Memory assemblages

My talk here at Burque last winter

I want to start by thanking you all and acknowledging the department of philosophy, the University of New Mexico and this land, as a visitor coming from the south of the border and from the land of many Macroje peoples who themselves live in a way that is constantly informed by memory, immortality and their ancestors, I strive to learn more about the Tiwas, the Sandia peoples and other indigenous communities of the area. I keep finding myself trying to find their marks around – and they seem quite well hidden. For reasons to do with this very talk, I welcome the gesture of directing our thoughts to the land where we are; both as an indication of our situated character and as an archive of the past which carries a proliferation of promises for the future.

In this talk, I will try to elaborate and recommend the idea of memory assemblage, a central notion in my current project around specters and addition. I begin by saying that I claim that I find those assemblages in many different places and forms, but I would only do that with a measure of care. The bite of the thought about memory assemblages is that things are not simply there, they are not kept or retained as they are by themselves. A first contrast to be made is with the very notion of presence that invokes the self-standing and self-enduring character of ousia (translated also as substance) and the present time that is associated with the disclosure done once and for all of a landscape of actualitates in reference to which what is real can be (at least in principle) detected. Memory assemblages are not present in this sense, they are rather composed of a diachronic coupling between the time when something is retained and the time when what has been in retention is retrieved. The difference between a presence and a memory assemblage is decisive. Presences are full contents that are the way they are and, if they can be at least partly accessed, their contents can be extracted and they can be simulated in an artificial instantiation and then, eventually, made redundant. This has been, in the reading of Heidegger I favor, the quest of metaphysics. Metaphysical thinking is some kind of hunting where what there is is the prey that once captured – and this can take several different attacks – is surrendered and placed under command. The world, facing metaphysical thinking, is, according to Heidegger, in danger – and it cannot do anything about this entrapment other than to try to escape. Escaping, to be sure, can take many paths, one of them being that of finding ways of confronting whatever sets the engine of metaphysics going. Once metaphysical thinking can be done by anything – it is oriented toward what Nietzsche posited as a Freigeist moved by the will to power – also humans are under its grip, both in the sense that we can be taken over by metaphysics and in the sense that we can see no other way but to perform it. We have been thinking about ourselves in those terms: what makes us the way we are is a landscape of actualitates – material or immaterial, substantial or relational and that include also capacities and propensities – that is eventually going to be properly disclosed. In contrast, memory assemblages with their diachronic poles are not contents that can be made transparent once and for all – it is only with respect to a situated retrieval operation that something like a content emerges. As with our ordinary employment of memory, what is recalled is later set in oblivion again, not being transparent for good. The (metaphysical) effort most common is that of trying to explain features of memory like this one by looking for suitable underlying presences that can do the trick. My approach can be perhaps described as the opposite one, the one of taking memory to be what makes possible the effects of presence altogether. In a memory assemblage, retention cannot do the trick of determining any content on its own, the pole of retrieval – that is often triggered and oriented by something that is added to the landscape.

Notice that other bifurcating strategies have been attempted to deal with the multiple problems with metaphysics – like distinguishing between phenomena and transcendental subjects or, more generally, promoting a bifurcation of nature and experience. While these bifurcating strategies acknowledged that the content we extract from things is not entirely there, they continued to posit full-blown presences like the subject, experience or nature. Each of those could be made transparent, and once gripped, replicated artificially. Memory assemblage proposes a diachronic bifurcation; this is to say, a bifurcation in time, the past that archives something signaling to its future and the future that will recuperate it while accessing its past. These diachronic poles introduce an anachronism common both to thought and to its objects. Importantly, it also introduces the situated element that addends coming from what is exterior (and a posteriori) provide. Thought, as any event, is hostage to these addends that will not only shape retrieval, thus forging a memory assemblage, but also promote remembering and forgetting. Instead of picturing thought as an extraction of intelligible content from things, as a network of memory assemblages thought retains with the help of the traces of the past in the world and retrieves prompted by addends from outside.

