bnr#2 => Invisible Airs, open-data book stabber, Bristol Council Chamber
Pits to Bits
Artefact Festival, Out of the Ground, 2011
Mattew Fuller - Interview with Graham Harwood
This interview follows on from a project called “Coal Fired Computers (300,000,000 Computers - 318,000 Black Lungs)” carried out in Newcastle in spring 2010 for the AV Festival. The project, by Graham Harwood, Matsuko Yokokoji with Jean Demars involved a means of producing a physical diagram between components in production as they undergo transformations across different kinds of time, politics, matter, knowledge, and vitality. The project found a way of working with such things that was particularly powerful. The interview begins with a discussion of CFC but also moves off into databases and a certain understanding of their material force.
One thing we don’t cover in the interview is the detail of the Coal Fired Computers project’s work with miner activists, including the inspirational Dave Douglass. (See information on his memoirs here http://www.freedompress.org.uk/news/2010/03/07/new-book-ghost-dancers-by...). More of this can be found in a booklet about the project here (), including links to all the groups involved.
The interview was carried out by email in May and June 2010.
Matthew Fuller: If we are to list the visible components of the project it would go something like this: pile of coal -> fair ground steam engine -> power transformer -> computer / software -> air compressor -> blackened lungs. But there are a lot of things missing from that set of components that are integral to the project, what are they and how do you see them?
Graham Harwood: This list should really start with Jean Demars who set up the collaboration with the miners, being French and youngish gave him little knowledge of the UK class struggle of the miners strike in the 1980's. He use his political enthusiasm, critical analysis to re-examine the strike in the context of globalisation with the people who struggles against it back then. If they had not been defeated they would have picketed every port to stop the everyday atrocity that powers our world.
Matsuko on the other hand is always the leader of all things organised, productive, efficient. I'm much too lost in my own space to ever accomplish much on my own. Matsuko takes this raw material of todays obsession and forms it into graphics, budgets and how the exhibition will look and act. She does not like to talk in public she just likes getting on with things.
MF: So perhaps you have some considerations about how the kinds of work we are speaking about involve collaboration, within this ‘list of components’, about yourself, about Jean and Matsuko and the kinds of collaboration you have been developing with them and others, with Richard Wright and earlier, as part of the group Mongrel
GH: Collaboration is a necessary minefield, if you’re interested in the place where media systems and the social clash, unfold and get really dirty then they are mandatory. It would be far too easy to claim everything under my own authorship but anyone with an ounce of nous would see that all imaginings are dependent on the context in which they arise or are seen, I'm just a bit more explicit about that.
MF: So to return back to the list of components…
GH: Then the place, Newcastle Upon Tyne a former mining and industrial district in England's North East with a geographical propensity for coal, then - people maybe – firstly the miners who displayed their literature and spoke about their lungs, then the Discovery Museum, with it's cleaners who were also Miners, it's exhibits, Charles Parson's 1884 steam turbines the descendants of which produce the worlds electricity. Then there is the 3000 visitors who had some familial relationship with Coal mines, lung disease.
It could be said that coal dust gets into everything. Sealed into the lungs of miners it forms visible blue streaks, like veins of coal. According to the World Health Organisation, 318,000 deaths occur annually from chronic bronchitis and emphysema caused by exposure to coal dust. The common perception is that wealthy countries have put this all behind them, displacing coal dust into the lungs of unrecorded, unknown miners in distant lands, however coal returns into our lives in the form of the cheap and apparently clean goods we consume.
Coal fired energy not only powers our computers here in the UK, but is integral to the production of the 300,000,000 computers made each year. 81% of the energy used in a computer's life cycle is expended in the manufacturing process, now taking place in countries with high levels of coal consumption. The UK currently produces less that one third of the coal it uses, importing the majority of it and therefore displacing 150,000 tons of coal dust into unknown lungs.
Then there’s the recent histories of media and my preoccupation with it's interrelation to death. But more about that later.
