On Špela Petrič’s Per­for­ma­tive Ethnog­ra­phy or How to Con­t­a­m­i­nate Automation

I am a granddaughter of a farmer and I stand in front of the hectares of “the product” growing, the ground is slippery and dusty, the wind strong, and I feel how my gaze can rest, as I long for this flattened and unified scenery which is quite unusual for my body. I am a new material researcher and I walk the corridors of a medical centre dressed in a disposable plastic lab coat, passing the windows that reflect laboratory tables, pipettes, DNA amplification machines, and storage freezers of generations of DNA. The dry air penetrates my nostrils, reminding me of my unaccustomedness as well as my biopolitical dependence on this sterile environment. I am an immigrant, standing in between lanes of pot plants; UV lights paint the landscape of a harmonious and automated control of this uninterrupted supply for the market shelves and despite my experience how all that I see is occupied by extractive economies, I find myself admiring its sublimity. A monoculture crop farm, a lab at a human genetic centre, a horticulture greenhouse—these are the automated spaces we, the urban middle-class bodies from the global north, are not supposed to encounter. These spaces are to be invisible, where the human is assumed to be decentred or absent altogether, because the space is run by machines. And yet, performing as hidden, automation is believed to condition not only urban prosperity and the nourishment of its bodies, but also our consumer desires and future redesigns of every mundane habit and tool. And while there have been voices criticising AI automation, for example, out of fear of humans losing control over things, out of fear of humans becoming enslaved by the machine, the unifying logic of automation spreads into imaginations and ways of thinking mostly unnoticed.

What happens when the facade of imperceptibility is penetrated, when the consumer not only sees the invisible but is allowed to experience it on their own terms, outside of the logic of consumption? What relations, desires, and values are revealed when one is exposed to automation not as an abstract question of one’s distinct human position in relation to the machine or as inevitable progress fuelling their consumerist desires, but as a complex and unequal experience of being already tightly interwoven with it, of not being innocent of it? Does anything change when you as a particular body stand in the automated monocrop field, in the DNA biobank, in the greenhouse and experience it not as something other, but as something close because felt?

Špela Petrič conceived a series of participatory expeditions into spaces of automation called Performative Ethnography (PE)—an embodied observation practice that creates a possibility for the lay audience to not only experience these spaces but also to employ a particular embodied perspective from which their observation will take place, e.g. a plant, a DNA, a ping-pong player, a vampire, a farmer, an engineer, etc.

PE is designed as a tour where visitors from mixed professional, cultural, and social backgrounds are given tentative instructions on how to situate their observations in an intuitive process of fieldnote making. This means that their observations and reflections are not governed by the regime of truth, but rather follow the principle of what becomes important for each participant when embodying the chosen perspective. In this way, PE becomes a novel artistic practice that combines methods taken from ethnography, role play, and participatory art. As in more experimental role play practices, the main narrative of PE is shaped and controlled by the participants (players),^[1] whose observations and reflections generate the meaning of each visit. Unlike some role-play approaches, however, PE does not operate within the gap between the space of the game and reality,^[2] nor does it ask its participants to simulate their roles, their perspectives, in order to prepare them for what is real.^[3] Rather, similarly to participatory art, each perspective is understood and treated as already real because it is already present and felt yet often ignored within these spaces or discourse.^[4] As such, PE creates a possibility for recognizing bodies and their agencies usually deemed irrelevant within automation, as well as the onto-epistemological consequences of automation that we learned to ignore.

All PE participants share their observations among each other which are recorded as statements after each visit. These statements and observations are then collected into the artist’s archive encompassing three years of research into the agricultural and medical sectors of algorithm-based automation implementations, of interviews and discussions with various experts, visits to hospitals, farms, greenhouses, food factories, medical research centres, and monoculture crop production. This archive became what Petrič calls a multibody.^[5] The multibody are an excess of experiences and values, concerns and stories of bodies that are otherwise hidden in order to fit the logic of algorithmic annotation. Their way of being is close to what Gilles Deleuze described as an assemblage of temporary relations of bodies that are in constant transformation^[6] because they are shaped and influenced by their relational encounters.

