Skip to content
Back to search results

Infrastructure Studies

Introduced by
Susan Leigh Star, Geoffrey Bowker, Karen Ruhleder
Year
1996
Type
Qualitative method for studying relational infrastructure
Family
Infrastructural inversion
Font size

What does it do?

The infrastructure studies method, developed by Susan Leigh Star, Geoffrey Bowker, and Karen Ruhleder, studies the normally invisible relational systems—standards, classifications, pipes, protocols, and installed bases—on which modern life silently depends.

Its signature move is 'infrastructural inversion': deliberately foregrounding the background, treating the taken-for-granted substrate as the object of analysis, and reading its standards, classifications, and breakdowns to understand how it shapes work, knowledge, and lives.

The intuition behind the method

Infrastructure is what you notice only when it fails. The electricity, the plumbing, the classification scheme behind a hospital record—these recede into the background precisely because they work. Star's insight is that this invisibility is not a problem to be ignored but the very thing to study. To do so you flip the picture: instead of looking at the people and work in the foreground, you make the substrate the figure. You watch what happens when it breaks, you read the standards and categories that organise it, and you ask whose work it eases and whose it strands. The boring becomes the most revealing.

History

The method emerged from Susan Leigh Star's work in the 1990s, building on her earlier concept of the boundary object (Star & Griesemer, 1989). Star and Ruhleder's 1996 paper 'Steps toward an ecology of infrastructure' reframed infrastructure as relational and proposed studying it through breakdown and the installed base. Bowker and Star's Sorting Things Out (1999) extended the programme to classification, showing how categorisation systems—from the International Classification of Diseases to apartheid race categories—silently order work, identity, and life. The approach seeded the interdisciplinary field of infrastructure studies spanning STS, information science, and organisation studies.

When to use it

Use the infrastructure studies method when the object of interest is a large, embedded, relational system—an information standard, a classification scheme, a network, a platform—whose workings are normally invisible and taken for granted. It suits research on how categories and standards organise work and knowledge, on whose labour and whose existence a system renders visible or invisible, and on the politics buried in apparently neutral technical arrangements. It assumes that infrastructure is relational and best studied through breakdown, ethnography, and historical tracing. It is less appropriate when you need a controlled comparison, a quantified outcome, or when the system in question is novel enough to lack an installed base or settled standards.

Limitations and considerations

  • The interpretive, ethnographic approach resists generalisation and quantified comparison across many infrastructures.
  • Studying invisible systems demands deep, time-consuming access to sites, archives, and practitioners that is hard to obtain.
  • Identifying what counts as 'the' infrastructure is itself a relational judgement, leaving the unit of analysis open to dispute.
  • Rich description of one infrastructure can be difficult to translate into actionable design or policy recommendations.

Source conditions do not replace independent scientific review or validation by an analysis engine.

Strengths

  • Makes visible the powerful but normally invisible substrate—standards, classifications, installed bases—that other methods overlook entirely.
  • Reveals the politics and morality of apparently neutral technical categories, showing whom they privilege and whom they erase.
  • Provides durable concepts—boundary objects, installed base, infrastructural inversion, invisible work—that travel across domains.
  • Uses breakdown and friction as a disciplined empirical entry point into systems that are otherwise transparent in use.

Applications

  • Analysing medical classification systems such as the ICD and nursing classifications and their effects on care and record-keeping.
  • Studying the development and consequences of information standards and metadata schemes in libraries, science, and government.
  • Investigating large data and computing infrastructures—cyberinfrastructure, cloud platforms, scientific databases—as relational systems.
  • Examining how the invisible work of maintenance, repair, and data cleaning sustains apparently automatic systems.
  • Tracing how identity and demographic categories built into administrative infrastructures privilege and marginalise populations.

How it works

How it works

1

Select an infrastructure and recognise it by its relational properties rather than as a thing. Star and Ruhleder argue that infrastructure is not a substance but a relation: something becomes infrastructure for particular people, embedded in practice, transparent in use, reaching beyond a single site, learned as part of membership, and built on an installed base. The analyst chooses a system—an information standard, a classification, a network—and studies when and for whom it is infrastructure.

2

Perform the infrastructural inversion: deliberately foreground what normally recedes into the background. Following Bowker and Star, the analyst treats the substrate—the categories, the standards, the wiring—as the central object rather than as inert support, surfacing the design decisions, politics, and labour that the smooth surface conceals.

3

Study breakdowns, friction, and workarounds, because infrastructure becomes visible upon breakdown. Moments when the system fails, when users improvise, or when newcomers struggle to learn it expose its hidden assumptions and the work required to keep it running, giving the analyst empirical access to an otherwise transparent substrate.

4

Trace the classifications and standards built into the infrastructure and their consequences. In Sorting Things Out, Bowker and Star show how categories—of disease, race, occupation—are moral and political as well as technical, privileging some entities and erasing others. The analyst follows how a classification was made, what it includes and excludes, and how it travels and sticks.

5

Analyse the infrastructure's relational properties and effects: how it embeds the assumptions of its designers, how it distributes visibility and invisibility of work, how boundary objects let different communities cooperate without consensus, and how the installed base constrains change. The analysis links the technical substrate to the social order it quietly sustains.

Common pitfalls

  • Reifying infrastructure as a fixed thing rather than studying it as a relation that holds for particular people in particular practices.
  • Treating standards and classifications as neutral and technical, missing the moral and political choices they encode.
  • Overlooking invisible work—the maintenance, care, and articulation labour—that keeps the infrastructure running beneath notice.
  • Studying the system only when it works smoothly and thereby forfeiting the analytic leverage that breakdown provides.

Frequently asked questions

What exactly is 'infrastructural inversion'?

Infrastructural inversion, a term Bowker drew on and developed with Star, is the analytic move of foregrounding the background—treating the normally invisible substrate of standards, classifications, and technical arrangements as the central object of study rather than as inert support for human action. It reverses the usual figure-ground relation, so that the boring, taken-for-granted infrastructure becomes the thing you most want to explain. It is the method's defining gesture.

Why study infrastructure through breakdown?

Because infrastructure is, by definition, transparent in use: it works best when you do not notice it. That very invisibility makes it hard to study directly. Breakdowns, friction, and the struggles of newcomers learning the system make the hidden assumptions and the sustaining labour suddenly visible. Star treated breakdown not as a nuisance but as the analyst's privileged window onto a substrate that otherwise refuses to appear.

What is a boundary object and how does it fit in?

A boundary object is an entity—a form, a specimen, a classification—plastic enough to adapt to the local needs of different communities yet robust enough to maintain a common identity across them, letting groups cooperate without full consensus. Star introduced the concept to explain coordination across social worlds, and it is central to infrastructure studies because standards and classifications often function as boundary objects, holding diverse practices together while meaning different things to each.

Related methods

Sources

  1. Star, S. L., & Ruhleder, K. (1996). Steps toward an ecology of infrastructure: design and access for large information spaces. Information Systems Research, 7(1), 111-134.

    Open DOI reference
    Copied.
  2. Bowker, G. C., & Star, S. L. (1999). Sorting Things Out: Classification and Its Consequences. MIT Press.

    ISBN: 9780262522953
    Copied.