buckminster fuller's world game and its legacy
Andrew Kirk, Counterculture Green: The Whole Earth Catalog and American Environmentalism (Lawrence: University Press of Kansas, 2007), 109
4 Felicity D. Scott, Outlaw Territories: Environments of Insecurity/Architectures of Counterinsurgency (New York: Zone Books, 2016).
5 Richard Ashley, âThe Eye of Power: The Politics of World Modellingâ, International Organization 37, no. 3 (1983), 495â535. Amy Dalmedico and Matthias Heymann, âEpistemology and Politics in Earth System Modelling: Historical Perspectivesâ, Journal of Advances in Modelling Earth Systems 11 (2019), 1139â1152.
6 Eva Lövbrand, Johannes Stipple, and Bo Wiman, âEarth System Governmentality: Reflections on Science in the Anthropoceneâ, Global Environmental Change 19, no. 1 (2008), 7â13
intro:
Even if Fullerâs plan had gone ahead, the Gameâs architecture and technocratic approach would have restricted its sought-for deliberative process.
The World Gameâs proposals to model planetary 'life support' systems, redistribute resources, and establish a distributed management network foreshadow the needs and dilemmas of current earth system governmentality. Far from replacing politics, and despite Fullerâs claims, the World Game was an early attempt to introduce planetary, science-led action and long-term thinking to a non-expert public. Discussions of planetary management and stewardship or âearth system governanceâ now shape our politics, as negotiations, conflicts, and alliances become âclimatisedâ.
the latest [iteration]\ being an updated World Game Workshop under development at the Schumacher Center for a New Economics where it aligns with projects such as the Green New Deal. It looks again at the migration of ideas and personnel, technologies, game architectures, gameplay, and information design and other techniques of visualization. With hindsight, the World Game might appear naĂŻve or utopian or technocratic. It was likely all, but not only, those things, and on closer examination, we might find that we share the World Gameâs problems, if not always its solutions.
about:
Two major architectural projects were proposed for the World Game, namely the US Pavilion for Expo 67 in Montreal and the World Resources Simulation Center (WRSC) planned for the Edwardsville campus of Southern Illinois University (SIU), envisaged as the first stages of a global network of command centres
For Fuller, although planetary systems might be âextraordinaryâ and âunpredictableâ (or âsynergisticâ, in Fullerâs jargon â what we might now call emergent), and not reducible to statistical or probabilistic analysis, they never work âin ways you canât modelâ. âThe most important and useful work Iâve been able to do will be achieving this return to modelabilityâ, he declared in a 1971 interview.
The World Gameâs âsystems approachâ, which Fullerâs research assistant John McHale (see Chapter 4) called âa new social instrument for complex planningâ, aimed to define the functional requirements of planetary systems and identify âwithin their design various feedback sub-procedures which regulate the system towards the desired optimal end functionâ.16 This approach adapted the general systems theory (GST) of the Austrian biologist Ludwig Von Bertalanffy, first outlined in the early 1950s,17 and combined it with economist John von Neumannâs game theory, which studies how strategic interaction between rational economic actors produces intentional and non-intentional outcomes.
