INHABITING THE HYPER-AESTHETIC IMAGE

JACOB LUND: I would like to talk to you about the role of images in the work of Forensic Architecture in general, and more specifically, about the notion of “the hyper-aesthetic image” that you develop in your forthcoming book co-authored with Matthew Fuller, Investigative Aesthetics (Verso).1 In the book you argue for the function of aesthetics beyond perception. We, along with animals, plants and other living organic cells, are not only sentient beings. We may also act as sensors for—or traces of—events at levels other than sentience. I am very intrigued by this idea about “aesthetics beyond perception.” What are the implications of this expanded notion of aesthetics for our conception of images and how they are generated?


JACOB LUND:
I would like to talk to you about the role of images in the work of Forensic Architecture in general, and more specifically, about the notion of "the hyper-aesthetic image" that you develop in your forthcoming book co-authored with Matthew Fuller, Investigative Aesthetics (Verso). 1 In the book you argue for the function of aesthetics beyond perception. We, along with animals, plants and other living organic cells, are not only sentient beings. We may also act as sensors for-or traces of-events at levels other than sentience. I am very intrigued by this idea about "aesthetics beyond perception." What are the implications of this expanded notion of aesthetics for our conception of images and how they are generated?

EYAL WEIZMAN:
Effectively, this notion of aesthetics and of the photographic that we developed comes to confront the modernist and, to a certain extent, colonial or imperial legacy of photography which is based on the interdependent relation between a frame and a viewer.
The frame cuts things from the context, isolates them and lends them to classification. So, the idea of looking photographically at things that are not photographs is, to a certain extent, an attempt to decolonize the photographic process through a certain theoretical reworking of aesthetics. Humanistic aesthetics predates photography by a few dozen years, not too long, but to a certain extent they are co-temporaneous, connected to the idea of contemplation and judgment, i.e. to a Baumgartian and Kantian sort of aesthetics.
When aesthetics is re-connected to the sentience of the environment, of things, it opens up a way to understand it as a relational practice between people and things, breaking it out from judgment. So, the very idea of aesthetic judgment and the history of photography are somewhat aligned at the turn of the 19th century-aesthetics being that kind of relation between yourself, a viewer, and a photograph in a relation of judgment.
The idea of sentience and aesthetics as a mode of relation between things, that all things in the world exist in an aesthetic relation, is connecting them to a premodernist or often to indigenous INHABITING THE HYPER-AESTHETIC IMAGE Eyal Weizman (in conversation with Jacob Lund) The Nordic Journal of Aesthetics, No. 61-62 (2021), pp. 230-243 conceptions of aesthetics. So, in a sense, it decolonizes both aesthetics and photography. The thing that is important about the photographic for Matthew Fuller and I, and also in Forensic Architecture, is that the photographic is a way of inscription, in which surfaces are inscribed with events that happened near them, in relation to them, etc. It's a process of continuous inscription and erasure, in which there is no privileged surface.
All surfaces do that inscription and erasure continuously. You know, silver salt halides, fixers and the like are built upon phenomena that exist in all material surfaces, from the skin of your body to the forest in the tropics.

JACOB LUND:
This has already led us to your reflections on "how to inhabit the hyper-aesthetic image." In Investigative Aesthetics you write of the emergence of a photographic milieu or image space where digital images join with material surfaces functioning as images in their own right, and propose the hyper-aesthetic condition to be an image-making process in itself, stating "Inhabiting an image means accepting the image-being of all material surfaces, and one's own actions, and constitution, within them." You also argue that we need to read reality as a self-referential image, a metaimage-which seems to resonate with your thoughts elsewhere on re-enactment as a method for turning reality into a model of itself, whereby it becomes an object of negotiation, you could say. Could you explain the notion of the hyper-aesthetic image, and how your work might be regarded as ways to inhabit such image?

