For a certain number of species of birds (passerines, palmipeds, waders), the presentation of a blade of grass (or of straw or moss) to the female by the male during courtship rituals, as a tribute seems to play a specific role in the linking of behavioural sequences. For example: the male zebra finch first sings and dances so as to attract the attention of the female, he perches on a branch and whilst balancing on it, brandishes a blade of grass in his beak. Then he imitates the characteristic position of the young of this species seeking food, tipping his head to one side, seeming to offer the blade of grass but without letting go of it.1 This use of the grass stem index, which seems to entail no improvisation is especially interesting to us to the extent that it could be related to the functioning of human faciality traits such as those described by ethologists with regard to ‘flirting behaviour’ and ‘welcoming behaviour’. It is a matter of rapid imitations, the encoding of which is probably hereditary and the details of which can only be detected by slow-motion filming. In particular they include imperceptible phases of eyebrow raising and widening of the eyes, which last no more than two- to three-tenths of a second.2 The birds’ grass stem ritual evidently doesn’t put the same components of expression into play as flirtation and welcoming rituals amongst humans, and we should perhaps talk of silhouetting traits instead of faciality traits. The difference matters, because unlike what has happened with humans, there has not been any deterritorialisation of a face, that is to say, a surface of inscription, in relation to the snout of the animal in the case of birds, on which gestural, postural, sonorous traits of expression as a whole would reverberate, be concentrated, articulated and hierarchised, by way of the anatomic freeing up of the lips, the particular development of facial muscles, correlative to that of the phonatory apparatus. The ‘comparison’ must not be made here to the detriment of the analysis of the specific traits of each assemblage. It would be relatively easy to interpret the bird’s grass stem and the faciality traits of the human on the basis of the same psychoanalytic algorithms: phallus, unary trait, bar of castration (without mentioning part and transitional objects, which are now a bit out of fashion!).
It is by ‘examining’ the differences, that is to say, by really doing analysis (contrary to what psychoanalysts claim to do with their stereotypical interpretations) that one will perhaps succeed in making the existence of abstract machinisms that are not common – since unlike ‘complexes’ they cannot belong to anyone – appear, which participate in the same deterritorialising processes, the same adaptive headlong flight, the same kinds of semiotic solutions … By starting from some phylogenetic landmarks, we will therefore try to grasp the ‘machinic sense’3 of the functional evolution of this grass stem ritual. Ethologists explain to us that it is a matter of an archaic ‘residue’ that is related to nesting behaviour. This does not imply that it might be reduced to a simple function of representation, stimulus, or reflex trigger. Rather than talking of signs here we would like to talk of a concrete machine (a machinic index or diagrammatic operator) that participates in machinic assemblages without necessarily referring to the hierarchised systems of the reflex arc, to a signifying structure or even to a manifest assemblage of enunciation. What has to be accounted for here is thus not the application of a universal topics that would have to ‘localise’ contingent singularities, but a ‘machinics’ which brings into play components that are very different from one another (hereditary, acquired, improvised …) and which has crystallised in a mode that is irreducible to any general formula. Perhaps it will be objected that we are displacing the problem of ‘universals’, by ourselves postulating a universal deterritorialisation instead of an order of rational progress. But the difference resides in the fact that this deterritorialisation doesn’t have any order ‘in general’ and doesn’t participate in a progress that is inscribed in the order of things.4 The semiotics of the grass stem results from the ‘refining’ of a deterritorialisation, a territorialised nesting behaviour. We will see that this local deterritorialisation has as its ‘consequence’ a change in the abstract formula that articulates the semiotisation of territory and that of sexuality. But this mutation doesn’t as such involve any ‘political’ progress for the species or a liberation of individual desire. Abstraction and dialectical determination always remain coupled with semiotic unevenness, archaisms, stratifications that result from interactions between phylogeny and ontogeny, ecological and historical ‘Accidents’ that specify them, without irreversibly attaching them to a context or to an evolution that is fixed once and for all.
This is the case, in particular, with what we might call the ‘abstract machine’ moving-towards-more-sociability. For certain number of species, the fact that it seems to involve the deterritorialisation of a series of components – as the following examples will suggest for finches – doesn’t automatically imply that they are linked to an idea of ‘progress’. Not that it might be necessary to give up on estimating the progress of inter-assemblage transformations, with all the risks that involves. But it doesn’t entail any univocal relation with one formula rather than another. If it exists, it is at the overall level of a rhizomatic process. It is political, not normative: in other words, it doesn’t arise from transcendental characteristics (example: individual freedom, which is manifestly lacking amongst ants), but it must be evaluated as a function of the rhizomatic expansion of assemblages, their lines of flight, their lines of creation, the elegance of their solutions – to talk in the way mathematicians do – and since we are not concerned to avoid the accusation that we are irresponsible idealists, why not also add that it must be evaluated as the function of a grace and a beauty to which it isn’t just human eyes that are sensible?
