“How do you catch a cloud and pin it down?” –Rodgers and Hammerstein, The Sound of Music, RCA Victor, 1965. CD.
Just back from this year’s Transmediale in Berlin. It has transpired that our little talk about the weather was picked up by Die Welt, one of the country’s established (and relatively conservative) national news outlets. The article does a fair job of capturing the general brooding nature of our panel, a result of the festival’s rather imposing theme of datafication and capture all. From what I can tell with my limited German, there are also some little playful, sardonic flourishes in the writing of the article that are indeed in keeping with the nature of the presentations, which included the serving of “weather drinks” by Fran Gallardo and a cloud drawing application by Minka Stoyanova, with which audience members could draw clouds and text that would then drift across the screen above us. Jane Prophet also got in on proceedings with a diffractive reading of Kenneth Goldsmith’s The Weather that interspersed extracts from that text with the affective results of Google searches on biometeorology. There is also an amusing subtext that could be extracted from the article, one in which we are apparently denying our German audience of a certain innocent relationship with the weather and bringing up the war again!
Amidst the serving of drinks and gathering user-generated clouds above us, Christian Ulrik Andersen and I carried out a back and forth conversation in which we shared some reflections on how one might read the conference’s theme of capture all through the weather. It doesn’t look as if there is a recording of the talk, so here is the script for my bits of the conversation (some of which were cut down for time reasons during the panel itself). Bits of this grew out of initial draft writing (“ecologies of capture all”) produced for a PhD workshop in Hong Kong, which we all attended a few months prior the festival itself.
A talk about the weather
//production of the weather (media as environment)
…As Christian is highlighting, there is a way in which the oversaturation or banalisation of things like weather prediction or datafication can gradually wear out or erase the important acts of mediation that help to produce and present something like the weather, or one’s datafied life, as something that can become an object of discussion.
At the same time, one can also consider the way in which mediations of things like the weather produce a kind of tone and atmosphere, literally in the sense of the weather on the particular day, but also in how it is streamed out to us in the form of weather reports, periodically updating weather icons on our phones, user comments on various social networks and during the various moments in our daily lives in which we might typically choose to discuss the weather. This is the sense of weather, and also its mediation, as environment, an atmosphere, as something that sets the tone and affects the daily rhythm of things. Marshall McLuhan is well-known for speaking of media as environments, and in relation to the readings of Goldsmiths’ transcriptions of radio weather reports, McLuhan has characterised radio as a medium that works as a “a kind of nervous information system,” capturing as it does the daily rhythms of weather, news, traffic and other selective soundbites of the day; how radio in McLuhan’s time emitted its own quality as a kind of atmospheric medium which its many users attuned themselves to. He says, “News bulletins, time signals, traffic data, and, above all, weather reports now serve to enhance the native power of radio to involve people in one another. Weather is that medium that involves all people equally. It is the top item on radio, showering us with fountains of auditory space or lebensraum.”
McLuhan’s use of the term Lebensraum is no coincidence. He highlights the role of radio in 1930s Nazi Germany as a kind of tribalising drum beating a somnambulist rhythm. Lebensraum, the German for habitat or literally “living space,” is a term that implies an expansionist mindset of capturing more land, more room, territory and the raw materials that will be a prerequisite for a self-inflating, colonising power. In a similar way, we might want to think about what is at stake if datafication is swelling and becoming in some senses a kind of environment or expansionist lebensraum. What does it mean to live in a present and be moving towards a future in which we are increasingly “couched in data” (to borrow a phrase from Mitchell Whitelaw), a data that will itself always be inflected with certain ontologies and vectors of intention?
To continue with what Christian was saying, this raises questions as to what is the nature of the mediation at play in a particular form of data production, what kind of potential feedback loops is this or that selection and computation of data putting into place, what are the metabolisms and daily rhythms that result from them? In the kind of erasure or banality at play in considerations of things as everyday as a status update on the weather, there are nevertheless ongoing machinations and decisions being made on the part of the dull but strikingly nebulous “gray media” that Fuller and Goffey speak of, machinations that produce and allow for such considerations, many of which can easily expand into other realms, rhythms and spaces of production.
