Placing Names
164 pages

Vous pourrez modifier la taille du texte de cet ouvrage

Obtenez un accès à la bibliothèque pour le consulter en ligne
En savoir plus
Obtenez un accès à la bibliothèque pour le consulter en ligne
En savoir plus
164 pages

Vous pourrez modifier la taille du texte de cet ouvrage


Well before the innovation of maps, gazetteers served as the main geographic referencing system for hundreds of years. Consisting of a specialized index of place names, gazetteers traditionally linked descriptive elements with topographic features and coordinates. Placing Names is inspired by that tradition of discursive place-making and by contemporary approaches to digital data management that have revived the gazetteer and guided its development in recent decades. Adopted by researchers in the Digital Humanities and Spatial Sciences, gazetteers provide a way to model the kind of complex cultural, vernacular, and perspectival ideas of place that can be located in texts and expanded into an interconnected framework of naming history. This volume brings together leading and emergent scholars to examine the history of the gazetteer, its important role in geographic information science, and its use to further the reach and impact of spatial reasoning into the digital age.

Preface / Peter K. Bol
Introduction / Ruth Mostern, Humphrey Southall, and Merrick Lex Berman

Section 1: What is a Gazetteer
1. Gazetteers Past: Placing Names from Antiquity to the Internet / Ruth Mostern and Humphrey Southall
2. Gazetteers Present: Spatial Science and Volunteered Geographical Information / Michael F. Goodchild
3. Gazetteers Global: United Nations Geographical Names Standardization / Helen Kerfoot
4. Gazetteers Enriched: A Conceptual Basis for Linking Gazetteers with Other Kinds of Information / Ryan Shaw

Section 2: Using Gazetteers in Combination
5. International Standards for Gazetteer Data Structures / Raj Singh
6. Place, Period, and Setting for Linked Data Gazetteers / Karl Grossner, Krzysztof Janowicz, and Carsten Keßler
7. The Pleiades Gazetteer and the Pelagios Project / Rainer Simon, Leif Isaksen, Elton Barker, and Pau de Soto Cañamares
8. Historical Gazetteer System Integration: CHGIS, Regnum Francorum, and GeoNames / Merrick Lex Berman, Johan Åhlfeldt, and Marc Wick

Section 3: Exemplars
9. Building a Gazetteer for Early Modern London, 1550-1650 / Janelle Jenstad
10. Digitally Exposing the Place Names of England and Wales / Paul Ell, Lorna Hughes, and Humphrey Southall
11. Standardizing Names Nationally: The Work of the United States Board on Geographic Names / Michael Fournier
12. The Yeosi Project: Finding a Place in Northeast Asia Through History / Youcheol Kim, Byungnam Yoon, Jonghyuk Kim, and Hyunjong Kim

Section 4: Doing History with Gazetteers
13. Mapping Religious Geographies in Chinese Muslim Society / Mark Henderson and Karl Ryavec
14. Core-Periphery Structure of the Nobi Region, Central Japan, With Reference to the Work of G. William Skinner / Tsunetoshi Mizoguchi
15. Gazetteer GIS and the Study of Taiwan Local Society and its Transition / Pi-ling Pai and I-Chun Fan
List of Contributors



Publié par
Date de parution 08 août 2016
Nombre de lectures 0
EAN13 9780253022561
Langue English
Poids de l'ouvrage 12 Mo

Informations légales : prix de location à la page 0,0025€. Cette information est donnée uniquement à titre indicatif conformément à la législation en vigueur.


David J. Bodenhamer, John Corrigan, and Trevor M. Harris, editors
Imagined Landscapes: Geovisualizing Australian Spatial Narratives
Jane Stadler, Peta Mitchell, and Stephen Carleton
Deep Maps and Spatial Narratives
Edited by David J. Bodenhamer, John Corrigan, and Trevor M. Harris
Geographies of the Holocaust
Edited by Anne Kelly Knowles, Tim Cole, and Alberto Giordano
Toward Spatial Humanities: Historical GIS and Spatial History
Edited by Ian N. Gregory and Alistair Geddes
Troubled Geographies: A Spatial History of Religion and Society in Ireland
Ian N. Gregory, Niall A. Cunningham, C. D. Lloyd, Ian G. Shuttleworth, and Paul S. Ell
Locating the Moving Image: New Approaches to Film and Place
Edited by Julia Hallam and Les Roberts
The Spatial Humanities: GIS and the Future of Humanities Scholarship
Edited by David J. Bodenhamer, John Corrigan, and Trevor M. Harris
Enriching and Integrating Gazetteers
Edited by Merrick Lex Berman, Ruth Mostern, and Humphrey Southall
Indiana University Press
Bloomington and Indianapolis
This book is a publication of
Indiana University Press
Office of Scholarly Publishing
Herman B Wells Library 350
1320 East 10th Street
Bloomington, Indiana 47405 USA
2016 by Indiana University Press
All rights reserved
No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying and recording, or by any information storage and retrieval system, without permission in writing from the publisher. The Association of American University Presses Resolution on Permissions constitutes the only exception to this prohibition.
The paper used in this publication meets the minimum requirements of the American National Standard for Information Sciences-Permanence of Paper for Printed Library Materials, ANSI Z39.48-1992.
Manufactured in the United States of America
Cataloging information is available from the Library of Congress.
978-0-253-02244-8 (cloth)
978-0-253-02256-1 (ebook)
1 2 3 4 5 21 20 19 18 17 16
Preface/Peter K. Bol
Merrick Lex Berman, Ruth Mostern, and Humphrey Southall
Section I: What Is a Gazetteer?
1 Gazetteers Past: Placing Names from Antiquity to the Internet
Ruth Mostern and Humphrey Southall
2 Gazetteers Present: Spatial Science and Volunteered Geographical Information
Michael F. Goodchild
3 Gazetteers Global: United Nations Geographical Name Standardization
Helen Kerfoot
4 Gazetteers Enriched: A Conceptual Basis for Linking Gazetteers with Other Kinds of Information
Ryan Shaw
Section II: Using Gazetteers in Combination
5 International Standards for Gazetteer Data Structures
Raj Singh
6 Place, Period, and Setting for Linked Data Gazetteers
Karl Grossner, Krzysztof Janowicz, and Carsten Ke ler
7 The Pleiades Gazetteer and the Pelagios Project
Rainer Simon, Leif Isaksen, Elton Barker, and Pau de Soto Ca amares
8 Historical Gazetteer System Integration: CHGIS, Regnum Francorum, and GeoNames
Merrick Lex Berman, Johan hlfeldt, and Marc Wick
Section III: Exemplars
9 Building a Gazetteer for Early Modern London, 1550-1650
Janelle Jenstad
10 Digitally Exposing the Place Names of England and Wales
Paul Ell, Lorna Hughes, and Humphrey Southall
11 Standardizing Names Nationally: The Work of the US Board on Geographic Names
Michael R. Fournier
12 The Yeosi Project: Finding a Place in Northeast Asia through History
Youcheol Kim, Byungnam Yoon, Jonghyuk Kim, and Hyunjong Kim
Section IV: Doing History with Gazetteers
13 Mapping Religious Geographies in Chinese Muslim Society
Mark Henderson and Karl Ryavec
14 Core-Periphery Structure of the Nobi Region, Central Japan, with Reference to the Work of G. William Skinner
Tsunetoshi Mizoguchi
15 Gazetteer GIS and the Study of Taiwan Local Society and Its Transition
Pi-ling Pai and I-Chun Fan
List of Contributors
T HE EBB AND flow of social life takes place in space. People are always somewhere. Governments claim territory; commerce moves goods between places. But social life also unfolds alone the axis of time. We live in the present, but the organization of places and spaces in which we live change over time; taken in aggregate the social organization of space is remarkably unstable. Places are created but they disappear. Places have names, but those names are not constant, and they are not singular. Named places have locations, but those locations do not stay fixed and the territory they encompass can expand and it can shrink.
We often act on the assumption that the situation at the current moment will last forever. In daily experience we know where we are going and we accommodate the changes in names, locations, and boundaries of the spaces in which we live. The relevance of these changes to us varies depending upon our interests. Although the tax collector pays attention to changes in territory that affect the tax base and the taxee needs to pay attention to the entity to which taxes are owed, few are aware of all the changes that are taking place in the spatial organization of their surroundings. And if one asks not only what has changed but also when it changed, personal experience and memory prove sorely inadequate. And the further into the past we go the harder it is to identify changes. This is of course, why we have authoritative reference works such as gazetteers.
The chapters in this volume are inspired by the view that contemporary gazetteers, whether produced by governments or private efforts, have given scant attention to time as an attribute of space, or to the complex historical and cultural contexts within which places have been named. Taken together they call for enriched gazetteers, consider various models, and propose paths forward. An enriched gazetteer, as this book frames it, is one that includes the best possible information on when changes in place names took place. But this is not quite so simple. Just as the representation of space depends upon scale, so does the identification of points in time. Did it happen on this day? In this year? In this century? What are the many names that a place has had over time? And there are names that refer to imagined places, or places for which the location in time and space is extremely uncertain.
Thus, the complexities of tracking historical place names goes beyond adding a time element, such as recording sequences of names used over time for a given feature or locality. From the perspective of historical research, each instance of a historical place name has its provenance: we need to know not just when a given name was used but by whom, and dates appear as part of the documentation of sources. Linguistic place name researchers can extract additional meanings from the earliest forms, for example revealing past landscapes by identifying names associated with woodlands, while archivists and librarians need place name authorities, which identify preferred forms among many variants. Accuracy in gazetteers must be equally dependent on the attestation of names, dates, and documentation of sources as it is dependent on locations.
