Péter's Digital Reference Shelf
May 2004

Title: Adherents.com
Publisher: Preston Hunter
URL: http://www.adherents.com
Cost: Free
Tested: April 19-20, 2004

The Context

These days when religion, as well as the religious affiliation of groups and individuals, are in the forefront of daily news and in the background of many decisions and actions, this free compendium is a useful ready-reference source. It complements the many encyclopedias and dictionaries of religions that feature essays, short and long articles about the history, theories, doctrines, practices, major figures, works and symbols of various religions.

Almanacs are excellent and inexpensive resources for some gross statistical data, but they usually cover only the major religions. For example, the Britannica Almanac's religion census data is provided by the foremost authority of religion demographics, David B. Barrett, the author of the World Christian Encyclopedia, which deals with 10,000 religions and sects.

The encyclopedias and dictionaries are often monumental works, decades in the making, with contributions by hundreds of theologists and other specialists. In contrast, Adherents.com is essentially the work of one person — Preston Hunter — with help form other people who are given credit for even minor contributions, an unusually generous approach. Hunter is neither a theologist nor a statistician, but a programmer interested in the factual/statistical aspects of adherence to faith.

A statistical compendium and fact-based Web site maintained by a single person may not sound exactly reassuring for quality ready-reference work, but it is not impossible, as there are some remarkable free Web sites posted by individuals who could not have done it in a print-only world. Such is the case with the outstanding Encyclopaedia of the Orient, which I reviewed four years ago and which remains a splendid resource. Adherents.com is another labor of love by someone who spared no time and efforts to trace down, extract, analyze, synthesize and document nearly 45,000 references from predominantly scholarly sources of religion demographics. It must have required the dedication of a monk.

The Content

The nearly 45,000 pieces of statistical data in this collection are for 4,200 churches, religions and faith groups (although religion is broadly interpreted and includes some groups that are more political and life-style touting than religious, such as the Adamites). The coverage is extremely broad-ranging, from the now-defunct Abecedarians to the former followers of Zurvanism.

The data have been collected mostly from print books (about 60%), while the rest is from journal articles, Web sites and personal correspondence. The 430 books and 25 journals that provided the backbone for the statistical compendium are listed in one of the many informative (sometimes a tad too talkative) FAQ pages. (If you are wondering why the numbering of sources begins at 0, I have a guess. Preston Hunter was working on his bachelor's degree in computer science when he started the project and in computer programming cyclic operations start counting with zero.) Both the books and the periodicals represent the most important sources of demographics and/or religion-related factual information.

Same Faith, Different Locations, Different Times

If more sources had information about the adherents of a religion, or if the same source had different information for different time periods or data were available for the population for different geographic entities (the predominant situation), a distinct record was created for each.

To understand the data organization and representation you must look at this screen shot, as it well illustrates the approach. It shows that Hunter had several sources regarding the Cao Dai religion. One of them was Edward Rice's book. On the same page, Rice mentions two population data for different periods. Hunter created a record for the reference to the 1934 population of adherents in Vietnam, and one for the 1940s. He uses 1945 for the decade and, based on the text, he lists three countries, as the religion spread to Cambodia and Laos. For two additional sources, he created two more records — one where the total number of adherents did not change but the years are different, and one because the number of countries changed as Britannica specifically mentions France and the U.S. beyond Vietnam and Cambodia, but omits Laos.

The records include the name of the religious group, their geographic location, the number of adherents, the number of countries where there are or were adherents, the year the data refers to, the source consulted (including complete bibliographic citations with the page number), and quotes from the sources and other notes. Obviously, not all records have each data elements.

If the data were gathered from a Web site, there is a link to the original location. Many of them are not free anymore, or are not at the original address, but can usually be figured out. For example, the links to the free Web version of the 1914 edition of the Catholic Encyclopedia are dead because it changed its base address from www.csn.net to www.newadvent.org, but the rest of the URL is the same and can be reconstructed. For example, the non-existent www.csn.net/advent/cathen/01035b.htm can be easily changed to the correct www.newadvent.org/cathen/01035b.htm address. Such predictable address changes could be taken care of by programmatic global search-and-replace commands.

Precious Excerpts

There are quotes from the original sources to show the context for the data, as well as useful notes and comments in the last column alerting users to possible errors or discrepancies in the source information. Once again, this is an arduous job and it spares users, who may not have access to the primary documents, a lot of time.

