the technology of religion: mapping religious cyberscapes · este art´ıculo combina estudios...

The Technology of Religion: Mapping ReligiousCyberscapes

Taylor SheltonClark University

Matthew ZookUniversity of Kentucky

Mark GrahamUniversity of Oxford

This article combines geographical studies of both the Internet and religion in an analysis of where and how avariety of religious practices are represented in geotagged Web content. This method provides needed insightinto the geography of virtual expressions of religion and highlights the mutually constitutive, and at timescontradictory, relationship between the virtual and material dimensions of religious expression. By using thespatialities of religious practice and contestation as an example, this article argues that mappings of virtual rep-resentations of material practices are important tools for understanding how online activities simultaneouslyrepresent and reproduce the material world. Key Words: cyberscapes, Internet, neogeography, religion,volunteered geographic information.

Este artıculo combina estudios geograficos tanto de Internet como de religion en un analisis sobre donde ycomo una variedad de practicas religiosas se representan en contenidos web ubicados en la categorıa geo. Estemetodo da la necesaria ilustracion en la geografıa de expresiones virtuales de la religion y destaca la relacionmutuamente constitutiva y a veces contradictoria entre las dimensiones virtuales y materiales de la expresionreligiosa. Utilizando las espacialidades de la practica religiosa y sus retos como un ejemplo, este artıculoarguye que los mapeos de representaciones virtuales de practicas materiales son herramientas importantespara entender como las actividades online simultaneamente representan y reproducen el mundo material.Palabras clave: cyberpaisajes, Internet, neogeografıa, religion, informacion geografica voluntaria.

O ver the past two decades, the Internethas evolved from the exclusive domain of

technophiles to a near ubiquitous aspect of ev-eryday life. As more and more people come on-line, the use of the Internet increasingly reflectsthe concerns and interests of the population asa whole. At the same time, however, the Inter-net serves to represent and reproduce society in

a variety of ways. Indeed, some argue that thereis no clear division between the online and off-line, the virtual and material, as the two arecoconstituted and mutually influencing (Zookand Graham 2007a). Particularly compellingare the new types of linkages made betweenonline activity and offline locations in which In-ternet users associate meaning—ranging from

The Professional Geographer, 64(4) 2012, pages 602–617 C! Copyright 2012 by Association of American Geographers.Initial submission, August 2010; revised submission, March 2011; final acceptance, March 2011.

Published by Taylor & Francis Group, LLC.

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2

The Technology of Religion: Mapping Religious Cyberscapes 603

the prosaic to the profound—to specific sitesin the material world. These novel phenom-ena, commonly referred to as geotagging (alsogeoreferencing or geocoding), provides an innova-tive means for studying the spatial contours ofthe virtual dimension of practically any subject,including religion.

Using geotagged data collected from theInternet, this article demonstrates that the mea-suring and mapping of online references to re-ligion can provide important new insights intodynamic spatialities of religion and how onlinerepresentations are simultaneously productsand producers of offline social processes. In thiseffort, this article integrates the two relativelydisparate literatures on the geographies of re-ligion and the Internet. Although academic re-search into religion and the Internet have foundsome common ground in other disciplines(Campbell 2005), it remains a relatively un-derstudied arena within geography. This articleseeks to address this lacuna via a cautious exam-ination of how the virtual dimension of religionis dispersed across the globe. This analysis canthus shed new light on both the geography ofreligion and the geography of the Internet.

This article argues the following: First,geocoded online data are valuable for iden-tifying the online visibility and prevalence ofdifferent religions as well as the offline locationand concentration of religious practices. Al-though this can be seen as a “proof of concept,”this approach also underscores the power andflexibility of geotagged data in illuminatinglittle-known religious practices as well as thevirtual dimensions of well-known religiousgeographies. For example, the technique intro-duced in this article effectively illustrates howoffline religious conflicts can be reflected incyberspace. Because of the Internet’s ability toprovide new sites for religious representationbeyond traditional sacred spaces, the ways inwhich these disputes occur online are deservingof further analysis. More generally, the cases ofreligion presented in this article highlight theintertwining of online interactions and offlinesociospatial practices.

Dynamic Approaches to Geographiesof Religion

The spatial distribution of religious practice isa long-standing interest in geography and is

one of the standard markers by which regionsare often identified (e.g., “the Islamic world,”“the Bible Belt”). As Kong (1990, 2001a, 2010)outlined, however, the trajectory of religiousgeography has been multifarious, concernedwith a variety of topics from a number of theo-retical perspectives. Despite the differences inapproaches and method, this work represents acollective attempt to understand the sociospa-tial configurations of religious belief and theirrelationships with the rest of society. As cus-toms, cultures, and technologies continue toevolve, studies of the geographies of religionalso need to reflect these changes.

From Ecclesiastical Geographies to CriticalGeographies of ReligionAmong the earliest forms of religious geogra-phy was “ecclesiastical geography,” or the map-ping and analysis of religious spheres of influ-ence (Isaac 1965, in Kong 1990). This so-calledecclesiastical geography had a decidedly appliedfocus, as it was primarily employed by Christianmissionaries to better understand those placesin which they sought to find potential converts.Although ecclesiastical studies often containedcolonial impulses (Kong 1990), there has alsobeen significant work on how religious beliefand practice overlay the human landscapewithout the pretenses of the Christian mission(Shortridge 1976; Crawford 2005; Warf andVincent 2007; Warf and Winsberg 2008).

