If we are to follow the likes of Anna Munster and Geert Lovink in problematising the concept of the network as a unified or stable object, then perhaps we need a different point of view. As they put it: “the very notion of a network is in conflict with the desire to gain an overview” (2005). Maybe it is that ‘overview’ that leads us astray. Perhaps we need an ‘underview’, a view not from a different level but at a different scale. Protocol studies offers such a perspective… if we account for protocols as objects without foundations. If we locate protocols and standards – Internet in Galloway’s case, Ethernet in Miyazaki’s or imaging in my own – as either the outcome of structural processes (capitalist development, control societies etc) or as some determining essence, we not only fail to address the specific operations of those protocol objects and the alliances within which they enfolded, but also open them up to the sort of counter-protocological struggles or ‘exploits’ that Galloway talks about with Eugene Thacker (2007).

Miyazaki’s focus on time, through rhythm, adds a certain level of dynamism to the imag(in)ing of networks and so network politics, but that needs to be framed within an ontology that also places the emphasis on movement. It is here where I believe an object-oriented philosophical (OOP) approach (drawing on Harman’s reading of Latour (2009) and later speculative realism) offers a way of tracing the connections (or in OOP terms, the alliances) between protocol and corporate interests and state institutions as well as other scales of software. What is particularly useful about OOP is not only its refusal to engage with foundations and its flat ontology (treating protocols as actants equal with other material and immaterial forces and players), but its demand that networks (in the ANT rather than the narrow technical sense) be seen as always in motion, folding and enfolding, fractal and generative. In terms of my own work. The jpeg protocol’s alliances with Google’s search algorithms and business practice, Facebook’s data mining and face-recognition strategies as well as Apple, Nikon and Adobe’s businesses, Police databases etc are themselves enfolded with other software, material and immaterial actants. Those alliances are continually remade as artists and activists imag(in)e through the jpeg scopic apparatus but also as Facebook friends who have no interest in “network politics” use its capacity for metadata to tag, geolocate, link, connect and share.

This movement, the sort of rhythm perhaps that Miyazaki is exploring, demands an analytical framework that is equally fluid. As Galloway says: “we require a method of analysis unique to protocol itself”.

But it is not just at the level of analysis that an object-oriented approach to protocol offers possibilities. Galloway and Thacker say that “to be effective, future political movements must discover a new exploit… an antiweb… an ‘exceptional topology’” (p22). This is not some alternative space but rather the sort of different view, conceptualisation or ontology that OOP allows. By framing protocol as object-actant setting network effects and affects in motion, we are able to not simply deconstruct dominant imag(in)ings but also open the black boxes (OOP’s term for objects that have become so everyday they have become transparent) to reconfiguration. Mashups can be seen as an example of this exceptional topology. Here protocol generates new imag(in)ings. In terms of my own case study (the 2012 Olympics), the same protocol that generates imaginaries around legacy can generate imaginaries of globalised sports-business. My hacking the jpeg metadata to geolocate an image of a sportswear sweatshop inside the Olympic stadium, an exceptional topology is set in motion as soon as a visitor to the opening ceremony uses their iPhone to view “images taken nearby”. This is not to say that such protocol-trickery, such semiotic samizdat is the sum total of scopic network politics. What is more important is that protocol is not a fixed tool but a field of possibilities.

• Galloway, A.R. & Thacker, E., 2007, The Exploit: A Theory Of Networks, Univ Of Minnesota Press, Minneapolis.
• Harman, G., 2009, Prince of Networks: Bruno Latour and Metaphysics, Anamnesis, Melbourne.
• Munster & Lovink, 2005, Theses on Distributed Aesthetics. Or, What a Network is Not, FibreCulture(7).

Added as contribution to Network Politics: Request For Comments.