MF: Part of the interest of the work it seems to me is that in a context in which ‘the world is too complicated to describe or to understand’, it provides something like a diagram, or a formula which shows how a series of things are joined together, how certain kinds of momentary connections are made, but does not renounce the difficulty of such work of abstraction, and really gets into the very different kinds of qualities, materialities, knowledges, histories and powers of the things that are nevertheless joined. The work doesn’t make it’s argument through affirming a set of categories but by drawing out these formulae in uncannily clear ways through this process of conjunction. What kind of process of searching and sifting goes into making this diagram of formula?
GH: It's probably best if it starts as a joke, a completely unrealisable funny fantasy that will not go away. Yet every time you tell someone new, you can see it connects and they recount stories or expressions that affix to the initial idea. Next you formulate a contraption who's structural operations can leak out into the domains you want to contest, play with and the areas people have spoken to you about. I leave as much work undone as possible, so as it unfolds it can contest the space in which it's showing and the space/geography can contest the contraption. As the physical/code machine begins to take shape it creates complex negotiations, apprehensions and upsets as the speculation grows.
Then there is fear, violence and the dead. I need to be scared of what I make, It needs to put me in embarrassing, difficult, hurtful and potentially violent situations or it's just not interesting.
MF: A number of projects you have been involved in over recent years work with ‘primary’ raw materials, stuff dug out of the ground and refined, such as the metals aluminium – in the film ‘Aluminium’ presented at Manifesta7 - and coltan – which is explored in the various iterations of the Tantalum Memorial and Phone Wars projects. What are the stakes in this coupling of elementary or primal materials with computational systems?
GH: What interests me is the material’s ability to recursively unfold possibilities, transforming the flesh, the social, political and economic. Essentially what a material makes possible and what it shuts down when it's ripped from the earth and it's context and contaminates human ecologies.
Simultaneous with the material properties, they are contagious concepts that move around technical cultures growing on the jelly of science embedded with it's own philosophical speculations about the nature of the world.
The materials also come into to existence as a force when the political, geographical and economic situations are right for them to do so. Aluminium ‘needs’ Italian Fascism to ‘need’ a national metal, It ‘needs’ Italy to lack coal, iron and have bauxite instead. Coal for a long time in the UK was dug from deep cast mines and the shafts required pumping out which creates the steam engine which in turn requires more coal and more labour. Tantalum ‘requires’ political unrest in the Congo, kids playing Sony games.
Then there is the flesh and death, the material bends the flesh to suit itself, miners lungs, bones shattering, light, fast munitions ripping into countless bodies, rapes and murders.
MF: And the place of mines in our clean modern world?
GH: Mines are everywhere in everything once you start looking, you cannot have humans without them, we seem to be preprogrammed to burrowing blindly underground like worms. The main difference from us and worms is that we have a compulsion to burn or explode whatever we find.
For the CFC project I wanted to look at the steam engine as a physical and conceptual machine simultaneously in a popular setting like the Discovery Museum.
In the 19th century the great engines of change at that time were built around coal-fed steam. This was a society that rested on its mines; its products dominated life and determined its inventions and transport infrastructure and its politics. In this way, the coal mines of England recursively transformed the bodies of those who touched them and redirected large parts of its society to feed its machines.
This is still the case, but the mines and production are displaced to India and China. It's like contemporary media tries to obscure it's origins. When we use an Ethernet cable we rarely think of the poor bastard who had to mine the copper or think about the effect of early copper mines on our cultural, social evolution.
I like to imagine the matter of contemporary media crawling out from the satanic pits of the early 19th Century, struggling to evolve in the winding towers. Then laying rails for itself to feed, spreading out creating denser and denser webs of interconnection for itself.
One you suspend seeing transport and communication in contemporary terms and think about them as the same thing, as they once were, then different histories of media emerge. Like in the 1840's, physical machines, steam engines force the compression of landscapes into manageable chunks of aligned time-tables, co-ordinating the bodies on to trains and into mass labour.
Submarine telegraph cables start to criss-cross the Atlantic, re-compressing the ocean’s trade routes into global markets realigning it into the rows and columns of the ledger, birthing scientific management and unifying markets. The mines transformed the body as the body transformed the mine, feeding lungs into the hungry boilers of empires.
