Feral Robotic Birds

Lou Terry

‘the ancients observed birds erupting from spaces in the sky and invented letters from the shapes they described.  the birds, in return, insisted on the possibility of song’                                     

Sean Bonney – on the relationship of hermits to the means of information

Abstract

Since the 19th Century, observational sciences have represented the world through “precision measurement, numerical data processing, and the representation of scientific information on a global or cosmic scale” (Gabrys 2013, p. 133).  In her research into how sensor-based technologies ‘in-form and “program” environments’ (Gabrys 2013, p. vii), Jennifer Gabrys describes how observatories have formed useful data repositories, from which environmental monitoring has been able to draw correlations between data over time, to explain unified geological concepts.  In this way, certain objects that the data represents are seen as causatory, forming a topological system with a ‘source’, or explanation.  More recently, climate change has steered monitoring towards figuring environmental change through a root cause, greenhouses gases, particularly CO2.  Greenhouse gases’ absorbing of solar radiation reflecting from earth is connected to increased tropospheric air and sea temperatures, melting sea ice and glaciers, increased sea level rise, and habitat change and destruction.  Through observational science, molecules, previously invisible to us, have been revealed as global actors of adverse environmental changes for much of life on earth (and also, paradoxically, why life can exist here at all).  However, in light of Donna Haraway’s, ‘it matters what stories we tell to tell other stories with’ (Haraway 2016, p. 12), it’s important to interrogate how the ways in which we figure environmental changes govern our approaches to environmental challenges.  Currently, the atmosphere is often viewed as an exterior entity, which we need to act upon, or ‘fix’.  In response, a host of technologies from less polluting cars, to renewable energy projects and even weather control systems have been developed.  However, these technologies can fail to address, or draw our attention to the myriad agents of environmental change (such as our behaviour), which are multi-dimensional, and are entangled with the historic apparatus that forms the nature and culture we swim in.  In 'Feral Technologies', Anna Tsing encourages us to consider the historic collateral of technologies; ‘You can’t just take the intent and say “that’s the technology, the thing we’d like to have work”’ (Tsing 2016).  The car is something that drives, but also something rusting in a dump.  She explains how ‘feral’ means to notice the history making of non-humans we get involved with, particularly those outside our purposes, which are crucial to understanding the environmental challenges we face.  To understand these challenges we need a different framework from unifying theories, one that dives into the noise, and listens to the confusing and heterogeneous stories of environmental change..  Through creatively led practice combining multi-species storytelling and actor network theory, this project explores the implications that engagement with non-human organisms, specifically through sound, and with regards to artistic modes of perception as well as scientific, has for forming critical insights into environmental and ecological change.

Actor Network Theory

How might we begin thinking about the landscape of environmental change in terms of feral histories? Actor Network Theory (ANT) is a change of topology, which encourages us to think not in terms of spatial dimensions or surfaces, but in ‘nodes that have as many dimensions as they have connections’ (Latour 1996, p. 370).  It ascribes liveliness to all entities in an environment, (hence ‘actors’), and explores the chaotic interactions (the net-‘work’) that are done between them.  The landscape becomes a dynamic stage, that the actors + the network terra-form.   If we follow ANT to its extreme, landscapes themselves do not exist, because there is nothing that isn’t actors or the net-work between them.  The landscape is in us, and we are in it.  This conception of nature echoes Deleuze and Guittari’s cultural theory of the rhizome, which ‘resists the organizational structure of the root-tree system’, and takes instead the rhizome, characterized by 'ceaselessly established connections between semiotic chains, organizations of power, and circumstances relative to the arts, sciences, and social struggles' (Deleuze & Guattari 1987, p. 6).  Through ANT we can start to see the unpeeling of natural landscape as an exterior entity upon which we enact, and instead see our entanglement, in it, from it, with it, as it.  Cultural and natural landscapes become wrapped up as one, removed from distinctions of ‘created-by-nature’ and ‘created-by-human’, both become stranger, entangled, the result of historical (and perhaps pre-emptive) work of multiple actors.  