I take memory assemblages to be an ultrametaphysical (or postmetaphysical) notion and, in this sense, it is perhaps not even properly an idea – it is rather a deconstructive gesture, a preparatory exercise or a template to conjure other pathways in thinking. By ultrametaphysics I mean the effort to move past metaphysics, to sideline or exorcize its presuppositions, to explore its possible aftermath. The endeavor, I take, was inaugurated by Heidegger who faced metaphysics as an epoch in a history of beyng, preceded by its fomenting and that could be succeeded by something that is under preparation since the endeavor exhausted itself with Hegel and Nietzsche. Heidegger himself rehearsed several paths for an ultrametaphysics, the most explicit perhaps being that of Ereignis and Lichtung where the event that arrives and takes place is not a disclosure of anything but simply to assert its right to appear. He contrasts the will to truth as correct depicture that presides over metaphysics with truth in a glance in which attends to the event only to immediately look away, always in the light of the event’s “own illumination” which keeps concealed where it comes from. Heidegger reckoned that the exhaustion of the metaphysical project stems from its very disclosure as will to power. That exhaustion, for him, ushered in another moment for thought and at the same time shows that metaphysics is itself an event that illuminates only itself – as such, metaphysics has not the final word about thought. But it is not simple to look for other words.

In any case, ultrametaphysics continued in different terms by philosophers such as Derrida and Malabou, from whom I borrow the phrase ‘ultrametaphysics’. While Derrida looked into ways to deconstruct presence by examining the anachronic elements that engenders it, Malabou proceeded to shed light on the plasticity of any intelligibility. Both made clear that the content extracted from things cannot stand alone – it is like a text that needs to be read and a lore in the hands of the heirs or a code that requires an environment to be expressed. Ultrametaphysics aims at forging a future for thinking after metaphysics and beyond its presuppositions while contemplating them from the outside. If for Heidegger metaphysics has terminated because it has been exhausted, for Derrida that simply worked as an invite to deconstruct it in an infinite task; as it is not clear that thought would ever be rid of it.

I see memory assemblages, thus, constituting an ultrametaphysical proposition. Its core is that memory is entangled with addends in a way that only while the coupling is at play there could be statements concerning how things have been. These couplings last only until the next addend comes into the picture. Each of these assemblages retrieves something that could fail to be in continuity with what has emerged before - like a new reading of a book can make something different of it. As a replacement for presence, memory assemblages focus on the past – but not the past of actualitates that are ready once and for all, like the historicist would maintain, according to Walter Benjamin. The past here requires retrieval – it needs an exercise of recuperation and these exercises change as the future comes along with different addends. Memory is taken to be a widespread phenomenon, it is found in the human mind, but also in archives, texts, genomes, tree rings, rocks, landscapes, traditions and natural laws. A land acknowledgment like the one I did some minutes ago is perhaps a memory assemblage in embryo: it states that in the land there is something in retention. In any case, the ultrametaphysics I propose looks at memories and archives as a point of departure for a more general approach to what the world is. It is not a metaphysics in the sense that it cannot offer a landscape of actualitates, it doesn’t deal in presences. It is, rather, ultrametaphysical. It is not a speculative move from something more known to something less known through a mirror projection because there is nothing to be known once and for all. It is, as I will claim in a while, ultraspeculative.

Before that, I’ll make further brief remarks about some features of memory assemblages. The entanglement of memory and addition is such that they activate each other to the point that none can subsist on their own. Addends can come from anywhere to compose memory assemblages. They also destitute memory assemblages and forge others. The issue of how to individuate them could then seem pressing as addends keep coming by, new couplings are made and we should then rather talk about networks of successive memory assemblages. The task of individuating them is difficult, but it only makes sense in relation to a specific memory assemblage (or a network of them). Individuation is situated, just as measurement is situated for Whitehead. For him, the selection of a congruence relation for measurement can only be done from the standing point of a location; it is only from a specific situation that the selection is suitable and appears as self-evident. It is from where we are that we measure the size of objects for our locomotion, it is only with respect to the traveling body that we can compare distances, it is for a purpose that we count books, cells or atoms. Analogously, the individual memory assemblage can only be identified within a memory assemblage. This is because memory assemblages form no landscape of actualitates. Memory assemblages, thus, cannot be viewed from outside, sideways-on, so to speak. There is no such thing as furniture of the universe to be exposed once and for all. This is why memory assemblages are ultrametaphysical – and we will see soon why they can only be ultraspeculative also.

