The universe has a remarkably consistent elemental composition: about 75% hydrogen, 24% helium, and 1% heavier elements. Stars, for all their element-producing abilities, cannot have created these abundances. This points to another cosmic oven, in the universe’s hotter past.

The farther we look, the redder the light from galaxies appears. This fact points to a remarkable feature of our universe: it is not static. It is expanding.

Various alternatives to the big bang model have been proposed, both inside and outside the scientific literature. We review some examples, and show how they deal (or not) with the evidence of astronomy.

This book was inspired by the cold glare of a shark, which happened to meet my eyes in a train station bookstore. A full spread on the cover of a diving magazine displayed the prize-winning picture taken by the underwater photographer Doug Perrine. The image is of two copper sharks (Carcharhinus brachyurus) having their way with a hapless school of sardines. With sardines still stuck between their teeth, they are darting through the evasively maneuvering swarm, and the glance of one of the sharks during this feeding frenzy seems to fixate on the diver's camera. What Perrine managed to capture here so impressively is the famous sardine run – the annual migration of immense schools of sardines along the coast of South Africa. Their morphologies and dynamics number among the most fascinating phenomena of the animal kingdom that Alistair Fothergill and his team of BBC filmmakers had documented so vividly around the turn of the millennium.

Not long after this encounter, I coincidentally came across this image again: Perrine's photograph happened to adorn the cover of a small brochure that, in 2005, was used to advertise an upcoming consumer trend conference in Hamburg. Given the title ‘Schwarmintelligenz: Die Macht der smarten Mehrheit’ (‘Swarm Intelligence: The Power of the Smart Majority’), the symposium featured the keynote speaker Howard Rheingold, who had recently published his study of smart mobs, and thus shifted the focus away from sharks and schools of sardines toward the dynamics of highly concentrated network economies:

The trend conference in 2005 was thus right on trend. Swarm intelligence was on everyone's lips at the time and had just lately entered the discourses of the humanities and social sciences with the publication of Rheingold’s book.

In media-historical terms – as I discussed in my introduction – swarms have been fused into biological, computer-technical hybrids that can best be understood with the concept of zootechnologies. For, contrary to biotechnologies or biomedia, they are conceptualized less on the basis of bíos (the notion of ‘animated’ life) than on the basis of zoē, the unanimated life in the swarm – a sort of life that can be technically implemented. The knowledge gained from this hybridization has helped to establish a highly technical lifeworld that is increasingly confronted with models of complex contexts and systems. Whenever one is faced with disrupted or constantly changing conditions, or whenever solutions need to be found for unclearly defined sets of problems, methods can now be employed that are commonly known under the umbrella term ‘swarm intelligence.’ Eric Bonabeau, Marco Dorigo, and Guy Theraulaz thus make the following remarks toward the beginning of their standard work Swarm Intelligence: From Natural to Artificial Systems:

Models as Media

As demonstrated in the previous chapters, Charles Breder was right when, in an early article, he referred to fish schools as ‘notoriously difficult laboratory materials’ – a characterization that applies just as well, if not more so, to studying them in the open sea. Alongside the efforts being made to gain empirical data about the control logic and organization of schools by means of various observational and analytic systems, a second (and complementary) strategy explored models and mathematical formulas in order to describe the dynamics and functions of biological collectives. Models, after all, provide the opportunity for drawing connections between different scales of observation and between different influential variables. They offer access to levels that the technical methods of visual or acoustic analysis cannot access, and they combine theoretical considerations and empirical data. However, two further sets of problems have to be considered in this context which exceed the difficulties that I mentioned in the previous chapter – namely the question of viewpoint, the feasibility of convincing experiments in research aquariums, and the revealing of visual ‘invisibilities’ by acoustic methods. One issue concerns the ways in which the relationality of schools could be converted into models and how to measure the epistemic value of such models in conjunction with other forms of knowledge production in the field of biology.

Fishy Business: Media Technologies of Observation and Experimentation

Schools of fish have not been studied as complex systems for very long. It was not until the end of the 1920s and the beginning of the 1930s that researchers first engaged with questions concerning the possible parameters and observational media that would be necessary to gain any knowledge about them. In his 1931 article on schooling behavior, Guy Malcolm Spooner remarked: “The phenomenon of schooling has received surprisingly little attention either from fishery investigators or from those studying animal behaviour.” The Russian biologist Dimitri Radakov similarly commented: “The phenomenon of schooling has undoubtedly been known since ancient times, at any rate since our ancestors began to catch fish. But it was not until comparatively recently that special investigations of fish were launched: at the end of the 1920s.”

At the beginning, investigations of the form and structure of fish schools were not as strongly motivated by the sense of fascination that can be found in literature or, as we have seen, in the work of early swarm researchers such as Edmund Selous.

