The oxymoron “digital humanities” embodies what is perhaps the fundamental human problem of our age, one that GA can hopefully help us to formulate with a modicum of supplementary clarity. What C. P. Snow called back in 1959 the problem of the “two cultures” has evolved into a stratification in which the firstness that drives cultural innovation is increasingly located on the digital side of the barrier. But DH does not simply crown the triumph of the digital over the humanistic; it has redeeming cultural features that can hopefully allow global civilization to transcend the victimary resentment that currently motivates so much of its political and cultural activity.
At the most superficial level, DH reflects the obvious fact that today the activities of production and commentary on the cultural works that are the concern of humanists involve the use of digital computers. Thus we compose and archive texts online, we search them, we scan printed text into digital files; and we can compile data about these documents and analyze it with software, just as we do with other data. But in doing this, we open the door to reducing the signs of language to numbers. This begins superficially with ASCII codes, but it is unclear where it ends; suppose all of human language, all of human experience, could be formulated in algorithms, reduced, that is, to complex forms of counting?
As I pointed out in Chronicle 477, associating the humanities with the digital takes us back to the origin of the human in the “digital” act of separating the significant/sacred from the insignificant/profane. The tension (fear, irritation, excitement) implicit in the term DH comes from that between the humanistic opposition sacred/profane with its secular-esthetic variants and the binary difference 1/0 that is the starting point of mathematics.
When we consider the originary sign as designating/representing the sacred central object from the “humanistic” perspective, what is significant/sacred is infinitely rich and precious, and no representation can fully do justice to it. As the unique central object is by this very fact a placeholder of uniqueness on the human scene, any other object can come to share its unique status and prompt us to explore its significance. In contrast, if we confine the sign’s “meaning” to the mere gesture of designation itself, reducing the significance that it confers to a single bit of information, we can abstract from its specific designatum and simply designate objects in turn irrespective of their individual qualities. This permits us to count them: one, two… And once one begins to count things, the rest of mathematics follows.
The history of the interactions of language and mathematics, meaning and counting, has gone through many phases since ancient times, notably the beginnings of modern science in the Renaissance, when mathematical formulas were first applied to experimental data, and the secularization of the Enlightenment, which dethroned sacred scripture and its anthropological intuition as a source of knowledge about the natural world. What is new in the era of DH is the emergence of artificial intelligence, the use of mathematical techniques to simulate the kind of reasoning that had previously been done with language rather than numbers. This process involves creating algorithms to digitize decision-making processes that had formerly relied on human intuition to evaluate alternatives.
For example, although chess is an objective exercise, the quality of human players depends not only on their capacity for pattern retention and manipulation but on their intuition of the board that allows them to choose which variations to follow. Computers play chess using brute force (although they have their own, cruder ways of pruning trees of useless moves). Humans think; computers just calculate. But in the end what counts is the result, and in the case of chess, that result has been settled in favor of the computer.
The advent of DH-AI poses to us a question never faced before: what in the human, if anything, is not digitizable or simply, programmable. In the previous Chronicle (487) I mentioned poetry as an example of a creative activity that strikes us as “too human” to be imitated by a computer. Yet I have seen output from poetry programs that is arguably of at least the quality of most of the poems published today. Indeed, my impression is that AI has exercised on the arts a not-quite-conscious mimetic attraction. Where one might expect artists of various kinds to avoid producing works that could appear to be the result of an algorithm rather than a human soul, the contemporary arts often seem to go in the opposite direction, reducing human specificity to our unique ability to make arbitrary esthetic decisions (“this urinal / this boulder will be shown as an artwork”). Which certainly affirms the (free) human as wholly other than a (mechanical) computer, but which paradoxically is the easiest kind of decision process to simulate on a computer: for Mallarmé’s Un coup de dés jamais n’abolira le hasard, just substitute a random number.
If the output of the human-scenic-sacred can be simulated without the experience of transcendence (except in the humans who produced the computers in the first place; but they would be analogous to the “watchmaker” God of the Enlightenment, not the living God of genuine religion), then even if the operation of the human mind is not at all comparable to that of a digital computer, we could not avoid the feeling of being reduced to an epiphenomenon, a discovery procedure for robots, all human religion and language and art just devices to get us to the point where all we have accomplished with representation since the origin can be analyzed as the output of digital algorithms. The extreme forms of the victimary that we witness in our era, from anti-Western Islamist barbarism to racial and sexual resentment to climate hysteria may perhaps best be understood as symptoms of an apocalyptic fear of the “end of the human.”