The semiotics of the grass stem amongst birds, like that of faciality amongst humans, doesn’t just have a function of representing, triggering or inhibiting. With other less ‘spectacular’ components of the rhizome of assemblages (investments that are hormonal – which we will come back to with regard to the refrain – emotional, perceptual and also ‘political’ at the level of the territory and the species), it works directly at the production of a style of life, at the semiotisation of a world. To try to illustrate the non-representational, a-signifying, diagrammatic character of this particular kind of semiotic component, we will now review two series of examples: the first taken from some very different species of birds, the second from among the variants of a very old species of chaffinch. However superficial our inventory may be, it should allow us to set out some hypotheses concerning the ‘machinic sense’ of this semiotics of the grass stem, which is to say that the deterritorialisation of the nesting behaviour into a symbolic ritual seems to be correlated with two other series of deterritorialisations, concerning:
•the mode of semiotisation of territory amongst the most ‘evolved’ species, which tends to open up to a development of gregariousness and an intensification of social life;
•the specific refrain function, which also tends to become less ‘territorial’ and to place itself at the service of more intimist assemblages, like those of courtship rituals, or even to give rise to solitary improvisations for ‘the pleasure of it’. On the one hand, then, the opening up to the socius, and on the other, to the individual.
In the mating season, the male grebe – a species of palmiped living in small groups, which nonetheless have a very strict conception of territorial defence – constructs a floating nest with the collaboration of a female. Throughout this activity, the courtship ceremony is punctuated by face-to-face intimidation, simulated grooming and the offering of bits of vegetation. The fact that this last form of behaviour is not ‘yet’ ritualised could be brought into relation with the relatively poorly developed sociability of this species.5
One finds a giving ritual that is already much more complicated amongst grey herons, a wader that lives in small colonies (although certain herons can boast up to one hundred nests) and coexists with sparrows as well as falcons and kites, without any problem. A spot for a nest having been chosen – already constructed or not – once a female has started to take an interest in the cries, the bobbing, the inclining of the neck, the ruffling of the male’s feathers, the latter stops his attempts at seduction so as to invite his partner to effectively participate in making the nest. To do this he holds out branches that she will place on the nest as it is being constructed: but any kind of sudden gesture or clumsiness can call everything into question and bring them to blows.6 One thus remains closer to reality than symbol here, and the conjugal assemblage (let us note in passing, so as to illustrate our previous remarks on this subject) has not yet been completely set on ‘genetic rails’: conjunctural tactics, improvisations on the fly can be associated here with innate codings and genetically conditioned learning.
These last two examples seem already to indicate to us certain correlations between, on the one hand:
•assemblages that open up the Umwelt of the male to the female (courtship ritual);
•assemblages that demarcate a territory for a couple and furnish a protected space for their young;
And on the other hand:
•the territorialisation of the machinic indices of the grass stem offering; and
•a certain ‘disposition’ towards gregariousness.
With the Troglodytidae, the wren family, which constitutes one of the least sociable of the sparrows (although when it is very cold a dozen of them will gather together to keep warm), the activity of demarcating a territory brings into play what Paul Geroudet calls a ‘music box refrain’, that is to say a highly formulaic chant, aimed as a constant warning to possible intruders. After having taken possession of his territory, the male builds up to a dozen nests. When a female arrives in the vicinity, he lowers the intensity of his song, which is then reduced to a mere trill. ‘He goes to a high point in front of one of his nests, sings and puffs himself up, lets his outspread wings hang and shakes his outstretched tail, then returns to his nest, and sings whilst looking out from it, leaves and returns several times in a row. The invitation is clear: if the female consents, she answers with a little cry, bobbing jerkily several times, and finishes by inspecting the nest.’7 It was necessary to cite Paul Geroudet’s description here in its entirety to shows the richness of the semiotic interactions of this courtship assemblage, which, it will have been noticed, doesn’t include a grass stem component. We haven’t ‘yet’ arrived at the mimicking of the building of a nest, but only at the presentation of a nest that has already been built. The courtship and territorialisation assemblages remain autonomous of one another. But what it seems we must retain from this example is the role of the passage component of the refrain, and this for two reasons. In effect, we see here that it participates in two successive functions and in so doing perhaps ‘announces’ a supplementary degree of deterritorialisation that leads to a more pronounced autonomisation of the vocal semiotic and to its more individuated subjective internalisation.