ANT's capillary-like nature echoes Haraway’s ‘string figures’ and ‘speculative fabulations’ (SF). However, whereas ANT specifically ‘does not attempt to fill in what is in between the filaments’, SF encourages us to think about our meaningful connections with our non-human actors;

‘pass-ing patterns back and forth, giving and receiving, patterning, holding the unasked-for pattern in one’s hands, response-ability; that is core to … staying with the trouble in serious multispecies worlds.  Becoming-with, not becoming … is how partners are rendered capable. (Haraway 2016, p. 12)

If we extend ANT through Haraway’s concept of becoming-with multi-species kin, we can think of ANT not just in terms of the historical apparatus Tsing refers to, but also as a potential-actor-network, where bodies, entanglements and stories are at every moment potentially becoming.  As actors with response-ability ourselves, we are implicated in rendering what becomes and what doesn’t in the potential-actor-network.  We are all, along with our multi-species partners, environmental citizens.  How might we articulate environmental change so as to render this citizenry and kin-ship, and also the multi-dimensional and entangled nature of landscape?

Sensing Environmental Media

In 2011, during a fieldwork residency in Lapland exloring how ‘environmental monitoring articulates distinct practices and politics of environmental citizenship’, Jennifer Gabrys considered ‘how creative and community-based monitoring projects might differently attend to expressions of climate change as they occur through connections made across entities and milieus’ (Gabrys 2013, p. 112). Mentioned are examples which monitor complex and lived relations between more conventional variables, such as the atmometer, which emulates the experiences and processes of transpiration within a plant (Ibid, p. 127), or the Exchange for Local Observations and Knowledge of the Arctic (ELOKA), which documents traditional knowledge and community monitoring activities (Ibid, p. 124). Also included are multiple examples of non-humans as expressions of environmental processes, such as the Pika, whose stress hormone metabolites in fecal samples help monitor habitat changes in mountainous regions.  Like technologies, organisms are assemblages of environments, quite literally inheriting, working through and living with changes.  Gabrys calls them ‘environmental media of sorts’.  Organisms as environmental media are interesting as variables because they are simultaneously the multi-dimensional nodes ANT proposes we think with, but also frame non-human kin as environmental citizens, whose lived experience is the stage onto which environmental change plays out.

This framing also implicates our ‘response-ability’.  There have been many instances of non-human environmental media across science and art, though not all unfortunately have fully considered this ‘response-ability’ with our non-human partners.  Like the canaries as carbon-monoxide detectors in mines, so-called ‘sentinel’ species are often canon-fodder for informing guides to human action, with human safety in mind.  This isn’t to say multi-species work is going to be clean, when ‘staying with the trouble’.  As Haraway notes regarding Batman Park pigeon loft (a birth-control hatchery occupying on a colonized wetland previously the Wurundjeri people’s territory, that encourages invasive pigeon species out of Melbourne), ‘the municipal pigeon tower certainly cannot undo unequal treaties, conquest, and wetlands destruction; but it is nonetheless a possible thread in a pattern for ongoing, noninnocent, interrogative, multispecies getting on together’ (Haraway 2016, p. 29).

An example of non-innocent and interrogative multispecies work is the mollusk Corbicula fluminae as sentinel species, used by ecotoxicologists for their acute detection of metal pollutants in rivers.  Through such practices, one would hope to highlight the sensitivity to which many non-human kin bare environmental change, and lead to legislation that considers their well-being with regard to damaging practices (the ‘possible thread’ Haraway talks of).  But we must also consider how non-humans are rendered in the process. The breeding and dissection in labs of Corbicula fluminae re-entrenches exploitative and dominant attitudes towards non-human species (attitudes which have previously driven hazardous environmental change), and so far guides authorities on human consumption of certain freshwater species.  Furthermore, the culture of precision measurement means many more Corbicula fluminae are bred and subjected to adverse conditions, than if we were instead a culture that focused on fostering action to a rougher premise that the river was becoming polluted.

I would like to see a shift of focus that more deeply considers the non-human modes of consciousness we work with in multispecies practices. We should not necessarily look to recognise the forms that accrue through partnering, as forms that work for us.  You don’t create a carrier pigeon, a pollution detector, racer or pet without severing other potential ‘becoming-withs’; non-human versions of dreams/desires/ideas about oneself, and oneself in relation to the community of other non-human actors. Art-science-works like Elaine Gan’s 'Fungal Clock' and Amelia Black’s 'Amphibious Architecture' approach multispecies worlds more in this way of care and attention to non-human consciousness, and frame us as listeners (and re-actors), as opposed to controllers/utilisers.  In 'Fungal Clock' we experience multiple modes of time/potentiality, intention/unintention, between pine, oak, humans and matsutake mushrooms.  In 'Amphibious Architecture' both human and non-human inter-actors are rendered as environmental citizens, where humans can ‘text fish’ and receive information on water quality, fish numbers and human interest, thus engaging people in perceiving a shared citizenry with urban non-human kin.

