Pleiades is an online database of spatial information modeled on the long tradition of gazetteers. It is most useful to people interested in Greek and Roman material but also includes a growing amount of information concerning other ancient cultures.
The user interface is simple and intuitive. First, type the name of a place into the search bar. Then, if Pleiades recognizes the name, it provides a selection of peer-reviewed information about that place: for example, alternate names, relevant citations, and chronological periods during which the place was active.
For casual visitors, Pleiades functions primarily as a search engine for ancient places. For specialists in GIS (Geographic Information Science), it also serves as an invaluable storehouse of spatial data. In this review, I focus on introducing Pleiades to a wider audience, rather than those in the Classics community who specialize in spatial data. Accordingly, I first provide a broad overview of the project, then offer some guidance for new users, and finally discuss the strengths and weaknesses of the site.
What is Pleiades?
Pleiades is among the oldest online projects committed to the study of the ancient world and predates the popularization of the term “digital humanities.” Pleiades contributors began accumulating place records in 2008, but the project’s roots go back even further—to the development of the Barrington Atlas (2000) at the UNC-Chapel Hill Ancient World Mapping Center (AWMC). Since that time, the library has grown to around 37,000 entries.
Pleiades makes an important distinction between two apparently synonymous terms: “place” and “location.” Place is the basic unit of the Pleiades database, and the site is designed to include the broadest possible variety of places. Each entry in the database is considered a place and given a unique place identifying number (PID). Once a place record is created, information can be added or edited continuously, by any of the roughly 150 members of the Pleiades community, so long as the changes pass an editorial review.
The simplicity of the term “place” is potentially misleading. Pleiades defines a place as “any locus of human attention, material or intellectual, in a real-world geographic context.” Places run the gamut of scale, from large geographic regions (e.g., Iberia), to individual structures (e.g., a tomb, altar, or villa). The database also includes both unnamed places (e.g., unnamed settlement, unnamed bridge, unnamed well, unnamed aqueduct) and places that are named in ancient sources but have not (yet) been located (e.g., Rhoda, Arene).
Roughly 33,000 places (89% of the database) also have a “location,” which in this context refers to real-world geographic coordinates. Pleiades provides these as latitude and longitude in decimal format (Datum WGS84). Users should pay close attention to the confidence rating of the location data they plan to use. In some instances, the location data is only a rough estimate. In others, it is accurate to the meter.
Figure 1 shows the geographic distribution of Pleiades places. Each colored pixel represents around 300 square kilometers of surface area. The map demonstrates a important characteristic of the Pleiades database: it is predominantly Mediterranean, with a strong tendency for regions within the boundaries of the Roman Empire.
Figure 1. Distribution of Places in Pleiades Database. Map by author.
Slightly under thirty thousand places (80% of the database) are also associated with one or more chronological periods. Pleiades recognizes over 100 such periods, many of which overlap chronologically and/or apply only to specific regions, such as the Middle-East or Egypt. Only 9 periods are associated with more than 1% of Pleiades places, and these periods are very general (e.g., Modern, Roman, Late-Antique). The frequency with which these commonly used periods appear provides a fair characterization of the chronological distribution of the Pleiades library (Figure 2).
Figure 2. Chronological distribution of Pleiades places. Figure by author.
Guidance for New Users
Pleiades can be found at pleiades.stoa.org (Figure 3) or by searching for the phrase “Pleiades Stoa” on any popular search engine. The central feature of the front page is a search bar. Type a few characters into the search bar, and a list of possible results appears, each of which includes the matching place name, a short description, alternate names (if applicable), and the PID, which also functions as a link to a page with additional information.
Figure 3. Pleiades Front Page.
A search for “Ephesus“, for example, returns only one result, entitled “Ephesus, Ephesus/Arsinoe(ia) = 599612.” Clicking on the PID takes the visitor to the results page, which offers location coordinates, alternate names and spellings, and connections to places in the surrounding area. This information is followed by a list of ancient and modern sources that reference the site. The results page also shares the original source of the information and a variety of useful file formats for the more technically inclined.
Pleiades searches do not work the same way as a search engine like Google. Some best practices will improve the quality of results. Using the search bar on the home page — reached by clicking “home” at the top of the front page (Figure 4) — will grant a greater degree of control. As a rule, searches from the home page work best using a proper name rather than an attribute. A search for “Temple of Apollo,” for example, will not produce a complete list of temples dedicated to Apollo in the Pleiades database. It will return some places with temples dedicated to Apollo — primarily because they have “Temple of Apollo” as part of their name or description — but key locations, such the temple at Didyma, will not appear. If you know a place’s proper name, add a space after the name for best results. Searching for “Roma” returns over twenty thousand results, because records that use the word “Roman” in their description are included. Adding a space after Roma eliminates over 99% of the results, and the imperial capital is at the top of the results list.
