Skip to content
LüükCase studies from the harvest field

Protecting people

From Lüük

Many people in case studies from the vault live where naming a person or a village can get someone hurt. This page describes how vault stories are prepared and checked, how historical case studies are documented, and how any case study is taken down.

Two layers

Lüük is kept in two separate places.

The vaultThis site
What it holdsThe full case study, with the real names of people and placesVault case studies with names changed or removed, and historical case studies that use names from their sources
Who can open itA small invited team. Each person signs in with a code sent by email and a second code from an authenticator appAnyone
What a person seesA writer or a reviewer sees only the works they are assigned toEvery published case study

The pages of this site are plain files. They are built from the list of published case studies and from nothing else in the vault. They hold no password or key to the vault. If this site were broken into, there would be no names from the vault here to find.

The name registry

Each work has a registry. It lists the real names in that work's stories and what the public version says in place of each one. People, places, groups and organizations all go on it, along with other spellings and nicknames.

For each name the writer chooses one of these:

  • Another name. A made-up name stands in for the real one.
  • A description. The name becomes words such as "a farmer" or "a market town".
  • Removal. The name is taken out.

The registry holds the real names, so it is the most closely guarded part of the vault. It is never published.

The scrub

When the writer is ready, the vault makes the public version in four steps.

  1. Every name on the registry is replaced, in every spelling listed, in every part of the text.
  2. The generational map is stripped. Names, leaders and places are removed. The shape of the map, the counts and the year each group started are kept.
  3. The writer edits the result so that it reads well.
  4. A check runs on the edited text. It looks for registry names that are still present, email addresses, phone numbers, links, map coordinates, exact dates, text in a script other than the Latin alphabet, capitalized words that may be names, and descriptions that could point to one person without naming them.

A registry name that is still in the text has to be fixed. For anything else the check finds, the writer fixes the text, or a person allows the term and writes down the reason.

The case study cannot be sent for review while any finding is open. It also cannot be sent without a note that says who agreed to the story being told and how they agreed. When it is sent, the database checks the text again for registry names, email addresses, phone numbers and map coordinates, and refuses it if it finds any.

The second person

Every case study from the vault has been read by two people. The person who approves it cannot be the person who wrote it or the person who sent it for review. The database enforces this rule, so a fault in the vault's screens cannot get around it.

The reviewer reads the public version beside the private one, reads each allowed term and its reason, and confirms eight things.

CheckWhat the reviewer confirms
NamesNo real name of a person appears, including in quotes and captions.
PlacesNo village, town, district or landmark is named. The region is broad.
DatesThere is no exact date. Years and seasons only.
DetailsNo detail singles out one person. Examples are the only teacher in a town, the mayor's son, a rare job or a public event.
NumbersEach number is marked as counted or reported, and no number identifies a small community.
MediaNo photo, film, voice, sign or landmark identifies a person or a place, or the media was left out.
ConsentThe people in the story agreed to it being told, and the consent note says how.
Allowed termsEvery allowed term was read and is safe.

The case study is approved only when all eight are confirmed. If one is not, it goes back to the writer with notes.

When an approved case study is published, the database checks it one more time against the registry. If it finds a registered name in any spelling, it refuses to publish.

That last check knows only the names on the registry. A name that nobody registered has to be caught by the check on the draft and by the reviewer. This is the reason a second person reads every case study.

What a published case study still says

  • A broad region, such as Central Asia.
  • Years or seasons.
  • Counts of people, groups and churches.
  • The shape of the generational map.

A film is linked only after a reviewer has watched it for faces, voices, signs and landmarks. This site links to a film and does not play it in the page, so nothing loads from another website unless you follow the link.

Historical case studies

Some case studies are drawn from the public historical record. The people in them lived long ago, and the sources name them. These case studies name people and places as the sources do.

On their maps, each church and group is marked with how well the sources support it.

  • Stated. The sources say so directly.
  • Inferred. A careful reading of the sources supports it, but they do not say it in so many words.
  • Tradition. A later source says so. The source is named.
  • Assumed. A working assumption of this case study, given with its reason.

Historical case studies hold no private names, so they do not go through the vault or the two-person review described above. Each one is kept as a file with the code of this site, and every change to it is reviewed before it is published. To take one down, the team removes the file and builds the site again.

Taking a case study down

Any member of the team who works on a case study can take it off this site. It takes one person and a written reason. The case study is gone from the site the next time the site is built.

To change a published case study, the team takes it down, edits it and sends it through the same review again.

Every page asks search engines not to keep their own copy. Some may keep one anyway, and a reader may have saved the page. Those copies are outside the team's control, which is why the checks come before publishing.

If you are in a case study, or you know someone who is, and you want it changed or taken down, tell the worker who wrote it.

This site

These pages set no cookies and run no analytics. Nothing is loaded from another website, and the fonts are the ones already on your device.

If you choose a light or dark theme, the choice is kept in your browser and is sent nowhere.

The company that hosts the pages can see which pages are requested, as any web host can.