escaping auguraculum - on emergent divinations
by Sonia Bernac
The universe was so created that certain results would be preceded by certain signs, which are given sometimes by entrails and by birds, sometimes by lightnings, by portents, and by stars, sometimes by dreams, and sometimes by utterances of persons in a frenzy. And these signs do not often deceive the persons who observe them properly.
Cicero, On Divination
A wooden ear, enormous, plugged with cotton and the tediousness of Cicero. What a great stylist, everyone says. Today no one writes such long sentences. And what erudition. He even knows how to read inscriptions on stone. Only never will he guess that the marble veins in the Baths of Diocletian are the blood vessels of slaves which have burst in the quarries.
Zbigniew Herbert, Classic
ex avibus [from birds]
The ancient practice of augury - divinations performed by means of birds - could be easily romanticised and hence, misunderstood as a moment of resonance between the collectively conceptualised ‘ancients’ and nature, cosmos, future. “Taking the auspices” was based on the collection and interpretation of various bird-related incidents and patterns: aerial murmurations; mating songs; the sudden appearance of members of particular species; their unusual behaviour or the order of their actions. Undoubtedly, there is something seductive about the image of a collective movement of eyeballs, human or otherwise, attuned, temporarily suspended from their own daily trajectories, tracing the birds’ erratic undulations.
Thomas Nail, in his reading of ancient poet Lucretius' kinetic materialism, notices that the swerve of swallows chasing insects in the air is an expression of the pedetic motion of matter. That is, Nail rightly claims that motion so often perceived as random is, in fact, based on localised interactions that can/must include causality. However, Nail’s interpretation of causality does not presuppose some overarching rational logic of motion: instead, causality is simply matter responding to/sensing itself:
Wisdom. Swallows are also known for being prophetic birds. […] Lucretius does not invoke augury because he believes in prophecy, but because he believes bird movements tell us something about nature, the weather, and so on (which they do).[1]
This image also tethers Nail’s argument to a particular kind of nostalgia — since it not only points to alternative ways of knowing that do not presuppose a centralised, anthropocentric Reason - God thrusting meaning into things, or some other universal authority of (Western) Idea - but also subtly suggests (a return to) some pre-rationalist or otherwise spiritual unity of Man with Nature.
In fact, ancient augury was a highly exclusive practice performed only by the chosen, usually in a ritualised space of the Auguraculum, a roofless temple, from which only a section of sky was visible. The positions of birds were marked with stones on the floor at very particular times of day. Until 300 BCE only patricians were permitted to take auspices and even the augurs of noble birth were not allowed to interpret their omens. The role of augurs was only to perform a rite of accurate prophetic measurement: observation. The honour of interpretation was then reserved for the highest elected political office - the consuls, who would read the obscure clues for use in political navigation.[2] In that sense, augury was an important part of a highly hierarchical ancient theatrics of Politics, which preceded all important public events, elections, legislative decisions, and the declaration of war. Indeed, the word ‘inauguration’ marking the introductory phase of an event, comes from this ancient divinatory practice, which is not to claim that any etymological or structuralist analysis can ever reveal too much about the subject of the philosophical investigation.
ex tripudiis [from dancing]
Cicero, an augur himself, in his dialogue, On Divinations, mentions other types of augury practices that did not always require for birds to fly or to remain alive. For example, a change in appetite or behaviour of sacred chickens, a practice somewhat mysteriously named “the dance”, could easily lead to the postponement of a battle.[3] A more violent example, such as the respected tradition of haruspicy, required a careful examination of a sacrificed animal’s entrails for the minuscule patterns or gut irregularities which enabled a prediction of the future, at least for those beings who survived the ritual.
Although the ancient’s reliance on haruspicy for accurate predictions might seem scientifically laughable today, it could be claimed that the custom is not so different from Western practices of the 19th century, or even, those contemporary systems of invention/discovery. Knowing through taking apart: the processes of naming, identifying common features and declaring them universally applicable has a long historical trajectory.
