Why Do Ghost Stories Work? Fear Built From Incomplete Information

Fear responses are driven far more by ambiguity than by threat, which is why a closed door is worse than an open one and why a described monster is less frightening than a suggested one. The brain generates predictions continuously, and when sensory information is insufficient it fills the gap, so a story that withholds detail recruits the listener's own imagination to supply something more personally frightening than any description could be. H. P. Lovecraft stated the principle in 1927, arguing that the oldest and strongest fear is fear of the unknown, and the craft advice that follows is consistent across traditions: show less, explain nothing, and stop before the explanation. The same mechanism explains why a story loses power on a second hearing and why horror films frequently deflate at the moment the creature is shown clearly, since the gap the audience was filling has been closed with something specific and therefore finite.

The devices recur across cultures and centuries because they work on the same machinery:

• •The unreliable witness, so the audience cannot determine whether an event occurred, which sustains ambiguity past the end of the story
• •The delayed reveal, in which something present all along is recognised only at the last moment, forcing the listener to reinterpret everything preceding it
• •The rule that must not be broken, which creates dread by making the audience anticipate a specific transgression
• •Familiar settings rather than exotic ones, since a threat in a bedroom or a corridor is closer to the listener's own environment than a castle
• •Auditory ambiguity, since hearing is the sense least able to localise and identify, which is why noises in the dark work better than sights
• •The frame narrative, where a character tells the story to other characters, which supplies deniability and models the audience's own reaction
• •The unresolved ending, which prevents the relief that closure would provide

Voluntary fear is a real puzzle, and several explanations have evidence behind them. Excitation transfer holds that arousal produced by fear does not dissipate immediately, so the relief that follows is experienced as intensified pleasure, which explains why the enjoyable part frequently comes just after the frightening part. Benign masochism describes the enjoyment of a negative sensation when the body's alarm can be overridden by the knowledge that it is safe, which also covers spicy food and fairground rides. Mastery accounts hold that practising fear in safe conditions is useful, providing a rehearsal of threat responses at no cost, which would explain why frightening play is common in children and other young animals. A study conducted in a commercial haunted attraction with physiological monitoring found that participants who reported enjoying it most showed substantial physiological arousal and afterwards reported improved mood, which supports the idea that the experience does something rather than merely passing time.

Several reported ghost encounters have well-documented physiological explanations, and knowing them does not eliminate the experiences. Sleep paralysis, in which a person wakes while the muscle atonia of dreaming sleep persists, produces immobility, a crushing sensation on the chest and a vivid sense of a malevolent presence, and it appears across cultures with locally specific interpretations. Hypnagogic and hypnopompic hallucinations occur at the boundaries of sleep. Carbon monoxide poisoning produces hallucinations, confusion and dread, and a well-known nineteenth-century case of a haunted house was traced to a faulty furnace. Infrasound, sound below the threshold of hearing, has been reported to produce unease and a sense of presence, though the evidence is mixed. Pattern recognition tuned to detect faces and agents produces false positives readily, particularly in poor light, and expectation shapes what is perceived, with studies showing that people told a location is haunted report more anomalous experiences in it.

Ghost stories work on people who do not believe in ghosts because fear responds to ambiguity rather than to belief, and a story that withholds detail makes the listener supply something worse than any description. The techniques are the unreliable witness, the delayed reveal, the forbidden rule, familiar settings and unresolved endings. Voluntary fear is enjoyable through arousal carrying over into relief and through safe rehearsal of threat, and many reported encounters match sleep paralysis or carbon monoxide exposure.