The case for enriched and integrated gazetteers is easy for geographically minded historians to make. They want to know where events happened, the boundaries of a district a population or tax record applied to, what places were linked by roads and railways, and so on. Historical geographic information systems such as those for Great Britain, China, Germany, the United States, and a number of other national territories, have shown that such information can by systematically collected and organized.
A historical gazetteer has value to historians, but there is a reason to use the more inclusive term enriched gazetteer. This volume begins to answer the question of how agencies and private organizations can integrate change over time into the current stream of toponymic authority systems, and how multiple gazetteers can be used together. In a digital environment the focus is often on staying up to date; there may be no paper records to archive. Our present is the future s past; without keeping track of changes that are now taking place, we lose the ability to carry out a longitudinal analysis of how our social and environmental contexts have changed. This is not a matter for historians alone. Planning cannot afford to ignore the trajectories of development. This volume, the first book-length work to address these matters, is therefore highly welcome.
Many of the chapters in this volume began as papers for a special three-day track on Gazetteers for Space-Time System Integration held at the 2010 Annual Meeting of the Association of American Geographers. There were further sessions at the 2011 meeting. The many participants from universities, government agencies, and industry in those sessions have informed the work here and most papers originate from those panels. As the organizer of that event, I am most gratified to see that this volume has come to fruition.
Peter K. Bol
T HE EDITORS WISH to thank the Association of American Geographers (AAG), and their Executive Director Doug Richardson, as well as the Harvard Center for Geographic Analysis, and its then-Director Peter Bol, for sponsoring the two workshops at the AAG s 2010 and 2011 Annual Meetings, where first versions of most of these chapters were presented. We also wish to thank Megan Overbey and Wendy Guan for helping organize these meetings, and the many individuals who participated in them.
We also wish to thank Indiana University Press for their patience and assistance in shepherding this book through several years of writing and production. In particular, David J. Bodenhamer offered prompt and excellent advice, intellectual direction, and logistical support at every stage of the process, as did the other editors of the Spatial Humanities series, John Corrigan and Trevor M. Harris. Janice E. Frisch and Raina Polivka have been our calm and expert guides at the press. Karen Kemp reviewed the manuscript with enthusiasm and offered suggestions that considerably improved the project.
Linda Hill was the person who first and most rigorously engaged the three of us in conversations about gazetteers more than ten years ago, and her advocacy for the importance of well-structured gazetteers continues to inspire our thinking. We three editors first met one another, as well as a number of the people who are named in these acknowledgments and whose chapters appear in this book, though meetings of the Electronic Cultural Atlas Initiative in the 2000s, and we warmly acknowledge ECAI Director Lewis Lancaster and Codirector Michael Buckland for organizing an important series of conferences about the spatial humanities.
We gratefully acknowledge the support of the UK Arts and Humanities Research Council for Humphrey Southall s work on this book, as part of their Digital Transformations Big Data program (award AH/L01002X/1).
Finally, we would like to thank one another and all the authors who appear in this book for the several years of excellent and collegial collaboration and stimulating conversation that have now come to fruition.
Merrick Lex Berman, Ruth Mostern, and Humphrey Southall
G AZETTEERS, SPECIALIZED INDEXES of place names, are invaluable works of reference and infrastructure for almost all spatial humanities projects. Since the 1990s, humanists, geographers, and information scientists have recognized the importance of standardizing and integrating these sources. After all, gazetteers are the basis for much of the spatial search and visualization that specialists and the public have come to take for granted. Unfortunately, most current large gazetteers are byproducts of topographic mapping that often consist simply of place names and coordinates. They do not adequately support much of the local, vernacular, historical, and cultural scholarship that the spatial humanities requires.
The purpose of this volume is to recognize and bring together new gazetteer research that better serves the needs of the spatial humanities. As the second chapter in the book explains, there was once an earlier tradition of richer, more encyclopedic gazetteers that described places as well as locating them. This book is inspired by that tradition of discursive place making as well as by contemporary approaches to digital data management that have revived the gazetteer and guided its development in recent decades. It is now possible, conceptually and technically, to reunite the two traditions. We have put this book together so that readers can explore the ways in which that work is being done.
The past fifteen years have seen a renewed interest in gazetteers that associate more extensive and descriptive content with place names, which have become a major research focus of the spatial humanities. One line of gazetteer research, largely from the perspective of information science, has focused on designing new frameworks and access mechanisms for existing gazetteer content, most of which was originally created by national mapping agencies, and is now sometimes enhanced through crowdsourcing. Meanwhile, humanists have been active in the field and the archives, researching places and place names, vernacular and historical, and designing computer systems around that content. That is the work that we are showcasing here.
In recent years, a series of international meetings brought together scientists and scholars from both gazetteer research traditions to collaborate with one another. The largest of those meetings, cosponsored by the Association of American Geographers (AAG) and the Harvard University Center for Geographic Analysis, were held at the 2010 AAG annual meeting in Washington, DC, and the 2011 AAG annual meeting in Seattle. Together they featured over a hundred authors. This book is organized around the most substantial gazetteer research presented at those workshops, supplemented by articles from other leaders in the field whom we invited to submit papers to this volume. Although there are people doing extraordinary work conceptualizing, developing and integrating gazetteers and performing research based upon gazetteer content, little of that work has been published. As a result, it is difficult for members of the spatial humanities gazetteer community to find and cite one another, to train students, or to identify best practices. We hope that this book will fill that gap, and that it will be useful to gazetteer developers, spatial humanists, and anyone seeking an overview of this important and emerging field.
Gazetteers Defined
The discussion that follows is intended as an admiring tribute to earlier work on gazetteer standards development, which remains significant to this day. However, it is also a critical analysis that explains why we believe that the new research presented in this book, which comes from a humanistic and semantic web perspective, is such an important new direction in the field.
Most simply, a gazetteer is a list of places. As Raj Singh s chapter in this volume explains, standards organizations and librarians have long worked on digital gazetteer standards. Foundational work by the International Standards Organization defines two alternative conceptual schemas for digital gazetteers. ISO 19111 covers spatial referencing by coordinates 1 and ISO 19112 covers spatial referencing by geographic identifiers. 2 The second of these allows for gazetteers that locate one set of named entities relative to a second set of named entities, by being within, between, or adjacent to them. Coordinates are not required, and a simple list of geographic names would meet the standard if all were contained within some territory.
The ISO defines conceptual reference models, while two other standards define precise formats for digital gazetteers. The Gazetteer Service Profile of the Open Geospatial Consortium s Web Feature Service standard, WFS(G), enables geographical features to be selected from a spatial database based on feature s names, location and type. 3 The Alexandria Digital Library (ADL), in work led by Linda Hill, defined a Gazetteer Content Standard, 4 a Gazetteer Service Protocol describing how another computer can query an ADL-compliant gazetteer over the internet, 5 and a Feature Type Thesaurus. 6 This foundational work brought gazetteer design into the computer era and made it possible to efficiently build standardized gazetteers and integrate them with one another and with other digital resources.
ADL standards development has a large impact on the digital humanities research presented in this volume. The following description and discussion is organized around ADL s three core gazetteer elements. Each of these-feature name, classification, and spatial location-can launch a research agenda for enriched and humanistic gazetteer developers. By examining these elements, we aim to define what a gazetteer is, and also to introduce the challenges that inspire this book. The present volume introduces many domain-specific exemplars that diverge from the ADL standard. It also proposes the development of Linked Open Data standards as a way to link gazetteers such that related concepts can be clustered in a less rigid way than the ADL standard requires.
First, the ADL standard requires that each feature include a name. In practice, each feature also has a unique identifier, usually a number. Since the same name may refer to many different places, it is hard to be sure exactly what qualifier makes a place unique. Boston unqualified most probably identifies the capital of Massachusetts, but there is a Boston in every continent except Antarctica and in over twenty US states. Boston, England most probably refers to the town in Lincolnshire, but there is another in Yorkshire. Further, the same place can have many different names, distinguished partly by historical period and by language but often used interchangeably. The ADL Gazetteer Content Standard allows for the inclusion of alternate as well as preferred names, but giving primacy to one particular form is often deeply problematic. This vast ambiguity of place names greatly complicates gazetteer construction.
Second, in the ADL standard, every entry is also classified by being assigned to a type, and the ADL Feature Type Thesaurus specifies 210 preferred terms. However, the thesaurus is language specific and also culturally specific. It is defined only in English and is in practice closely based on the set of features symbolized on US Geological Survey maps, including terms such as court houses and capitol buildings. Given that almost everything has a location, and can therefore be classed as a geographical feature, any particular feature-type thesaurus can be criticized for being incomplete or lacking detail. However, and especially when working with historical events rather than landscape features, the priority may be to bring together the different associations of the same named location, not carefully separate them. In particular, a fixed typology leads almost inevitably to a simplistic treatment of administrative units. For example, the ADL system starts with countries, 1st order divisions, and goes down to 4th order divisions. It adds statistical areas to cover almost everything else. This does not effectively reflect the actual political geography of any nation or empire.
