Richard Lemmens website

Copyright:
PublicDomain
This text content and maps on this page are in the public domain. This means you are free to copy, adapt, share, distribute and even capitalize on it.

Knowledge management

Theoretical background and hands-on tips

Introduction

This document gives a definition of the term "knowledge" and explains the phenomenon "knowledge management". It shows why the latter is so important and so difficult at the same time.

Management summary

In the modern world and especially the ICT sector, knowledge is the key factor in discerning oneself among other companies and gain a competitive advantage. Knowledge is of a higher order than information. Information is the carrier, knowledge the ability to act on it. ICT systems can facilitate information sharing, but knowledge sharing is the exclusive domain of humans - at least for the coming decennia.
The most important aspect of knowledge management is facilitating and promoting knowledge sharing, teaching and learning. This is difficult because it is an activity with more long-term than short-term rewards. However, with a good vision and insight the difficulties can be overcome.

The importance of knowledge sharing in the modern economy

Throughout history, workers and companies have focused on different kinds of activities to "earn their pay". There has been a steady evolution, that has accelerated in the last two centuries.

Nowadays methods of agricultural and industrial production and providing service in the western world show little difference between companies. Information handling on the other side, has not yet professionalized to a similar high degree and it is here that competitive advantages may be created. Knowing how to optimize logistics, what state a company is in, what the desires of current and even potential customers are, how to combine existing knowledge and resources into new products before the competition does - those are examples of decisive knowledge that can make the difference between loss and profit. Quote from Nonaka & Takeuchi: "In an economy where the only certainty is uncertainty, the sure source of lasting competitive advantage is knowledge."

This is especially true for companies operating in the ICT, which after all has information and communication as its core business. ICT companies need to develop continuously, adapt to constantly changing environments, adopt and integrate emerging technologies, etc. They must be able to absorb a huge stream of information, select what is relevant and spread it around to the right people.
Most readers have realized that the most important assets of an ICT company, especially software vendors and service companies, are not its buildings, machines or stock, but its people. What some are just beginning to realize is that the people themselves are not that important - modern employees tend to do a lot of job hopping - but the knowledge inside their heads is. What a few have really understood is that expert knowledge is all too often tied to the expert and that in order to make it a company asset, it must be shared to other experts in the organization. Geniuses are rare and will stay rare, teams of experts don't have to be - and can be just as ingenious.

What is "knowledge"?

bits
The term "knowledge" is frequently confused with "information". However, it is of a higher degree of abstraction, as the following list of makeshift definitions shows:

See appendix A for detailed descriptions of these terms. The most important difference is that between information and knowledge, so that will be explained in some more detail below.

Information is a code-stream that can be understood because its syntax and semantics are known, as the definitions above show. It is something that can be written down, put in a course, translated or stored / transmitted in other ways. Information can be shared and copied easily, as the 20th century has shown clearly.
Knowledge on the other hand is personal, i.e. it is different for every person. One can gain knowledge by not only receiving information, but also interpret it. This interpretation depends on previous experience, personal habits and methods, attitude and intelligence. In a formula:
    "knowledge" = "information" multiplied with "personal mindset"
Because of this, two persons will interpret the same piece of information differently and each gains a different insight/understanding/knowledge. A very important deduction of this is that it is easy to copy information, but quite impossible to copy knowledge - how would you copy years of life experience?

By now it is hopefully clear that it is not possible to transfer knowledge from one person to another as such. To "transfer" knowledge, one has to "condense" it into information, transfer the information over to the other party and have them interpret the message. It is obvious that the knowledge gained on the receiving side depends heavily on two factors:

These two factors relate to two well-known but seldom understood processes, called teaching and learning.

Learning

Learning is obtaining knowledge, by gathering, absorbing and interpreting information. There are roughly four methods of learning:

In real life humans don't limit themselves to one method of learning, but practice all four methods in a mix, using the best method available for given circumstances. A person can be teacher on one subject and a student on another. A good knowledge management should therefore embrace all four methods and allow for varying roles.

The most important aspect of knowledge management is knowledge sharing, i.e. condensing knowledge into information and spreading it around so that multiple persons can learn it. Of course the information that is being shared must be relevant to the audience but that is usually of less importance, as people are generally well able to filter the daily stream of information.

The difficulties of knowledge sharing

knowledge sharing
Sharing knowledge is harder than it looks. The need to "translate" knowledge into information and back into knowledge hinders the rapid transfer of knowledge.

This is but one obstacle. In older companies, one may still find the "attitude of the invaluable expert". This is adopted by some people who have become the single expert in a certain field of knowledge. They are of immense value to the company, because if they would for some reason stop their work, all activities that make use of their knowledge cease. The experts realize this and anxiously guard their private knowledge, thus ensuring that they can maintain their position of power. By making company knowledge their own knowledge, they have made themselves indispensable. This kind of situation is hard to change on short notice, but easy to turn around in time - all knowledge ages and becomes less relevant through time.

The expert unwilling to share his knowledge is rare in modern companies. Rather there are plenty of experts who are willing to share their knowledge, but are not very good at doing so, just don't have the time or don't see the relevance. All too often one encounters the attitude of "They can always ask me if they want to know something!". However knowing what's available to know becomes rather hard in this case.

Effective knowledge management

A definition: "Knowledge management is managing the process of knowledge sharing among people". The previous sections have shown the difficulties of this process, but all problems can be overcome. To set up a good knowledge management, one must:

It is a flaw of human nature that we tend to have a short-term view of most things. Knowledge sharing is a long-term activity: you first have to sow and can reap only some time later. Many people want to see rewards quickly and discard anything that does not match that criterion, including knowledge sharing. This is probably the single-most important difficulty to overcome with knowledge management. It takes vision and dedication to overcome, but aren't these values that all companies like to promote among their employees?

