Systematic Indexing by Kaiser: A Structured Approach to Information Retrieval

When most people think about library indexes, they picture an alphabetical list pointing to documents. Julius Otto Kaiser flipped that assumption on its head. Working as a special librarian in the early twentieth century, he argued that an index should organize information itself, not just the documents that carry it. His method, published in 1911 as Systematic Indexing, introduced a disciplined, rule-based way to break subjects into parts and recombine them into predictable strings. More than a century later, his ideas still echo in faceted classification, thesaurus construction, and database design.

What Kaiser set out to solve 🔗

Kaiser worked in commercial and technical settings where librarians had to answer very specific questions quickly. A general classification scheme like Dewey could tell you which shelf held books on trade, but it struggled when a manager wanted a precise fact buried inside many different reports. Kaiser’s response was to index the units of information scattered across documents rather than the documents as whole objects.

This shift matters. A traditional subject catalogue treats a book as a single item with one or two headings. Kaiser instead pulled out the individual ideas contained in the literature and filed each one separately in a card index. His system was originally developed for special libraries as a way of indexing pieces of information drawn from many documents within a single card file. He also insisted that index terms be as specific as possible, because in a business context a precise fact is far more useful than a broad one.

Concretes and processes: the building blocks 🔗

The heart of Kaiser’s contribution is the idea that almost any subject can be split into two fundamental kinds of terms. He called them concretes and processes.

Concrete terms 🔗

A concrete is a thing, an entity, or a commodity. In Kaiser’s commercial world, a concrete was anything that could be the object of trade. Examples include steel, copper, cotton, or railways. Kaiser actually distinguished three kinds of concretes: movable concretes (objects that can be shifted from place to place), immovable concretes (things fixed in the landscape), and the curiously named abstract concretes, which largely covered human labour viewed as something with exchange value, such as a manager or an engineer.

Process terms 🔗

A process is what happens to a concrete, the action it undergoes or the condition attached to it. Kaiser defined processes as either the conditions attaching to concretes or the actions they undergo. Examples include manufacture, export, analysis, or distribution. On its own a process tells you little, but paired with a concrete it produces a meaningful, searchable idea.

The third category: countries 🔗

Kaiser later recognized that geography mattered enormously in business information. A trader did not just want to know about copper; they wanted to know about copper in Chile or Spain. So he added a third category, countries (or places). This gave him a richer scheme. In his fullest version, a subject string could combine concrete, country, and process. Kaiser even treated the three-part form as the archetype, stating that a complete statement strictly speaking should consist of a concrete, a country, and a process.

Statements: how the parts combine 🔗

Kaiser did not stop at sorting terms into categories. His real innovation was combining them into what he called statements, which are compound subject strings built according to strict grammar-like rules. This is where his method reads almost like a language with its own syntax.

The rules were precise. The first term in a statement had to be a concrete or a country, and the final term always a process. Concretes and countries were interchangeable in that opening position, but the process always closed the string. So an indexer might build statements such as:

Copper – Manufacture (concrete + process), or Copper – Spain – Export (concrete + country + process), or Spain – Copper – Mining (country + concrete + process).

Because the position of each term was fixed by its category, two different indexers working on the same material would tend to produce the same heading. Kaiser saw this consistency as the foundation of a reliable, ever-growing card index. He compared these correctly formed statements to the fixed points that hold the whole system together.

The double-entry method for place-based subjects 🔗

One practical feature of Kaiser’s system deserves special attention: how he handled subjects tied to a particular place. Because a country and a concrete were interchangeable as the lead term, the same piece of information could be entered under more than one heading. This produced a kind of double entry.

Consider a subject like the economic development of a particular nation. Under Kaiser’s logic, this could be filed once with the place as the lead term and again with the activity reorganized into a concrete-and-process form. A reader looking under the country name would find it, and so would a reader approaching from the subject side. The same applies to a topic such as the literature of a country: an entry under the place name and a corresponding entry built around the activity itself both point the user to the same material.

This multiplication of access points is the core benefit. Kaiser understood that people search from different starting angles. By generating more than one entry for place-related subjects, he raised the chance that a user would meet the information wherever they happened to look. The trade-off was a larger index, but in a special library where precision mattered, that cost was acceptable.

A few of Kaiser’s term-formation rules 🔗

Kaiser was a stickler for consistency, and his book laid down detailed rules for how terms should be written. Terms for concretes were generally to be expressed in the singular form, although countable goods in his trade indexes sometimes appeared in the plural. Inverted noun phrases, the kind of awkward reversed headings common in other systems, were strictly avoided. Where a term was ambiguous, a qualifier in brackets was added to pin down its meaning. These small disciplines all served the same goal: a predictable index where the form of a heading followed automatically from the nature of the subject.

Why Kaiser still matters 🔗

Kaiser’s specific scheme is rarely used in its original form today, but his thinking shaped the field in lasting ways. He is widely regarded as one of the first to use general categories as the basis for structuring compound subject strings, a technique that later became a hallmark of facet analysis.

The bridge to faceted classification 🔗

The most important line of influence runs to S. R. Ranganathan. By separating subjects into categories such as concrete, process, and place, Kaiser anticipated the faceted approach that Ranganathan would develop into Colon Classification. The librarian and theorist Elaine Svenonius described Kaiser as rightly the precursor to Ranganathan, crediting him as the first to see the usefulness of facets in building expressions within a synthetic index language. Researchers today position Kaiser as a pioneer of the analytical-synthetic method and therefore a starting point for Ranganathan’s faceted theory.

Influence on modern indexing 🔗

Beyond facets, Kaiser’s emphasis on consistency, specificity, and the relationships between terms fed into later work on thesaurus construction and controlled vocabularies. His insistence that headings follow systematic patterns improved agreement among different indexers, a principle that underpins modern subject indexing in specialized databases and business information systems. The idea that a subject can be analysed into its components and then synthesised into a structured string is now standard practice in many retrieval tools, even when the people using them have never heard his name.

A quick summary 🔗

Kaiser’s contribution can be reduced to a few durable ideas. He indexed information rather than whole documents. He analysed every subject into concretes (things), processes (actions or conditions), and later countries (places). He combined these into rule-governed statements with a fixed grammar. And he used multiple entries, including the double entry for place-based subjects, to give users several routes to the same fact. Each of these moves pushed indexing toward the analytical, synthetic, and faceted methods that dominate knowledge organization today.

What do you think? If Kaiser were designing an index for a modern digital database, how do you think his concrete-and-process categories would map onto tags, metadata fields, or search filters? And can you identify a present-day search tool that, knowingly or not, relies on the same idea of breaking a topic into parts before recombining them?

References

1. https://www.ideals.illinois.edu/items/46775
2. https://journals.lib.washington.edu/index.php/nasko/article/view/15867/13297
3. https://journals.lib.washington.edu/index.php/nasko/article/view/16373/14079
4. https://www.researchgate.net/publication/395493094_From_Card_Index_to_Book_Index_Julius_Otto_Kaiser's_Application_of_Systematic_Indexing_to_Book_Indexing
5. https://www.inlibra.com/document/download/pdf/uuid/66cdf291-e2de-32b9-bb49-722badcb6cac
6. https://www.iskocus.org/NASKO2013proceedings/Sales_Guimaraes_ContributionsOfJuliusOttoKaiser.pdf