The Mesh We’re In

With a mildly ironic nod to its philosophical energies, I would say that this book is not only important but necessary. Giving it a notice is the least we can do. The Digital and Its Discontents is, however, a difficult book to review precisely because it is so clearly and concisely written, not only with the necessary words but also with engaging flourishes, and it all but reviews itself in what proves to be a brief initial introduction. Given an insightful foreword by Alexander Galloway and a more extensive introductory chapter, “Approaching the Digital,” readers are well prepared for the book’s careful expositions.

One reason why The Digital and Its Discontents is necessary (as the introduction explains) is that it there are few if any books quite like it. There are many discussions of specific digital technologies and once-new media, taking into account their cultural and sociopolitical impacts and implications. Instead, this book addresses the digital as such, philosophically. It allows us, clearly and directly, to consider what the digital is. What is its ontology? To this extent, the book sets aside certain questions of when and how the digital came to be, historically, but it does engage with the question of why, in that it helps to account for what Evens calls, at one point, its “stunning efficacy” (a more considered choice of adjective than might at first be apparent). As to the where of the digital, together with all of us in the developed world, the book assumes ubiquity, in the sense that devices housing and allowing us to provoke the operations of the digital are everywhere, in myriad forms of actual, real-world hardware. Evens’ question is, what do all these devices harbor? His answer is clear and relatively straightforward. These are, all of them, mechanisms for the representational logic of a single difference which we humans represent to ourselves as that between 0 and 1. This one difference can be variously exposed to the sixteen possible logic gates that determine one or the other output given any possible two inputs of one or the other value. The digital is an abstraction, any interpretation of which depends on representations that we give it when we decide, for example, to represent numbers or letters as sequences of the digital difference, or that certain sequences are commands to a mechanism, requesting it to move to some other point of address in a devices’ constitutive sequences and remove or inscribe other sequences. These are then its representational logics. They provide the digital, once commanded to ‘execute’ – since it would be otherwise inert – with an ability to move within itself and to alter its sequences and their extent during the time of execution.

That’s it. There is nothing else to the digital. It is (with Evens’ original emphasis) “discrete, mechanistic, and inert” (p. 26). Standing guard at its gates (in Evens’ own phrasing, ibid.), all computational hardware mechanisms – including Quantum computational mechanisms – are entirely devoted to preserving and literally perfecting this ontology of the digital. They are designed and constructed to remove, precisely, a crucial characteristic of our actual world: contingency, the term that Evens singles out in ontological contradistinction to digitality. ‘Actual’ and ‘contingent’ (the latter usually cited in its nominal form) are key words for Evens. The world in which we humans find ourselves is actual and all its constituents are emmeshed (‘mesh’ is another key to Evens’ thinking) with contingencies of all kinds. These contingencies are those which make our world seem and feel to us:— so rich, so filled with “meshworks” of interconnected meaning, substance, beauty, value, and evolutionary change – along with all their opposites—in dimensional continua—likewise enmeshed. By contrast, Evens, speaking in his more technically philosophical (if nonetheless value-laden) register, says that contingency is precisely what is banished from the digital, this “banishment” constituting its “defining deficit,” “ensuring its poverty, its formal senselessness, its flattening of value, its endless spinning in place. (p. 59)”

Returning to the where of the digital, Evens’ thinking locates it in another world, a world of the perfectly symbolic, either divorced from ours or connected to it by rather expensive hardware whose rule-based use may respond to perceived user needs and desires but is, ultimately, designed and governed by its makers. Who were and are these makers? They are – with contingencies of actual history taken into account – post-war scientists and engineers, often in the service of the war-like interests of what Friedrich Kittler conveniently calls “the upper echelons.” “The language of the upper echelons of leadership is always digital.”¹ The question of how, in real material terms, the digital connects with our world of contingency is not quite central to Evens’ book. He does not engage directly with the psychology or the politics that this might entail, providing us instead with a chapter, “From Bits to Interface”, which looks at how “bits [the digital itself] are employed in digital machines to connect those machines to the human world and make them useful (p. 9).” This chapter demonstrates how layers of digital representations are created, from the lowest—exclusively constituted by its bits’ single, purely abstract difference—up through the letter-sized layer of bytes, to those of words, until we pass through more layers of digital constructs that Evens theorizes through his particular conception of “icon.” The upper layers allow human users to grasp their constituents as the iconic handles and control levers of some instrumentally “useful” machine … or weapon.

