Why Categorising Information Sources is Important?

Walk into any library or open any database, and you are surrounded by books, journals, reports, encyclopaedias, and datasets. On the surface they all look like “information.” But information professionals never treat them as one undifferentiated pile. They sort sources into clear groups-most commonly primary, secondary, and tertiary. This act of sorting is called categorisation, and it is far more than a tidy housekeeping habit. The way a collection is categorised tells you how good it is, what kind of work it supports, and how easily you can find what you need. This post explains why categorising information sources matters so much, and where the system still struggles.

What categorisation actually means 🔗

Categorisation is the practice of grouping recorded and unrecorded sources according to the nature and originality of their content. The most widely accepted scheme was popularised by the British information scientist Denis Grogan, who divided documents into primary, secondary, and tertiary levels based on how much the original information had been reorganised. Primary sources are first-hand records: original research articles, raw data, patents, conference papers, theses, and eyewitness accounts. Secondary sources analyse, repackage, and comment on primary material-think review articles, textbooks, and indexes. Tertiary sources, such as encyclopaedias, directories, and bibliographies, condense and point to the other two.

The key insight is that these labels describe the relationship between information and the context in which it was created, not just the physical format. A peer-reviewed journal can hold a primary research paper or a secondary review article. A textbook is usually secondary, but if you study the history of textbooks themselves, that same book becomes a primary source. Categorisation forces us to ask what role a source plays, and that question is the foundation of everything that follows.

Evaluating an information collection 🔗

One of the most practical uses of categorisation is judging the quality of a library or research collection. A list of total holdings tells you very little on its own. A library could own a hundred thousand items and still be weak for serious research if almost all of them are tertiary-general encyclopaedias, guidebooks, and popular handbooks.

Reading the balance of a collection 🔗

When you categorise a collection, you can see its true profile. A strong research library is expected to hold a healthy proportion of primary sources-original journals, conference proceedings, dissertations, and datasets-because these are the raw material of new scholarship. Primary sources provide direct, first-hand evidence, while secondary sources show how that evidence connects to the wider field. A collection heavy on primary and secondary material signals depth; one dominated by tertiary works signals a collection built for general reference or beginners rather than advanced inquiry.

This evaluation guides real decisions. A librarian planning acquisitions for a university with active doctoral programmes will look at the existing balance and deliberately strengthen the primary and secondary layers. A public library serving a general community may quite reasonably stock more tertiary and popular material. Without categorisation, there is no honest way to assess whether a collection matches the people it is meant to serve.

Matching the collection to its users 🔗

Categorisation also exposes gaps. If researchers in a particular department repeatedly request original studies that the library does not hold, the imbalance becomes visible the moment you sort the collection by type. This makes categorisation a quiet but powerful tool for collection development, helping institutions spend limited budgets where they will do the most good rather than buying more of what they already have in plenty.

Understanding research trends in a field 🔗

Categorisation does not only describe a single library. When applied across the published literature of a subject, it reveals how that subject behaves and how mature it is.

How the literature reveals a field’s character 🔗

Information in a discipline moves through a predictable cycle. Findings are first shared informally, then presented at meetings, and finally published as primary sources. Only later are they indexed, reviewed, and absorbed into secondary and tertiary works. By categorising the literature of a field, you can read where that field sits in this cycle. A subject producing a steady stream of original research papers, fresh datasets, and new conference proceedings is clearly research-oriented and active. A subject whose output is mostly review articles, compilations, and reprints of established knowledge is more settled or consolidating.

This is why the mix of source types is treated as a signal of research maturity. A young, rapidly advancing field is rich in primary literature because so much remains to be discovered. A mature field has had time to review and synthesise its findings, so secondary and tertiary works accumulate. Neither pattern is “better,” but categorisation lets analysts and policymakers tell them apart at a glance.

Why this matters for students and planners 🔗

For a research student, knowing this helps decide where to begin. Secondary research is usually done first to learn the vocabulary of a field, identify leading theories, and find gaps, before turning to primary sources for fresh evidence. For institutions and funding bodies, categorising the literature shows which fields are growing, which are stagnating, and where investment in original research is most needed. In short, the categorised shape of a literature is a map of intellectual activity.

The challenges of categorisation 🔗

For all its usefulness, categorisation is not a neat, finished science. Several stubborn problems remain, and they have real consequences for how easily information can be found.

The boundaries are blurry 🔗

The line between categories is rarely sharp. Documents such as monographs, conference proceedings, and theses can carry the features of both primary and secondary sources at once-presenting new findings while also reviewing earlier work. Because of this overlap, some information scientists argue the three-fold division has limited practical value. Even Grogan acknowledged that there is no rigid boundary between the levels; his scheme is the most accepted, but it is a useful convention rather than a law of nature.

Categories shift between disciplines 🔗

A second difficulty is that the meaning of “primary” changes from one subject to another. In the sciences, a primary source is typically a study reporting original experimental data. In history, it is a document or artefact from the period being studied. In literature, it is the creative text itself. A biology textbook is secondary within biology, yet becomes a primary source if your topic is the history of science education. Categorisation therefore depends on the question being asked, which makes any single universal rulebook hard to enforce.

No universal system, and what that costs retrieval 🔗

This brings us to the deepest challenge: there is no one universal categorisation or classification system that everyone uses. Libraries rely on different schemes-Dewey Decimal Classification, Universal Decimal Classification, and Library of Congress Classification among them-each with its own logic. These systems were largely built in the bibliographic world and have rarely been used outside their field of origin, so they struggle to link similar information scattered across different domains.

The cost falls on information retrieval. When categorisation is inconsistent, a user searching one system may miss material that a different system would have surfaced. The explosion of digital content makes this worse. The online landscape brings a huge volume of interdisciplinary material, weak standardisation, and inconsistent metadata, all of which confuse traditional category-based methods. A source that does not fit cleanly into an established category can effectively become invisible to a searcher who relies on that category to find it.

Efforts to bridge these gaps continue. Classification schemes have been adapted to organise electronic resources and improve subject browsing, and newer approaches using artificial intelligence and machine learning are being explored to handle large, messy datasets. But the underlying tension remains: categorisation is most powerful when it is consistent, yet knowledge is too varied and too fast-moving to be captured by any single fixed system.

Why the effort is still worth it 🔗

Despite these challenges, categorisation remains indispensable. It gives librarians an honest way to evaluate and build collections, gives analysts a way to read the health and direction of a research field, and gives every user a structure for navigating an otherwise overwhelming mass of material. The imperfections do not cancel the value; they simply remind us that categorisation is a thoughtful judgement, not a mechanical sorting task. Understanding both its strengths and its limits is what separates a skilled information user from a confused one.

What do you think? If you were assessing your own college library, would the balance of primary, secondary, and tertiary sources tell you it was built for serious research or for general reference? And as more knowledge becomes digital and interdisciplinary, do you think a single universal categorisation system is achievable-or even desirable?

References

1. https://library.uwsp.edu/InformationSourcesInTheSciences
2. https://www.scribbr.com/working-with-sources/primary-and-secondary-sources/
3. https://libguides.jcu.edu.au/scholarly-sources
4. https://www.researchgate.net/publication/271196954_Proposed_Maturity_Assessment_Framework_for_a_Research_Field
5. https://research.com/research/primary-research-vs-secondary-research
6. https://umb.libguides.com/PrimarySources/secondary
7. https://arxiv.org/pdf/2205.01395
8. https://www.lisedunetwork.com/knowledge-classification/