Linguistic Entropy and Chinese Whispers

I am certainly not the only one observing that the workplace has a tendency to make any word that described an interesting or innovative concept almost meaningless over time.

The word “Agile” is an often cited example of a concept that has taken on so many different meanings over time that it’s information content has gone down to almost zero. Personally, I found the evolution of the expression “Minimum Viable Product (MVP)” especially striking: I first came across the concept in Eric Ries’ book, so I associated the word with the idea of building something that is just good enough to attract early adopters with the stated purpose of validating or disproving a number of hypotheses about the market and the problem space. Ries calls this “validated learning”.

I was surprised that even though the word MVP has made it into standard business jargon, no one seems to employ the word in the same sense that I had initially seen it used: Usually it just seems to mean “version 1.0 of whatever we are building”. Evidence for this difference in meaning is that the person mentioning the MVP almost never goes on to listing the hypotheses that they would want to test.

Language is not static, if everyone agrees that MVP simply means “version one” then that is not by itself a good or a bad thing. I do not get to decide right or the wrong use of a word. People also use the word “literally” in a way that is different from its original meaning. That might trigger some grammar-nazis but, at the end of the day, we still understand each other.

What can happen, however, is that we all of a sudden find ourselves without a clear word to describe the original concept. As soon as the same word takes on more than one meaning, it becomes almost useless. When I want to talk about an MVP by Ries’ definition, I now have to resort back to saying “a release that is principally designed to test the following hypotheses…”. That makes our language a bit less precise. We could talk of a form of linguistic entropy.

How did we get here? One thing I noticed is a trend to define a word by Chinese Whispers: A word is used frequently in meetings and instead of going back and trying to find a definition of what that word means, participants assume they know what it referred to (e.g. “a release that does not yet contain a lot of functionality”) and decide to go with that definition.

I find this understandable. We are bombarded with tons of jargon every day. We cannot expect people to read a business book for every meeting that they attend. A second reason why I find that a natural response is that, in most cases, it does not matter what the dictionary thinks a word means. I want to find out what my boss wants me to do and therefore it makes sense to try to guess their definition of the word.

An Anti Entropy Function

To counteract entropy, I would like to propose the following strategy:

  • Don’t use jargon if you can avoid it.
  • If you need to use jargon, provide a short and clear definition of the word at the top.

Consider the following example of the finest business babble: “We will leverage the cloud to adopt lean and agile development practices”.

And replace this with the following: “We want to allow teams to iterate quickly and autonomously. As there are a lot of things we do not know about the market we are in, we want to formulate hypotheses which we will test with small controlled experiments (‘Lean Development’). Both goals cannot be attained if teams are blocked by a gatekeeper and we think that the bottleneck created by our own server infrastructure is such an obstacle. We will therefore ask teams to maintain their own infrastructure resources using the data centers of a third party vendor (‘Cloud’).”

Of course, this is is more verbose. However, I would argue that the information content of the original phrasing is not only zero but negative: It contributes to linguistic entropy in a way that reduces the information content of all subsequent conversations on the topic. The new phrasing makes it clear to the reader, how the expressions are used. The fact that the rephrased sentence contains a higher information value is highlighted by the fact that there now is something to argue about. For example: Is server infrastructure really the factor limiting the development team’s cycle time?

In summary, vague use of jargon is a practice that will not go away. Shaking my fist and complaining about a decline of language will not help. What I can do is deal with the consequences: As entropy increases, the only remedy is to be very mindful in one’s own use of jargon and to never assume that the everyone in the audience understands an expression in the same way.

Leave a comment

Blog at WordPress.com.

Up ↑