Earth as image and operation

Bryan Norton (BN): In 2020, Geocinema (Asia Bazdyrieva and Solveig Suess) released an astonishing documentary on China’s Digital Belt and Road Initiative, ‘Making of Earths’, for the ‘Critical Zones’ exhibit at the Center for Art and Media (ZKM) in Karlsruhe. The curators, Bruno Latour and Peter Weibel, referred to the exhibition as an ‘observatory’ of critical zones, those networks of biological, ecological and geological processes that are vital for maintaining life on planet Earth. Earlier this year, Jussi, you released a wonderful new book, Operational Images: From the Visual to the Invisual (Parikka, 2023), which interrogates the long history of scientific image making, relating it to current regimes of algorithmic control and contemporary artistic production, among other topics. What was the impetus for these projects? Would you describe each of these as continuations of previous projects and interests?

Asia Bazdyrieva (AB): ‘Making of Earths’ was our first attempt to test out key questions that sat at the core of Geocinema. We started the project with the idea that the Earth is sensing and recording itself through multiple distributed and decentralized processes of signal transmissions. These recordings contribute to a massive assemblage of data that exceeds human capacities to comprehend and make sense of it. For us, there was an obvious tension between the popular image of the Earth, the Blue Marble, and these outcomes of sensing and recording that might or might not take visual form, that might or might not be seen, understood and known. So we decided to scrutinize the representational power of the planet, which is being used and abused on multiple levels from technopolitics to environmentalism, and more. Our working method was to switch from the ontological mode (what is the Earth?) to an epistemological one (how is the Earth?) – how it is being sensed, transmitted, packaged, recorded, represented and archived (Figure 1). We decided to focus on infrastructures for Earth observation, in order to see what happens when this image of the Earth is coming into being, what imagery, agencies and desires are folded into these processes.

OPEN IN VIEWER

Figure 1.

Geocinema. “The Making of Earths.” Beijing Ancient Observatory. 2021.

Jussi (Parikka, 2020) aptly theorized our work:

over the past few decades imaging and sensor-based techniques have proliferated and are emblematic of an alternative sense of visuality, but also of the political formulation of issues of territory and futurity . . . This visuality is less about what is seen but about what is being actively measured and calculated. As well it is less about the images as such, but about the architectures, labor, processes and infrastructures through which the operations unfold.

Again, these imaging techniques are operationalized in ways that escape immediate visibility or perceptibility. They are part of larger constellations that include migrating forms of governance which spread through deterritorialized flows of global markets, and have become further obscured by the seemingly invisible and not necessarily territorial regimes of data. In the ‘Making of Earths’ we wanted to enter these layers of architecture, labour and infrastructure.

Jussi Parikka (JP): The projects, Geocinema’s ‘Making of Earths’, and my Operational Images are indeed very much entangled. I was able to be in dialogue with Asia and Solveig during their ‘Digital Earth’ fellowship and observe and discuss their ways of elaborating the operationality of not just images but large-scale infrastructures of sensing as part of the Chinese Digital Belt and Road Initiative. This nailed one of the things I was (and still am) interested in: what is the ‘operational’ in operational images, what are the infrastructures that are not visible in this format of visual culture as they recede to the background and yet are of crucial importance? Their fieldwork, including interviews and shooting, across different locations from China to Thailand, was a really interesting contemporary counterpoint to the earlier notions of the operational image. In their case, the particular examples of visual cultures of monitoring, sensing, observing, calculating and predicting were also contextualized beyond any single apparatus to encompass large-scale systems. Our discussions about ‘what is “cinema” in Geocinema was one point of reference for the genealogy of imaging that stemmed from the 100+ years of surveillance, military, scientific and other technical modes of imaging that now finds its contemporary versions in different geopolitical contexts (see Holl, 2017). What is more, I wanted to focus on the operationality of imaging and images beyond the most direct references to surveillance and war, and in those slightly more subtle sites of sensing that are themselves spatial and architectural – as Geocinema shows in their work that deals, for example, with institutions and landscapes of remote sensing. These approaches to infrastructure and architecture come to define the broader material sites of Earth; here, the Earth becomes operationalized in this circuit of built and visualized environments that builds on what Asia just mentioned as the epistemic question of ‘how is the Earth?’