More recently, Kong (2001a) has arguedthat much of the current focus of religiousgeography has turned away from simpleattempts to understand where religious prac-tice takes place. Instead, there has been aconsiderable body of research united in theuse of religion as a valuable lens throughwhich larger questions of culture and politicscan be understood. For example, Zelinsky(2001) examined the multiplicity of religiouslandscapes in the United States as an exampleof “American exceptionalism,” a term that em-bodies a number of sociopolitical contentions.This research, along with that of Proctor(2006a, 2006b), further highlights the complexinterconnections among religion, place, andpolitics. Other scholars have even criticizedthe notion of religion as being ontologicallyprior (Ivakhiv 2006). Likewise, religiousgeography has greatly broadened its scope of

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2

604 Volume 64, Number 4, November 2012

what constitutes the “sacred” to more fullyunderstand the spatialities of religious commu-nities and practices (Brace, Bailey, and Harvey2006). This has required a rethinking of thepotential sites of study for religious geographybeyond the traditional categories of organizeddenominations or sacred places. These trendshave significantly broadened the scope of whatfalls within the realm of religious geography.

Religion and TechnologyOne of the key directions in this expanding fieldof study is the role of information and commu-nication technologies (ICTs) in mediating andrepresenting religious practice (Kong 2001b).Far from having a single predetermined appli-cation of a communications technology to re-ligion, applications and technological practicesare as varied as those engaged in their use. Thisvariation is reflected by the wide variety of top-ics studied by those interested in religious geog-raphy and ranges from one-to-many televisionor radio broadcasts of religious programming(Stump 1991; Kong 2006), the use of the Inter-net to promote religious education (Brunn andLong 2000), or the many-to-many construc-tions and contestations of religious narrativeson the Internet (Campbell 2007; Cheong et al.2009). This article builds on this work and seeksto understand the mutual constitution of tech-nological practices, digital representations, andthe material world with respect to religion.

Reflections on the links between material andoffline practices and the virtuality of the Inter-net are not new to geographic scholarship. In-deed, there is a long tradition of arguing againstideas that new ICTs would result in the “deathof distance” (Cairncross 1997) or the renderingof geography irrelevant, as had been positedabout earlier communication technology ad-vancements (Fischer 1994). Far from makinggeography less important, it is clear that newcommunications technologies amplify the im-portance of space, place, and distance (Castells2001), albeit not necessarily in accordance withhistorical patterns.

For instance, the infrastructure necessary toproduce the world’s high-speed telecommuni-cations network grounds flows of informationin a distinct material reality that determineswho has access and who does not (Grahamand Marvin 1996, 2001). The concentration of

these physical infrastructures in dense urbanenvironments is in no way surprising, althoughit reaffirms the connection of the digital to thephysical (Gorman and McIntee 2003). Even inplaces with dense communications networks,there remain significant socioeconomic imped-iments to access, which are only reinforced bythe inability of these segments of the popula-tion to participate in the exchange of infor-mation through the Internet (Chakraborty andBosman 2005; Crang, Crosbie, and Graham2006). Because of these persistent inequalities,the production of online content is oftenconcentrated in urban areas (Zook 2000;Townsend 2001) and divided along lines ofrace and class (Warf 2001; Crutcher and Zook2009).

Emerging Religious CyberscapesPerhaps more important than focusing on theconnections between ICTs and geography isthe fact that the two are mutually constituted,in much the same way as religion and society asa whole (Cooper 1992). In other words, ICTshave both the potential to reinforce existing so-cial hierarchies and the capacity to enable newforms of spatial practice.

This potential has become more readily visi-ble with the advent of Web 2.0, the name givento the rise of user-generated online content inthe form of blogs, wikis, photo sharing, and so-cial networking sites like Facebook and Twit-ter. Web 2.0 remains a contested concept withfuzzy boundaries, but its ultimate importancelies in the shift from viewing the user as con-sumer to viewing the user as producer (O’Reilly2005). Due to the decentralization of contentproduction, it is frequently argued that Web2.0 allows for the production of a qualitativelydifferent, not to mention more open and demo-cratic, online social world (Shirky 2008). Ratherthan the existing practices of one-to-many dis-semination of traditional media such as tele-vision, radio, and print, Web 2.0 relies on amany-to-many network of content productionand consumption, in theory allowing anyonewith a computer and an Internet connection tocreate content that is just as accessible as majormedia outlets.

Of particular relevance to the discipline ofgeography is the rise of Web 2.0 spatial appli-cations. With the release of tools like Google

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2


Earth or Wikipedia, Internet users are able tocreate content (text, photos, etc.) that can beassociated with a particular point on the Earth’ssurface (i.e., geotagging). These practices havebeen given many names, perhaps most notablyneogeography (Turner 2006; Goodchild 2009;Graham 2010) and volunteered geographicinformation (Goodchild 2007; Elwood 2008),to highlight the distinct spatial content ofsuch Web 2.0 content. The adoption of neo-geographic practices and use of volunteeredgeographic information has been astounding.Google Earth alone had been downloadedmore than 700 million times by July 2010(Birch 2010). Although not all users are alsocontent creators, the fact that Google Earthand its companion Web application, GoogleMaps, allow users to search specifically for user-generated content highlights the importanceof these new Web 2.0 spatial applications.