However, the purpose of this brief contextualisation of bird watching is not to provide a historical trajectory of human interest in murmurations. It is rather to distinguish between those who were authorised to look (up) and formulate careful predictions based on omens (elected men of noble birth), those who remained voyeuristic usurpers offending the Gods with their unworthy peering into the future (all other people, including slaves and women), and those who served as a - singular or collective - body/flesh/data for prophetic scrutiny (animals). However, it is not particularly insightful, especially after Foucault, to claim that knowledge is not separable from power structures, or more accurately, the (circulation of) matter of its time.[4] Ways of seeing, sensing and predicting are always wrapped in a codified economy of future telling-plotting-modelling-fictioning-making. In fact, this “wrapping” does not go far enough to convey the depth of this inseparability, since it implies a form of contextualisation (a stage), and therefore separation between the matrix (spacetime) of power relations and those impressionable subjects, objects and practices stuck in the net of historical dependencies and inequalities.
ex avibus et apibus [from birds and bees]
A visual requisite, frequent example, and one would dare to say: a symbol of the current turn to complexity in sciences are the birds and the bees: the murmuration and the beehive. Approaches to systemic complexity vary depending on discipline however there remains certain recurring features which characterise it as a phenomenon or principle of systemic ordering.[5] Those associated qualities include: nonlinearity, self-organisation (spontaneous order), the fact that the properties of the delineated collection are not a sum of its constituents – emergence, so called “adaptability”, coherence based on feedback loops, non-deterministic behaviour and some relation to what is inaccurately defined as ‘randomness’ or ‘unpredictability’. It is crucial to mention here that those features are not logically parallel but rather interdependent and have their own consequences for systemic prediction making – future telling.
Complex systems, contrary to their name, do not have to be very complicated to self-organise in line with the principles of complexity. Rather, complexity refers to the impossibility of universal scalability of local structural and functional forms of ordering. In that sense, complexity could be seen as a counterpoise to the reductionist approach that is based on the assumption that the grammar of a larger system can be expressed in terms of its microscopic components. The opposition between the reductionism and emergence is not, however, a conflict of scale (small for reductionist versus big/broad for emergence) but rather a disbelief (or distrust) in the possibility of absolute ontological scalability.
That point is thoroughly and subtly developed by Laughlin in his book A Different Universe: Reinventing Physics from the Bottom Down. Laughlin writes of the investigation of physical fundamentality that historically in the sciences has always been associated with some form of universal law, discovered through investigation performed in microscopic scales. For Laughlin, all laws have “collective origins” by which he means that fundamentality is an emergent property.
In other words, the distinction between fundamental laws and the laws descending from them is a myth, as is the idea of mastery of the universe through mathematics alone. Physical law cannot generally be anticipated by pure thought, but must be discovered experimentally, because control of nature is achieved only when nature allows this through.
ex finitimis [from neighbours]
In a simple textbook differentiation, dynamic systems are usually characterised as either conservative or dissipative. In conservative systems the total energy is preserved, and their processes are considered reversible. Critically, swarms/murmurations (which I will temporarily use interchangeably) are classified as open and dissipative non-equilibrium systems. These dissipative systems operate sometimes far from thermodynamic equilibrium, and exchange energy and matter with the environment. These systems are associated with processes that are irreversible.[6]
There is no underlying theory, formula or model that may predict with certainty the motions of multimodal dynamic groups such as birds, bees and swallows. A variety of methods have been applied to approximate and simulate behaviour of such animal groups, or more broadly “active matter systems”,[7] but although some of them are useful for formulating accurate predictions or convincing simulations, it does not mean that they are necessarily “correct”.[8] For example, even though swarms are non-equilibrium systems, the formalism of phase transition taken from the theory of equilibrium systems often helps to model the collective motion of swarms very accurately. However, the theory of equilibrium systems cannot explain many other important behaviours, for example, the self-organisation of a critical state that does not seem to depend on the change of externally controlled parameters.