Sound As Sensing

For monitoring multispecies actors-in-networks/networks-in-actors, with a power to reconstitute our ontology of who/what can count as an actor, or environmental citizen, I want to suggest sound as particularly rich and multi-faceted.  Sound-artist/ecologist Jana Winderen interprets the idea of organisms as environmental monitors by listening to them.  In ‘The Listener’ Winderen says if we identify the sounds of specific species, we can understand river health through focused listening’ (Winderen 2016). Through technology she extends our acoustic horizons and amplifies inaccessible sounds of landscapes, such as glaciers her piece ‘Isolation/Measurement’.  Sound can be understood as an abstract, non-specific and simple language shared across insects, icebergs and humans, not unlike the ‘infralanguage’ Latour refers to in ‘On actor-network theory’ (Latour 1996, p. 377).  As artist Anri Sala says,

“Sound is one step away from language…but once you are in the street, most sounds are untamed, they are in the act of becoming.’ (in. Toop 2019, p. 170)

Sound is not a text, or a visual object, but is a relational mapping through fluctuations in air.  It passes through circumstances, traversing micro-macro scales, echoing and reflecting states of landscapes, and recomposing itself in and out of beings through voice, and through time and space.  Thus sound as voice can be understood as a deconstruction of individuated expression and of self, and instead an entity unto its own, a relational history of sonic reflections, prompted and provoked across multiple bodies, entities and experiences.  You might say ‘I am hungry’, but the bacteria in your gut make your belly rumble.  Simultaneously however, sound reterritorializes individuals in space, it’s a ‘here I am’ of something there.  It’s this double-edged quality of sound as both relational history, but also the culmination of this history into the seeming expression of oneself as voice exits the body, that makes it interesting as environmental media.  As Dominic Pettman puts it, sound is an ‘aural punctum’ with ‘the potential to pop or connect…ontological bubbles’ (Pettman 2017, p. 75) and furthermore voice is a ‘sonic exploration of ontological conditions’ and ‘[tests] the world and one’s location, role and value in it’ (Ibid, p. 54).

Soundscape ecologist Bernie Krause demonstrated sound’s power to tell the story of environmental change.  In his ‘Great Animal Orchestra’, various sonic niches of frequency and time are filled with living sounds, compared to recordings taken a decade later after deforestation.  Silence is a powerful indicator of effects on biodiversity.  Despite this, not one of the 50 essential variables used to inform the UNFCCC and the IPCC is sonic.  This is most likely because, although intuitively meaningful in an individual sense with known context (as with Kraus’s work), the feasibility of deriving meaning from sound on global scales, with current scientific methods, is extremely difficult.  So what steps are being taken in order to overcome this, whilst maintaining an engagement with human and non-human environmental citizenship?

Over the past few years, Dan Stowell’s ongoing research into automatic birdsong recognition through machine-learning has resulted in an app, Warblr, that allows people to engage in identifying bird species present in an area through their songs.  The algorithms focus on categorization, and so figure sounds and birds through species-groups, as opposed to expressive sonic agents.  However sonic analysis is relatively new in machine learning, and unlike the further developed analysis of human-made sounds, where machines are able to understand, to an extent, the verbal and emotional content of what is being said, analysis of non-human-made sounds isn’t there yet.  An issue which makes analyzing sound with machine learning difficult, as Stowell describes, is that unlike visual objects, sonic objects are not contained as continuous spatially entities (Stowell 2020). When presented on a spectrogram, they spread across one other, due to fundamental frequencies existing above the main frequency of any occurring sound.  The language and infrastructure around sonic analysis still need attending to. This is where, when using a framework like ANT, which, as opposed to a ‘descriptive vocabulary’, rather ‘opens, against all a priori reductions, the possibility of describing irreductions’ (Latour 1996, p. 375), we must take care to navigate the ‘untamed’ and abstract qualities of sound Sala refers to which give it its power in cross-species storytelling, whilst allowing specific languages to grow from it, be they bird languages, machine-learning languages etc.