Whether or not user-generated geographicinformation represents a truly democraticalternative to traditional cartographic practiceor just another example of corporate omnipres-ence (Harvey 2007; Boulton 2010; Graham2011a, 2011b), Web 2.0 mapping tools haveallowed Internet users to construct new repre-sentations of place through georeferenced text,photos, and video (Zook and Graham 2007b).These representations—also referred to ascyberscapes—act as interfaces between virtualcontent and material experiences. Cyberscapesare constituted by any number of types of user-generated, georeferenced content, includingGoogle Maps placemarks, tweets, Flickr pho-tos, or geotagged Wikipedia articles and blogposts. These subjective representations of usersare then often digitally reordered by searchengine ranking algorithms, producing, re-producing, and shaping particular worldviews(Zook and Graham 2007a). It is thereforeimportant to begin to more closely study howuser-generated cyberscapes are impacting bothsociospatial practices and our ability to studythose practices.

This article pays particular attention to theimplications of virtual representations for socialresearch and asks how cyberscapes can provideresearchers insights into the geographies ofreligion. By comparing religious cyberscapesto the wealth of scholarly knowledge aboutthe offline geography of world religions, it ispossible to test the extent to which material

religious practices are reflected within onlinerepresentations. Even more important, thismethod allows one to study aspects of religiouspractice (e.g., interfaith conflicts and newconceptions of the sacred) that have hithertobeen extremely challenging to study usingmore traditional means. At the same timeas cyberscapes are used to analyze the ge-ographies of religion, it is possible to see thisexercise as allowing the geographies of religionto shed light on how cyberscapes are producedand might or might not be representative of thematerial world and what factors might help toproduce these incongruities between the two.

A Method for Measuring ReligiousCyberscapes

To collect data on cyberscapes, a speciallydesigned software program was used, whichtallies the number of references to particularkeywords (e.g., Christianity or Allah) that havebeen geotagged to particular locations andare indexed in Google Maps. The softwareconducted queries using a variety of religiousterminology in February and July 2010. At eachlocation, the number of hits for each keywordwas collected and recorded. In essence, theprogram automates the usually manual processof searching for a particular keyword (e.g.,Catholic) in a specific place (e.g., Chicago,IL).1 Figure 1 provides an example of such amanual search with a result of 7,519 geotaggedreferences containing the term Catholic nearChicago, Illinois.2

One of the chief advantages of this methodis that it avoids many of the pitfalls of relyingon fixed, hierarchical conceptions of scale (cf.Marston, Jones, and Woodward 2005). By us-ing points and buffers to sample, we are able toeschew a local-to-global organizational frame-work and instead collect data based on a sam-pling architecture that is more concerned withthe density of sampling points. For example, thedata presented in this article cover the entiresurface of the globe and are sampled from ap-proximately 250,000 latitude and longitude co-ordinates on the Earth’s surface derived from a1/4" grid.3 It is, however, also possible to utilizea range of gradations in grid size that are moreappropriate for metropolitan and regional-levelresearch, as is done later in this article.

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2


Figure 1 Google Maps search for Catholic near Chicago, Illinois. (Color figure available online.)

Another important strength is the unlimitedflexibility in the selection of keywords tobe searched. One need not limit terms toestablished religions or orthodox dogma. Forexample, this article explores the results ofkeywords for the names of religions (e.g.,Christianity or Islam), denominations withinreligions (e.g., Lutheran or Sufi), buildingsassociated with particular religious practices(e.g., church or mosque), and the names ofdeities or otherwise important religious figures(e.g., Jesus or Allah). Although these keywordsearches do not embody the full range ofreligious practice, they are representative of adiversity of religions and ways of representingthose beliefs textually. For instance, both main-stream religions (e.g., Christianity, Judaism,Islam, and Buddhism) and less populous beliefsystems (e.g., Scientology or Zoroastrianism)were queried, as well as a variety of denom-inations within religions—from Anglican,Catholic, and Baptist to Sunni, Shi’a, and Sufi.

Through a comparison of the relativenumber of references to keywords at differentlocations, it is possible to examine religiousinscriptions on the virtual landscape. In otherwords, by measuring the amount and diversityof content written about religion by believersand nonbelievers alike, including the locationof places of worship, concentration of faithcommunities, or other sites imbued withspiritual meaning, we can begin to engagein a critical analysis of how these virtualrepresentations either do or do not appear tocorrelate to expectations based on material

practice. Using a series of vignettes, this articleargues that although online religious represen-tations open up new spaces for, and ways of,measuring religious practice, online contentultimately and necessarily offers only selectiverepresentations of the world. So although thegeographies of user-generated online contentare broadly reflective of offline practices,including that of religion, user-generatedcontent remains subject to the same unevenpower relations at work in the material world.

Mapping a New EcclesiasticalGeography

Just as the earliest geographers of religion wereinterested in mapping the spatial extent of re-ligious beliefs around the world using surveys,census reports, and attendance at religious ser-vices, this article uses geotagged references toreligious terminology on the Internet to cre-ate a new ecclesiastical geography, albeit onefocused on the virtual extension of materialplaces. The maps from this analysis highlightthe variable concentration and scarcity of on-line religious representations, allowing one tosee how the online dimension of various placesis influenced by a variety of factors.

