Planetary alterity, solar cosmopolitics and the Parliament of Planets

Bronislaw Szerszynski

Introduction: how to gain and lose a planet

On 18 February 1930, our solar system gained a planet. Clyde Tombaugh, a 24-year-old amateur astronomer and son of a farmer from Illinois, had started work at the Lowell Observatory in Flagstaff, Arizona just a month before. The observatory had recruited him to help in the search for a hypothesised planet that had been going on since Percival Lowell had calculated in 1905 that only the existence of a distant, unseen ‘Planet X’ could explain observed perturbations in the orbits of Neptune and Uranus. Tombaugh had been given the task of using a blink comparison technique to detect any differences between pairs of photographic plates taken on different nights of the same area of the night sky (Schindler et al. 2018: 64–68). His successful blinking, by discovering a point of light that had changed position between 23 and 29 January, thus took the complement of planets in the system from eight to nine.

But then in 2006 the number of planets in our system fell back down from nine to eight again. The International Astronomical Union finally agreed a new definition of a planet, in an accommodation between competing scientific ‘ways of knowing’ in the astrophysics community – between a ‘structuralist’ focus on what a planet is made of and a ‘dynamicist’ focus on how planets move and interact with other bodies (Messeri 2010:190). The new definition was ‘Aristotelean’ in form, in that it consisted of a list of necessary and sufficient conditions for something to be classed as a planet (ibid.: 191). Thus, the IAU agreed that a planet is a body (i) that is big enough to have made itself spherical through its own gravity (i.e. achieved hydrostatic equilibrium), (ii) that orbits a star, and (iii) that dominates that orbit (Soter 2006, 2007). Pluto passed the first two criteria but failed the third, since it moves in an elliptical orbit that crosses that of Neptune – and is not even the largest body in that region of the solar system. Pluto was thus demoted to the status of a ‘dwarf planet’ or ‘Kuiper Belt object’, a decision that raised much controversy – not least among the wider public, who had got used to the idea of Pluto as part of the complement of planets.

What is ‘due process’ in the process of deciding what planets are and how many we have in our solar system? And how might exploring that question help to inform our thinking about alterity in non-human realms? In this chapter I will explore these questions by drawing on the cosmopolitical proposals of Isabelle Stengers and Bruno Latour: proposals for ways to determine what things exist in the world, and how they should coexist, which do not prematurely foreclose or divide questions of fact and value. I will repeatedly return to the two core cosmopolitical questions that Latour poses in Politics of Nature (2004a): how many are we? how shall we live together? I will explore the idea that deciding, reckoning, counting and accommodating might be operations that can be carried out by matter itself.

In doing so I will draw a great deal on planetary science, but also on the thought of Henri Bergson, Gilles Deleuze, Felix Guattari and Gilbert Simondon, which will help us develop a geophilosophical understanding of the modes of alterity exhibited by matter under planetary conditions. We will see that the concrete, individual planet that presents to us is a mere effect or phase in a wider ongoing process of ontogenesis, and that the material isolation of planets never wholly sunders them from immanence and possibility. We will also see that each planet exhibits alterity and multiplicity not just in relation to others, but internally, in that planets are always out of step with themselves, which is what enables their becoming. And we will see that the way that planets become, endure and interact involves various modalities of alterity, difference and multiplicity, modalities that are inextricably bound up with planetarity itself: with the particular mode of existence of planets.

Geophilosophy, transduction and planetary being

This chapter is an exercise in ‘geophilosophy’, a term that originates with Deleuze and Guattari (1994). By analogy with Annales historian Fernand Braudel’s argument that all history is geohistory (ibid.: 95), Deleuze and Guattari regard geophilosophy not as a particular branch of philosophy, but as a way of doing philosophy that pays attention to the Earth ‘as a milieu that determines philosophy from within, an earth that intrinsically belongs to philosophy, an earth that is the turf of philosophical thought’ (Gasché 2014:16). Geophilosophy is thus thinking not about but through the Earth, reflecting on how being for us is conditioned by planetarity.1

Although my account will draw a lot on the empirical findings and theoretical understandings of the natural sciences, the planet that Deleuze and Guattari’s geophilosophy summons for us is not simply the planet as known to science. Deleuze and Guattari argue that the sciences approach entities such as planets on what they call the ‘plane of reference’, cataloguing the lawful behaviour of the ‘actual’, of already constituted entities (Deleuze and Guattari 1994: 118). By contrast, they argue, philosophy – and I would say specifically geophilosophy – approaches entities on the ‘plane of immanence’, because it is concerned with the intuition of a ‘Whole’ that is full of the ‘virtual’, multiple possible states that may never actualise. Thus, for example, whereas the sciences employ the logic of exclusive disjunctions and the excluded middle (in which things are either this or that but not both), geophilosophy, for Deleuze and Guattari, employs a logic of the inclusive disjunction (things can be ‘both-and’). In this chapter I suggest, however, that planets even as understood by the natural sciences are entities that require us to think of them on the plane of immanence, of the virtual. And crucial to this is one important aspect of planetarity: that of multiplanetarity, that planets exhibit multiplicity – they are other, different, alterior.