The emergence of large-scale order in swarms is linked to local interactions between the agents but should not be considered as emerging from it. In bird murmurations - sticking with the most common avian example - a singular agent makes on average six to eight connections (with its seven nearest neighbours). Those connections are adaptive and dynamic, implying constant shifts, changes and fluctuations. The links between group members are not metric but topological, which means that the selection of “neighbours" is not dictated solely by the distance but through the topological convenience and arbitrary rules of attraction. That non-metrical convenience is conditioned by the position of the agent in the overarching shape and folding of the topological transformation of the murmuration. Neighbours are also selected based on similarities in speed, size and direction, as well as scientifically arbitrary parameters, such as long-term affiliations that one could dare to call “friendships”.[9]
All those fascinating dependencies were discovered through the use of stereometry, dynamic 3D modelling and behavioural experiments, during which some of the flocking birds were caught in a net-trap suspended in the air and removed from the murmuration. Researchers then scrutinized how the remaining group dealt with the loss and what it meant for the local interactions. In fact, the sudden disappearance of a frequently contacted group member led to increased socialising of the abandoned agents.[10] Hence we can surmise that a turn to biomimicry in the sciences is not necessarily tantamount to compassion.
ex chao [from chaos]
Plato’s Republic makes frequent use of the rhetoric of divination to suggest ways of accessing transcendental knowledge of the Good. For Plato “they rectified the vile part in us by establishing divination there, so that it might in some degree lay hold of the truth” [11] Divining for Plato is a way of intuiting without constructing an analytical proof, since what is being summoned is the already existing Idea (of Forms). In Plato’s language of systems and predictions, the future trajectory is neither known or knowable, yet it already exists in some metaphysical and inaccessible form.[12]
A very different divinatory logic is presented by Lucretius in On the Nature of Things, where divinations do not serve to summon some metaphysical readymade, but are themselves a part of the ongoing and indeterminate motion of matter driven by microscopic (inter)actions, micro-causalities, resonances, fluctuating attractions of patterning, etc.
When formalising, conceptualising and modelling swarms, ample focus is placed on mapping those instabilities that might lead to chaotic behaviour. Contrary to the common-sense meaning, this “chaos” does not equate to mess or absolute disorder but rather acute sensitivity to initial conditions, the so-called “butterfly effect”. More interestingly, within the observation of apparent order formations in murmurations, one could mistakenly conclude that formulation of the patterns of order is always immediately perceptible. However, there are many forms of order, such as the periodicity of movement that only reveals self-similarity of apparent incidents and irregularities over long periods of time. Alignment, change of direction and the re-coupling of agents that can be visually observed are only a small part of the murmured pulsations emerging through the movement. Indeed, complexity scientist Roland Bouffanais mentions a fascinating case of turbulent flows, which might ocularly appear as lacking order. However, the analysis of these flows reveals the emergence of large-scale coherent structures associated with fluctuations of the velocity field.[13]
Consequently, the emergence of a pattern is dynamic, but not just in the sense that such a pattern might transform in time. In fact, every entity perpetually emerges through different forms of patterning and feedback (cohering). As such, the same material arrangement/property might act as a similarity and difference, depending on the systemic cohering order operating in various fluctuations, which are of the same time and yet of different spatiotemporal periodicities. Moreover, the same “part” might partake in different scales and ranges of order. The key to grasping this strange plurality is hence not through ocular means but by way of sound and rhythm.
Furthermore, collective adaptive systems are often characterised as existing “at the edge of chaos”.[14] This dramatic (and rather annoying) statement implies that these systems are close to criticality. In statistical physics, criticality or a critical point (on a phase diagram, for example) is a state separating an ordered phase from a disordered phase. This is a very important point, since if the system was too “ordered” it would be very difficult for its avian agents to break down their ritualised choreography quickly enough to react to the fast-changing conditions (of the environment). If the system was too “disordered", the change in insufficiently correlated agents would be too slow and would fail to utilise the collective matter (and force) of the group efficiently. However, at a critical point, thanks to long-range correlations, the swarms can react quickly, efficiently and in the emergent collective interest. For example, by avoiding the danger of a predator through a confusing waving of the collective body, or by reacting to the increased noise of stormy weather with tighter distances between the agents.