EYAL WEIZMAN:
In Mengele's Skull (co-authored with Thomas Keenan, Sternberg, 2012) we were regarding aesthetics as a process-any processof inscription and registration on material surfaces. We wrote that the bones function as a kind of photographic surface in long exposure, in which the life of the person is the image exposed onto the bones and registered within them. So there is a kind of fascination that started there. But there is a problem also, because if every material does it, where is the act? And where is politics? If registration is ubiquitous, sentience is ubiquitousbetween materials, organic or non-organic, minerals, codes, papers, surfaces, metals, human memory-it becomes a world that lacks decisive action. To engage with that Matt and I came up with a category of hyper-aesthetics, which is a practice, to hyper-aestheticize something. It works when you have accepted not only that aesthetics is ubiquitous in relations between things in the world, but also that politics operates in increasing the sentience of surfaces or your ability to read into them. It's like turning things into more tuned senses or regarding them as more tuned surfaces. The first aspect of hyper-aesthetics is that you actually say: well, everything registers, but the wall of that building in Gaza after this attack registers, right now, something crucial for our political strategy, that which needs to be hyperaestheticized in order to confront some sort of dominant narrative or force: everything registers, but in hyper-aesthetics you need to amplify. Second is that you need to network sentient surfaces into other senses. You need to multiply, not only amplify. The second part of hyper-aesthetics-again, it's a political act-is to say that that surface of that building in Gaza that registers a particular chemical residue from an attack could be connected to the material surface of other sentient surfaces: a video, could be connected into the sensation and memory recollection of a witness, into the blurry, still blurry, satellite images over Gaza at that time, or indeed to chemicals in the sea. That connecting and bringing together unlikely sensors is that kind of political act of hyper-aestheticization, with always a danger of that flipping into what we call hypersthesia where an over-registration basically starts racing the signals that are accrued otherwise.
When we are speaking here about Gaza, it's very topical. It's a very painful time. We are speaking now a week or so after the end of yet another gut-wrenching bombing. Again, things are denied.
Again, the truth is also a casualty of the war, and the interpretation of the facts are a casualty, in the way that Israel and its supporters claim that no crime has been committed.
What we are doing there now is understanding, and quite directly in reference to your question, we are not looking at the rubbles of homes destroyed. We are looking at cracks in buildings very far away from the source of bombings. What happened in Gaza during this attack is that the subsoil has been blown. Groundpenetrating bombs enter into the subsoil and blow up some tens of meters under the surface, creating-and in one day in particular on the 14th of May with half of the Israeli Air Force, hundreds of planes dropping ammunition over a very small area-an artificial manmade earthquake, basically liquefying the ground, turning it into a kind of gelatinous thing. It's no longer a scratch as a trace.
But the consequence of the earthquake in a place that was not built for earthquakes could be cracks in buildings miles and miles away from the place of the impact. Cracks that would remain there and that at some point would surrender to the force contradiction that exists on any built fabric and would tear apart and bring the building down. These cracks are also evidence for a war crime, for the indistinct subjugation of an entire population to the colonial rage of Israel at that moment. It exists in those faraway cracks, rather than in the nearby piles of rubble that you see. It exists in the most unexpected small things.
These cracks as registration, as a hyper-photographic record, if you like, means that Gaza now inhabits that image of the war, and not only the sort of image of the war as in what the war has sensed in the material surface of the city. It's precisely those things that are far that could allow us to claim for the indistinct subjugation of an entire population rather than the pinpoint accuracy that Israel wants to claim in doing so.
So, everybody has become forensic architects recently: New York Times, The Washington Post. They adopt our methods.
And it's wonderful: they should. But we are already somewhere else. No, we don't want to analyze those piles of rubble. We are looking at and checking the leaves along the border to find banned munitions. Because leaves are excellent sensors. We hyper-aestheticize the leaves of all sorts of local vegetables; they-for instance variations of lettuce-are very sensitive surfaces when herbicide fall on them. Mobilizing the wind, Israel sometimes spray herbicides along the border waiting for the wind to carry them into Gaza. They register that. They are senses, they are photographs in that sense. (Fig. 1) During that war, and again checking with plants, Israel has bombed several chemical facilities in Gaza, creating also airquakes: the kind of clouds of lethal substances that move quasi-unpredictably across the strip and posit their toxic residue across buildings and kindergartens, streets and mosques, etc. But the leaves register it best, and we have a connection from previous work with a network of farmers in Gaza that allow us to take those measurements. We have the sensors and we have the measurements. So here you have a kind of hyper-aestheticization of the ground, of the crack and of the leave as a political act.

