Protocol: How Control Exists After Decentralization
Source: https://archive.org/details/protocolhowcontr0000gall ↗
Galloway's thesis is that the internet is not, in any politically meaningful sense, a space of freedom — it is a space of protocol.
He argues that TCP/IP and DNS constitute a new form of control that operates not through hierarchy or centralized authority but through the voluntary adoption of shared technical standards.
The book draws directly on Deleuze's concept of "societies of control" and on Foucault's analysis of distributed power, applying both to the actual engineering of network architecture.
Galloway reads RFCs as political documents and routing tables as instruments of governance.
The result is a theory of power that takes the OSI model seriously as a diagram of how authority is exercised in distributed systems.
It remains the sharpest political reading of the internet's technical infrastructure.
Central argument
Galloway argues that the internet is not a decentralized space of freedom but a highly controlled one, where control operates through the mandatory adoption of shared technical protocols — chiefly TCP/IP and DNS. Drawing on Deleuze's societies of control and Foucault's distributed power, he reframes RFCs and routing tables as political documents that govern behavior without a central authority issuing commands. The key thesis is that protocol — voluntary, horizontal, technically encoded compliance — is the dominant form of power in networked systems, more effective precisely because it appears neutral and infrastructural.
Critique
Galloway's framework is most persuasive when applied to the network layer, but it struggles to account for the market and legal mechanisms that have since concentrated internet power into a small number of platform gatekeepers — a dynamic that looks far more like traditional hierarchical control than Deleuzian modulation. By 2004, his argument was already racing against the emergence of Google and the App Store model, where protocol compliance is necessary but insufficient, and where a single actor can effectively veto participation. The book's insistence on protocol as the primary register of power may understate how capital and intellectual property law have reasserted older, more vertical forms of authority on top of the distributed substrate.
Why it matters for product
For a CPO, the book reframes a familiar product problem: the standards, schemas, APIs, and data contracts a team publishes or adopts are not merely technical choices — they are governance decisions that shape what other teams, partners, and users can and cannot do. Designing a platform API is an act of protocol-making in Galloway's sense, quietly encoding assumptions about who has agency and who must comply. This is particularly sharp when thinking about internal platform teams, design systems, or data contracts across a product organization: the architecture you choose is a power structure, and the teams who write the specs hold a form of authority that rarely appears on any org chart.