MF: Coal Fired Computers doesn’t attempt to resolve the problem of energy, but using a wonderful but rather inefficient engine turns coal into heat, into movement, into electricity, this in turn transmuted into a machine that handles data, and drives a compressed air machine feeding a pair of blackened lungs. The machine is a diagram, but also composed of an enormous different kinds of things, timescales and eras, of sorts of stuff, and of different kinds of expertise and ‘states of nature’ things that are worked and transformed in various ways. The project is also, as you say, very much about transformation, of matter, time, knowledge, media systems, communities, flesh.
GH: Yes it's a dark futurist contraption – a strange, unnecessarily intricate, improvised machine, dreamed up to bring power, media, histories and flesh into proximity with each other. When I plugged the electricity from a hundred year old steam engine into the computer, I was elated to feel the symbolic power of that, I did not care what anyone else thought – I needed that fix.
Then bringing the miners who dug the coal that was shovelled into the boiler to watch the diseased lungs inflate with every database record made it orgasmic. The miners have a fantastic vision of class power that I recall from when I was a child and they bought down two Governments in the 1970's. The melancholy of all those lungs, death, disease, power, electricity - we just don't have a vision of power like that anymore.
I deliberately wanted to burn as much coal as possible, pollute a massive area for no purpose other then to feed my contraption, I needed to see what it felt like to be completely wasteful. Originally we wanted to gather the coal from child labour in India but this proved too difficult, but it led ultimately to our discovery of the nameless labour... the lack of datasets that fuel our wealth and power.
MF: And the connections run on?
GH: I suppose the other fix was the lack of separation between flesh and the machine. The lungs hanging on the front of the steam engine with wires poking out and pulsating. For me, this reflects my own reality of having big bits of steal screwed into my body with nylon screws that I have carried for the last 35 years, and having endless cameras and other bits of medical technology inserted into my flesh, or conversations with kidney dialysis patients about where their life ends and the machine begins, and the simple reality of those bodies that feed the machine of our power.
MF: You have also worked with databases that provide statistics on the conditions of work, (such as the Lungs: Slave Labour project of 2005). Work records, health records, the registrations of populations in figures becomes something that you see both as means to tell some kind of truth or story about the conditions of life, but also to make them physically palpable, through breaths, but also tender, bodily and ephemeral. These are two different means of registering peoples’ lives, two ways of knowing the world but here they are brought together in a way that is both very sad, mournful, but also somehow irrefutable. What are your thoughts on the relation between statistics, record keeping, the infrastructural cruelty of the systems you record, and the kinds of expression that they yield in the systems you assemble at a tangent to them?
GH: Death and media excite me, it's one of my kinks. In what might be an unhelpful nutshell, Memorial is where the database combined with death changes conduct.
MF: Could you explain that a little more?
GH: Record keeping is still seen by many as being separate from lived experience, a model, a trace, residue if you like. But we are transformed by the use of indirection, modeling, creation and implementation of our record keeping or by not keeping records at all.
Simply put... the database, the need to create a conceptual-view for our records, necessitates the implementation of sets of formal rules that are contained within the database. These theoretical machines are used to dissect an enterprise into sets of discrete normalised fields from which comparisons can be made which, in turn influences the conduct of the records input.
You can see the raw power of the database at The Tower Hill Memorial, Trinity Square to the Merchant Navy's 28,000 War dead in London's East End. The ordering of names, ships, dates forces you to iterate over the data in specific physical ways. The enlistment system records its victims by inserting a date in the death column. The collection of the data, to include commonwealth dead, but not those of the USA, echoes empire and the order of international relations at the time.
MF: Yes, this is a neo-classical monument that conflates masses of dead with architectural masses, columns covered in metal plates bearing the names and details of dead sailors, which in turn support a roof structure. The allegory is there for the turning.