The map of Christianity in Figure 2A offers asomewhat unexpected ecclesiastical geography.Virtual references to Christianity are highlyclustered in Western Europe and North Amer-ica, and only a relatively small number of ref-erences are found in the rest of the world. This

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2


Figure 2 Major world religions in Google Maps: (A) references to Christianity; (B) references to Bud-dhism; (C) references to Hinduism. The exact size of the symbols in these maps is neither the sameacross all three maps, nor is it associated with any particular quantitative value. These maps show onlythe relative concentrations of references to each particular keyword. This same method of representationis repeated in Figure 6.

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2


is particularly surprising given the extremelylarge number of Christians in sub-SaharanAfrica and South America, among other places.This would seem to indicate that the concen-tration of references to a particular keyword isnot associated with mere population density.Instead, as previous work on the geographies ofuser-generated content has suggested, the cor-relation between levels of Internet access andthe degree of engagement with user-generateddigital content in a place seem to be the largestfactor in why some places are underrepresentedin cyberscapes (Graham 2009). It is impor-tant to highlight the uneven representationsoffered by this method because of the po-tentially significant material consequences ofbeing omitted from these virtual ecclesiasticalgeographies.

In many ways, however, religious referenceson the Internet mirror corresponding offlinereligious practices. For example, Figures 2Band 2C illustrate significant clusters of ref-erences to Buddhism in East and SoutheastAsia, and the largest cluster of references toHinduism is located on the Indian subconti-nent. Both maps are thus largely consistentwith the offline distribution of religions. Inboth cases, however, a significant number ofreferences can be found in Western Europeand North America, an observation thatperhaps owes much to the practice of thosereligions by large diasporic populations onboth continents. The fact that searches wereconducted only in English (an issue discussedin more detail later) likely also contributes tothe number of results found in Anglophonecountries.

Absolute measurements are only one wayof understanding the location of differentreligious practices. Also of interest to this newecclesiastical geography is mapping the loca-tions where virtual references to one religiousbelief are more prevalent than all others. Assuch, it is instructive to compare relative countsof religious references that have been overlaidonto the same places (see Figure 3). Figure 3compares references to five prominent Chris-tian denominations in the United States. Ateach point on the Earth’s surface, the numberof references to the given set of keywords wascompared. Each point was then assigned a colorvalue based on which of those keywords had the

most references at that particular point.4 Forinstance, in Figure 3, a green dot indicates thatthe number of references to Baptist is higherthan the number of references to Catholic,Lutheran, Methodist, and Mormon in the samelocation.

As the previous example illustrates, thisparticular method of visualization can presentstriking contrasts among different regional re-ligious preferences. References to Baptist dom-inate the U.S. South; Lutheran prevails in theupper Midwest; and references to Mormonare strongest in the Rocky Mountain regionof Utah and Idaho. Quite unsurprisingly,these concentrations correlate strongly to boththe prevalence of adherents in a particularplace and popular conceptions of the religiousculture in these particular areas. This canpotentially be attributed to the more evendistribution of the Internet across the UnitedStates, as opposed to the global scale, whereaccess is significantly more unequal. Somewhatmore surprising is the belt of Methodists thatdivides the largely Baptist South from a mixof Lutherans and Catholics to the north. Incontrast to these clusters is the diversity ofgeocoded religious references that are apparenton the West Coast of the United States. Thesefindings affirm previous work by Shortridge(1976) on the diversity and heterogeneity ofreligiosity in the American West. Likewise, thespatial concentrations of each of the previouslymentioned religions actually conform quiteclosely to a more advanced centrographicanalysis undertaken by Crawford (2005) usingcensus-like data at the county level, furtheraffirming the validity of these methods in thiscase.

Figures 2 and 3 thus illustrate the oftenclose relation between online and offlinepractices, simultaneously calling attention tohow issues of inequality and scale can affectthese visualizations. Nevertheless, this methodhelps to illuminate the possibilities of a newecclesiastical geography focused on the onlinerepresentations of religion. More important,it shows that the virtual sphere is an importantsite for religious activity and suggests thatthese references can constitute new digitalspheres of influence that can be claimedand contested by practitioners of differentreligions.

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2


Figure 3 Christian denominations in the United States. (Color figure available online.)

Uncovering Cyberspaces of ReligiousConflict

Heeding Kong’s (2001a) call for more researchon how the Internet provides new frontiers forreligious conflict, this section extends our com-parative analysis of online religious geographiesto a number of sites of religious contention. Inthis instance, references to religious keywordscan illustrate the relative strength of religiousrepresentation on the Internet as well as howthe online presence of different religiousgroups conforms to (or transgresses) sites ofreligious contention in the material world andthe official borders of nation-states.

These representations within virtual space,marked by overlapping but contradictory andcontentious depictions of the same places,are often new expressions of conflicts inphysical space and add a new layer to alreadycomplex and variegated representations ofplace (Graham 2010). In Figure 4, references

to Buddhism, Hinduism, and Islam are mappedin Asia, employing the same method as inFigure 3. Although cyberscapes are rarelyuniform in their conformance to any kind ofpolitical boundary, they do exhibit distinctpatterns that are indicative of the influenceof nation-states on the construction of onlinerepresentations. For example, the clustering ofreferences to Hinduism almost exactly mirrorsthe boundaries of India, a country wherethe majority of citizens are Hindu. Similarly,references to Buddhism are concentrated inJapan, Thailand, South Korea, Nepal, andSri Lanka, and references to Islam are mostconcentrated in Malaysia, with smaller pocketslocated in and around Pakistan.