Evens makes particular note of the isomorphism (and homoousia—same substance—although Evens seems to fold this sense into his use of isomorphism) of these layers, taking this to be a key, problematic characteristic of the digital. Unlike analogous structural layers in the actual world, where, for example, walls may not be regularly or exclusively built and encoded in terms of and in the same substance as bricks, “all of the digital layers are isomorphic, each tied deterministically to the others, such that there can be no substantive tension, no disagreement in these layers, and ultimately nothing available on one layer that isn’t strictly captured in the others (p. 148, original emphasis).” This is indeed an important property of digital ontology and structuration. It makes the human grasp of whatever is construable in an interface complicit with the regular, formalized manner in which this ‘grasp’ is translated ‘down’ into the lower layers of digital operation and also ‘up’ when digital control mechanisms are explicitly connected to physical—and only then contingent—effects in our world. Think of the drone operator and their grasp of the drone’s interface and how the digital works to remove any contingency down into, for example, trajectory computation and then up into the actual effects of the physical device. Compare this to a pilot and what they grasp and perform in the cockpit of their fighter. What value (or values) do we attach to the type of contingency-implicated control of the fighter pilot as opposed to that of the far more digitally formalized control of the drone operator? The answer is multifaceted and variously valanced. It includes the sense that digital control mechanisms tend to remove some of the ethical engagement—of both operator and those who give them their orders—because they have less contingency with which to contend. And, with respect to Evens’ particular address to these issues, one answer is that the more digital, less contingent mechanism is ‘better’ because it is more efficacious.

Such efficacity, “so profound and appealing,” is what Evens suggests is perhaps the chief reason that the digital “has become a significant component of human reality” (p. 6). The digital for Evens is, within its closed, abstracted world, the epitome of “an exemplary positivism, a devout rationalism, and a compliant instrumentalism” (p. 14, original emphasis). Otherwise inert, the digital can only proceed, in its representational logic, to do what it does in terms of posits, rationally, and with—ultimately, even if this is denied by human operators and their contexts—a pre-determined purpose. For the digital and its artifacts, these three terms—an assertion that the world is constituted by self-contained posits, a formal logic for their interrelation, and a functional teleology for their operations—are necessary and definitive. On the other hand, they resonate with humanity, according to Evens, as a post-Enlightenment ideology, an ostensibly scientific but actually scientistic, claim about the world and how it should be studied, understood, and controlled. The actual, for Evens, eludes and exceeds all three terms. But actual humans prefer to affect the world in which they dwell as if this were not the case, using the most efficacious tools that they have manufactured to do so. They root out contingency with the digital, with its “machine of representational possibility” (p. 115, original emphasis). Under capitalism, they design for the most profitable possible outcome.

This seems to me to be a fairly clear, insightful critical statement concerning the circumstances in which we now find ourselves. Evens proceeds with a penultimate chapter on ‘What Does the Digital Do to Us?’ summed up, perhaps, by his saying that it allows us to relinquish “the anxiety of responsible subjectivity” by encouraging us “to insert [ourselves] into a circuit that establishes and satisfies” our desire (p. 176, original emphasis). Our “insertion” into this “circuit” calls up a number of possible engagements with the digital that might take us a little bit further than Evens has chosen to go and, reflecting on his thinking, I’ll conclude by exploring a couple of these.