GH: Or to put it another way, the normalisation and categorisation of the experience of an enterprise distilled into the conceptual-view creates an encoded expertise of the enterprise which reproduces its power in new and unexpected ways.
In Coal Fired Computers we tried to unpack this materialist view of software, its histories and engines. Open it up to a live experiment, see with others how the conceptual machines of the 19th Century have unfolded in to the everyday conditions that are now defined by perpetual crisis management, in the economy, ecology, security and financial systems.
MF: How important then has the key requirement in statistics and database design for data normalization to be maintained had an effect on other kinds of normative process, such as social normalization?
GH: There is almost no separation. If we think again about the Tower Hill Memorial as a physical manifestation of a database laid out in space. The body of the visitor is moved to access information, by ship, name, date. We order ourselves to read the fields as the ships, crews were ordered by the records kept on them.
The space between data and the management systems that processes the data points to a history of conceptual machines at least going back as far as Samuel Pepys’ days at the admiralty where he introduced examinations rather then class privilege as a means of evaluating officers, standardising ship types across the country, the provision for officers' pensions and payments for sailors' widows; amongst other things. His great innovation in all this was a distinct separation of information from the methods of its own representation... scrupulous, absolute record keeping as a machine to produce Empire. In contemporary terms we would see this as a form of standardisation. In turn building the ability to reference something using a name, reference, or container instead of the value itself.
MF: One aspect of a number of the projects you have produced in recent years is that of the incorporation of pseudo-code, bringing instructional sequences, written in an idiom that is close to Perl, sometimes with a degree of functionality, sometimes not: what are the stakes in working with this material?
GH: Hmm, there is no great difference between pseudo-code and functional code. There is just one level of abstraction or another. Maybe I'm too old but all my early experience of coding was with algorithms written in pseudo-code to get over the problems of language specificity. I have produced pseudo-code that has done much more processing then the more functional stuff. Maybe another way to see this is that I build software contraptions that enable me or whoever I'm working with to speculate about the world.
I'm completely uninterested in software that’s useful or works too well. I have no desire for a seamless integration in to my desktop and the systems it implies.
MF: To return to CFC then, the question of seamlessness is one that often occurs in the rational discourse on sustainable energy, in terms of creating energy systems that don’t loose power, that don’t leak. You are saying that in computational terms, another kind of consideration arises, that leakiness creates the possibility for excrescences, for imagination, the expressivity of data in relation to slightly mismatched algorithms or visualization schemes?
GH: Code leaks all the time, that’s what’s worrying, hopeful about it when it's received uncritically. You create it with intention, a technical fix, but in implementation it leaks into the social enabling, disabling as it iterates over the social, cultural, economic and political conditions that formed it.
In my own work I exploit this by creating assemblages of code, hardware, histories, people and materials. Particular datasets have particular resonance in certain geographical, social and political situations. In CFC we used a UK database of over 164,000 records containing the details of coal mining accidents and deaths in the UK from 1600 to the present day. This was created and/or paid for by Raleys Solicitors - specialists in workplace accident and disease compensation - a way of accruing knowledge. During 2003 and 2005 when the scheme was at its peak, Raleys' annual profit rose from £2.5 million to £15.7million. During this period two Raleys partners, Ian Firth and David Barber, made personal profits of £9.9 million and £7.2 million respectively. To reuse this dataset in other ways allowed us to play with Raleys as part of our contraption.
With the Lungs project in ZKM, the original dataset of records of slave workers was conceived within a Hollerith/IBM paradigm of punch cards, a mechanism of census taking that unfolded into racial hygiene. To take a Nazi dataset of the number of slaves used in the armaments factory in the building that now houses ZKM, to calculate the air that was in each set of lungs at the point of death, and re-breathe it into that factory was a way of unleashing new knowledge from fascist systems.
MF: One of the underlying arguments I think in CFC, but also in Lungs: Slave Labour is about the power that vast accruals of data can have. Databases are no longer called Data Banks, but there is something about the agglomeration of large amounts of data that gives it an affinity, if not quite to capital, to something common in a power of amassment to create distortions of power and understanding around it. In which ways might we need to reshape our understandings of data?