We may define scientism as the postulation that the human brain is “nothing more than” a computer; a “scientism-ist”—I will abbreviate these words as s-ism and s-ist, (interesting that the term scientism seems deliberately designed to prevent its advocates from having a name)—is someone who denies in principle not the originary hypothesis itself but its very relevance. An s-ist is a Platonist who believes that mathematics “exists” in our absence, and that all that we do and think can in principle (and presumably will fairly soon—the latest date for Kurzweil’s “Singularity” appears to be 2045) be understood in terms of numerical algorithms. The paradoxical aspect of language that, for example, Kieran Stewart emphasized in his examination of the compatibility of GA with the theories of Mircea Eliade in the last Anthropoetics, or in Sartre, the néant that separates the subject from the object of his representation, or in Derrida’s terms defers (diffère) the choice of the sign that designates the referent—all this is dismissed as window dressing. If all phenomena in the universe are reducible to input-output operations expressible in mathematical equations, then the paradox inherent in the fact that humans alone have been able to discover these equations cannot be given any special weight without putting the entire edifice in question.
Let me present this point in all its nudity. GA postulates that the human comes into being by means of the deferral of violence through representation. In denying the irreducibility of human representation, s-ism is effectively claiming that violence—”instinct” if you like, horizontality as opposed to verticality—is all there is, but that nature’s “violence” itself, through Darwinian evolution alone, is able to maintain the human social order in equilibrium just as a natural ecology does—as animals do, for example, through the instinctual inhibition of aggressive behavior before a more powerful adversary. The existence of representation in its various guises would be of interest only as a tool for manipulating the algorithms of “nature” to improve production, health, etc., but without being ontologically different from the instinctual reactions of animals. Chomsky’s famous victory in his debate with Skinner would have been effectively reversed.
Expressed in this fashion, scientism might seem coextensive with worldly realism; can one reject s-ism without relying on the supernatural? But rational thought is not limited to the algorithmic. GA provides an intellectual basis for those who don’t want to kick the human upstairs but consider that our very dependence on the transcendental is what makes us human. The transcendental is revealed on the scene of representation. This scene has so far not been conceived as an object for computer simulation, although if GA one day leaks out of the GASC, this may eventually be attempted. But the scene of representation concentrates in itself the untotalizable reality of human desire. In a word, a digital device that could fully simulate this totality would, indeed, be human.
Although monotheism could be discovered only once, every attempt to understand the One God repeats this discovery; that is what we mean by an act of faith. The concentration of all meaning in a unique being is the antithesis of the one-bit meaning embodied in counting. God’s refusal of a (non-declarative) name should be understood as a critique of “names” in general; no particular use of language is meaningful in itself, but depends ultimately on the unknowable and undecidable complexity of the scene that forms around its center. The Saussurian notion of the paradigm central to Derrida’s concept of différance must be referred to its ontological basis in the scene of representation. It is not enough to say that “language is a system of differences” in which every individual term depends on the whole; the “whole” of language is not a given totality to be divided up, but the virtual totality of meaning concentrated in the sacred center of the scene that defers the potential violence of the periphery.
It would be a mystical exaggeration to claim that every use of language is a revelatory equivalent of the originary event. We should say, rather, that every use of representation is experienced on a continuum between revelation and counting, the experience of the sacred and the mere gesture of designation, or simply: humanities and the digital. Art, which provokes the esthetic oscillation between the sign and its (imaginary) referent, is the institution whose purpose is to recall to us outside the ritual context the root of our shared representations in the originary sacred, to show us that we all contain within ourselves the means to experience revelation.
Reflection on the digital makes us realize by contrast the non-algorithmic complexity of our reaction to a work of art. One relates to an artwork as to a Subject; whence film criticism’s insistence on the director as the auteur, despite the undoubtedly collective nature of film production. But film should also teach us that, precisely, a “Subject” is not coextensive with a human individual; the originary Subject was collective and scenic, and we are all as individuals reflections of this scenic collectivity, a fact that Girard’s term “interdividual” reflects, though I think rather one-sidedly, by appearing to deny individuality altogether.