//prediction of the weather (weather as model)
…As both meteorologists and all of us “on the ground” are well aware – weather is stubborn. It is always on, always indifferent; predictable and unpredictable. The latest weather gadgets only serve to heighten this sense of the aporia and gap of an almost-real-time prediction, the nearly always accurate prediction whose inaccuracies are all the more glaring and galling to a mediated will towards accuracy. As Mitchell Whitelaw puts it in a post subtitled Notes on Watching the Sky, “every month brings new data, but the more we know the less certain we become; in fact the only consensus seems to suggest more uncertainty.” In Tarkovsky’s filmatisation of Stanislaw Lem’s sci-fi novel Solaris, we are provided with a series of haunting images of this aporia and stubbornness that results in the gaps between uncertainty and the will to anticipate, including the memorable scene of a rain shower perforating the interior of a childhood home. A memory leak in every sense of the word.
In being placed on a panel tasked with talking about the weather in relation to this year’s theme of datafication and capture all, one task that comes to mind is to attempt to unpack some of the drives and metabolisms of prediction, particularly in regards to their tight relation to capture. In the seemingly unattainable object of desire that is the stubbornness of weather one tends to find an ever stronger will to forecast and predict, a drive that not only produces regimes and machines of anticipation but also prescriptively lays out a justification for capturing all in the name of predicting all.
In the case of programming for and computing the unpredictable, one can trace the metabolism of one particular lineage of prediction back to the very first days of computer production and the stirrings of a discipline called “numerical weather computing.” Beginning around 1904, Norwegian physicist Vilhelm Bjerknes kickstarted the field when he characterised the problem of weather forecasting with that of solving equations that express the laws governing the behaviour of the atmosphere. In the early 1920s, polymath Lewis Fry Richardson proposed how such equations could be solved by numerical computations. By the 1940s meteorology had established itself as a science, and yet forecasting still remained a relative mystery or art form of sorts.
It was on the initiative of John von Neumann – already hard at work with his team on assembling some of the first electronic general-purpose computers at the Institution for Advanced Study – that meteorology moved well and truly into the domain of electronic computation. Von Neumann recognised forecasting and numerical weather prediction as what he described as “the most complex, interactive, and highly nonlinear problem that had ever been conceived of—one that would challenge the capabilities of the fastest computing devices for many years.” In 1948 he forebodingly summed up the task ahead as follows: “The part that is stable we are going to predict. And the part that is unstable we are going to control.”
Von Neumann appointed Jule Gregory Charney – now generally considered as the originator of modern dynamical meteorology – as principal meteorologist for the project and gave him and his team access to the Institute’s ENIAC general purpose computer. The first predictive calculations were carried out by laborious punch-card operations. As the team reported, “the computation time for a 24-hour forecast was about 24 hours, that is, we were just able to keep pace with the weather.” Only two years later, the team could produce the same 24 hour prediction in five minutes.
In preparing some for being placed on this panel, I have only just begun to become acquainted with the intricate workings of things like numerical weather computing, but there are some small local details from these early days of weather computing that might be of interest on an exploratory panel like this. For instance, consider the following little aside from George Dyson’s book Turing’s Cathedral, where he discusses these early days of numerical weather prediction:
while trying to simulate weather inside the computer, the meteorologists were plagued by the weather outside the computer. The “York” refrigeration units continued to become overloaded in the sultry Princeton heat, and during thunderstorms the Williams tube memory often failed. On one very hot day in May there was trouble with the IBM card equipment, and the machine log records: “IBM machine putting a tar-like substance on cards.” The next log entry explains: “Tar is tar from roof.”
Here we see the restless, processual and stubborn entanglement of things, how weather drips into hardware that is busy predicting the same weather that keeps it from functioning properly. One can readily point to many such concatenations of weather and machine, such as the massive amounts of energy required simply to run supercomputers such as the UK Met Office’s double football pitch size systems, whose 12,000 tonnes of CO2 emissions each year inevitably bleed and work themselves into future weather behaviour.