Adrain Mackenzie’s subject is ‘wirelessness’, the assemblage of practices, forces, ecologies and affects set in motion by Wi-Fi’s technical and cultural workings. He defines it as “an experience trending toward entanglements with things, objects, gadgets, infrastructures and service, and imbued with indistinct sensations and practices of network-associated change”(Mackenzie 2010: 5). His reading of William James drives him towards a focus on experience as a way of approaching the practical “inseparability of thinking and the things” (p14). Wirelessness here is a matter of “transitions and feelings of being in transition” (p39) the same themes that are at the heart of James’ radical empiricism. He often cites James’ insistence: “The relations that connect experience must themselves be experienced relations, and any kind of relation experience must be accounted as ‘real’ as anything else in the system” (James 1996: 42). James dynamic philosophy in motion, with its flat ontology is, for Mackenzie a useful tool to approach the dynamic spaces and relations of wireless urban experience which requires an acceptance of things and thinking as objects of analysis and, it could be argued object-actants.

What is important is how he sees code as part of this conceptual structure. Mackenzie, as in his previous work, is not afraid to deal with the “monstrously complicated” scales of software studies (Mackenzie 2008: 48) whether that is video codecs (2008), languages (2005, 2006) or algorithms (Mackenzie 2006). Here too Mackenzie feels it necessary to draw an account at the scale of algorithms in order to not only present a full picture of the experience of wirelessness but also in order to be able to trace the power relations in operation that his radical empiricism demands remain open and at the centre of analysis.

He argues that Digital Signal Processing (DSP) algorithms, “if acknowledged at all, {…] are treated as the most abstract aspect of electronic media and communication technologies, the part that lies closest to mathematics. We need a much more sensitive treatment of their becomings” (Mackenzie 2010: 66-67). Here algorithms, or protocols, are not looked at as background or foundation or dangerous supplement. Rather it is their ‘becomings’ that is at the heart of the arrivals, departures and transitions within networks and wirelessness as well as network and wireless experience. “They transduce realities,” he says (p67). In my terms they imag(in)e. We will come on to address the subtle difference between Mackenzie’s reading of radical empiricism and object-oriented philosophy but what is important to note here is first that Mackenzie treats software as implicated and in/enfolded in the “ness” he is looking at, within the different scales. Secondly it is important to note that he treats the algorithmic/protocol scale of software as at the heart of the issue. It is here where imag(in)ing works. It is here where analysis and ontology can develop and it is these objects (in his case the fast Fourier transforms, that also under[pin jpeg, and the convolutional coding-Viterbi  decoding (pp72-77)), that open up that “ness” experience, that black box, to analysis.

• James, 1996, Essays in Radical Empiricism, University of Nebraska Press, Lincoln.
• Mackenzie, A., 2005, The Performativity of Code : Software and Cultures of Circulation, Theory, Culture & Society, 22, pp. 71-92.
• Mackenzie, A., 2006, Cutting Code: Software And Sociality, Peter Lang, New York.
• Mackenzie, A., 2006, Java: the practical virtuality of internet programming, New Media & Society, 8(3), p. 441.
• Mackenzie, A.,  2008, Codecs, in Fuller (ed), Software Studies : A Lexicon, MIT Press, Cambridge, Mass., pp. 48-54.
• Mackenzie, A., 2010, Wirelessness: Radical Empiricism In Network Cultures, MIT Press, Cambridge, Mass..

Discussing the operations of protocol leads inexorably to the issue of ‘depth’. It is tempting to talk of protocols as the building blocks for software or software relations, to speak of different levels within software architectures or within analysis. Galloway for instance talks of the four nested layers of the Internet suite (2004: 39), themselves part of the ISO’s seven-layer Open Systems Interconnection Reference Model. In engineering terms this makes sense but in ontological, and arguably political terms it is problematic.

Graham Harman reads Bruno Latour as advocating a “flat ontology” (2009). He goes on to argue for a more nuanced account of two-types of objects and a fourfold ontology, but for the present purposes, the key thing is the removal of depth metaphors, levels and foundations. ‘Irreduction’ is central to this Latour (Latour 1993) (of course trying to pin down a single “Latour’ object is similar to trying to pin down the wave/particle object, or indeed any other object-actant). Harman says: “nothing can be reduced to anything else. each thing simply is what it is, in utter concreteness” (17). This gives rise to the famous Latour litanies where lists of ‘objects’, material and mythical are made to bump against each other creating the sort of dialectical images Benjamin the collector would have been proud of (Leslie et al 2007).

