Practical Applications of Bradford’s Law in Library and Information Science

Walk into any large university library and you will find thousands of journals lining the shelves and many more available through digital subscriptions. No library, however well-funded, can subscribe to every journal in existence. So how do librarians decide which journals are worth the money and which can be skipped? A surprising amount of this decision-making rests on a pattern discovered in the 1930s by a British librarian named Samuel C. Bradford. His observation, now known as Bradford’s Law of Scattering, has quietly shaped how libraries build collections, how abstracting services operate, and how researchers search for literature. Let us look at how this law actually works in practice.

A quick refresher on what Bradford observed 🔗

While working as a librarian at the Science Museum in London, Bradford noticed something consistent about how articles on any given subject are spread across journals. A small handful of journals publish a large share of the important articles on a topic. A larger group of journals publishes a moderate share. And a very large group of journals publishes only the occasional article on that topic. Bradford published these findings in 1934 and later expanded on them in his 1948 book Documentation.

The pattern can be stated mathematically. If you arrange journals in order of how many articles they publish on a subject and then divide them into three zones, each containing roughly one-third of the total articles, the number of journals in each zone follows the ratio 1 : n : n². Here, “n” is called the Bradford multiplier. So if the core zone has 5 journals and the multiplier is 4, the second zone will hold about 20 journals, and the third zone about 80 journals, with each zone still delivering the same number of useful articles.

The three zones are usually named the core (or nucleus), the middle zone, and the peripheral zone (or tail). The core journals are the most concentrated and valuable sources, while the peripheral journals are scattered widely and contribute little individually. The top third of journals represents those most frequently cited in a subject’s literature and therefore of the greatest interest to researchers.

Optimizing abstracting and indexing services 🔗

Abstracting services scan published journals, summarise the articles, and supply those abstracts to subscribers. The challenge is obvious: there are far too many journals to scan all of them, and scanning costs money. Each journal that gets monitored requires staff time, subscription fees, and processing effort. This is exactly where Bradford’s Law becomes a practical planning tool.

Because the core journals deliver a disproportionately large share of relevant articles, an abstracting service can cover a third of all the literature on a subject by scanning only the small core. To cover the next third, it must scan a much larger middle zone. To cover the final third, it must scan an enormous number of peripheral journals, each contributing almost nothing on its own. The service therefore faces a question of diminishing returns.

The cost-effectiveness trade-off 🔗

Consider a simplified illustration. Suppose a subject has 1,000 relevant articles spread across 70 journals, divided into three zones of roughly 333 articles each. The first zone contains 10 journals, the second contains 20, and the third contains 40 (following the 1 : 2 : 4 pattern). By scanning just the 10 core journals, which is only about 14% of all the journals, the service captures a full third of the available articles. To capture the next third, it would need to monitor twice as many journals. To capture the final third, four times as many.

For an abstracting service trying to balance comprehensiveness against running costs, this is decisive. If the cost of scanning the peripheral journals exceeds the revenue those extra abstracts will bring in from subscribers, the service is better off limiting its coverage to the core and middle zones. As the teaching material on the law explains, providing truly comprehensive coverage would mean scanning every journal, and the cost of doing so may exceed what the service can ever recover from subscribers. Bradford’s Law gives managers a rational basis for deciding where to draw the line.

This problem was important enough that the founder of citation indexing, Eugene Garfield, built his entire approach around it. Garfield observed that a relatively small core of journals reappears across many fields, an idea he called “Garfield’s law of concentration”. This insight allowed his Institute for Scientific Information to index a manageable set of journals while still covering most significant research, and it eventually fed into the design of large citation databases.

Library resource planning and journal subscriptions 🔗

The same logic that helps abstracting services also helps libraries decide how to spend limited acquisition budgets. Journal subscriptions, especially to international scientific titles, are expensive, and prices rise every year. A library cannot afford to subscribe to everything, so it must prioritise.

Bradford’s Law tells the librarian that subscribing to the core journals in each subject the library serves will satisfy the large majority of user demand. These core titles are the ones most likely to be requested, cited, and read. The peripheral journals, by contrast, are used rarely, and subscribing to all of them would consume a large part of the budget for very little return.

