Culture does not move cleanly from past to present to future. It circulates. Images survive their eras. Predictions become designs. Designs become habits. Habits become institutions. Institutions become ruins, and the ruins acquire meanings that their builders never intended. The future is imagined through fragments of the past, while the past is repeatedly rewritten by the futures that actually arrive.
Within this circulation, identity is never entirely stable. Personal identity depends upon memory, recognition, narrative and continuity. Cultural identity depends upon much the same things, only distributed across populations. Both are haunted by what has disappeared, what might have happened, and what was once expected to happen but did not.
Fear enters precisely where these narratives fail.
The modern individual tends to imagine the self as coherent: one body, one biography, one name, one history. Yet even before digital technology this coherence was partly fictional. Memory edits. Families suppress. Nations mythologise. Institutions archive selectively. Individuals reconstruct themselves according to circumstances. We inhabit stories about ourselves because without narrative continuity experience would be only a succession of disconnected events.
Technology has made this instability more visible.
A person now exists simultaneously as biological body, legal identity, employment record, social profile, photographic archive, search history, credit score, behavioural model, biometric template and statistical prediction. These representations overlap but are not identical. Some are generated by the individual, some by institutions, and others by machines inferring characteristics that the person may never consciously have expressed.
The self has become distributed.
This produces a peculiarly contemporary form of horror. The threat is not necessarily that someone will steal your identity. It is that identity has already ceased to belong entirely to you.
A system somewhere may possess a version of you that determines whether you receive a loan, enter a country, see an advertisement, qualify for insurance or become the subject of investigation. That computational double need not resemble the person you understand yourself to be. Nevertheless, in practical terms, it may become authoritative.
Here personal anxiety begins to resemble cosmic horror.
H. P. Lovecraft repeatedly constructed stories in which knowledge destroys the apparent boundaries of the self. A protagonist discovers an ancestry that alters who he believes himself to be, a consciousness that can migrate between bodies, an intelligence older than humanity, or a universe in which human categories possess no privileged significance.
The terror does not simply come from encountering a monster.
It comes from discovering that the category called “human” is less secure than expected.
The same structure survives throughout modern horror. The alien becomes parasitic. The body becomes unreliable. Memory becomes questionable. The copy becomes indistinguishable from the original. The environment becomes sentient, or perhaps reveals that the distinction between environment and organism was always artificial.
Eventually the question ceases to be, “What is the monster?”
It becomes, “What am I?”
This movement from external fear towards ontological uncertainty is one of the defining features of modern cultural anxiety.
Yet Lovecraft himself is now part of a larger feedback process. His stories were influenced by earlier anxieties about science, degeneration, empire, race, archaeology and humanity’s shrinking position within an expanding universe. Those anxieties entered fiction. Fiction transformed them into a distinctive symbolic vocabulary. Later writers, filmmakers and game designers reused that vocabulary. Their works then shaped how subsequent generations imagined the unknown.
The fiction returned to culture as a framework for perceiving reality.
This is where hyperstition becomes useful.
A hyperstition is more than a prediction that happens to come true. It is an idea whose circulation contributes to creating the conditions under which something resembling it becomes real.
Cyberpunk provides particularly obvious examples.
Early cyberpunk did not predict contemporary technological life in any simple prophetic sense. It exaggerated existing tendencies: computer networking, multinational corporations, prosthetics, surveillance, consumer electronics, virtual environments, artificial intelligence and fragmented urban identities.
But those stories were subsequently consumed by people who designed software, interfaces, games, corporations and networks.
Fiction provided metaphors.
The metaphors became expectations.
Expectations influenced designs.
Designs produced technologies.
Those technologies were then described using the terminology of the fiction.
The loop closed.
“Cyberspace” is perhaps the exemplary case. A fictional spatial metaphor helped people conceptualise computer networks. Once the metaphor circulated widely enough, actual technological development began to be understood through it.
The fiction did not predict reality.
It participated in constructing reality.
This process also operates culturally rather than technologically. A society produces stories about itself. Those stories influence behaviour. Behaviour produces circumstances which appear to confirm the stories.
A nation that continually describes itself as threatened begins constructing institutions around threat. Those institutions generate surveillance, boundaries and enemies. Their existence then seems to prove that the original fear was justified.
A culture that tells itself technological innovation is inevitable directs money, prestige and labour toward certain technologies. Their eventual appearance is then presented as proof of inevitability.
Narrative becomes infrastructure.
