From data to action: Leveraging insights to make more informed collection decisions

“How many books are there on Islamic law?”

This was the deceptively simple question posed by a colleague while brainstorming the potential scope of a digitization project on Islamic legal history. For non-librarians, it might seem like something one could just “look up.” But as you and I know, that kind of analysis only begins with a simple question and branches out into many more complex issues of data analysis, accounting for duplicate titles, different editions, and multiple languages, etc.

But I decided to give it a go and survey a selected number of libraries in North America using the tool Choreo Insights. I published the in-depth results of my findings in another post on the Islamic Law Blog, but today wanted to talk more about the process.

Defining the whole, exploring specifics

My source is Choreo Insights, an analytics tool developed by OCLC for collection development, shared print, and other management purposes. It allows us to analyze and compare library collections by accessing data directly from WorldCat.

When first starting to use the tool, I wanted to work toward that “big number.” Something close to the total of all Islamic law texts in library collections, or at least a place to start. So, I chose 14 large North American libraries with holdings in WorldCat (listed here by number of holdings that matched my search): Harvard, Library of Congress, Princeton, Yale, Columbia, NYU, UC Berkeley, University of Michigan, Stanford, University of Pennsylvania, UCLA, University of Chicago, Cornell, and McGill.

I searched for “Islamic Law” in the FAST heading filter—and when I searched across all these libraries I returned a set of 46,032 distinct title editions.

Now, this is obviously not a complete answer to the original question. But it gives us a reasonable “collective collection” to then perform deeper, specific explorations.

Different lenses on the collective collection

Once I had my initial list, I could start to do some research based on what’s important to my discipline. And also, what I’m personally curious about. So, I dove into:

  • Top languages (Arabic, English, Persian, Indonesian, and Urdu)
  • Countries of publication (Egypt, Lebanon, Iran, Saudi Arabia, and Indonesia)
  • Dates of publication over the 19th, 20th, and 21st century

These queries yielded really interesting results! For example, it revealed the relatively high standing of Indonesian language titles on Islamic law, as compared to languages like Urdu and Turkish. It also showed that around 65% of the titles have been published since 1990. Which is confirmation of something I expected to find—that the growth of the library collections within this subject corresponds with the growth of Islamic law as a field of research in North American universities.

More titles vs. more holdings

With a little more digging, I found another interesting detail. Even though the majority of titles (30,671 out of the aforementioned 46,032) were published in Arabic, the most popular titles were, by far, ones published in English, with several being owned by more than 500 libraries across WorldCat. The most popular Arabic title about Islamic law was held by only 257 libraries, and with only a few more held by more than 100. The numbers for other languages were even more limited.

Even more fascinating to me was that ALL the authors of the works published in Arabic are authors from the medieval period, not works by modern scholars. This is, again, consistent with the growth of Islamic legal scholarship in the west, by modern authors, publishing predominantly for a North American and/or English-speaking audience.

Acting on insights

The ability to easily probe into the collection information from a number of large academic libraries on a specific subject proved very useful. In the past, doing research like this would have required working with an API or doing many queries and then sorting and cleaning them up. Both of which require specific expertise. This tool was very easy to use, and the spreadsheet exports allowed me to dig even further into the results using Excel.

While the data in my research does not represent “everything” on the subject, it’s good enough to make some conclusions and recommendations:

  • The largest proportion of the collections are from Lebanon, Egypt, Iran, and Saudi Arabia; their publications dominate the study of Islamic law. Librarians looking to diversify their collections might look for resources from other areas, including especially Africa, South and Southeast Asia, Central Asia, Turkey, and the Balkans.
  • The bulk of these holdings are only around 30 years old. This provides an opportunity to expand our attention to materials published in the early 20th century and before. These materials will often be more vulnerable to physical disintegration due to the nature of their paper and structure, and less easy to locate. They should be prioritized for retention, both locally and as part of shared print processes.
  • This data also makes priorities for digitization more apparent. The information here can be compared against Islamic law materials in digital collections such as the HathiTrust and Arabic Collections Online to see which titles still haven’t been digitized.

It’s important to work toward actionable insights like this, because what is the good of new knowledge if we don’t put it to use?

An impossible question with very possible results

What started as an impossible question—and still probably is—yielded some very helpful information. While large groups and consortia can certainly do things together to “operationalize their collective collections,” with tools like Choreo Insights, any individual researcher, educator, cataloger, or collection development specialist can go in and do some quick, easy research to get an idea of strategies that can help:

  • See what’s unique to your library—and possibly important for retention and digitization
  • Diversify your collection based on language, period, and country of publication
  • Compare to libraries whose programs you admire, or with whom you could collaborate or extend resource sharing opportunities
  • Make sure you have what you need for teaching purposes (using CIP codes, available as a search criteria within Choreo)

It’s an exciting way to pursue new avenues of research and collection development. Not just based on our own experience—or interesting questions asked at conference—but backed up by a huge amount of shared library metadata.