Expo 67 and the WRSC aimed to visualise and operationalise this âbiggest systemâ in their dome architectures and immersive displays. In short, they aimed to translate GSTâs âscience of wholenessâ into an informational architecture, or what Fuller called âa total information integrating mediumâ, that was global in ambition
This parts/whole correlation made the World Gameâs strategies âas comprehensive and correct as is presently possibleâ.24 Knowledge of the whole, or what Fuller called the âbiggest systemâ, meant that players could âautomatically avoid leaving out any strategically critical variablesâ
Among clippings collected by Fuller between 1967 and 1968 was a section by the psychologist Robert H. Davis from Psychological Research in National Defense Today, a technical report published in June 1967 by the US Army Behavioural Science Research Laboratory, which criticised game theory, especially its zero-sum games, for assuming both that players are âperfectly rational and motivated by greedâ and that cooperation is always a loss rather than a gain. Instead, 'mixed-motive' games introduced uncertainty about the player, âhis value system, his strategyâ, and allowed that cooperation provides mutual beneft.30 As noted by Max Ackerman, one of the graduate assistants employed at Central Illinois University for World Game research from 1969, the Game âgoes beyond zero-sum, non-zero-sum, and mixed-motive theory and it may be a new order of game theory: meta-motivational gamesâ.31 The World Game was a strategy game, but promoted cooperation to answer the basic needs of all aboard Spaceship Earth and thus promoted self-actualisation, or what Abraham Maslow called âmetamotivationâ.32 With their basic needs met through the World Game, the astronauts of Spaceship Earth could devote themselves to more metaphysical or spiritual pursuits
physical installations:
show a rectangular space frame truss raised on four seventy-foot-high âtensegrity columnsâ with suspended catwalks. At its centre was a large, low relief Dymaxion map to show historical and live data streams. Above the catwalks, from which visitors could view scheduled displays, were personal consoles through which players could request displays, print relevant data sets, focus upon regional problems and trends, ask questions, or test solutions. McHale summarises: The whole map complex would be treated as a dynamic display surface capable of showing a comprehensive inventory of the worldâs raw and organised resources, together with the history and trending patterns of world peoplesâ movements and needs.35
Whereas the Temcor dome was largely a workspace, with a display space at its centre, the larger dome aimed at a âmarriage of man/displayâ (Figure 1.2
Such an interface would intensify the sensory experience and expand the awareness of design scientists, assist their long-term strategies for resource utilisation, and advance their capacity for planetary stewardship. The WRSCâs âbasic hypothesisâ, Fuller asserts, is that decision-making and strategising improve in direct proportion to the intensity of the visual environment in which they occur. 43 With the WRSC, as Mark Wigley notes, Fuller pushes architecture further into the electromagnetic spectrum, as it dissolves into networked image: âImage has become envelope ⊠Yet again, architecture as the space of images and the science of architecture as the art of the imageâ
the World Game were necessary to understand our âdynamic, ever-regenerating environmentâ and to âplug in our sensory awareness mechanisms to the switchboard of âuniverseâ, to get in sync with the metabolism of this spaceshipâs environmentâ
As McHale noted, the compound curvature of such geodesic domes produced âfinite systemsâ, suited to viewing the world in its entirety.45 Necessarily, this was a game environment, which would minimise risk to playersâ decisions. [The WRSC] must be perceived, truly, as a âgameâ environmentâ. Errors of judgement must not be catastrophic for players or spaceship passengers. ⊠It must be an environment, in short, which allows him, the decision-maker, to learn ⊠to hone his awareness of what is possible, of alternative courses of action.
By 1980, Fuller declared, the WRSC would have the technology necessary to input data rapidly and spontaneously and to project images across the âentire inner skin surfaceâ of its larger dome. Automated and non-automated pattern recognition systems would identify correlations and trends to planners.51 As such, the WRSC provided the architectural template for a command centre to pilot and manage Spaceship Earth, in which computers would replace the messiness and fallibility of human decision-making with objective assessments and strategies that would command universal agreement.52 With time, Fuller believed, it would be possible to extrapolate out playersâ biases and âstylesâ from the assessment of resource scenarios.53 Seated in their contour chairs, comprehensive design scientists would be the responsible crew of an increasingly automated and streamlined Spaceship Earth
Visitors to La Ronde discovered materials for the Game alongside a fairground, theatre, musical performances, sports events, art workshops, a puppet festival, and âhappeningsâ
design students tested out displays and interfaces for planetary trends in energy production and use, infrastructure, agriculture, resource extraction, and so on. The theme of their first exhibition, beside the Orangerie in the Jardin des Tuileries, Paris, from 2 to 9 July 1965, was âworld literacy regarding world problemsâ
sought to identify global problems, their interrelations, and their possible solutions.