An adjacent theorist, N. Katherine Hayles, offers a more positive assessment of the digital in at least two ways.² For Hayles computational media are nonconsciously cognitive, and can form assemblages with human agencies. Much of Hayles’ recent work focuses on cognition— “a process that interprets information within contexts that connect it with meaning”— which she finds in everything that is animate or that is created and animated by, typically, human animals.³ Humans themselves, for Hayles, are governed more by their biological cognitive nonconscious than by their conscious cognition, and it is certainly possible to agree with Hayles that the human cognitive nonconscious is continuous with and no ‘better’ than its instantiations throughout the biological domain. But Hayles also distinguishes and acknowledges a technological cognitive nonconscious which may be either mechanical or digital (the latter necessarily interfaced with the mechanical). She proposes that we should explore if not embrace the sense of our human post-war history as moving toward one in which the human—never for Hayles distinct from non-human cognitive nonconsciousness—becomes actually integrated into ‘cognitive assemblages’ with cognitive nonconscious machines of its own manufacture. In the cognitive assemblage, there is contingency and creativity because the human is ‘always already’ integrated with cognitive nonconscious technics, and here Hayles can appeal to the kind of (Bernard Stieglerian) evolved technics that the human derives from, for example, bipedalism and opposed thumbs. Opposing this, Evens (and yours truly) would insist on the impossibility of any actual integration of the digital with human contingency, on the basis of their opposed ontologies. The digital can only connect with our world through our own adoption of or adaptation to some abstract representational logic that we ourselves design and encode. Hayles’ assemblages entail intercommunication between their chiefly cognitive nonconscious constituents as if these constituents were more or less commensurate. She usually characterizes this intercommunication in terms of feedback loops, crediting cognition—in the abstract and regardless of its biological or technological, its conscious or nonconscious instantiation—as what keeps the assemblage together and gives it interpretable meaning. She goes so far as to give a special priority to “computational media” as the quintessentially “cognitive technology [original emphasis]” and thus “able to impart to a huge variety of other technologies the advantages that cognition bestows—flexibility, adaptability, and evolvability [emphasis added].”⁴ In Evens’ (and yours truly’s) thinking, any contingency implied by these three terms cannot be represented, let alone “bestowed,” by the digital. Our association with the digital cannot be one of assemblage such as Hayles proposes, it can only ever be an instrumental relation, with the animated party either dominant or willfully enabling some kind of co-dependency.

In actual practice, representations and representational logics are made with what we usually refer to as a computer language. In this context, this word becomes, perhaps, what the more (Gilles) Deleuzian Evens might accept as a fold, where language as such—a human faculty that is actually, contingently integrated with the human over temporal scales that are both evolutionary and historical—meets and touches, on the basis of its own symbolic aspect, formal language’s timeless purely symbolic ontology. If the ‘language’ of Large Language Models (LLMs) is similarly ‘folded’ does this provide us humans with an actual connection to the digital? Evens resists this possibility or potential, although he would be ideally placed to explore it in future work. For now, he sees an LLM as “a database in the form of a neural network” (pp. 189-90). If any and all weighted, pseudo-randomness were denied to all of the dimensions of an LLM’s vector space after training, then this would be exactly what it is. The prompt-as-database-query would simply return whatever it returned, deterministically. Whatever we perceive as creative—or as varying in some contingent manner—in LLM output is, as Evens points out, simply a phenomenon of scaling not-so-mysterious linear algebra. Linguistic data points are fed into the models plotted along human-construable dimensions, such as absolute frequency of tokens, or frequency of token collocations (which can be calculated by the model) and then Deep Learning is programmed to generate billions more digital dimensions with data points which the model can use subsequently to inflect outputs. In all of this ‘training’ nothing actually new has been produced by the model. Introducing a little pseudo-randomness in any or all of the dimensions, whether derived from world-scraping or computed, produces nothing more than a simulation of contingency in actual human language. Whereas this contingency in actual human language is derived from—amongst other things—the embodied lived experience that concerns any human interlocutor in the temporally specific moments that they produce or receive traces of language. Most obviously, they must ‘evocalize’ these tokens, integrating them with whatever actual language they speak.⁵ So far, I believe, Evens’ account of the LLMs would accord with mine and he would go with me at least this far.

But the tokens of language that we use—speaking or hearing-understanding—are transcribed in forms that lend them a symbolic ideality, they seem to be identical with those with which a ‘chatting’ model transacts. The models’ symbolic, digital world seems to ‘fold’ onto ours at these points of ideality. Underlying this seductive folding, we may recall certain attempts to treat human psychic structure as symbolic with, for example, an unconscious (not a nonconscious, but then who’s to know?) that is “like a language.” Psychoanalysis-as-science seems to have moved into its own winter during our current AI-hyperSpring. Perhaps we should give it a warming thought.

