a preservation source may look similar in a directory while serving very different purposes.

The review also needs the supplier’s actual commitments. Where storage is external, identify which location or resource information is disclosed, which operations are included in the service, and what the library can obtain as evidence. An unknown should remain an unknown rather than becoming an invented sustainability claim.

Compare choices through the service they sustain

Suppose a team proposes changing a storage arrangement. The comparison should include the preservation and access functions each option can support, together with the verified resource information available for the relevant period.

A lower storage total alone cannot establish an improvement if the change removes a necessary recovery function. Conversely, the importance of preservation does not automatically justify retaining every intermediate file indefinitely. The institution needs to explain the purpose of the material and the means required to sustain it.

I would ask the responsible staff to identify the smallest reliable arrangement that meets the collection’s declared requirements, then compare that arrangement with the current one. Any claimed saving should state what was measured, what was estimated, and which assumption governs the difference.

The Library of Congress recommended formats guidance can inform decisions about the formats retained for continuing access. It is one source of preservation advice, rather than an instruction to discard a collection’s meaningful material features in favor of a smaller file.

Hogan’s article keeps the wider question visible. Information stewardship involves a continuing material commitment whose costs and authority are located somewhere. The library’s account should be able to connect that commitment with the public purpose it supports.

At the next collection review, retain the storage roles, responsible parties, and evidence for any proposed change alongside the preservation requirements. A future reader’s access should be supported by an arrangement the institution can describe, rather than by an unexplained accumulation of copies.