To pick up from the earlier mention of our shared interest to go against the circulation of the ‘Blue Marble’ genre of Earth images, Operational Images is basically pitched as a project about the multitude of images that are more administrative, bureaucratic, instrumental, grey and fundamentally, indeed, operational. In this ‘genre’ of inconspicuous images and cultural techniques of sensing, they are approached as drivers of central elements of managerial power as well as epistemic coordination that comes to be channelled through technical images. Hence, the sorts of examples my book outlines include DBAR, but also astronomical photography (the opening case of the book) as an additional site of counting and calculation, AI and data-enabled images where the infrastructural invisuality becomes central, historical pre-digital cases of data images in photogrammetry and many other liminal examples of how images are changing.

My book emerges from a five-year research project at the photography department at FAMU, part of the Academy of Performing Arts (Prague). ¹ The centrality of the question of photography in all this was a productive restriction; not just going to analysis of the most recent networked photography and the digital imaging-data complex, I worked in the context of theories and histories of visual culture and transformations of photography. This involved asking, alongside the whole research project team, questions such as: what are the sites, techniques, platforms, and institutions where the transformation of visuality and images can be observed? How to write archaeologies of contemporary imaging backwards, recursively, in terms of their constitutive techniques that go beyond photography proper? Hence, it was specifically interesting to include such pioneering work as Alan Sekula’s into the mix – the link between his instrumental images and Farocki’s operational images is of course central, as is Sekula’s work on 19th-century notions of photography, archives and control – while shifting from photography as visual images to techniques of sensing and detection. Here the actual image becomes secondary to the process through which different modes and models of sensing are being developed and which then finds an interesting counterpoint gradually since the 1960s and 1970s with remote sensing. This allows us to shift from the usual military and weapons focus of operational images to the vast infrastructures of planetary sensing that also come to play a central role in environmental management as well as, of course, processes of the financialization of natural resources. This is not a dismissal of the centrality of war but rather a recognition of the multitude of different modes of power and control of which operational images are a part. This also includes a recognition of the different kinds of violence that might be included in these operations (Alliez and Lazzarato, 2018) where the image is not necessarily one of crosshair targeting but all sorts of inconspicuous screen-based interfaces that work upon transformation of material landscapes. Finance would be one (Terranova, 2022: 50–52) – a truly powerful agent of material redesign of planetary surfaces through a variety of screen interfaces. Remember how Farocki (2004) talked of ‘phantom images’, perspectives that were impossible to be occupied by humans? The sort of phantom images are thus not only located on the nose of a missile but across the ghostly apparitions of all sorts of odd causalities that appear and disappear in finance or environmental models. They are massively distributed abstractions that nevertheless have the operational power of radical transformations of large-scale areas, chemical states, ‘habitabilities’ and ‘inhabitabilities’.

BN: I think the counterpoint to the blue image genre both of you mentioned really became apparent when the first images of a black hole, if we can even call them images, were made public at the end of 2019. The telescope array that was used to create these pictures, the Event Horizon, used radar cameras capturing electromagnetic radiation rather than light. In a very real way, it was on the basis of sound rather than sight that scientists were able to algorithmically visualize the black hole at the centre of the galaxy M87. This brings to mind one of the most memorable parts of ‘Making of Earths’ – a conversation at the Remote Sensing and Earth Institute with the head of the DBAR Big Data Working Group, Li Guoqing, who describes the process by which data is taken from satellites and eventually distributed as knowledge for public consumption. In this interview, Guoqing takes pains to carefully articulate the difference between a satellite signal and a picture. The subtitle of your book, Jussi, is From the Visual to the Invisual. What on Earth, if you will excuse the pun, is actually happening to the image in these distributed networks?

JP: I actually think that would be a great alternative subtitle for the book: Operational Images: What on Earth Is Happening to the Image! Here the ‘happening’ comes to take a central place: all of this is not necessarily about the ontological question of what is an image so much as the operations that enable images as well as images that enable operations that have transformative effects, e.g. on landscapes whether rural or urban. And, as you hint, much of those conditioning contexts take place as signals that are not necessarily visual. This is the paradox of contemporary visual culture: it operates on a variety of wavelengths other than the visual spectrum where some of it might be visualized, but it also reaches out beyond a particular human notion of visuality or images. In Operational Images, I pitch the idea that while most of Arts and Humanities has until now focused on that tiny little stretch of visible light, in what ways can we extend to the other spectra and other signal realms as a significant challenge for aesthetics and knowledge? So if, for example, most of the history of photography has been about sorting the kinds of wavelengths that are considered the most relevant, for decades now we have been in a rather different ballgame with techniques and infrastructures of remote sensing – both the ones turned toward the Earth and the ones turned toward outer space.