In some cases, these differences are simplyillustrative of the cultural differences amongnations based on religious beliefs. In others,however, these cyberscapes can be riddled withsociopolitical contentions, like in the ongoingstruggle between India and Pakistan over the

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2


Figure 4 Buddhism, Hinduism, and Islam in South and Southeast Asia. (Color figure available online.)

Kashmir region. They can also reflect the vary-ing levels of online engagement and the pro-duction and reproduction of relevant culturalinformation (e.g., note the lack of representa-tion in Bangladesh). It is unclear whether or notthe larger number of references to Hinduismthan Islam in this region is a result of languagedifferences, varying levels of Internet access, ora concerted effort to claim virtual territory.

Another example of how virtual informationcan shed light on real-world conflict is a closeexamination of the ongoing conflict within thestate of Israel and the occupied Palestinian ter-ritories. This analysis uses the same basic ap-proach of collecting data, but its specificityallowed for three important refinements tothe method for measuring cyberscapes. First,rather than searching for names of religions,this effort uses keywords based on the namesfor places of worship for each of the three re-

ligions that have deep and ongoing ties to theregion: synagogue for Judaism, mosque for Islam,and church for Christianity. Second, searchesfor the first two keywords were conducted inboth English and the local language of the pop-ulation: Hebrew for synagogue and Arabic formosque. Third, the search grid for these querieswas much finer than the one utilized at theglobal level, with a search radius at each pointof approximately 0.8 km (0.5 miles). This ap-proach allows for the documentation of placesat which geocoded references to one religion’splace of worship are greater than the numberof references to places of worship for either ofthe other two religions.5 The results, illustratedin Figure 5, show one aspect of the contours ofreligious cyberscapes in the region.

The clustering of references within andaround the city of Jerusalem is a clear indica-tion of its importance to all three Abrahamic

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2


Figure 5 Places of worship in Israel and the Palestinian territories. (Color figure available online.)

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2


religions. Jerusalem is important both polit-ically and spiritually to Jews, Muslims, andChristians, as all lay claim to spaces within thecity with particular spiritual significance. Fig-ure 5 provides an illustration of this conflictin the relative diversity of references to eachof the three religions within and around thecity.

Although Figure 5 shows that the virtualterritory of the state of Israel is well popu-lated with references to synagogues, it is alsomarked by a diversity of references to Mus-lim and Christian sites of worship. Of partic-ular importance is that geocoded references tomosque and synagogue do not conform to thepolitical boundaries between the state of Israeland the Palestinian territories of the West Bankor Gaza Strip in the same way that referencesto Buddhism, Hinduism, and Islam correlatedto national borders in Figure 4. This is in partbecause the political boundaries within this re-gion do not necessarily correspond to ethnicand religious population patterns. Not only doa significant population of Muslims live out-side of the Palestinian territories but so do asignificant number of Jews and Christians livewithin them. Indeed, the predominance of ref-erences actually corresponds quite closely withthe offline presence of various ethno-religiousgroups. The concentration of references tomosques in much of the Northern District ofIsrael is reflective of the fact that the regionis predominantly Arab, and the scattered refer-ences to synagogues throughout the West Bankcorrespond to a number of Israeli settlementsin the area, especially around East Jerusalem(Btelsem 2002). In this sense, mapping refer-ences to various keywords helps to reflect thelived reality of these spaces in a way that thesimple drawing of borders and claiming of ter-ritories cannot.

An important qualification to this mappingis the recognition that the politics of represen-tation in Google Maps are not straightforward.The exclusion of information about Palestinefrom many Web 2.0 spatial applications canpotentially preclude any representation at all(Ben-David et al. 2008). For instance, withinIsrael and the Palestinian territories, referencesto mosque and synagogue make up 21 percentand 60 percent, respectively, of the entire sam-ple, with mentions of church representing the

remaining 17 percent. These numbers stand instark contrast to the actual demographics of theregion, where nearly 55 percent of all people areof the Jewish faith, over 40 percent are Muslim,and but a small fraction are Christian.6 The dis-continuities between these two indicators showthat Christians seem to possess a disproportion-ately powerful voice in the Web 2.0 world ofGoogle Maps, as compared to Muslims. Thereasons behind this, however, are less clearand highlight some of the constraints to thisapproach.

Methodological Limitations

In the examples presented, it has beenshown that the mutual constitution of thematerial world and online representations ofit can be better understood through an analysisof geotagged Google Maps content. There ex-ist, however, a number of issues related to themethods employed in this article. These con-cerns, although important to address and beaware of, should not be seen as negating the im-portance and relevance of the methods demon-strated in this article. As with most methods,mapping references to geotagged Web contentis meant to offer specific insights into a certainset of practices and processes but not to be anall-encompassing means of understanding of allaspects of the geography of religion or of theInternet.

First, virtual representations tagged tospecific places can, in theory, be created byanyone from anywhere. One need not bein India to create content about Hinduismgeotagged to the River Ganges or in Jerusalemto code a reference to the Church of the HolySepulchre. Indeed, one need never have evenvisited a place to create such content. Nordoes a reference necessarily connote a positiveendorsement. Hypothetical representationscould reference the destruction of a synagogue,mosque, or church, just as equally as they couldhope for the construction of more religiousbuildings in a place. Representations of placeare always multiple and are tempered by anynumber of social, political, and economicfactors in the world (Graham 2010). Becauseonly a minority of people have access to the In-ternet and an even smaller proportion still are

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2


active creators of content (Ortega, Gonzalez-Barahona, and Robles 2008), a relatively smallnumber of people have a disproportionatedegree of power when it comes to constructingthe virtual landscape. We therefore see thatalthough measurements of online referencesare by no means perfect representations of theperspectives of all individuals or groups, theyare, in many ways, accurate reflections of thepersistently uneven power dynamics at workin both the online and offline worlds. In otherwords, the religious cyberscapes capturedin this article are not intended to present aperfectly representative reflection of offlinereligious practices. Nonetheless, they do offeruseful insights into how inequalities in theoffline world are reflected in the unevenness ofrepresentation in the online world.

