Seeking, documenting, and preserving knowledge have deep roots in Muslim history. This can be seen in prepress-era libraries developed by Muslim scholars and dynasties. Examples of such libraries include the library Al-Sahib Ibn ‘Abbād during the fourth century of Islamic calendar (tenth century AD, Iran), which held a collection of 6,200 books with a ten-volume catalog. Al-’Aziz Fatimid (tenth century AD, Egypt) had a collection of 1.6 million volumes in his library.1 The historic Dar al-Hikmah library during the Abbasid caliph Ma’mun al-Rasheed era (ninth century AD, Iraq), Khazainulqusoor during the Fatimid period with 1.6 million books, and Hakam II’s library with 400,000 books in tenth-century Spain are remarkable examples. This trend carried on during the Muslim rule in different parts of the world throughout history.²

Islamic studies continue as a vibrant field, evidenced in the volume of currently available publications. The author searched Amazon’s online catalog (www.amazon.com) (using “books on Islam” as the search phrase) in April 2010 and found 17,726 titles on Islam published since 2000. Breakdown of this search result discovered the following:

• Titles published 2001–2: 2,897
• Titles published 2003–4: 3,515
• Titles published 2005–6: 3,361
• Titles published 2007–8: 4,050
• Titles published 2009 to date: 1,286.

The same search carried out in December 2011 resulted in 21,292 titles, an increase of 3,566 titles despite the possibility that a number of titles available in the previous search could be out of stock. Another search, conducted at the same time to verify this fact with a parameter of available titles published on Islam since April 2010, resulted in 5,004 titles. This shows a continued growth in publications on Islam. Dar Al-Kotob Al-Ilmiyah, a book publisher in Beirut, Lebanon had more than 4,400 titles available on more than one hundred main topics of Islam, in the 2009 catalog.³ The same publisher’s 2011 catalog shows 5,394 titles available on Islam.⁴ A search using the term “books on Islam” in the Barnes and Noble (www.barnesandnoble.com) online catalog retrieved 24,307 titles available. A wide range of publications on a variety of Islamic subjects are available from local publishers in Muslim countries, although data of this output is not as easily available as that reported here.

According to Riazuddin, the devisors of all standard classification systems were from western parts of the world, except Ranganathan, who was from India but was not a Muslim.⁵ The environment of the devisors and their interests were the probable reasons that all these systems lacked attention to eastern areas of knowledge, especially Islam in the class of religions. DDC, which is the most popular standard classification system in the libraries of the east, provided only one of one thousand notations in the third summary. This was quite insufficient. Subsequently, many expansions were made in DDC to provide adequate placement for collections on Islam. These expansions were made either within the given space (i.e., notation 297) or alternatively utilizing the notations that were originally provided for Christianity.

Qaisar also mentioned the deficiencies of DDC regarding the class of religion.⁶ He proposed a reasonable amount of expansions in notation 297 for Islam in the sixteenth edition of DDC to cater the collections on Islam. Khurshid commented on the efforts made by Shafi for the expansion in DDC class for Islam.⁷ The summary of classes, along with subclasses devised by Shafi, is included in Khurshid’s book. These recommendations were sent for the incorporation in the eighteenth edition of DDC, but were not added.

Sardar devised a classification system for Islam.⁸ Sardar’s two objectives were encouraging a debate on the need for contemporary classification schemes on Islam and presenting a model for consideration and criticism. He invented a scheme comprising four major parts: (1) pre–main class; (2) main class; (3) post–main class; and (4) auxiliary schedules. Main classes have been denoted by capital letters, and small letters have been used for subclasses.

Chishti presented a historical review of classification in relation to Islamic materials, referencing the works of Farabi, Kawarzami, Ikhwan-Al-Safa, Ibn Sena, Ibn Hazam, Ibn Abdul Birr, Raghib Asfahani, Ghazali, Ibn Al-Nadeem and Ibn Abi Al-Rabie as the earliest efforts to classify Islamic knowledge.⁹ Chisti also described Razi’s system with sixty main classes and many subclasses. According to Chishti, the theory and practice of classification developed historically beginning with knowledge classification, moving to bibliographic classification, then book classification, and finally utility classification.