In particular, first, planets are multiple in multiple ways: they manifest numerical difference (they are countable), but also qualitative difference and forms of ‘internal’ difference, which bring them into relation with each other. Second, planets exhibit multiplicity in a way that is distinctive to planetary (and other astronomical) bodies; even numerical difference means something different to planets. Third, each planet as it forms and develops has, in a sense, to discover and invent planetary alterity anew. Fourth, we as human thinkers about planets have to follow an analogous process to the becoming of planets.

To expand on that last point, we can draw on the work of Simondon.2 Simondon (1992) criticises the categories of conventional logic for the way that they deal with the ‘already constituted individual’, treating the process of individuation through which it came into being as not important, and dividing the individual entity from its milieu. He argues that we should ‘understand the individual from the perspective of the process of individuation rather than the process of individuation by means of the individual’ (ibid.: 300). To help us to grasp this, Simondon uses the concept of ‘transduction’, an idea that points in two directions. First, transduction is a ‘psychic process’ and a ‘logical procedure’ that occurs within the analyst of a phenomenon; but in contrast to the standard logical procedures of deduction and induction, it occurs as a continuous process of discovering ‘the dimensions according to which a problematic can be defined’. However, second, transduction is also the process through which an entity itself individuates from the pre-individual state, and out of which emerges not only the individual itself (always only a partial resolution of the latent potentials of the pre-individual state) but also the structures and dimensions within the individual that will determine its development and the individual-milieu relation (ibid.: 313). So, as well as us as thinkers being engaged in transduction in thinking about them, planets themselves can be understood as being engaged in transduction in coming into being. Planets as they develop compose themselves into certain kinds of individuals, and then into different categories, and do a kind of accounting or reckoning with each other.

Due process in the heavens

How do we understand what happened in the two cases discussed in my introduction? What happened to the number of planets in our solar system in 1930 and 2016? I want to start to address this question through the idea of cosmopolitics, as developed by Isabelle Stengers. Stengers takes the concept of cosmopolitics from Kant (1903) but develops it in a decidedly un-Kantian direction: for Stengers, adding ‘cosmo’ to ‘politics’ does not point to a kind of universality, and nor is cosmopolitics beyond politics; instead, it is a recognition that there is something in politics that exceeds the human will (Stengers 2011: 351–62). As Latour puts it, in Stengers’ redefinition of ‘cosmopolitics’ ‘the strength of one element checks any dulling in the strength of the other. The presence of ‘cosmos’ in cosmopolitics resists the tendency of ‘politics’ to mean the give-and-take in an exclusive human club. The presence of ‘politics’ in cosmopolitics resists the tendency of ‘cosmos’ to mean a finite list of entities that must be taken into account. Cosmos protects against the premature closure of politics, and politics against the premature closure of cosmos’ (Latour 2004b:454).

In Politics of Nature (2004a), Latour develops a particular cosmopolitical proposal, and we can use this to distinguish three possible ways of treating ‘how many planets are there’ as a cosmopolitical question. The first would be to consider it within the frame of what Latour (1993) had called the ‘modern constitution’, an understanding of the world that makes a sharp distinction between nature and fact on the one hand, and politics and value on the other.3 A modern would locate the two changes in the numbers of planets described above – from eight to nine and then back to eight again – purely in the realm of ‘culture’ rather than ‘nature’. Such a position would be to say that Pluto as an astronomical body did not ‘blink’ into existence on February 1930; the only thing that blinked was Clyde Tombaugh himself. Pluto, the ‘thing in itself’, had been there for billions of years, although unknown to humans; however, over time, technology and science developed so that by 1930 we humans were able to perceive it and incorporate it into our model of the solar system. In this version of solar cosmopolitics, it was only our model of the solar system that changed, not the solar system itself. Then, we might say that the more recent change downwards was merely the result of change in a human classificatory system, as this was adjusted to reflect improvements in human understanding. Out there, in space, once again nothing had really changed.

