Routes, journeys and networks¶
Not everything a gazetteer holds is a place in the everyday sense. A Roman road, a pilgrimage way, a king’s progress through his realm, a river and its tributaries, the cities a merchant wrote to: each is made of places, in an order or in a pattern, and each has a history of its own. PLATO records all of them with the same few ideas, and the World Historical Gazetteer will show them as routes, journeys and networks.
This page explains those ideas. The worked examples then show each one on real sources:
the Antonine Itinerary: two Roman roads from London to the Kent ports, a route;
King John in 1215: the King’s movements around the sealing of Magna Carta, a dated journey;
the lower Idle: a river from Mattersey to the Trent, a network with a direction;
the Datini letters: the post between a merchant’s offices, a network of correspondence between cities.
Three kinds of thing¶
A route is a way through places in order: a road, a sea lane, a pilgrimage route. If it has dates, they say when it was in use, not when anyone travelled it.
A journey (PLATO and the World Historical Gazetteer call it an itinerary) is a route someone actually travelled, with each stop dated as the source allows: arrived on this day, left on that one.
A network is a set of places and the connections between them, in no single order: a river system, a canal network, the cities that exchanged letters.
Each of these is itself a row in the places sheet, with a row in the types
sheet saying which kind it is. The kinds have fixed identifiers, listed on the
vocabularies page: put route, itinerary, network
or segment in type_label, and the matching identifier in type_uri.
Members, in order¶
A place on a route is a member of it. In the relations sheet, a row with
relation_type MemberOf says so: the place in place_id is a member of the
route in related_place_id. Its position goes in sequence, as your source
orders it: 1, 2, 3.
The order belongs to the source, not to the place. Two itineraries of the same road can list its stations differently, and one town can be the third stop on one route and the ninth on another; each row keeps its own source’s order. Two members with the same number are alternatives at that point (two branches of a road). A member with no number is in an order the source does not give. The members of a network usually have none; a river’s reaches, which the water puts in order, can be numbered downstream.
The stretch between two places¶
The road between two stations, or a reach of a river between two
confluences, is often something your source describes in its own right: it
gives its length, or says when it was built, or when it silted up. Such a
stretch is a segment: a row of its own in the places sheet, typed
segment, and a member of its route or network like the stations.
Its ends are relations too: BeginsAt and EndsAt for a segment with a
direction (the way the itinerary runs, the way the water flows), or HasEnd
twice for one without. Its length and anything else your source says about it
go in the properties sheet, like any other fact.
Platforms may leave segments out of lists and searches of places, so that “Road from Londinium to Durobrivae” is not offered as a place to visit.
Connections¶
Sometimes two places are simply connected, with nothing to say about the
stretch between them: two cities that exchanged letters, two ports with a
regular crossing. Then a relations row joins them directly: ConnectedTo if
the connection runs either way, LeadsTo if it runs one way, from the place
in place_id to the other.
Direction is always in the word you choose, never in a separate column. If a
connection is known in both directions but from different sources, record two
LeadsTo rows, each with its own source, not one ConnectedTo.
A source may give figures about a connection: how many letters went from one city to another, how many days they took. Those go in the connections sheet, one figure per row, so that the figure is about the connection and not about either city.
Your project may need a more particular word, such as “flows into” for a
river. The spreadsheets take only PLATO’s own words, so a project’s own kinds
of connection need the JSON format. There a document declares
each one once, in relationTypes, with the PLATO word it narrows, so that
any software that knows only PLATO’s words still follows it the right way.
How long¶
A source often says how long without saying when: “where I stayed six weeks”.
The relations sheet’s duration column holds that length, written in a
standard form: P42D for six weeks (the form has no weeks, so write them as
days), P1D for a day, PT12H for twelve hours. Put the source’s own words
in date as usual. It may be given with dates or without them.
People, objects, events, images and records¶
A place also has a part in things that are not places: the town where someone
was born, the field where a hoard was found, the site of a battle. The person,
object or event is usually described somewhere else already, such as in
Wikidata or a museum catalogue. So a relations row can point to it by its web
address, in related_uri, with a name to show it by in related_label, and
relation_type BirthplaceOf, DeathplaceOf, ResidenceOf, WorkplaceOf,
FindspotOf or SettingOf.
The same goes for what an archive or collection holds about a place: a
photograph, map or drawing that shows it (DepictedIn), or a file, site
record, report or publication about it (SubjectOf). That is what lets a
collection be explored by place. It is different from citing a source: a
source you cite is the evidence for a statement, while “this photograph
shows the site” is a statement in its own right, with its own source,
usually the catalogue that identified it.
Each relations row names exactly one thing it relates the place to: either a
place in related_place_id or an address in related_uri, never both and
never neither. PLATO tools reports a row
that breaks this rule.
What not to enter¶
Platforms work some things out for themselves: a journey’s overall dates from the dates of its stops, a route’s line from its stations. When they export data they mark such values as computed, because they are not evidence. Record what your sources say; do not copy a computed value back into your own data as if a source had said it.