Stowell describes how answering the questions, ‘are there any birds here?’ and ‘which birds are present?’ are equally difficult with machine-learning, despite the former seeming easier (Stowell, 2020).  From this we can grasp the leap between machine and human logic, but also the exciting potential of asking seemingly difficult questions like, ‘what are the birds saying?’  For now the app is useful in engaging people in environmental monitoring, and for conservation, as audio can be archived to detect changes in populations and migration patterns, particularly where birds can’t be seen but are easily heard.  The project also aims to help biologists understand the evolution of bird language, and identify social cues of various groups/individuals.  This is part of a wider area of research into bird (and animal) communication (such as research into Japanese Great tits’ use of compositional syntax to produce various meanings from individual calls/notes (Suzuki 2016, p. 1-7)).  Machine learning and cybernetics are part of a history of attempts at cross-species communication, from John Lilly’s experiments with dolphins to computer games developed for pigs.  However as feral thoughts teach us, communication may already be occurring on the periphery of technology.  Hanna Wirman’s communicative experiments with orangutans through gaming consols showed that the multi-species language of ‘play’ manifested itself not through the consoles, but through the orangutans grabbing ethernet cables, cage doors and human hands.  Although many times stronger than humans, the orangutans respected ‘the physical capabilities of the other.’ Thus a ‘meta-language’ formed between orangutans and humans, ‘establishing the other in relation to oneself’ (Wirman 2013, p. 2-3).

Intervention

In the rhizomatic spirit of ANT, I wanted to explore non-humans as environmental media, though not as pollution detectors or to answer other human enquiries, but simply to listen to non-human environmental experiences, ‘outside our purposes’ as Tsing says.  Through sound’s reflexive nature, how might non-human purposes then relay back to us, as response-able environmental citizens?  With the help of machine-kin, I explore a situated story of environmental change through the vocalizations of birds, whilst also speculating on future vocalizations and environments. Realizing sound’s ability to shift ontologies and pop/connect ontological bubbles as Pettman testifies, I think an intervention here is important, in:

1)          Continuing the sensing of environmental change through organisms as environmental media, and its trajectory towards concretizing notions of shared environmental citizenship with non-human kin.
2)         Showing how through creative listening and imagining of future environments, the felt and lived non-human experiences of environmental change can be rendered visible (and audible), and our own actions (response-ability) with respect to these experiences can be internalized.

Fig 1. Feral Robotic Bird

Adapting Natalie Jeremijenko’s pollution detecting Feral Robotic Dogs, I named my intervention Feral Robotic Birds.  It’s a speculative artwork that expands notions of climate modeling and imagines future birdsong, through listening to and registering current environmental soundscapes.  Part of the inspiration for the project was a radio podcast on starlings living in a barn, who were reported to be mimicking an engine.  The engine, long disused, must have been mimicked by the starlings’ ancestors, who then passed this strange song through the generations. I began to think about the starlings as both sonic reflections of their environments, but also artistic agents, inspired by their surroundings and expressing a cultural change in the environmental landscape, through song. I say cultural change referring to the many ways birds’ lives and relations to one another change with environmental change.  For instance the fact that birds now sing louder to match road traffic and aircraft noise, that less birds are being born due to partners not hearing one another’s songs, or are ‘re-housed’ and ‘dispossessed’ by building development projects, can all be understood as cultural changes, created by environmental changes we as actors constitute (even if such changes are not figured as ‘environmental’ changes in mainstream climate modeling).  Such interactions between birdsong and mechanical sound have been explored by Alexandra Ginsberg, using generative adversarial networks to create machinic birdsongs with a critical angle to the effects of human-made sounds on birds.

The starlings and the engine shows how sound can map a history of relations.  Just as geological sedimentation leaves historical traces, so too will echoes of us and the machines we engender exist in future birdsongs.  Similarly, echoes of the past exist in birdsong now.  Analysis of ice core’s composition yields climatic information dating back 800’000 years.  What could the history of sonic relationships existent in multi-species voices tell us about the past, if only we had the means to dissect and understand them?  By cataloguing the histories of their sonic environments and gradually incorporating them into each new vocalization, Feral Robotic Birds encourage us to consider sound as historical record, whilst simultaneously being an artistic speculation of future birdsongs and environments.  Feral Robotic Birds are also preservationists.  Like Alan Lomax’s field recordings of American folk songs, which led to UNESCO’s ‘Appeal for Cultural Equity’, defending every culture’s right to safeguard, express, and develop its artistic heritage, Feral Robotic Birds are archives of birdsongs.  When these recognizable songs begin to change, we are encouraged to consider the songs as a developing heritage of cultural artifacts, that deserve preservation.  

