Studio for Design and Research
The Cloud, the State, and the Stack: Metahaven in Conversation with Benjamin Bratton


Benjamin Bratton is a theorist whose work spans Philosophy, Art and Design. He is the Director of the Center for Design and Geopolitics at the University of California, San Diego. In a previous life, Bratton was the Director of the Advanced Strategies Group at Yahoo! His forthcoming book, The Stack: On Software and Sovereignty (The MIT Press, 2013), deals with the political conditions brought about by what Bratton calls “planetary-scale computation.” The interplay between the thick digital spaces that overlap existing geographies, including the nation-state and its borders, will result in new polities which may look completely different from traditional sovereign units. In this ongoing geopolitical experiment, Bratton argues, the accident will be the most important evolutionary form. The Stack sets out to describe a model for such events to unfold; it deploys an architecture of seven layers of sovereignty and computation which wrap up the earth like a skin.
This is no easy stuff to grapple with. On the Nomos of the Cloud: The Stack, Deep Address, Integral Geography, a lecture Bratton presented at the Berlage Institute and at the École Normale Superieure in 2011, formed the starting point for a dialogue in which we inquired about The Stack. Some citations from his talk are formatted here as (fake) iPhone screenshots. We’ve previously interviewed Bratton in a special section we edited for Print magazine in 2011. That interview can be read here. We’ve also lectured twice at the annual Design and Geopolitics Conference, organized by Calit2 and UCSD, alongside Nathalie Jeremijenko, Lev Manovich, Larry Smarr, Daisy Ginsberg, Ed Keller, and others. This interview is part of the research for our essay series, Captives of the Cloud, and our forthcoming book, Black Transparency (Sternberg Press, 2013). 


What is The Stack?

Benjamin Bratton
Instead of viewing the various scales of emergent ubiquitous computing technologies as a haphazard collection of individual processes, devices and standards (RFID, cloud storage, augmented reality, smart cities, conflict minerals, etc.), it is more illuminating to model them as components of a larger, comprehensive, meta-technology. The Stack is planetary-scale computation understood as a megastructure. The term “stack” is borrowed from the TCP/IP or OSI layered model of distributed network architecture. At the scale of planetary computation, The Stack is comprised of 7 interdependent layers: Earth, Cloud, City, Network, Address, Interface, User. In this, it is an attempt to conceive of the technical and geopolitical structures of planetary computation as a “totality.”

What are the “geopolitical structures of planetary computation”?

Examples of the geopolitics of the Cloud might range from anonymous server routers from Egypt, The Google-China conflict, the ITU United Nations governance controversies, Anonymous (the group) going up against Mexican drug cartels, WikiLeaks, the Facebook/Twitter/YouTube stack in Cairo, TOR users building on the Amazon cloud, MPLS level 2 dark fiber networks connected trading centers for optimal position, to trading floors gaming the speed of light, the microeconomics of transcontinental bandwidth… all these things exist already, but my interest is as much in thinking about what comes next.
Right now the geopolitics of the Cloud are defined by a clash between State and non-State organizations—China vs. Google, US vs. WikiLeaks, etc.—with decentralized networks in between. In the coming years, the geopolitics of the Cloud will be defined by the tension and conversion of States and Clouds into one another: States becoming Cloud-based platforms, and Cloud-based platforms taking on more and more practical forms of sovereignty—identity, geography, mobility. So a Cloud Polis, we might imagine, would represent other weird kinds of “sovereign” software/State platforms that are probably not bound to specific spots of territory in the same way as the Westphalian nation-state was. They might be more effective and democratic, or more totalitarian, more or less elective than the models we have now. Some asymmetrical mix is most likely. I retain the word “geopolitics” in that these would be defined by different relationships to planetarity as a global/local condition, and would I assume continue to stage a great deal of inter-sovereign conflict between some polities that are brand new and others that are primeval but which have learned to make use of the Cloud—such as religions.
I mean polis more as a body of citizens, whereas really a polity is a component of the body of the State. In addition to what we’ve discussed already it refers to the important role of global networks of cities. The Stack sees urban layers of information, energy and concrete interchangeable. It sees cities as amalgamations of these three infrastructures even if their histories and politics are very different from one another. Each has its own history, momentum, alliances, contradictions.
Further, this aggregate urban condition means that instead of each of us living in “a city,” we all live in “THE city.”

