Technical Introduction to Places
The Concepts
A Place is a geographical and historical context for Names and Locations. Places may have within their core some features of the physical world – a sea, a bay, a river, a mountain range, a pass, a road, a settlement, or an ethnic region – but their primary quality is that, in the words of Yi-fu Tuan [1], they are constructed by human experience. Places may be no larger than a family dwelling or as big as an empire, be temporally enduring or fleeting. They may expand, constract and evolve over time. A place may be unnamed, unlocated, falsely attested, or even mythical.A Location is a current or former, concrete spatial entity. The midline of a river channel is a location. The center of a bridge's span is a location. The perimeter of a walled settlement is a location. Every location belongs to a place. The highest point of a mountain summit, for example, would be a location while the entirety of the mountain: its faces, ridges, couloirs, and forested slopes – and its significance in human history – would be the place context. The location entitled "Temple of Vesta" is but a small part of its context, Rome, while the place entitled "Hafir far west of Araba" provides only the barest semi-anonymous context for its sole desert waterhole location.
A Name is a current or former, abstract textual entity. Like a location, a name belongs to a place. The Πάπρημις of Herodotus 2.59 is one of many with no known locations in the same place. At the other end of the spectrum, Ἀφροδισιάς belongs to a place rich in locations and names.
These Pleiades concepts are somewhat different from those of other conceptual systems in the cultural heritage and geographic information domains. The entity of the CIDOC Conceptual Reference Model (CRM) labeled E53 Place, which
... comprises extents in space, in particular on the surface of the earth, in the pure sense of physics: independent from temporal phenomena and matter,
is almost exactly equivalent to the Pleiades concept of Location. The Pleiades Place has no single equivalent entity in the CRM. Many places are localized (settlements, stations, temples and monuments) and have much in common with the CRM's E27 Site. Others like ethnic territories, areas of centuriation, or mining districts are rather different.
The concept of Feature in the OGC/ISO 19101 system on which many enterprise GIS systems are founded overlaps with both Pleiades Location and to a lesser degree Place. Pleiades Names are nothing like ISO 19101 Features and, as entities in their own rights, are much more than simple feature attributes.
The Objects
Instances of the Pleiades Location, Name, or Place entities are maintained in a database and come to life (so to speak) as needed as objects in the Pleiades web application. The relationship depicted with a line and a filled diamond between a place object and a location or name object in the abridged Unified Modeling Language (UML) diagram below is containment.
Every object has a short name (or id) unique within its containing object. The container for places is itself an object and its id is "places". Our convention is that ids contain only lowercase letters of the Latin alphabet, numbers, and dashes. Places have numerical ids such as "1043", those of locations are derived from their titles ("barrington-atlas-location", in many cases, or "castellum"), and the ids of names are formed by transliteration of their attested spelling ("ad-fines", for example).
Easy disambiguation and unique identifiers are properties of this object model. There are many instances of "Alexandria" names in the ancient world. The object with id "alexandria" in the context of place "727070", or places["727070"]["alexandria"] in a notation like that of the Python or Javascript programming language syntax, is distinguished from another "alexandria" object in the context of place "59669" (places["59669"]["alexandria"]).
The other important aspect of an object's existence is workflow state. Pleiades has a simple publication workflow that keeps drafting-state objects semi-private and protects published-state objects against corruption.
The Web
Pleiades users do not interact directly with the objects described above. The interface through which researchers and the public (including machine users like web crawlers) access our information follows the architecture of the World Wide Web [2]. Every instance of a Pleiades entity has a corresponding Resource on the web. The resource is identified using a Uniform Resource Identifier (URI) that maps to object names in their containment context. The resource corresponding to the object places["727070"]["alexandria"] is identified by and addressed with the URI http://pleiades.stoa.org/places/727070/alexandria.In other words: Pleiades provides a web page for every place, name, and location. The content of these pages are generated for a user from objects stored in the database, filtered according to their workflow state and the user's roles. Reviewers get a somewhat richer view than authenticated users, who in turn get to see more than anonymous users.
Pleiades places are thoroughly crawled and indexed by search engines of the web and can be found readily via queries like "aphrodisias+pleiades".
The Catalog
The catalog of Pleiades objects can be searched via simple or advanced forms and is also written every morning to files that can be opened in a spreadsheet program.References
[1] Y. Tuan, "Place: An Experiential Perspective," Geographical Review, vol. 65, Apr. 1975, pp. 151-165.
[2] Jacobs, I. and N. Walsh, "Architecture of the World Wide Web, Volume One," W3C Recommendation, 15 Dec. 2004. http://www.w3.org/TR/webarch/