PLATO: a guide to organising data about places

Why PLATO

Research on historical places produces a great deal of evidence: names read in documents, positions taken from maps, statements about what kind of place something was and what it belonged to. That evidence is usually kept in whatever shape one project needed at the time, which makes it hard to reuse, to combine with other projects’ work, or to hand on to others.

PLATO, the PLace ATtestation Ontology, gives that evidence a shape. It is a published standard for recording what sources say about places, designed so that data in its shape can meet the FAIR principles, and organising your data in its shape gives you three things.

A structure to think with.

PLATO asks the same few questions of every piece of data: what is it about, what does it say, which source says so, and when does it apply? Answering them is often the most useful step in organising a dataset, because it separates what your sources say from your own interpretation of them, and shows where your evidence is thin.

It also leaves room for what historical evidence is really like. Sources that disagree are recorded side by side, and nothing forces you to choose between them; a scholar’s judgement that one statement contradicts, supports or supersedes another can be recorded as evidence in its own right. And PLATO keeps apart the different ways of not knowing, rather than collapsing them all into “uncertain”:

• how sure you are of a statement, which better evidence could change;

• how firmly the source itself says it: a claim it only reports (“it is said”), hedges, or raises and leaves undecided;

• a date known only within limits, or given only as “c. 925” or “before 1300”;

• a place with no sharp edge, such as “the Levant”, which no evidence will make crisp;

• a location given only relative to somewhere else, such as “two leagues north of the ford”;

• evidence that is certain but weaker in kind, such as a name found only inside a person’s name;

• a reference an editor has had to work out from “ibid.”;

• a form read from a damaged or misread document, judged for how well it was read;

• a mention that could be one of two places, recorded as alternatives, at most one of them right, rather than as two separate claims;

• a match with a record elsewhere that is close, but perhaps not exact.

A source can also say that something was not so, such as “no market here”, and that is recorded as a denial, not left out.

Ready for other systems.

Data in PLATO’s shape can be read by any software that implements the PLATO standard, without a conversion written for your project alone. You do not need to know in advance which systems those will be.

The World Historical Gazetteer

The World Historical Gazetteer (WHG) is moving towards implementing PLATO, so that data organised this way can more easily be welcomed there.

Its Map your Data tool, reached from the WHG site, reconciles your places with established reference gazetteers such as Pleiades, GeoNames and Wikidata: it finds the record for the same place, so that you can take its coordinates where yours have none. With WHG’s agreement, Map your Data is moving out of WHG and into PLATO tools, and the PLATO tools version is the one now being tested. There it is the first guided workflow in Methodos: the regions your places lie in are matched level by level, then the places themselves, and each place is located from its match or drawn, including over a georeferenced historical map, before the result is checked and written out in PLATO. Like the rest of PLATO tools it is experimental, runs entirely in your browser without uploading anything to WHG, and can be used whether or not you contribute to WHG.

Nothing is flattened.

Every statement keeps its source and its date, so your data can sit alongside other people’s evidence about the same places without either overwriting the other, and a reader can always ask who says so, and when?

This guide shows how to put your data into that shape. You do not need to know anything about ontologies, JSON or linked data: if you can fill in a spreadsheet, you can use PLATO.

PLATO and FAIR data

The FAIR principles ask that research data be Findable, Accessible, Interoperable and Reusable, by people and by software. Funders and repositories increasingly expect them. PLATO is built to help, and here is how, principle by principle.

Findable.

Every place, source and statement in PLATO’s shape can have a permanent web address of its own, so it can be found, cited and linked to. PLATO itself has one, https://w3id.org/plato, and each release is archived on Zenodo with its own DOI.

Accessible.

PLATO’s definitions are openly licensed (CC BY 4.0) and published at that address in the standard formats software asks for. Data in PLATO’s shape is plain spreadsheet tables, JSON or linked data: open formats that anyone can read without special software, and that PLATO tools checks and converts.

Interoperable.

PLATO is a formal standard (an OWL ontology) built on widely used vocabularies, among them PROV for provenance, DCAT for datasets, SKOS for vocabularies and CiTO for citations. Your places can be matched to shared gazetteers such as Pleiades, GeoNames and Wikidata, so your data joins up with other people’s.

Reusable.

Every statement keeps its source, its date, how sure anyone is of it, and how firmly the source itself says it. That is the provenance someone else needs before they can trust and reuse your data. A licence can be stated for the dataset and for each source.

PLATO gives the structure; a few steps remain yours when you publish:

• choose a licence for your dataset and state it;

• give it permanent addresses, under a base you control and will keep, such as your own w3id address or a DOI, not a temporary one;

• deposit it in a repository that gives it a DOI, such as Zenodo;

• describe it: a title, who made it, and what it covers.

In the spreadsheets, the licence, the base of your addresses and the description all go in one place, the about sheet; in JSON, in the document’s gazetteer header.

Taken together, those steps and PLATO’s structure make a dataset FAIR. PLATO tools reports what your description still lacks, writes the files a repository needs, and makes a website and permanent addresses for your places.

Where to start

I have a list of places, or notes about places, to organise.

Read Organising data in spreadsheets, then follow a first dataset step by step.

I want to understand the ideas first.

The ideas in five minutes explains places, attestations, sources and dates with one example.

My sources are philological or archival, with copies, editions and “ibid.”

The place-name survey example shows how to record manuscripts copied later, names found only inside personal names, editorial headwords and inferred references.

I have annotated texts or maps in Recogito.

Annotations from Recogito shows how the places you linked become a PLATO dataset.

My data is a road, a journey, or a network of rivers or letters.

See Routes, journeys and networks, with four worked examples.

I want to check a file, or convert it to another format.

Use PLATO tools: drop the file on the page, and see checking, converting and comparing for what it tells you. It runs in your browser, so nothing is uploaded, and it works at any size. To check many files at once, it also runs from the command line.

My dataset is ready, and I want to publish it.

See publishing your dataset: a report on what its description lacks, the files for depositing it, a website for every place, and permanent addresses.

I produce data with my own software, or work with RDF.

See JSON formats and Linked data.

My sources are statistical tables: census volumes, returns.

See Statistical tables.

I have met a word I do not know.

The glossary explains the terms used in this guide.

PLATO is developed by the Pelagios Network’s Place Working Group and the World Historical Gazetteer. The formal definition of every term is in the ontology reference; this guide is the friendlier way in.