This is an assumption critiqued by Pier Vittorio Aureli in The Possibility of an Absolute Architecture. He wants to discern between a walled city and a continous city, and pertains that every political form, every political association, hence every city, must have a limit, is absolute, not so much in it being a totality as it being clearly defined as a shape. An island, perhaps—standing out against its surroundings. Is your notion of “THE city” not evoking more the end of any distinct polity, than announcing a new one?

To a degree the “City” chapter of The Stack works to suspend and unwind this distinction between the walled and the continuous.
I discuss how common infrastructure provides common citizenship. The notion of civis romanus sum, where the being inside the city was the condition of membership. The wall provides interiority and dwelling, and it can be physical or virtual. It can be a concrete barrier or an interface or a living skin. Mobile networks link one site to the other and let some people walk right through walls, and others right into new ones. Those networks substitute for “the wall” and also provide for new kinds of continuity and connection, or discontinuity and segregation. It is augmented by these non-linear links and gateways, as described by Deleuze’s late essay, Postscript on the Societies of Control. I would not say that this suggests the end of distinct polities, because there may well prove to be many more ways to subdivide the superimpositions of real and virtual walls, real and virtual continuities.
If you presume that the division is what provides the space, and therefore the spatial identity, then its virtualization into mobile networks could be seen as the removal of one identity and the provision of another.
The idea is less about some “authentic” continuity than the simple observation that in San Diego, Google Maps does not care, as of now, whether its user is an “illegal alien” or not. It provides the City to him just as quickly. Concrete or Cloud-based, commonly accessible infrastructure provides common access to urban space, and to the extent that any one city is really part of a much larger urban network condition.
So again, instead of the one city and its single polis, perhaps it is more that from global urban network another—weirder —form of cosmopolitanism is derivable from it? How could these be leveraged, formalized, defended?

Are we in an imagined community with someone because we disagree with him or her on Twitter?

Perhaps, but to what effect? The imagined community that could ground any Polis might take to the Cloud to re-animate their archaic theology , or it might be a function of the Cloud that is new and largely native to the Cloud era. 
Just because you are talking to someone and in the same “mental space” or whatever doesn’t qualify as a real imagined community, but doesn’t exclude it either.
Let’s not forget that the realm of the imagination gets built on very real backbone.
The hard technical cabling of the cloud, its voracious energy appetite, its bizarre landscaping footprint, its water and coal requirements, etc, are central to what it is as a global project. It is dependent on the Earth layer for gigawatts of energy, and vulnerable to the vagaries of climate change. The Cloud is very heavy.

How do the layers of The Stack work as a system? Can someone cross all seven layers in one action?

Every message from a user to another user of The Stack goes all the way down from top to bottom and then back up the stack again. In TCP/IP, that would be from the application layer, like e-mail, down to the fibre wire, and back up to the application layer. That’s how stacks work in general.
So in my model of The Stack, a link goes from user, interface address, network, city, cloud to earth and then back up again to user. The important image to keep in mindis that the whole system is invoked in any one connecting gesture. That is, any one message tunnels up and down then back up the vertical Stack, and so User A speaks to User B by activating all seven layers down the Stack and all seven back up the Stack. In this, the whole, the totality, is invoked and activated in an instance.

In terms of the present day examples you cite, what do these varied examples have in common? For example, the WikiLeaks saga is not one of “pure” cloud storage; part of their hosting always was about circumventing storage in any single jurisdiction. Could you shed light on how, for example, TOR bridges in the Amazon cloud are an argument for the Stack or fit in it?