Theory: Noise

Amalgamations of Perplexity

“In the beginning was the noise.” This is not the opening sentence of Michel Serres's The Parasite(that would have been too prosaic). It rather concludes a paragraph in which he emphasizes the productivity of interference – the bruit parasite, as the French goes. Thus, instead of beginning this chapter with one hackneyed pronouncement or another about the fascinating nature of flocking birds, schooling fish, their tendency to ‘hover’ and ‘dance,’ and the apparently mystical beauty of such activity, I have decided to take the diametrically opposite approach. This is because a phenomenological or anthropological ‘view’ obscures the fact that, in the swarm itself, perception is always disrupted. Swarms shimmer in a field of tension between interference and organization whose discursive and historical dynamics are resistant to any subject-oriented or analytical perspective. Or, in Serres’s words:

Fish and Chips

A book investigating the role of fish schools at the intersection of biology and computer science can hardly refrain from using the subtitle ‘Fish and Chips.’ Although fish-school research has done much, as we have seen, to assist the fishing industry, I am less interested in British comfort food here than I am in the operative and performative function of agent-based computer simulations of schools. Whereas the previous two chapters dealt with the ‘fishy business’ of experimental and empirical studies, their optical and acoustic methods of observation, and their physico-mathematical modeling – over the course of which the ‘naturalness’ of fish schools was gradually subtracted and replaced by sets of functional principles – the aim of the present chapter is to elucidate the final stages of this subtraction. In addition to fulfilling this aim, however, I will also have to focus on a historiographical and epistemological recursion that marked a transformation in our knowledge of swarm collectives: against the backdrop of an epistemology of computer simulation, a biologization of computer science coincided with a computerization of (swarm) biology.

In what follows, I intend to examine the liminal area between fish and chips, within which biology and computer science overlap. The technology of computer simulation expands the realm of addressable problems by increasing the applicability of quantitative analyses. Simulations treat multiple variables simultaneously and make them manageable in time. In real time, that is, they directly address the behavior of complex systems without needing to be based on any specific reference to empirical data.

The self-organization of swarms – their opaque and nonlinear global effects, which are created by the local interactions of numerous agents – has both zoopolitical and zootechnological implications. Would it not be nice if human social processes and political decisions could take place as simply and quickly as the decisions of swarming collectives to change direction? And what would be more appealing than no longer having to rely on old analogies with colonies of ants and hives of bees as positive or negative examples of collective organization, but instead being able to describe the dynamics of ‘human swarms’ with reference to technical interfaces for communicating and making connections? Since the 1990s, swarms have been reformed into technologized, rationally implementable, and effectively visualizable zootechnologies, and thus it is easy to see why they have been applied as a powerful metaphor to describe various processes of “social swarming.” The condition of possibility for such metaphorical transferences, which differ from previous zoopolitical or anthropomorphic comparisons, was the reevaluation of swarms as figures of knowledge. Reference is no longer made to the biological ‘life form’ of various swarms but rather to their media-technological control logic and their available technical applications.

Any critical description of such transferences, however, should closely examine the extent to which swarms have been equated with networks, for instance, and where they are cited for their emancipatory potential despite the fact that human beings are fundamentally different ‘agents’ from those in swarm simulations, flocks of birds, and schools of fish. Perhaps it can be said that swarms should not be regarded as the most advanced form of older collectives – such as crowds or social groups – but rather as structures of organization and coordination that, in light of our media-technologically charged culture of intransparency and the ever-changing nature of so many areas of life, have become effective optimization strategies in these very areas. Or, better: they have become effective as self-optimization strategies whose specific governmentality is especially worthy of investigation.

“The Leviathan has had its day, and a swarm of hornets has appeared in its place.” With these words, Bernhard Siegert summarized the precarious relationship between polis and nomos and its effects on naval warfare during the twentieth century. The swarms that have since been established as figures of knowledge no longer confront governmental principles of order as something politically uncanny.

The development of a genuinely biological and ethological perspective on animal collectives by no means represented a clean break from the related fields of animal psychology and mass psychology, which had been heavily influenced by the well-known works of authors such as Auguste Forel, Francis Galton, Alfred Espinas, Gustave Le Bon, Gabriel de Tarde, and Scipio Sighele. Beginning around 1990, a methodological arsenal was gradually developed that made it possible to investigate the behavior of social insects, flocks of birds, and schools of fish without having to rely on mere anecdotal overviews or to focus on interactions between humans and animals. Systematic approaches and experimental methods found their way into zoological research practices, and only then did the data produced by zoologists become comparable and ‘scientific.’ The protagonists of the early stages of ethology were especially concerned with formulating a natural-scientific basis for their research that could serve as a counterposition to the vague psychological attributions – the poorly defined and anthropomorphizing concepts of instinct, emotion, and intelligence – that had previously characterized descriptions of animal collectives. Of course, these older approaches were not ousted altogether. Theories about the quasi-metaphysical levels of understanding possessed by collectives and theories about their miraculous collective instincts and ‘souls’ persisted even within the newly institutionalized field of behavioral biology.