In a general fashion the chaffinch is considered to occupy a special place in the finch family. They bring together species that are relatively the most ‘territorial’ in this family. Unlike other finches – canaries, bullfinches, etc. – chaffinches only live in groups for a part of the year: during mating the territorialisation component becomes autonomous and dominates the sociability component. Curiously, it seems that the male chaffinch defends his territory all the more ferociously for abandoning himself to limitless gregariousness outside of this assemblage of sexual territorialisation. The Australian finches studied by K. Immelman and M.F. Hall allow the evolution of the grass stem ritual to be followed across vestigial behaviours that are fixed amongst a whole range of species and constitute in some way a series of ‘living fossils’:
•Males of the genus Bathilda and Aejintha cannot court females without actually having a piece of straw in their beak. But by contrast they only mimic the construction of the nest.
•Same scenario for the genus Neochmia, but the male uses a material that is different to what he will use when building the nest. The semiotisation of the grass stem has therefore become autonomous.
•The male of the genus Aidemosyne only uses a grass stem in the initial phases of the courtship.
•With the genus Lonchura it is only prior to deciding to court that a grass stem is sometimes carried.
•The male of the genus Emblema only pecks at grass stems but doesn’t use them.
•Courtship with grass stems only occasionally appears amongst the genus Poephila, especially young males.
What particularly interests us about the evolution of Australian finches is that parallel to a deterritorialisation that makes the grass stem more and more symbolic, even ending up in its disappearance, one witnesses the emergence of a new kind of refrain. Thus the phylogenetic articulation of the visual semiotics of the grass stem with the sonorous semiotisation of the courtship refrain is indicated. In this regard, Eibl-Eibesfeldt writes that ‘[c]arting nesting material for nest building evolved into the male courtship actions using grass stems. This was again secondarily reduced in some species and became rudimentary, while at the same time the song, which originally served the function of staking out territories, also underwent a change in function. These animals are gregarious and hardly territorial. Instead of courting with grass stems, these males sing softly while sitting next to the females.’8
In the previous chapter we insisted on the fact that the ‘matters of expression’ put to work by the assemblages did not simply play the role of something that ‘fills’ semiotic forms or the ‘channel’ of transmission, in the information theoretical sense. They participate actively, according to all sorts of modalities, in modelling, in catalyses, ‘choices of rhythms’, stratifications, lines of flight … They are ‘inhabited’ by abstract machines that ‘opt for’ one connection rather than another. In short, when we talk about the components of an assemblage, what is in play is not just forms and quantities of information or differentiations, but also irreducible material traits such as the ‘viscosity’ of a transmission channel, the rhythms, inertia, the black holes that are proper to a biological, social, or machinic stratum, etc. As soon as one tries to take the point of view of machinic assemblages, of formative assemblages, the brute opposition form-amorphous matter has to be abandoned to the profit of a deterritorialisation that works forms as well as matter, deterritorialising forms and deforming matters. Certainly one can always account for the quantity of movement and the translation of forms on the basis of ‘purified’ spatio-temporal coordinates. But taking into consideration the intensity, the mutations, of regimes of deterritorialisation implies the intervention of other ‘existential’ coordinates, which one might call the coordinates of substance. What characterises components of passage like faciality and refrains is that they work within both norm and deterritorialisation: that is how they allow for the passage from one assemblage to another. They do not belong to space and time ‘in general’, they effectuate particular spaces and times. Let’s go back to our last examples concerning bird silhouetting traits and refrains: because of the ‘material’ characteristics that are proper to them, one can see that these components, which nevertheless sometimes have the same sort of function – in courtship rituals, for example – do not entertain the same kind of relationship with the deterritorialisation that traverses them both. Silhouetting traits are, in some way, ‘carried off’ by a phylogenetic deterritorialisation that bears on nesting behaviours and grass stem rituals, and they subsequently tend to efface themselves to the profit of an indexical semiotisation which is integrated into other semiotic components (dance, posture, etc.). In sum, the effect of deterritorialisation is to dissolve them as an autonomous assemblage, an assemblage that was, at the outset, rather plastic, ‘sticking’ to the territorialities of the species that it concerned, thus putting into play highly heterogeneous (morphological, iconic, mimetic, postural, etc.) components, highly varied ‘tools’ and procedures (grass stems, twigs, mosses, fish, etc.).9 The situation is very different with the birdsong component. It is also ‘originally’ territorial but the more it is deterritorialised, the more refined, specialised and autonomous it becomes. It ends up playing a very particular role in processes of evolutionary selection, as it can be considered that amongst certain sparrows, for example, the consequence of the existence of different ‘dialects’ has been an ‘ethological isolation’ of different populations and the division of certain species.10 Besides articulating intra-species refrains – centred on the territory or on courtship, the ‘catalytic’ behavioural function of birdsong can also return in a much less specific system of warning cries. When birds of prey hover over them, for example, finches will emit cries that resembles those of other species of bird, trait for trait, who will, if they are in the area will not fail to make use of the information.