As I put it in the book (Parikka, 2023: 58):

Tell me what spectrum of electromagnetic radiation you are concerned with, and I will tell you what your discipline is. Is it the visible spectrum that registers people, objects, and landscapes? Or is it the invisible spectra observed by other remote sensing systems on satellites and the ground that register people, objects, landscapes – but also a multitude of other layers of reality from temperature to humidity, mineral resources to the chemical composition of vegetal surfaces.

Nothing less is at stake than the fundamentals of disciplinary objects, which are accepted and standardized as objects of knowledge! What would be, for example, an art history or aesthetics of ‘the invisual’? Some of this work has already been done – to me, a lot of the discussion around the algorithmic and networked image has been premised on deeper assumptions about where exactly the image is even located (see Dewdney and Sluis, 2023).

In this manner, the black hole image features nowadays in many theoretical writings as both an emblem of planetary scale sensing as well as the extended sphere of artificial intelligence. From Orit Halpern (2021) to Benjamin Bratton (2019) as well as my own book, the black hole stands as this particular example that, as you point out, is a different kind of a visual object than the Blue Marble. A significant part of this discussion relates to the massive scale apparatus of sensing that is employed to detect particular signals, as well as the way signals are being composed as what we will then call an image. Here, the process of machine learning and devising a ‘model’ image comes to the fore as an interesting epistemic object: what is the status of an object that is entirely real but not verifiable with a ‘ground truth’ of the sort we intuitively thought of it in relation to the calibration of remote sensing of minerals, for example, which Geocinema also use in their ‘Making of Earths’ project? What is more, as far as distributed networks are concerned, I try to mobilize the notion of operational image to discuss AI culture and think about how the stakes are less about any individual image and more about the processes taking place at the thresholds of visibility: training, data, annotation and, of course, the question of the model in computer vision and other AI systems (see Crawford, 2021; Malevé, 2021 and Lin 2022).

AB: Let’s dwell on the difference between the kinds of visual objects represented by the Blue Marble and the black hole. Our research leading to the ‘Making of Earths’ was, first, preoccupied with understanding how the (geo)political is woven into the representational by the production of a certain kind of image – a product of specific power relations supported by particular infrastructures. We were looking at how these infrastructures operate – how they are mobilized spatially and bureaucratically, and asking what the technical specifics of imaging are. We were asking what this image is and more importantly, what it is not. Responding to such questions means unwinding the image-as-modern-fantasy, from an image that is an actual photograph and most importantly, from an image that is not exactly an image, but rather data as image, or image as data. Remote sensing, being a predominant technique of imaging the Earth from space, was already at the moment of its implementation designed to focus on signals even before they become images. These vastly distributed infrastructures allow for the acquisition, transmission and archiving of data that are not inherently visual and likely to resist the kind of representation brought forth by the Blue Marble. Sputnik 1, launched in space in 1957, signalled a different type of relationship with the Earth and therefore a different type of knowing, where the image or the photograph was no longer a stable object of reference through which the Earth can be known.

This responds to your question, Bryan: what has happened to the image? The information that can be extracted through the processes of sensing and imaging is not a fixed entity that exists as such. It is in itself a plethora of possibilities that can be recognized in one way or another, and are largely dependent on material, sociopolitical and cognitive aspects of knowing through their acquisition and interpretation. Here we see a new image, or rather a new type of entity altogether, which triangulates measurement, scale and temporality within the production of data – ‘a topological process that conforms relations between images to the commodity form’ (Cubitt, 2021: 30). This new type of relationship with the Earth is a different type of epistemic practice, which is a still largely articulated type of mathematical knowing. Not only does this type of knowing raise questions of scale and temporality with regards to how the world is perceived and understood but it is also particularly interesting as it has operational qualities which introduce an important topological shift. If earlier scientific understandings of the sensing and imaging of the Earth’s surface implied fixed categories by which the Earth can be represented, this emerging mathematical knowing implies that knowledge is generated by new means of sensing and imaging. This importantly includes historic situatedness, the way images are embodied and embedded, and the possibility of including the desired outcome in the proposed socio-economic scenario. To understand this shift epistemologically is crucial in order to inquire about the political implications of such processes. Mathematical knowledge, as opposed to representing the world as a particular stable-scale space–time, generates speeds and spaces of dynamic scales, which never exist just out there in the world. As a part of larger geometries of power, they are designed, produced and created so they can then be occupied.