GH: Yes, I have never quite got to the bottom of the name change from data banks to databases – Codd who produced the first relational database still refers to data-banks in the late 1960's. I suppose it’s something like there was no separation between the data and the code that produced it in the data-bank, leading to a repository of information and the methods of accessing that data. After Codd’s idea of relational database management systems, data sets and the code that process them are separate. So, the DBMS becomes an engine for the production of knowledge and power, changing conduct from processing the sets of information.
I'm working on health records at the moment in Liverpool and trying to think about the aggregation of 60,000,000 health records in the UK... forget about the privacy issues for a moment.
The aggregation and structure of this information will produce new knowledge with a measurable power to change conduct as I described earlier. This will disrupt older forms of health authority like the British Medical Association, based upon professional knowledge, with a new kind of power formed from a software-mediated return of the masses in the form of health records. This is where the leaks get interesting, potentially on the road to new tyrannies.
An example from Liverpool, is the “Joint Strategic Needs Assessment” document developed by the Liverpool NHS Primary Care Trust (PCT) dated 2008. The PCT had found that it had a strong indication that 10,000 people were out there somewhere with Hypertension. They had no direct knowledge of this, but it was indicated by comparing their records with other records around the country. If those people could be found, then morbidity rates throughout the city could be reduced. The argument for this interpretation was created by comparative analyses involving many datasets. These datasets coalesced as new forms of authority that in turn could direct PCT priorities. A further convolution in the reading of the data was that Public Health advisors also thought that if you put the same money as it would cost to take the measures against hypertension into promoting the health of 16-25 year olds, this would have greater long term benefits – unfortunately however the evidence for this would take longer to gather then the lifetime of a parliament and so had to be discounted.
MF: This sounds like a story with many possible turning points in it, many moments when decisions were made, resources were joined, work was done, in one way or another. What kinds of connection and combination can you imagine for such datasets to yield new figures of truth and potentially a new politics of this new kind of mass?
GH: I remain hopeful that vast datasets will ooze new forms of power from the aggregation of mass records which have the potential to dislodge established forms of professional knowledge before they unfold into new modes of tyranny further down the road. The problem with this optimistic model of transformation is that it depends on datasets being 'rationally' built by people who understand the flows of information.
Recently when I was working with a Health Trust I noticed that the fields within the five competing datasets were politicly driven and the system was undesigned to protect the competing political/financial interests of the Hospitals, Health Trusts, Government and General Practitioners. The system was not live in that records were at least two months old, had to be requested over night and arrived in a flat file with one table of more then a 1000 fields in a table. This would be shocking to any elementary computer science undergraduate.
I find myself becoming a data puritan, well designed, ruthless information, using open systems will allow for much better regulation of data privacy then any sloppy, propriety and politically determined system.
MF: I like this term ‘contraption’ that you have started using recently. It seems to couple a kind of intentionality with a bit of the looseness required to keep things going. What is a contraption?
GH: In French, Jean says, contraption is pronounced "Machin, truc, bidule": something that one cannot or refuses to name. Its quality as “passe-partout” (passe-partout is a device that opens all doors) is to be unqualified, thus connecting elements and revealing sets of relations that are not evident or sometimes hidden. Its in-between states allows for a practical exploration and/or understanding of power and media ecologies that surround it.
A contraption in English is were the domain of the technical overlaps the imaginary, an experiment with nothing to prove. Usually strange, unnecessarily intricate, unfinished, inherently unstable, improvised machine.
'Strangeness' enables it to become a place of experimentation and fun. 'Inherently unstable' refuses easy utility, normalisation and emphasis the forces at play in the machine that break it. 'Unfinished' is about provoking thought, emotion rather than wanting to show it how it is/should be. 'Improvised machine' implies a playful assemblage of pre-exiting parts. 'Unnecessarily intricate' allows for a geeky self-expression or the elegance, aesthetics we find in complex code.
I suppose what I'm hinting at is the unstable state of invention before the 'machin' becomes normalised.