But this is to enact a disguised cosmopolitics, to hide the political aspects of planetary designation. In Politics of Nature (2004a), Latour suggests how we might assemble a cosmos of humans and non-humans in a way that accords with what he argues would be ‘due process’ rather than arbitrary power. He proposes a modern, cosmopolitical constitution that would require the setting-up of a more-than-human ‘Parliament of Things’ in which politics and science were not seen as separated powers but were done together. However, Latour’s parliament would have its own distinctive separation of powers. The first, upper house would have ‘the power to take into account’, be oriented to the question ‘how many are we?’, and obligated to keep as open as possible questions of truth, consistent with the due process rules of ‘perplexity’ and ‘consultation’. The second, lower house with ‘the power to arrange in rank order’ would be about closing things down again, guided by the question ‘can we live together?’ and involving procedures and norms of ‘hierarchization’ and ‘institution’ (Latour 2004a: 109).

Asking about planets in the more transparently cosmopolitical way described by Latour would involve a multi-disciplinary speculative planetology that kept as open as possible the question of how we think about planets. This view would ask whether the decision about Pluto followed ‘due process’. Following Latour, the question of ‘how many are we’ would need to involve skills from the humanities as well as the sciences, and to involve wide consultation with publics to open up the widest set of questions about what planets are, how many there are, what they might do, and what their significance is for the cosmos. Then, the processes of hierarchisation and institution would also involve diverse skills and conversations in order to come to a shared agreement – and crucially, the process would never be regarded as completed once and for all. As Latour insists, ‘all Republics are badly formed, all are built on sand. They hold up only if they are rebuilt at once and if the parties excluded from the lower house come back the next morning, knock at the doors of the upper house, and demand to participate in the common world’ (Latour 2004a: 183).

But in this chapter, I want to suggest a third option, one that radicalises Latour’s idea of the Parliament of Things in a way that is even more open to the agency of the non-human. One limitation of Latour’s parliament is that it is less a parliament of things than a parliament of people who can represent things – as if things can only be lively if they are enrolled into networks with humans. Elsewhere Latour makes it clear that he is very happy for the world to be doing things when we are not there. In ‘Irreductions’, for example, he writes that ‘[h]ermeneutics is not a privilege of humans but, so to speak, a property of the world itself’ (Latour 1988: 245).4 So how can we be more Latourian than Latour when it comes to convening our planetary system? What if we were to ask, not ‘what human skills and knowledge do we need to gather together to open up and then close down the question of what is a planet, and thereby how many planets we have’, but instead, ‘how do planets themselves decide “how many are we?” and “how should we live together?”’ Is it really only humans that ‘make planets count’, or do planets do it themselves – make themselves countable things? And do planets decide how to live together, and if so, how? Can we speak of the solar system as itself a solar Parliament of Things – and one that does not have to pass through the minds of human beings to constitute itself? And what is the role of alterity in the countability and coexistence of planets?

The changing countability of planets

Of course, when Bruno Latour asked, ‘how many are we?’ he was not simply thinking about the sheer number of beings admitted to the collective. But let us start by thinking about numerical alterity – about how planets may be counted. We will see later that the question of what it means to say that planets or anything else are countable is not so simple, but let us start with a simple working definition, with two parts. First, countable things have to be sufficiently individuated that they can be counted – they have to be other to each other. Second, they need to belong to a boundable class of things – so that we know when to stop counting: as a class they need to be other to the things that we don’t want to include in our count. Thus, to speak of Earth and Mars being two planets is to speak of two separate members of the class of things that are planets. And in some ways planets seem eminently countable, as they are objects that are materially isolated and thus ‘other to each other’ to an extent that is never found with entities on Earth.

Planets have not always been seen as such clearly demarcated things. Defined initially as lights in the sky that seem to move slowly against the fixed stellar background, it was nevertheless not always obvious that different lights seen in the sky on different occasions were in fact the same celestial body. The Mesopotamians seem to have known that ‘the morning star’ sometimes seen before sunrise in the east, and the ‘evening star’ sometimes seen after sunset in the west, are the same entity (the planet that we now call ‘Venus’). The Sumerian myth of the goddess Inana’s disappearance from the sky and her sojourn in the underworld is at least partly a narration of the synodic movement of Venus in the sky as seen from the Earth, by which it spends eight months visible in the east, then disappears for about two months when it goes behind the sun, then reappears in the west (Cooley 2008). But in ancient Greece these two celestial lights were still regarded as different things, named Phosphoros and Hesper or Hesperus respectively. Book XXII of Homer’s Iliad is full of astronomical allusions; the glint of Achilles’ spear as he prepares to slay the giant Hektor is likened to Hesper, the evening star heralding the onset of night (Genuth 1992: 295). The recognition in ancient Greece that they were the same entity is traditionally credited to Pythagoras around 500 BCE (Dreyer 1953: 48).5