Some argued that the mastering of nature happened through this mastering of the image. This does make a certain amount of sense when we talk about the Earth and the colonial extractivist logic through which it becomes known. Yet what does the imaging of a black hole do? For me it is less a question of whether the phenomena is visible or made visible algorithmically (we’ve had that moment in history with the discovery of x-rays, for example, haven’t we?). The question, rather, is this: does an image of a black hole still sit within the attempt to understand what is out there, so to speak, or is there already something else that is at play at the moment of this particular scientific and visual inquiry?

BN: I think this issue of scale that you have both mentioned is important. A black hole is something so immense it borders on a sort of cosmic sublime, yet its representation is created through an infinite number of micro-temporal processes (see Cornford, 2023). There seems to be a short circuiting of the everyday and the readily available, the mesoscopic, in favour of micro and macro-level processes here. Is this ‘something else’ you mention here a matter of media history, Asia? Of politics? Something different entirely? For you, Jussi, is this where the issue of the Invisual enters the picture?

JP: To discuss such images (or even ‘unimages’) as in the Event Horizon project, or remote sensing, or many other similar examples reminds us how broad the field of media and visual studies can be: while the everyday empirical media objects have become the standard go-to of critical theory as they concern crucial issues of social politics of culture – intersectionality, for example – it is also interesting to extend to asking what those seemingly technical images consist of that affects our sense of not just representations of the world, but again, those operative modes of intervening in that world. An image is only a temporary conceptual hinge through which to discuss modes and models of intervention. The paradox haunting contemporary visual studies is to work with the current notion of the image in order to get rid of the image as the primary referent entirely.

And here is where it gets really interesting: what are those modes that seem to concern ‘images’ but are not of the order of the visual? Asia already mentioned emergent geometries of power. These informational or data rich images contain a multiplicity of potentials of use, interpretation and mobilization, and while we persistently refer to them as images – sometimes even photographs – they are actually barely images anymore. The notion of the invisual, developed earlier by Adrian Mackenzie and Anna Munster (2019), is in this sense a useful way of framing this level of operationality that in their case emerges from questions of data platforms and the format of ‘seeing’ that is embedded in the data operations. In this regime of images, seeing is not returned to a gazing subject, but distributed across observational events, they argue. Furthermore, it is not the individual image at play but the collection, the aggregation, the ‘image ensemble’, that is: ‘images, not simply quantified, but labeled, formatted and made “platform-ready”’. So any image and its sensing are already in some way prefigured into a formatting, thus already hinting that there are other operations at play too that are not just about ‘the visual’. A cascade of operations that prepare and mobilize images as data comes to the fore as a particular diagram, to still stay with Mackenzie and Munster. Allow me to quote from them, as I think it nails it well – including this aspect of the multiscalarity of the particular diagram at play:

Invisual image ensembles move across any clear contrasts between eye, lens, sensor, file, screen or database. The corollary of the processing of collections is that the platform itself as a raised, visible surface of intermediating cuts off the capacity to look from a position because it distributes seeing transversely: pixel, device, and hardware architecture are conjoined through the operations of this invisual diagram. (Mackenzie and Munster, 2019: 18)

A multitude of contemporary practices can be addressed through such an idea of the diagrammatics of the invisual: AI and images are about the invisuality of labour, datasets and the extremely interesting epistemic project of ‘the model’; lidar as a particular technique of light that does not return representations in the usual sense but models again – models constructed of light pulses; such histories of imaging hover between representation and data extraction, like photogrammetry way before digital modes of imaging, and so forth. My modest proposal then is to see operational images as one particular way of understanding invisuality, and vice versa.