Beside this were panels with photographs showing these same problems accompanied by texts offering design solutions. There was also a Geoscope, which allowed a viewer to enter and see dynamic correlations of trends and resource data
Trends Exhibits such as this were pedagogical tools, prototypes, like the World Game itself, for the âflexibly dispersed learning centers for various purposes interlinked to centralised libraries and other major facilitiesâ demanded by McHale for the purposes of world design science.61 Without these major facilities, however, the World Game existed as a distributed network, where different constituencies would play with future scenarios without end and therefore without âone best wayâ. Although the World Game was supposed to integrate scenarios towards an optimal future, it gave no priority to any one scenario. It was, in Fullerâs words, an âaggregate of non-simultaneous and only partially overlapping eventsâ.62 Each scenario modelled by World Game was, in practice, suboptimal. At the heart of the World Game was an ambivalence. It was an optimal scenario-building game in which nobody wins, or, to extend Fullerâs metaphor of mountaineering, a mountain without a peak.
*Rather than resemble the system that it sought to replace (that triad of Manichean sciencesâ â operations research, game theory, and cybernetics â criticised by Peter Galison),64 âFullerâs project reorients the âsystemâ from within by playing games with the very idea of a graspable, collective futureâ.65 To do this, the Game displaced politics to cartography (and, I would add, to information design more broadly) to cognitively map the present and extrapolate multiple futures. Its models were provisional and experimental, tested through its data visualisation and its scenarios, which struggled to visualise this âbiggest systemâ. Likewise, its problem-solving tended to uncover further, often more complex problems
The World Game and Its Critics
Fuller abandoned politics for design science and the World Game forsook the âpolitical expedient of attempting to reform man and commits man to reforming his environmentâ through more efficient and egalitarian use of resources aboard Spaceship Earth
Fullerâs abandonment of politics has drawn much criticism. His appeal to universal humanity and planetary togetherness masked uneven responsibility and consequence of the environmental crisis, just as freeing the worldâs resources from private or common ownership, land rights, labour costs, and other externalities,
assumed US sovereignty over government, infrastructure, and computing technologies needed to establish this âtrue direct planetary democracyâ. More broadly, design for the purposes of âplanetary housekeepingâ (Wardâs phrase), more commonly known as sustainable development, has often meant intervention into and management of human populations, settlements, and habitats that eliminates vernacular and endogenous sustainable design practices in the name of development.74 What is more, the ethnocentric and classist biases of the Spaceship Earth metaphor became evident in concerns that the planet might soon exceed its âcarrying capacityâ, prompting a moral economy in which only optimum (or optimisable) organisms and systems would justify their place aboard. The âcabin ecologiesâ at the heart of the spaceship metaphor, too, brought ecology into line with the Cold War theatre of operations.75 Nisbet is especially scathing, claiming that World Game transforms ecological responsibility for most into âsimply following the rules predetermined by the top-down principles of global effciencyâ
ullerâs faith in the objectivity and effciency of computers, âlegitimated new forms of controlâ, Massey argues.78 What is more, the Gameâs âeffciency-optimisingâ architecture and âone best wayâ approach to resource redistribution would have restricted the deliberative process even as it sought to globalise that process. The Expo 67 proposal appears to exemplify that coupling of âcontrol in open sitesâ with the âquest for âuniversals of communicationââ that once made Gilles Deleuze shudder.
he World Game was suboptimal, despite its grand architectural designs. In addition, its supporters and players often held that its redistributive and egalitarian ambitions could not be reduced to top-down management structures. This becomes more evident once we turn our attention to where and how and with what the World Game was played. We fnd that in the World Game Seminar, to which we now turn, Spaceship Earth was crewed by amateurs who knew very well their limitations and blind spots, but who muddled along, nonetheless. Indeed, muddling along continues to characterise much of the World Game up to the present
Design and Gameplay of the Seminar
The four steps to the World Game were: (1) Inventory, (2) Trend, (3) Strategy, and (4) Scenario-Building. Players first identifed a resource to research, then projected that research into future trends, to provide the worldâs âvital statisticsâ. To identify trends was to bring the inventory âout of a static state into a dynamic oneâ and to help players to infer general principles. 10 Once players found principles, they then developed scenarios to provide a synoptic view of planetary developments
The principal game rule was cooperation toward more comprehensive and accurate scenarios. The only penalty was disqualifcation for any player who tried to dominate another or to introduce competition
The playersâ first move was an âEnergy Scenarioâ, âto design the blood system [of the Earth] so that everybody was fedâ, Schlossberg wrote.14 Players plotted the worldâs energy networks and sought to make them more effcient to meet basic human needs. They calculated the latter in internal metabolics (calories and protein) and external metabolics (kWh), then graphed these from 1965 to 1980 according to current provision to show that most of the worldâs population will only reach a bare minimum of energy. Their redistribution of off-peak energy, along with more effcient production and transmission, would bring âeveryone on earth to a minimum of 2000 kwh per year by 1980â.15 This provision, equal to Europe in the late 1960s, was far more than the bare maximum of 1,500 kWh per capita per year needed to make humans âtruly consciousâ and able to function collectively as a âworldwide organismâ, one player declared.16 Or, as Schlossberg wrote, When we meet the physical needs of all of humanity then the metaphysical [thought and invention] becomes the realm of activity â fear gives way to longing â survival to curiosity â and the spaceship earth begins to care for its crew.