Returning just briefly to Lewis Richardson. In his influential 1922 book Weather Prediction by Numerical Process, Richardson provided a notable (though perhaps predictable) piece of design fiction and dreamy blueprint for a model of capture all. At the end of the text, Richardson permits himself to envision a speculative future for weather prediction that involves “partitioning the earth’s surface into 3,200 meteorological cells, relaying current observations by telegraph to the arched galleries and sunken amphitheater of a great hall, where some 64,000 human computers would continuously evaluate the equations governing each cell’s relations with its immediate neighbors, maintaining a numerical model of the atmosphere in real time” (Dyson, p.216). Here is an extract from Richardson’s description (keep in mind that “computer” at this time refers not to a machine but rather an individual using a computing device such as an abacus, logarithmic table or slide rule):
Imagine a large hall like a theater, except that the circles and galleries go right round through the space usually occupied by the stage. The walls of this chamber are painted to form a map of the globe.
Myriad computers are at work upon the weather of the part of the map where each sits, but each computer attends only to one equation or part of an equation. The work of each region is coordinated by an official of higher rank. Numerous little ‘night signs’ display the instantaneous values so that neighboring computers can read them. Each number is thus displayed in three adjacent zones so as to maintain communication to North and South of the map.
From the floor of the pit a tall pillar rises to half the height of the wall. It carries a large pulpit on its top. In this sits the man in charge of the whole theater; he is surrounded by several assistants and messengers. One of his duties is to maintain a uniform speed of progress in all parts of the globe. In this respect he is like the conductor of an orchestra in which the instruments are slide rules and calculating machines.
Already in the 1950s though, Charney and his team were finding that even the exponential power of computers would struggle to get a hold on the dynamic complexity and unpredictability of the weather. Referring to Richardson’s imaginative model, Charney suggests that even if one increased the granularity of the kind of cellular Laplacian lattice suggested to a mesh size smaller than one millimetre, it would still not be possible to make accurate long term forecasts, because “an error introduced by a turbulent eddy can grow exponentially, from 1mm to 10km in less than a day, and from 100km to planetary scales in less than a week.” Even a preeminent cybernetician like Norbert Wiener felt it folly to pursue medium to long-term weather prediction, let alone something like “infinite forecast.” But the weather forecasting beat goes on, as does the expanding lattice of distributed, planetary computing, the model here potentially becoming itself a hyperobject – like the very hyperobject it aims to model.
One metabolism of capture all would be this ready one to one, self-referential model of parallelism that the will to predict prescribes. And such a logic can be readily mapped onto and infused with other wills and logics. In the case of the weather, we find such regimes of prediction fed by two particularly strong, existent wills or regimes: that of economics (with better weather prediction ostensibly enabling a further maximising of profit returns) and that of crisis (avoiding the catastrophes of “heavy weather”). Crisis in particular seems to have taken hold of the current moment. As Wendy Chun (“Crisis, Crisis, Crisis, or Sovereignty and Networks”) points out, with its prescribing of one catastrophe or another, crisis has in the last few decades proved particularly effective as a driver for further prescribing codes and coded logic, whereby “computational codes are increasingly privileged as the means to guarantee ‘safe living’ because they seem to enforce automatically what they prescribe.” The point here is to again emphasize the way in which prediction is also production. It involves decision making, and the constant material infusions within these decisions stubbornly set in motion the need for further decisions, any of which might be automated, debated or glossed over.