About the multiscalarity of the diagram at play: this form or format of image and data practices concerns scale, and can be understood through newly reconceptualized notions of scale. This is visible both in recent literature on the topic – Zach Horton (2021) and Joshua DiCaglio (2021) for example – as well as several curatorial narratives of the past year or so. transmediale (2023) had ‘scale’ as its theme, as did Aksioma cultural and art centre in Ljubljana. In the transmediale case – full disclosure, I was part of the curatorial team – the digital art and visual culture festival shifted its focus from the centrality of scale as measurement and control, as central operations that pertain to the history of design and architecture to also include how scale relates to the question of simulations and the seemingly free-scale shifting whirlwind politics of digital media. Contemporary politics of scale as simulation, as well as multiple digital visual culture examples like holograms, augmented reality and other additive visualities circulating across urban and online space, are a case in point here about the ontogenetic power of such a fundamentally invisual sort of process. Ranjodh Singh Dhaliwal (2023) writes nicely in his text for the transmediale publication about this splintered distribution of fragmented images that occupy a particular place in visual rhetoric: ‘As fictions become our facts in this era of late capitalism, we find ourselves sifting between the many preserved bodies (corporeal and corporate) and constructed images (public and technological) of those in and around power.’ Such images at play – the Queen, Modi, etc. – are less interesting as images and more noteworthy for their operative capacity to circulate, capture, affect and modulate particular kinds of socio-political relations that are not necessarily any more visible on the spectrum of white light. Although, to add, I am more interested in the sort of scientific and epistemic context of simulations that Geocinema has focused on, as well as Digital Twins, software-enabled modelling practices, etc.

AB: Jussi mentioned diagramming a few times, so I will add – in continuation of what I said earlier about new types of mathematical knowing – that in recent years there has been a growth of scholarship on mathematical epistemology and diagramming as a reasoning tool. While I was working on a text for Svitlana Matviyenko and Kayla Hilstob’s upcoming book, Cyberwar Topologies, Svitlana recommended looking at Silvia de Toffoli’s work (see Priest et al., 2018) in this regard. The very basic idea is that diagrams are not so much a stable representation of things as they exist in the world. Instead, they ‘have been and are being used not merely to represent knowledge, but as aids to generate knowledge’ (p. 58) Diagramming is not neutral, it is a material and dynamic practice; diagrams ‘can be manipulated to generate, explore, and test hypotheses. This manipulability is central to their ability to generate knowledge.’ Moreover, it is a social and cognitive practice, which means diagrams are both embodied and embedded, where the knowledge they generate is always context-specific and largely shaped by the objectives of their creators. ‘There are many ways to interpret data. Only through a particular criteria data becomes legible and otherwise.’ So Toffoli proposes the term ‘manipulative imagination’ to describe this process of enhancement through diagramming practice. Since this type of mathematical knowing is becoming a very widespread tendency, it has triggered the attention of scholars of the history and philosophy of mathematics with the central idea ‘that mathematics, the most idealized and seemingly disembodied expression of thought, should be considered also as a human activity with a particular history rather than exclusively as a domain of eternal truths.’

For ‘Making of Earths’, this was particularly relevant because our case study, the DBAR project, was pitching the idea of ‘digital earth’ both as a scientific tool and as a platform. A tool that was proposed by the Chinese Academy of Sciences (and supported by a vast international community of scientific institutions willing to join forces with the platform) was not meant to represent a single solid world (although it heavily employed the Blue Marble, with all the representational power of that image), but rather to see the world as a manipulable spacetime full of many conflicting possibilities with multiple scenarios to which software, hardware, international policies and logistical planning can be proposed. Here the topological shift, as we argue in our text for Cyberwar Topologies, is in the type of relationship with the Earth (and its future) (see Figures 2 and 3). For example, in environmental management, the earlier topographic approach would mean visualizing possible scenarios of the Earth’s future based on simulation models run on past geo-referenced data. China’s Digital Earth currently gears towards a different type of future forecast. Their Digital Earth allows for the modification of parameters or boundary conditions to render and evaluate a number of alternative scenarios, while feeding back and adjusting parameters in real-time towards them. So there is a turn from prediction based on data to data based on prediction.

OPEN IN VIEWER

Figure 2.

Geocinema. “The Making of Earths.” Overlay of a Diagram of the “Cartographic Process” onto shots of a volcanic field in Yunnan Province. 2021.

OPEN IN VIEWER

Figure 3.

Rosa Menkman, Decalibration Target: Start of Image Marker, Drone footage, produced by transmediale with the support of the Stimuleringsfonds Creative Industries NL, January 2023. Photo by Rosa Menkman. https://beyondresolution.info/TARGET

BN: From prediction based on data to data based on prediction – I really like this formulation. One thing that always strikes me in ‘Making of Earths’ is how contentious these forecasted futures can be. Perhaps this links back to Jussi’s point about formatting happening on an unseen, almost a priori level for digital platforms. At one point in the film there is an interview with a DBAR scientist who uses the wonderful euphemism of ‘harmonization of the database’ to describe efforts to standardize the formatting of data collected from all of the different countries along the New Silk Road. Could you say a little bit more about how you see this new relationship to the Earth taking form? What types of scenarios are being modelled? What exactly is being predicted in such a project?

