Michael Richardson’s War, Data, and Ecology after Nonhuman Witnessing reimagines the act of witnessing in the age of AI as a human-nonhuman relational process, exploring technological warfare, ecological crises and algorithmic hierarchies. Janica Ezzeldien lauds the book’s interdisciplinary depth, conceptual innovation and its critical engagement with knowledge production, ethics and agency beyond traditional human-centric frameworks.
At a time when human and nonhuman modes of communication are becoming more complex and fluid, with countries such as Germany introducing the use of AI in asylum procedures, research that aims to understand these shifts is critical. Michael Richardson’s Nonhuman Witnessing: War, Data, and Ecology after the End of the World takes up this task, making a timely and essential intervention into dominant conceptualisations of witnessing and perception.
The book reveals the limits of traditional frameworks and confronts the paradox of knowledge production within dominant European thought (8), questioning the reduction of nonhuman forms of witnessing to “evidence” and their dependency on human interpretation. Instead, he elaborates witnessing as the entanglement of human and nonhuman entities regarding technoscientific warfare, ecological crises, and algorithmic control (3). Observing how “witnessing is and always has been nonhuman,” he suggests reconfiguring witnessing as a shared human-nonhuman process of knowledge production and thus proposes “nonhuman witnessing” as an analytical concept to explore “diverse technical, media-specific, and situated relations that produce ethicopolitical knowledge without privileging human actors” (4–8).
Tracing these entanglements across technological, environmental, and infrastructural systems that mediate crises, Richardson illustrates their role in shaping knowledge, justice, and political subjectivity (3–7). He develops the concepts of “violent mediation,” “machinic affect,” “ecological trauma” and “radical absence,” to capture the complexities of nonhuman agency (8). Structured around conceptual “doublings,” the book offers a reconceptualisation of witnessing as a relational process that recognises its emerging tensions and the issues of accountability it poses. Opening with the February 21, 2010, Uruzgan helicopter attack, in which the US Army killed 20 Afghan civilians, Nonhuman Witnessing poses the central question of ” who – or what – bears witness?” This point of departure leads him to shed light on the range of nonhuman agencies implicated in the event and its documentation i.e. the material and technological infrastructures entangled in violence, the drone, its sensory apparatus and algorithmic surveillance tools.
Using Microsoft Flight Simulator (2020) to illustrate how algorithms generate realistic 3D visuals, Richardson highlights how these algorithms produce ‘techno-affective milieus of witnessing’
Chapter One uses to A. Bousquet’s (2018) theorisation of a “martial gaze ” – a mode of perception and destruction through a “range of sensorial capabilities,” imaging and mapping “relevant to the conduct of war” (40–42) to develop the notion of “violent mediation.” These mediations exemplify the invisibility and diversification of targeting systems characterised by “layers of simulation, data, modelling, and algorithms” (39). Focusing on interconnected technology, bodies, and environments, the chapter illustrates how these technologies bring to light a shift in agency from the human to networked relations (38–39).
Chapter Two considers machine learning algorithms’ capacity to create worlds. Using Microsoft Flight Simulator (2020) to illustrate how algorithms generate realistic 3D visuals (80-1), Richardson highlights how these algorithms produce “techno-affective milieus of witnessing” – machinic affects that lead to “relations within and beyond technologies, bodies, and ecologies” (10). A look at Deepfakes and Triple Chaser further emphasises how machine learning blurs reality and shapes perception in areas from digital aesthetics to forensic investigation (93-5), shaping power structures while remaining inscrutable.