Third, every ADL entry must include a spatial location. One nonissue is how to represent these: any global gazetteer, or gazetteers for continental states like the United States, will usually hold locations using the World Geodetic System or WGS 84, meaning decimal degrees of latitude and longitude; many smaller countries define their own local coordinate system based on a local flattening of the globe, such as Britain s OSGB national grid; but converting between these is very well understood. Less clear is whether to represent everything as points; the ADL content standard allows for entries to be alternatively represented by lines or polygons. However, historical and cultural gazetteers may include many names that certainly refer to places, but ones for which we have only very limited information about location. The most obvious examples are semi- or wholly mythic places such as Avalon, Mount Meru, or Eldorado, but historical sources such as taxation lists contain plenty of prosaic examples as well.
Behind these issues, and a key question for many of the chapters which follow, is the issue of space versus place. Cultural geographer Yi-fu Tuan, who has spent his career explaining how the concept of place emerges from lived experience, has famously asserted that the number of places in the world, that is the locations that someone cares enough about to describe or name, is potentially infinite. 7 The enriched gazetteers that this volume propounds are intended to more effectively model the kind of cultural, vernacular and perspectival idea of place that Tuan alludes to than current gazetteers are able to do.
Because existing modern gazetteers derive from topographic mapping, they are grounded in a geospatial approach that prioritizes physical features which exist in the landscape, and whose locations and types are often far more significant than their names. By contrast, humanities researchers and cultural heritage workers commonly start, not with the landscape but with texts, and consequently with places that are largely defined through their names and associated discursive and cultural information about them; places that change over time, are contested among many constituencies, and rarely have one authoritative descriptor. This calls for a very different approach to gazetteer building than the ADL approach propounds, one that is geosemantic rather than geospatial, and one focused on systematically identifying all possible named places. Over time, many events will happen in any locale of much significance, and various features of differing ephemerality will be named after it, but the place itself cannot be fundamentally associated with any one feature type, and its location, if known at all, will be hazy.
Gazetteers Enriched
So, the purpose of this book is to make the case for enriching gazetteers beyond the ADL framework, and especially to do so from the perspective of historical and cultural scholarship, global comparison, and spatial humanities agendas. The chapters we have assembled specify at least four directions for enriching gazetteers. Each one allows place name lists to be integrated with other kinds of information. The first is to identify and link gazetteers to one another and to other resources that are themselves rich in place names. The second is to provide a framework for integrating gazetteers with information from specific domains, such as historical statistics or tables of historical events. The third is to add a time dimension to gazetteers through both dates and named time periods. The fourth is to open gazetteers to user comment and annotation. Through such extensions the gazetteer itself becomes more clearly a work of scholarship. All of these directions suggest a gazetteer development agenda that is quite different from that specified by ADL, and which raises new challenges of design and infrastructure.
First, the ADL and other gazetteer data standards described here do not meet the usual expectations of scholars because no information is included on sources. Each entry is an unsupported assertion that a certain name is associated with a particular location. This may be unproblematic when the gazetteer records current names and is authored by a national mapping agency, but poses clear problems with crowdsourced gazetteers, most of whose contributions come inevitably from visitors to areas rather than residents, and is nonsense once we start assembling variant names from diverse sources. As Michael Fournier from the US Board on Geographic Names explains in his chapter in this volume, even official agencies work with vast quantities of documentation when they assign place names.
One example of a gazetteer-based information system that is enriched according to the principles we are laying out here is the Pelagios project, described in chapter 7 . It focuses on gathering name attestations, each of which links a place, identified by the web address of its entry in an existing online gazetteer, with a geospatial document, meaning an old map, historical gazetteer, or itinerary in which the place is named. The source is also identified by a web address, and where possible this should link, not to a catalogue or bibliography entry, but rather to a digital copy of the source itself. Here the potential is to create, not just a gazetteer with variant names, but an integrated corpus of geographical knowledge-a specialized kind of knowledge organization system, as Ryan Shaw puts it in his chapter in this volume, such that places in the online gazetteer directly hyperlink to textual and cartographic representations of those places, making the gazetteer a name index to the map collections. 8
Similarly, the Great Britain Historical Geographic Information System (GIS) and its web site, A Vision of Britain through Time, incorporates a large collection of historical travel writing with a two-way relationship with the gazetteer: variant place names have been harvested from the writers, but the gazetteer s place pages then provide access to place-references within the writing. 9 Systematically linking to source definitions incorporates dates of publication or attestation into the gazetteer, creating a naming history. This approach is described in detail in a number of the chapters in this book.
Second, the gazetteer can become an organizing framework to which a wide variety of other information is linked. Some of the clearest potentials are with statistics, especially historical statistics: census reports, taxation listings and many other sources consist mainly of columns of numbers but usually preceded by what are clearly geographical names. However, these are usually not the places that appear on topographic maps and in travel writing but some kind of area , sometimes defined for statistical reporting but often with some administrative function.
With modern data we would expect to work, not with a gazetteer, but with a Geographical Information System, and the relevant digital boundaries can be downloaded from national mapping agencies or statistical offices. However, as we move deeper into the past boundary data become progressively more problematic: more than forty or fifty years back, or more recently in less developed countries, born digital boundaries will not exist and the researcher will need to digitize paper maps; more than a hundred to two hundred years back, there will be no systematic paper boundary maps; going further back, settlements did not necessarily adjoin, instead trailing off into uninhabited areas, so boundaries were not necessarily defined at all; and we will encounter some areas, possibly associated with abandoned settlements, which cannot be located at all.
Somewhere in the this progression even a basic gazetteer of named places linked to point coordinates has a role, and even where boundary maps exist, the cost in money or the researcher s time may make a quicker and simpler approach preferable: a current small project to map intra-urban data from the British 1961 census will be partly gazetteer-based for cost reasons. Further, the previous section described only a minimal implementation of the ADL Gazetteer Content Standard, and a fuller implementation can include, not just a point location, but also dated sequences of boundary polygons and explicit hierarchic relationships, although as noted previously the ADL Feature Type Thesaurus does not enable us to distinguish between the multiplicity of kinds of district existing at roughly the same spatial scale but in different periods and with different functions. Of course, once we add a large number of polygons to a gazetteer it starts to become a GIS, and we can hold it only in a database with object extensions.
Although situations where we cannot locate statistical data more precisely than with a point coordinate taken from a conventional gazetteer, we should generally avoid directly linking the data to the place record: almost always, the data actually relate to a currently unknown polygon, which further research may be able to reconstruct, or possibly simulate. In most cases, we have some knowledge of the kind of area a dataset covers: settlements plus some wide hinterland or market catchment ; settlements plus the areas the inhabitants farm; just the built-up area; or, where the data relate to specific corporate bodies or are essentially decreed by particular corporate bodies, particular buildings. It would be very poor scholarship to associate all these different scenarios simply with places.
The Great Britain Historical GIS is the clearest example of indirectly linking statistical data to a gazetteer. The highest level consists of about 20,000 places, untyped but mostly towns and villages, and all with point coordinates. About 75,000 units exist below the places, with a very detailed and extensible typology; these can have detailed boundary polygons, but are required to have only a hierarchic relationship, enabling them to be added to the system based only on their inclusion in a statistical listing. Finally, at the base, each of c. 14m statistical data values are linked to the units. The resulting system has all the usual capabilities of a GIS for twentieth-century data, but can also hold data from the 1334 Lay Subsidy for Vills located only as points, or not at all. It also includes Urban Labour Markets as a unit type, holding nineteenth and twentieth century unemployment and wage rate data which are fundamentally point-based. 10 Although at heart a gazetteer, the resulting system describes places as well as locating them, and one key advantage over a conventional GIS is its ability to present statistics as time series as well as map them. This is of course a single monolithic computer system, but again the larger potential is for gazetteers to serve as knowledge hubs for data distributed over the internet-an infinite knowledge space.
An enriched gazetteer must go farther than modeling just modern places and their historical names. In addition it must explicitly represent change over time. Thus, one research and development challenge is to incorporate dates and time periods from many modern and historical global systems. Consequent issues include the standardization of dates derived from different calendar systems, the common usage of named time periods in historical sources, and the problem of temporal scale. All of these historical or temporal modeling challenges are unavoidable in the construction of historical gazetteers, and here we survey cutting edge implementations.
There are, for example, several possible methods for standardizing the dates being recorded in the world s many historical calendar systems, but very few have actually been implemented. One successful case integrated the calendars of China, Japan, and Korea by harmonizing those related but unsynchronized calendar systems based on Julian Day Numbers. 11 The consistency of the Julian Day Numbers enables the indexing of any cyclical binary day from the Chinese calendar, as well as looking up any corresponding day in the Japanese or Korean calendars. Mapping particular cyclical binary days to Julian Day Numbers (JDN) also provides a method for defining the time spans of named time periods. We could, theoretically, make use of the same mapping to Julian Day Numbers to construct a timeline for all the various calendars in human history, but this becomes progressively more difficult as one adds more systems with their own distinct dating logics. When we look at original source texts, we find that the majority of historical dates are not necessarily stated in calendar dates at all, but are more often than not referenced by means of named dates, named events, or named time periods. The nature of the source materials subsequently drives the compilation of gazetteers in the direction of integration with chronologies of events, calendar systems, and the cross-indexing of named time periods.