The triggering of these relatively undifferentiated cries is highly progressive and it seems ‘conceived’ in such a way as not to allow the bird of prey to establish binaural comparisons helping it to locate the birds emitting the cries. The latter’s territorial song or courtship song, which are different for each species, because of the sharp variations of frequency that they put into play, are by contrast, easy to localise. The singing of the finches can thus play in two registers: one of alarm and territorial scrambling or of specification and localisation. But it also allows for combinations that make of it a sort of a-signifying language. But the song components can also enter much more elaborate rhizomatic combinations, which tend to function as a sort of signifying behavioural language. We have seen that in passing from territorial to courting behaviour, the wren could inflect its refrain – a lowering of intensity, reduction to a trill – this change of direction constituting a signalling and triggering system at the heart of the same component. We have also seen – in the phylogenetic order, this time – the refrain being substituted for the grass stem system amongst Australian finches. It thus seems that the most deterritorialised component – here that of the song – tends to impose itself at the heart of the rhizome of assemblages. Tinbergen’s description of the courtship behaviour of the albatross, the highly complex scenario of which is as if ‘crowned’ by a song component, seems to confirm this.11 Or equally that of Lorenz, for grey geese, where one also finds this same sort of ‘victory cry’ at the conclusion of their courtship ritual, marking the neutralisation of aggressive assemblages and the establishment of a ‘defence community’ at the level of the couple.12
The ritualisation of a behavioural assemblage is not synonymous with automation. A semiotisation can become machinic without for all that being mechanical. And all sorts of approximations, variants, lines of flight, black holes, always remain possible. We have evoked the [examens rates] of the wrens and the domestic scenes of storks but gratuitous acts such as the imitation of the song of the buzzard by the blue tit13 or the incredible chattering of the excited starling caricaturing – in the absence of a real talent for mimicry – the blackbird, the oriole, and even farmyard animals.14 Without mentioning the well-known exhibitionism of the nightingale, which leads to it taking the risk of exposing itself 5 or 6 metres from the ground so as to be sure its extraordinary vocal performance has the maximum impact.15
But nor is this ritualisation synonymous with a release of or a break with more ‘determinist’ components, even in the case in which super-deterritorialised components like that of birdsong are brought to the fore (and for humans, that of speech and religious rituals). Let us borrow a few more examples from ethology to illustrate this dependency, or rather this system of rhizomatic interrelations between components. Let’s come back to our first example, the zebra finch, who, it will be recalled, combined a ‘grass stem’ component and a ‘return to childhood’ component in its courtship ritual. To assemble its territory, it also uses two other semiotic components, to keep other males at a distance: one that is visual – highly coloured plumage,16 and one that is sonorous – a stereotyped refrain. Young zebra finches acquire this refrain by learning with their congeners. But if one amongst them is raised in a family of white-rumped munia (known in aviculture as the striated finch), it will learn the song of its foster father.17 Let us note that this learning is carried out during what is called a ‘sensitive’ period, long before the young bird is in a position to sing effectively. One must therefore distinguish between a purely auditory phase of semiotisation (through imprinting) and a phase of active phonic semiotisation. Additionally, ‘behind’ these two components biological components of an entirely different kind appear, as is shown by the fact that a female zebra finch, who doesn’t ‘normally’ have a territorial song, acquires one when given male sexual hormones. Obviously, she only reproduces the song of the species with which she has been imprinted during the thirty-five day ‘sensitive period’ of her life.18
That a component like the refrain is more deterritorialised than the others doesn’t in the least imply that it has taken its distance from more ‘determinist’ components like those of learning, imprinting, or endocrine transformations. And perhaps one is justified in expecting that the more a component is deterritorialised, the more closely it ‘meshes’ with more molecular components of behaviour and life itself. There is no doubt, for example, that for man, linguistic semiotics, in parallel with their magic conjuration function and of social subjection, have assembled a ‘omnipotence’ over his behaviour, his environment and numerous living species, of a new kind for him. And the supplementary degrees of deterritorialisation that the successive phases of deterritorialisation that the taking off of a ‘mecanosphere’ from the biological, linguistic and social order represent have taken on such an importance that without them, the survival of man would be inconceivable. (On a biological plane in particular, industrial man only ‘maintains himself’ by his capacity to discernabilise, semiotise, to artificially diagrammatise the pathological agents that threaten him.) But how are things at the relatively elementary level we placed ourselves at, with a semiotic component like that of refrains amongst birds? We cannot insist enough on the fact that even in such a domain the relations that are established between biological and semiotic components do not function in one direction only. One can better understand the complexity of this kind of relation by examining a graph such as that proposed by R. Hinde19 to describe the interactions between different factors that intervene in the reproductive cycle of the canary, which put into play:
•physical components such as the length of the day and the degree of light;
•biological and morphological component, production of hormones, growth of the gonads, development of the brood patch and oviduct;
•perceptual components, iconic stimuli emitted by the image of the male and his changes of posture;
•behavioural assemblages that are individuated such as egg laying, and social, such as courtship, nesting, etc.
