Knowledge Decay
A couple nights ago, I went to the Knoxville Agile Group meeting and listened to a SCRUM-Lean talk, given by Joe Little of Kitty Hawk Consulting. He’s a great presenter on Agile topics, particularly Lean. Unlike some of those other Agile jokers, err…I mean experts, he seems to know what he’s talking.
Anyway, he hit me with a term/paradigm that I had never heard of before:
KNOWLEDGE DECAY
Knowledge decay (in my mind) refers to how quickly knowledge and ideas tends to disappear from people’s consciousness. When he presented this term to me, I began to ponder the definition and ramifications of the concept he was teaching. I was starting to panic. It shook the foundation of my core assumptions about the world in general. I had always had the idea that knowledge is truly eternal, infinite, and no one can shake or destroy it. People could choose to ignore it, but once out there, it’s out there forever. Instead, I started to think about the book The Time Machine by H.G. Wells. In the book, you have a whole group of people that have no concept of the knowledge of the past . I also start to think about the Dark Ages that occur after the fall of the Roman Empire. The barbarian hordes descend on Europe, and there’s rapid knowledge decay. Only during the renaissance (“rebirth”) do the citizens of Europe start to get seriously back on track.
How do we preserve knowledge and keep it from decaying? More specifically, let’s apply why to a software development shop: development, testing, users, etc. Well, to do this, we really need to know what kind of knowledge we are preserving. Joe mentioned that there is knowledge that’s explicit and knowledge that’s tacit.
Explicit knowledge is knowledge that is concrete, fully formed, and not ambiguious. For this kind of knowledge. The written word is always the best! I hate heavy documentation, but I’m a strong advocate for writing stuff down that we might need later. The more specific the item, the more crucial it is that you write it down.
Example: In Powershell, you can launch powershell.exe and pass in the script name you want to run in the same line. However, if you have a path with spaces in it, Powershell will treat those spaces as extra parameters:
powershell.exe c:\Foo\Powershell Scripts\testscript.ps1
This script breaks because it thinks that the space between Powershell and Scripts denotes a new parameter. However, if you put an & and single quotes around the path, it’s fine:
powershell.exe & ‘c:\Foo\Powershell Scripts\testscript.ps1′
The point is that this knowledge is very explicit. It might be something you need to write down so that it doesn’t catch the next person. Granted, some people can remember this kind of stuff easily, but no person’s head is a good replacement for Google when it comes to explicit knowledge.
Tacit Knowledge on the other hand is more of an apparition. You can’t clearly define it, but you know it exists.
Example: I have testing knowledge, but if someone says, “How do you test?” I’m usually at a loss for words. The reason is that software testing knowledge is tacit. There are principles that you learn to test, but you can’t really learn to test. I learned by doing. If I were to teach someone how to test, I would teach them principles of testing, and then I would have them sit with me and watch, ask questions.
Using both written record and face time with people, you can slow Knowledge Decay by transfering your knowledge to people. The only way to keep knowledge alive is to share it. Trees bear fruit and drop seeds so that they can propagate, producing new trees for one day those mature trees will die from lightening, disease, or old age. Are you a tree that bears no fruit? Are you hording your knowledge to make yourself more valuable to your company? If you think that’s what you’re doing, then think again. Management knows that good trees bear fruit and bad trees don’t. When layoffs come, the bad trees will be the first to be chopped up and thrown in the fire. Be a knowledge pipeline moving hundreds of gallons of knowledge through you. Don’t be a swamp, holding your stagnant knowledge to yourself. Let knowledge flow through you. Do it for yourself. Do it for your company. Do it for all of mankind!
Marisa Seal
Daniel Brown
Josh Carroll
Daniel. What if someone says “How do you test process x?” Would the tacit knowledge of testing process x be if the tester just knows that when an error happens it’s fine because of learned knowledge due to the time spent testing x. Is the Utopian goal of any process to lower the amount of tacit knowledge by removing “oh yeah” from the knowledge gathering process.
Shannon "The Sensai" Turner
August 14, 2009
Wow, Shannon, that’s a very sexy thought: “Is the Utopian goal of any process to lower the amount of tacit knowledge by removing ‘oh yeah’ from the knowledge gathering process?”. I will have to ponder that paradigm. I may have to respond to your question as another blog post once I’ve spent a couple hours reflecting and meditating. Dynamite thought! Thanks man!
Daniel Brown
August 16, 2009
Good post!
So, the solution is easy, yet hard: Just-in-time knowledge creation. Especially where knowledge will decay quickly.
For more see Wikipedia, Takeuchi & Nonaka articles (by one or both), or even my blog.
Joe Little
August 17, 2009
To Shannon’s point. I think this best “utopian goal” is to get the most from the tacit knowledge (and explicit knowledge) that we have. Sometimes that means converting tacit to explicit (Takeuchi and Nonaka discuss this), sometimes just using the tacit. …My 2 cents.
Joe Little
August 17, 2009
This makes perfect sense; thanks for an excellent post. isn’t this knowledge decay a common occurrence in human history? Just ask modern day Egyptians or Indians descendants of ancient civilizations who doesn’t know anything about languages and culture of their ancestors. You can see this happen in companies as people leave and new people get hired.
Some times it could be a good thing. How much of that ‘this is how we have always done’ do you really want? Speaking of which, I think knowledge decay is more like knowledge atrophy and you are arguing for regeneration and I like it.
Rao Kasibhotla
August 25, 2009
Yes, Rao, thanks for your comment. I like the fact that you said that “sometimes it could be a good thing”. You are right! Very poignant! Sometimes, it’s better for us to just forget the knowledge of the past, especially when it keeps us from innovating and acquiring new knowledge.
Daniel Brown
August 28, 2009