The Narrative Killers – A Psychological Autopsy of the Flannan Isles Mystery

The up-to-date record and the popular story do not match. In 1901, officials described an orderly room after a finished meal. Later retellings installed an untouched table and a fallen chair. This investigation asks how the story beat the record, and what that tells us about the way narrative travels.

A Tale of Two Kitchens

Joseph Moore, the relief keeper, landed on Eilean Mòr on 26 December 1900 and found the living quarters quiet and orderly. The compound gate and outer door were shut, the hearth was cold, and the beds were empty. Crucially, the kitchen told a simple story of routine. Utensils were washed, pots and pans put away, and the space tidied after use.

Superintendent Robert Muirhead confirmed this in his formal report of January 8, 1901, noting signs consistent with duties being carried out until dinner time on December 15. There was no mention of half-eaten food or chairs out of place. If we start with the documents, the picture is dull, precise, and professional.

Twelve years later, Wilfrid Wilson Gibson published the poem ‘Flannan Isle’. It conjured a vivid interior. The poem describes ‘a table, spread for dinner… But, all untouched’ and, at the head, ‘a chair / Lay tumbled on the floor’. The scene is compact and cinematic. It tells the reader, without saying it, that violence or panic cut across a meal. None of this appears in Moore’s account or in Muirhead’s report, and both predate the poem. The poem is a work of art, not a source document. Its images, however, are the ones most people remember.

Two kitchens cannot be true at once. The superintendent’s record describes order after a finished meal; the latter poem describes chaos in the middle of one. This clash drives the case. It sets the baseline question of why the invented scene beat the documented one in cultural memory. Once we lock that contradiction in place, the rest of the investigation has a clean point of focus. Either the officials misdescribed a room they saw, or a poet invented a room he never visited.

Gibson’s scene functions as a ready-made ‘snapshot’. It compresses a large mystery into a single domestic interruption. Readers can picture the bread, the knife, and the chair on its side. The poem sets the event at a human scale, a meal that will never be finished. The factual description offers none of that, because it is not trying to. A procedural note is meant to be checked, not felt. The emotional gear engages because the imagery is so easy to see.

The official language is procedural and ‘disfluent’, meaning it is harder for the brain to process. ‘The pots and pans had been cleaned and the kitchen tidied up’ is accurate but dull, and it does not attach to a plot. The poem offers simple, high-contrast objects that stick in the mind. In cognitive terms, pictures and concrete nouns are processed quickly and recalled reliably. When a detail is easier to imagine, it often feels truer. That fluency advantage will matter across the rest of the case.

Scripting the Terror

The keepers’ main logbook for mid-December 1900 is missing from public archives.

Contemporary notes indicate that routine observations for 14 and the morning of 15 December were recorded on a chalk slate, a temporary board used to jot down observations before copying them into the permanent logbook. The original volume appears to have passed through court custody during the 1901 inquest, then fell out of view.

That administrative failure removed the men’s documentary ‘voice’ at the key hour. The absence does not prove anything dramatic happened. It does explain how a forged voice found room to grow. Archivally, the gap is as important as any surviving page because it weakens the chain of custody, the official trail of who held the evidence and when, for the case file.

In 1929, an American pulp title printed a story under the byline ‘Ernest Fallon’ that claimed to reveal the keepers’ final log entries.

Pulp magazines sold sensation as fact. This one offered first-person notes from the island, as if a witness had scribbled through fear. The magazine had no duty to evidence, only to a good yarn. The byline is likely a pseudonym, making its origin impossible to verify. As a delivery vehicle for a borrowed mystery, the format was perfect.

The fabricated entries follow a horror arc. One man grows quiet. Donald McArthur ‘was crying’. All three ‘had been praying’. The storm is ‘the worst I have ever seen’. Then came the curtain line, ‘Storm ended, sea calm. God is over all’. Each beat simplifies the men into types and escalates the mood on cue. Regional weather reports for those hours do not support those claims. None of these phrases match the mundane slate notes Moore reported.

Gibson’s poem supplied a stage set. The 1929 story supplied a psychological script. Together, they created character, inner states, and an ending that feels like fate.

The mixture is powerful because readers can now imagine not only the room but also what the men thought and felt. The invented log even positions itself as a first-hand account, which adds borrowed authority. It is a neat package because each piece fixes the other’s gaps.

The simple facts push back.

The slate recorded duty and weather, not prayers or panic. Muirhead’s investigation found heavy storm damage at the west landing, but nothing inside that suggested a struggle. The timetable of reported gales in the area supports rough seas, not a once-in-a-century supernatural blow. The most persuasive ‘document’ in the popular tale is the one we cannot inspect because it never existed. The most boring documents are the ones we can, and they show routine.

Mental Trap 1: Narrative Bias

Narrative bias, the mind’s tendency to prefer a coherent story over raw data, is a simple yet powerful concept with significant effects. Given a choice, the mind prefers a tale about identifiable people with causes and a resolution. Data offers none of that on its own. Once a story locks in, later facts tend to be bent to fit it. The Flannan case shows that tilt clearly.

Muirhead’s official theory, filed on 8 January 1901, reads like an accident report. He inferred that two men went to secure equipment at the west landing and that a great sea climbed the rock and swept them away.

There are no villains in this account. There is also no real agency or power to act, beyond weather and position on a cliff. It explains the physical damage and the clothing noted at the scene. It leaves little room for a human arc.

The fabricated log gives the reader protagonist roles. Ducat is stoic, McArthur breaks, all pray, and the storm becomes an antagonist. The sequence creates rising action and an ominous finale. It supplies motives where none are recorded. It also lets readers step into a drama rather than sift procedures. That is the core advantage of narrative in the mind.

