The family resemblance of technologically mediated work practices

Practice-based perspectives in information systems have established how, in every instance of use i.e., work practices, the user exercises considerable discretion in their appropriation of the technology with local workarounds and situated improvisations. We analyse the relationship between technologically mediated work practices separated in time and space. Specifically, we analyse how similarity in work practices is achieved. Achieving absolutely similar or best practices is unattainable. Drawing on a longitudinal 2007 to 2011 case of ambulatory maintenance work in the oil and gas sector, we identify and discuss three constituting strategies called differentiation, assembling and punctuation through which a family resemblance of similar but not the same work practices is crafted. We discuss how, in the absence of an essentialist criterion, similarity is subject to pragmatic but also political negotiations.

💡 Research Summary

The paper investigates how work practices that are mediated by technology can appear similar across different times and locations without ever achieving exact uniformity. Drawing on a four‑year longitudinal case study (2007‑2011) of an ambulatory maintenance crew operating on offshore oil‑and‑gas platforms, the authors demonstrate that “family resemblance” – a concept borrowed from Wittgenstein that emphasizes overlapping features rather than strict identity – is the operative notion of similarity in such contexts.

The study begins by situating itself within the practice‑based tradition in Information Systems, which has traditionally focused on how individual users appropriate technology through local work‑arounds and improvisations. The authors argue that this focus leaves a gap: we know little about how disparate sites, each with its own constraints, manage to sustain practices that are recognizably alike. To fill this gap, they adopt a qualitative, multi‑method approach that combines prolonged field observation, semi‑structured interviews with crew members, supervisors, safety officers, and document analysis of work orders, incident reports, and the mobile maintenance management system.

Through iterative coding, three constitutive strategies emerge that together generate a “family resemblance” among practices:

1. Differentiation – Teams consciously adapt standard procedures to the particular physical, regulatory, and organisational characteristics of each platform. For example, a platform with higher pressure ratings may require an additional lock‑out step not present elsewhere. Differentiation acknowledges site‑specific risk profiles and embeds them into the procedural artefacts, thereby preventing a one‑size‑fits‑all approach.

2. Assembling – Past improvisations and locally successful work‑arounds are captured in a shared knowledge base within the mobile system. When a new site is encountered, the crew “assembles” a bespoke workflow by selecting, filtering, and recombining relevant cases. This process creates a modular repertoire of practices that can be re‑used across sites while still respecting local nuances.

3. Punctuation – Unexpected events—such as equipment failures, safety incidents, or contractual changes—act as punctuation marks that interrupt the ongoing flow of practice. In response, the crew temporarily suspends existing routines, conducts a rapid review, and introduces revised steps. The punctuation event thus triggers a re‑negotiation of both differentiation and assembling, leading to a refreshed pattern of resemblance.

These strategies are not linear; they interact cyclically. Differentiation may generate new local variants that later become candidates for assembling elsewhere; punctuation can force a wholesale re‑differentiation; and assembling can smooth out the disruptive effects of punctuation by providing ready‑made alternatives. The result is a dynamic equilibrium where practices are never identical but remain recognizably similar enough to be called a “family”.

A crucial insight is that similarity is not decided purely on technical grounds. The authors show that the determination of what counts as “similar enough” is a pragmatic and political negotiation among multiple stakeholders—field supervisors, engineers, safety officers, and external regulators—each wielding different priorities, resources, and authority. Budget constraints, staffing levels, safety compliance requirements, and power relations all shape the final configuration of practice. Consequently, the authors argue that similarity should be understood as a negotiated standard rather than an essentialist one.

The paper contributes to IS theory in two major ways. First, it reframes the notion of standardisation: instead of seeking an unattainable uniformity, organisations should aim for “adjustable uniformity” that tolerates local variation while preserving a core set of overlapping features. Second, it highlights the role of technology not merely as a conduit for disseminating procedures but as an active “coordination mechanism” that can enable or inhibit differentiation, assembling, and punctuation. System designers are therefore urged to embed features that capture work‑arounds, support rapid re‑configuration after incidents, and facilitate knowledge sharing across sites.

In practical terms, the findings advise managers of distributed, safety‑critical operations to invest in knowledge‑capture tools, to formalise processes for rapid post‑incident re‑design, and to recognise the political dimension of practice alignment. By doing so, they can sustain a family resemblance of work practices that balances safety, efficiency, and local adaptability over time and space.

Overall, the study demonstrates that technologically mediated work practices achieve a “family resemblance” through a triad of strategies—differentiation, assembling, and punctuation—mediated by both pragmatic constraints and political negotiations. This framework offers a robust lens for understanding how similarity is continuously produced, contested, and reshaped in complex, distributed work environments.