As you suggest, one thing that ties them together is the ways that they puncture and perforate Modern models of contiguous sovereignty and jurisdiction. But many things do that. The Kantian-Westphalian geographic system never was stable. Its enclaves and exclaves, not to mention colonies, were always there.
Regarding the examples, in some cases these are matters of deliberately invested design—such as dedicated lines between trading floors—whereas in other cases they emerge in relation to some kind of emergency. I think that this is probably the more important process. Perhaps the more important gestures and networks of sovereignty emerge not only as formal acts of drawing but through more improvised states of exception that become, over time, norms and forms. No one planned for Facebook/Twitter/YouTube to work the way they did in the Arab Spring, thought I tend to agree with Evgeny Morozov on this. Morozov is very critical of what he calls the Google Doctrine narrative, which says thatthe Arab Spring was caused by Twitter, YouTube, etc, as opposed to already existing opposition networks, and that the internet’s role in regional conflict is unambiguously positive. Morozov argues that the whole platform can just as easily, perhaps more easily, be used as a tool for oppression and surveillance. The same ambivalence about any of the examples is wise.
For example, TOR on Amazon shows how the two sides of the “anarcho” vs. “totalitarian/corporate” divide are not so neatly separated. These processes are mutually parasitic, cannibalizing each other’s accidental accomplishments and affordances.
To answer your specific question, yes: WikiLeaks circumvented centralized Cloud services as they exist, but in doing so made it plainly obvious how different network topologies of ubiquitously accessible storage don’t just evade jurisdiction, they produce it.


To me what is most interesting in the very long time are not Manichean struggles between light and darkness, our team/their team, ork/not-ork―all the messianic, absolutist rhetoric is surface-level noise―but how truly alien, unforeseen, even indecipherable geopolitical forms can appear and accelerate the post-Copernican rotation in geopolitical thinking in ways that we can scarcely anticipate. It’s not just about redrawing the flat world map but of thinking political geography in multiple dimensions and temporal scales.

We agree. Isn’t Anonymous itself an example, if not the clearest one, of this? All their videos are on YouTube. Anonymous, as opposed to the cyber-vegetarians of yesterday’s “critical media,” are the biggest all-you-can-eat omnivores of the internet… Otherwise your point is very clear.

As someone who very much appreciates a lot of what the two groups do, let me speak somewhat critically of Anonymous and WikiLeaks. My points are somewhat obvious and not meant to heap scorn.
Anonymous does some interesting things, but the 4chan angry teenager roots also steer them off into predictable dead-ends. Their YouTube videos seem straight out of a 1999 videogame about “Hackers in Cyberspace vs. The Man.” Nevertheless, they are a clear example of what a Cloud Polis might look like, and to me much of what they do is about clearing the way for something more substantial to emerge.
Hopefully their model of Cloud Polis is one among many, because in the constitution of deep alternatives, it’s not only about subverting and liquifying Oedipalized power, it’s also about knowing what to do with power once you have it, how to compose with it.
WikiLeaks is a method and ethos raised to the level of an ideology—Assange’s interview in e-flux with Hans Ulrich Obrist is great, by the way—but it’s not clear what the next move might be, even if Assange were free to move about. Open it all up? OK. Everyone will see what’s really happening? OK. This is a basic Enlightenment aspiration, but the idea that once the primal evil is exposed the public will not tolerate it is insufficient to say the least. Even as they evade and escape legal harrassment, I don’t see alternative architectures other than an anarchist tabula rasa, and some fragile and absolutist paranoias. Building what is next is perhaps not Assange’s job, I understand. But I am sorry, it’s obvious that a world without authority and opacity is entropic; it would have no music (or very bad music). It’s not only not possible, it’s not desirable. A field without solid lines cannot sustain interesting forms of tension for very long. I am more interested in the very long expanse of composable time than in any purifying first principles regarding the legitimacy of authority. By the way, apropos of nothing, “cyber-vegetarianism” is always a great phrase. Thank you.

How about this idea of “totality”?