Several individuals have developed classification systems focusing on Islamic works. A classification scheme for the oriental languages collection on Islam in the Punjab University Library in Lahore, India, was developed by Labhu Ram.¹⁰ In this scheme, “Ar” for Arabic, “P” for Persian, and “U” for Urdu collections have been used as notation prefixes. Bajwa, in his master’s thesis, described contributions in the field of knowledge classification made by four ancient Muslim scholars: Al-Farabi, Ibn Al-Nadeem, Ibn Khuldun, and Tashkubrizada.¹¹ Gacek, head of the Islamic Studies Library at McGill University, described the classification system developed by Wilfred Cantwell Smith, the founder of the institute, explaining that

the Smith classification system was developed by Prof. Smith, the first founder of the Institute. The system has not been used since 1982. It was abandoned in favor of the Library of Congress Classification. It is entirely on Islam (in its broad sense) and is divided into four main classes: A. Reference works; B. Extra-Islamic subjects; C. Classical Islam (until about 1800 AD); and M. Modern Islamic world (from about 1800 AD).¹²

A project of King Abdul Aziz University, Saudi Arabia, translated DDC into Arabic, with some amendments and expansions.¹³ The class of religion (200) was the core of this project. Hassan translated DDC into Urdu with some additions in the class of religion for Islam.¹⁴ Shaniti of Egypt translated the abridged edition of DDC into Arabic and expanded the Islamic topics using DDC classes 210–60 for Islam.¹⁵

The LC revised its class of K (Law) in 2002.¹⁶ An expanded class of KBP has been specified for Islamic law. The International Islamic University of Malaysia (IIUM) adapted this expansion in LCC to cover the areas of Islamic law more granularly.¹⁷ Notations KBP140–58 have been used for this purpose and cover more than 380 topics and subtopics of Islamic law. This expansion can be extremely helpful if an independent classification scheme is developed for Islam to deal with Islamic law. Usmani and Sabzwari mentioned shortcomings of standard systems and emphasized the need for a standard classification system for Islam.¹⁸

Quaid-e-Azam Library, in Lahore, Pakistan, developed an expansion with some alterations in the structure of DDC, nineteenth edition, using 220–80, seven classes for Islam.¹⁹ The Islamic Research Institute, in Islamabad, Pakistan, has devised an local scheme, covering 152 subjects.²⁰ A list of subjects followed by Arabic numerals in ascending order has been made, and every book of a particular subject that arrives in the library is given the next serial number of the subject. Soltani has mentioned translations and expansions made in DDC and Library of Congress Classification (LCC) in Arab countries and Iran.²¹ These expansions were sent to DDC and LCC for incorporation. Unfortunately, the expansions could not be incorporated into either scheme, despite being appreciated by DDC and LCC teams.

The Indian Institute of Islamic Studies, New Delhi, developed a classification system based on UDC.²² Hundreds of topics and aspects of Islamic knowledge and literature were included in this scheme, which uses mixed notation. “IS” (for Islam) is used as a prefix in the notation. Usmani and Qudsi revealed that Shafi played a vital and leading role in developing an appropriate classification system for Islam, covering all gaps in the standard DDC system.²³ According to Fatima, the central library of Karachi (Pakistan) University was using Shafi expansions for their collection on Islam.²⁴

Riazuddin compiled an extension plan in the DDC for Islam within original notation designated by Dewey (i.e., 297).²⁵ Because this was only an extension to one number, its scope is very limited and problems like lengthy notations and congestion are present in these extensions. Gondal devised a classification scheme for Islam, which used by the libraries of the Mosques Department of Punjab University’s Quran Academy (Lahore, Pakistan), and partially in the Punjab University Central Library.²⁶ Eraqi worked on an expansion of DDC eighteenth edition’s class for Islam (i.e., 297).²⁷ The Tehran Book Processing Centre also devised an expansion in DDC.²⁸ These two expansions are similar except that the first is in Persian and the second in English. Standard subdivisions have been provided with every class. Sabzwari devised an expansion for Islam using the DDC eighteenth edition classes 210–60.²⁹ Khan devised an expansion in DDC twenty-first edition for the life, family, and companions of Muhammad (PBUH) using notations 297.63 and 297.64.³⁰

According to S. S. H. Rizwi and S. J. A. Rizwi, Dickinson’s expansions in DDC made for Punjab University Library in 1916 are considered the oldest expansions of DDC, followed by Indian Library Association, Yousufuddin Ahmad, and Molvi Shafi expansions.³¹ Altaf Shaukat also devised an expansion for Islam in DDC, which is similar to the Shafi expansion.³² Rizwi noted the extensive collections in the libraries on Islam and inadequate classification schemes to classify them.³³ He reviewed the Colon classification, Bliss classification, DDC, and LCC, and found them insufficient. Rizwi devised a schedule of Islamic topics, without provision of any notation, based on Shafi’s expansion with some amendments and additions.

