mapping religion

The Pitfalls and Promises of Mapping World Religion

I have long been dissatisfied with world religion maps, especially those that are available on the internet. To be sure, mapping religion is an inherently difficult task. Many areas contain multiple faiths, just as different places often vary tremendously in regard religiosity itself. Changes in the religious landscape, moreover, are often difficult to capture. Most of Europe, for example, is appropriately mapped as Christian when it comes to its religious heritage, but in the 21st century such a depiction is no longer completely accurate. Over much of Europe, nonbelievers now greatly outnumber believers, and in quite a few places practicing Muslims outstrip practicing Christians. Some reports go so far as to claim that in terms of actual practice, France is now more Muslim than Christian,* although this assertion is probably exaggerated.

Religious “mixture,” moreover, can characterize not just regions but also individuals. An anthropologist friend of mine once characterized the West African country of Guinea as “90 percent Muslim and 90 percent animist,” which could well be true. But animism and so-called tribal religions more generally usually get short shrift in world religion maps. The same is true for syncretic faiths such as Candomblé, which might be the dominant faith in parts of northeastern Brazil, although only around five percent of Brazilians overall report themselves to be adherents. But such numbers are themselves suspect, as it is often difficult to enumerate religious adherents. Polling and census data are partial or non-existent over much of the world, and people often fail to be forthcoming about matters of faith when asked. As a GodWeb post argues, “To put it bluntly, when asked about religious belief and practice, ordinary citizens lie. And they lie about their faith to a greater degree then they lie about their sex life, or political activity.”

World Religion Map 1Another common problem in the mapping of religions is the inconsistent division of major faiths into their constituent branches. If Christianity is divided into its Roman Catholic, Protestant, and Eastern Orthodox branches, as it often is (see the map posted to the left), then by the same token Buddhism should be broken down into its Mahayana and Theravada forms, just as Islam should be divided into its Sunni, Shia, and Ibadi branches. Making such divisions, moreover, should be done in a rigorous manner. The so-called Oriental Orthodox Christian churches, such as the World Religion Map 3Armenian Apostolic Church and the Ethiopian Orthodox Tewahedo Church, for example, should not be mapped with the Eastern Orthodox branch, as they often are (see, for example, “World Religions Map 2006” posted here) for the simple reason that they do not belong. As the Armenian Apostolic Church and the Ethiopian Orthodox Tewahedo Church reject Branches of Christianitythe Creed of Chalcedon that was adopted by the Christian mainstream in A.D. 451, they stand apart from Eastern Orthodoxy as well as Roman Catholicism and Protestantism. But as the diagram posted here shows, this situation is complicated by a number of subsequent unions of theologically disparate Christian branches.

But if the mapping of religion is inherently problematic, that does not mean that all maps of world religions are of equal value—or lack of value. Some basic maps are, of course, much better than others. Recently, moreover, a number of highly innovative and extremely detailed world maps of religion have appeared on the internet. Several GeoCurrents posts next week will examine these maps in some detail. Before doing so, however, I cannot resist pointing out how amusingly bad World Religion Map 2maps of religion can be. I would be tempted to nominate the one posted to the left for the booby prize of the worst world map on the internet.

To begin with, the map deeply distorts basic patterns of both physical and political geography. Note the seaway between North and South America, the misplacement of New Zealand, the division of North Korea into two World Religion Map detailcountries, and so on. A detail of the map’s depiction of central southeastern Europe reveals how laughable it is. But more to the point, consider its portrayal of religion: Guatemala and Costa Rica are non-Christian; Jordan is Jewish: Armenia, Ethiopia, Uganda, and Lake Victoria are Muslim, as is Taiwan; Japan is Christian; both North Koreas are “Chinese”; and Sri Lanka is Hindu. Interestingly, the site on which it is posted, which includes some fine maps of religion, merely notes that it “is a much more generalized map of world religions.” In actuality, this map verges on intellectual malpractice.

*According to a 2012 report by the Gatestone Institute:

Although 64% of the French population (or 41.6 million of France’s 65 million inhabitants) identifies itself as Roman Catholic, only 4.5% (or 1.9 million) of those actually are practicing Catholics, according to the French Institute of Public Opinion (or Ifop, as it is usually called).

By way of comparison, 75% (or 4.5 million) of the estimated 6 million mostly ethnic North African and sub-Saharan Muslims in France identify themselves as “believers” and 41% (or 2.5 million) say they are “practicing” Muslims, according to an in-depth research report on Islam in France published by Ifop.

Taken together, the research data provides empirical evidence that Islam is well on its way to overtaking Roman Catholicism as the dominant religion in France.

In Britain, Islam has overtaken Anglicanism as the dominant religion as more people attend mosques than the Church of England. According to one survey, 930,000 Muslims attend a place of worship at least once a week, whereas only 916,000 Anglicans do the same.


