Ideas do not live in minds alone. They inhabit instruments, archives, cables, libraries, databases, streets, rituals, bodies, software platforms, scholarly conventions and the infrastructures that make thought durable, transmissible and contestable. To ask where ideas live is therefore to reject the romantic fantasy of knowledge as a purely internal, individual possession and to recognise instead that ideas are ecological: they emerge through relations between persons, media, technologies, institutions and material environments. Across the texts considered here, a shared proposition becomes visible: knowledge is never simply produced and then communicated; it is shaped by the systems through which it is recorded, circulated, preserved, visualised, monetised, forgotten or made public. The life of ideas is thus inseparable from infrastructure, whether that infrastructure takes the form of a scholar’s memex, a digital repository, a public library, an undersea cable, a laboratory apparatus, an urban wall, a data archive or a contested social media platform. Ideas live wherever evidence becomes meaningful, wherever memory is organised, and wherever communities construct the conditions under which something can count as knowledge.
Vannevar Bush’s “As We May Think” offers one of the most influential twentieth-century answers to this question. For Bush, ideas live in associative trails rather than in rigid systems of classification. His imagined memex is not merely a storage device but an intellectual architecture, designed to mirror the mind’s movement from one item to another through relevance, analogy and connection. The scholar studying the Turkish bow does not simply retrieve isolated documents; he creates pathways, annotations and linked sequences that make knowledge navigable. Bush’s central insight is that the problem of modernity is not scarcity of information but the inadequacy of inherited mechanisms for selection. Ideas live, then, not in the accumulation of records but in the relational structures that permit meaningful retrieval. His essay anticipates later hypertext systems because it understands memory as active, connective and social: once trails can be shared, thought becomes a collaborative infrastructure rather than a private act of recollection. Christine Borgman extends this infrastructural understanding of scholarship by showing that knowledge in the digital age depends upon sociotechnical systems. In Scholarship in the Digital Age, she argues that digital networks transform not only the circulation of research but also the grounds of legitimacy, preservation and trust. Ideas live in documents, data, publications, repositories, disciplinary norms, intellectual property regimes and institutional practices. A preprint, a dataset or a conference paper is not simply a container for thought; it is part of a system of scholarly validation. Borgman’s later Big Data, Little Data, No Data deepens this argument by refusing the mythology that bigger data automatically produce better knowledge. Data are not self-evident facts waiting to be mined; they are representations of observations, objects or phenomena used as evidence in particular contexts. Thus, ideas live differently in astronomy, archaeology, Buddhist studies, sensor-networked science and the social sciences because each field constructs evidence through different instruments, standards, scales and practices. The absence of data, the refusal to release data or the inability to reuse data also shapes what can be thought.
Susan Leigh Star and Karen Ruhleder’s account of infrastructure makes this point even sharper. In their analysis of the Worm Community System, infrastructure is never a thing in itself but a relation. What appears to a software developer as a simple technical instruction—download a file, use UNIX, configure X Windows—may become, for a biologist, a complex institutional and epistemic problem involving funding, laboratory space, tacit knowledge, technical support and disciplinary culture. Ideas therefore live in arrangements that may remain invisible until they break down. The Worm Community System was designed to democratise access for a scientific community, yet its actual use depended on laboratories’ resources, skills and local contexts. A system built for collaboration could reinforce inequality if only well-funded or technically sophisticated laboratories could participate. This reveals a crucial principle: an idea’s location is not merely the place where it is stored, but the ecology that permits it to become usable. Infrastructure is the condition of possibility for thought, but also the site where exclusion, friction and power are reproduced. Hans-Jörg Rheinberger’s historical epistemology shifts the question from communication to experimentation. For Rheinberger, ideas live in experimental systems, where epistemic things emerge through apparatuses, materials, procedures and uncertainty. Scientific objects are not simply given to thought; they are produced through practices that stabilise them enough to be investigated while preserving enough ambiguity to provoke further inquiry. Drawing on Cassirer and Bachelard, Rheinberger argues that philosophy of science can no longer impose a universal system from above. It must follow the differentiated historical practices through which sciences generate their objects. In this sense, ideas live in risk. They live where reason exposes itself to the unexpected, where an experiment can reorganise what counts as a problem, an object or an explanation. The laboratory is not merely a place where theories are tested; it is a generative environment in which knowledge becomes possible through the tension between control and surprise.