The songs are not a unified explanation.  Sound doesn’t tend towards linear causes, but trickles sideways, reflecting, re-appearing, dying.  As Francis Alys describes, sound is ‘a certain way of delaying the narration or postponing the conclusion…a progression always, but a kind of going back or forth’.  He says how the sonic nature of dynamic cities enforces adaption, where once the urbanist had to ‘plan ahead the expansion of the city’ it is now ‘to react to given situations of spontaneous urban growth…there is an absence of any master plan’ (Toop 2019, p. 174). This mirrors the turn of 20thC composers and musicians de-centering themselves from controlling masters of sound, to listeners and collaborators.  Science that works with organisms as environmental media could take this on board.  Sound talks back at you, and it’s this reflexive nature of sound, leading us as responders to it, as a means to a conversation with the environment, that I want to capture with my artwork.

My prototype, presented here, takes nightingale songs as its initial repertoire.  Nightingales are birds I’ve worked musically with before on a research project exploring human migrant stories and nightingale song, and I had field recordings of them, which I used for this prototype.  The technical aspects of are handled in Max/MSP.  Although a future version of this project could involve machine-learning to analyze archived recordings of birdsongs changing over time, and to allow the robot to adapt to its sonic environment, I took a simpler approach for the purposes of this project.  The patch contains 5 initial nightingale calls, triggered semi-randomly.  The character of the surrounding environment is recorded and incorporated into the birdsong through convolution (explained in the paragraph below), via a gradual process.  The longer you leave a Feral Robotic Bird in its environment, the more it begins to adapt its call in response to the sonic history of that environment.

Convolution multiplies the frequency content of two signals over time, producing an impression where one signal ‘sits inside’ another.  An issue that occurred was that after a few hours, Feral Robotic Birds became very noisy, and began to simply ‘sound like convolution’.  I filtered out background sounds below certain amplitudes and without distinct sonic profiles, and added a high-pass filter, as the convolution was exaggerating lower frequencies.  I also added an automatic volume controller that compensates in response to the loudness of the bird’s call and its environment.

Fig 2. The main patch window

Fig 3. Subpatch of 1 birdcall.

Fig 4. Convolution subpatches. 
Left: True convolution - correct multiplication of complex numbers, directly multiplying the cartesian form, then multiplying amplitudes and adding phases.  This sub-patch is lifted from a tutorial patch provided by Max/MSP.
Right: This subpatch combines the frequency spectrum of incoming audio with the amplitude profile of the bird song.  This was more of an artistic decision that maintained a ‘bird-like’ character to the sound, whilst incorporating the spectra of the surrounding sound.  This sub-patch is also lifted from a tutorial patch provided by Max/MSP.

Fig 5. Automatic volume controller subpatch.

Below is a video of Feral Robotic Birds in action.  Part 1 shows the bird's inital calls.  Part 2 shows the bird after listening to its environment for about an hour.  Part 3 shows an a modification I made to the calls, by slowing them down.  Nightingales' auditory recall is faster than ours, meaning they perceive each other's songs slower than we do.  By slowing them down, we hear them closer to how a nightingale would hear them, and I think this demonstrates how much information is actually in the calls, they sound very much like a language.  Part 4 shows an experiment where the Feral Robotic Bird has been listening to Thelonius Monk. The other day I heard a nightingale in a field where there used to be a village festival called May Fair.  Male nightingales arrive in the UK in May, and begin singing competitively with one another.  They may be missing (or not) the competition with the bands, maybe there’re elements of the bands imprinted in the birdsongs now.

Video 1. Feral Robotic Birds, Youtube.

In the vein of ‘public amateurism‘ (Costa 2008, p. 173) of artists doing science, my bird is a DIY stitching-together of household objects. The head is a small painted kaleidoscope, echoing a camera lens, and giving a machine-like, even surveillance-like appearance.  The feet are cut from tags of rice cake packets, attached to a picture-frame hook.  A Bluetooth speaker, which is connected to a laptop containing the Max/MSP patch, functions as both the microphone (or ears of the bird) and speaker (mouth).  A string attached round a hook allowd the bird to be hung from branches or other objects.  It can also stand on surfaces, or ‘sit’ in the lid of some shoe spray!