In four points, the author thus makes explicit the ‘principles’ that regulate incontestably rhizomatic relations:
1The causes and consequences of sexual behaviour are strictly linked to those of nest construction and one cannot consider them separately.
2External stimuli (the male, the nest) create endocrine modifications the effects of which are added to these factors.
3Hormone production is governed by diverse checks.
4Hormones have multiple effects.
The distinctions that we have been led to establish within behavioural rhizomes, between assemblages of semiotisation and semiotic or coding components, are entirely relative and do not imply any priority of one over another, no a priori hierarchy. Certain assemblages can be stratified, automated and ranked as components in another assemblage, whilst certain components can start to ‘bud’ and to produce new assemblages. In addition, certain hyper-stratifications can entail zones of semiotic collapse, black holes which, in turn, will generate super-deterritorialised lines of flight (example: the explosion of ‘Eternal Russia’ between 1905 and 1917).
The connections between assemblages and the components of a rhizome thus do not necessarily respect the existence of layers that would be staged in a pre-established order – the order of deterritorialisations between the ‘physical’, ‘chemical’, ‘biological’ and ‘semiotic’ … There are certain ‘transversals’ that thus connect the ‘more social’ to the ‘more biological’ or to the ‘more ecological’ in the animal order. But isn’t this rhizomatic organisation ‘doubled’ by a less visible hierarchy which, this time, no longer concerns the assemblages and components, but the very texture of these latter, what (following glossematicians) we have called the traits of matters of expression and of coding? In this regard, one might consider that social faciality, which we have classified amongst Collective micro-equipment and to which it pertains to manifest the demarcations of power between the ‘acceptable’ and the ‘licit’ and which is charged with globally memorising the ‘graphs’ of binary choice borne by the dominant significations,20 in fact rests on the innate faciality traits that ethologists are currently studying.21 In another order of ideas, one might consider that the two kinds of memory that have been brought to light by psycho-physiologists – short-term memory, which capitalises information for a period of a few tenths of a second, and long-term memory – are entirely tributary to sensory memory, which retains information for two-tenths to three-tenths of a second. But to what extent does this molecular memory not also depend on the more molar memories, which seem to rely on them? The scientistic refusal to admit that the most deterritorialised of existents, such as faciality, refrains, ideational processes, abstract machines, are also just as real, just as closely ‘meshed’ with reality as the visibly material processes results in an a priori privileging of systems of linear causality and dualisms that go from the chemical towards life, from matter towards mind, etc. If they have any reality at all, one certainly cannot doubt that faciality and refrain components have something to do with the brain. One could even ‘localise’ them in an approximate way, along with other, globally visual and tactile memory components, in the left anterior part of the temporal lobe, in ‘opposition’ to the discursive memory components that intervene in language, ‘localised’ in the right-hand side of this same lobe.22 But, on the other hand, the inverse hypothesis, that faciality components, musical components can also intervene in the body, modify the brain, transform metabolisms would appear rather unscientific. And yet it is probably in this direction that ethological research will lead when it has finished with its infantile disorders (taxinomism, reflexologism, behaviourism, vitalism, etc.). We always come back to this same question: what makes assemblages and their heterogeneous components hold together? A transcendent hierarchy of spatio-temporal forms, a propping of physico-chemical effects, or the contingent montage of certain components which ‘take on’ specialised transcoding and deterritorialising functions (which we have been calling ‘components of passage’ or ‘diagrammatic components’)?