If Rheinberger shows that ideas live in experimental systems, Shannon Mattern shows that they also live in cities. In “Deep Time of Media Infrastructure”, she argues that media infrastructure did not begin with electronic communication. Cities have always been communicative environments, shaped by voice, inscription, architecture, ceremony, public address and writing. The agora, the Roman Forum, Fatimid Cairo’s public texts, Chinese stone inscriptions, Yemeni spiral urban forms and New York’s Union Square all demonstrate that ideas live in spatial arrangements. Streets, walls, plazas, facades and acoustic volumes become media systems. The city is not merely represented by media; it is itself a medium. Mattern’s concept of deep time is essential because it resists the novelty myth of digital culture. Contemporary media do not replace older infrastructures; they layer themselves upon oral, graphic, textual, sonic and architectural forms. Ideas live in this layering, in the residual presence of past media that continue to structure present communication. Mattern’s broader work reinforces this critique of computational reductionism. Her scholarship on libraries, dashboards, maps, urban intelligence, maintenance and care insists that cities cannot be understood as computers. They are not problems to be optimised through data extraction but plural, sensory, historical and political environments. Ideas live in public libraries as social infrastructures; in maintenance practices that sustain everyday life; in maps that organise spatial knowledge; in dashboards that promise control but often conceal ambiguity; and in forms of urban care that resist the fantasy of seamless automation. Her corpus therefore relocates intelligence from the algorithm to the public realm. Urban knowledge is not simply processed; it is negotiated through institutions, material supports, embodied experience and collective memory.
Nicole Starosielski’s The Undersea Network brings the location of ideas into the ocean. The introduction, “Against Flow”, challenges the dominant imagination of the Internet as wireless, dematerialised and deterritorialised. Undersea fibre-optic cables carry the overwhelming majority of transoceanic digital communication, yet they remain largely invisible. Ideas moving across the world as emails, research files, financial transactions, images, videos and social media posts pass through rural and aquatic infrastructures: landing stations, coastal communities, seabeds, cable ships, maintenance regimes and geopolitical chokepoints. Starosielski’s maps of transpacific cable routes show that digital networks often follow older telegraph, telephone, colonial, trade and military pathways. This means that ideas live in routes as much as in platforms. Their apparent fluidity depends on fixed, vulnerable and unevenly distributed material systems. The “cloud” is grounded in the ocean, and global connection is sustained by strategies of insulation and interconnection that protect cables from turbulent ecologies while anchoring them in local environments. David W. Lewis’s bibliographic scan of digital scholarly communication infrastructure brings these questions into the contemporary academy. His mapping of 206 projects across researcher tools, repositories, publishing, discovery, assessment, preservation and general services shows that scholarly ideas now live in a highly fragmented ecosystem of platforms and organisations. The central political question is whether this ecosystem will be governed by commercial consolidation or community-controlled open infrastructure. If major publishers integrate the entire research workflow—from writing and submission to data analytics, discovery and evaluation—universities risk losing control over the conditions of knowledge production. Yet open projects face their own difficulties: sustainability, governance, collective action and funding. Ideas live, therefore, in business models. They live in the difference between a community-owned repository and a proprietary platform, between bibliodiversity and corporate capture, between open standards and closed analytics. The future of thought depends not only on access to content but on who owns and governs the systems through which scholarship moves.