Then, in the ancient cosmos of Ptolemy’s second-century Almagest, the planets became a group of seven lights in the heavens, including the sun and the moon, that were all understood to circle the Earth (Ptolemy 1984). The Ptolemaic universe defined dominant elite views of the heavens in Europe for centuries to come, and implies that the planets are a ‘closed class’, closed in a similar way to that in which some word classes are closed, in the sense that it is difficult to add new members to them (Dixon 2004). The class of all of planets was not seen as being capable of being expanded by discovery or fiat; the seven-ness of the planets was not accidental but somehow essential to the concept of planet.

Copernicus famously displaced the Earth from the centre of the universe and set it on the move. A planet was now defined not as a body that moved against the star field as seen from the Earth, but as one that orbited the sun. In the Copernican system the sun and moon were removed from the list of planets – but the Earth was added, to make six in all. In the longer term, the work of Copernicus was to signal a revolution in thought which turned the closed world into an infinite universe. Alexander Koyré describes this revolution thus: ‘the disappearance, from philosophically and scientifically valid concepts, of the conception of the world as a finite, closed, and hierarchically ordered whole … and its replacement by an indefinite and even infinite universe which is bound together by the identity of its fundamental components and laws, and in which all these components are placed on the same level of being’ (Koyré 1957: 2). In this new ‘infinite universe’, the sun became a star, a mere member of an open class of ‘suns’ with potentially infinite membership.

But was the class of planets around the sun also an open class? Was the six-ness of the planets an empirical fact, that could be overturned by new discoveries, or an a priori truth? In 1596 Johannes Kepler noted that the ratios of the orbits of the six planets seemed to correspond to what would be the case if the five Platonic solids (tetrahedron, cube, octahedron, and so on) were nested between each pair of orbits in turn; he called this the ‘Mysterium Cosmographicum’ – the secret of the cosmos (Kepler 1981). This seemed to imply that the convening of the planets around the sun followed some kind of logic of necessity – not just regarding their arrangement (how shall we live together?) but also their number (how many are we?). For in three-dimensional Euclidean space there can only be five Platonic solids, so if each gap between planetary orbits can be uniquely assigned to one of them, there can a priori only be six planets. Such thinking also resurfaces in later post-Copernican thought – for example the Titius-Bode Law of 1772, that observed that the semi-major axes of the planets are of the form a = 4 + x, where x = 0, 3, 6, 12, 24, 48 etc., which was apparently confirmed by the discovery of Uranus in 1781 (Jaki 1972).

The Titius-Bode law also seemed to predict that there should be a planet between Mars and Jupiter, where instead there seemed to be an empty space. Between 1801 and 1845 the hunt for this apparently necessary planet resulted in the discovery of first one, then many objects orbiting the sun, classed first as the missing planet, then downgraded to asteroids (Hilton 2017). Once the huge difference in size between asteroids and planets was determined, a formal definition was not seen as socially necessary. The definition of ‘planet’ was prototypical (e.g. a planet is something like the Earth), and took a loose, ‘family resemblance’ approach, able to accommodate the different ways of knowing of different communities of natural philosophers, then scientists and the wider public (Messeri 2010). It was this informal definition that was to last until 2006.6

Planets and numerosity

Counting (form the Latin ‘com-putare’, reckoning together) is not the only way that planets might relate to number. Cognitive psychologists and anthropologists tell us that, for humans and other animals, ‘number’ is not really a single system or phenomenon, but a collection of separate numerosity systems with quite different logics (Dehaene 1992: 34–35). Counting is just one of these – a particular form of numerosity which is closely linked to verbal ability and relates as much to parole (specific embodied performances of language) as to langue (language as an abstract system of relations); counting is at core an embodied action, linked to recitation, rhythm, cadence, one that is often gestural and indexical, involving pointing towards countable things (Maurer 2010). This is most clear when people count on fingers, knuckles or other body parts – or when shepherds use sheep-counting rhymes, or children use counting-out rhymes to generate remainders to choose people and items (Bolton 1888). Counting is linked to ordinal numbers, which are semantic (they denote – e.g. we say ‘the third planet’) and paradigmatic (are potentially interchangeable with other predicates that also denote the same entity, such as names) (Crump 1990: 39).