The present paper is thus a part of the multibody that learn how to navigate within these mutating relationalities of automation, recognizing, after Yuk Hui, that automation is not “something that happens only in factories” but something that conditions “a broader reality”.^[7] The task is therefore not to criticise alienation of the human that automation generates but to learn how to reappropriate technologies of automation into different practices, including those related to how we gather knowledges, what we understand as relevant epistemological methods within not only art practices. Here, I propose to consider PE as a response to what Hui called for, namely the need for reappropriation of technologies “beyond industrial, consumerist applications”.^[8] As he writes, “to reappropriate […] would mean to create alternatives based on different ontologies and epistemologies”.^[9] If we are to overcome the human alienation caused by automation, more needs to be done than simply repurposing existing technologies, as this would only be a change of functions that keeps the conditioning ontologies and ways of understanding them unshaken. Reappropriation demands an ontological shift of what technologies are and can do. And this, I argue, is what Petrič does in her Performative Ethnographies: she does not propose an alternative that would leave the status quo unchanged, but rather contaminates the ways of knowing that automation fosters. From within the seemingly smooth and clear relations and processes of automation, Petrič proposes ways of knowing that, rather than escaping it, unravel automation as a colonial logic of knowledge universalisation, an imperial desire for risk control in order to find possibilities of its transformation.

As a part of multibody, I participated in four Performative Ethnographies, numerous events organised by Petrič and shared discussions with her, and will therefore take a voice of many and look closely at the ways of how to condition encounters within automation that Petrič exercises in her work. This paper is thus not an art theoretical and formal analysis of an artwork but a multibody reflection on how PE becomes a differing practice of knowing.^[10]

*

Performative Ethnographies take place in spaces that are said to be invisible, inaccessible to the average, middle-class consumer. Thus, when entering these environments as a participant of PE, you tend to first feel like a voyeur—a tourist with a special task, equipped with a notebook and a phone camera, eager to document what you see and think upon seeing and experiencing these spaces. You are told to embody a perspective within which your observation is to be grounded, being invited to embrace a particular notion of fabulation that is close to Erin Manning’s definition, namely “the language spoken by the fugitive public, the public that does not know its place”.^[11] Through the experience of being misplaced, you are to reclaim the tools of misinterpretation and penetrate spaces deemed to be unknown, only to quickly be faced with the awkward realisation that they are familiar.

The concern regarding automated technologies, data collection, and the use of AI for their predictions and operations, as Petrič argues,^[12] is typically discussed in relation to global economic infrastructures, social networks, and geopolitical implications. Instead, Petrič looks at automation in spaces that—because they, as she puts it, have to do with more “primal and basic” aspects of living bodies, such as eating and sustenance—are usually not considered to be of public interest. As she lives in the Netherlands, known for its globally influential horticulture that not only exports goods and high-tech technologies but is also to globally implement the whole design and mindset behind them,^[13] Petrič has a particular opportunity. In her work, she focuses on greenhouse landscapes and medical research centres that implement AI automation as a shared, yet unequal, condition for humans and plants—more familiar to each other one would initially think.^[14] In order to understand Performative Ethnographies as a practice of contaminating the universalising logic of automation, it is important to first map assumptions and challenges of algorithmic automation initially encountered.

*

Within agriculture, AI is often understood as an obfuscated agency without identity, neutral, apolitical, just functional. Defined as “different from natural intelligence found in biological organisms as it is created, artificial, and digital”,^[15] AI is believed to be universally applicable and efficient wherever data can be collected and annotated. However, as Kate Crawford argues, rather than neutral, AI is a combination of various technological practices, tools, and operational systems, which makes it “metamorphic”, i.e. in a constant state of transformation, mutation by its own multiple components, which include “manufacturing, transportation, and physical work […] data centres […] personal devices and their raw components […] datasets […] computational cycles.”^[16] Not recognizing these complex correlations, how they are co-dependent and shaped by each other, obfuscates the material and labour costs that sustain them, making AI an extractivist industry. As Crawford argues, “AI is neither artificial nor intelligent. Rather, artificial intelligence is both embodied and material, made from natural resources, fuel, human labour, infrastructures, logistics, histories, and classifications.”^[17] Under the “promiscuous term” of AI lies what Crawford names as “colonial impulse” to universalise, to centralize power and control over bodies while disguising any political character of its processes.^[18]

