The Operative Image - an Approximation

operative image, digital warfare, locative media, computer vision, medical imaging, interface criticism, digital image theory

Contributed by Ingrid Hoelzl Assistant Professor at School of Creative Media, CityU Hong Kong
February 03, 2014
ingrhoel's picture
Part of the Cluster:


On our computers, laptops, tablets and smartphones, the image as the termination (fixation) of meaning has given way to the image as a network terminal (screen). Images are no longer limited to a political and iconic representation; they are not only an interface, but play an active role in synchronic data exchanges. They become what media artist and theorist Harun Farocki has called “operative images”: “images that do not represent and object, but rather are part of an operation” (2004: 17).[1]

Farocki’s concept of the operative image, coined in the context of his discussion of automated warheads, where the image functions as a guiding tool for target tracking and the real-time adjustment of a missile’s trajectory, proves to be very operative, too, when placed within the field of locative media and digital surveillance. “Location services” such as Google Street View, enabled by real-time data processing and continuous exchanges between user location, GPS sensor, software, network and database, are based on the principle of the users’ trajectories feeding back into the database. The result is what we could call, with reference to Paul Virilio (1989), a ‛reverse operativity’ which proves to be the more problematic side of locative media applications: It is not only that we are operating the world through Google’s images, it is also and primarily that Google’s images are operating us (Hoelzl/Marie forthcoming 2014).

The digital image is an object difficult to apprehend, not only because its status is somewhat unclear, constantly oscillating between visual entity and digital data, but also because it seems infinitely malleable and to stand in a somewhat arbitrary (or generic) relationship to its object. What is more, with the advancement of image-based software applications, the function of the image has changed; it is not only showing things, but doing things. For this The New Everyday cluster I asked five researchers/artists from different fields (cyber forensics, critical software studies, new media art, critical theory, and medical image studies) to cross-examine the “operative image”, three of five choosing to work in interdisciplinary teams.

Aud Sissel Hoel and Frank Lindseth, in their essay entitled “Differential Interventions: Images as Operative Tools,” use the term to describe the role of imaging methods in medical operations, while at the same time extending it in what they call a “differential approach.” Operational approaches, they argue, allow us to understand the image in terms of “doings and happenings,” “open-ended processes of becomings,” and in terms of distributed agency, “as humans, apparatuses, and tissues form an integrated system.”

David Gruber and Daniel Howe explain their current research project and installation Gesture::Language::Mirror, a critical exploration of the mirror neuron hypothesis and its use in psychotherapy, via what they call “mirror algorithms.” Their definition of the mirror neuron as an “automatic, unconscious flickering of the brain,” a “collective movement of electricity, glia, light, and fatty surfaces,” brings them to understand the image (contrary to the Western tradition of representationalism) as “the bursting of simultaneous activity” and as a “collective event,” linking Farocki’s proposition of images being part of a process, to the philosophical project of Gilles Deleuze, who considers things not as stable and self-contained but as “a continuous variation of matter” (1988/1993: 19).

George Legrady describes the functioning and aims of his installation Swarm Vision (developed with Danny Bazo and Marco Pinter), which consists of three camera-robots that collectively explore the installation area and whose output is displayed on two screens. While the first screen contains three separate views, the second places all three picture sequences in a 3D reconstruction of the installation area. Legrady explains that “imaging systems function to capture presence, to record change, to see beyond human range, to create evidence, to stimulate action.” The image produced in this particular imaging system is part of a feedback-loop; it is both the output of camera movement and an input to image processing that calculates further camera movement.

The feed-back loop is also at the heart of Christian Andersen and Søren Pold’s “Manifesto for a Post-Digital Interface Criticism.” The interface, they argue, is a “multimedia that integrates sound, images, text and interaction in feedback-loops,” where the registering (or generating) and representing (or visualizing) of data occurs simultaneously. The interface is not merely a tool for human-computer interaction, as it is often understood, by users and designers alike, but “constitutes the sensible (even beyond the human),” i.e., “the way we sense, what we sense, and how we act upon this.”

Richard Overill’s contribution, finally, shows the “layered” nature of the operative image. While to the human eye, a digital photograph may function as a simple holiday photograph (or porn video), on a sub-representational level it is executing the complex operations of image compression and decompression (Hoelzl/Marie 2013). Overill’s point is that digital images can also be used, in a method called digital steganography, as covert, “subliminal” channels where information is embedded into the “least-significant bits” of each Red, Green and Blue channel (or 8-bit sequence) for the purpose of secret transmission (where the receiver of the cover image will then decode the hidden message), or for the purpose of infiltrating malware into organisations (in order to secretly exfiltrate information, for instance). To the human eye, the changes in the image are barely noticeable; only statistical software is able to detect a potential steganogram.

That also means that “operative images” are not necessarily made for the human eye. Is their aim “to see beyond human range” (Legrady) and to “expand the human action range” (Hoel/Lindseth)? If humans operate machines and machines operate humans through images, it is evident that images, as image-programs, operate machines as well as humans. But machines also operate machines, as in the case of automated sensing systems where the measuring and processing of light, heat, or sound data is no longer dependent on its output in visual form for human interpretation and action. In this process, where human eyes (and operators) are no longer needed, the question of what is an image (if not seen) is extremely difficult to answer (but philosophically necessary).

Farocki’s stance on this issue is somewhat ambivalent: In his 2004 essay entitled “Phantom Images,” where he coined the term of the operative image, he relates that his interest in images, “taken in order to monitor a process that, as a rule, cannot be observed by the human eye,” lies in their non-intentionality,[2] the US military’s tactical warhead pictures approaching what he calls the “unconscious visible” (2004: 18). But he also argues that such images, even if not made by man, are made for man, since “[…]there are no pictures that do not aim at the human eye. A computer can process pictures, but it needs no pictures to verify or falsify what it reads in the images it processes” (21).

Ten years later, the development of computer vision techniques seems to indicate a turn towards what we could call ‛post-human operativity’: while the imminent task at hand is to perfectly simulate how humans see and make sense of the world, the ultimate goal are fully autonomous systems of image creation, analysis and action, capable of substituting human observers and operators altogether. But then we will need a radically new definition of the image (or have no more need for it).


[1] Farocki refers here to an earlier work of his, Eye/Machine (2001) where he had first coined the term for images that are “made neither to entertain nor to inform” (17). The film analyzes images of the Gulf War, where the US Army used missiles that philosopher Klaus Theweleit has called “filming bombs”. In Eye/Machine III (2003) Farocki further explores his concept of the operative image (renamed operational image), tracing it back to the 1980s cruise missiles whose software could compare stored photos of a landscape with actual photos taken during the flight.

[2] An aspect explored in his contribution to the 2001 ZKM exhibition CTRL [Space] curated by Thomas Y. Levin. The “Phantom Image” article published in 2004 is based on a talk Farocki gave at ZKM in the context of this exhibition.



Deleuze, Gilles 1988/1993 The Fold: Leibniz and the Baroque. Trans. Tom Conley. Minneapolis: University of Minnessota Press

Farocki, Harun 2004 “Phantom Images,” PUBLIC 29: 12-22

Hoelzl, Ingrid and Remi Marie (forthcoming 2014) “Google Street View: Navigating the Operative Image,” Visual Studies 29:3

Hoelzl, Ingrid and Remi Marie 2013 “CODEC: On Thomas Ruff’s JPEGs,” Digital Creativity, 19 September 2013,

Virilio, Paul 1989 War and Cinema: The Logistics of Perception. London: Verso