Fig 6. Feral Robotic Bird Seated.

Historically nightingales have represented foreboding and change, and their songs metamorphose in Feral Robotic Birds just as Procne metamorphoses into a nightingale in Tereus, forever mourning in the forest.  As in Keats’ ‘Ode to a Nightingale’ both the ethereal ‘already there’ of nature can be heard in Feral Robotic Birds, but so too can its fleetingness, the pace of change and fragility of existence.  Today nightingales are known to be migrants, each year they fly from Africa to Europe, and are met at the borders with guns and mist nets for hunting, and in the UK their habitats (low scrubland) are destroyed by military training grounds and housing development projects (Birkhead 2018).  It is through the ‘untamed’ abstract qualities of sound Sala referred to, the intonations, the ‘infralanguage’, the music, that we come to understand nightingales as part of our cultural landscape, and have been able to hear our existence reflected in them.  Nightingales sing hundreds of highly complex, socially learned and individually adapted songs.  Through ANT we are not required to partition those songs as being the individual’s alone, but can treat them as the net-work of multiple actors, thus becoming sonic reflections that tell stories of us, our multi-species kin and the environments we share.  Importantly, presenting their songs through the ‘robot’, unhinges ideas of voice as being the individual’s alone.  Animal-machine-environment become interconnected actors; Feral Robotic Birds is a kind of performative ‘becoming-with’ one another.  An engagement with birds and their environments in this way provides insights that the hides of science do not provide, and informs a ‘deep and critical knowledge of the ecosystem…achieved through not only the scientific mode of thinking but also other forms of knowledge and perception, such as the humanities and the arts’ (Brown & Timmerman 2015, p. 1).

There are of course limits to this project.  Although it helps us think about frictions between ourselves, birds, songs and environmental change, it is not able to inform normative guides to action like the 50 variables informing the IPCC do, and does not suggesting we stop measuring such variables.  But through reframing the story of environmental change through sound, Feral Robotic Birds gets us to think about an extended sense of self and environmental citizenry, that stretches across species boundaries and machines, and frames environment change as occurring internally as well as externally.  Furthermore, Feral Robotic Birds are sentinel species-machines that tell us something about themselves, and whose direct concern is not human wellbeing or action.  They get us to think about the collateral of our behavior, and the technologies we create on our non-human kin.  They help us see that the natural world is not an ‘out there’ to be measured, but something we are always already part of, reminding us of our response-ability.

Feral Endnotes

An unintended outcome from Feral Robotic Birds occurred when I took my prototype on a nearby walk where some of the first nightingales have arrived to nest near an estuary in Essex.  The male birds, who arrive first, sing competitively to establish territories before performing for the later-arriving females.  The Feral Robotic Bird seemed to encourage the nightingales to sing, and is a form of cross-species communication, which was one of the things I thought I might initially research.  As ‘feral thoughts’ teach us to notice unintended potentialities, a cross-species communication of sorts occurred through modeling future bird song.  With the development of apps like Warblr and their use for conservation, I wonder if Feral Robotic Bird’s ability to ‘bring out’ the nightingales could be developed, and potentially help in counting numbers.

The communicative element of Feral Robotic Birds also makes me think of potential actor networks.  Like the practice described by Shiho Satsuka, of disturbing the landscape, in the hope Matsutake mushrooms might potentially grow, Feral Robotic Birds digs a little into the dirt of nightingale-human-machine-environment inter-action and communication.  At a lecture titled ‘Ask the Birds’ in 2018, bird scientist Tim Birkhead talked of a bullfinch he’d met who had mastered folk songs better than human whistlers, and greeted its owner with the songs more affectionately than a dog.  So sonically different from their usual ‘squark’, this suggests extensive brain activity, and although that activity is hijacked here from its usual uses (in the typical messiness of multispecies worlds), it suggests unknown and rhizomatic potentials in ways sound can manifest itself through non-humans and humans alike, and in the ways animals could think, act and express themselves, in worlds where we interacted positively and fully with them.