The revelations of art and religion are irreducible to the digital because they are open-ended. The turning-back-upon-itself that transforms the originary gesture of appropriation into a form and thence into a sign that communicates to the others the transcendent status of the central object cannot be reduced to the transmission of so many bits of information. (It is a mistake to think of the first sign as simply pointing; pointing is more like counting, something done once the “mutual shared attention” established by the first sign has been achieved.) The sign-as-form is meant to model and draw attention to its referent’s numinousness that makes its appropriation unthinkable. In the dynamic of the scenic configuration the precise “meaning” of the central object cannot be fixed as so to speak the “dictionary definition” of the sign. I have called the first sign the “name-of-God” to make clear that there is no limit to the ideas and emotions that may be invested in the central object, whose position at the center of the scene delimited by the deferred desires of the participants is what “defines” it constantly anew throughout the scenic experience, which is always at least virtually collective. What is essential is not the “meaning” of the central object but the effectiveness of the scenic configuration in deferring collective violence.
The scene may usefully be contrasted with the laboratory, a thought-category associated with the legendary French epistemologist Gaston Bachelard. A laboratory is a controlled environment where the number of parameters to be measured in a particular experiment is reduced to a minimum. The laboratory is the privileged realm of the digital, where counting takes place; on the scene, in contrast, there is no way to limit the circulation of meanings, even when there is but a single spectator before an artwork.
Does this guarantee that s-ism is refuted, that computer simulation can never equal or excel the creativity of human representation? Computers were invented in my own lifetime. Imagine ten, one hundred lifetimes into the computer age (and imagine, if we can, that we haven’t destroyed ourselves during this time); can we be certain that HAL-27 or HAL-1382 won’t be able to relive all human history and simulate the human scene in all its richness, that Andrew Bartlett’s impossible human won’t have become fully human, not to say superhuman?
Here speculation becomes useless. But we should recall how we distinguish genuine art and revelation from human simulations. Distinguishing good from bad art involves an intuition of authenticity, a term Heidegger defined in terms of being-for-death, but for which GA gives us a much more concrete criterion: authentic humanity is what appears on the originary scene (where one faces death, and chooses deferral and life) and on every subsequent scenic/cultural experience that recalls the originary experience of the sacred. The authentic experience of the scene is never a simple repetition, just as no authentic work of art merely recycles old formulas.
Society depends on mimesis, paradoxically, always producing something new. The reciprocity of the moral model depends on the constant renewal of firstness. Firstness always arouses resentment, and I have already suggested that the current disquieting intensity of victimary trends is linked to the growing dominance of digital as opposed to “humanistic” firstness in our era. But this is the moment at which to recall our earlier insight that DH, whatever else it may be, is a refuge from victimocracy.
Faith in humanity is faith that the common world of linguistic reciprocity, of cultural unity, can be maintained. And if we have come to an era where this unity breaks down over the differential ability of some, aided by whatever “privilege,” to operate in the digital world, then the digital itself must be called upon to fix the problem, and not merely by insisting on the “diversity” of its workforce.
In this context, we should think of DH as more than merely an academic activity, but as a central participant in the broad cultural movement of our society. In particular, a frequent concern of DH is the phenomenon of “computer-generated art.” But such activity is not limited to AI computers programmed to write poetry or paint masterpieces. Its chief manifestation is in the exponential enhancement of individual and crowd-based representations to be shared, argued over, abandoned, and recycled—the world of Twitter and Facebook and YouTube and Instagram and Meerkat and newer apps those of my generation haven’t yet heard of—in a word, the Internet, the One Medium that is perhaps best understood, to pastiche an old Apple slogan, as the sacred for the rest of us. For it is just when the digital risks dividing humanity into castes of creators and drudges that it at the same time opens up to vast sectors of the public the possibility of participating in new forms of esthetic creation—forms that many DH projects have indeed focused on. The number of people making and posting videos, photographs, blogs, tweets, and what have you, increases by an order of magnitude every few years. The educational potential of such things as video games for acclimatizing the young to digital learning and creativity is only beginning to be tapped.