But not every imagined future arrives.
This is where hauntology enters.
Hauntology concerns the persistence of absences. Its most compelling cultural form is the lost future: something that never existed but was once confidently expected.
The twentieth century generated vast archives of futurity.
Monorails crossed conceptual cities. Nuclear electricity promised abundance. Automation promised leisure. Supersonic aircraft promised a shrinking planet. Space stations promised expansion beyond Earth. Public architecture promised enormous civic confidence. Computers promised liberation from bureaucracy.
Some of these futures arrived partially.
Some arrived in distorted form.
Others simply disappeared.
Their images survived.
A faded architectural drawing of a pedestrian city can therefore produce a strange emotion. We are not merely looking at the past. We are looking at the past looking toward a future which never became our present.
The object contains two times simultaneously.
It is historical and futuristic.
This is the territory of the semiotic ghost.
A semiotic ghost is not a supernatural apparition but a sign that has outlived the cultural structure which originally gave it meaning. The chrome rocket, the streamlined tower, the computer terminal glowing green in a darkened room, the flying car, the transparent videophone, the optimistic corporate family standing beneath a geodesic dome: each carries an embedded future.
When encountered later, these objects appear haunted.
Not because anything supernatural inhabits them.
Because an abandoned expectation does.
They are evidence from timelines that did not occur.
Personal identity is structured in much the same way.
Everyone possesses lost futures.
The career never pursued.
The relationship that did not continue.
The city never moved to.
The version of oneself imagined at seventeen, thirty or fifty.
These unrealised selves do not simply disappear. They remain available as comparisons against the present.
One can therefore be haunted by someone who never existed.
This is one reason nostalgia can become painful. Nostalgia is usually described as longing for the past, but much nostalgia is actually longing for a past future.
We remember not merely what happened but what we once believed would happen.
Childhood memories are filled with this peculiar temporal structure. The remembered summer contains the adult knowledge that the future anticipated by the child will not arrive.
The remembered object—a toy computer, a magazine, a science-fiction programme, a school textbook—can suddenly recover the expectation attached to it.
For a moment we remember tomorrow.
Cultural identity operates through the same mechanism at larger scale.
Nations construct narratives from remembered victories, defeats, migrations, industries and institutions. Yet they are equally defined by abandoned futures.
The industrial future.
The imperial future.
The socialist future.
The technological future.
The European future.
The post-industrial future.
The global future.
Each political generation inherits not merely history but unfinished predictions.
Consequently public debate is often conducted among ghosts.
People argue over versions of society that once seemed possible. Political movements promise restorations of conditions that never existed quite as remembered. Technologies are promoted using imagery borrowed from earlier technological revolutions. New crises are interpreted through wars, depressions and cultural conflicts whose original circumstances have disappeared.
The past is therefore continually projected forward.
At the same time, imagined futures travel backward.
Once something happens, earlier history begins to appear as though it were always leading toward it.
The arrival of the internet made previous networking experiments look like precursors.
The arrival of smartphones transformed pagers, PDAs and early mobile computers into evolutionary ancestors.
The arrival of generative AI causes decades of fictional computers and artificial minds to be reread as prophecies.
This retrospective reconstruction is powerful.
The successful future colonises the past.
Technological history becomes populated by inventions which appear inevitable only because we know what eventually succeeded.
Failures disappear.
Alternative paths become curiosities.
The past is reorganised around the present.
This creates another feedback loop:
the present selects its ancestors.
Those selected ancestors then justify the present.
The resulting narrative makes the present appear inevitable.
Culture becomes haunted not simply by what came before, but by the paths it has erased.
This is where fear becomes politically and culturally important.
Fear simplifies possibility.
When frightened, individuals and cultures seek strong categories: friend and enemy, human and nonhuman, authentic and fake, normal and abnormal, safe and contaminated.
Horror repeatedly attacks precisely those distinctions.
The possessed person looks normal.
The double possesses the original’s memories.
The machine imitates human speech.
The photograph may be synthetic.
The trusted institution contains the threat.
The familiar landscape has subtly changed.
The horror lies in failed classification.
Contemporary technological life increasingly generates similar uncertainties without requiring supernatural causes.
A photograph is no longer straightforward evidence that something occurred.
A voice recording is no longer straightforward evidence that someone spoke.
A message may have been written by a human, generated by a model, generated by a model and edited by a human, or copied from something generated years previously.
An online identity may belong to a person, organisation, bot, agent or composite.