As we have seen, humans can count the planets. Doing so can be described as making a one-to-one correspondence between the planets and the first eight members of the set of natural numbers – 1, 2, 3, 4 etc. – or with sets of other ‘numerons’ such as tokens, words or body parts (Crump 1990: 32). Because of the way that planets ‘live together’, in near-circular concentric orbits, it might feel more natural to count them ordinally – ‘first’, ‘second’, ‘third from the sun’; or we could even count them without numbers – we could recite the names of the planets, perhaps aided by a mnemonic. But planets themselves do not count – the lack of language or an articulated body would make it hard (though maybe they could be said to have counted themselves if they each were able to attract only a single moon as a numeron). However, planets can certainly be interested in some features of ordinality – if you are a planet that has a big planet near you, whether the orbit of that planet is inside or outside your own can make a big difference to how you live with them.

A second form of numerosity is absolute number, about which planets seem less interested. Dehaene identifies arithmetic or calculation as a distinctive system of numerosity. Like counting, the process of calculating seems dependent on language but goes beyond finding the number that corresponds to a collection of items to the manipulation, combination and application of abstract numbers – numbers as concepts that are detachable from any particular manifestation (Dehaene 1992; Pica et al. 2004). If counting is related to ordinal numbers, calculation is more tied to the cardinal numbers, which are inherently abstract, syntactic, operational and ready for manipulation (Crump 1990: 39). Indeed, when we count a class of objects we turn the last number in the count into the number that represents the whole set – we make it cardinal, ready for calculation (Dehaene 1999: 119).

But we can relate planetarity to certain other forms of numerosity that allow some forms of calculation but resist absolute number. One way to do this is by approximate numerosity, which is available to animals and preverbal infants as well as adult, verbal humans. This involves processes such as ‘subitising’ (recognising the numerosity of small numbers of items by eye) or ‘estimation’ (judging the size of larger groups). Planetary systems can be said to estimate. Like all non-linear self-organising systems, if they can be said to be computing, they do so using analogue computation, based on continuous physical quantities such as speed, mass and distance, rather than distinct natural numbers (Pickering 2009). When a planetary system forms, there will be an approximate number of planets that the system is ‘trying’ to form, as the interactions among the planets ‘estimate’ a reasonable number of planets for their system.

But ethnomathematics can help suggest other ways to think about planets calculating without absolute number. In many number systems, numerosity – for example ‘threeness’ or ‘fourness’ – is not seen as an abstract concept that can be transferred between different classes of entities: the ‘threeness’ of one class of thing, for example, might be seen as quite distinct from the ‘threeness’ of another. For example, Helen Verran argues that whereas in English numbers are simply a subset of qualities or predicates that can be applied to any kind of spatiotemporal particular (objects, places and so on), in West African languages like Yoruba numbers are modal terms (i.e. modes of presentation) that are applied to sortal particulars (things which are defined as inherently possessing certain qualities). As she summarises, ‘[a]n English speaker who talks of spatiotemporal particulars might say these particulars have qualities. A Yoruba speaker, however, talks of sortal particulars, and since these particulars have been defined by categorisation around sets of characteristics, these objects cannot be said to have qualities, but they can be said to have modalities, or modes’ (Verran 2001: 137). Numbers in Yoruba work less like adjectives than adverbs; they describe how things appear, whether as a group or as a collection of individuals, and also as a collection of how many (Verran 2001: 67). Another feature of number in many languages that breaks with the logic of abstract number is that of ‘numeral classifiers’ – terms that are included when numbers are attached to nouns and are specific to that class of noun. Sometimes there are only two sets of numerical classifiers, for animate and inanimate things respectively – as if the three-ness of ‘three rocks’ and of ‘three cows’ are different, but other languages have dozens (Ascher 1991: 11–13).

Such features of non-Western number systems have been taken to mean that non-literate peoples simply do not understand abstract number; however, treating numbers as adverbial modes of presentation or using numerical classifiers can be read as a recognition that quantity can be meaningless without quality – that it is not always reasonable to abstract numerosity from the specific qualities of the type of entity in question. Indeed, with the adoption of the 2006 definition, planets in Western scientific thought became something much more like one of Verran’s ‘sortal particulars’, a kind of entity that inherently has a particular set of relational properties. Just as a living thing’s interactions with its environment is not accidental but constitutive of its status as a living thing, similarly what makes something a planet includes the role it plays in a wider assemblage and set of processes.