Searching for a partial name can be useful if the user is uncertain about the spelling or transliteration of a place. For example, searching for “Narb” returns Narbonne, Narbon, Narbasoi, Narba, and Narbata, among other places, mostly associated with the name “Narbonensis.” This technique, however, only works for the beginning of place names. “Arb” does not locate any of the records above.
Figure 4. Pleiades Home Screen
There is also an advanced search method, available through the link just under the search bar on the site’s front page. This tool permits the visitor to search the catalog according to a number of criteria, such as categories of place (e.g., Bath, Settlement, Island), and to search within a region defined by a southwest and northeast coordinate (latitude/longitude; Figure 5). There are too many search parameters available to discuss them here, but this area is worth exploring. The advanced search feature can help visitors sift through the Pleiades library, but user beware — the advanced search tools rely on the details included in the place records, and not all records are equally detailed or similarly constructed. The advanced search options achieve better results when used individually. For example, an advanced search for sites categorized as temples, tagged with Mithras, and located within a box that includes Italy (southwest 37.51, 9.01, northeast 47.16, 13.95) returns zero results, but simply searching for sites tagged with Mithras returns sixteen places, four of which are in Italy.
Figure 5. Advanced Search Home Screen.
Strengths and Weaknesses
Pleiades is a valuable resource — certainly worth the minimal investment of a quick search if you want to learn more about a place in the Ancient Mediterranean. The gazetteer often provides information that is difficult to find elsewhere, and each record has undergone some level of scholarly peer review. The site has something to offer a broad range of users, from visitors with only a casual interest, to historians with specific research questions, to digital humanists with GIS experience and skills. An informal poll among fellow Classicists suggested that many use the site to learn more about locations they encounter in the course of their research. Pleiades is an excellent resource for this sort of query. To those interested in more substantial scholarship, the entries often provide useful citations of both ancient sources and modern scholarship.
Pleiades has shortcomings, many of which have to do with the project’s scope and how it has developed over time. The massive library was, for the most part, built by incorporating large collections of data from other projects, which were compiled according to different sets of priorities. This process has produced inconsistent coverage of the ancient world and a level of detail that varies substantially depending on the source of any particular place record. Additionally, Pleiades is in large part a product of scholarship in the Classics community, and it reflects the wider bias in Classics towards the Mediterranean in general and especially the areas around Athens and Rome. The database is probably better understood as a record of what Classicists tend to study, rather than a representation of ancient activity in the Mediterranean. One should not assume that the absence of a record means a place does not exist, nor that the place is not in the database under a different name. For obscure places, where arguably the database would be most useful, the available information tends to be sparse. Finally, and perhaps most importantly, one should always assume that Pleiades provides only partial information about any place. The records are highly selective; they contain only information that contributors have chosen to include.
It should be noted that the challenges described above are common among historical gazetteers, and Pleiades has negotiated them better than a static publication could. Despite its grandfatherly status in terms of digital humanities projects, the team behind Pleiades continues to improve both the website and its underlying data. The user interface, for example, was recently updated. The site could be annoyingly slow in the past, but that problem has been remedied. Pleiades has linked its data to other digital humanities projects and online resources around the web, such as Peripleo and Antiquity À-la-carte. And it is now possible to build geographic context for sites using the “connections” feature.
More than a decade in, Pleiades continues to grow, develop, and refine its database. After an initial period of rapid growth, expansion has slowed (Figure 6). Since 2012, when the site started to track not only additions to the database, but also edits to existing records, the latter have substantially outpaced the former (Figure 7). Taken together, these two figures show that, for some time now, the Pleiades team has been investing more in quality than quantity. This is good news for those who were frustrated in the past by slow load times or poor search results. Pleiades is worth visiting again.
Figure 6. Growth of the Pleiades Database. Figure by author.
Figure 7. Number of Modifications per Month (Blue Diamonds) vs Number of Additions per Month (Red Circles). Figure by author.
DESCRIPTION: Historical gazetteer
NAME: Roger Bagnall, Tom Elliott, Lindsay Holman, Richard Talbert, et al.
PUBLISHER: The Ancient World Mapping Center, The Stoa Consortium, and The Institute for the Study of the Ancient World
COLLECTION TITLE: none
DATE CREATED: 2008–pres.
DATE ACCESSED: May 2020
AVAILABILITY: Free, no subscriptions
RIGHTS: Creative Commons Attribution 3.0
CLASSIFICATION (1-10 keywords): mapping, databases, Greek history, Roman history, linked open data, GIS
Header image: Pleiades front page.