Click to learn more about author Ronald G. Ross.
In this second post of a four-part series about crafting great business definitions, I will identify common pitfalls in creating definitions. I will explain how to detect when a definition is data-ish, how to avoid circular definitions, and what the difference is between a definition and a description. Read Part One here.
Definitions with overt or subtle IT or data bias are a curse for effective business communication. Good business definitions are oriented to what words mean when used by real business people talking directly about real business things. Definitions pertain to data only if the thing under consideration is a data thing.
Suppose someone defines an organization as: a data type for a body of people organized for a particular purpose.
In the real world, an organization is not a data type! It’s a group of people.
A better definition for the real-world concept of an organization would be: a body of people formed for a particular purpose.
By the way, the faulty definition above is a real example from industry standards work(!).
Digging Deeper: Perhaps the standards group was defining fields in a data design. Then missing from the definition is context.
Organization [data store]: a data type for a body of people organized for a particular purpose.
If you’re defining data, say you are defining data!
Definitions should never be circular. Circular definitions provide little or no concept clarity. Example:
Cordless phone: a phone that has no cord
This definition is circular because it includes the word cord. Does the definition actually tell you anything? No! Here’s an improved definition.
Cordless phone: a phone whose handset uses radio signals rather than electrical signals through a wire to communicate with a base.
Digging Deeper: A more subtle form of circularity is the use of the verb form for the term being defined. For example, the faulty definition of organization above includes the word organized, merely a verb form of the term. Not good.
Definitions vs. Descriptions
If a concept is adequately defined, it has been differentiated from all other concepts in scope, especially ones close in kind. What its instances can and can’t be is also clear.
Providing a definition doesn’t necessarily mean you will understand how an instance acts, looks, or feels. For example, a cat is warm and fuzzy. That’s simply a description. It doesn’t distinguish cats from dogs or panda bears.
Imagine you are a witness to a crime. The police would ask you for a description of the suspect. In response, you would tell them what the suspect looked like, how they acted, whether they had an accent or tattoos or other visible marks, and anything else that would help the police with an image of the person sought. Descriptions add details that help you form a mental image. The police would not ask you for a definition of the suspect(!).
A description creates a verbal portrait of something. Its purpose is to convey a mental image, an impression, or a better understanding. A description complements a definition but cannot substitute for it.
Digging Deeper: The suspect in a crime is an individual. In general, you should never try to define real-world individuals — whether people (e.g., William Shakespeare), places (e.g., The Eiffel Tower), things (e.g., the planet Mars), or organizations (e.g., American Red Cross). That’s a recipe for trouble. Imagine the difficulties you’d get into, for example, if you defined William Shakespeare as the person who wrote Hamlet. Then research discovers that William Shakespeare actually did not write Hamlet. Bad day! The person called William Shakespeare is still William Shakespeare, but your definition actually talks about (misclassifies) someone else.  So, describe real-world individuals all you like, but don’t define them.
Important exception: You do need to define individual things that do not correspond to anything in the real world — e.g., Santa Claus. Unfortunately, such notions are only in your mind (and the universe of discourse). Be prepared — business knowledge typically includes more such things than you might expect.
Extracted from Business Knowledge
Blueprints: Enabling Your Data to Speak the Language of the Business, by
Ronald G. Ross, 2020.
 The Stuff of Thought, by Steven Pinker, Viking, 2007, pp. 9-11