In the background to the problem of the refrain, another problem, that of the synchronisation of biological rhythms is posed, which before resulting in a new science – chronobiology – has given rise to innumerable metaphysical developments. One of the founders of graphology, for example, Ludwig Klages, had been led to oppose a vital rhythm to more cultural cadences. He considered that the human alone was able to assemble elementary rhythms in free spatial and temporal cadences. ‘Life’, he wrote, ‘is expressed in rhythm: Mind, by contrast, forces the rhythmic impulse of life to bend to its law, by means of metric cadence.’23 But rather seeking to ‘attach’ trans-rhythmicity to mind and to culture, chronobiology endeavours, on the contrary, to derive it from a rhythmic composition with a molecular base. Thus it currently considers that circadian rhythms24 result from a generalised coupling of what A. Reinberg calls a population of molecular oscillators, with an inhibiting effect.25 It is interesting to find here the same method of research into ‘molecular packs’ as we have signalled with regards to memory.
This ‘logic of packs’ certainly ought to help us escape from formal categories such as Life, Mind, Matter, but will it for all that allow us to progress with a problem such as that posed by Klages with regard to the articulation between vital rhythms and more complex ‘cadences’? The fact that heterogeneous systems are ‘traversed’ by the same kind of molecular element – infra-biological molecular rhythms, for example – indicates to us that between them there exist systems of articulation ‘from the inside’, as it were, but it doesn’t succeed in enlightening us on what makes qualitative differences crystallise at a molar level, or on what characterises the functioning of what we have called components of passage. To illustrate this kind of difficulty, one last example from the ethology of birds. In the course of his study of the chaffinch refrain, W.H. Thorpe has been led to distinguish two types of rhythmic and melodic level in its internal organisation: one which concerns a certain ‘finish’ in its structure, which allows the song to be differentiated into three strophes, and to be articulated according to a given order (true song).26 But as we will see, this distinction is far from intersecting with that of Klages, between elementary vital rhythms and socialised cadences! In effect, the basic material here is already highly elaborated on the ‘musical’ plane and it is, in addition, impossible to clearly distinguish between what would arise from hereditary programming and what would arise from social programming. Raised in isolation, young finches spontaneously discover the number and length of the basic syllables, but they also have available to them a sort of ‘recipe’ for learning or, more precisely, as Thorpe emphasises, selecting the melodies that they have to imitate. (If one gives them several different song recordings during their sensitive period, they will retain ‘those which, by the quality of tone and the form of the strophes, resemble the typical song of their species’.) Let us also signal that one part is also left for improvisation and competition, since, as W.H. Thorpe remarks, the details of the final phrase, with their fioritura, are apparently not learned, but ‘worked’ with other members of the group (‘worked out by competitive singing’). The diagrammatism of codings is manifested here by this constant entangling of heredity, learning, experimentation, and improvisation. And by means of this example, one may notice that what ‘passes’ from one domain to the other, are not just the basic materials or universal schema, but highly differentiated forms, sorts of singular keys for opening and closing a territory or a species, which we have proposed calling abstract machines.
It shows us that the stage of an analytic, quantitative and statistical study of basic elements – of rhythm or faciality, for example – ought necessarily to be followed by a more qualitative stage, of the specification of assemblages and, correlatively, of the definition of machinic ‘procedures’ resulting in changes of form and mutations of structure. Having brought to light the back and forth of packs of molecules and signs that link together a set of chemical, biological, ecological, technical, economic components on the same machinic phylum, it remains for molecular analysis to determine the paths by means of which a living and social thing selects, assembles, and normatises the circuits and rhythms of these packs. But if it is true that the essence of the living being ‘sticks’ at one and the same time to ‘matter’ and to the ‘semiotic’, it will then necessarily be from the first moment – of the homogenisation of intensive molecular fields – that the question of the ‘reconstitution’ of spatio-temporal localisations, totalisations and stratifications will be posed. If the ‘molecular machine’ does not wish to crush and reduce all the material and semiotic rough edges in an undifferentiated continuum (the Cartesian res extensa27) parallel to the rhizomatic connections of flows and the generalised intersection of assemblages, it will have in effect to bring to light the kind of interaction that ensures that ‘there will be’ assemblage. We rediscover here a problem that is similar to the one that we evoked in the second part of this book, when we were led to relativise the distinction between ‘generations’ and ‘transformations’ in the semiotic domain, because in the last analysis it is indeed a question of the same micropolitical ‘optional matter’ it is a question of. One might even consider that the semiotic relation transformation/generation is only a particular case of the molecular/molar relation that is established at the level of the ensemble of what we have called ‘machinic propositions’. Interactions between the molar and the molecular are constant but they result from assemblages that in certain cases majorise the ‘power’ of visible passage components to the molar state and, in other cases, ‘invisible’ molecular processes.