Second is the central issue of language. Be-cause particular terms of interest differ acrosslanguage and space, it is difficult to accountfor the full possibilities of how religion is rep-resented online, especially at the global level.For example, Jesus is both the central figure inthe Christian faith and a common first name inSpanish-speaking countries. Such differencesin meaning can easily lead to the recordingof false positives when performing automatedqueries, whether in Google Maps or any otherWeb 2.0 platform of interest.

A further illustration of how languageshapes cyberscapes is shown in Figure 6, whichvisualizes the result of searches conductedusing a 100-m grid, limited to Bangkok,Thailand. At each point, searches for thekeyword temple in both English (Figure 6A)and Thai (Figure 6B) were conducted. Basedsolely on the differences in language use, thereare obvious spatial discontinuities between thetwo maps. Clearly, English speakers and Thaispeakers envision and inscribe different spacesassociated with the respective terms for temple.The analysis of language use in user-generatedgeographic information has great promise inthat it can reveal how the creation and useof this information vary across cultures andplaces. For instance, returning to Figure 6,English-language references to temple inBangkok mostly refer to high-profile touristdestinations, whereas references to the wordin Thai are in many cases closely associ-ated with alternate sites of local religiouspractice.

These differences reveal the thorny problemof language in selecting keywords, but thereare a number of methods through which theseissues can be addressed. A straightforward wayis to simply restrict analysis to regions withlargely uniform language use, as was done inFigure 3, which examined Christian denom-inations within the United States. Althoughthis sharply limits the range of cases studiespossible, it does serve to constrain the variabil-ity in results based on language differences.Second, one can continue to use terms fromone language (e.g., English) across multiplelanguage areas but be extremely cautious ininterpreting the results, as was shown in theglobal distribution of religious referencesin Figure 2. This is primarily a means toidentify new avenues for research across arange of languages and subjects. Finally, onecan conduct searches using the same keywordtranslated into multiple languages that arerelevant both topically and geographically.This is largely restricted to clearly defined casestudies of specific cities or regions (as in thecase of Bangkok or Israel and Palestine).

The final issue of the analysis is the ques-tion of scale. Although it is possible to mapgeoreferenced information at any number ofscales—from the urban to the global—and thusexamine a variety of different spatial patterns,the analysis remains necessarily constrainedby the fact that data collection is limited tothe aggregate of this information. That is,although we can find where the highest con-centrations of the keyword Christianity existand can conceive of a range of possible reasons,we cannot understand any context beyond itsbeing tagged to that particular location. Whocreated a reference to Christianity? Why?Was the reference positive or negative? Whatother things were included in this particularplacemark? What potential effects would thisparticular representation have on anotherperson? All of these questions, although inter-esting and important, exist at the microscale,affective level, which cannot be examined usingthe methods employed in this article. Thisweakness, however, can easily be translated intoa strength by pursuing in-depth case studies ofreligious cyberscapes that interrogate the issuesraised here. Indeed, an important goal of thisarticle is to call for exactly this kind of extendedanalysis.

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2


Figure 6 Temple in Bangkok, Thailand: (A) references in English; (B) references in Thai. All imagery ©OpenStreetMap contributors, CC-BY-SA. http://www.openstreetmap.org. (Color figure available online.)

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2


Conclusion

This article offers an innovative look into new,informal spaces of religious practice online thatKong (2001a; 2001b) contended are a key com-ponent of a new religious geography and howthese virtual spaces are coconstitutive of thematerial spaces and places with which they areassociated. As religion extends outward fromchurches, mosques, and temples into these in-formal online spaces, it is important to devisenew means for measuring and understandingthese emerging practices to complement andinform existing methods used to study religion.Moreover, the approach and insights offered inthis article on religion can easily be extendedto a variety of other cultural markers and socialactivities.

Although the first step of this work is simplymapping the contours of this phenomenonin a way similar to the earliest ecclesiasticalgeographies, it is done without the intentof evangelizing. Rather, it is done to betterunderstand the multiplicity and mundanityof religious practices and how the everydayand the sacred, the material and the digital,are made and remade. The analysis presentedhere clearly shows how offline practices areinscribed into online representation, but it alsoillustrates how these reflections have a logicof their own. Old material patterns persist butare layered and filtered through these newactivities within the digital realm.

Of particular interest are spaces of religiousconflict in which virtual religious practice of-fers digital extensions to existing disputes overvoice and claims to territory. As online informa-tion becomes more deeply integrated into dailylife, understanding the intersections of the ma-terial and the digital becomes increasingly rele-vant. Especially important to spaces of conflictare the ways in which certain voices or reli-gious inscriptions are alternatively promotedor excluded from online and offline represen-tations. In short, to study religious geographiesin the twenty-first century, one must necessar-ily understand the blending of age-old materialpractices with newly formed digital platformsand representations. !Notes1 For further explanation of this method, see Graham

and Zook (2011).