It would have been useful if during the research of this project and in reflecting on my artwork, I could have met with the staff and ranger at Fingringhoe Wick Nature Reserve, renowned for nightingales, and spent time with the nightingales and with Feral Robotic Birds there, but I look forward to doing this in the future.  Also, to really get a sense of my piece as an intervention, I need to try it out in lots of different locations, such as busy roads, construction sites, parks, shopping centres etc, and leave it there for quite some time, whilst documenting the sonic changes, and see what people think about it, though this is of course not possible at the moment.  During this epidemic, many have remarked on the prevalence of birdsong in cities, where normally there would be traffic and other more traditional urban sounds. For now, Corona Virus, an actor and life-form itself, has made us consider our sense of place amongst technological noise and birds.

Video 2: Feral Robotic Bird singing with nightingales.

Annotated Bibliography

Gabrys, Jennifer. Program Earth: Environmental Sensing technology and the making of a computational planet, Minneapolis: University of Minnesota Press, 2013.

Gabrys’ research formed the basis of my thinking around how technological measuring instruments and the measuring of solitary variables render certain processes and practices relevant, and furthermore what politics and ontologies about oneself with respect to environments concretize. Her ideas about environmental citizenship and non-human organisms as environmental media were vital in how I situated Feral Robotic Birds. Also useful were her thoughts on creaturing data, community sensing and artistic modes of measurement, which helped me situate my piece amongst DIY artists doing science, such as Beatrice de Costa.  This disruptive element also situates my piece within critical technical practices, and if I am to continue this project, it would be nice to explore how Feral Robotic Birds could be more forceful and disobedient as a work.  Whitehead’s Modes of Thought, which Gabrys mentions quite a bit, would have been nice to look at too, but I could not find an online version from our library during this pandemic.

Latour, Bruno. ‘On actor-network theory: A few clarifications’ Soziale Welt 47. Jahrg., H. 4 (1996), pp. 369-381.

This (alongside Casey D. Allen’s ‘On Actor-Network Theory and landscape’) helped me visualise the chaotic nature of landscapes, and to deconstruct boundaries such as biomes, nature/culture, and even the self, and start to think about these as assemblages of myriad agents (both macroscopic and microscopic) across networks of time and space.  It made me comfortable exploring such a situated aspect of environmental change, because I saw it as one of many heterogeneous but interconnected stories, that needed telling, and wasn’t captured by the over-arching explanations of systems theory/observational science.

Haraway, Donna. Staying with the Trouble: Making Kin in the Cthulucene, Durham: Duke University Press, 2016.

The ideas of ‘becoming-with’ as a collective and a decentering of individualism and anthropocentrism chimed with me.  As did a sense of stewardship having to do with responding and listening to other beings, delving into non-human modes of being, and not leading the way as such.  A strong believer in animal rights, I often find it difficult when artists or scientists use animals in their work, because however well-intentioned they are, you cannot get consent from the animal, and the utilization of an animal for a means reinforces ideas of animals as inferior and the exploitation of animals worldwide due to this idea is unimaginable.  However Haraway helped me see the situated messiness we are in, and to realize that in order to begin building ‘serious’ multispecies worlds, shortcuts and assumptions are sometimes inevitable.  It was important to hear this from someone who clearly genuinely cares about non-human life.  However this messiness is not an excuse to be idle toward non-human rights, and this is something we need to be vigilant against.

Pettman, Dominic. Sonic Intimacy: Voice, Species Technics (or, How to Listen to the World), California: Stanford University Press, 2017.

Exploring voice, this book looks at the creaturely voice, the technological voice, the gendered voice and the world voice, or vox mundi. This helped me think about sound as a medium that crossed humans, non-human species, machines and environments.  The book questions ownership of voice, talking about the reflective nature of shared voices across environments.  This linked with ideas in ANT, and Anna Tsing’s feral thoughts, in that sound is also an assemblage of actors, histories and relations.  However his ideas on ‘reterritorialization’ through voice also lends validity to animals as individual and expressive beings, and this was important in informing my thinking of birds as artistic storytellers, and an important consideration I think in how we might approach multispecies collaboration.

Note: Despret's 'What Would Animals Say if We Asked the Right Questions' was something I think could have been very useful for this project but was unable to get a copy due to the library being closed.  Also missed was Rob Harling's film at Cable Depot 'Exile in the Kingdom is Beautiful'.  This would have been interesting to see as housing dispossession is looked at and told through ash trees.  Dispossession is something nightingales face, and it would have been interesting to consider this as something occurring across birds, trees and humans. Perhaps this is something I can do in the future though.