Tonia Sutherland’s “Making a Killing” reveals a darker location for ideas: the racialised digital afterlife of bodies, trauma and evidence. Her essay examines the circulation of digital records documenting the deaths of Black Americans and asks how such records function as memory, spectacle, ritual and commodified content. Here, ideas live in images that wound, in videos that bear witness, in social media systems that replay trauma, and in archives where Black bodies become both evidence and object. The deaths of Michael Brown, Alton Sterling, Philando Castile and others show that digital visibility is ethically ambivalent. Images can mobilise protest and counter official denial, yet they can also reinscribe white supremacist spectatorship by making Black death endlessly shareable. Sutherland’s crucial contribution is to show that records are not neutral. Their circulation depends on ritual, embodiment, consent, context and power. Ideas about justice, mourning and memory live in the contested space between witnessing and exploitation. Joanna Bryson’s “Patiency Is Not a Virtue” adds a normative dimension to the question by asking where responsibility should live in systems involving artificial intelligence. She argues that AI moral status is not a fact to be discovered but a decision societies make through design and law. Because machines are artefacts, humans are responsible not only for how they are treated but for what kinds of entities they are built to be. Bryson rejects the idea that intelligence alone should generate moral patiency or agency. To attribute responsibility to AI may allow designers, corporations or governments to evade accountability. Her argument therefore locates ethical ideas not inside autonomous machines but in the human institutions that design, deploy and regulate them. Responsibility must live with those who create systems, not with artefacts deliberately constructed to appear agentic. In this view, ideas about moral agency are infrastructural too: they are embedded in law, design principles, transparency requirements and social norms.
Taken together, these works answer the question “Where do ideas live?” by dismantling the distinction between thought and support. Ideas live in associative trails, datasets, laboratories, repositories, cable routes, city squares, digital platforms, archives, experiments, bodies, machines and institutions. They live in media, but media are not merely channels; they are environments that shape what ideas can become. They live in infrastructures, but infrastructures are not merely technical systems; they are political, historical and social relations. They live in data, but data are not raw; they are contextual, curated and interpreted. They live in archives, but archives can preserve violence as well as memory. They live in machines, but machines do not absolve humans of responsibility. The most important conclusion is that ideas survive only when the conditions of their circulation are cared for. To ask where ideas live is also to ask who maintains them, who profits from them, who is excluded from them, who is harmed by them and who has the power to make them disappear. The contemporary crisis of knowledge is therefore not simply epistemological; it is infrastructural. We often speak as though ideas fail because people misunderstand them, suppress them or disagree about them. These things matter, but ideas also fail when infrastructures decay, when records are inaccessible, when data cannot be interpreted, when repositories are defunded, when cables break, when platforms commodify trauma, when laboratories cannot sustain experiments, when libraries are weakened, when commercial systems capture public knowledge, or when ethical responsibility is displaced onto artefacts. Conversely, ideas flourish when infrastructures are open, durable, plural, accountable and cared for. They flourish when communities can trace provenance, preserve context, maintain access, resist commodification and build systems that support interpretation rather than mere circulation. Ideas live, finally, wherever human beings build worlds capable of holding them.
Borgman, C.L. (2007) Scholarship in the Digital Age: Information, Infrastructure, and the Internet. Cambridge, MA: MIT Press.
Borgman, C.L. (2015) Big Data, Little Data, No Data: Scholarship in the Networked World. Cambridge, MA: MIT Press.
Bryson, J.J. (2016) ‘Patiency Is Not a Virtue: AI and the Design of Ethical Systems’, Proceedings of the AAAI Spring Symposium Series, pp. 202–207.
Bush, V. (1945) ‘As We May Think’, The Atlantic Monthly, July.
Lewis, D.W. (2020) A Bibliographic Scan of Digital Scholarly Communication Infrastructure. Atlanta, GA: Educopia Institute.
Mattern, S. (2015) ‘Deep Time of Media Infrastructure’, in Parks, L. and Starosielski, N. (eds.) Signal Traffic: Critical Studies of Media Infrastructures. Urbana: University of Illinois Press, pp. 94–112.
Mattern, S.C. (2022) Curriculum Vitae. New York: The New School.
Rheinberger, H.-J. (2018) ‘On Science and Philosophy’, Crisis & Critique, 5(1), pp. 341–347.
Star, S.L. and Ruhleder, K. (1994) ‘Steps Towards an Ecology of Infrastructure: Complex Problems in Design and Access for Large-Scale Collaborative Systems’, Proceedings of the 1994 ACM Conference on Computer Supported Cooperative Work, pp. 253–264.
Starosielski, N. (2015) ‘Against Flow’, in The Undersea Network. Durham, NC: Duke University Press, pp. 1–25.
Sutherland, T. (2017) ‘Making a Killing: On Race, Ritual, and (Re)Membering in Digital Culture’, Preservation, Digital Technology & Culture, 46(1), pp. 32–40.