2 The use of representations within the Googlesearch engine only provides insights into a subsetof the total amount of content available online and,indeed, alternate sources of content like Twitter,Flickr, or Wikipedia could also have been used tomeasure ecclesiastical geographies. Google is theworld’s most popular search engine, however, andthus the ways in which places are represented in itare a key source of spatial knowledge online.

3 Each search also specified a radius around everypoint to delineate which geocoded data to includein search results. The exact value of the radius pa-rameter varied according to latitude to adjust forthe contraction or expansion of distance betweenlongitudes near the poles and equator, respectively.In the case of the global searches, the average radiuswas approximately 16 km (10 miles).

4 Points without a colored dot either possessed noreferences to any of the keywords being analyzedor had two keywords with an equal number ofreferences greater than that of each of the otherkeywords.

5 For synagogue and mosque, whichever version ofthe keyword (English or Hebrew/Arabic) returnedthe most references at a point was used as the value.For example, if synagogue received twelve hits atone location but the Hebrew text for synagoguerecorded twenty hits at the same location, twentywas used as the value to be compared to referencesto mosque and church.

6 These percentages are aggregates of the statisticsfrom the CIA World Factbook entries for Israel,the West Bank, and the Gaza Strip.

Literature Cited

Ben-David, E., G. Langlois, A. Mager, and E.Weltevrede. 2008. Cyberlands. Digital Meth-ods Initiative. http://wiki.digitalmethods.net/Dmi/CyberLands (last accessed 16 August 2010).

Birch, P. 2010. Announcing Google Earth 5.2.Google LatLong blog. http://google-latlong.blogspot.com/2010/06/announcing-google-earth-52.html (last accessed 26 February 2011).

Boulton, A. 2010. Just maps: Google’s democraticmap-making community? Cartographica 45 (1):1–4.

Brace, C., A. R. Bailey, and D. C. Harvey. 2006. Re-ligion, place and space: A framework for investi-gating historical geographies of religious identitiesand communities. Progress in Human Geography 30(1): 28–43.

Brunn, S. D., and E. Long. 2000. The worldviews ofsouthern seminaries: Images, mission statements,and curricula. Southeastern Geographer 40 (1):1–24.

Btelsem. 2002. Jewish settlements in the WestBank: Built-up areas and land reserves. http://

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2


www.btselem.org/download/settlements map eng.pdf (last accessed 16 August 2010).

Cairncross, F. 1997. The death of distance: How thecommunications revolution will change our lives. Cam-bridge, MA: Harvard Business Press.

Campbell, H. 2005. Making space for religion inInternet studies. The Information Society 21 (4):309–15.

———. 2007. Who’s got the power? Religious au-thority and the Internet. Journal of Computer-Mediated Communication 12 (3): 1043–62.

Castells, M. 2001. The Internet galaxy: Reflections onthe Internet, business, and society. New York: OxfordUniversity Press.

Chakraborty, J., and M. M. Bosman. 2005. Measur-ing the digital divide in the United States: Race,income, and personal computer ownership. TheProfessional Geographer 57 (3): 395–410.

Cheong, P. H., J. P. H. Poon, S. Huang, and I.Casas. 2009. The Internet highway and religiouscommunities: Mapping and contesting spaces inreligion-online. The Information Society: An Inter-national Journal 25 (5): 291–302.

Cooper, A. 1992. New directions in the geography ofreligion. Area 24 (2): 123–29.

Crang, M., T. Crosbie, and S. Graham. 2006. Vari-able geometries of connection: Urban digital di-vides and the uses of information technology.Urban Studies 43 (13): 2551–70.

Crawford, T. W. 2005. Stability and change onthe American religious landscape: A centrographicanalysis of major U.S. religious groups. Journal ofCultural Geography 22 (2): 51–86.

Crutcher, M., and M. Zook. 2009. Placemarks andwaterlines: Racialized cyberscapes in post-KatrinaGoogle Earth. Geoforum 40 (4): 523–34.

Elwood, S. 2008. Volunteered geographic informa-tion: Future research directions motivated by criti-cal, participatory, and feminist GIS. GeoJournal 72(3–4): 173–83.

Fischer, C. 1994. America calling: A social history of thetelephone to 1940. Berkeley: University of CaliforniaPress.

Goodchild, M. 2007. Citizens as sensors: The worldof volunteered geography. GeoJournal 69 (4):211–21.

———. 2009. Neogeography and the nature of geo-graphic expertise. Journal of Location Based Services3 (2): 82–96.

Gorman, S., and A. McIntee. 2003. Tethered con-nectivity? The spatial distribution of wireless in-frastructure. Environment and Planning A 35 (7):1157–71.

Graham, M. 2009. Wikipedia’s known unknowns.Guardian Unlimited 2 December. http://www.guardian.co.uk/technology/2009/dec/02/wikipedia-known-unknowns-geotagging-knowledge (lastaccessed 16 August 2010).

———. 2010. Neogeography and the palimpsests ofplace: Web 2.0 and the construction of a virtualearth. Tijdschrift voor economische en sociale geografie101 (4): 422–36.

———. 2011a. Cloud collaboration: Peer-production and the engineering of the Internet. InEngineering earth, ed. S. Brunn, 67–83. New York:Springer.

———. 2011b. Time machines and virtual portals:The spatialities of the digital divide. Progress in De-velopment Studies 11 (3): 211–27.