Whatever the structuralist efforts to overcome the ancient separation between the psychic and the somatic might be (from Goldstein’s ‘structure of the organism’, Merleau-Ponty’s ‘structure of behaviour’ to Lacan’s ‘symbolic structure’ …), and to articulate what von Weizsaecker called the ‘ontic’ and the ‘pathic’ in life, might be, they have been given ‘weight’ by the epistemological models of classical physics.28 They have considered that maintaining and even accentuating an opposition between, on the one hand the laws of matter and on the other, those of life, the mind and the socius, goes without saying. As material assemblages, assemblages of biological encoding, assemblages of enunciation, propositional assemblages, etc., constitute phenomenally distinct worlds, they have refused to venture into what for them would only have been a return to an outdated metaphysics, that is to say, the exploration of the ‘machinics’ that crosses through all these ‘regions’ of experience. Every system of enclosure, of the looping back onto themselves of physico-chemical laws and causalities in parallel prohibits any genuine opening up of the organism, the socius or the signifier onto reality. To our mind that is where the fundamental impotence of structuralist theories and their political responsibility resides: they accept much too easily the stratifications that they come up against in the order of components of material, biological, and social encodings. There is no escaping the primacy of a subjectivisation and an assemblage of enunciation that is based on a transcendental cogito with them.
But once the principle of an exceptional existential status has been accorded to this kind of subjectivity, it is subsequently not surprising that no inter-component diagrammatic connection can be established without being haunted by it in one way or another. The Subject, Form, Structure, the Signifier relay each other in contemporary thought, so as to resist an inanimate matter which has, in any case, become imaginary in relation to effective scientific research. By means of the celebrated formula ‘a signifier represents the subject for another signifier’, the hegemony of the Lacanian signifier tends to make subjectivity proliferate universally. But not no matter what subjectivity, only that of individuated enunciation, of signifying centring, of power over the self – the myth of mastery by symbolic castration, the subjectivity, in fact, which serves as a relay for capitalist power formations and their tentacular network of collective equipments. Now, the subject, we repeat, is evidently not something that exists solely where there are autonomous individuals, conscious language, a responsible discursivity … Precisely, it will be objected, psychoanalysis has clearly seen that the subject did not coincide with consciousness or with the exercise of responsible discursivity. But to make unconscious subjectivity, which it is additionally claimed is being liberated, essentially depend on speech and the field of language, really does mutilate it.29 There is subjectivity in the group – whether territorialised or not, there is subjectivity in the economy, at the stock exchange, for example, in politics, in factories. There are also subjectivation functions that are deployed in living matter and in machines, with or without human hands, with or without a cogito. And of course, it is not a matter each time of the same subject, who would miraculously make messages, decisions and laws pass from one component to another. A little subject in my head, like a minuscule manager on the top floor of a building! Processes of subjectivation correspond to complex assemblages, knots of deterritorialisation that associate heterogeneous components – and thus never a pure and universal signifying substance opposed to a no less pure and universal matter of content. The serial production and massive exporting of the white, conscious, adult, male subject, master of himself and of the universe, has always had as its correlate the chasing away of intensive multiplicities which essentially escape any centring, any arborescence. But once one has decided to abandon the model of the cogito or its derivatives as the implicit reference of assemblages of semiotisation, it becomes possible to discern the real play of the machinic indices, lines of deterritorialisation, abstract machine, the infinite diversity of modes of subjectivation, reflexivity, and discursivity.30 It ceases to be surprising that molecular packs and populations ‘claim’ to machine a creative order at their own level.
We must constantly guard against our conceptual instruments starting to function as simple blades that binarise objects and ‘arborise’ problems. Let us insist once again on the fact that the ‘molar’ must not be opposed here to the ‘molecular’ as the bigger and more passive would be opposed to the smaller and more active. There is a passive molar faciality – that of the imago and psychoanalytic identification – and an active molar faciality – that of schizo-analytic faciality traits. There is a ‘mechanical’ molecular faciality – that of ethology – and a molecular faciality that transmutes the coordinates of perception and desire – as described by Proust, for example, with the ten faces of Albertine, which get successively closer to the narrator, at the moment of their first kiss. But one can also pass from one component to another so as to safeguard an assemblage – further on we will examine the to and fro of Swann from a refrain to a faciality, for example. Besides, there are direct interactions, on ‘this side’ of closed assemblages and substantialised components, at the level of matters of expression. Thus whilst one can have the impression of remaining ‘in place’, of being established with a signification, a solid system of redundancies, one can be torn between warring components.