Other students presented on agriculture, food distribution, power infrastructure, mineral and metal resources, levels of industrialisation, and alternative energy sources (wind, solar, geothermal, and tidal). They asked themselves practical questions about needs, materials, and distribution systems, and more speculative questions, such as âHow far ahead can we conceive a future lifestyle?â and
When is a game a game?â18 They were troubled, Schlossberg noted, by both the amount of data that was required, and to which they had no access, and with the âimaginative leapâ needed to conceive of Spaceship Earth. After these first moves, they discussed âpossible synergetic scenariosâ that ranged from naĂŻve to utopian to prescient. These included what we now call âsmartâ mobile housing, electric cars, biofuels, âthe laser beam transmission of power and informationâ, and âthe possibilities of a world guaranteed annual incomeâ. They also speculated about what humans would do when they were freed from the âdrudgery of having to earn a livingâ
For information design, players used a four-foot-high and sixty-foot-long tri-axial grid that showed energy needs, population, population density, access to water, and âvital statisticsâ of Earthâs inhabitants and two ten-by-ffteen foot Dymaxion maps upon which were placed acetate overlays to show global resources, networks, and infrastructures. Mimicking the jargon of Fuller, Gabel described these as, The graphical, functional and mathematical orderings and simplifcations of the omni-complicated and interrelated processes of the World. The conceptual simplifcations of ârealityâ into the vectors of an interacting process which can be dealt with on a scientifc basis.20 This information design was fundamental to the gameplay of these amateurs âin a feld where there are no professionalsâ. Players used these maps and grids as what Johanna Drucker calls âknowledge generatorsâ, or Graphical forms that support combinatoric calculation. Their spatial organisation may be static or mobile, but their spatial features allow their components to be combined in a multiplicity of ways. They make use of position, sequence, order, and comparison across aligned felds as fundamental spatial properties.21 Both Dymaxion map and grid were generative and dynamic. They provided players with working models of planetary systems
The Dymaxion projection's advantage is to reduce distortion and to allow for three-dimensional and, when unfolded, two-dimensional representation. For Fuller, the minimal distortions of the Dymaxion projection provided players with a âfaithful background upon which to show dataâ.