The informational world becomes populated by ambiguous entities.
Historically, cultures interpreted ambiguous agency through ghosts, spirits, demons, doubles and possession.
Modern technological society uses algorithms, agents, models and synthetic media.
The ontologies differ enormously, but the psychological problem has similarities.
Something acts.
Its internal state is inaccessible.
Its intentions are uncertain.
Its presence must be inferred from traces.
The ghost returns as an epistemological structure.
This does not mean contemporary machines are supernatural. It means older supernatural narratives remain useful because human beings repeatedly confront situations in which agency is visible while the agent is obscure.
Technology does not eliminate ghosts.
It manufactures new conditions in which ghost metaphors become persuasive.
Generative artificial intelligence intensifies this effect because it is fundamentally archival.
Its outputs appear new, yet they emerge from statistical relationships extracted from enormous bodies of previous cultural production.
The future is generated from the past.
Ask for a future city and the machine returns accumulated fragments of previous future cities: cyberpunk neon, modernist towers, Gulf megaprojects, science-fiction spacecraft, corporate concept art, Japanese signage, twentieth-century utopian architecture.
The supposedly unprecedented future arrives pre-haunted.
AI therefore creates an unusual temporal machine.
The archive produces the future.
The generated future enters the archive.
Future models learn from those generated artefacts.
The distinction between cultural memory and cultural prediction begins to blur.
This could be understood as an automated semiotic feedback loop.
Humanity imagines the future.
Machines learn humanity’s imagery of the future.
Machines produce new images of the future.
Humans consume those images.
Designers incorporate their aesthetics.
Products begin to resemble them.
The generated image gradually acquires causal power.
Hauntology and hyperstition converge.
One concerns futures that failed to arrive.
The other concerns imagined futures that help themselves arrive.
Between them sits contemporary culture, continuously deciding which ghosts to resurrect.
This is also why cultural identity is becoming increasingly unstable.
Identity requires forgetting as well as remembering.
A person who remembered every humiliation, contradiction and abandoned ambition with equal intensity might struggle to sustain any coherent sense of self. Cultures likewise depend upon selective memory.
Digital technology weakens forgetting.
Old photographs return automatically.
Messages remain searchable.
Embarrassments survive.
Dead websites persist in archives.
Abandoned communities remain indexed.
Voices can be reconstructed.
Images can be generated of people at ages they never reached.
The dead increasingly leave behind interactive informational residues.
We are entering a culture in which absence itself may become technologically difficult.
The consequences may be profound.
Traditional mourning depends partly upon the gradual transformation of presence into memory. Digital systems complicate that transition because traces can continue behaving as though presence persists.
A dead person’s photographs can appear in automated reminders.
Their account can remain contactable.
Their voice may survive in recordings.
Their written language may become sufficient to approximate a conversational persona.
The cultural ghost approaches technical implementation.
Again, the supernatural metaphor should not be confused with literal supernaturalism.
But the emotional structure is ancient.
The dead speak.
The past answers.
The image moves.
The voice returns.
The boundary between memory and encounter becomes uncertain.
At this point personal hauntology, technological hyperstition and cultural identity converge.
Human beings construct themselves through stories.
Cultures construct themselves through stories.
Stories generate imagined futures.
Some futures influence behaviour and become real.
Others fail and become ghosts.
Those ghosts enter later stories.
Later stories alter later behaviour.
The cycle continues.
Past and future are therefore not opposite directions.
They are participating systems.
The past provides materials from which futures are imagined.
The future determines which parts of the past become significant.
Fear concentrates attention upon certain possibilities.
Fiction gives those possibilities form.
Culture repeats the forms until they become familiar.
Technology sometimes embodies them.
Experience then appears to validate the fiction.
Eventually the distinction between prediction, memory and design becomes difficult to locate.
Perhaps that is the deeper social haunting of the twenty-first century.
We inhabit environments crowded with unrealised futures, reconstructed pasts, digital doubles, artificial voices and stories that have escaped the conditions that created them.
Our ghosts no longer need graveyards.
They inhabit databases, interfaces, brands, ruins, archives, algorithms and expectations.
And among them moves the most persistent apparition of all: the imagined coherent self, attempting to preserve a continuous identity while being constantly rewritten by memory, technology and culture.
We are haunted by what happened.
We are haunted by what did not happen.
We are haunted by what we once believed would happen.
And increasingly, we are haunted by futures that may be becoming real precisely because we imagined them first.