An enriched gazetteer, in one way of framing it, is a list of events and eras that are located in both time and space. As the cultural geographer Doreen Massey puts it, place is the meeting up of histories in space. 12 Although the study of named time periods is an emerging field of research not specifically covered by the present volume, it is a conceptual space directly intersecting with the construction of historical gazetteers. For example, PeriodO has been established to create a gazetteer of period assertions, which combines both chronological and geographic extents of historical and archaeological periods. 13 Moving toward the semantic web for enriched gazetteer development allows for more flexibility than traditional indexes, such as library catalogs can permit. 14 It instead takes advantage of an open linked data ecology for the definition and retrieval of information about historical time periods. 15
The definition of historical periods raises thorny issues about accuracy and fuzziness. To give just one example, the term Civil War refers to 1642-1651 in England but 1861-1865 in the United States. Just as the normalization of geographic areas in concrete terms, such as coordinates in Cartesian space, poses difficulties when dealing with uncertain locations and areas in space; so too does the uncertain date or span of time pose problems when we attempt to pin that fuzzy information to begin and end times using a system such as the ISO 8601 date standard, or to Julian Day Numbers. Dealing with this issue in the context of digital humanities has resulted in some interesting solutions, such as the definition of inner and outer boundaries of possible dates, and consequently, the computational challenge related to managing spans of fuzzy historical time periods is being met with novel solutions. 16
This book does not deal with this work in detail, but we mention it here to explain how research on enriched gazetteers is being addressed by experts with domain knowledge that spans geographic information systems, semantic web ontologies, and more traditional historical methods. 17 A part of the vision of this book is to showcase and applaud these collaborations and publicize the results they have achieved thus far. The study and practical development of gazetteers has become further entwined into these new approaches, and as a result we can see a greater number of working implementations that demonstrate how the character of historical places can be inferred from multiple assertions in time and space, from a series of events, and by indirect references to named time periods. 18
In the end, enriched gazetteers matter because of what they allow us to do. At the intersection of research on modeling both geographies and time periods, such gazetteers have become a meaningful component of systems that attempt to characterize and answer such questions as: which civil war are you talking about? What areas are covered by the particular Bronze Age being referred to? The ways that places change can (and should) be modeled as series or a cloud of events, each of which exerts a change to the characteristics of the place. In this sense the places described in historical gazetteers may merge, split and transform in ways that are more like communities than individuals. The introduction of modeling time for gazetteers expands and enhances our conception of places into systems with complex morphologies. Just as a time period might relate to multiple events or locations, our concept of a place in historical gazetteers might need to be considered as a cloud of attestations including place names, time periods, and historical sources, rather than individual localities that existed for uninterrupted spans of time.
The Organization of This Book
The remainder of this book is organized into four sections: section 1 : What Is a Gazetteer? ( chapters 1 - 4 ), section 2 : Using Gazetteers in Combination ( chapters 5 - 8 ), section 3 : Exemplars ( chapters 9 - 12 ), and section 4 : Doing History with Gazetteers ( chapters 13 - 15 ).
Section 1 expands on the issues that we have raised in this introduction. It further exploring the basic notion of what a gazetteer is, has been, and could become. It investigates gazetteers past, present, global, and enriched. Ruth Mostern and Humphrey Southall explore the early history of gazetteers, arguing that while the twentieth century gazetteer may have been the map s poor relation, in earlier periods textual representations of geographical knowledge were more important and far more widely used than the cartographic. Michael F. Goodchild summarizes more recent work in both geographical information science and neogeography, and the role of gazetteers in collapsing the old dichotomy between space and place. As he explains, gazetteers are tools of a platial world in which names, rather than positions, provide linkages. Helen Kerfoot writes about global gazetteer integration from an official perspective, describing the United Nations work on place name standardization and the need for it. She identifies impediments that continue to stand in the way of the vision of a global gazetteer, setting up the need for new directions in technical and conceptual research in this field. Ryan Shaw, an information scientist, relates gazetteers to other kinds of knowledge organization systems, and explores how they can maintain their conceptual autonomy from similar indexes while becoming both broad in what they cover and deep in what they describe.
Section 2 consists of four chapters about standards and infrastructure development work aimed toward integrating multiple gazetteers. Raj Singh traces international gazetteer standards development, culminating with the sharing-centric and linked data oriented Open Geospatial Consortium Points of Interest standard. The next two chapters further explain and advocate for a linked data approach to enriching gazetteers. The first, by Karl Grossner, Krzysztof Janowicz and Carsten Ke ler, discusses general principles for linked open data gazetteers and advocates for the ideal coverage for a global cultural and historical gazetteer. The second, by Rainer Simon, Leif Isaksen, Elton Barker and Pau de Soto Ca amares, reports on the work of the Pelagios project, which integrates gazetteers not by merging them, but by semantically linking them, growing a globally navigable graph of documents and places that establishes connectivity through common references. The last chapter in the section, by Merrick Lex Berman, Johan hlfeldt, and Marc Wick, confronts the practical challenges of name and record matching across three fairly different gazetteers.
The second half of the book focuses on building and using gazetteers in particular historical and geographical contexts. Section 3 presents four different enriched historical gazetteers, focusing on progressively larger world areas and attuned to different objectives. Janelle Jensted s work on early modern London demonstrates how to develop a gazetteer for a very small but historically significant area, and explains how to build a hybrid historical gazetteer that is integrated with encoded literary texts. Paul Ell and Lorna Hughes describe the long-term project of the Survey of English Place-Names, started in the 1920s and still ongoing, and the DEEP project to construct an online database from the Survey s varied reports. Michael Fournier describes the work of the US Board on Geographic Names. He demonstrates the detailed work and policy infrastructure needed to maintain a national gazetteer and to transition it into the digital realm. Finally, Youcheol Kim, Byungnam Yoon, Jonghyuk Kim, and Hyunjong Kim describe the construction of a large historical gazetteer for all of northeast Asia, with its many kingdoms and shifting imperial domains. They are creating a gazetteer that is both a freestanding digital resource and the framework for a print historical atlas.
Section 4 demonstrates that historians can produce substantive scholarship by treating gazetteers as research sources in their own right. All three chapters in this section concern different parts of East Asia, and all three are based on research with difangzhi , encyclopedic printed gazetteers from the Chinese tradition. In addition, all three chapters take the work of the American anthropologist G. William Skinner as a significant reference point. Mark Henderson and Karl Ryavec study the spread of Buddhism and Islam into western China through the middle and late imperial era (roughly eighth to nineteenth centuries CE). Tsunetoshi Mizoguchi analyses central place geography in central Japan in the Edo Period (1603-1868 CE). Pi-ling Pai and I-Chun Fan have gathered place names from historical and modern Taiwan, beginning with the seventeenth century, integrating text analysis and field survey methods. They have been able to develop new insights about past landscapes under regime change by studying place name etymology.
1 . International Standards Organization (2007).
2 . International Standards Organization (2003).
3 . Open Geospatial Consortium (2012).
4 . Alexandria Digital Library Project (2004).
5 . Jan e (2009).
6 . Alexandria Digital Library (2004).
7 . Tuan (1977).
8 . . See Southall and Pridal (2012).
9 . Southall (2014).
10 . Southall (2011, 2012).
11 . Bingenheimer et al. (2008).
12 . Massey (2005).
13 . Rabinowicz et al. (2014).
14 . Petras, Larson, and Buckland (2006).
15 . Doerr, Kritsotaki, and Stead (2005).
16 . Ore (2013).
17 . Janowicz (2010).
18 . Grossner and Meeks (2014).
1 Gazetteers Past: Placing Names from Antiquity to the Internet
Ruth Mostern and Humphrey Southall
In recent decades, and especially in a computing context, gazetteers have often been defined as simple pairings of place names and coordinates, sometimes including feature types as well. However, in earlier centuries, texts organized around named places included a much richer range of information and occupied a more prominent position relative to that other and more familiar expression of geographical information, the map.
Today, the map apparently reigns supreme. Gazetteers are not just less familiar to most people than maps, they have become largely ancillary to them, so the archetypal gazetteer is usually derived from digital-mapping systems. It is little more than a dump of labels and their associated coordinates, and its primary purpose is as an index to a set of maps. Today s print gazetteers are indexes at the backs of road atlases and tourist guides, while online, GeoNames describes itself as a geographical database rather than a gazetteer, and Google Maps and its rivals simply present users with unlabeled search boxes that are intended to generate maps, though these, of course, access gazetteers. 1
As we go back in time, the dominance of the map becomes less clear, but so do definitions. For instance, expressing coordinates graphically on a map is unambiguously different from listing them in a gazetteer index. However, an itinerary, which is a list of places in order of visit, shares a topological logic with the visual representation of a route on a map, which itself may be heavily annotated with text. Historical East Asian place name encyclopedias are organized according to the spatial and hierarchical structure of administrative geography rather than the alphabetical format typical of the European tradition, so they too link the structure of the text with the spatial relationships encoded in it.
A systematic account of how all the earth s past cultures have documented places and their relationships is far beyond the scope of this book, and beyond our expertise. Indeed, there would be little academic literature for us to draw on if we did wish to write such an article. While there are scattered articles about particular gazetteers or gazetteer traditions, there is no historical overview about gazetteers more detailed than the Wikipedia article on the topic, although that can be recommended. 2 Maps, mesmerizing visual and ideological artifacts that they are, have distracted scholars from the fact that they are only one aspect of the global and historical field of strategies for spatial discourse, orientation, organization, and wayfinding.
What follows can therefore be only a promissory note for future research, with three main elements: first, a general discussion of the practicalities of geographical writing relative to mapmaking in past times; second, an exploration of the Chinese tradition of administrative gazetteers; and third, an account how the European geographical text tradition evolved from linear itineraries into alphabetical gazetteers before vanishing almost completely at the start of the twentieth century.