Aaedi compiled an Arabic translation of abridged edition of DDC 21 with some expansions in the classes of Arabic language, literature, geography, and history, and the Islam religion. It is similar to the one made by King Abdul Aziz University in 1977.³⁴ Idrees and Mahmood determined that DDC is the only standard classification system being used in Pakistani libraries that have rich collections on Islam.³⁵ According to Sabzwari, the situation is the same in other Muslim countries.³⁶

The multitude of approaches to classifying Islamic materials described above make clear that many have sought to address the problems. A brief comparison between three local expansions (developed by the Quaid-e-Azem Library, Muhammad Shafi, and S. Mahmood Hasan Qaisar) of DDC along with a picture of historical developments in different editions of DDC regarding Islam is presented in figure 1. The figure shows that the early editions of DDC until the sixteenth edition provided only one notation for Islam without its further extensions. A noteworthy point is that the term “Mohammedanism” was used as the heading instead of “Islam” until the fifteenth edition. The later editions used the term “Islam” along with its further extension, standard subdivisions, and number-building options. The twenty-first and twenty-second editions of DDC included some drastic expansions in the notation for Islam. Additional revisions and expansions have been made in the twenty-third edition, in which Hadith (statements or actions of Muhammad) has been relocated to 297.125 from 297.124. Some topics for future development are under discussion on 025.431: The Dewey Blog (http://ddc.typepad.com), which is edited by Jonathan Furner, assistant editor of DDC. Nevertheless, the independently developed local expansions made to the earlier editions of DDC still provide more space and options for Islam compared with the DDC twenty-first, twenty-second, and twenty-third editions.

The 30 libraries from which the data were collected represent almost all types of libraries, i.e., academic, special, public, and national libraries. A majority (20) are academic libraries at universities. Five are special libraries (court, bank, and specific research institutes), 2 are public libraries, and 3 are national libraries. Table 1 shows the libraries in this study by type and location.

The participant libraries hold collections of varying size. The collections in the sampled libraries ranged from 15,000 to 2,065,000 volumes. Four libraries have collections of fewer than 50,000 volumes, all being special collection libraries focusing mainly on Islam. Five libraries have collections between 100,000 and 200,000; 11 libraries have collections between 200,001 and 500,000. Three libraries hold collections between 500,001 and 1,000,000. Seven libraries have collections of more than 1,000,000 volumes. Frequency distribution of libraries by their collection has been displayed in table 2.

Collections on Islam in these libraries range from 15,000 to 400,000 volumes. Nearly 37 percent (11 respondents) have Islamic collections in excess of 100,000. Seven libraries have collections of between 15,000 to 20,000 volumes. Eight libraries hold collections of 20,001 to 50,000. Four libraries have collections of between 50,001 to 100,000, while 9 libraries have collections of between 100,001 and 200,000 volumes. Only 2 responding libraries contain collections of more than 200,000 volumes on Islam. The composite total collection of all 30 libraries approaches the figure of 15,144,000 with an average of 504,800 volumes per library, whereas the composite total collection on Islam is 2,543,000—an average of 84,767 volumes per library.

The author contacted one person (either the chief librarian, the individual responsible for technical services, or the section head of Islamic collections) at each library for the interviews. The author sought to interview the most appropriate people who are aware of and in touch with the problem in hand. Five out of 30 interviewed subjects are working as chief librarians of their libraries, and the other 25 are working as librarians. The academic qualifications of the 30 interviewed library personnel are as follows:

• Nine (30 percent) have PhD degrees.
• Nineteen (63.4 percent) have masters degrees.
• One (3.3 percent) has a bachelor of library science (BLS) and a post–BLS diploma.
• One (3.3 percent) has a BLS degree.