Moreover, the growing fear surrounding AI concerns the specialist knowledge and technology it demands as well as worry of possible manipulation and monopolisation of access to it. The solution is often seen in the democratization of AI technology to prevent, for instance, companies from manipulating sensitive data and use it for their own profit.^[19] However, as Crawford stipulates, “to suggest that we democratise AI to reduce the asymmetries of power is a little like arguing for democratising weapons manufacturing in the service of peace”.^[20]

For Kelly Bronson, the actual problem with AI is not access but, rather, our tendency not to recognize its inherent political embeddedness. AI automation is based on harvesting and optimizing information gathered from data abstracted from political, cultural, and social contexts and perceived as raw. This belief in the neutrality of data, in its capacity to “provide truth about the world as it really is”^[21] renders automated systems of decision-making deeply powerful, because naturalized. In this way, AI has, as Bronson argues, a “performative force” because the assumption of the givenness of data it operates with can generate a socio-economic reality of universal quantification that sustains this very ideology by making it habitual, omnipresent, and therefore legitimated.^[22]

A way to challenge violent assumptions inscribed in the way how AI is understood and used might be to unravel the hegemonic ways of seeing and knowing embedded within it.^[23] The question is thus not how to make AI beneficial, how to minimize its risks, how to make it more accessible and therefore democratic. Rather, the task is more mundane and, as such, more difficult—it is about exposing that automation lacks neutrality by mapping the political and cultural desires already inscribed in it, but not for the sake of criticism alone; the task is to recognize these desires in order to change them.

Something thus transforms when confronted with the unquestioned, self-explanatory manifestation of the desire for an omnipresent AI implementation; in the greenhouses, in the medical centres, the concerns of each body whose opinions are deemed irrelevant or ignored are voiced—such as that of a giant, who wonders:

Everything became about patterns and colours. Then I thought about … Yeah … in the Netherlands, the Dutch people really seem to like beautiful patterns in everything.^[24]

The automated landscapes start to reveal their material costs when we hear a rationalist perspective saying how

the greenhouse, in a way … is a success story of the plant, spreading its seed across the Netherlands. But everything is controlled … You do preserve the DNA but you do not allow for mutation. So the plant is just frozen in time … They are doing all this crazy technological gymnastics, but in the end, they are growing flowers, which is something so … It is not futile, but … I don’t know, I don’t buy flowers, I usually prefer to have potted plants, for instance, because I just find it sad that you buy flowers and then they die after a couple of days … I was very overwhelmed by how crazily controlled everything is, but in the end you are just making … These beautiful flowers, you know, we don’t need them, actually.^[25]

Or when we hear a person afraid of the medical gaze, who asks:

How does this information get into the system, what are the categories involved in that …? Do I have any ability to change those categories? Do I have the ability to even see those categories …? For example, someone who is trans, like myself … It is probably important for them to know that I was born male for certain types of medical procedures. But I do not want to be discriminated against for that reason … How is that information released among people? Who gets that information later?^[26]

Something transforms when in the seemingly smooth space of unquestioned progress that justifies all means we hear a terminally ill African patient, who asks:

In the context of the myth of foreign medicine as toxic and good for the white man … Not a black dude like me … So I am thinking of the costs of all this … How much would I pay?^[27]

The transformation that is happening within Performative Ethnographies as the multibody materialises and shares the participants’ observations concerns the previously mentioned metamorphic and performative character of automation. Here, automation is suddenly discussed and reflected upon not as an autonomous entity, but as something very close, co-dependent on socio-political and economic ideologies and their implications, as something that has many relations of importance, as something that is felt. Here, within such encounters with automation, what becomes exposed are differing ways of seeing and desiring. As such, I argue, PE creates conditions in which we can experience the transformation of the unifying and optimizing logic that drives automation into something more complex and murky—we can experience what I call contamination.

