Related Links

Web Seminars

Taxonomies are everywhere in information management, but they are hardly ever formally acknowledged and managed.

Most code tables (lookup tables) contain taxonomies - perhaps even the overwhelming majority of them. Yet we seem to have no consistent guidelines on how to design the taxonomies that will be used to populate these tables.

Also poorly understood is the difference between those who create taxonomies and those who have to actually use them. A case in point: Anyone confronted with filling in the average government form will likely come across a question where a choice must be selected from a list of alternatives that are difficult to interpret. No doubt the choices were intelligible to the designers of the form, but that is no guarantee that they can be understood by anyone obliged to complete the form.

Semantic Humor

It is unfortunate that little is taught about taxonomy, or many other aspects of traditional logic, in the educational systems of the West. However, it is possible to find some interesting articles scattered across the literature. One of them is by Jorge Luis Borges, entitled “The Analytical Language of John Wilkins”. In this piece Borges claims to produce a taxonomy from a Chinese encyclopedia called “The Celestial Emporium of Benevolent Knowledge.” This is a hoax - no such encyclopedia ever existed. However, the taxonomy that Borges produces is rather interesting. It purports to classify animals as listed in Figure 1. (Please note that the original essay was in Spanish and several variants of the taxonomy exist in English. The table is my translation from the Spanish.)

What's Wrong with This?

Borges' taxonomy seems rather exotic and has sparked a lot discussion in academic circles. Much of this centers on comparisons of Western and Non-Western thought. For instance, Michel Foucault is quoted by Wikipedia as describing his reaction to the taxonomy “...that shattered, as I read the passage, all the familiar landmarks of thought — our thought, the thought that bears the stamp of our age and our geography..."

I disagree. If the academics who made such comments about Borges' taxonomy had bothered to look into a few of the code tables implemented in databases right here in the U.S., they would have discovered examples of taxonomies at least as bizarre. And they would have been told by the creators of these taxonomies just how logical and necessary they were. I remember one code table I reviewed that was for customer credit level and had entries for "Gold," "Silver," "Bronze," "Employee," and "Suspended." How could "Employee" and "Suspended" be in such a table? But they were, and program logic was built around them.

Even in publicly available taxonomies we can find issues that echo what we see in Borges. For instance, in Figure 2 is a classification of Financial Instrument extracted from the venerable International Monetary Fund.

What is a Taxonomy?

If we are to criticize taxonomies - including what Borges produced - we better know what a taxonomy is. This is difficult to find out because, as noted above, the subject that deals with taxonomies, traditional logic, is hardly taught in the West anymore. However, if we look into texts of traditional logic, we find that taxonomies are formed by breaking a generic concept (a "genus") into more specific concepts ("species") that compose it. There are some rules about how this should be done.

For instance, "the basis of division must remain constant," and "the species must exhaust the genus." This is not the place to get into a treatise on traditional taxonomies, but rather to note that there is literature about how they should be created and governed.

However, beside the traditional top-down method of forming taxonomies (properly called "logical division"), there is also the bottom-up process of classification. This can be used to group any number of objects according to any need we have. For instance, the 10 things I would take out of my house if it were on fire have no commonality among them as such (e.g., the cat, my children's photographs, my PC, etc.) other than my purpose to prevent them being destroyed by fire. In classification we are not bound by the same rules as logical division. Rather, there is some common way in which we deal with the objects in a classification.

Now, Borges' taxonomy does not make any sense from the top-down perspective. However, it might be more allowable as a classification if some purpose could be found for it. And perhaps that is what Borges is hinting at - not that the universe is so constituted that animals fall into his classification, but that some people have a reason (admittedly unknown) for grouping animals in this way. If we could find the reason, we would understand the taxonomy.

Unfortunately, in information management, taxonomies often seem to be little more than grab bags of concepts thrown together, neither dividing up a general concept, nor aggregating distinct concepts for a specific purpose. Hopefully, as semantics progresses, we will see a lot more clarity brought to bear in this area.

Malcolm Chisholm, Ph.D. has over 25 years of experience in enterprise information management and data management and has worked in a wide range of sectors. He specializes in setting up and developing enterprise information management units, master data management, and business rules. His experience includes the financial, manufacturing, government, and pharmaceutical industries. He is the author of How to Build a Business Rules Engine and Managing Reference Data in Enterprise Databases and Definition in Information Management. He writes numerous articles and is a frequent presenter on these topics at industry events. Chisholm runs the websites http://www.bizrulesengine.com, http://www.refdataportal.com and http://www.data-definition.com. Chisholm is the winner of the 2011 DAMA International Achievement Award.