No uniform practice regarding the usage of classification systems for materials on Islam was found among the libraries in this study. A significant majority (25 or 83.3 percent) use multiple systems, i.e., standard systems for general collections and local systems or expansions made in standard systems for materials on Islam. This is because the standard systems are convenient for general topics, but they do not optimally fulfill the classification needs of materials on Islam. Of the 5 libraries that use single standard system, 4 are using DDC, while 1 is using LCC. The 25 libraries that use multiple classification systems break down in the following manner:

• 18 use DDC and local expansions in DDC
• 3 use LCC and local expansions in LCC
• 1 uses LCC and DDC, plus local expansions in LCC and DDC
• 1 uses DDC and local expansions in DDC, plus an independent local scheme
• 1 uses DDC and an independent local scheme
• 1 uses LCC and an independent local scheme
• 20 are using 8 different expansions made in DDC
• 4 are using 4 different expansions made in LCC
• 1 is using 1 expansion made in UDC

Three different independent local schemes are being used for Islam. Five groups of libraries were using the same classification systems and their expansions. One group consisted of 5 libraries and the other 4 groups consisted of 3 libraries each. Only 2 of these 5 groups had coordination between the members. Figure 2 presents a summary of the classification systems in use by the sampled libraries.

Interviewees were asked whether the classification systems in use were suitable for the organization of Islamic collection being acquired in the libraries and if the libraries were satisfied. Libraries’ level of satisfaction has been classified into three main categories: satisfied (the systems are optimally fulfilling the requirements), partially satisfied (the requirements are being fulfilled so far, but, for certain materials, systems is use do not have the options needed, and unsatisfied (the available systems do not fulfil the requirements and present numerous problems and shortcomings when classifying materials on Islam). Only 1 library was satisfied with standard classification systems and 1 was partially satisfied. None of the 3 users of independent local schemes on Islam were satisfied; 2 were partially satisfied. Two out of 23 libraries using local expansions were satisfied and 17 were partially satisfied. Figure 3 shows the level of satisfaction with classification systems currently in use.

The participants were asked about the problems they had been facing and that were causing their dissatisfaction. They identified the following specific problems or shortcomings in the available standard combined with local systems:

• Allocated space for Islam in standard classification systems is too limited to cover all Islamic subjects minutely (27 participants)
• New and emerging topics are not found in local expansions of standard systems (10 participants)
• Standard systems do not comprehensively cover Islamic knowledge, so new and emerging topics are not found there (2 participants)

Additionally, the following problems were identified by 1 participant each:

• No standard was followed in developing the local system being used in his or her library, with only a list of subject terms developed.
• Ineffective and noncomprehensive classification systems are creating a result in form of improper organization and inefficient browsing of library materials.
• No uniformity exists with other libraries with similar collections because different and dissimilar classification systems are being used.
• No coordination with other similar libraries is present.
• The types and variety of materials being published on Islam exceed the provisions in systems.

The researcher asked participants for their suggestions for an optimal solution to the problem in hand. A majority (22) suggested developing a new, independent, and comprehensive classification system for materials on Islam. Two suggested amendments and expansions in standard classification systems. Two suggested both amendments and expansions in standard classification systems and a new, independent, and comprehensive classification system for materials on Islam. They further suggested that the first approach would be suitable for libraries having small or medium-size collections on Islam, and the second could be used by the libraries having larger collections on Islam. Three librarians suggested amendments and expansions in standard classification systems as their first preference if it could be incorporated in the original systems. They mentioned a new, independent, and comprehensive classification system for materials on Islam as their second priority if the amendments and expansions were not incorporated in the original systems. Finally, 1 proposed a new, independent, and comprehensive classification system for materials on Islam as the first priority and amendments and expansions in standard classification systems as a second preference.

Participants generally shared the same views on a new, independent, and comprehensive system for Islamic materials. Twenty-eight agreed that the variety and depth in Islamic topics and the volume of published literature indicate a real need for developing an independent and comprehensive classification system for materials on Islam. Twenty-eight also agreed that developing an independent and comprehensive classification system exclusively for Islamic materials was technically possible. The term “technically” did not mean technologically, but instead referred to the classification techniques, methods, and approach to develop a scheme for a specific area of knowledge. Twenty-seven agreed that guidance for developing an independent system for Islam was available both previously developed standard and local classification systems and the published literature on their shortcomings.