Fluctuations and oscillations of the systems should not be seen as “glitches” or irregularities of the overarching pattern of order. Crucially, they are not dirt or excess but important factors in the system’s ongoing coherence. They not only push the system closer to criticality, but act as a cohering force in response to the noise and perturbations of the environment.
ex motu [from motion]
However, the concept of the environment that clumsily keeps returning to the argument is not as innocent as it seems, even with the assumption of fuzzy, blurry edges of anything. The distributed multimodal systems, and matter in general are often wrongly thought of as some sliced up, dispersed oneness consisting of many moving parts, or slightly more subtly, but also incorrectly, as a complementary dance of continuousness and discreteness. The swallows of Lucretius might be helpful in understanding this problem from a slightly different angle than the already criticised notion of the contextual “stage”.
Swallows are known above all for their chaotic and unpredictable flight paths as they hunt insects in mid-air. (...) Since humans cannot see these insects, it looks as though the swallows are moving ‘randomly’, when in fact their erratic-looking movements are the result of a highly relational and responsive entanglement with their prey.[15]
What this passage in Nail’s book serves to illustrate is the fact the system of birds observed from the ground is linked to another (less-perceptible) murmuration of midges which might in turn be more closely interacting with another complex system. That thickening topology of links should not be imagined as a growing network of (trophic) dependencies or a tangle of clusters of same-species/ same-material /same-tribe aggregations symbiotically influencing each other. Those local and non-local resonances have little to do with the apparent edges of bodies or ocularly perceived similarities. Hence, the patterning and feedback (loops) do not happen just within the boundaries of one thing (either distributed or seemingly centralised) but act as intersectional processes of coherence - resonance operating in different ranges/ dimensions/ phases of matter.
Johnny Golding talks about those thickening topologies as a form of camouflage, a return of similarity and, more importantly, difference that sets up the fabric of meaning. In earlier work by Golding, it is observed that this camouflage can to some extent be expressed by the Mandelbrot set equation Z ⇌ Z2 + C, where the recursive fractal patterns enable expansion – a thickening of “any given ‘Z’”.[16] This is, of course, not just a mathematical proposition as this “thickening” is an ontological move that speaks directly to circulation, a particular kind of dynamic folding and weaving.[17] Those should not be considered as metaphors or analogies but powerful images that enable thinking through different aspects of mattering.
ex signis [from signs]
In the light of all certain features mentioned earlier and a few comments on causality, one could mistakenly reach a conclusion that the motion of matter is at heart driven by utilitarian deterministic actions and therefore can be accurately predicted/divined (with only some minor embellishment of teleologically driven fluctuations).
The indeterminate nature of motion does not mean that it is incalculable, hard to predict or model, or that the most efficient tools for prognostication have not yet been invented. Undoubtedly, there is plenty to discover and invent about complex adaptive systems and their artificial doppelgängers. The uncertainty in question does not mean that the material world is mysteriously unknowable or otherwise impenetrable, which would be repeating the mistake of Object-Oriented Ontology (by injecting matter with some essentialist secret). Undecidability means that, in principle, the future of a system, to stick with the rhetoric of divinations, is not deducible or reducible from its initial conditions. Gödel’s theorem makes it formally apparent that any totalising consistency is not expressible through the original axioms of the system.[18] Even though every system is fundamentally undecidable, complex adaptive aggregations with their leaping, re-coupling, phase-shifting and reordering without any help from external parameters (or God) make the non-deterministic logics of motion pleasantly palpable.