Following Denis Wood and John Fels, we might consider the propositional logic of this âfaithful backgroundâ. Like all maps, the Dymaxion projection was a vehicle for âthe creation and conveying of authority about, and ultimately over, territoryâ, whose authority is âthe social manifestation of what the map presents as its âintrinsicâ and âincontrovertibleâ factualityâ.26 Evidently, Fuller believed the factuality of the Dymaxion projection to be incontrovertible. This is what made the map a âfaithful backgroundâ. Likewise, he and the players assumed the factuality of the data presented upon this background, which gave the World Game authority. The Dymaxion projection presented a world unifed as a game board (or theatre of operations) and as a closed system. Wood and Fels argue that maps are âsignifcant players ⊠in the world of actionâ.27 Maps appear as players in a complicated social game defning the relationship of our species to the rest of existence. Pretending to be no more than scorekeepers, maps stand revealed as more like the ball, the very medium through which the gameâs moves are made.28
Fullerâs projection would visualise the world as an integrated system not for the purposes of geopolitics but to redistribute resources that were bounded, integrated, and exhaustible. What is more, in anticipation of the now-current ecological metaphor of a blue planet, the Dymaxion projection showed a unifed ocean. Life declared that the Dymaxion projection showed the world as âone continentâ, as the sinuses were placed through the oceans to maintain continuity of land areas. This was a principal feature, explained in detail in the patent (Figure 2.2).32 Rather than a continent, however, Fuller would later call it a âone-world island in a one-world oceanâ.33 In truth, it is more of an archipelago than an island
Its âepistemological pretenceâ is clear: the planet is knowable and available as a game board and arena of voluntary, global action. In more detail, however, as polyhedral globe, it offers âan object of contemplation, detached from the domain of lived experienceâ.35 With a globe, argues Tim Ingold, âTo know the world ⊠is a matter not of sensory attunement, but of cognitive reconstructionâ
Another source of uncertainty in the World Game is its use of data. In addition to the acetate overlays placed upon a âfaithful backgroundâ, Seminar players produced what Schlossberg called a âtriangulated interactive gridâ to show correlations between data sets. This provided a âworking storehouse for informationâ on energy needs, population, population density, access to water, and so on, divided into 120 sections (right diagonals) and 23 âmajor areasâ (left diagonals)
This chart was a triangular grid on which one of the three axes were the 22 major geographical areas of the world and their individual countries. The second axis consisted of, in fve year increments from 1965 to [2000], fgures in population, population density, calorie and protein intake, total kwh, metric tons of coal equivalents and energy slaves. The last axis could indicate up to 20 possible world-trends for each area and country. We use thirteen: fossil fuel potential, life expectancy, mortality rate, arable land, housing, amounts of copper, aluminium and steel, food literacy, reinvestable time and hydropower.40 The players used this grid as âan object to think withâ, a computational device for building scenarios
the World Game includes no opportunity for players to refect upon how this data is collected or processed, its accuracy, its relevance to defnitions of human needs, or to what extent the design parameters of the Game predetermined the problems in response to which this data would be mobilised.
Returning to Drucker, we might argue that the data shown in this grid âare capta, taken not given, constructed as an interpretation of the phenomenal world, not inherent in itâ.42 The grid depended upon already parameterised data sets, themselves the product of an âinfrastructural globalismâ discussed by Paul N. Edwards in his history of how meteorology and climatology constructed an extensive infrastructure, a âvast machineâ to collect, model, and disseminate environmental data.
However, the World Game includes no opportunity for players to refect upon how this data is collected or processed, its accuracy, its relevance to defnitions of human needs, or to what extent the design parameters of the Game predetermined the problems in response to which this data would be mobilised
Gabel asserted that in a closed system all variables could be controlled,45 but the open structure of the grid could not model a closed system because it could not include all those âstrategically critical variablesâ of which, Fuller and other Game advocates believed, automated computation would give a full account. As a result, the grid modelled open systems where the critical character of variables could only be ascertained once they were correlated with other variables. Added to the limitations, omissions, and biases of the playersâ data sets, playersâ use and representation of data through information design further introduces uncertainty and complexity. Crucially, the Seminar players seemed to know this.