Place Description and Map Making
It is by now a commonplace to state that maps have always been artifacts of persuasion as much as tools for spatial orientation. They are a special way of communicating spatial dominion with an arresting visuality, simplification of real-world phenomena, and aerial perspective that grant a sensation of inevitability and permanence to the landscapes they depict. 3 However, there are many scenarios for which maps are not the most effective tool for supporting navigation and spatial decision making. Historically, and especially before the modern era, they were used in limited ways, and they have served only to communicate about certain kinds of spatial arrangements. However, because they are so visually compelling, their existence has made it difficult for geographers and historians to think about the other genres of spatial knowledge organization with which maps have always coexisted. In fact, maps were problematic both to create and to use for most purposes during much of human history, and name indexes and itineraries were more common than maps.
One common way to represent the experience of space and place is in the form of narratives of journeys. Itineraries bring structure and meaning to records of travels and serve as guides for others. The most basic itinerary is simply a sequence of named locations, and the second most basic adds journey times measured in days. Such listings can be absolutely without error even without the use of surveying equipment or timepieces. By contrast, for most of human history maps were always and necessarily simply wrong. Considered as coordinate data, well into the nineteenth century maps lacked the accuracy for wayfinding or cadastral purposes. 4 Even the production of accurate sketch maps requires visits to elevated viewpoints usually away from major routes and the possession of large sheets of paper. Since maps were not only hard to produce and store but also unreliable, they were relatively rare, and few people knew how to read them. Maps existed, to be sure, but they were generally surrounded by text or viewed with expert interlocutors, they were often used for display rather than practical planning, and they were rarely made to be carried.
In an important article, geographer Michael Curry has examined spatial discourse in classical Greece (Curry, 2005). He explains that there were then three ways of thinking spatially about the world. Choros was the field of describing names and regions, and it is the basis of what we now think of as the gazetteer. Topos was the term that encompassed travels, itineraries, and the ways that discrete places existed in relation to one another. Only the third concept, Geos , was focused on mathematically oriented maps of continuous space. In other words, in the Ptolemaic imagination, for most purposes the world was envisioned as a collection of named places, not as abstract space, and therefore texts, not maps, were the most common forms of representing it. Curry notes that to this day, many phenomena are routinely and effectively represented as lists of named places or addresses, not as geographical maps, and as an example he explores postal delivery routes.
The Pelagios Project, described in a later chapter of this book, has undertaken a systematic survey of all surviving geographical writing from ancient Greece and Rome. At the time of this writing, forty-four documents from the Latin tradition have been identified and six documents have been identified from the broadly earlier Greek tradition. The earliest is Homer s Illiad (c. 760-710 BCE), a narrative of the Trojan War. Strabo s Geographica (c. 20 BCE-23 CE), an encyclopedia of geographical information, is spatially comprehensive and organized thematically. 5 Pausanias s Description of Greece (100-200 CE) is a systematic work of chorographic reference. Each of its ten volumes describes a different region of Greece in the form of a tour, but with many digressions that focus on the natural and human history of each area rather than Pausanias s personal experiences as a traveler. 6 The Periplus of the Erythraean Sea is a mid-first century CE Greco-Roman itinerary describing navigation and trading opportunities throughout the Indian Ocean region and the Red Sea as far as the mouth of the Ganges and overland to the Han capital at Luoyang.
In the Latin tradition, the Antonine Itineraries , dating from the early third century CE, lists places along Roman roads, with distances. The inclusion of a place is an indication of importance, listing in sequence makes it relatively easy to associate Roman names with modern locations, and the multiplicity of routes means we can start to treat this as a systematic enumeration of places of a certain importance; however, there is no descriptive information at all. 7 The earliest Christian itinerary is the Itinerarium Burdigalense , describing a pilgrimage from Bordeaux, France, to the Holy Land and back in 333-334 CE, almost purely through a very detailed list of the places through which the pilgrims passed. 8
Catherine Delano-Smith explains that in medieval and early modern Europe, people of all social classes were highly mobile and there was significant demand for information about routes, but this continued to be supplied mainly as written itineraries, not maps, while famous maps of road networks, like the Peutinger Map and the Gough Map, were intended for armchair readers, not travelers. Even after the invention of map printing, maps were rarely attached to practical travel guides until the late seventeenth century, and when maps started to appear more frequently, they served the needs of commercial traffic, not private travelers. 9 The commonest medieval assemblies of geographical information were tax lists and inventories of property holdings, organized hierarchically. For example, the Domesday Book of 1086 groups places and people together according to the feudal lord who owned them, and these holdings often did not form contiguous blocks (Darby, 1977).
Maps are ideal for simplifying and organizing the abstract expanse of space, but text does better at conveying information that differentiates particular and similar places from one another. 10 As Paul Carter explains in the classic spatial history Road to Botany Bay , the first stage of eighteenth-century European colonization of Australia was a concerted and deliberate practice of place naming. It was by naming places and writing a narrative about them that James Cook and his contemporaries developed a story that made the unknown space they encountered into a set of places they could possess and inhabit. It was the names and the text about them, not the maps themselves, that clarified and ordered the events of exploration into an account of conquest. 11 As Carter emphasizes, spatial history begins and ends in language (Carter, 1987, p. xxiii).
In short, texts that catalogue and list named places along with their locations, relationships, and other characteristics have been at least as ubiquitous and significant as maps, despite being less visible to historical geographers. People who came of age in Europe or America in the twentieth century lived during the brief reign of the road map. However, that abundant, portable, and workaday exemplar of cartography, depicting a novel landscape of paved roads built for automobile travel, was a serious historical outlier (Delano-Smith, 2006; Akerman, 2007). With that exception, maps were often made to be mnemonic, entertaining, or hortatory, and when they were practical (as they were, for instance, during wartime and for property surveys), they were most often fragile, precious, and ephemeral objects in manuscript form. Some maps represented cosmic, metaphorical or fully imaginary worlds. Until well into the industrial era, these were the kinds of purposes that maps served best, at least at the regional scale. 12 The historical record reinforces Curry s conclusion that in most times and places, gazetteers and itineraries were the main tools for workaday spatial decision making, reasoning, and wayfinding. Maps were too political, too difficult to make, too unwieldy to transport, and too difficult to preserve for them to be used for these purposes.
Case Study 1: Maps and Gazetteers in China
Cartography has an illustrious history in China, beginning with exemplars from the fourth century BCE. The famous third-century BCE silk maps from the Mawangdui tomb site in Hunan include topography, military settlements, and administrative seats. Classical and medieval texts make consistent reference to maps. Dozens of extant stone-carved and block-printed maps date from the Song dynasty (960-1276 CE), in one famous case even including a graticule and a depiction of the empire s river systems with an accuracy that rivaled that of the modern era. However, as Cordell Yee demonstrated in his masterful survey of Chinese cartography, Chinese maps were rarely intended to stand alone. Whether carved on stone or embedded in books, they were generally accompanied by extensive documentary material that described in text the same landscapes that were depicted on the maps. 13
Moreover, as in other locations in the preindustrial world, workaday maps were a source of frustration. Eleventh-century documents detail unsuccessful efforts, spanning decades, to create accurate maps of the Chinese-Tangut battlefields on the northwestern frontier, only to find that the maps rapidly became obsolete or went missing before they were even filed. During the next century, efforts to create land registration maps in the aftermath of war with the Jurchen Jin regime met the same fate. In both cases, text documents fared better than large, fragile maps, which were difficult to make and reproduce. 14
Gazetteers were a different matter. Their history in China has been documented for English-language readers in important articles by historians James Hargett (1996) and Peter Bol (2001), which are the basis for much of the following discussion, and Ruth Mostern (2011) has written about the spatial discourse of Chinese gazetteers. 15 The fourth-century BCE Book of Documents ( Shu jing ), one of the Chinese Five Classics, includes one section, The Tribute of Yu ( Yugong ) that describes the provinces of China, their commodities and tax responsibilities, and their water courses (Mostern, 2011, pp. 61-69). Official dynastic histories, which each ruling house compiled to summarize its predecessors activities from the second century CE onward, generally included monographs about both administrative units and water courses. The oldest Chinese gazetteer dates to the fourth century CE, and over 8,000 Chinese gazetteers survive today. Beginning in the Sui dynasty and extending through the Tang and into the early Song (a period covering approximately the seventh through the tenth centuries), local government officials were required to compile official gazetteers known as map guides (tujing ) every three to five years, succinct indexes totaling a few chapters each, which recorded information about the population, tax revenue, and personnel of official administrative districts. There were approximately 300 to 400 prefects at any one time, and they based their documents on submissions from county magistrates, who generally numbered about 1,200 to 1,500: thousands of map guides would have been completed during these centuries. Individual map guide submissions were compiled into massive imperial compendia.