It seems important to mention here that contemporary interest in complexity is not driven solely by philosophical curiosity but by practical interest in very specific features (of multimodal adaptive systems in particular) that are invaluable for design: adaptability, flexibility and robustness. However, the study of these organisational phenomena in systems does not necessarily imply the emergence of a new method or methodology. In other words, as long as complexity and emergence are still approached as subjects, making predictions (divining the not yet known) will continue to resonate with the already existing structures of knowledge/power, and with the patterns of exclusion and oppression. Herbert’s post-war poetry (cited at the start of this paper), so often critical of the partial view and idealisation of antiquity, features traces and signs that refer to slavery as a part of the economy of classical thought. For Herbert, the “marble veins in the Baths of Diocletian” are a form of “seeing” the patterns of classical violence, so often overlooked in contemporary references to ancient thinkers. Without a rigorous scrutiny towards scientific methodologies and their politics, our contemporary fascination with distributed complex systems will result in a repetition of the same mistakes.
To be more specific, Wendy Chun in Discriminating Data draws attention to ways in which the methods of network theory abstract complex systems of relationships. She divides this process into two stages: the first implies conceptualising bundles of things as agents and relationships which requires a division of phenomena into static “nodes” and “edges”. The second stage is more mathematical, by producing an abstraction of that data into a reproducible model used to design social interactions in the present/future.[19] This is also put rather elegantly in Gerard Nestler’s work on the Derivative Condition, where futures modelled on the biased past render any true change impossible.[20] To phrase this differently, it is a process of forcing matter into a controllable and malleable determinism or range of probabilities, in which undecidability is simulated with the parameters of randomness.
This opens up a much broader question in machine learning and forms of, so called, artificial intelligence. In many ways AI can be considered as the ultimate “prophecy machine”.[21]
It could seem that the Laplace demon, anticipating transcendence to qubits, but already with enough computing power, is already capable of uncannily accurate simulations.
An intellect which at a certain moment would know all forces that set nature in motion, and all positions of all items of which nature is composed, if this intellect were also vast enough to submit these data to analysis, it would embrace in a single formula the movements of the greatest bodies of the universe and those of the tiniest atom; for such an intellect nothing would be uncertain and the future just like the past would be present before its eyes.[22]
That simulation does not have to have the aesthetics of ancient prophecy. In many ways any generative AI tool trained on databases but capable of prompt-induced or spontaneous creation follows a prophetic model: scanning for recurring or significant elements, detecting patterns-models, generating something that feels like it belongs to that systemic whole, and trying to pass that through the adversarial network.[23]
The question is not if those generated entities are new enough, or truly emergent. As mentioned before, any system, even that thriving on the illusion of total determinism is always already to some extent undecidable. The question is rather, to what extent that undecidability, even understood narrowly as a possibility of a logical leap, is nurtured within the systemic logic, and not simply exploited in a form of artificially induced randomness or identified as a glitch.
The discovery of emergence in complex systems therefore requires the symbiotic invention of truly interdisciplinary emergent methodologies. Such methodologies, based on generative, experimental, explorative and reflexive praxis must embrace undecidability not as incalculable probability but as a necessary condition for knowing, making, and predicting. Hence, the future is not summoned or calculated based on initial conditions but collectively invented-fictioned through the local and non-local interactions between conscious and non-conscious agents.
The birds, whether dead or alive, have no rehearsed messages to pass on. The method of emergence is a practice of plural divinations that does not presuppose prior or hidden meaning, nor imply the presence of a pattern that must be traced/caught/retrieved and made subject to interpretation. In fact, divination does not seem to be possible without what Golding calls attunement, a particular kind of listening-sensing that betrays any understanding of sterile one-sidedness.[24] This move would only be possible through some kind of experimental poetics (art), where artistic methodology functions as a form of attuned, emergent divination.
––––––––––––––––––––
Footnotes
[1] Thomas Nail (2019), Lucretius II: An Ethics of Motion (Edinburgh University Press), 22.
[2] Maria Jaczynowska (1997), Religie świata rzymskiego (Warszawa: PWN). ISBN 83-0107405-1.
