Structured data
Statements, identifiers, labels and coordinates from Wikidata, which places them in the public domain. Every statement on a profile links to the statement it came from.
Honest about the evidence
How this site reads its sources — written as the mechanism rather than as a policy, because a mechanism can be checked.
The rest of the statements are imports, and the site says so on every row rather than in a paragraph here.
Where each claim
comes from.
Sources and licences
01 · Licences
Tracked separately, and credited separately on every page that uses one.
Statements, identifiers, labels and coordinates from Wikidata, which places them in the public domain. Every statement on a profile links to the statement it came from.
The opening paragraph quoted on a profile, and the article read at /wiki/, come from Wikipedia under share-alike terms. The article link is the attribution, because the article’s history is its list of authors. Text is quoted, never merged into our own prose, and never reconciled across accounts that disagree.
Each file on Wikimedia Commons carries its own licence, named under the picture with its author. A file whose licence cannot be read is dropped, with its subject’s picture: being on Commons is not a licence.
A modern translation or edition is not public domain because the work behind it is ancient. Where this site shows a witness to a text, it names that witness’s own date and its own rights.
Structured data, article text and pictures come from three places under three different licences, and a single line saying “content from Wikimedia” would be wrong about all three. They are tracked separately and credited separately on every page that uses one.
02 · Relationships
Every relationship belongs to a mode, and the mode decides how it may be drawn.
The drawing code is given the mode and never the relationship, so a teacher cannot be rendered as a father by an oversight. Under each group on a profile is what that relationship does not claim — one example of each:
Parentage as a named source tells it. Not an independently established biological relationship, and different accounts disagree.
“parent of” does not imply an independently established biological relationship · agreement between accounts
A documented human family relationship. Not a succession mechanism, and not religious transmission.
“parent of” does not imply a right to an office · transmission of a teaching
Who held an office after whom. Not parentage, not discipleship, and not universally recognised — offices are contested.
“held the office” does not imply recognition by every tradition · an uncontested claim
Teaching passed from one person to another. Not biological ancestry and not automatic institutional succession.
“taught” does not imply biological ancestry · institutional succession · agreement between teacher and student
Association with a tradition or school. Traditions branch, overlap and re-merge; this is not a clean single-origin tree.
“belongs to” does not imply personal belief at every point in a life · the practice of a population
One figure treated as another in some context. Not literal identity in all contexts, and usually a claim by a third party.
“identified with” does not imply literal identity in all contexts · that either tradition made the identification
A transmission argued from evidence. Not shared symbols, and not a short path through a database.
“influenced” does not imply shared symbols as evidence · a short path through a database
Two figures share a trait. Not proof of influence and not proof of common origin.
“shares an attribute with” does not imply influence · common origin · that either tradition draws the comparison
A ruler or institution supported something. Not the belief of a population, and not the patron’s own practice in every period.
“was a patron of” does not imply the religion of a population · personal adherence in every period of a life
Where something is, or is documented. A location is not by itself evidence of activity.
“is located in” does not imply that anything is documented to have happened there
A work, its editions and its witnesses are distinct; their dates are not interchangeable.
“appears in” does not imply that the work is the only account · that the work is contemporary with what it describes
A domain, symbol or epithet as sources record it.
“is associated with” does not imply an exclusive domain · the same association in every account
A figure is venerated somewhere or by someone. Not a claim about how many people did, or when they stopped.
“is recognised as” does not imply recognition by any other church or tradition · a judgement about the historicity of the life · the same status in every branch of the church that confers it
Events Wikidata records in a subject’s history, and the day a festival is kept. Not a complete history, and the day a festival falls on is not when it began.
“recorded event” does not imply that these are all the events in its history · that Wikidata ranks the events by importance
03 · Dates
A single date field is the most common way a reference work about religion misleads, so there is no single date field.
It happened, or is attested, then. A life, a reign, a building.
A work was composed, copied or printed then. The Theogony’s date is not Aphrodite’s, and a manuscript’s date is a floor on the work rather than a date for it.
Time inside an account. It has no historical counterpart and cannot be compared with one.
A deity’s earliest surviving mention is not its birth. A shrine’s building date is not the beginning of a tradition. A missing end date means unknown far more often than it means ongoing, and the pages say which.
Which clock a date is on comes from the kind of thing it belongs to, not from the field it was read into, which is why a text cannot accidentally acquire a historical date.
Precision survives to the comparison. A date known to the century stays an interval, and a year filter answers certainly, possibly or no date recorded — three answers, drawn three ways, rather than a midpoint invented so a marker had somewhere to sit. Internally years are astronomical so that arithmetic across the era boundary is correct; nothing outside one module prints a year, so no reader ever sees a year zero.
04 · The atlas
Every shape on the atlas carries a role, and the legend names it.
A blank area on the atlas means no evidence in this collection. It does not mean the absence of belief.
The roles above are the whole list: a site, a find, a community, documented activity, a route, political context, or a reviewed estimate. A point for where an object is kept today and a point for where it was venerated are different claims that look identical, so they are never the same mark.
There is no layer of religious distribution, and there will not be one derived from what this site has. Current borders, a ruler’s affiliation and a count of points cannot produce one: that needs a dataset with its own methodology. Where such a layer is unsupported the atlas says so and everything else keeps working — which is honest coverage rather than a deferred feature.
05 · Standing
There is no checkmark anywhere on this site. Each claim carries its standing as a word.
Taken from Wikidata as it stands. Nobody here has read it.
The statement cites something outside Wikimedia. That citation has not been checked here.
A person here read it and signed it with a date. This label is unreachable in the code without such a record.
What a tradition teaches, presented as a teaching rather than as a finding.
Sources disagree, and the disagreement is kept.
The source said circa, allegedly or presumably, and this does not turn that into a date.
Worked out from other records here. A chain of records is a fact about this database, not an argument about history.
The import did not say which tradition or source the claim holds for. Most imported kinship rows truthfully say this.
Wikipedia and a Wikidata statement imported from that article are one source, counted once. Where a statement’s only reference is a Wikimedia project, it is not counted as sourced — which is why the number on each profile is lower than the number of references.
There is no checkmark anywhere on this site. “Imported”, “has references” and “reviewed” are three claims about how much anyone has checked something, and one tick would flatten them into an endorsement two of them have not earned.
06 · Privacy
No account is needed to read, to save or to keep a history.
You are never asked what you believe, and nothing on this site infers it: what someone reads about religion can imply what they believe, or are considering, or are afraid of, and the right design is to make the question unanswerable rather than to answer it carefully.
What you save and where you have been are two keys in this browser’s local storage. There is no endpoint they could be sent to. Settings counts what this browser holds and clears it.
Some requests do leave a page, and none carries your shelf or your trail. Pictures load from Wikimedia Commons, the reader fetches its article from Wikipedia, and a search this site cannot answer is also put to Wikidata, so Wikimedia’s servers see which file, article or search they were asked for, as they would for anyone reading Wikipedia. Each page also loads HistorIQly’s account script, which asks the account server which plan this browser is signed in to; it names this site, not the page you are on.
07 · Upstream
Discovered from Wikidata, harvested once and committed — so a published page cannot change under a reader.
Records are discovered from Wikidata by class and by religion — deities through the subclass tree, places and figures through a search index ranked by incoming links — then filtered against each item’s own statements. The harvest runs on a developer’s machine and its output is committed, so a deploy never talks to Wikidata and a published page cannot change under a reader.
Next
Counts of records, citations, places and undated entries for every collection, with the gaps an editor wrote down.
Counted,
never scored.