Graham, M., and M. Zook. 2011. Visualizing theglobal cyberscape: Mapping user generated place-marks. Journal of Urban Technology 18 (1):115–32.

Graham, S., and S. Marvin. 1996. Telecommunicationsand the city: Electronic spaces, urban places. Londonand New York: Routledge.

———. 2001. Splintering urbanism: Networked infras-tructures, technological mobilities and the urban condi-tion. London and New York: Routledge.

Harvey, F. 2007. Just another private–public part-nership? Possible constraints on scientific infor-mation in virtual map browsers. Environment andPlanning B: Planning and Design 34 (5): 761–64.

Isaac, E. 1965. Religious geography and the geogra-phy of religion. In Man and the earth, University ofColorado Studies, Series in Earth Sciences No. 3.Boulder: University of Colorado Press.

Ivakhiv, A. 2006. Toward a geography of “religion”:Mapping the distribution of an unstable signifier.Annals of the Association of American Geographers 96(1): 169–75.

Kong, L. 1990. Geography and religion: Trends andprospects. Progress in Human Geography 14 (3):355–71.

———. 2001a. Mapping “new” geographies of reli-gion: Politics and poetics in modernity. Progress inHuman Geography 25 (2): 211–33.

———. 2001b. Religion and technology: Refiguringplace, space, identity and community. Area 33 (4):404–13.

———. 2006. Religion and spaces of technology:Constructing and contesting nation, transnation,and place. Environment and Planning A 38 (5):903–18.

———. 2010. Global shifts, theoretical shifts:Changing geographies of religion. Progress in Hu-man Geography 34 (6): 755–76.

Marston, S. A., J. P. Jones III, and K. Woodward.2005. Human geography without scale. Transac-tions of the Institute of British Geographers 30 (4):416–32.

O’Reilly, T. 2005. What is Web 2.0. O’Reilly Me-dia. http://oreilly.com/pub/a/web2/archive/what-is-web-20.html (last accessed 16 August 2010).

Ortega, F., J. M. Gonzalez-Barahona, and G. Rob-les. 2008. On the inequality of contributions to

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2


Wikipedia. Paper presented at the 41st annualHawaii International Conference on System Sci-ences, Manoa, HI.

Proctor, J. 2006a. Introduction: Theorizing andstudying religion. Annals of the Association of Amer-ican Geographers 96 (1): 165–68.

———. 2006b. Religion as trust in authority: Theoc-racy and ecology in the United States. Annals of theAssociation of American Geographers 96 (1): 188–96.

Shirky, C. 2008. Here comes everybody: The power of or-ganizing without organizations. New York: PenguinGroup.

Shortridge, J. R. 1976. Patterns of religion in theUnited States. Geographical Review 66 (4): 420–34.

Stump, R. 1991. Spatial implications of religiousbroadcasting: Stability and change in patterns ofbelief. In Collapsing space and time: Geographic as-pects of communications and information, ed. S. D.Brunn and T. R. Leinbach, 354–75. London andNew York: Routledge.

Townsend, A. 2001. The Internet and the rise of thenew network cities, 1969–1999. Environment andPlanning B: Planning and Design 28 (1): 39–58.

Turner, A. 2006. Introduction to Neogeography. Se-bastopol, CA: O’Reilly Publishers.

Warf, B. 2001. Segueways into cyberspace: Multiplegeographies of the digital divide. Environment andPlanning B: Planning and Design 28 (1): 3–19.

Warf, B., and P. Vincent. 2007. Religious diversityacross the globe: A geographic exploration. Social& Cultural Geography 8 (4): 597–613.

Warf, B., and M. Winsberg. 2008. The geography ofreligious diversity in the United States. The Profes-sional Geographer 60 (3): 413–24.

Zelinsky, W. 2001. The uniqueness of the Ameri-can religious landscape. Geographical Review 91 (3):565–85.

Zook, M. 2000. The web of production: The eco-nomic geography of commercial Internet contentproduction in the United States. Environment andPlanning A 32 (3): 411–26.

Zook, M., and M. Graham. 2007a. From cyberspaceto DigiPlace: Visibility in an age of informationand mobility. In Societies and cities in the age of in-stant access, ed. H. J. Miller, 241–54. New York:Springer.

———.2007b. Mapping DigiPlace: Geocoded In-ternet data and the representation of place.Environment and Planning B: Planning and Design34 (3): 466–82.

TAYLOR SHELTON is a doctoral student in theGraduate School of Geography at Clark Univer-sity, 950 Main Street, Worcester, MA 01610. E-mail:[email protected]. His research interests lie at theintersection of economic and Internet geographies.

MATTHEW ZOOK is an associate professor in theDepartment of Geography at the University of Ken-tucky, 1422 Patterson Office Tower, Lexington, KY40506. E-mail: [email protected]. His research focuseson the impact of technology and innovation on hu-man geography with a particular focus on digital andgeospatial technologies.

MARK GRAHAM is a research fellow in the Ox-ford Internet Institute at the University of Ox-ford, 1 St Giles, Oxford OX1 3JS, UK. E-mail:[email protected]. His research focuses ongeographies of the Internet, and the role of infor-mation and communication technologies within thecontext of economic development.

Dow

nloa

ded

by [U

nive

rsity

of K

entu

cky]

at 0

5:27

24

Oct

ober

201

2

the technology of religion: mapping religious cyberscapes · este art´ıculo combina estudios...

Documents