Modes of planetary alterity

In the last section we focused on planetary alterity as involving numerical and quantitative difference. But in order to develop a fuller geophilosophical account of how planets live together we also need to attend to different modes of alterity. Firstly, planets are alterior, other to each other, not just because they are materially separated and countable, but also because they take different forms. When we talk about Earth and Mars, we are not just talking about two numerically distinct things, like two electrons; they also look different and behave differently – they exhibit qualitative difference. But secondly, in order to understand this qualitative difference geophilosophically – not just as a statement of fact, but as something that binds them in ontogenetic relations – we also need a notion of alterity that does not just involve contrasting a thing with something else. We thus need to mobilise concepts of ‘internal multiplicity’ – ‘non-oppositional difference’, or ‘difference in itself’ (Deleuze 1994: 28–69). This further complicates the idea of multiplanetarity and planetary alterity as I have developed it above; we will see that, on the dimension of internal difference, the ‘multi’ in multiplanetarity does not simply divide planets from each other but also divides them from themselves.

This internal difference within planets is not simply a descriptive state of affairs: it is a dynamic, active force. We can develop this idea using three concepts developed by Deleuze – ‘multiplicity’, ‘the virtual’ and ‘the intensive’. In his Bergsonism (1988), Deleuze takes from the philosopher of vitalism a particular idea of ‘multiplicity’ as the differential ground of existence. Bergson insists on looking at difference not as a secondary property derived from a metaphysically prior ‘identity’ but as itself originary. For Bergson (1921), difference is an explosive force within things that creatively and inventively generates novelty. Deleuze also takes from Bergson the idea of the ‘virtual’ as a way to understand the genesis of new forms. Unlike ‘the possible’, which we think of as imagined counterfactuals with no reality, the virtual is no less real than the actual. Virtualities are already present in the world, real but latent, and simply may or may not be actualised. Finally, the distinction between the ‘intensive’ and the ‘extensive’, discussed by Deleuze in his later work Difference and Repetition (1994), was first posed by the physicist Richard Chase Tolman (1917). ‘Extensive’ properties are divisible properties such as length, volume and mass that together comprise the stable, actual, completed form of an object. ‘Intensive’ properties such as temperature, pressure and density, by contrast, cannot be divided or altered without introducing asymmetries and qualitative change. Deleuze links Chase’s concept of the intensive closely to Bergson’s concepts of multiplicity and the virtual.

We can use these concepts, combined with Simondon’s focus on ontogenesis as a continuous process of individuation, to understand the role of alterity in planetary becoming – including the emergence of the numerical alterity discussed above. Before there are any countable planets orbiting around a given star, there is a smeared, immanent solar nebula, a spinning protoplanetary disc of hydrogen, helium and metals. Then, fluctuations, generated largely by the dynamics of the rotating disc itself, disrupt the latter’s homogeneity, triggering a process whereby intensive forces starts to generate extensive form, and the disc organises itself into separate clusters of matter. Initially these are simply areas of greater densities of particles, but some of these will clump into ‘planetesimals’, some of which then combine into ‘planetary embryos’, which either separately or through combination form the cores of a number of planets circling around the central star, made variously of rock, ices and captured liquids and gases.7 So planets fall into being, self-assemble, create their own gravity wells, forming dense, approximately spherical bodies orbiting in stable near-circular orbits separated by empty space.8 They turn themselves into countable things.

But by such processes of ontogenesis, planets also come to have qualitative alterity, because of the specific intensive conditions under which their emergence takes place. Such conditions include the type of star and the metallicity of the accretion disc; how far from the sun the planets form; and the presence of other planets that might affect the process of formation or their movement. Analyses of the distribution of the mass and size of known exoplanets (planets around other stars) suggest that planets tend to fall into three groups: rocky planets (like Earth) with at most a thin atmosphere; middle sized planets (like Neptune) with a solid rocky or icy core and massive, thick atmospheres; and gas giants (like Jupiter) with metallic cores (Buchhave et al. 2014; Chen and Kipping 2017). These qualitative groupings are a product of emergent (we might say with Simondon ‘transductive’) patterns in their ontogenesis. If a planet is in – indeed is – a constant process of becoming, of taking form, then it is a form that it generates ‘on the fly’ from its own internal inconsistencies.