By the eleventh century, in the context of rapid growth in the population and economy of an urbane but militarized and politically factionalized state, the Song court ceased to update the official map guides. Local officials were no longer required to submit new information about the jurisdictions they governed. Map guides had been written in manuscript form, their only purpose was to fulfill a bureaucratic mandate, they had no readership, and none survive today. However, as the map guides declined, a new genre of gazetteer arose in their place, the local encyclopedia ( difangzhi ). In contrast to map guides, local encyclopedias were funded and written by residents of the jurisdiction in question. Like the map guides, they described administratively recognized units, primarily counties and prefectures. Also like the map guides, they did include some information that would have been of interest to the state, such as the population and tax revenue of each township in a county, the spatial extent and boundaries of the jurisdiction, and the location of the government offices. However, they also focused on the characteristics of the unit s unique local identity and sources of its distinction: poems and literary works that mentioned the jurisdiction or its scenic sites, biographies of its past and present illustrious residents, and famous historical occurrences that transpired there. They were printed works sold for profit in a commercial economy. Many items in the local encyclopedias were spatially referenced, including locations of towns, waterways, mountains, and bridges. Generally, each chapter in a geographical encyclopedia was a list of named places in the jurisdiction with a topical article about each place. For instance, one chapter might describe each township s temples and their locations, and another its characteristic agricultural products and the places where they were produced. As both Bol (2001) and Hargett (1996) conclude, the shift reveals a great deal about the declining role of the state in local affairs and the rise of a local gentry class.
Over time, geographical encyclopedias often grew to dozens of chapters in length, and they were block-printed for a reading public. By the time of the Ming dynasty (the fourteenth through seventeenth centuries), each of the counties in the realm (generally numbering around 1,500) had been the subject of a gazetteer, and by the end of the imperial era in 1911, some places had been chronicled many times over. Monasteries, mountains, cities, individual townships, foreign lands, the realm as a whole, and other kinds of places all had their own gazetteers written by amateur geographers. The genre did not cease in the modern era. The Republican government sponsored a set of revisions of the county gazetteers in the 1930s. A recent gazetteer initiative began in the 1980s and still continues today. The last section of this book includes three examples of spatial history scholarship that can be conducted using databases made from Chinese (and the similar Japanese) local gazetteers.
This brief account has demonstrated that place name indexes were an intellectually significant and extremely common way of organizing and communicating spatial information in one of the world s largest and most literate civilizations. Although maps were also part of the Chinese tradition of spatial information, and although gazetteers themselves included plenty of maps, maps were not the most significant part of the spatial information tradition, and they never stood apart from gazetteers and other works of text. There were a vast number of gazetteers, and that fact alone is notable. Moreover, the transformation of the gazetteer genre from the medieval map guides to the early modern geographical encyclopedias demonstrates, as Bol (2001) and Hargett (1996) concluded, that gazetteers, like maps, have an intellectual history and content worth studying.
Case Study 2: From Itineraries to Descriptive Gazetteers in Britain, 1500-1900
Turning to Western Europe, it was in the sixteenth century that itineraries began to be transformed into gazetteers. Authors began to aim for systematic coverage of areas, but at first they still organized their work as descriptions of journeys, even when the travels were hypothetical. 16 For instance, John Leland (c. 1503-1552) gathered detailed information on England s topography and antiquities while touring the country rescuing monastic libraries for the King s Library, though his notes were not published for another 200 years (Hearn, 1768). William Camden (1551-1623) drew very extensively on Leland s work in his Britannia , published in Latin in 1586 and in English in 1610. This was a work of chorography aimed at a systematic enumeration of Britain s significant places (Mayhew, 2008). Although it was organized into sections about counties, which are then described by following each river in turn, Britannia is in no sense organized around Camden s own journeys (Crane and Waterhouse, 2010). Daniel Defoe (1660-1731) took the same approach, but where Camden focused on antiquities, Defoe was much more concerned with the present and with the first stirrings of the Industrial Revolution.
The first alphabetical gazetteer in the British Library is the Villare Cantianum of 1659, which was limited to the single English county of Kent, describing the main towns with great detail of land ownership and including at the end a separate alphabetical list of all the parishes and hundreds, with detail of their names etymology, derivation, and definition (Philpott, 1659).
Strikingly, the gazetteers of the late seventeenth century already include most of the features found 200 years later. For example, these early gazetteers were, if anything, more likely than later ones to include coordinates, and the Index Villaris of 1680 consists essentially of one alphabetical table of approximately 24,000 locations, including information about the local gentry of each place and an appendix that goes into more detail about the higher nobility. 17 The Gazetteer s or Newman s Interpreter , of 1693, is broader in scope, Being a Geographical Index of all the Considerable Cities, Patriarchships, Bishopricks, Universities, Dukedoms, Earldoms, and such like; Imperial and Hance Towns, Ports, Forts, Castles, c. in EUROPE. Shewing in what Kingdoms, Provinces, and Counties they are; to what Princes they are now subject. This is a slim volume with entries all of about the same length. 18 Similar books began appearing in French, 19 Portuguese, 20 and Spanish over the next century. 21
From the mid-eighteenth century to the end of the nineteenth century, commercial publishers in the United Kingdom and United States steadily produced gazetteers that did not just locate places but also described them in ever-growing detail. The Great Britain Historical GIS project has converted several of these gazetteers into a systematic database. The Imperial Gazetteer of England and Wales (1870-1872) consists of six volumes, 29,411 entries, and about 5,000 words; the Edinburgh entry in Groome s Ordnance Gazetteer of Scotland (1885), by itself, contains over 110,000 words, making it longer than the present book; and the Gazetteer of the World, or Dictionary of Geographical Knowledge (1856) is in seven volumes consisting of 5,913 pages, about 87,000 entries, and about 7,000 million words. 22
This brief account poses two questions. First, why did gazetteers, as distinct from itineraries, emerge, apparently more or less fully formed, in the mid-seventeenth century, given that shortly before Leland and Camden, even authors who had personal knowledge of Britain that was arguably far more systematic than later gazetteer compilers presented that knowledge as linear itineraries? The key change was the development of mapmaking, and perhaps specifically the work of John Speed (1552-1629), who was responsible for both the first set of maps of the individual counties within Britain and also the first world atlas produced by an Englishman. It is not just that Speed worked in precisely the period between Camden and the first gazetteers; in fact, it is known that Speed was encouraged in his work by Camden. 23 Similarly, the vast listing of places and coordinates in the Index Villaris was a direct by-product of John Adams s work to integrate and adjust Speed s maps to create a single large map of England. 24 While the first section of this chapter argued that geographical text became widespread in advance of the map because it required less technology, an important caveat is that specifically alphabetical English gazetteers required an enumeration of places to first be created by a systematic cartographic survey.
Second, why did descriptive gazetteers become increasingly content-rich up to the late nineteenth century but then simply stop appearing? A full answer would require an investigation of the decisions of the commercial publishers involved, but one suggestion is that these increasingly lengthy and unaffordable books were supplanted by several other, more useful forms. For instance, the thematically organized regional geography textbook emerged around the turn of the twentieth century with Vidal de la Blache s archetypal France de l Est , which appeared in 1917. The shorter tourist guidebook was a mid-nineteenth-century invention, with Baedeker s first book appearing in 1854 and the company rapidly expanding in the 1870s. Finally, the larger gazetteers of the late nineteenth century were no longer geographical dictionaries but really place encyclopedias, and as new genres of reference works matured, organizing detailed information on, for example, architecture or wildlife entirely within articles each defined by a place no longer really made sense.
While gazetteers as printed books are now rare, place name indexes are resurgent in the digital realm, with spatial representation pivoting back to a world without maps even as location awareness is becoming an increasingly salient aspect of contemporary life. The world s largest and most heavily used encyclopedia is, of course, Wikipedia. Though not by design, it is very arguably also the largest and most widely used gazetteer, and it is a rich gazetteer, not just a list of place names and coordinates. 25 Roughly 20 percent of Google desktop queries are where-type questions, meaning that they result in a map being generated as part of the response, 26 but while the responses involve images, the queries that generate them are formulated as text strings involving place names or other entities having location. New mobile technologies similarly mean that while the volume of spatial data in the cloud is growing, it becomes steadily less visible to most humans: today our phones tell Google where we are, so we ourselves do not need to know. In the future, Google may take us places in our autonomous vehicles with even fewer of us keeping track of our locations in space. As one recent commentator put it, We re fast approaching an endgame in which the capacity to read a map could become a lost art. It s no exaggeration to describe the smartphone as the equivalent of a cursor moving through a one-to-one-scale map of the world. Today, turn-by-turn navigation is the quintessential map app. Already, some maps exist as voices that tell you where to go: Turn left, turn right . When cars drive themselves, the map will have been fully absorbed into the machine. 27 If we think of maps as two-dimensional displays designed for humans, rather than as visualizations of coordinate data held in the cloud, then the era of map dominance lasted only a few generations and appears to be ending. This chapter has argued that the current interest in gazetteers represents a return to forms of knowledge organization that were the norm for most times, places, and purposes in human history.
1 . .
2 . Gazetteer, , accessed June 23, 2014. David Harley and John Woodward s magisterial History of Cartography series (1987-), for example, simply covers the ground named in the series title, and while Cordell Yee s excellent essay on Chinese cartography argues for the centrality of text in the cartographic tradition, it does not look at works of geography that are purely textual like those we are discussing in this article (Harley and Woodward, 1994). The European volumes in the series take over 5,000 pages to get to the Renaissance while still not giving textual statements of geography equal space. (Actual page counts: Vol. 1, 656 pp.; Vol. 2.1, 604 pp.; Vol 2.2, 998 pp.; Vol 2.3, 500 pp.; Vol 3, 2,272 pp.)
3 . Some classic works about the power of maps include Wood (2010), Harley (1988), Winichakul (1994), Latour (1987), and Monmonier (1991).
4 . See, for instance, Edney (1997) on the difficulty of trigonometric surveying and accurate cartography in British India at the turn of the nineteenth century.
5 . An online edition of a complete English translation is available via the Perseus Project: see Hamilton and Falconer (1903).