//war and weather (prediction as production)
…In emphasising earlier how prediction is production, one can also consider the flip-side of this suggestion of Latour’s: how matters of concern can produce seeming matters of fact. We were discussing Wendy Chun’s writing on crisis and the way in which computational codes are increasingly privileged because they automatically enforce what they prescribe – namely protection from and rapid response to crises. Philosopher Brian Massumi (“National Enterprise Emergency: Steps toward an ecology of powers”) goes so far as to outline a way in which today’s modes of crisis-oriented governance establish what he describes as a “full spectrum” and continuum of “war and weather.” In making this link between the US government’s response to weather and the war on terrorism, Massumi wants to outline how there is a shared ontology between the two, one in which war, like weather, stands in relation to an always immanent, unknowable threat, whether it be a hurricane in Katrina or a terror threat from above. This sense of threat radiates throughout and produces crisis-oriented modes of being in an atmosphere of ongoing disruption and response. As the tagline from a show of Martin Beck’s aptly highlights, “nothing better than a touch of ecology and catastrophe to unite the social classes…”
Massumi suggests that the particular power at play in such an ontology of immanent crisis is the way in which this ontology sucks the futural into the present. In this way it fuses ontology (what is) with politics (what ought to or should be) in such a way that this mode of being is always actively trying to control the potential and unpredictability of emergence, what Massumi calls an “infracolonisation” of the “prototerritory.” Here we might remember that for many of those working on the early numerical weather prediction projects, the ultimate goal was not only “infinite forecast” but the ability to actively control the weather itself. Recall also the von Neumann quote mentioned earlier of the cybernetic use of prediction: “The part that is stable we are going to predict. And the part that is unstable we are going to control.”
A crisis-oriented politics of prediction readily relies on the spectres of threat it conjures up, at the same time that it can very easily conjure up spectres all of its own. Consider, for instance, the case of selective US policing statistics, whose “matters of fact” take little account of the many subcurrents and age old feedback loops of fear and racism that have informed their production. In these systems, a particular spectre of criminality has been projected upon black individuals from the very beginning. Policing forces then diligently pursued this spectre for decades, dutifully marking up its datasets and feeding the same metabolisms of fear and prejudice that projected this threat.
Is it any wonder then that in today’s generalised atmosphere of immanent crisis and threat, we now arrive at a situation whereby the deployment of lethal choke holds and the militarisation of police departments are treated as “only natural” responses by those in power. In such cases we see how an “ontopower” (as Massumi terms it) – in this case the militarised US economic war machine – readily makes itself at home in the ground it so ably prepared. One is reminded of what Peter Sloterdijk describes as a “military climatology,” an air conditioning that actively conditions. Consider the conditioning and full spectrum of war and weather that is highlighted in a hashtag like #ICantBreathe. Or the militarisation of something as seemingly harmless as the blue sky in regions such as Afghanistan, where inhabitants describe how sunny, cloud free days have for them now become a considerably less cheery phenomenon given that blue skies are the optimal conditions for military drones to operate in. Or the struggles on the streets that we witnessed during of our stay in Hong Kong for the workshop, where the humble umbrella becomes a resonant symbol for the Umbrella Movement, shielding its bearers not only from rain but also from the batons, water canons and tear gas of powers that seek to extend themselves into the environment. An environment that a host of powerful actors – governments, private entities, algorithms, etc. – are continuously working to explicate and capture in predictive, comprehensive fashion.
Again, of particular interest for this panel is how talking about the weather can act both as an allegory for but also as a potentially correlative layer of abstraction for talking about datafication and logics of and practices of capture all. We should regularly reckon with the hallucinatory potential of predictive datafication and its easy slide towards overdetermination, self-sufficiency and, indeed, self prophecy. In Whitelaw’s words, “When data swarms and flows with apparently inherent dynamics, it’s easy to forget how data is created, or even that it is created. This is especially true when the data source is the network itself; self-referentiality gives an impression of self-sufficiency, again a world in which data is given, rather than made.”
And also to let Chun’s message disperse and accumulate… like the weather. That in an age of automation, the experience of the unpredictable, the unquantifiable or undecidable is not a crisis to be resolved, but rather converges into instances of responsibility, decision and experience. It is an event. Often a political one. As Chun points out, it reminds one that, “we must also deal with — and emphasize — the precariousness of programs and their predictions.” To frame the stubborn gaps and aporias of prediction as calls for responsibility, as matters for concern – but also as spaces for the kind of resistance, speculation and play that dwells in the unpredictable.