AB: At one point in the film, we focus our attention on a project which plans to use the extraction of data in the Polar region. This project serves for us as an entry-point into understanding how mathematical forms of knowing mobilize the creation of speeds and spaces within global capitalism. One of many DBAR collaborations is with the Center for Integrated Remote Sensing and Forecasting for Arctic Operations in Norway that is developing a research project to integrate remote sensing for sustainable development in the Polar region. This project includes international collaborations with state and private entities across Norway, Finland, Russia and China. It also includes the prospective opening up of underwater resources and development of the Arctic Ocean Northern sea route as a new transport corridor as part of the Belt and Road initiative, while laying the groundwork for scientific research projects and infrastructure development in partnership with DBAR. Sustainability within this project implies the parallel processes of the modelling of melting glaciers, the calculation of underwater resources and their potential use, and, finally, speculative risk management following the extractive activities. Such international and quasi-corporate collaboration predisposes certain outcomes over others, in this case, towards desired scenarios within the Belt and Road Initiative.

BN: This question of intention seems particularly important when we think about how contemporary artmaking practices are situated in relation to this question of modelling and the relationship between data, model and environment. When reading much of your work, Jussi, I get the impression that you find a large degree of historical continuity in terms of media infrastructures and the capabilities provided by new technology to image production, hence the longer timespan of the story you tell in books like Geology of Media (2015) and Operational Images (2023). Do things shift in terms of the relationship between image and world when media historians turn to artistic practices? Or is this expecting too much from contemporary art?

JP: Let’s start with your last question first: artistic practices get easily mythologized in contemporary art markets, even at times in critical discourses. There is still a hint of solutionism in particular parts of media arts too – ‘let’s invite an artist to solve this or that planetary scale issue.’ The other pole is just as bad: art reduced to ‘raising consciousness’. Somewhere in between are really interesting, detailed, rigorous art methods that help to investigate the creation of problems and how they can be ‘sourced’ and what aesthetic knowledge does in this assemblage of methods. What I call aesthetics, mind you, is closer to Matthew Fuller and Eyal Weizman’s materialist reading in Investigative Aesthetics (2021). Aesthetics here closely relates to human and non-human inscriptions and also, importantly, to the methods that can make such inscriptions visible. Fuller (2023) has also referred to the idea of artistic practice and art as a ‘meta-discipline’, ‘a mode of work whose operation includes both working in other disciplines’ and acting upon them. Hence, it is not only art seen as a separate practice but asking the question of how it is involved in existing institutions, even with types of ethnographic methods, in different methodological combinations (art–science has been one of those), and with varying scales of impact. Think, for example, of Forensic Architecture and other similar agencies like INTERPRT as they address questions of human and environmental rights, art and architectural methods, and investigative work with different source bases, from witness testimonies to image-complexes. ²

For me, Operational Images is thoroughly embedded in the sorts of practices that have already been rehearsed by artists and architect groups for decades. Farocki’s own approach to operational images was already a second-order framing of those grey, administrative images: to show images via images, to frame control screens, data dashboards, military and surveillance apparatuses through the soft-montage that he practised in some of his works, and through the cinematic essay narrative that unfolded through his particular media historical – some might argue ‘media archaeological’ – work. Elsaesser (2017 p. 219) says this explicitly – Farocki was sort of a cinematic media-archaeologist who composed historical arguments using jump-cuts from photogrammetry to modern automated weaponry: ‘film history, as one of the overlooked genealogies of the cinema, must count among Farocki’s most consequential contributions to media archaeology, as well as an essential part of the prehistory of digital images.’ So it is not media history or art, but in the manner above – artistic composition, montage, work with images, but also with heterogeneous sources of other kinds – that inform particular ways of reading media history anew. In my book, I also try to cultivate this idea of operational images with a conceptual force for reading historical sources or themes in ways that problematize or re-historicize particular traits of contemporary digital imaging.

We can also broadly speak of an archaeology of technical images, a topic that has made many rounds. Now, operational images, as they are part of architectural modelling (and other adjacent practices), are a really intriguing case. I am specifically interested in how software environments become multi-scalar reality-creating engines in this manner, enacting one aspect of operational images as acting upon their own conditions of emergence. This recursive trait is visible in many cases where the modelling of landscapes, built environments, urban and rural sites becomes a way to transform those very sites. One could speak of operational images as recursive in their role in changing planetary surfaces from cities to rural landscapes and one could refer to the differing time scales of this recursion: we often consider operational images as this almost-instant automated action, but the focus changes if we start looking at those operational loops at a time scale of months or even years.