Mapping Islam: Bad and Good Efforts

Mapping the distribution of religious groups is often a frustrating exercise. Good data on the numbers of adherents of any particular faith or sect, let alone the intensity of their beliefs, are often lacking, while the spatial intermingling of different religions presents formidable cartographic challenges. As a result, even the best maps of religion at the global scale are deeply flawed. Britain, for example, is almost always mapped as solidly “Protestant Christian,” even though by 2005 it was estimated that more Britons were attending weekly services in Muslim mosques than in Anglican churches. As British Muslim populations are highly concentrated in urban areas, and as many non-church-goers still think of themselves as vaguely Christian, depicting Britain as a Christian land may be roughly acceptable. But more accurate cartography would portray Britain, like most of Europe, as largely secular and partly Muslim.

If mapping religion is always challenging, some cartographers meet the challenge far more adeptly than others. Some poor maps of religion may simply be a product of misinformation, but others clearly champion certain faiths or sects while disparaging others by exaggerating or minimizing their geographical extents. On the internet, such misleading maps can spread virally from one site to another, and are often viewed uncritically as straightforward depictions of religious realities. In contrast, some of the best maps of religious distribution languish on obscure web sites.

Exhibit A in my case against deceptive religious cartography is the first map reproduced above, entitled “World Muslim Distribution (Sunni and Shia) 1995,” archived in the Perry-Castañeda Library
Map Collection at the University of Texas at Austin. This map is commonly deployed as an objective depiction of the distribution of Islam; I have seen it used on dozens of websites without critical commentary. The map admittedly has a few positive features; it is aesthetically pleasing, it is constructed at a high level of resolution, and it is readily available at the Perry-Castañeda website, an indispensible source of public-access cartography. Its content, however, is intellectually hazardous.

The map has two main problems: it exaggerates the geographical expanse of Islam overall, and it minimizes the extent of Shiism in the Muslim world. As the cartographer does not specify the threshold that must be passed for an area to be mapped with the green shades of Islam, most viewers would probably assume that a majority Muslim population is required. In actuality, vast areas in India, Indonesia, and sub-Saharan Africa that are colored light green for Sunni Islam are overwhelmingly non-Muslim. In India, the cut-off point seems to be around five percent; only areas with Muslim populations below that figure are left unmarked. In Indonesia, Christian, Hindu, and animist areas are uniformly depicted as adhering to Islam. Such portrayals would not be a problem if the map specified its inclusion of Muslim-minority areas, but it does not.

The minimization of Shia Islam here is equally problematic. To be sure, the cartographer captures the main Shiite areas: Central Iran, southern Iraq, Azerbaijan, north Yemen, south Lebanon, western Syria, and the Hazara region of central Afghanistan. Many other Shia zones, however, are either left unmarked or unduly diminished. Such errors are readily apparent in the depiction of the Middle East, posted in expanded format in the second map above. Here oil-rich Khūzestān in southwestern Iran is incorrectly depicted as Sunni, the Shiite areas in Saudi Arabia’s oil-rich Gulf region has been reduced almost to the vanishing point, the majority Shiite population of Bahrain is invisible, the Zaidi Shi’ite zone of northern Yemen is incorrectly depicted as terminating at the Saudi border, and the Alevi (highly heterodox Shiite) area of eastern and central Turkey is reduced to a few splotches. Minimizing the Shiite presence in the Arabic-speaking countries of the Gulf area is especially problematic, as it is the focus of much international and internal tension. In mid-September of this year, for example, violence erupted in Bahrain after its Sunni government cracked down on Shiite dissidents. One cannot understand Saudi Arabia’s fears of Iran and of a Shiite-dominated Iraq, moreover, without grasping the extent of its own deeply marginalized Shiite population living in its main oil-producing area.

Vastly better mapping of religion in the Middle East is easily accessible. See, for example, the third map posted above, produced by Mehrdad Izady as part of Columbia University’s Gulf 2000 project. Izady’s map is intricate. It simultaneously depicts religion and demography; virtually unpopulated areas are left white, while sparsely settled areas are shaded more lightly than densely settled areas. It also shows zones of religious mixture. Izady’s differentiation of Wahhabism from Sunni Islam on the map may be controversial, but note that he acknowledges such controversies on the map itself rather than trying to paper them over.

To my knowledge, Izady’s cartographic works for the Gulf 2000 project are not simply the best available portrayals of the geography of religion in the Middle East, but are actually the finest examples of cultural cartography, at this scale of analysis, ever produced. Yet this body of work is little known, even in academia. Even Izady’s Wikipedia article focuses on the controversial nature of his work on Kurdish history and religion, ignoring his outstanding cartographic contributions. I suspect that it is because we demand simplicity in maps, feeling overwhelmed by the complexity of the world. But for those who have the patience, there is a tremendous amount to be learned from Izady’s gorgeous maps.