If DH and the digital world as a whole can provide us via such para-humanistic enterprises the means to love each other a little better, we will have all the revelatory material we need to preserve our civilization from the barbarism and apocalyptic destruction that unchecked resentment risks provoking. And if on top of this, as I still believe, humans remain able to create true art in the traditional sense—universal, revelatory, originary-scene-restoring—so much the better.
DH and Violence: A Footnote
The Girardian question to GA is always: “But where is the scapegoat?” Since DH’s attempted reduction of the human to the algorithmic removes specifically human violence from the equation, taking for granted that “violence” in the sense of “nature red in tooth and claw” can handle itself through animal defense mechanisms to which human representation adds nothing ontologically new, we would do well to unpack GA’s notion of “the deferral of violence” that might otherwise be accused of sweeping violence under the rug.
In speaking of the scene of representation as in principle irreducible to digital analysis, I described it as non-totalizable, as the locus of an unpredictable set of interactions that maintain the numinousness of the center and prevent immediate appropriation and conflict. But this deferral of violence is, in the originary event and in principle in all successor events, only temporary. The originary deferral is succeeded by the sparagmos in which the central object becomes the focus of the controlled violence of sacrifice, the “scapegoat” in Girardian terms, and (a key element not in Girard) supplies the community with nourishment. In the artwork that “imitates” the scene, this violence remains in the realm of the imagination, whether directly in tragedy or indirectly in comedy, which always involves some form of sacrifice or expulsion.
In both the digital and the GA model, we may say with Girard that violence expels violence, that Satan drives out Satan, but in the latter case this action, key to human survival, is dependent on the scene of representation, which we access primarily through the deferring signs of language, whereas an algorithmic analysis of human behavior simply understands language as one more “instinctive” phenomenon, a “code” like the DNA embodied in living matter. This is indeed the key point of contention between GA and “s-ism,” and to the extent that DH deals only with algorithmic aspects of culture, it can be associated with s-ism, although in principle passively, since insistence on ideas such as “free will is a myth” or “the mind functions like a digital computer” is hardly within its purview.
To the extent that DH is indeed a refuge from the victimary thinking that dominates the academy today, it is precisely in its refusal to understand cultural works within a paradigm that assigns to a predetermined “oppressor,” whom we might just as well call the work’s “scapegoat,” responsibility for the violence suffered by the “victim.” In such analysis the sacrificial mechanism of the work is projected onto the social reality within which it was created, effectively denouncing the Aristotelian “catharsis of pity and terror” as a mask for oppression.
Just stepping back to investigate some aspect of how paintings, narrative plots, or musical sequences operate on their audience, what patterns of words, notes, or images they employ, is an act of deferral that, even within an overall victimary framework, would be a step in the right direction. And most DH work that I have seen avoids imposing such a framework, since it is the analytic detail that is truly “digital.” Just within the context of literary/cultural study, this is a “deferral of violence” that we should appreciate.
But it should also remind us that “violence” in the sense that we understand it as heirs to Girard and beyond him, to the Judeo-Christian religious tradition, is not accessible within the paradigms of DH. Of course the category of “violence” can be included in a DH analysis as well as anywhere else, but the notion of a breakdown of categorial distinctions into a “primordial” disorder (Girard’s crise sacrificielle) from which order must be restored cannot be understood in digital terms. (The characteristic violence of video games that enact in the most extreme form the resentment-purging formulas of popular culture illustrates the insensitivity of the digital world to the problematic relationship between violence and human culture; violence/aggression is treated merely as a “drive,” and the more it can be satisfied, the better.)
The notion of the originary event presupposes a chaotic “time of troubles” following the breakdown of the animal pecking-order hierarchy from which the proto-human community learns to escape via the discovery/invention of the sign’s scene-creating deferral of violence. But whereas we can propose a minimal model of the originary event focused on the sacralized central object of desire on the scene of representation, one that concludes in the controlled sacrificial violence of the sparagmos, no such model can usefully be constructed for the chaos that preceded the event; its chaotic nature is precisely what makes it irreducible to a model. This is the “singularity” from which the human emerges and that poses the ultimate (albeit not necessarily insuperable) obstacle to the digital simulation of the human. Neither the crise nor its resolution can be understood as algorithmic; only once the scenic deferral of violence has been established can the scene be used as a laboratory where counting becomes possible.
(added May 28, 2015)