This is what is shown by the results of English researchers into the interferences between auditory and faciality components in spoken language, which they have brought to light by modifying a message read on someone’s lips in relation to what is given to be heard.31 In effect, encoding components and semiotic components do not properly belong to one of these levels of analysis. Under certain conditions, as a function of certain machinic formulae (abstract machines), certain amongst them can play an essential role in assemblages. Passage components, arising from refrains or facialities, can, for example, trigger new passional assemblages, make new components proliferate, block others, make black holes resonate and focalise their effects … The same components, under other circumstances, will fall back to the rank of intra-assemblage, subjugated, stratified components. In the same way as natural, territorialised assemblages or artificial technical assemblages, assemblages of semiotisation, assemblages of subjectivation, assemblages of conscientialisation, assemblages of ‘alterisation’ and so on, result from machinic montages that are localised over the ensemble of phyla of (semiotic and material) deterritorialisation, and territorialised on the rhizome of stratifications – on the plane of consistency of abstract machines having to make this diachrony of stratifications and this synchrony of deterritorialisations ‘hold together’. Thus one cannot pose the problem of the subject, or the Other, or consciousness, in general. This kind of assemblage will produce a black hole effect, the effect of a territorialised collective or individual subject, of subjection, etc. The cogito as empty consciential subjectivation corresponds to a black hole assemblage, a correlative semiotic laying bare of the growth of capitalist flows, whilst the subject of the Freudo-Lacanian unconscious marks a supplementary degree of semiotic deterritorialisation – monemes progressively giving way to phonemes, graphemes and ‘mathemes’. But other politics, other societies, other montages will assemble other subjectivations, other more social or more molecular semiotisations, or both at once, more ethological or more revolutionary, etc.
As has been seen, a flux of hormones can ‘trigger’ an unexpected competence in the matter of refrains, a flux of DNA can transform a memorisation process, or enlarge circadian rhythms. Intersections, marriages, that are apparently the most unexpected, the most ‘against nature’, always seem possible, but on condition that they are compatible with a set of machinic propositions, the montage of which, without being properly speaking universal, as it is ‘dated’ (because it marks irreversible choices on the phylum of deterritorialisations), nonetheless imposes on them a sort of ‘reality threshold’.32 One corollary of the contingency of abstract machines is that no type of molecular population, no universal rhythm, no energy equation, can account once and for all for the infinite variety of what one might call ‘assemblage convertors’. Certain amongst them will seem of an elementary simplicity – like the ‘magnetic effects’ of rhythms that E. von Holst describes, the effect of which is to impose one rhythm on others,33 others of a great complexity – like the human brain, which not only selects schemas and rhythms so as to ‘paradigmatise’ them on deterritorialised mental representations and on systems for inducing ‘passages to the act’ but additionally make them susceptible of entering into a combinatory of unlimited richness. Is this to say that the scale of complexity of these convertors is parallel to that of phylogenetic evolution? Not at all. In fact we know that at apparently the least differentiated, the least ‘evolved’ level, extremely sophisticated systems of interactions of heterogeneous34 components can exist, whereas inversely, at the most differentiated, the most ‘evolved’ level, mechanisms of a wretched poverty can appear – fascist gregariousness, for example. The elementary, the binary, feedback, black hole-abolition are not the property of one evolutionary stage. The elaboration of complex codings can borrow many other paths than those of individuated, conscious enunciation. Why not admit that a genetic knowledge exists? Why not admit that a machinic consciousness exists – for example in the case of the enslavement of the driver to his machine? Grass stems, refrains, faces for birds and for our passions – but for our intelligence too – are instruments of knowledge, pragmatic operators, in the same way that spoken or written words, figures, graphs, plans, equations or informatic memories can be in a factory. Once one wants to grasp them outside of the dominant redundancies, the signification of the world, the sense of desire, demand that one broaden the range of the semiotics we resort to. A thousand machinic propositions constantly work over every individual, over and underneath his speaking head.35 If we place the accent on faciality and the refrain in the components of the passage of human desire it is because one of their principal specificities is in some way to take other components ‘against the grain’ by short-circuiting their rhizomatic connections, by recentring them on black hole effects, by making the latter echo one another.
In as much as a certain abstract perception of time and space and, as a consequence, of work and the socius, rest on the establishing of these two components, essential components of capitalist subjectivation depend on the prior emptying of intensities of desire (the values of desire) of their substances and the prior reduction and gridding of rough edges of the world as a function of dominant redundancies and norms (the use-value-exchange value couple). We have tried to show elsewhere how, in order to explore the desiring coordinates of a new type of bureaucratic capitalism, an author like Kafka was led to becomings-animal, musical, perceptual deterritorialisations, etc. On the basis of the numerous pages of prodigious analytic work that Proust’s oeuvre constitutes, we now propose to examine the impact of certain capitalist mutations at the start of the twentieth century on an amorous passion, that of Charles Swann for Odette de Crécy.