Now, the philosophy of memory goes back at least to Aristotle who formulated the question of how to distinguish memory from imagination – arguably the ultimate hard problem for the area. I claim that Aristotle’s answer, while often seen as insufficient, is not off the mark: memory, he claims, “is of the past”. To say that the object of memory is the past is not to say that it captures the past as a presence or even that the past is there to be grasped once and for all. To be sure, imagination itself can have the past as its object and the past could prove to be oblivious to the efforts of memory and imagination. This perhaps indicates that imagination is often memory-involving and that no memory assemblage could redeem the past on its own – as a text is never read once and for all, as Derrida puts it, reading is spending while saving. Further, the past is not unaffected by the moths of the outside. If the efforts of memory are oriented toward the past, they are oriented toward forging a memory assemblage. The past is itself situated, it is indexical, which enables me to say that the philosophy of time cannot be but a department of the philosophy of memory. This, in turn, indicates that retrieval is always open-ended and that forgetting something once and for all is as impossible as retrieving something once and for all. The distinction between remembering and imagining is not a clear-cut one: we sometimes believe we are imagining what we are remembering and doubt our

remembrances wondering if they have been invented. The problem is itself a consequence of the entanglement of retention and retrieval: what is consigned to memory is orphan and retrieval is to a large extent beyond prediction. Attempts at solving the problem – and drawing a line distinguishing remembering and imagining – involve full-blown skepticism concerning memory, not far away from the modern-day constructivism in the philosophy of memory, and the opposite view that there is no fiction but only recollection from a long past that could involve an eternal return. A better approach could be the dissolution of the problem: while we cannot eliminate either of the poles, there is no general and principled line we can draw between them. Such line would entail separating out pure retention – which would be memory itself and the object of remembering – and pure retrieval – which could have no hold on what is in retention – which would constitute mere imagining. Rather, there is an amalgam between remembering and imagining because imagining triggers remembering; while arguably the opposite is also the case. It is only at a given time – and this with respect to some memory assemblages - that the separation between memory and imagination would make sense. To be sure, often we can have an assemblage that brings together different archives placed in retention and those decide what is, say, both in the stories being told and in the archaeological data. Evidence, conceived in terms of memory assemblages, is about what is in retention, always according to a retrieval operation.

Bergson has to be credited for placing together memory and duration. But he confined the importance of retention to the human spirit, and his recoil from the metaphysics of presence is therefore limited. Further, he associated active recollection with consciousness, making memory hostage of a presence. Still, he paved the way for the idea that perception, as much as thought, is memory-involving. Now, if we claim that the two poles of retention and retrieval as entangled, we are committed to the seemingly correlationist thesis that one cannot either think or perceive one without the other. But this is a diachronic correlationism that ensures there could be no full-blown presence. It is, I argue, broader than the one between thought and the world – or between perception and things. If we further assume that the res vera in an episode of perception is an archive that is now read by the sense organs, we see how the poles of the diachronic correlation of retention and retrieval are broad enough to encompass at least some of the consequences of assuming a correlation between a fully present res vera on the one hand and a perceiver ready to articulate, coordinate or conceptualize. That we cannot know or conceive things-in-themselves appears then as an unhappy formulation of the claim that retention without retrieval is blind (while, surely, retrieval without retention is empty).

It could help to consider memory assemblages in the context of contemporary debates in the philosophy of memory. I’ll limit myself here to a single comment. The neurobiology of memory is experiencing an engram renaissance – a renaissance of the idea of Richard Semon that what humans retain is kept somewhere in their body. Engrams are now sought among neural structures and the issue is whether what is stored there could be something like full-blown self-standing propositional contents. In the last few years, the idea of memory traces has received a lot of attention – they appear as attempts to describe what is in retention as something that cannot be apt to be true or false on their own. Now, from the point of view of memory assemblages, it could be useful to model the engram, or the memory trace, as an archive that places in retention what can only become a content when retrieved. This is a point in which ultrametaphysics, especially the one influenced by Derrida’s deconstruction, can help neurobiological research. The starting point would be to think of neuronal structures as texts that always require a reading, or as genetic code read into proteins with the help of environmental variables. Retrieval – and indeed addition – is always expected.

The notion of memory assemblage is at odds with the metaphysical assumption that things could be in principle disclosed and made transparent. Heidegger found in physis the starting point of a saga that would engender metaphysics. The physis of something makes it appear, unveiling and concealing itself, and truth within the thought of physis (aletheia) is what ensures the right of anything to withdraw. Memory assemblages are not thought of in these terms because physis itself requires no engagement with what is exterior to promote retrieval. In the protometaphysical regime of physis, there is no landscape of actualitates, but each thing retrieves itself of its own accord. There is no entanglement between memory and the coming of exterior addends. Perhaps we can say that while through physis we can say that the right of beings to be is respected, through memory assemblages it is the right of the exterior addends that is asserted. It is primarily not about a faithfulness to being (or beyng), it is above all about an attention to what lies outside it and can change what it is.