Probably the most well-known reference would be Fredric Jameson’s work on utopias being speculative models of social totalities. Jameson discusses this in his “Utopia as Enclave” essay. Islands, for example, are total jurisdictions, self-contained and therefore so suitable for literary and political utopian fictions. So the model of The Stack, while not a proper a social scientific observation about the state of globalization, is nevertheless a gesture toward a kind of "total" concept-model-technology through which we might glimpse very different kinds of processes and events as interrelated, interconnected, systematic, coherent—and therefore plastic and available to higher-order sorts of modifications.
In that it works with with how geopolitical-technological systems can be thoroughly recomposed, The Stack as an idea is probably also a utopian project, especially in that it claims possible futures as designable things.
Specifically with regards to “design,” it also identifies planetary computation as part of a lineage of architectural megastructures, and therefore works with the utopian as a way of teasing out the dystopian potential of gestures at that scale.

Which megastructures in particular, and do you imply megastructures that were somehow “realized,” ones that weren’t (or some of both)?

We probably don’t have to go further than some canonical utopian megastructural projects such as Archizoom’s No-Stop City, the Metabolists’ Tokyo Bay, Superstudio’s Continuous Monument, etc. These may have found their ultimate realization, however twisted, in today’s Stack megastructures, such as Cisco and NASA’s Planetary Skin, Foxconn, the oceanic fibre backbone, etc.

Why Foxconn?

First, I would underscore that even these apparently noxious developments—like Foxconn as one future of human labor—may contain the seed of something more “utopian” that we have to imagine and compose into existence using the raw material of catastrophe. Just as the progressive utopian megastructures of the 1960s and 1970s contained the seeds of these Stack incarnations, perhaps an opposite conversion from dystopian to utopian is just as feasible. If Archigram’s Plug-In City can mutate into Foxconn, then Foxconn can mutate into something else.
So to think of The Stack as a megastructure means to think of its ongoing construction as a giant act of collective architecture, which means theories, models, prototypes, testing, construction, occupation, decay, retrofitting, etc. The whole of planetary computation is architecture, not like architecture. What does that observation acccomplish, other than making things more complex for designers? It clarifies that because it is a work of design, it can be redesigned, and that it has a wide archive of precedents to draw upon .
If we see each of the layers of The Stack as parts of a whole, it may make it easier to adjudicate the complex geo-design decisions that they demand of us.
I characterize each layer in terms of its effects and potential, but also in terms of the “integral accident” that it brings about. As Paul Virilio had it, “the invention of any new technology is also the invention of a new kind of accident.” Each layer brings its own accidents, and The Stack’s layers together generate more amalgamated accidents, particularly to natural and political ecologies.

Since all the layers of the Stack bring about such a sprawl of new accidents (supposedly on the many points where they, like tectonic plates, are not shaped to interact smoothly) how can the Stack as a whole then make it easier to decide on geodesign issues? Isn’t it rather that it might make those decisions more informed of their (un)intended possible consequences? We’re also thinking of Vinay Gupta’s notion of the goat rodeo—a situation where different players collaborate with different goals in mind and the outcome is a complete mess.

Because the inverse of Virilio’s axiom is equally true, and new accidents also produce new technologies. In a way, the book, The Stack, should be read as a long, intricate, challenging design brief, for me and for others. It draws the world as it is today, so as to illustrate certain possibilities about its future that are not true today, and may never be true ever. And the opposite strategy is equally at work, in that the book describes future possibilities just so as to underscore some point about the present moment. It’s not then a book of predictions. It’s a book of rationales for a series of projects, some by me and some by others, most of which the book itself does not explicitly name.
Each layer of the Stack will go on generating accidents whether we see them as part of a larger whole or not. We can deal with them locally as we choose.I prefer to think in terms of integrated models, especially interrelated catastrophes. Even if those models are speculative, this is integral to my own design thinking. As a theorist, I think it is more useful to provoke larger gestures, ones that cut or link diagonally across layers in the system that otherwise might have remained unconnected. As a model The Stack is vertical, it is a sectional totality. I invite the reader to turn it on its side if they choose.
Back to the geopolitics of the Cloud, as an example, it is partially defined by the ”accidental” de-lamination of traditional Westphalian geographies of sovereignty through the realization of other topologies, including bound networks. As said,for me the Sino-Google Conflict is emblematic. Here we have the vertical superimposition of two possibly incommensurate logics of geography and governance. One, a globally distributed, cognitive capitalist, NSA-protected polis predicated on data rationalization, and two, a geographically circumscribed central command which sees the Cloud as an extension of the body of the State. The topological difference between the two makes them incommensurate, and the friction caused by the grinding of these two “layers,” each demanding acquiescence of the other, will characterize the geopolitics of the coming decades. Each of us will live our lives walking through the mist of multiple superimposed jurisdictional claims on any site. How we manage the final undecidability of sovereignities will characterize a good part of our legal, public identities: Rubik’s Cube passports.
Other accidents bring about different kinds of jurisdictional emergencies. AMO did a project where it redivided European borders according the kinds of energy which different areas produce. We see at IPCC that very different polities align and federate according the self-interest of energy production. At the same time, and at the same meeting, groups like the Alliance of Small Island States appears to consolidate the interests of those effected by the energy production of the producer groups. In the alliances of producers and those effected, climate change redraws the map as well.