So if, philosophically, we need to treat all object-actants as ‘equal’ – not in terms of value so much as in status – how are we to map the relations between protocol and ‘the Internet’, between jpeg and Google? Latour and Harman are clear that they need to be treated in their specific locations, as occupying positions within actant-networks particularly when an actant becomes so everyday and transparent in terms of its alliances with other actants that it becomes a ‘black box’. Adrain Mackenzie, although wary of what he sees as a microemphasis in social constructivist accounts (Mackenzie 2010: 218) reads William James’ focus on ‘conjunctive relations’ as an antidote to “most social and cultural theories that tend to cut realities into things, selves, locations and relations” (p39). Analysis and politics cannot collapse all actants into an ontological soup. It is not just thinking of objects in irreductionist terms, it is finding a language to articulate those relations. Mackenzie argues that ideas (our conceptions of levels and depths), “have an immanent function in concentrating and short-cutting transitions” (p32) (the sort of dynamic movement that he sees at the heart of wirelessness and can be seen in the operations of protocol). Ideas are a way of “establishing a trajectory or modulating a movement by substituting experiential short-cuts” (p35). Ideas like any other actant, do things in the world. That is why we need a conceptual apparatus that can account for a dynamic flat ontology.

As many writers following Deleuze (Deleuze &Guattari 2004: 537) such as De Landa (2000) and Bennett (2010) have shown, the language of fractals and complexity offers a way of conceptualising these alliances and black boxes without recourse to foundationalism. A fractal is self-similar. As James Gleick has it: “self similarity is symmetry across scale. It implies recursion, pattern inside of pattern” (1988: 103). It does not make sense to speak of levels as one falls into a Mandlebrot set or a Lorenz attractor. Each scale at which the equation works is self similar. The fern-like branches at one magnification are recursively present at others.

This is not just a metaphor. Actant networks are complex systems. The relationships, alliances and translations within software and media ecologies and assemblages are complex in a scientific as well as an everyday sense. They are fractal. The alliances and relations around jpeg are self-similar to those around Apple’s imaging software and Google’s image search. They are not the same. They are at different scales but one does not determine or act as the foundation or context for the other.

To speak and think in terms of fractal-scales within and across software ecologies is to leave questions about actant-network-relations open, to raise the importance of alliances as a focus of analysis and to potentially open up black boxes to what Galloway and Thacker call ‘counter-protocological struggle’ (Galloway & Thacker 2007).

• Bennett, J., 2010, Vibrant matter : A Political Ecology of Things, Duke University Press, Durham.
• De Landa, M., 2000, A Thousand Years of Non Linear History, Swerve Editions, New York.
• Deleuze, G. & Guattari, F., 2004, A Thousand Plateaus : Capitalism And Schizophrenia, Translated by Massumi. Continuum, London.
• Galloway, 2004, Protocol: How Control Exists after Decentralization, MIT Press, Cambridge, Mass.; London.
• Galloway, A.R. & Thacker, E., 2007, The Exploit: A Theory Of Networks, Univ Of Minnesota Press, Minneapolis.
• Gleick, J., 1988, Chaos, Cardinal, London.
• Harman, G., 2009, Prince of Networks: Bruno Latour and Metaphysics, Anamnesis, Melbourne.
• Latour 1993, Irreductions, in The Pasteurization of France, Harvard University Press, Cambridge, Mass.,.
• Leslie, E., Marx, U. & Archiv, W.B., 2007, Walter Benjamin’s Archive : Images, Texts, Signs, Verso, London ; New York.
• Mackenzie, A., 2010, Wirelessness: Radical Empiricism In Network Cultures, MIT Press, Cambridge, Mass..