OpenHistoricalMap/Good practice

This is a rough draft of a proposed guideline. Please discuss any areas for improvement on the talk page.

OpenHistoricalMap depends on its contributors to tell the world's geographical history in full. We recognize that historiography is a nuanced and challenging endeavor in academia, let alone for volunteer mappers without any formal historical training. We hope these rudimentary guidelines will help you map history responsibly so that OpenHistoricalMap can remain a useful, durable resource.

Usability

Whenever possible, every feature that has tags should have a start_date=* tag so that data consumers know when to make it first appear. After all, nothing in the database started at the beginning of time. If you do not know a specific start date, set start_date=* to a rough estimate and indicate the start date as a range using start_date:edtf=earliest/latest (with a literal / between two timestamps).

Objectivity

Avoid unfounded speculation. Technically, you can map a feature that starts or ends in the future, but you should only map specific, concrete aspects of the future. For example, if a shop has announced that it will go out of business next month, you should set its end_date=* to next month, even if there is a chance that their plans may change. However, the mere fact that a bridge was built with a 50-year life expectancy does not mean you should set its end_date=* 50 years after its start_date=* – indeed, it is very normal for a bridge to remain in service long after its expected lifespan, and some are destroyed or replaced early.

 To do: For natural features, start dates can be difficult to estimate, and the historic paths of watercourses and waterbodies can also be uncertain. What degree of precision is expected?

Verifiability

Main article: Open Historical Map/Hypothetical vs Verifiable

For any detail you add to the map, it should be possible for another mapper to somehow verify that detail. Some details can be verified "on the ground" as in OpenStreetMap, such as a building's dedication date etched into a cornerstone. Unfortunately, you probably lack a time machine to verify many other kinds of historical details. Instead, you are free to cite secondary and tertiary sources, as well as contemporary primary sources. You may also copy from historical maps as long as you don't infringe on any copyright. Either way, to facilitate verification and avoid plagiarism, please cite your sources using source=* tags on individual features or changesets.

Prefer sources that are competent authorities on the topic at hand and in general (see Gell-Mann amnesia effect).

It is acceptable to map uncontroversial details from memory based on local knowledge. You should explain your connection to the subject in the feature's source tags, a changeset comment, or your user profile, as appropriate. Try to corroborate the details with other sources, because memories can be faulty.

Neutrality

OpenHistoricalMap does not shy away from depicting unsavory historical facts. Do not attempt to paper over disasters, conflicts, occupations, or real-world manifestations of discrimination by deleting accurate details. Your accurate mapping sheds light on the magnitude of these problems so that they are more difficult to ignore or exacerbate. At the same time, a marginalized or victimized community need not be solely defined by what has happened to it; the time dimension enables us to depict a community's normality as well.

Whenever there is a serious dispute about some geographical fact, strive to include every noteworthy claim in the database, tagged with a claimant (claimed_by=*), such that a data consumer can reliably present each viewpoint in isolation.

 To do: How to indicate whether a claim is de facto or de jure?  To do: Do debunked claims belong in the database at all, even if they're noteworthy?

Primary, secondary, and tertiary sources all contain biases. A source's bias could render it unsuitable for inclusion (e.g., some forms of propaganda). Consider whether the source's bias affects its credibility on the specific topic, or whether an overreliance on the source could harm the project's trustworthiness.

As in OpenStreetMap, name=* tags should generally be written in the local language, when known; exonyms go in language-qualified name:*=* tags. This is particularly important when mapping occupied lands, indigenous cultures, or other peoples whose cultures have been suppressed.

 To do: Do we prefer contemporary names or names applied retroactively in historical accounts? For example, should a historic building's name be applied from the time the building was built or only once it was listed by that name in a heritage register?

While OpenHistoricalMap is in a very early state of development, some degree of editorial imbalance is natural and unsurprising. Some mappers contribute broad coverage of a geographical region or time period, while others focus on narrow coverage of a place throughout time. Either approach yields glaring omissions; let these omissions motivate you to keep mapping and invite others to map diligently alongside you!

See also

• Good practice in OpenStreetMap
• OpenHistoricalMap Code of Conduct – also applies to contributors

References

This article is a stub. You can help OpenStreetMap by expanding it.