This is not about the “classical” juxtaposition between solid and fluid, state and network, is it?

No. It is possible to read that opposition into this, but that doesn’t help much when States become more like Networks and Networks more like States. The geographic-topological distinction between fixed node or fluid edge, central or de-central, could locate States on either side of the distinction and that amphibiousness is only increasing. I take “State” to mean a kind of formal jurisdictional structure that absorbs a geographic domain into its purview, that can assign identity and address, which can control mobility within a certain field, and which can claim final authority over the legitimacy of violence within a specific domain. But “networks” do all those things too.
As said, Google-China is a key example here, but US superjurisdiction is another to be sure. The US can, it seems, claim a right to inspect any data seemingly remotely related to a machine on its soil. In fact the limit of US jurisdiction can’t be defined by a contiguous land mass or even a body of law. Pick your favorite example: drones, the Megaupload seizure, the monopolies of Visa/Mastercard, ICANN, etc.
This tension drawns on archaic distinctions with very different histories of (the drawing of land vs. the drift of the sea), but, again I am most interested in how such conflicts can give rise to weird, novel jurisdictional forms. Your Facestate project shows what is menacing and also perhaps weirdly progressive about certain combinations.
We are at an impasse in our geopolitical thinking about what comes next, and I think speculative design can provide a vocabulary of alternatives. I am not a great fan of States per se, but neither of markets and corporate gardens, nor particularly faithful in anarchic autopoiesis and absolute commonwealths.
It seems that in practice we perhaps cynically lean on one of these three when the accidents caused by combinations of the other two become too awful. States are invoked and tolerated when markets and commons contradict each other Commons are championed when states and markets horde and flail, while we turn to the apparent dynamism of markets when states and commonwealths can’t get it going. A kind of triangulation of bad faith?
Instead, I am interested—aren’t we all?—in the accidental design and deliberate composition of alien, alternative platforms of mass sovereignty.

How exactly are Clouds de facto States? They can bring together a rather large polity (like the Facebook population, for example), but this polity has no rights within that Cloud and does not share its (for example) Facebook citizenship with other Facebook users as their primary sense of citizenship yet… correct?

I will rely on the characterization of The Stack as a design brief. You are correct that no private Cloud platform existing today has the full power of a State, and perhaps they never will.But Cloud platforms will gain in influence and ubiquity and so may result in modes of sovereignty that are very different but equally impactful as those of States. Perhaps your Google ID will mean more in terms of your effective ability to migrate and trade and communicate than your passport or State ID. Furthermore, and equally important States will themselves increasingly become Cloud-like in various ways. The conversion works both directions.
The curious example of the Google Maps war between Nicaragua and Costa Rica, where a change in Google Maps almost initiated a conflict between the two countries, is telling. The naming and measuring of the ground over which and into which politics might maneuver used to be a core function of the State. In this example, States defer to Google Maps to draw the nomos.

A Google ID might “mean” more but the protocol of border crossing and/or claiming of state benefits is still served by the passport. How do Cloud-issued identity certificates pass the test from “soft” to “hard” currency in this respect? For example, when one’s Facebook ID becomes one of the possible entries into one’s “digital profile” with the state? And if not, how does something like a Google ID loosen, affect, or otherwise change the coordinates of passport/national identity within the context of the Stack?