Considering the intertwined aestheticism and affectivity, and between trauma and mediation, Richardson identifies the witnessing of what is otherwise unperceivable
Chapter Three then centres the ways nonhuman witnessing connects with ecological trauma during climate catastrophe and nuclear threats. The author draws on B. Massumi’s (2021) conceptualisation of affect as the entanglement of “the virtual in the actual and the actual in the virtual,” to capture the affective dynamic of ecological trauma and G. Deleuze’s (2001) works to engage with “witnessing ecologies” (147-8). Chapter Four extrapolates this theoretical line to focus on the ISIS 2014 video A Message to America and the disappearance of Malaysian Airlines Flight 370, also in 2014, to examine the affective intensities of absence present via technological mediation of digital media, “affective ruptures” (156). Returning to Massumi, Richardson reflects on how digital mediation not only shapes trauma but also informs collective memory and fluid subjectivities.
Considering the intertwined aestheticism and affectivity, and between trauma and mediation, Richardson identifies the witnessing of what is otherwise unperceivable – events that fail to materialise yet remain affectively potent through more-than-human mediation (6–7, 157). The chapter emphasises the nonhuman nature of such “radical absences,” as both traumatising and capable of articulating trauma, an act of witnessing absence and the absence of witnessing (151).
Richardson conceives of witnessing as a form of converting diverse forms of information, involving the identification, filtering and reproduction of patterns of communication, a form of translation
The book concludes with a consideration of the COVID-19 pandemic as a reminder of how “nonhuman agencies […] impinge upon and transform us and our ways of living in profound and immeasurable ways” (174). Throughout this chapter, as well as the book more broadly, Richardson demonstrates how algorithmic systems perpetuate historical hierarchies, reinforcing structures of exclusion and control. Using S. Wynter’s critique of colonial knowledge systems and the European conception of the human alongside S. Noble’s study of search engine bias, he formulates an in-depth critique of algorithmic “neutrality” (3-6, 83). Forced to confront these human-nonhuman entanglements in all its variety, the book thus recognises nonhuman witnessing as a distinctive communicative modality challenges conventional politics by expanding the scope of agency beyond human actors (174).
Nonhuman Witnessing provides a conceptually innovative resource for students, scholars, and practitioners in media studies, environmental humanities, philosophy (applied ethics), and security studies, looking at the role of nonhuman entities – such as machines, data systems, and ecological processes – in witnessing.
The book’s thoroughly presented, in-depth explanation of its key conceptual argument makes a strong case for paying attention to the processing, interpretation and translation of knowledge beyond the realm of semiotics (6). Richardson conceives of witnessing as a form of converting diverse forms of information, involving the identification, filtering and reproduction of patterns of communication, a form of translation. It would be interesting to see whether this conceptual mode could be extended to the notion of translation and interpretation more broadly. It could be brought into generative conversation with Douglas Robinson’s (2024) argument concerning “(machine) translatability,” considering the performative power of language to reshape, rather than to passively reflect, reality.
Departing from anthropocentric and Western-centric ideas of conception, this book invites the reader to rethink witnessing. We are asked to look at what it does in its relationality in the process of the production of knowledge (8). The book brings together key insights from the literature on affect, works on trauma and memory with a well-developed account on computer vision, machine learning and the workings of algorithms. In so doing, it offers an essential, critical, and original intervention into dominant understandings of witnessing. Alongside an empirically-diverse selection of case studies, it is rich in analytical depth, bringing the nonhuman to the fore. Nonhuman Witnessing provides a conceptually innovative resource for students, scholars, and practitioners in media studies, environmental humanities, philosophy (applied ethics), and security studies, looking at the role of nonhuman entities – such as machines, data systems, and ecological processes – in witnessing. It creates space for an interdisciplinary engagement and invites further research into contemporary developments in communication, knowledge production, and the construction and dissemination of registers and data.
- This article originally appeared at the LSE Review of Books blog.
- Subscribe to LSE USAPP’s email newsletter to receive a weekly article roundup.
- Image: Free_styler on Shutterstock.
- Please read our comments policy before commenting.
- Note: book This review gives the views of the reviewer, and not the position of USAPP – American Politics and Policy, nor the London School of Economics.
The post Book Review | War, Data, and Ecology after Nonhuman Witnessing first appeared on USAPP.