JACOB LUND:
So, you hyper-aestheticize the ground, the cracks, the leaves, etc.
in order to generate an image. You are the agents of this imagegeneration, and that image is then addressed to different forums: to visitors to the places where you do exhibitions, to lawyers and judges in courts of law, etc. It seems to me that there are two where you investigate, build operative models, etc., but after that you also address the image to someone and thereby reinstall human perception and the human mind in the equation. You could say that "aesthetics beyond perception" belongs to the lab and the studio, whereas the findings are exhibited/addressed to a public; even though you strive to collapse the two in a process of "Open Everyone interested in these matters would of course immediately think of the work of Forensic Architecture as a way to actually counter suppressive regimes of visuality, a way to produce counter-images or counter-narratives; making use of images, video, etc. available online or "belonging" to or stemming from state and corporate authorities, which you then make testify against themselves (to paraphrase Farocki-who was actually working on a film on Forensic Architecture when he sadly passed away in 2014).
But when we talk about images there is also this distinction between pictures and images. A picture, according to W.J.T.
Mitchell, is a material object, which can be destroyed or broken, whereas an image is what appears in a picture, and what survives its destruction. The picture is the image as it appears in a concrete material support or a specific place, it is the embodiment of an image. To this I would add, that an image is dependent on recognition-in line with Walter Benjamin's notion of "the image in the now of its recognizability"-an image is dependent on the involvement of a human agent; it is something that develops out of the spectator's relation with the image. The spectator, the one who is being addressed, is invited to take part in a process of signification and sense-making (this means that machines, for instance, do not see images, they merely register and process It's always about extending the conversation. When we put this work with AI at the Whitney, the Triple-Chaser work, 3 obviously it was built based on relations with activists in Tijuana and with Emily Jacir, another activist in Palestine who collected things for us. We had like a big hoovering machine: from all over the world people were sending us images of those canisters. (Fig. 2+3) But, when we composed that film, we said that it has two viewerships. We always said that very clearly. It's put in a museum. You cannot avoid the fact that it's made for those people that would go to the museum, would hear about it through the media, and whatever that would activate. But the video was built to include most of the training set necessary for the algorithm itself to learn how to see. There are thousands and thousands of still images. That is why it strobes so fast. We trained the algorithm classifier to learn how to see the triplechaser, what is not a triple-chaser, what is, etc., against different backgrounds. Like teaching a child to identify images. So, it has a viewership, which is algorithmic in as much as a viewership that is human.
The idea is always to extend that conversation. By the way, an interesting thing for you might be that we realized when we were strobing that image that we were going close to the legal threshold of how fast you can strobe things. They have regulations in different countries because there are people who are "photo-sensitive". Epilepsy is only one part of it, photosensitivity is kind of a medical condition, but photo-sensitivity is also our work condition. This is what we are all about, photosensitive, increasing our sensitivity to images. Rather than take this message out of the film and put it on the wall, we put it in the film because photo-sensitivity is interesting for us conceptually, it is the condition of inhabiting the hyperaesthetic image. This is a condition of our reality, like being flooded, strobed with endless images, and in seeing images through images; images being doorways into other images, an image only being a link, a hinge between multiple, sometimes dozens, sometimes hundreds, sometimes thousands of images, clouds of images that need to be linked together in order to see things and perceive things. We use models as ways of viewing.
Models are optical devices for us in the way that they allow for a navigational viewing that places images in a simultaneity. You don't cut, you move in-between them. Sometimes we build those spaces physically ourselves in order to create the reenactments that are necessary to interrogate the image itself. You were talking about reenactment, probably you were referring to the Halit Yozgat case. The reenactment was a way to examining an image. There was a video reenactment produced by the police.
We thought the reenactment as manipulative. We considered that video is not a representation of a crime, like a reenactment is a mode of representation. That video was itself the crime because we thought that this secret service agent was lying. The crime is in the image making, the making of the video. In order to interpret that image, we had to enact it, we had to put it into practice, we had to practice that image in the real physical space in order to destabilize it and to confront it. And that is how we did it. So, we reenacted the reenactment if you like.

JACOB LUND:
Your use of machine imagery and training these algorithms to detect the canisters from tear gas grenades can be seen as a kind of counter-narrative from within. You are making use of the same tools as the state authorities and...

EYAL WEIZMAN:
Evil people, yes. Like with satellite images, like with photograph, like with video, these things have all military histories, or at least repressive use in colonialism, or in military, or in policing, etc. This is why we have a very important principle in Forensic Architecture, which is, when we use technology, we will simultaneously use it as part of a composition for our case-in a kind of nest building, to refer again to Farocki. 4 But we will also use the case in order to shed critical light on that technology. So, we use satellite images to show us what they can, but we interrogate the threshold of detectability: what they don't show and why. What's the political theory of resolution? For example. (Fig. 4) With AI it's the same.
We followed the Triple-Chaser with a piece called Model Zoo at the de Young Museum in San Francisco, 5 where we said, okay, if we can control the machine vision and machine learning process by feeding training sets synthetically produced, models looking like realistic objects, to tell a computer how to see things: What is it? Like a child as I mentioned earlier. You see this? This is a train. This is also a train. This is a train. That's also a train. This is not a train. Right? You have this kind of process of training, and then you have the results on the other side. Now, what that gives you, creating synthetic training sets, is being able to interrogate, or what we call introspect, the algorithm itself. So, we put like systems of variations, and we see when, in the variation, the machine stops seeing. How does it learn faster? We realized very interesting things in this piece called Model Zoo: if we create "wild" variations, extremely colorful, bizarre, tear gas canisters, tear gas canisters that have zebra stripes on them, or that are imprinted with images and twisted out of shape, those outliers, what we call extreme images, make the identification of "normative" images easier. If you are operating an AI surveillance system or predictive systems on people, what you need as training set is not just the kind of "normative" behavior by many people. You need extreme variations in behavior. You need people to act "crazy" in different ways. That allows you to predict, not the behavior of the outliers, but to predict also the behaviors of the mainstream. We realized that that is the perfect logic to validate why internet companies encourage extreme behavior online, because extreme behavior online provides the predictive algorithm with those outliers. We therefore did two things using machine learning: We interrogated human rights violations in a physical material world, and we investigated human rights violations in the algorithmic world.
That's how one needs to operate across those fields. 4 When Harun Farocki was collecting source material for a film on Forensic Architecture, he wrote to Eyal Weizman to ask for the found media they were working with: bits of blurry user-generated video and screen grabs of software in action, documentation of physical traces and aerial and satellite images, stating: "Instead of designing a film in the way a building is designed I prefer to build a film in the way birds build a nest." See Matthew Fuller and Eyal Weizman, Investigative Aesthetics: Conflicts and Commons in the Politics of Truth (London: Verso, forthcoming). 5 See https://forensic-architecture.org/investigation/ model-zoo.