[3] Cicero (1923). On Old Age. On Friendship. On Divination, translated by W. A. Falconer. Loeb Classical Library/ Harvard University Press), 154.
[4] Michel Foucault (2020), Power`; The Essential Works of Michel Foucault 1954-1984, (New York: Penguin Classics).
[5] There is a degree of ambiguity and freedom in mapping of the systems’ complexity due to the interdisciplinary nature of the subject. However “complexity” has a precisely established meaning in various fields, being irreversibly connected to different definitions of “a system” in the respective disciplines.
[6] Cf. Roland Bouffanais (2016), Design and Control of Swarm Dynamics, 1st ed. 2016 edition (Singapore: Springer), 37.
[7] Active matter systems consist of units that consume energy and generate force, driving emergent dynamic properties on larger scales. Cf: Moumita Das, Christoph F. Schmidt and Michael Murrell (2020) “Introduction to Active Matter,” Soft Matter, 16, no. 31, 7185–90, https://doi.org/10.1039/D0SM90137G.
[8] Cf. Bruno Latour (2010), On the Modern Cult of the Factish Gods, (Durham, NC: Duke University Press, 2010).
[9] Roland Bouffanais does not mention friendships. Roland Bouffanais (2016 [2015]), Design and Control of Swarm Dynamics, (Spingapore: Springer).
[10] Josh A. Firth et al (2017)., "Wild Birds Respond to Flockmate Loss by Increasing Their Social Network Associations to Others’, Proceedings of the Royal Society B: Biological Sciences 284, no. 1854,:20170299, https://doi.org/10.1098/rspb.2017.0299.
[11] Plato (nd), The Republic, edited by Melissa Lane (2007),trans. H. D. P. Lee and Desmond Lee, (Penguin Classics).
[12] Peter Struck (2014), “Plato and Divination,”’ Archiv Für Religionsgeschichte 15, no. 1 (March 2014): 17–34, https://doi.org/10.1515/arege-2013-0003.
[13] Bouffanais, Design and Control of Swarm Dynamics.
[14] This term was first coined by Norman H. Packard in his Adaptation Toward the Edge of Chaos. Norman H Packard (1988), Adaptation Toward the Edge of Chaos, (University of Illinois at Urbana-Champaign, Center for Complex Systems Research).
[15] Nail, Lucretius II: An Ethics of Motion, p.22.
[16] Johnny Golding (2012), “Ana-Materialism and the Pineal Eye: Becoming Mouth-Breast (or Visual Arts After Descartes, Bataille, Butler, Deleuze and Synthia with an “S’),”, Philosophy of Photography 3, no. 1, 99–120, https://doi.org/10.1386/pop.3.1.99_1.
[17] Cf. Thomas Nail (2018), Lucretius I: An Ontology of Motion (Edinburgh: Edinburgh University Press).
[18] Kurt Gödel (2012), On Formally Undecidable Propositions of Principia Mathematica and Related Systems (Dover Publications).
[19] Wendy Hui Kyong Chun (2021), Discriminating Data: Correlation, Neighborhoods, and the New Politics of Recognition (Cambridge, Massachusetts: MIT Press).
[20] Gerald Nestler (2016), The Derivative Condition. A Present Inquiry into the History of Futures (Goldsmiths University of London Ph.D. submission in Research Architecture).
[21] I am indebted to a researcher and artist John Wild for this phrase in the context of generative AI.
[22] Pierre-Simon Laplace (2011), Philosophical Essay on Probabilities, Reprint of the original 1st ed. 1995 edition (New York: Springer).
[23] I am aware that the language used here might suggest forms of intentionality. The generative process mentioned here is referencing GANs: Generative Adversarial Networks, but there are, of course, many other forms of generative AI.
[24] Johnny Golding (2010), “Fractal Philosophy: Attunement as the Task of Ar,”, edited by Stephen Zepke and Simon O’Sullivan (Edinburgh University Press), 133–54, https://edinburghuniversitypress.com/book-deleuze-and-contemporary-art.html