*

Contamination as an event and an idea has been positioned against the category of purity and is therefore something that science as well as humanities have avoided within their regimes of disciplinary boundaries, their notion of truth, and their need for clarity in definitions of bodies and phenomena. Understood as a transformation and change, contamination would only be harvested as long as the risk of a change can be mitigated and controlled.^[28] For instance, within the implementation of digital automation, as Nishant Shah reminds us, AI in particular has been promised to “avoid unwanted contamination”^[29] and to ensure cleanness, reliability, and transparency of its technology. Imagined as neutral and incorruptible by any cultural values and desires, AI became a myth of cleanness that disguises “how computational technologies are ontologically and manifestly produced through multiple layers of contamination”.^[30]

But what happens when we embrace contamination as a relational way of being of bodies that excludes any possibility of purity and neutrality, when the risk of how these relations might cause bodies to change becomes an ethical guide rather than a thing to manage? “Contamination as such, rather than being opposite to purity, becomes a way against ideologies and practices that manage risk according to given assumption of who deserves to live, to mutate and to change and in what way.”^[31] In PE, contamination thus becomes a possibility for reclaiming automation from the regimes of predictability and algorithmic purity of annotations by affirming the risk of the many that cannot be predicted and fully controlled. As Shah argues, AI automation and its algorithmic governance are already “messy, leaky, and violently militant in their everyday practice”.^[32] Rather than cleaning automation or insisting on its purity, PE undertakes to multiply that which becomes important, affirming in this way the ontological plurality and “messiness” that comes with this affirmation.

PE embraces contamination which involves affirming bodies and their relationalities that generate data. Through the exposure and subsequent assertion of the lack of neutrality of data, we realize that they are messy and cannot be smoothly quantified and ordered according to the logic of efficiency and uninterrupted production. PE exposes how this data is made by various bodies that are complex and have different desires and relationalities which are equally important. In this way, in PE, contamination is an embodied practice of reclaiming data collection for values and meanings that are close to the multibody—to participants experiencing and voicing their reflections and concerns—in order to make their reflections tangible. Through the multiplicity of embodied and partial voices that the multibody becomes, data as matter of facts is transformed into matter of concern—it becomes opaque, in the sense as Édouard Glissant described it: “The opaque is not the obscure, though it is possible for it to be so and be accepted as such. It is that which cannot be reduced, which is the most perennial guarantee of participation and confluence.”^[33]

Petrič’s PE thus creates a possibility to embrace AI opaqueness—not only repurposing it, but making it felt, affirming cracks, qualitative meanings, and uncertainties in order to complicate the smoothness of the rhetoric of algorithmic progress and utility. In PE, Petrič conditions experiences and relations that take neither the position of the outsider nor the expert insider, but embodied particularity. As Linda Smith would argue, this particularity of bodies deemed without a voice or its relevance expresses connection “to land, place, stories, context rather than claiming a universal authority over experience and people who can speak for themselves”.^[34] In this way, PE facilitates embodied political encounters where the belief in the need for data collection has been unquestioned. Without dismissing algorithmic automation, or criticising the data it collects, Petrič contaminates the narratives of automation—she reclaims what counts as data, performing, in this way, a feminist work, and a decolonial work.

*

Špela Petrič created a protocol of an encounter that has been repeated and used for various automated spaces during all the series of Performative Ethnographies. The protocol relies on each participant embracing and embodying their chosen, imagined perspective of observation. By inviting fabulation as a driving force of observation and meaning, PE escapes generalisation, opening a mode of thinking and experiencing that resists the priority of the value of truth as the measure of what is considered to be of importance and meaning. As such, PE becomes a “fictioning practice” as defined by David Burrows and Simon O’Sullivan, namely as a practice of “performing, diagramming or assembling new and different modes of existence through open-ended experimentation”.^[35] It affirms the narrative of the multibody: complex, awkward, and not fitting into any “deliverables” as recognized by a discipline. It embraces bodies whose experiences and meanings are likely to be considered as irrelevant, such as a con artist who

came here looking for opportunity,^[36]

an ECG cable who

also wants to be treated with respect and admiration like the AI system,^[37]

an immigrant who is concerned about the privileged access to the spaces of automation, a patient who in the face of the claim that the newly designed hospital is to feel like home reflects how it

does not look like hospital […] so this feeling at home is filled in here as a corporate space […] It is more like a shopping mall, where we can consume. So we are at home when we are consumers.^[38]