The players knew the limits of their information, their models, and their scenarios. The triangulated interactive grid and the accumulated acetate overlays show an iterative and complex systems approach, which suggests that although the Seminar players were encouraged to imagine themselves as the crew of Spaceship Earth and often deferred to a future environment of total, real-time information, and computational power, they nonetheless few in twilight, as it were, feeling their way from one Black Box to the next. Rather than a failure of the Game, this returned players to the messy world of politics, education, and dissemination, and to a non-holistic modelling of planetary systems
When flying with Fuller and Michael Binelli (a geodesic dome engineer) in a light aircraft from St. Louis to Carbondale, Youngblood realised at some point that they were flying blind. âI was thinking how appropriate it was to be with Bucky this way, trusting our lives to the very design integrity that he has spent a lifetime defending as humanityâs only hope for successâ. In this situation, fying blind celebrated computational assistance and reliability.47 During the Seminar, however, players few blind because of their modelsâ limitations and the character of their visualisation strategies, a blindness that would have been eased only partly by greater computational power
Again, it seems that uncertainty and complexity were structural and productive rather than incidental, because still the Seminar players modelled and played. In doing so, they made, to borrow from Bruno Latour, a âpassage through connectionsâ without assuming a Whole or a âbiggest systemâ. They performed âconnectivity without holismâ.48 This is signifcant because, if the Earth System itself (what James Lovelock famously called Gaia) is âanti-systemicâ, in the sense that there is no superior level above and beyond its parts, then the metaphor of a spaceship, with its pilotâs chair, mission control, and design scientists, fails.49 Instead, players faced the problem of how to get along with complex planetary systems that self-regulate as they co-evolve without âforesight, planning or teleologyâ.50 In such a case, the Game would be no cybernetic apparatus worthy of name, it seems, as there would be no âtiller of environmental controlâ.51 Yet, as noted above, perhaps there is another cybernetics in operation here, with its Black Boxes and âontology of unknowabilityâ, a more experimental cybernetics.
The concept of a Black Box, which arose to describe problems in electrical engineering, extends to any systemic entity that performs some task by means largely obscure to an observer. To paraphrase Ross Ashby, when faced with a Black Box, the question âWhat is in it?â becomes redundant. Other questions are more pressing, such as âHow should an observer proceed?â âWhich elements of the Box are discoverable and which are not?â âWhat methods can be used to investigate the Box?â53 Perhaps the only way to engage unpredictable Black Boxes, what mathematician RenĂ© Thom called their âcatastropheâ, is to play with them.54 Seminar players admitted that many of their models were reductive and lamented their lack of total information about the systems that they researched. Yet, they accepted their ignorance and played on. They planned to run simulations to see how different systems would respond to their moves, but again, were reluctant to do so because they lacked information. This uncoupled the Game from what McHale called Fullerâs âfanatical belief in complete over-all pre-planningâ and made it into a distributed pedagogical tool
The Game provided an âaerial view of ecologyâ. Players were like foresters looking from a tower to see âwhat is occurring in the forestâ.57 The WRSCâs âtotal world enveloping displayâ supported this planetary metaphysics as it elevated the design scientist in his âpilotsâ contour chairâ to the status of a sovereign decision maker. It is, of course, a poor ecologist, or a poor student of environmental studies, who investigates the forest ecosystem only from above the canopy. The grid, as noted, modelled systems differently. It offered only accumulated parts, and so returned players to the âforest foorâ (as did the accumulative and distributed nature of the Gameâs learning centres)
Their play offers a different modelling of earth systems to the mainstream environmentalism in the United States in the late 1960s and 1970s, which tended to treat the earth as a closed, self-regulating system. The players knew the limits of their scenarios. They also knew that they were amateur pilots. During the public demonstration, one audience member remarked that, however much the players overlooked these political conditions, World Game would have to be implemented. âHow can this be done without politics?â he asked.58 Gabel replied that the players were playing a âmountain-climbing gameâ to show that they could get everybody to the top as rapidly as possible. In his diary, Schlossberg recalls: âwe answer that we were not planning implementation but only exploring whether it is feasible to consider that man can be provided with physical support by the year 2000â.59 Showing that this was possible, the players believed, would prompt action without waiting for crisis
Pace students explore the interdependence of natural, physical, and human resources on the global scale. Perceived as a game, each âplayerâ learns to integrate his own experiences and research with those of others.66
requesting his support for a World Game handbook.65 Winsey organised an undergraduate course Dynamics of Change at Pace College, based on the World Game (Figure 2.3). She describes the course thus: Pace students explore the interdependence of natural, physical, and human resources on the global scale. Perceived as a game, each âplayerâ learns to integrate his own experiences and research with those of others.
They presented the World Game as a ânew educational simulation systemâ that would provide government executives with âa complete reality of what we mean by âWorldâ, of what problems that world faces and what strategies might solve them. The brochure quotes at length from Fullerâs dismissal of the Malthusian doctrine and the militarised geo-politics that derived from it