Somewhere in your family there is a person who vanishes from the paperwork. They appear in a 1930s letter, or in the margin of a church record, or on the back of a photograph, and then they never appear again. Most families have at least one. And most of us hold the paper that would explain that person without knowing it, because we don't yet know what to look for.

That is the strange part of genealogy. The document that breaks open a family mystery rarely looks important when you first handle it. It looks like a receipt, a school form, a crumbling certificate in a language you half read. You glance at it, decide it's nothing, and put it back in the shoebox. Years later, a cousin mentions a name, and suddenly that "nothing" is the only proof you have.

Why you can't judge a document by what you know today

Researchers call it a confirmation filter: we notice details that fit what we already believe and skim past the rest. In family history, that filter works against you. You scan the names you recognise and ignore the ones you don't. But the unfamiliar name is exactly where a new branch of the tree is hiding.

So the first rule of scanning genealogy documents is counterintuitive: capture everything, including the parts that seem irrelevant. The witness who signed a marriage record. The neighbour listed on the same page of a census. The address on the envelope. Genealogists lean on this idea constantly. Sometimes called the FAN principle (friends, associates, and neighbours), it holds that people in old records rarely stood alone. Immigrants often travelled, settled, and married among people from the same village, so the stranger beside your ancestor on a page is often the thread that leads back to where they came from.

If you only photograph the line with your great-grandfather's name on it, you throw that thread away.

Records are fragile, and some are already gone

Not every record can be replaced. Most of the 1890 US census was destroyed, first damaged by a fire in a Commerce Department building in 1921 and then lost in the aftermath. Researchers have worked around that gap ever since. It's one of the best-known examples of how a whole layer of documentation can disappear, and it happened to a government archive, not a family drawer.

Your household papers are far less protected. Paper family documents face the usual enemies: acidic paper that yellows and crumbles, ink that fades, folds that split along the crease, and water, which is the fastest of all. A baptism certificate that survived a century in a trunk can be lost in a single flooded basement.

There's a quieter risk too: the person who understands the papers. A grandmother who can say "that's my uncle, the one who changed his name" is worth more than the certificate itself. Once she's gone, the document becomes a puzzle with no key.

Spelling is not a mistake you need to fix

One of the biggest errors beginners make is "correcting" what they scan. They type the name the way it's spelled today, or they tidy a transcription so it looks consistent.

Don't. Before spelling was standardised in official records, and often well after, clerks wrote names the way they heard them. Literacy varied, accents differed, and the same person could appear under three spellings across three documents. Those variants are evidence. They tell you how the name sounded, who recorded it, and which records belong together.

A related myth is worth clearing up. The popular story that officials at Ellis Island changed immigrants' names on arrival is widely rejected by historians. Passenger lists were typically compiled at the port of departure, and inspectors worked from them rather than writing new names. Name changes more often happened later, when people chose to adopt a more local spelling themselves. The practical lesson: preserve the document exactly as written, because the spelling on it is a clue to when and where it was made.

What a good genealogy scan actually captures

A useful scan is more than a readable image. For family history, it has to preserve the original as an object you could return to years from now.

Both sides, always. Backs of certificates carry stamps, endorsements, and handwritten notes. Backs of photographs carry names and dates, which is often the only identification they have.

The edges and the damage. Don't crop tightly. A torn corner, a water stain, or a notary's seal at the margin tells you about the document's history, and a missing piece is easier to understand if you can see where it went missing.

Enough resolution for handwriting. Old scripts get hard to read when pixels run out. Scan or photograph in good, even light and zoom in on faded areas to check they hold up. Glare from glossy or laminated paper can erase a signature, so tilt the page or shift your light source.

Searchable text, with the image preserved. On-device OCR can make a typed or printed record searchable, which helps when you have two hundred pages and need every mention of one village. Treat it with care on old handwriting and unfamiliar scripts. It will misread, so keep the original image as the source of truth and use the text only to find things.

How to keep it findable

A folder named "scans" holding two hundred files called IMG_4471 is almost as lost as the shoebox. Give each file a name that says who, what, and when, such as Kowalski_Anna_naturalization_1938.pdf, and use the date as written on the document even if you suspect it's wrong. Put your doubts in a note, not in the file name.

Group multi-page items into a single PDF so a three-page will doesn't scatter, and keep a simple log of where each original physically lives. Future you, or the cousin who inherits the box, will want to know.

Your next moves

  • Start with the most fragile item today. Pick the single paper that is thinnest, most water-stained, or most faded, and scan both sides this evening before touching anything else.
  • Photograph the back of every family photo you find in the next week. Capture any writing before you worry about the picture itself.
  • Name three files using the pattern Surname_Given_document_year. Don't fix spellings; write them as they appear on the page.
  • Call one older relative and scan while they talk. Hold up a document, record or note what they say about it, and save the note beside the file. Do this before the next family gathering, not after.
  • Make a second copy in a second place. Put the scans on a drive or a trusted backup, and tell one other family member where they are.

Saving the question, not just the answer

Genealogy is a long game, and the person who solves your family's mystery may not be you. It may be a grandchild with better tools and a name you've never heard. What you can do is hand them complete evidence instead of a summary of what you thought mattered.

That's the reason LumenScan was built around keeping pages exactly as they are. It scans on your phone, reads text on-device, and doesn't send your family's papers to anyone's servers, which matters when the documents hold birth dates, addresses, and names of people who can no longer consent. If you'd like to start with that one fragile certificate, you can find it at https://lumenscan.lumenlabs.works.