Although there are some typos, the data collection and recording are incredibly meticulous, providing a gold mine of historical perspectives and corroborative data. Researchers and reference librarians can't help but appreciate Hunter's analysis and synthesis of so many sources. Hats off to him for religiously providing details that must have taken an enormous amount of time. For Cao Dai alone, for example, he has 20 references and thus 20 records. There are about 45,000 references processed. Using a spreadsheet for recording and presenting data was not a good idea, as I discuss in the software section, because it has negative consequences for the users, no matter how impressed they might be by the work.

Indexes

There is an alphabetical index listing the religions/beliefs. If there is more than one reference, their number appears next to the index entry. If there is a number between the index entry and the number of analyzed references, such as 300,000 next to the index entry for Quakers, it indicates the numbers worldwide for the religion. If there is more than one reference with worldwide data, then the most recent is listed in the index. For example, in the case of the Quakers, the 1998 data appears in the index. The index links to the exact location of the first record in the spreadsheet for the religion, so you will have to scroll down to get to the world data, as they are usually the last entries (unless there are statistics for Zambia, Zimbabwe, etc.).There you can see the other references to the number of worldwide adherents for the religion.

The references are sorted within religion by geographic names (country, region, state, province, etc.) and within that by date. So the records for Shi'ites begins with data for Afghanistan from 1992 to 1998 and have a total of nine records, as for some years there were multiple references, and ends with the Yemeni Shi'ite population counts in different references.

The other approach is looking up the index by location. The index entries here are governed by the references, but the country names are "normalized," using the current official name of the country. Therefore, even if the source refers to Burma, it appears under Myanmar — a convenience for the geopolitically challenged. This consolidation is not always possible. For example, there are two entries for Yemen from 1992 and 1994 (after the merger of North Yemen and South Yemen), and one entry for the Yemeni Arab Republic (a.k.a South Yemen) from 1984. This entry does not include data for North Yemen, it just mentions that its population is equally divided between Shi'ites and Sunnis.

Other Predefined Access Points

There are other predefined, interesting summary lists, including the religious affiliations of U.S. politicians and other famous people. Many of the detailed statistics and lists are deep in the site, but with some clicking around they can be found. The best lead-in for this is to get to the page that lists 22 of the world major religions and follow the links for most details. You can find revealing statistics at a deeper (but still aggregate) level that are usually not available in the almanacs.

The Britannica Almanac, for example, has details broken down for the six standard ecclesiastical groups of Christians, but has only a single line for Muslims. It does not even break down the data by the two major sects (the Sunnis and the Shi'ites), even though the sectarian division was the pretext for the Iran-Iraq war, as well as for the decimation of the Shi'ite population in Iraq. The table that presents the distribution of the adherents to Islam shows that about 88% are Sunni. One does not need pundits to understand how the minority Iraqi regime could get away for decades with the massacre and brutalization of the Iraqi and Kurdish Shi'ite majority in the country — without condemnation by Arab League members.

The site is rounded out with links to external sites if you need further information.

The Software

Beyond the indexes, there is also a search option, but it is the weakest point of the software. It yields a result list that is totally uninformative, showing cryptic URLs that contain your search term somewhere on the humongous pages. You are much better off searching the site using Google as a proxy as its result list will provide context and the cached version of the page is displayed by automatically highlighting the matching words on the page.

The underlying problem with the software is that a spreadsheet program was used to record the data elements. Using a spreadsheet must have made the process of entering data enormously redundant. A good text management software with relational database management functions, such as DB/Textworks, could have saved significant amounts of time for manual data entry. It could have also provided authority control for book and journal titles, author names and some other data elements, and presented the data in a more versatile and more efficient way.

The tabular format could have been excellent for at-a-glance identification of key data in the records. However, by forcing the long bibliographic entries and the precious, but even longer, notes and quotations into wrap-around cells of 10-12 positions wide, few tables can preserve the tabular look, many of them practically fall apart, losing the advantage of at-a-glance comparison of related entries.

DB/Textworks could have produced the same tabular look by limiting the wrap-around fields to three rows maximum, making the rest of the content to be scrollable within the cell, and preserving the layout. It could have also offered many advantages, including authority files, look-up tables for recurring values, like often cited authors and titles, and great flexibility in formatting result display and formatting.

While some of the above advantages cannot be exploited with the existing record structuring, it is still possible to load the nearly 45,000 records into a relational database.

This is a worthy project that really does deserve some financing by one of the many publishers that make good profits on religious publications. While at first glance it is a statistical database about religions, it is much more than that. It is a valuable compendium of religion demographics with impressive sourcing, excerpts and notes providing context for the raw data.