Building a core collection 🔗

In practice, a librarian can identify the core journals for a subject through a citation analysis. By examining which journals are most frequently cited in the theses, dissertations, and research papers produced by the library’s own users, the librarian can rank journals by productivity and pick out the nucleus. This makes it possible to set a lower bound on the number of journals a collection devoted to a given subject must contain.

A concrete example comes from a study of doctoral theses in Library and Information Science submitted to universities in Maharashtra. The researchers analysed 5,467 references drawn from 138 theses, spread across 798 different journals. When they ranked these journals, a small group emerged at the top: Annals of Library Science and Documentation led with 207 citations, followed by College and Research Libraries and Herald of Library Science. A library serving LIS research scholars in that region would gain the most value by ensuring these high-ranking titles are always available, rather than spreading its budget thinly across hundreds of rarely-cited journals.

This approach also guides cancellation decisions. When budgets are cut, the law suggests that peripheral journals can be dropped with minimal disruption to most users, since the core continues to meet the bulk of demand. The library avoids spending on titles that offer little value to its actual readers while protecting the sources that matter most.

A real-world example of Bradford’s application 🔗

To see how the law performs with genuine data, consider a study that applied it to identify the core journals of pediatric surgery. The researchers gathered the reference lists from articles published by pediatric surgeons, which contained nearly 200,000 references in total, and analysed the top 20 cited journals. When they fitted the data to Bradford’s three-zone model, the model identified the core journal with a remarkably good statistical fit. The core of the field turned out to be the Journal of Pediatric Surgery, alongside the leading specialist journals and high-impact general titles in surgery and medicine.

The practical value here is direct. A surgeon trying to stay current, a medical librarian building a specialist collection, or a database manager deciding which journals to index can all use this ranked core as a starting point. They do not need to monitor every medical journal in existence; covering the identified core delivers most of the relevant literature.

How a researcher uses the multiplier in practice 🔗

The law is just as useful at the level of an individual researcher conducting a literature search. A worked example often used to explain this involves a researcher with five core journals for their subject. Suppose those five journals yield 12 relevant articles in a month. To find the next 12 articles, the researcher would have to consult around 10 additional journals, giving a Bradford multiplier of 2. To find another 12 after that, they would need roughly 20 journals, then 40, then 80. Very quickly, the researcher realises there is little point in chasing the scattered journals further, because each additional dozen articles costs twice as much searching effort as the last.

This is the strategic insight Bradford’s Law offers to students and scholars. Start with the core journals to rapidly build a solid understanding of a topic and capture its most influential studies. Move into the middle and peripheral journals only when you need specialised or unusual perspectives. The law has long been regarded as an objective tool for selecting the most central sources in a field, and this strategy turns an overwhelming search into a manageable one.

Knowing the limits of the law 🔗

It is worth being honest about where Bradford’s Law falls short, because applying it blindly can mislead. The law is not statistically precise and is best treated as a general rule of thumb rather than an exact formula. Real datasets often deviate from the neat 1 : n : n² ratio, and some studies find the pattern holds only loosely.

There is also a conceptual difficulty. The law depends on knowing what counts as the “subject” of an article, yet the meaning of subject is rarely defined clearly. Scholars have pointed out that scattering can be understood in terms of words used, meaning, or topic, and these are not the same thing. The way a librarian defines a subject can therefore change which journals end up in the core. Despite these caveats, the law remains one of the most widely used and practically valuable tools in informetrics, precisely because it captures a genuine and stable pattern in how knowledge is published.

What do you think? If your library had to cut a quarter of its journal budget, would you trust Bradford’s Law alone to decide which subscriptions to drop, or would you weigh in other factors as well? And as more research moves to open-access and online-first publishing, do you think the neat zones of the core, middle, and tail will hold up, or will the pattern of scattering change?

References

1. https://en.wikipedia.org/wiki/Bradford%27s_law
2. https://www.sciencedirect.com/science/article/abs/pii/S0022480418302336
3. https://www.researchgate.net/publication/280218558_Bradford's_Empirical_Law
4. https://egyankosh.ac.in/bitstream/123456789/11366/1/Unit-6.pdf
5. https://arxiv.org/pdf/1312.3872
6. https://core.ac.uk/download/pdf/19726224.pdf
7. https://link.springer.com/chapter/10.1007/11495222_9
8. https://limbd.org/laws-of-bibliometrics-application-of-bibliometrics-laws/