Bibliography

Allen, D. Casey. ‘On Actor-Network Theory and landscape’ Area
Vol. 43, No. 3 (September 2011), pp. 274-280.

“BBC Earth” Facebook (Via BBC Radio 3). https://www.facebook.com/bbcearth/videos/1630414800313735/?hc_ref=ARR9xP9BBHgTEpsWbJdKxxwxfFGDBVNHyiLuiX0q0kbvoaVxPrUsHD68armTtR4so6o. 2017; accessed 4 May 2020.

Birkhead, Tim. ‘Ask the Birds’, Lecture at. Whitechapel Gallery, 24 Feb 2018.  

Brown, Peter, G. and Peter Timmerman. Ecological Economics for the Anthropocene: An Emerging Paradigm. New York: Columbia University Press, 2015.

Costa, Beatriz. da. & Kavita Philip. Tactical Biopolitics: Art, Activism, and Technoscience. Massachusetts: The MIT Press, 2008.

Deleuze, Gilles. & Felix Guattari. A Thousand Plateaus. Minneapolis: University of Minnesota Press, 1987.

Gabrys, Jennifer. Program Earth: Environmental Sensing technology and the making of a computational planet, Minneapolis: University of Minnesota Press, 2013.

Gramaglia, Christelle. ‘Sentinel Organisms: “They Look Out for the Environment!” limm, https://limn.it/articles/sentinel-organisms-they-look-out-for-the-environment/ Accessed 4 May 2020.

Haraway, Donna. Staying with the Trouble: Making Kin in the Cthulucene, Durham: Duke University Press, 2016.

Keats, John. John Keats: Selected Poetry ed. Elizabeth Cook. Oxford: Oxford University Press, 1994.

Krause, Bernie and J. Skinner. ‘THE GREAT ANIMAL ORCHESTRA: A Performance and Dialogue in Soundscape and Poetry’, Harvard University, 2012.

Latour, Bruno. ‘On actor-network theory: A few clarifications’ Soziale Welt 47. Jahrg., H. 4 (1996), pp. 369-381.

Pettman, Dominic. Sonic Intimacy: Voice, Species Technics (or, How to Listen to the World), California: Stanford University Press, 2017.

Roads, C. ‘Convolution’ The Computer Music Tutorial, MIT Press, 1996

Stowell, Dan.  ‘Who's singing? Automatic bird sound recognition with machine learning’ https://www.youtube.com/watch?v=pzmdOETnhI0. PyData, 2018; accessed 28 April 2020.

Suzuki, Toshitaka N. ‘Experimental evidence for compositional syntax in birdcalls’. Nature Communications. (2016), pp. 1-7.

Toop, David. Inflamed Invisible: Collected Writings on Art and Sound 1976-2018. London: Goldsmiths Press, 2019.

Truax, Barry ‘Convolution Techniques’ https://www.sfu.ca/~truax/Convolution%20Techniques.pdf. Accessed 28April 2020.

Tsing, Anna. “Feral Technologies: Making and Unmaking Multispecies Dumps Interview with Elaine Gan, Bettina Stoet” https://www.anthropocene-curriculum.org/project/campus-2016/feral-technologies. Anthropocene HKW, 2016; accessed 28 April 2020.

Tsing, Anna. The Mushroom at the End of the World. New Jersey: Princetn University Press, 2015.

Winderen, Jana. ‘Isolation/Measurement’, Energy Field, 2010.

Winderen, Jana. ‘The Listener’ 2016.

Wirman, Hanna. ‘Orangutan Play on and Beyond a Touchscreen’ n Cleland, K., Fisher, L. & Harley, R. (Eds.) Proceedings of the 19th International Symposium of Electronic Art, ISEA2013, Sydney. (2013) pp. 2-3.

Wirman, Hanna.  ‘The playing Other and what we cannot help learning from the study of animal play’ Transactions of the Digital Games Research Association. Accessed 28 April 2020.  

World Meteorological Organization. ‘Essential Climate Variables’
https://public.wmo.int/en/programmes/global-climate-observing-system/essential-climate-variables. WMO, 2020. Accessed 28 April 2020.

Yi, Anicka. ‘Anicka Yi on non-human ecologies and embodied machines’, https://podcasts.apple.com/es/podcast/e-flux-podcast/id1332021431
Podcast on e-flux, 2020. Accessed 2 May 2020.