Amongst media studies theorists who anticipated the effects of the digital, Kittler read Jacques Lacan as the only psychoanalytical theorist prepared to engage with the problems of a world governed by digital hardware and media.⁶ Lacan put forward the thought that the psyche was governed by commensurate structures. For Kittler, there was no software: the pilot would become a drone operator. For Lacan, the analyst had the formulas and must find a way to translate the linguistic ambiguities of Freud’s talking cure into the psyche’s symbolic order. But neither Lacan nor Kittler lost touch with the Real. Evens references Lacan’s “tripartite schema to describe the human world: the Real, the Imaginary, and the Symbolic” (p. 115, original emphasis), and maps it onto Evens’ own schema within which: the digital is the Symbolic; while, through the Imaginary, “the computer can say something about the actual world and can receive instruction from and offer instruction to people”; and “the Real corresponds well to the notion of contingency … [it] is the complement of the digital, that not-quite-something that makes everything what it is but hides from all scrutiny, marking the digital’s inevitable limit” (ibid.).

It is the Real, however elusive, that is of consequence for all of these thinkers. Lacan, in my reading, allowed us to imagine a tripartite schema but did so, “by distinguishing the symbolic from the imaginary in their relationship to the real.”⁷ His schema is more of a Real-related-two-ways triumvirate. The Imaginary can only ever consist in what actual humans imagine about the real world as they live and live in it. The Symbolic, on the other hand, turns mediated, softworn life itself, for Kittler, into machines (and weapons), since, for Lacan, it structures our neurotic or psychotic psychic (sint)homes. Their Symbolic was never in some other, abstracted digital world. This is expressed, in Lacan’s thinking, in a number of ways, including when he says, “The structure of the [psychoanalytic] symbol is that of a knot that one cannot flatten out” because this “privileged” symbol is created and defined by “the division that the signifier engenders in the subject.”⁸ Evens might agree that psychoanalytic symbols are thus knotted to contingency, the contingencies of embodied psychic structures. Yours truly would go on to say that the symbols of actual human linguistic practice—of language as such—are also knotted to the Real, whereas the symbolic tokens of the so-called language models are not, with Evens helping us to clarify this difference and its circumstances. The models are orthographic text models, poorly enriched with, at best, idiosyncratic usages and other(ed) miscellaneous linguistic data. Real language is, at the very least, enmeshed (and not separately tokenized) with its actual phonological performance and these enmeshed realities are also knotted to our psyches. Neither have we taken up the question of temporality with respect to language as such, or to Evens’ digital discontents, for that matter. For humans, the when of what is said is just as important as its what. How does the digital represent the kinds of time that are pertinent to human or animal life? These questions are not gone into in Evens’ book and there would be too much to say about this here.

Evens troubles and is troubled by the digital quite enough as it is. What will be the fallout from global infrastructural relations that are ontologically disjunctive and thus actually dysfunctional? While writing with what sometimes comes across as an evenhanded and dispassionate clarity, it is clear that Evens already laments our loss of contact with the contingencies of anything actual once it becomes represented by and subject to the digital and its logics. For many of us, more or less all cultural objects have reached this state, accessed only through digital interfaces. Cultural objects, in this sense, include those forms that mediate politics and the so-called public discourse that provides its media. These are now dominated by digitalized social media as has often been pointed out. It may be that a number of social democratic polities are experiencing, in actual recent history, the abyssal consequences of social media’s digital representational logics. People may no longer care to imagine what might follow from what they ‘say’ and ‘do’ in their non-actual digital representations, part generated and digitally transformed by ahistorical models. Evens’ apparently calm understatement is devastating: their “dampened contingency weakens … participation in subtle but essential human values (p. 182).”

 

 

Bibliography

Hayles, N. Katherine. Postprint: Books and Becoming Computational. New York: Columbia University Press, 2021.

Hayles, N. Katherine. Unthought: The Power of the Cognitive Unconscious. Chicago: University of Chicago Press, 2017.

Kittler, Friedrich A. Gramophone, Film, Typewriter. Translated by Geoffrey Winthrop-Young and Michael Wutz.  Stanford: Stanford University Press, 1999.

Kittler, Friedrich A. ‘The World of the Symbolic: A World of the Machine.’ In Literature Media Information Systems, edited by John Johnston. Critical Voices in Art, Theory and Culture, 130-146. Amsteldijk: G+B Arts International, 1997.

Lacan, Jacques. Écrits: The First Complete Edition in English. Translated by Bruce Fink.  New York & London: Norton, 2007.

LaRocca, David, and Garrett Stewart, eds. Attention Spans: Garrett Stewart, a Reader. New York: Bloomsbury Academic, 2024.

Stewart, Garrett. Book, Text, Medium: Cross-Sectional Reading for a Digital Age.  Cambridge: Cambridge University Press, 2021.

Notes