What the Dead Refuse to Leave: Spectral Evil and the Afterlife of Violence

The previous entry ended with an afterimage. Representation had done what the pharmakon always threatens to do: preserve enough of the poison to make of its structure an efficacious agent. But preservation immediately creates another problem. Once the image remains, what exactly remains with it? The event itself cannot be repeated without falsification; the dead cannot be restored by narrative; suffering cannot be redeemed by being remembered. And yet something survives the moment of violence—not as the event returned intact, but as a disturbance in the present that has lost the right to call itself merely past.

Derrida’s hauntology, developed in Specters of Marx, offers a vocabulary for this disturbance because the specter unsettles the reassuring opposition between presence and absence. A ghost—the specter of a massacre or a pogrom or a hate crime—is not simply here, like an archeological trace that can be inventoried, nor simply gone, like an object removed from the room. It arrives from a time that refuses chronological obedience. It addresses the living from a place the living cannot occupy, binding inheritance to a debt that cannot be discharged simply by declaring the historical account closed (Derrida, Specters of Marx).

To bring hauntology into a theory of evil therefore requires precision. It does not mean that evil possesses mystical immortality, still less that every memory of injury is a traumatic haunting. The wager is narrower: arbitrary violence has a peculiar capacity to leave a remainder because the destruction was never adequately absorbed by the reasons offered for it. Where justification collapses, closure becomes difficult. The wound may heal, the trial may end, the regime may fall, the perpetrator may die; nevertheless the unanswered disproportion between what was done and what could possibly have warranted it continues to exert pressure on meaning.

This is the temporal consequence of the Sadean scandal. The less an act can explain itself, the more insistently its effects may demand explanation afterward. Arbitrary violence inflicted upon the scapegoated other produces not only suffering in the instant but an interpretive debt—a hauntological residue. The victim is harmed once by the act and risks being harmed again when the living make that act intelligible too quickly—when memorial language turns destruction into (manifest) destiny, necessity, human fallibility, heroic sacrifice, therapeutic lesson, or the inevitable price of history. Haunting begins where the explanation fails to make the dead stay buried in the meaning assigned to them.

The specter, then, is not the dead person returned as symbol for our use. It is the failure of the living to “own” the past completely—to possess it diagnostically, analytically, in honesty and humility. That failure matters ethically. It prevents commemoration from becoming ownership and interpretation from pretending that the event has finally yielded its truth. Evil’s afterlife lies precisely in this resistance to completion: what was negated continues to trouble the categories through which the negation was made acceptable.

The five cultural works guiding these notes do not stage one generic ghost. Each invents a different formal mechanism through which violence survives its scene. Their value is diagnostic precisely because the mode of return changes with the kind of evil being represented.

In Heart of Darkness, colonial atrocity persists as narrative opacity. Kurtz dies, the river journey ends, and Marlow returns to Europe; yet the encounter cannot be left in Africa as a completed episode. Kurtz’s final recognition—the horror condensed into words as powerful as a primal invocation—migrates into Marlow’s narration, his evasions, his encounter with the Intended, and the darkness gathering over the metropolitan world. Colonial sovereignty returns not as a ghostly character but as a contamination of narration itself. What happened at the Inner Station becomes impossible to confine to the geography that imperial discourse had designated as an atopian elsewhere of the imagination.

Hitchcock’s Shadow of a Doubt makes the afterlife domestic. Uncle Charlie’s nihilism enters Santa Rosa through familiarity, charm, kinship, and the “telepathic” bond with his niece. The film’s terror depends upon the fact that revelation cannot restore the household to its previous innocence. Once ordinary gestures, rooms, dinner tables, and affectionate names have acquired a second meaning, normality itself becomes estranging and suspect. Freud’s uncanny is useful here not as a diagnosis of the characters but as a formal description: what should have remained hidden returns inside what was most familiar (Freud, ‘The Uncanny’). Evil does not haunt the suburb from outside. It makes the suburb remember what its surfaces had allowed it not to know.

Dogville turns haunting into architecture. Its chalk lines refuse the consolations of a fully furnished world: walls are visible as conventions, privacy as an agreement, property as an inscription that can be crossed. Grace survives the town’s abuse and eventually orders its destruction; the argument therefore cannot rest on the false idea of her annihilation. What persists is the trace of a community that had converted hospitality into exploitation and moral language into permission. The emptied stage does not erase Dogville. It leaves the spectator facing the Sadean diagram of the perverse relations constructed by the denizens of a town whose entire identity has come to be articulated around the staging (Darstellung) of a scapegoat (the Passion of Grace Mulligan).

No Country for Old Men locates haunting in the contingency of the other as viral agent (the “bubonic plague” of organized crime) and belated comprehension. The killer’s ritualized violence gives chance the appearance of nemesistic law, but the deeper afterlife belongs to Sheriff Bell, whose language circles an evil he cannot make proportionate to the moral narratives available to him. The novel’s losses do not culminate in catharsis or explanatory mastery. They return as recollection, bafflement, and dream. The past survives not because Bell discovers its secret but because the world of transnational capital has become less readable to him without becoming meaningless—a network of dark “principalities” and “evil in high places”.