Well, yes, how indeed. We will have to prototype it. There are thousands of possible routes to and from hard and soft in these cases. Again, I am not making a claim that this has taken place, but working backwards from an emergent future we can see where this leads.
Considering the inroads of with the local, State, and Federal agencies, it is not unthinkable that much of the back-end of the Federal Government’s user/citizen management will be partially sourced to Google (as opposed to sourcing such tasks to far less competent contractors).
The Feds have had no end of trouble with Internet ID schemes that would allow for trusted real identity mechanisms online, such that finally a reasonable electronic signature could be done once and for all.
Clearly Google+ and Google Wallet are ways to build Google ID into a macro-identity across purposes and sectors such as banking. Perhaps it becomes more politically feasible for the Feds to quietly adopt or offer Google ID support for certain important functions, and let that absorb the goals of real Internet ID programs. In such a case, my passport can also buy my plane ticket, book my hotel, etc.
But thinking long term, we could envision competing states/cloud platforms with competing services and protocol lock-ins. It may be that California offers far better digital identity services than Arizona and so non-residents choose to be part of that platform, effectively paying taxes to a state other than their own. People might become more intertwined with the services and content, and political conflicts of California, because the “state cloud” is actually providing them with identity, economy, even schooling, who knows. Perhaps they choose to live under California’s data laws/ platform even if they are in Dubai, and perhaps no one can stop them. Perhaps they don’t give up their Dubai passports but it might not matter. Maybe Taiwan’s services will be, for whatever reasons, deliberately designed to prevent interoperation with California. So the walled garden problem becomes one of real competing feifdoms. We can make up scenarios, and it’s probably well worth doing. I am waiting for California to set up its own embassy in Brussels.
The optimistic scenario is the emergence of “new modes of sovereignty” that would let people assemble and connect in ways that better serve their real needs and wants. Perhaps these are not recognizable as states, platforms, corporations, or commons, but some bizarre hybrid of all four plus three new things we don’t know yet.
Equally likely is what we can call Cloud Feudalism. In this scenario, the walls of some gardens are hard and thick. The mechanization and routinization of everyday life is amplified beyond measure and all politics (including biopolitics) reduce user-citizens reduced to mere personnel.
Those without means to purchase their way into a Sky Club Sovereignty are left to the wilderness: no privacy, poor services, easily curtailed access, highly restricted channels of online work, etc. Perhaps that is simply to say that Cognitive Capitalism creates its own bourgeoise, proletariat and lumpen proletariat, and that the highly centralized nature of Cloud platforms to date suggests that their architecture is Feudal.
Again, the Cloud very well could evolve into a horrible totalitarian world of inescapable stupidity. One version of it probably will. But it will also engender its own counter-hegemonic forms.
Here in California, the privilege of drawing up such lunatic schemes is part of the culture.

Thank you.

    1. 81 notesTimestamp: Sunday 2012/12/16 23:46:00cloudstatenetworkinternetsovereigntygeorgraphymetahavenbrattongeopoliticsdesign
    1. enki2 reblogged this from mthvn
    2. danielaolszewska reblogged this from germansierra
    3. germansierra reblogged this from mthvn
    4. transmediale-afterglow reblogged this from mthvn
    5. deep-cave reblogged this from mthvn
    6. rogierbak reblogged this from mthvn
    7. researchmaster reblogged this from mthvn
    8. thetousle reblogged this from mthvn and added:
      today: interwub-theory.
    9. tvaa reblogged this from mthvn
    10. dddsssmmm12 reblogged this from modernandmaterialthings
    11. modernandmaterialthings reblogged this from mthvn
    12. notational reblogged this from culturaltelemetry
    13. culturaltelemetry reblogged this from mthvn
    14. stacktivism reblogged this from mthvn
    15. aarongemmill reblogged this from mthvn
    16. facadomy reblogged this from andreasangelidakis