By making unlikely bodies speak while performing the work of observation and sharing, PE creates a reality where the assumed flatness of the world made by the universalising logic of automated annotation, the clear-cut borders between what is real and fictional that it sustains, becomes rather porous and complex, and its tentacles spread into many urgencies.

Affirming fabulation, worlding, “connections that matter”,^[39] PE mobilizes the truth rather than questions. It affirms that what is neglected and made invisible in these spaces—it “creates a trajectory, and from there the potential of the what else emerges”.^[40] In this way, it conditions transversality of an encounter as defined by Félix Guattari. It is not a higher model, a more abstract way of representation, but rather it incorporates multiplicity, chaos into the existing models, contaminating them from within, disrupting their representative claims of clarity, stability, universality.^[41] PE betrays the method. It contaminates and hacks the institutional distribution of power and what counts as meaningful. It multiplies understandings without seeking their hierarchical organisation into neatly ordered categories. Here, PE appears to be research into modes of being of bodies and into their relations within automation rather than a method of their control and data clarification.

There is thus a paradox that PE is situated in as it uses the tools of ethnographic mastery in order to refashion another tool of mastery—systems of automation. As it belongs to anthropology, ethnography is argued to be an imperial discipline that helped to dehumanise indigenous peoples by shaping the representation of indigenous knowledges as radically other from what is known to be scientific and modern. Enacting the “ethnographic gaze”, as Smith argues, anthropologists exploited lands and their people as part of the science of imperialism.^[42] Petrič applies ethnography as a tool of imperial knowledge universalisation as well as an artistic tool of intervention into the spaces these knowledges helped to create—the greenhouse, the hospital.^[43] The artist thus uses the ethnographic method that historically fetishised data and neutralised the performative aspects of knowledge it produced^[44] to reclaim its transformative and performative power. This reclamation happens firstly by acknowledging the existence of different bodies and their stories (those not recognized by Western science), and secondly by affirming these stories as important data. As Fabian and de Rooij explain, “questioning ethnography not only pertains to the ‘how’, the manner in which knowledge is produced, but also to the ‘what’, the kind of knowledge one seeks, and ‘what for’, the purpose such knowledge is made to serve”.^[45]

PE contaminates the method of ethnography, explores its etymological roots by saturating the performative aspect of it and by embracing the situatedness, ambiguity, and possible contradictions it may lead to. It is thus not surprising that Petrič’s work has been criticised by anthropologists and ethnographers. The main argument against her use of ethnography in PE is that it is not rigorous enough, that the artist does not follow its methodological clarity, that she mocks what they, the scientists, do.^[46] However, considering the decolonial urgency of investigating how knowledge-making practices are shaped, it is exactly the lack of rigor in its practice that makes PE powerful. By betraying the regime of the method, Petrič exposes the partiality of meanings created by PE participants; they become invested with the spaces they encounter when embodying a perspective, but they are also already shaped by it. What is more, PE is designed to avoid impartiality. Rather than mediated and mediating, participants experience relationality and the transformations emerging from it that were both caused by their new embodiment; consequently, what is already known becomes questioned and untrustworthy.

Through PE, Petrič explores “a new politics of knowledge”^[47] where what is research and what counts as knowledge and its tools is redefined and reclaimed. She challenges the universality of knowledge and its seeming objectivity that it claims to produce, and in this way embraces the practice of decolonization within automation—in PE, she exposes the questions of for whom and by whom the knowledge, values, and meanings are made in these spaces, and thereby creates the possibility for their change. PE thus decolonizes not only the very institutions of automation implementation as the spaces of inaccessibility and abstraction, but also the unquestioned authority that is distinct from the unskilled, lay person body.