First-person framing is persuasive because it simulates testimony. A line such as ‘I have never seen seas like this’ feels like a confidant speaking. Readers grant empathy and trust more readily to a voice than to a table of timings. The pulp piece exploits that effect by faking a logbook voice. Once that voice is accepted, contradiction feels like nit-picking.

The official account is coherent without being dramatic, which makes it easy to overlook.

The forged account is dramatic without being evidenced, which makes it easy to remember.

In judgement tasks, that imbalance often decides what people pass on to others. The better told story survives the peer-to-peer retelling. Over time, the story becomes the fact.

Mental Trap 2: Cognitive Fluency

Cognitive fluency is the ease with which the brain processes information. When processing is easy, the content feels right. That misreading of ease as truth is often called the ‘truthiness effect‘. Simple language, concrete images, and repetition increase fluency. Technical phrasing and procedural detail lower it. The Flannan material splits cleanly along that line.

‘Untouched meal’ is a simple phrase with a full picture inside it. ‘Overturned chair’ works the same way. They are concrete, visual, and emotionally loaded. No specialist knowledge is required. A reader can hold both in mind and retell them in one breath. That is what makes them sticky.

Administrative wording prioritises accuracy over imagery. ‘Kitchen utensils very clean’ and ‘kitchen tidied up’ are clear but sparse. They do not invite a plot or a feeling. Their value is in being checkable, not in sticking in the mind. For most readers, those lines resist immediate mental playback. That is the fluency penalty of procedural prose.

When two accounts compete, the easier one often wins. Fluency is a shortcut that saves effort, which the brain likes. In this case, the shortcut backfires. A plausible image is not the same thing as a true one. The poem’s pictures ride the shortcut to credibility. The report’s phrasing trips over it.

The poem and the pulp logbook are not just more dramatic. They are better engineered for recall and sharing. They pack images and emotions into short lines that a reader can quote without notes. The official record is harder to share, even though it is the one anchored to named people, dates, and a chain of custody. Once the fluent version circulates, it becomes the default.

The Laundering Machine

Information laundering describes how a claim travels from a low-credibility source into mainstream summaries. Each retelling by a more respectable venue reduces the urge to check the origin. Over time, the claim appears everywhere and the origin vanishes.

Both engines of the popular tale began in places with no evidential duty. A book of poetry minted the kitchen scene. A pulp magazine minted the terrified log. Neither was produced under investigative standards. Neither offered documents nor names to verify the claims. Yet both supplied irresistible details to any later storyteller.

By the mid-twentieth century, newspaper features on the case were already blending the poem’s imagery into the narrative as if it were a report. Later television segments and popular history websites repeated the chair and the meal as if they had always been in the file. Each repetition increased familiarity. Familiarity made the claims feel safe to repeat again. The cycle is self-fuelling once it starts.

When a detail shows up in print, then on television, then on a museum explainer, readers assume there must be a document behind it. Often there is none. The badge of a new platform acts like a fresh endorsement. That is the laundering mechanism at work. The story earns unearned authority by simply staying in circulation.

The lost logbook worsened the problem. Without the volume, the institutions cannot display a page that undoes the fiction at a glance. Researchers must piece the truth back together from secondary summaries and surviving notes. That lack of a single artefact prolongs the life of the forgery. The vacuum keeps the stage set for the wrong play.

The Narrative That Killed the Truth

The disappearance itself remains a maritime tragedy with a plausible official explanation. Our focus is how that explanation lost ground to dramatic inventions. The displacement happened in steps. A poem reframed a room. A pulp story reframed the men. The mind did the rest.

Gibson’s 1912 verses supplied a crisp picture of crisis. The 1929 pulp item supplied a logbook voice and a plot. The pairing created a complete, self-reinforcing fiction that combined scene and script. It was easy to absorb and easy to pass on. It outcompeted a dull truth.

Two well-known mental shortcuts did most of the lifting. Narrative bias made the story with agency feel richer than an accident report. Cognitive fluency made simple images feel trustworthy. Neither shortcut cares about provenance. Both reward clarity and emotion.

A good story does not need to be true to spread. It only needs to be easy to picture, easy to repeat, and emotionally satisfying. The Flannan case shows how quickly such a story can displace the record. Once inside guidebooks and museum displays, it becomes default memory. The costs are small for the storyteller and large for the archive.

The Final Question

Knowing how quickly narrative can kill a fact, what habits do we need to resist it? Which documents should we demand, and which details should we refuse to repeat until we have them? That is the practical work if we want the record to beat a story.

Method

Muirhead’s 1901 wave hypothesis remains the most evidence-based explanation. This investigation does not propose an alternative cause. It shows how the official account lost ground to inventions, and how information discipline might stop the same process happening again.

Sources

Sources include: primary source documents from the 1900-1901 investigation, including the initial landing report by relief keeper Joseph Moore and the formal inquiry by Northern Lighthouse Board (NLB) Superintendent Robert Muirhead; the two foundational narrative forgeries, specifically Wilfrid Wilson Gibson’s poem ‘Flannan Isle’ (1912) and the fabricated logbook attributed to ‘Ernest Fallon’ in a 1929 issue of True Strange Stories magazine; modern historical analyses and reconstructions of the event by researchers including Mike Dash and Roy Bainton, and summaries from institutions such as the National Records of Scotland; academic papers on the relevant cognitive biases, including foundational work on Narrative Bias and the ‘truthiness effect’ of Cognitive Fluency; and a range of secondary media sources from the 1940s to the present day, such as newspaper features, popular history websites, and television documentaries, used to document the process of information laundering.