Here the work of Simondon is useful again. For Simondon, individuation is ‘a partial and relative resolution manifested in a system that contains latent potentials and harbors a certain incompatibility with itself’ (1992: 300). Becoming is thus ‘a capacity beings possess of falling out of step with themselves …, of resolving themselves by the very act of falling out of step’ (ibid.: 300–1). Viewed this way, a planet’s becoming is determined on the virtual plane by its internal incompatibilities or ‘singularities’, non-linear thresholds which act as attractors or tipping points in the dynamics of the system.9 The trifurcation of planets into Terran, Neptunian and Jovian worlds suggests that there are singularities around 2.0 Earth masses and 0.41 Jupiter masses, unstable saddle points that divide the possible futures of emerging worlds. In theory, all planets contain the same virtual structure of singularities and possibilities and are simply separate ‘actual’ instantiations of it. But because of the way that planets in their forming diverge into these different classes of planet, and then follow a particular developmental course within that class, while increasingly isolated in the vacuum, it is truer to say that they each instantiate a particular subset of that common virtual structure.10

Planets, once formed and kept ‘out of step with themselves’, are constantly involved in generating new forms of otherness within themselves. For example, diverse forms of immanent, intensive alterity are constantly being generated in their vast, extended regions of solid, liquid or gaseous ‘continuous matter’, without clear borders or interfaces, crosscut by intensive gradients of pressure and temperature, and in constant motion on different timescales. In fluids in particular, the ‘actual’ and the ‘extensive’ is also full of ‘intensity’ and ‘virtual’ possibility. Fluids are constantly generating and dissolving form (Schwenk 1965). This kind of planetary alterity has three distinctive characteristics: rather than being transcendent, absolute and completed, it is immanent (internal to the region of the planet), gradual (it manifests as gradients) and generative (it is constantly producing form and new gradients).

But planets are qualitatively other to each other not just because of conditions which they are subjected to in their formation, but also because to some extent they can take hold of their conditions as they ‘fall into being’. This point can be seen as a generalised version of Gaia theory (Lenton and Wilkinson 2003), one that sees the significant shaping powers that biological life can have over a planet’s fate as merely a specific example of a wider set of planetary powers of self-organisation. Planets can do this because of certain specific features of planetary being: planets are assemblages of baryonic matter, chemically diverse, intermediary in temperature between stars and interstellar space, and gravitationally differentiated into different strata and compartments. As they orbit their star they are also subjected over long timescales to metastable flows of energy from the star as well as from their hot cores, with patterns of heating and cooling. They are thus maintained away from equilibrium, bringing their parts into active relation with each other, and able to do work on themselves (Kleidon 2016).

Thus, the concatenating internal differences within planets mean that they divide internally into different ‘spheres’ and substances and entities with different properties, which are maintained in dynamic relation. So, planets become historical entities, other to each other in a more than numerical way; their characteristics cannot be simply understood as the working out of universal laws; they are qualitatively unique, path-dependent entities, whose powers and possibilities are dependent on the particular course of development through which they have passed. Planets bifurcate, go through revolutions (Lenton and Watson 2011), and thus become other to themselves in a diachronic sense. Terrestrial planets in particular can retain the power to evolve and change, and they do so at their own pace. The divisions between the four great aeons of Earth’s geohistory – the Hadean, Archean, Proterozoic and Phanerozoic – are identified first of all by the signs left in the geological record that a significant transition in the Earth has taken place, and only secondarily assigned chronological dates. These are thus internally generated forms of time (Adam 2004), as planets develop in unique, path-dependent ways, as they undergo Simondon’s ‘transduction’ and themselves discover the way that the general laws of planetarity will be true in their domain. Planets are never solely at the mercy of external forces; even the effect on them of a collision with another astronomical body or a change in the brightness of their star will depend on how the planet has developed so far.11

Planets also react gravitationally to each other’s presence, which can shape how many planets there are and where they position themselves in the planetary system. This is particularly significant in the early stages of a planetary system’s life, when planets have formed but the disc has not yet been cleared of gas and planetesimals. During this time, larger planets interact gravitationally with the remaining protoplanetary disk, creating spiral density waves that have the effect of pulling the planets inwards towards their star (Morbidelli and Raymond 2016: 1967–1969). This migration is likely to destabilise the orbits of other planets and potential planets, especially smaller ones, which may be forced to merge, to become minor planets or ‘trojan satellites’ within the orbits of larger planets or be lost from the system altogether. But the remaining larger planets will tend to lock each other into orbits characterised by mean motion resonances between immediate neighbours in the ordinal sequence (for example, with the inner planet orbiting three times for every two orbits of the next planet outside it), and then be shuffled into new positions as the resonance starts to disrupt itself (Levison et al. 2011). Through this and other dynamic processes, the planets of a system arrive at a long-term arrangement of how they will share the space around their sun.

Conclusion: How many worlds are we? How shall we live together?