6 . Jones and Ormerod (1918).
7 . Cuntz (1929).
8 . Online English translation at .
9 . Delano-Smith (2006); Akerman (2007) covers related ground.
10 . This article is not the place to engage the vast literature on space and place. The classic work is Tuan (1977). See Hubbard and Kitchin (2011).
11 . Carter (1987). Especially see chapter 1 , An Outline of Names, pp. 1-33.
12 . An excellent survey of the range of landscape depictions that can be considered maps and the ways that they have been used is Turnbull (1993).
13 . The best introduction to the history of cartography in China can be found in the five chapters by Cordell Yee in Harley and Woodward (1994).
14 . Mostern (2011).
15 . Hargett (1996); Bol (2001), pp. 37-76; and Mostern (2011), chapter 3 , pp. 57-99.
16 . To be sure, travel writing persisted as well. However, from the eighteenth century onward, writers were more concerned with particular journeys they had made and specific people they met, Boswell s The Journal of a Tour to the Hebrides with Samuel Johnson, LL.D . (1784) being typical.
17 . INDEX VILLARIS: OR, AN EXACT REGISTER. Alphabetically Digested, of all the Cities, Market-Towns, Parishes, Villages, the Hundred, Lath, Rape, Ward, Wapentake, or other Division of each County [etc.] (1680).
18 . (1693).
19 . Guyot (1782).
20 . Cardoso and Pinto Rodrigues ([1747] 1751).
21 . De Alcedo (1786-1789).
22 . The entire text of the first two gazetteers can now be accessed online at , while the text of the Gazetteer of the World will be part of the PastPlace global gazetteer.
23 . Speed, John (1885-1900).
24 . Ravenhill (1978).
25 . The Wikidata project assembles and links content from the different language editions of Wikipedia, and at the time of writing identifies over 1.3 million entities, each the subject of at least one Wikipedia article, that have a latitude and longitude.
26 . Fisher (2013).
27 . Fisher (2013).
2 Gazetteers Present: Spatial Science and Volunteered Geographical Information
Michael F. Goodchild
The names we assign to features and points of interest provide an important basis for wayfinding and for communicating geographic knowledge. While the average adult English-speaker might possess a vocabulary of twenty thousand words (Zechmeister et al., 1995), it is clearly possible for an adult to be familiar with a similar number of geographic names (see, for example, studies of the knowledge acquired by London taxi drivers, e.g., Maguire, et al. 2000). These names might include those of countries and their subdivisions, cities and towns, streets, buildings on a university campus, retail outlets in a shopping center, nearby parks, travel destinations, and many more types.
Over time societies have developed arrangements for ensuring agreement among their citizens over the assignment and spelling of names, and for regulating the process of naming. The traditional gazetteer-a list of officially recognized names, their locations, and often the type of feature or point of interest being named-is the outcome of this process. Over the past decade, however, very rapid growth in the availability of geographic information, especially new types of information, together with parallel growth in the use of such information in online services, has radically changed the world of geographic names and opened up many new possibilities. The workshop on digital gazetteer research that I helped to organize in 2006 (Goodchild and Hill, 2007) and the special issue of International Journal of Geographical Information Science that followed (Goodchild and Hill, 2008) marked a significant watershed in this evolution. At that time the research community was just beginning to see the potential of these developments. Today, profound changes have occurred in the way society conceptualizes, acquires, assembles, represents, and uses geographic names. In this chapter I examine those changes and summarize some of the recent findings of ongoing research.
The chapter is organized as follows. In the next section I describe the concept of neogeography and the nature of volunteered geographic information. The next section describes the extension of gazetteers to vernacular names and the development of point-of-interest (POI) databases. The third major section discusses the growing emphasis on names in human discourse, and the shift from a spatial to a platial paradigm. The last major section discusses the effects of widespread citizen participation when maps and geographic information are regarded as social constructions. The chapter ends with a brief conclusion.
The term neogeography was coined by Turner (Turner, 2006) to describe an emerging world in which engagement with geographic information was no longer confined to experts-for example, professors of geography or geodesy or employees of mapping agencies or corporations-but extended to virtually anyone. The tools of geography were no longer expensive, and geographic information could be acquired at almost no cost by using the Global Positioning System (GPS) or widely accessible fine-resolution imagery. Geographic information could be downloaded free from websites, and processed using cheap software. These changes made it possible to create maps of virtually anything, at very little cost. Maps could be made for the sole use of an individual, centered on that individual s current location, and showing information valid only at the instant of creation, such as real-time traffic conditions. Maps could also be made from perspectives other than the vertical-oblique or even ground-level views. Maps and mapping thus became a personal, rather than a corporate or an agency, concern. The general-purpose paper map disappeared from the glove compartment, replaced by in-vehicle or smart-phone applications.
Neogeography thus implies a radical rethinking of the nature and distribution of geographic expertise (Goodchild, 2009). Citizens discovered an interest in maps of new kinds of phenomena that had never been a focus of traditional mapmaking: maps of trash or graffiti, or maps of cultural heritage. Citizens also became active participants in the creation of more traditional forms of geographic data. Wikimapia, for example, is a project to assemble a comprehensive database of named features, including locations and potentially extensive descriptions. In effect, Wikimapia is a form of gazetteer that has been created by the voluntary activities of citizens rather than agency experts. Significantly, it is not limited to officially recognized features and allows for extensive descriptions of features in addition to the traditional name and type. In a 2007 article I proposed the term volunteered geographic information (VGI) to distinguish this from traditional methods of geographic information production.
VGI is a form of crowdsourcing, relying on a large number of volunteers to provide acceptable coverage, and to correct glaring errors. While Wikimapia and other projects provide the kind of information traditionally associated with gazetteers, Open Street Map focuses on crowdsourcing the content of topographic maps. Besides projects that replicate or augment the production of traditional forms of geographic information, citizens are also involved with the creation of many other forms, in some cases involuntarily. Locations can be associated with photographs that are stored online in services such as Flickr; tweets, blogs, and other forms of citizen-contributed text can be georeferenced, and the tracks of individuals through space and time can be captured and uploaded in numerous ways.
These last may be instances of involuntary citizen contributions, since the individual may or may not be aware of the tracking. While many services are careful to allow their users to opt out of being tracked or of having their locations captured in other ways, there are many exceptions. A vehicle equipped with an automated toll payment system, for example, is logged every time it passes through a toll-gate, and such records have apparently been subject to subpoena in litigation, as evidence of an individual s location in space and time.
Although the vast majority of citizens are well-meaning producers of geographic information, the issue of data quality inevitably arises. While there is a long tradition of trusting mapping agencies and corporations, it is not clear that the same level of trust should be extended to citizens acting voluntarily, or even involuntarily. The protocols and standards that define processes and workflows in mapping agencies and corporations are generally missing, and there are few if any rigorous programs of quality control. In an earlier paper (Goodchild and Li, 2012a) Li and I argued that there might be three possible approaches to quality assurance. First, the ability to edit and correct might ensure that in the long term VGI converges on the truth, a principle sometimes known as the wisdom of the crowd (Surowiecki, 2004) or Linus s Law (Himanen, Torvalds, and Castells, 2001). Second, well-organized projects might include moderators or gatekeepers-citizens appointed to check and edit purported facts-that would in effect mirror the structures of mapping agencies and corporations, but on a voluntary basis. Third, purported facts might be checked to determine whether they were consistent with existing knowledge; since such checks might be automated, this third solution might be operationalized in near-to-real time.
The advent of neogeography has brought profound changes to the relationship between citizen and expert; to the behavior of traditional mapping agencies and corporations; and to the options available to users of geographic information. Open Street Map, for example, is now widely used in applications worldwide. Though debates continue about its quality (Coleman, 2013; Haklay, 2010), it is easy to find instances where it is better than the best official data, and equally easy to find instances where it is worse.
Extending the Gazetteer
As Hill and I noted in our earlier paper (Goodchild and Hill, 2008), a gazetteer is often considered in essence a collection of triples N,F,T where N is a name, F is a footprint, and T is a type. The three components address respectively what is it called, where it is, and what kind of thing it is. In the traditional gazetteer, such a triple is compiled only for officially recognized features: features named on official topographic maps, in most cases. F is most often defined as a point in two dimensions, most likely latitude and longitude. When a feature is distinctly extended, such as a major city or river, a series of conventions have been adopted to allow the designation of a representative point. A river, for example, is sometimes represented by a point at its mouth, and a city by the location of its central administrative offices. The type T is drawn from a list of officially recognized types, as established by the compiling agency.
In the world of neogeography, all of these assumptions and practices no longer hold. The set of features enumerated in a gazetteer is no longer limited to features named on official maps. In the United States, national mapping programs focused on scales of 1:24,000 and coarser, and thus omitted names associated with individual buildings, minor streets, businesses, and many other smaller features, as well as any features that were not likely to persist in the same location for many years. Many of these categories are of great interest to citizens, however. Online mapping, search, and wayfinding services quickly recognized the importance of such features, and began the compilation of what became known as point-of-interest (POI) databases. Even today there is a persistent separation between the concepts of gazetteers and POI databases, despite their obvious similarities in form and objectives.