Also with regard to recursive operations and imaging, the mode of ‘data capture’ that was inaugurated in early photographic practices – even in many pre-photographic practices – such as photogrammetry can be seen as part of the genealogy of technical images that pertain to architecture and design. Such examples imply something important about images, drawing and measurement-lines, as well as data that has come to play a central role in different ways in recent practices (for example, photographic images flattened into ‘content’ for datasets). Here, the notion of temporality comes to the fore again in how it is packed into an image that hovers between representation and measurement, between image and data: while there are bespoke practices for the creation of photogrammetry images, any image is always latent in terms of what can be read from it. I am not claiming this as a new revelation of my own – this has been voiced in visual studies in many ways – but it bears an interesting resemblance to the multiple temporal aspects of operational images. ³ Just to add: Andrew Witt’s recent book Formulations: Architecture, Mathematics, Culture (2022) is absolutely fantastic in respect to such practices of complex mathematical and data realities captured inside the image-complex of architecture.

AB: I would like to go back to what Jussi said before about wanting to move from the usual military and weapons-focus of operational images to the vast infrastructures of planetary sensing. Here I see a problem of almost intentional separation of war and climate discourses – the problem that I have addressed in a text that will be published in Aksioma’s PostScript series later this year. What I experience here in Western humanities is this tendency to move away from war as something that belongs to the past, while the actual pressing need is to directly theorize changing temperatures, environmental management and the financialization of resources. When I look at the videos of the gas leakages following the explosions of the Nord Stream 1 and 2 pipelines – this was considered one of the biggest climate threats in the Baltic Sea area – I wonder, how can the interconnectedness of the climate and war not be at the centre of our conversation? I understand this intention to move away (or further) from the media theory and subsequent visual practices that were defined by Cold War paranoia. Yet, if we want to talk about planetary sensing and its architectures that allow particular operations to unfold, we must also address specific instances of wars that are a direct outcome of the accelerating weaponization of datasets accrued from geological formations, natural disasters, social media behaviours and citizen biometrics alike. While these wars are safely confined within the borders of the states in the global south or east, they are waged by political decisions in Europe, where the rhetoric of sustainable development, energy security and energy sovereignty – supported by numbers, images, diagrams and visualizations – is used to justify action or inaction in response to imperial acts of violence that are unfolding elsewhere.

In this regard, to answer your question, Bryan, about the relationship between image and world when media historians turn to artistic practices, I would like to refer here to the work of my dear friends and collaborators, Elena Vogman and Olexii Kuchanskyi (2023). Over the past year they have been curating a number of film programmes, while developing working concepts that they call ‘mental ecologies of war’ and ‘mental geologies of war.’ I find it fascinating how they were able to articulate, following Guattari's ideas on ecology, the capacity of images to create ‘milieus [which] are not merely carriers of information or affect, but impersonal architectures of subjectivity, sluices of desire’. For me, it is a very important proposition to move away from (or to not limit ourselves to) art practices that are preoccupied with revealing structures of oppression, but to practices, that, as Olexii (2023) puts, propose ‘an alternative approach to a medium – as a milieu of interrelated mental, social, and environmental processes rather than as a pure means of information exchange.’ Within their theoretical and curatorial work, Vogman and Kuchanskyi (2023) have been looking at short films that were made in Ukraine recently, especially in rather extreme conditions after the big Russian invasion in February 2022. The production of a video work in the conditions of the war is very different from any other form of filmmaking. It unfolds within practices of survival, resistance and activism, and inscribes, as Elena Vogman (2023) wrote: ‘trajectories of production, distribution, and perception . . . under the extreme conditions of war, certain works transformed their quality of medium into a concrete milieu: a set of relations which retrace a territory, reconfigure the gaze, and reinscribe a relation.’