In Order out of Chaos, Prigogine and Stengers (1984: 305–306) suggested that a complex shift of cosmopolitics happened in early modern science, one in which the direction of modern science swerved as consequentially as any giant planet engaged in a ‘grand tack’ across its planetary system. For the classical science of Aristotle, pure mathematical descriptions had only been applicable to the incorruptible ‘superlunary’ world of the heavens and the gods; Earthly, ‘sublunary’ nature, by contrast, was regarded as a world of becoming, life, change and decay that did not admit of mathematical or deterministic understanding. The original intention of many early modern scientists seems to have been to extend the logic of the sublunary sphere to include the heavens: to show that celestial bodies were in principle no different to bodies on the Earth. Galileo thus caused controversy by arguing that the spots revealed by his telescope on the surface of the sun were not bodies orbiting in front of it, but actual spots or imperfections on the Sun itself (Galilei and Scheiner 2010). Yet the focus of Galileo and Newton on the dynamics of pendulums and planetary orbits had the effect of reversing this trajectory: instead of extending the sublunary realm of impermanence and change into the cosmos, they extended the timeless perfection of Aristotle’s heavens to include earthly things, with the ideas of lawful predictability, reversible (and thus timeless) change, and a detached, objective observer.

However, Prigogine and Stengers argue that with the rise of new sciences such as non-equilibrium thermodynamics another shift is happening, one that places the observer within a world characterised by inherently unpredictable change and emergent order. These new sciences, Stengers argues in a later book, are ‘open to a dialogue with a nature that cannot be dominated by a theoretical gaze, but must be explored, with an open world to which we belong, in whose construction we participate’ (Stengers 1997: 39). The idea of solar cosmopolitics developed in this article can be seen as illustrating and expanding on this claim, by showing how we can approach the findings of planetary science through the lens of geophilosophy, thereby revealing the heavens to be a realm of becoming, of negotiation and accommodation – and thus, by ‘resisting the tendency of cosmos to mean a finite list of entities that must be taken into account’, bringing a kind of politics into the cosmos (Latour 2004b: 454).

We have seen how the ethnomathematical findings of Crump, Verran and Ascher and others can help us clarify that the ways in which planets are individual countable, estimable or calculable entities, and also the ways that planets themselves do numerosity – counting, estimating, calculating – are specific to the mode of existence that is planetarity. Even bare numerical difference takes a particular form when it comes to planets; and each planet, in interaction with the planets and protoplanetary disc around it, has to discover planetarity and individuality for itself, and in its own way. The contemporary planetary sciences, when approached geophilosophically, further suggest that, while we no longer think that the number of planets in the solar system is an a priori, necessary fact in the way that it was for Ptolemy, neither is it simply a contingent one, as would be suggested by Newtonian, ‘classical’ physics. How many planets inhabit a given planetary system, and how they live together, are questions that have to be worked out by the planets themselves, jointly and severally, as they explore the virtual structure of singularities and possibilities that they inherit from the protoplanetary cloud out of which they form and seek a regularised mutual accommodation.

Drawing on the geophilosophical thought of Bergson, Deleuze and Guattari, we have also seen that different modalities of planetary alterity are a central part of planetary being and becoming – and that it is non-oppositional or ‘internal’ alterity that is more fundamental to shaping the emergence and character of numerical and qualitative alterity in planets. A planet is not just actual (the state of affairs at any one time) but virtual, in that it inherits a virtual structure of singularities and potentialities that may remain latent at any one time. And this virtuality of a planet, its power of becoming, creativity and historiality, derives from the way it manages to keep incompatible with itself, particularly in terms of its intensive properties, kept away from equilibrium by flows of energy.

We have also seen that planetarity weaves together alterity and relationality in distinctive ways. I drew on the ideas of Simondon to suggest that a planet is not simply the more-or-less completed solid ball of matter that is its clearest presentation to us, but a process of individuation, stretching back to the pre-individual stage before the planets were distinguishable from each other, and involving its milieu even after it has differentiated itself from it. Planets continue to interact through gravitational – and occasionally collisional – encounters, in ways that enable a ‘parliament of planets’ to answer the questions ‘how many are we’ and ‘how shall we live together’ for themselves.

But we also know that planets have the potential to encounter each other in other ways than the gravitational and collisional. As materially isolated bodies that can transductively take hold of their own development and pass through creative bifurcations, planets are able to carry out separate, diverse experiments in the self-organisation of matter. In the case of the Earth, this particular experiment has resulted in the emergence of organic life, the power of sight, and now the interplanetary movement of machines and potentially living things – all of which offer new ways in which the independent material experiments carried out by planets can start to interact with each other. It is surely the case that other planets will have produced very different new material powers and possibilities for interplanetary relationality, made possible by their own working-through of planetary alterity. However, thinking through the wider possibility space of how the being and becoming of different planets might weave together in radically different ways, activating new singularities and generating forms of alterity not available to single worlds alone, is a project for another day.