Eventually this separation may disappear, yet neither term is especially suitable as an umbrella. Gazetteer has no intuitive meaning, and its etymology is obscure (deriving from the process of gazetting, that is, the recognition of a name in an official publication). Point-of-interest implies a feature small enough to be represented as a point, and thus precludes more elaborate footprints. The consequences of this are easy to document. When asked, for example, to provide driving directions from Colorado to Wyoming, Google Maps, at the time of this writing, provided a detailed description of a route of 394 miles, including a segment that may be seasonally closed, from a point in the mountains southwest of Denver to a point in an unpopulated area of Wyoming. Of course the length of such a trip might be as little as a meter, Colorado and Wyoming being adjacent, or more than 800 miles from the southeast corner of Colorado to the northwest corner of Wyoming.
Previously, the contents of a traditional gazetteer were limited to officially recognized names. Moreover the widespread adoption of geographic information systems (GIS) for processing geographic information led to an additional resistance to the representation of features without rigorously defined boundaries (Burrough and Frank, 1996). Yet citizen discourse frequently includes references to vernacular names such as downtown, and it is clear that such names will grow in importance as neogeography evolves. Many methods have been researched in an effort to create spatial representations of such features that can be handled in GIS, including work with human subjects (Montello et al., 2003) and analysis of references to such names in social media and online sources (Jones et al., 2008). Such methods might lead to vector representations as points, lines, or areas or to raster representations with regular grids.
Many names of features are globally unique, but in other cases the same name may refer to many distinct features. While Great Britain usually refers to England, Scotland, and Wales, it might also refer more locally to the SS Great Britain , a historic ship docked in Bristol. According to official sources there are sixty-two populated places in the United States named Springfield. Thus a name might be associated with a domain over which it is known to be unique, such as Springfield, Illinois. Such issues are sometimes resolved by knowledge of context; for example, at the time of this writing a Google search for McDonald s from my home in Seattle yielded only three (of the more than thirty-two thousand worldwide), all within a short distance, because Google is able to filter the search based on its knowledge of my location, which it derives from my internet address.
It is clear that the number of named features on the Earth s surface is enormous. At the time of this writing the crowdsourced GeoNames database contained over 8 million, while Wikimapia contained over 22 million georeferenced features and descriptions. Yet a single university campus might include on the order of a thousand named features, and every business is a potential entry in a POI database. If the concept of a gazetteer/POI database is extended to include every identifiable feature then it might well include every building, every person, every vehicle, and even every farm animal and tree. There are abundant applications for such information, and despite the obvious social issues there appear to be no technical limits to obtaining it.
Space and Place
The concept of place has a long and notable history in the discipline of geography (see, for example, Tuan, 1977). The literature on place emphasizes its role as a social construction, a means by which humans simplify and organize the world around them. Places are identified primarily by name, and the meaning of these names can be highly individualistic and context-dependent. From the perspective of GIS, place lacks the kind of rigor and objectivity that characterizes official sources of geographic information; instead, GIS emphasizes the importance of space, with its standard methods of measuring location and standard methods of description. Thus the past few decades of rapid growth of GIS have shifted the balance away from place. But in the past few years the engagement of the citizen, and the availability of new forms of textual information from blogs, tweets, and other media, has placed new emphasis on names and informal human discourse, and in turn on concepts of place.
Recently substantial effort has been expended in adapting concepts of place to the computational environment. Reference has already been made to work on the spatial representation of place, though this approach merely seeks to bridge the gap between the two paradigms, necessarily adapting and modifying the concept of place in the process. More productive, perhaps, is research on a purely platial as distinct from spatial technology. In this section I review recent work in this area, extending the earlier discussion of Goodchild and Li (2012b).
A cornerstone of spatial technology is the linking of data by geographic location. Topological overlay was one of the first and most fundamental functions of a GIS (Foresman, 1998), while the far less sophisticated operation of graphical overlay, or visual superimposition of one layer of data on another (often described today as a form of mashup), provides abundant potential for insight on the part of the user. Yet these operations have always been challenged, by the lack of sufficient positional accuracy on the one hand, and by lack of semantic standards on the other. In a platial world it is names that provide linkages, rather than positions. Names are related hierarchically: for example, Notre Dame, the Eiffel Tower, and Montparnasse are all parts of Paris. Yet such hierarchical relationships need not have the same geometric rigor as the spatial concept of containment: Disneyland Paris is to most people hierarchically related to Paris, yet is not within the formal definition of the Paris municipality.
Several other aspects characterize a platial approach, and serve to distinguish it from GIS. First, the planimetric accuracy of traditional maps is less relevant, and sketchmaps, strip maps, and other rough approximations become potentially adequate to support wayfinding. The famous Beck map of the London Underground (Leboff and Demuth, 1999), now replicated in virtually all urban transit systems worldwide, sacrifices planimetric accuracy in the interests of clarity, thus distorting distances and directions. But it preserves those aspects of topology that are critical to wayfinding, including the ordering of stops along lines and the connections between lines. Strip maps (MacEachren, 1986) similarly sacrifice planimetric accuracy, in part to create a paper-based guide that is more easily handled in the cramped space of a vehicle. In general, sketch maps work by preserving those aspects of geographic information that are important to wayfinding, including the ordering of features, connections between linear routes, and the presence of named features, while distorting or deleting unnecessary detail. The automated creation of sketch maps from spatial data is a challenging research topic (e.g., Kopf et al., 2010).
For example, how should a smart-phone app create a sketch map for an individual that (a) responds directly to the individual s known preferences, (b) distinguishes between night and day and their appropriate landmarks (e.g., Nothegger, Winter, and Raubal, 2004), (c) combines the potential for multimodal travel, including transit and walking, and (d) is suitably adapted to the small screen of a personal device?
Second, it is interesting to speculate on how standard and widely used GIS functions might have parallels in the platial world. In GIS the preposition near is often implemented using a buffer, to identify features within a prescribed distance of a given location. But in a platial world distances are at best ordinal (as in the ordering of stops in the Beck map; on that basis Hounslow West is nearer Heathrow Airport than is Hounslow Central). One possible implementation would be as follows: a place B is near a place A if both A and B are parts of the same higher-level feature C. This implies that near is implemented as a difference of levels in the place hierarchy.
Consider, for example, a tourist in Paris interested in seeing the Eiffel Tower, the Mus e d Orsay, and Notre Dame. All three are in Paris, so are near at that level. But the Eiffel Tower and the Mus e d Orsay are both in the same administrative subdivision of Paris, the 7th arrondissement, while Notre Dame is in the 4th, therefore, we can say that the Mus e d Orsay is nearer the Eiffel Tower than is Notre Dame. However this approach cannot determine that the Mus e d Orsay lies approximately on a line between the Eiffel Tower and Notre Dame-although another platial source, the standard map of the Paris M tro, shows that the RER line along the Left Bank includes stops with those names in that order.
Third, a platial approach that placed emphasis on linking information through names would have to deal with the fundamental ambiguities of naming. While a similar problem exists in the spatial world because of differences among coordinate systems, map projections, and geodetic datums, as well as measurement inaccuracies, the comparable platial problem is potentially much more serious. Differences exist in spellings, translations between languages, and changes and movement of named places in time, as well as in the meaning of names under changing context (Los Angeles when mentioned by an Angelino in Los Angeles is not the same as Los Angeles to a native of New York). A platial system would need to equate the English Channel with La Manche, and be aware of numerous other equivalences.
Hastings (2008) has studied the issues involved in conflating gazetteer information, dealing with differences not only in naming but also in locations and types. He argues that location is the most important property in determining whether different records refer to the same feature, since alternative locations given to the same feature are necessarily similar. Type is next, but name is last, since alternative names given to the same feature can be entirely unrelated.
Maps as Social Constructions
I referred earlier to the nature of geographic information as a social construction. Language has a significant impact on a name, as does the script used to represent it. For example, Beijing, Peking, Shuntian, Peiping, and are all among the many names that have been used in various eras to refer to the same city, the current capital of China. Type is also socially constructed, since the purpose or meaning of a feature can vary among different linguistic and cultural contexts. Moreover the very existence of a feature is socially constructed: what is worth naming and mapping depends on who is doing the naming and mapping, and for what purpose. Mark and Turk (2003) have documented the ways in which a small aboriginal tribe in Western Australia identifies features on its landscape, and how it differs from common practice in, for example, US topographic mapping.
Online maps of the Himalayan region illustrate this point very well. Google s mapping service, as accessed through , shows many borders as dashed, indicating that they are in dispute. Kashmir, for example, has been contested between Pakistan and India ever since the partition of British India in 1947; and Arunachal Pradesh and several other areas are the subject of ongoing disputes between India and China. However, a request to Google from an internet address in India will produce a different map, in which Kashmir and Arunachal Pradesh are bounded by solid lines indicating that they are part of India; and similarly a request from an internet address in China (except Hong Kong and Macao) will show China s claims as fact. In short, there is not one map of the Himalayan region, but at least three, depending on national policies.
In 1913, at the height of British imperial power, an international conference was convened in London to initiate production of an International Map of the World, at a scale of 1:1,000,000 (Rhind, 2000). Standard protocols were adopted for the feature types to be represented and their symbology, and close to half of the sheets were completed over the next several decades before the project was abandoned. In its absence mapping remains primarily a national responsibility, with obvious differences between national practices; and for certain feature types practices differ widely at well below the national level.
Yet the concept of global standardization has resurfaced again in recent decades. The Global Spatial Data Infrastructure Association promotes international collaboration and cooperation in mapping, though it stops short of advocating global standards. However remote-sensing satellites gather global data without regard to national boundaries or cultural and linguistic differences, encouraging global standardization of feature types. Moreover, projects such as Open Street Map (OSM), which aim to create global mapping, also encourage uniform tagging of feature types.