Falling Man brings spectrality closest to the atmosphere of our own media world. The terrorist act does not end with the collapse of the towers because catastrophe survives in bodies, images, mishearings, pharmakonic performances, routines, and fractured memories. The children’s “Bill Lawton” turns historical terror into a linguistic ghost; the falling-man performance forces the image back into public space; memory loss and repetition disturb the confidence that an event can be stored as stable knowledge. DeLillo’s novel does not offer trauma as a cure. It shows how catastrophe enters circulation, becoming intimate and mediated at once (Falling Man).

Hauntology becomes morally dangerous the moment the dead are treated as poetic resources. There is an esthetic temptation in spectral language: absence is beautiful, ruins are eloquent, memory acquires atmosphere, and the victim becomes the occasion for the survivor’s profundity. A theory of spectral evil must resist precisely this conversion. The dead do not exist in order to deepen the living.

Levinas helps name what is at stake because the ethical encounter begins with a singularity that cannot be exhausted by my categories. His philosophy of the face does not provide a ready-made theory of ghosts, and it should not be recruited as one. But it sharpens the problem of representation: the other person exceeds the image, concept, institution, or story through which I attempt to contain them. When violence erases a face into a number, category, corpse, statistic, or historical function, remembrance becomes ethical only insofar as it resists repeating that erasure (Levinas, Totality and Infinity).

The haunting trace therefore cannot stand in for the victim as a substitute presence. It marks the insufficiency of substitution. A name on a wall, an empty chair, a photograph, a ruined building, a testimony, or a sentence in an archive can preserve an address to the living, but none possesses the person whose absence it records. The ethical power of the trace lies in this failure. It keeps memory from becoming resurrection theater and prevents knowledge from claiming that the lost other has been fully recovered.

Blanchot’s writing of disaster intensifies the same difficulty. Disaster is not simply a hyperobjectal event large enough to overwhelm representation. It disorganizes the confidence that representation arrives afterward, serenely intact, to describe what happened. The catastrophe enters the language that would contain it. Hence the negative capability and broken temporality characteristic of so much writing after violence: repetition, ellipsis, silence, dislocation, the sentence that approaches an event and then recoils. Such forms are not guarantees of ethical seriousness, but they can register the damage done to the very instruments of intelligibility (Blanchot, Writing of the Disaster).

This is also why “trauma” must not become a decorative synonym for anything remembered painfully. Clinical trauma belongs to persons and requires forms of knowledge these notes do not presume to supply. When the term appears here, it names the representational and historical problem staged by the works themselves: the return of an event that cannot be incorporated into a continuous narrative without a hauntological remainder. The distinction protects both theory and suffering from inflation.

A specter never arrives without an inheritor. Someone receives the testimony, preserves the photograph, enters the ruined building, hears the family story, reads the file, watches the replay, or decides that the archive can be closed. Haunting is therefore not simply persistence. It is a struggle over transmission: who may speak for the dead (and in/on what terms), what institutions will preserve their traces, what forms of authority remembrance will acquire, and what new violence can be committed in the name of an injury that has become sacred.

This last possibility is crucial. Memory is not automatically ethical. A community can inherit victimhood as permission, convert mourning into entitlement, weaponize the dead against new others, or turn historical injury into a permanent exemption from reciprocity. The specter can interrupt sovereignty, but it can also be conscripted by sovereignty. The fact that suffering persists does not determine what the living are entitled to do with it.

Here the Sadean problematic reappears in a less obvious form. Jouissance does not need to belong to the original perpetrator in order for violence to acquire an afterlife. Later institutions, audiences, political movements, and cultural markets can extract intensity from repetition: from the image replayed until horror becomes icon, from the grievance whose renewal produces identity, from the memorial spectacle that places the living on the righteous side of the dead. This is not evidence that remembrance is secretly sadistic. It is a warning that no practice of memory is exempt from the question that has governed these notes from the beginning: what surplus enjoyment appears when the declared purpose no longer accounts for what the practice has become?

The ethical alternative is not forgetting, but an inheritance capable of remaining answerable to what it cannot possess. Such an inheritance refuses both closure and fetishization. It asks what the trace requires now, which repetitions preserve recognition and which merely reproduce domination, which stories restore singularity and which convert the dead into instruments. The ghost matters because it denies the living the innocence of saying that what has passed has therefore been settled.

And yet ghosts do not remain weightless. Sooner or later their traces enter institutions. Testimony becomes deposition; names become lists; memory becomes archive; injury becomes category; responsibility becomes case number. The afterlife of violence acquires clerks, forms, schedules, procedures, databases, and rules for what can count as evidence. At that point another danger appears: the specter may be preserved so efficiently that no one any longer has to encounter the person whose trace is being administered.

The next entry follows evil into that colder register—from the ghost to the file, from the afterimage to the procedure, from the unresolved claim of the victim to the institutional language through which responsibility can be distributed until it appears to belong to no one.