*

It is after encountering the spaces of automation, when participants are asked to share their observations, that PE exposes the multitude of its implications. In this moment of voluntary sharing, the multiple perspectives collide with each other—no one is asked to agree or disagree so as to compose a unified picture. Rather, the participants affirm the moment of voicing their experiences, worries, moments of appreciation, wonder, pleasure, and concern. Some perspectives contradict each other, some fold on one another, giving new layers to the complexity of realities made flat by the ideology of functionality. The porosity and stickiness of things and bodies become more and more tangible with each new story, each new body telling their observations and feelings. And while listening, you realise how much we are all implicated, albeit unequally, in the problems voiced, in the worries and desires materialised.

In this way, Petrič’s PE responds to Stengers’s call for cultivating connoisseurs, the ones who can hold the science and those who produce it accountable. Stengers calls them a “distributed amatorat”—a multiplicity of those who care. Science systematically distances itself from the public on the false and distorted assumption that the public is incapable of grasping its importance and nuance, argues Stengers, creating the need for connoisseurs “to hold scientists to the task of taking care when making normative judgements about what does or does not matter”.^[48] This distributed amatorat does not represent any position that would be recognized by a given institution or discipline as relevant, they do not seek “professional recognition”, they do not seek to find a solution for a given problem and be credited for it.^[49] Rather, they weave their stories into situated and embodied matters of concern, materialising voices and meanings of what becomes important.

Here, while sitting in the elegant conference room of the Department of Genetics, UMC Groningen, we hear amatorat concerns expressed by a DNA, plotting a possibility of their thriving:

I am the source of extraction […] that kind of produced certain conflicts within me. I could see compartments, where my brothers arrived and land, from here, they were also extracted. There was a lot of talk about microscopic perspective, while for me, there was a lot of macro perspective […] I do not know how I feel being copied so many times, 100 000 of times, and that there is also some selection process happening, that some parts of me are not as desired as some others. I think this species have a very certain perspective of what is good within me and what is not. I entered this space thinking I am unique […] it was striking they only keep part of what they need […] It seems they avoid organic matter splitting everywhere. I imagine that in my desire, I would have blood and organic liquids everywhere, while here […] everything is so white and clean. I miss the messy, organic multiplication. They are not able to read us, when there is not enough of us, this is our strength.^[50]

And while walking through the fields of the monoculture crop production, among the voices of a farmer’s granddaughter, who is jealous of these monoculture owners because she lost her heritage, an organic farmer, who is not satisfied by the kind of innovations implemented, a trans-multispecies researcher, and a technical person, we hear a deer, who reflects:

I don’t hear anything about animals … Animals are being allowed to be on plants when they are useful, but when they are not, we have to get rid of them. As a deer, I like to have everything around me and I think everything is useful and everything is necessary to live for all of us, people and animals … And even when I look around, everything is flat with only cabin for his machines. And when I looked at his place, where he lived, it was a nice house with lots of trees around it. And my house, the house of animals, was just flat.[51]

The multibody of PE—the amatorat—is not here to create a coherent analysis of automation. It is not here to demonstrate standards and results that would be ready for automated annotation or traditional artistic commodification of object creation either. The amatorat is not to be credited according to the priority given to the disciplinary authority. The authorship is distributed, but not according to the logic of democratization. In PE, the multibody’s multiple voices are embraced not in order to create a cumulative and unified meaning, but to expose embedded asymmetric power relations and values that they generate. PE reveals that there is something else more important than the representation of what is there, of a given truth, and identification of credits.

PE as an art work becomes an ongoing event. Lacking any traditional deliverables, it resists disciplinary categorization and expectations, because there is something more important—something that is situated in bodies’ encounters and the stories and worries they voice. In PE, Petrič thus seems to follow what Melanie Ford Lemus and Katie Ulrich argues is at stake when experimenting with ethnography, namely that “it entails the risk of not arriving at a predictable outcome, of not arriving at any one outcome that could have been imagined previously”.^[52] In doing so, PE methamorphs what is already methamorphic—it contaminates our understanding of what automation is and does into something else, something shared and multiple.