BN: This return to the context of war naturally brings us back to Farocki, who introduced the central concept of the operational image during his early 2000s video work on drone footage from the Gulf War. Of course, he also hinted at a notion of counter-operation in subsequent work. This means that visual art doesn’t just reproduce existing world relations by circulating images of destruction and oppression (the issue of labour, of course, was central to Farocki’s understanding of the history of cinema). Images are potentially capable of producing new worlds as well. This is a very powerful thought and one that has been very useful for my own work with artists like Gregory Chatonsky (Norton 2024). By way of concluding, I am wondering, Jussi, and Asia too perhaps, if you might also indulge us in a bit of speculation about where visiting these different sites of technological, political and geological image-production may lead. Are there any particular images that are now occupying your thoughts? Are there any burning questions that these have opened up for you in terms of operations and counteroperations? It would be great to hear more about your present thoughts and future projects.

JP: I appreciate the question that I read as one version of the ‘worldmaking’ idea but with a particular bend to counteroperations, a bit like Forensic Architecture and others speak of ‘counter-forensics’.

I am recently mostly interested in images and imaging practices that are not necessarily artistic or ‘fine arts’. All sorts of inconspicuous screen practices from the history of the military to current architectural practices is where I see the interesting links of operations emerging. To echo again what I hinted earlier: for example, the question of the ‘model’, as in modelling practices especially in relation to software and data, is where I see many of the stakes of operational images really crystallized. Much of the history of modelling has been subsumed into their operational role – for example, in design. Besides physical models that are and have been such curious objects that tie abstraction and concrete materiality together, like diagrams, there is also the work of software and digital techniques. From everyday software such as Blender and Grasshopper to the various plugins that operate there to import, for example, environmental data is to me a truly fascinating part of direct and indirect ‘actions’ that images have upon the world. Multiple kinds of scale condense, different forms of dynamic entities are packed into these technical environments. They are barely images. They are more like image-processes, images in formation and images as they inform part of a broader chain of operations of transformation (the building industry, the wider construction industries, modelling environmental data in landscapes, etc.). ⁴ This mode of operation is not the usual impact of ‘direct’ operation which is one of the reasons why the temporality of the ‘what’ and ‘where’ an operation is located is important: is the operation at the work desk of imaging practices, in the software, or is it in the broader technologica–institutional chains of where the image becomes itself one agent of particular kinds of transformations?

In more technical terms, Abelardo Gil-Fournier’s (2018) work (see Gil-Fournier and Parikka, forthcoming 2024) on next-frame prediction is a parallel interest of mine that we also collaborated on (Figure 4). As one machine learning technique, next-frame prediction is an interesting kind of phantom image, too: the predictive capacities of projecting possible movement or dynamic of an image sequence infers possible futures on the micro-scale of a second or two. What exists here except the image? That uncertainty is the key to understanding such micro-scale predictive techniques. The image itself becomes the source, the dataset, of possible next steps that also are embedded in the world. Take, for example, a sensor that is in movement, such as in autonomous vehicles, or some other micro-scale prediction mechanism. As cinema, this is quite the intriguing thing of considering this unit of the ‘frame’ as also its own dataset, thus also expanding the question of spatio-temporal depiction into the vocabulary of data and inference, but also hallucination, not all of which is unreal either. It is this sort of control of the potentialization of futures that becomes itself an image-problem. As a brief note, let me add that our forthcoming collaborative book with Abelardo, Living Surfaces (Gil-Fournier and Parikka, 2024), also addresses some of these questions.

OPEN IN VIEWER

Figure 4.

Abelardo Gil-Fournier, “An Earthology of Moving Landforms 2018” [work-shop documentation]. A series of predictions of the movements of the fast meandering Ucayali River in Perú. Used with permission.

Of course, some of these images are ones that we see regularly, even as mainstream. Take, for example, game engines. Their role has expanded radically beyond gaming into their use in landscape modelling and other epistemic practices. There was an inkling of this present in Farocki’s Parallel I–IV series, but the importance of game engines, to state the obvious, has just massively expanded, as has their role as part of the pipeline of modelling and simulation from massive multi-agent systems to landscape dynamics. It’s their bracketed nature as infrastructure, of not just visual arts but this epistemic transformational apparition, a phantom image and a phantom model, that specifically intrigues me. They are at the back of the circulation of not just specific images, but types of images and conditions of images that have a specific epistemic and operative role. In more prosaic terms, I am now enthusiastically playing Humanity on PS5 thanks to a tip from Farzin Lotfi-Jam (2023).

BN: These sound like fruitful projects which not only explore images, but also highlight how generative the operation concept can be in many contexts, from war and surveillance to everyday software and video games. I would love to discuss this further with both of you but, for now, let me just say I am really looking forward to reading and learning from these forthcoming works. It has been a pleasure interviewing you, Asia and Jussi. Thanks for your time.