September 23, 2026
Every lighthouse has a story, and a good number of them are wrong. Not invented outright — legends rarely are — but somewhere between a keeper's diary, a Victorian guidebook and a website that copied both without saying so, dates drift, towers merge into one another and a simple beam acquires a shipwreck, a ghost and a rescue that never happened. Researching a station's past is mostly an exercise in subtraction: working out what can actually be supported. Here are the mistakes that cause the most trouble, and the habits that avoid them.
Mistaking legend for record
The most common error is not a wrong fact but a missing question. A story appears in three books, five websites and a leaflet at the car park, and it begins to feel sourced. It usually isn't. When you keep meeting the same anecdote, ask a different question: where does it first appear?
Trace it back and you often land on a local newspaper piece from the 1930s, or a guidebook written for day-trippers, or a memoir set down fifty years after the event. That doesn't make the story worthless — oral tradition is evidence of a kind — but it changes what you can claim. If the earliest reference you can find is a century later than the event, write "local tradition holds" rather than "in 1861". Wrecks in particular attach themselves to lighthouses: a ship lost on the rocks nearby becomes a ship lost on the lighthouse, whether or not the light was even lit that night. Contemporaneous inquiry reports, inquests and newspaper accounts are where you test that.
A date without a source is a rumour with a number attached.
Dates that don't mean what you think
Lighthouse dates are slippery because a station has several of them, and sources rarely say which one they mean.
Authorised is not built, and built is not lit
An Act of Parliament may authorise a light years before the first stone is laid. The tower may be finished well before the light is exhibited, because the optic, the lantern and the machinery arrived later and by sea. And the date a light was "established" in an authority's list may record when that authority took over an existing private light, not when the tower went up.
The tower is often not the first one
The Eddystone has carried four towers; the one standing today is not the one John Smeaton built. Other stations have been rebuilt on the same rock, replaced by a new tower further along a crumbling cliff, or heightened in stages as the lantern was upgraded. Keep a separate note for each event:
- the first exhibition of the light, and by whom it was shown
- each rebuild, re-siting or heightening, with the engineer if known
- changes to the optic and to the character of the light
- automation, and any later discontinuance — rarely the same date
Keeping these apart prevents the classic error of describing a twentieth-century tower as the work of a nineteenth-century engineer.
The lighthouse, the light and the lantern
Everyday language blurs things that records keep distinct, and the blur creates bad history. A "lighthouse" in a newspaper report might be a light vessel, a harbour light on a pier, a daymark, or a beacon with no light on it at all. The lantern is the glazed room at the top; the optic — often a Fresnel lens of a particular order — sits inside it. A lightship's name and the name of the station it served are not the same thing, because vessels were moved between stations, sometimes more than once.
If your subject is a lightship, the paper trail runs through different collections again. Keep the question "which light, at which date, on which structure?" running quietly in your head as you read.
Leaning on secondary sources — and on each other
Most online lighthouse histories are copied from a handful of printed sources, and copies inherit their errors. Two websites agreeing with each other proves nothing if both came from the same 1970s gazetteer.
Go back to records made at the time. Depending on the station, these include:
- the lighthouse authority's own archive (Trinity House, the Northern Lighthouse Board, the Commissioners of Irish Lights)
- Board of Trade and Admiralty papers, including wreck inquiries and notices to mariners
- Ordnance Survey maps and Admiralty charts in different editions
- local record offices, harbour commissions and estate papers
- newspapers, trade journals and engineering periodicals
- census returns, parish registers, and staff or pension records where they survive
Where an authority holds a station file, a short, specific email to its archivist will often settle in ten minutes what months of scrolling will not.
Photographs, postcards and undated images
Images are the easiest evidence to misuse because they usually carry no date. Postcards were reprinted for decades from the same negative, and photographs appear in books with captions added by someone who wasn't there. Read the picture rather than the label: the shape of the lantern, the number of keepers, the clothing, the vehicles, the boundary wall that is present in one view and absent in the next. Compare those details against dated maps and against the fabric of the building itself. If a photograph shows keepers, ask whether the station was still manned at that time — a single detail can shift an image by thirty years.
Keepers, names and the census trap
Family names are where careful research most often comes unstuck. Common surnames repeat within a service that recruited from the same coastal towns, spellings vary between records, and keepers moved station every few years. A man on census night may be at the light while his wife and children are recorded in the nearest village, so the household you find may look nothing like the family you expected. Follow the person across several records — census, civil registration, parish registers, staff records where they exist — and treat any single document as a lead rather than a conclusion.
A workable research
Photo: Pexels / Pixabay