*

I left the building of the medical centre. A strong cool wind welcomed me while in my head, I still heard one of the guides who, upon explaining the intricacies of applying AI to genetic medical research and newborn screening and being met with various questions and concerns from the amatorat, quickly refuted: “But these are ethical questions.” I stood for a while, looking how people enter and leave the hospital, moving as if in a machinic rhythm of exchange of information, wondering about their proximity to becoming a data sample. Overwhelmed by how, in the eyes of automation, we have so much in common with greenhouse plants, as the language and processes constantly optimize the messiness of our bodies, I hear Petrič asking: “Who is actually driving the change?”

Performative Ethnography is a pluriverse practice that resists “the modern monification of the world”,^[53] it embraces the many that come with affirming the uncomfortable. This discomfort is created because PE does not give answers. Rather than becoming any moral imperative, it makes visible a-symmetric relations, awkward dependencies, intimate desires, and worries that need to be acknowledged. And as such, it becomes a practice of onto-epistemological differing that reclaims realism as not “belonging only to monistic methaphysics”.^[54] It breaks with distantism as a necessary condition of knowledge and instead proposes listening and sharing as a practice of affirming “a plurality of epistemic and aesthetic realities”^[55] without the analysis of their hierarchical order and epistemological validity.

PE pluralism also manifests in what Erin Manning described following Deleuze as minor gestures, movements of practices that escape capture, that induce change and transformation from within by introducing a shift, “unmooring its structural integrity, problematizing its normative standards”.^[56] By initiating a fraction that multiplies and thus disrupts the major, the universal, a minor gesture creates new relations that are to be felt, experienced rather than annotated and defined. And as such, PE is not an object, an artwork, but that which makes an art work; it is “absolutely singular and infinitely multiplicious”, it is “a gesture that opens experience to its limits”.^[57] That is why it is risky, it is difficult; and to be visited by the amatorat is troubling because it does not offer predictable outcomes, it does not allow for the predetermined logic of its annotation—and this is what makes PE necessary.

PE necessarily troubles “institutional frameworks in the same way they trouble existing forms of value”.^[58] Being a non-art work, being a minor gesture, it does not offer any fixed form, any conclusion, because it multiplies realities, or rather shows how reality is already multiple. As Manning helps to explain: “When the art work exceeds its status as an object, when the work becomes relational rather than simply interactive, when there really is a sense that what is at stake is more than the sum of the artwork’s parts, present in each of the fields of action the work emboldens, is not the work as such, it is how the work works.”^[59] PE thus becomes an activist philosophy, a vector for “decolonial imagination”.^[60] As Manning would argue, it creates new ways of practicing that are attuned to change and movement of bodies—she calls it “choreographing the political”, and it “begins here, in the midst of shapeshifting speciation”,^[61] in the spaces seemingly depoliticised by the convenience of alikeness. By creating conditions for the amatorat to challenge the perspective of the one, by its affirmation of the multiplicity within automation itself, PE “keeps things unsettled, [it is] a push that ungrounds, unmoors, even as it propels”.^[62] Bodies practicing PE create a possibility for things and bodies deemed as irrelevant to be felt, experienced, and voiced differently than the disciplinary methods and algorithmic logic of smoothness allow. The amatorat affirms the complexity and tension, making ignored problems pertinent because embodied and told in their own way and order of relevancy. In this way, Petrič’s PE decolonizes the imperial desires for mastery and universalisation of automation not only within institutions and disciplines that enforce and justify its power relations, but also within the mundane spaces of daily habits of bodies overlooked that automation inevitably shapes.

By “striating more than smoothing”,^[63] Petrič does not look for an alternative. Instead, she creates a possibility of exposing how automation that governs our bodies is already often a violent part of us, how we are already closely implicated in its various desires and mechanisms of extraction and creation. But rather than despair, the artist contaminates, affirming an uneven relationality and reclaiming automation for the amatorat in its opaqueness, in its uncomfortable, murky, and obscure because intimately interwoven multiplicity, inviting us to feel how we might become differently.