An important point about the memory assemblages is that whatever is archived is hostage to the future; there is an opening to exteriority, to the outdoors, to what is beyond being. Whatever has been could be part of a memory assemblage and, as a consequence, cannot be fully forgotten (or fully remembered) because the addends bring in new possibilities of retrieval (think of archaeological research but also of frozen germs becoming effective with climate change or the megafauna coming back to life through genetic, and epigenetic, research). There is a sense in which this vindicates some writings of Parmenides concerning the impossibility of lapsing into inexistence. If we place Parmenides not in the framework of physis, but rather in that of memory assemblages, we can understand his remarks about being – that it is finite and it cannot fail to be – as remarks about what is in retention. Being ought to have an outside because it is retention, it is always visited by addends from beyond it. Because nothing is fully forgotten, one cannot understand lapsing into inexistence. Emanuelle Severino has taken Parmenides to be claiming that being is eternal, and whatever there is will not stop being, it is placed in retention somewhere. His is an ontotheological reading assuming presences – retention without retrieval. But this can be the beginning of a reading according to which what is permanent is only retention which is always hostage to retrieval.

I’ll close by saying something about the ultraspeculative, as I indicated. The picture of ubiquitous memory assemblages contrasts the metaphysics of presence. It posits that nothing subsists without a situated retrieval, a saying of the said, so to speak. It claims that the past is not composed of self-standing, indifferent, separate, staccato units that can dispense the anachronic action of memory. Memory assemblages straightforwardly break with the metaphysics of presence by offering an alternative to the sufficiency and completeness both of what is present and of presences – they cannot stand alone, they are not fully there. Instead of presences in the present, memory assemblages are retrievals prompted by addends that stir up what was incompletely (and insufficiently) placed in retention. Now, these claims can be seen as the result of a speculative argument that moves from the workings of some specific memory assemblages – say the human, animal or geological memory – to a general image where instead of presences the main intertwined protagonists are memory and addition. The problem with this speculative argument is that it goes from a found point of departure – that of memory assemblages – to something close to a landscape of actualitates. The resulting picture ends up betraying the central features of a memory assemblage because those memory assemblages that compose this open totality seem to be in retention without requiring retrieval. The speculative move takes us towards a complete picture, and that completeness is what precludes retrieval.

In Indexicalism, I addressed a similar friction by proposing that the emerging picture was paradoxical. There I argued that the furniture of the world is deictic, best described by expressions like ‘here’, ‘there’, ‘me’, ‘you’, ‘now’, ‘here’, ‘other’, ‘beyond’, ‘past’ etc. The situated picture, if made complete, forms a landscape that is not itself situated. At the time I thought that swallowing the paradox was the price to be paid to be faithful to exteriority. But I think another way out is possible. Instead of thinking of a landscape of memory assemblages, we can think of a memory assemblage of memory assemblages. In other words, the move is an ultraspeculative one that deconstructs whatever resulting general landscape that the speculative move could generate.¹ There is no general thesis about memory assemblages, rather what is proposed is itself like a memory assemblage. With this move, exteriority is also not exorcized by a totality. The conclusion cannot be one that is itself invulnerable to addition. It is as if the step was from some assemblages to another, broader – and this broadening is a result of addition. The broader picture cannot be achieved unless it is itself, as much as what it is about, a memory assemblage.

Memory assemblages are not a ladder towards actualitates that would then enable a view of everything without any memory assemblage. It is ultraspeculative because the inclusion of addends in reality transforms whatever was conceived as being real. We can consider this ultraspeculative effort as a form of storytelling – and take the metaphysics of presence as another form of storytelling, one where an epilogue could be reached where every plot and every character are fully disclosed. The ultraspeculative and ultrametaphysical story would not be one where a moral can be revealed and the ultimate significance brought to light by the end of the story. It is a story of addends that come in and retrieve what was previously lost, losing track of what was previously made explicit. But the story provides no immunity against addition. Rather, a story where the beginning is always retold because it is seen from different lights at each episode, at each retrieval. A story where no part of the plot is eventually fully explained and then redeemed from any obscurity. Rather, the story is never fully told. Not even its beginning is fully told. A non-ending story.

1Compare with the use of the term ‘ultratranscendental’ in Derrida (Of Grammatology, pp. 58-59).