Further reading:

Appendices

Appendix A: A definition of knowledge

This section gives a working definition of knowledge. As with many concepts, there is no single right definition for it, as opinions differ. However, the definition given in this document stresses the more important parts of this concept.

To arrive at a definition of knowledge, we must start at the bottom. Here we find so-called symbol-streams. In a definition: Symbol-streams are (sets of) coded representations of phenomena from the world around us.
This does not say anything about the code being used, in fact we tend use many different codes to describe things. Examples of symbol streams are files on a computer disk, but also telephone lists, train timetables, demographical time charts, series of flag signals used by navies, shopping lists, a message in Morse, a letter in French. You can no doubt think of many more.

Symbol streams in themselves are not really useful. If you don't understand the code that was used, it's all gibberish to you. For instance if you don't know the internal format of a PDF file, you'll find it very hard to read. This can happen with non-computer codes too. Take one example from the previous paragraph, if you don't know French, a letter in that language is impossible to read, unless you have some knowledge about other (preferable Roman) languages. Once you know the code you can actually read the symbols and decode them. With that, they become data. Data are symbol-streams combined with their underlying code, that defines their structure and inter-relations (syntax).

Having (and grasping) the code allows you to read data, but still it's not very useful. Consider the following small code fragment from a computer program:
    $i = 1;
    $y = $x;
    while ($i < $n) do
    {
        $x = $x * $y;
        $i = $i + 1;
    }
If you knew the syntax of this language, you could check that the code is syntactically correct. But what does it actually do? To grasp that, all the statements, operators and variables in it must be given a meaning. If I tell you that ...

  • $i, $n,$x and $y are variables that act as containers for values,
  • "$i = 1" means "the value 1 is stored in $i"
  • "$y = $x" means "the value of $x is stored in $y"
  • "while A do B" means "do B again and again until A is not or no longer true, don't do B if A is not true at the beginning",
  • that "$i < $n" means "the value of $i is smaller than the value of $n",
  • "$i = $i + 1" means "the new value of $i is the old value of $i increased by 1",
  • "$x = $x * $y" means "the new value of $x is the old value of $x multiplied by the value of $y"
  • and that - besides the looping - all statements are processed sequentially from top to bottom

... then you should be able to understand it, assuming you have some knowledge about arithmetic and English (sounds like a safe assumption).
With this set of "meanings", the raw data of the computer program has become information. Information is data with a meaning (semantics) assigned to it.
Things start to become interesting know. We can send and receive messages, code and decode them and also understand them. There is a next and last step, however.

That step is to make use of the information. Process it and deduct its context and implications to arrive at a thing called knowledge.
If your read the computer program example above with an eye for its effects, you'll see that it computes "$x to the power of $n". Now here's an understanding of the code, something that allows use to make use of it, for instance make it a part of a larger computer program.

Another example. Take the following shopping list:

  • 1 Chinese cabbage
  • 2 apples
  • cottage cheese
  • 100 grams of raisins
  • pot of honey

But what does this mean to you, if I tell you (supplying some additional syntax and semantics) that all items on the list are to be used in the same meal and that it is complete except for herbs and spices? You could conclude (1) that this is a vegetarian meal and that maybe the persons who are going to eat it probably have at least one vegetarian among them. If you combined this knowledge with some information on the protein content of the ingredients listed, you could also conclude (2) that the person who assembled this recipe did know how to create a vegetarian meal that makes up for the protein loss caused by the absence of meat.
These two conclusions can be called knowledge. They allow you to actually do something with the message. In a definition: Knowledge is information combined with understanding that allows a person to act on it.
You see that you cannot arrive at the two conclusions without some interpretation knowledge already at hand. Knowing that all the ingredients of the recipe exclude meat is rather common, but knowledge of their protein content is not. Some people know, other's don't. Thus it is part of their "private" knowledge.

Appendix B: Tacit knowledge

Some knowledge is very hard to transfer at all, as it is deeply integrated with one's mindset. To "learn" the knowledge the audience must adapt their mindset, a thing that may be accomplished only gradually.

Consider the example of karate. I could show you a "nagashi jodan junzuki with gaiwan uke jodan" (a punch at the head combined with an evasive body movement and a deflection with the other hand) but somebody new to the martial arts wouldn't be able to grasp that in an instant.
I could spend a quarter of an hour showing every move individually and explaining important aspects and the purpose of each - that would lead to understanding with an intelligent audience. But you still wouldn't be able to duplicate it properly ... lacking the balance, speed, agility and coordination to perform it well. You would need to have the knowledge not only in your head, but also in your body. That is something that can only be trained in time, being almost "physical knowledge".
Physical knowledge is hard to transfer, especially when it has become so much a part of you that you don't even realize anymore it is there. Not all knowledge is like that. Instructions on how to operate a machine, for instance, can be coded well into a manual. Read it, memorize it, get some understanding of the function of each control and you'll be able to handle it. Still, there is advanced knowledge involved. Anyone can learn to operate a mechanical crane in a few hours. But to able to open beer cans with it requires months or years of training and exercise.

"Physical knowledge" is but one example of tacit knowledge. Every person knows the example of the professor who can devise the most marvelous theories, but fails to explain them. When he tries, he just "skips" steps in the reasoning process that his students need to grasp the concept. Why? Because to him, these steps are "obvious", he knows them without paying any conscious attention them. They are his "subconscious" or tacit knowledge.
If the students want to understand the theory, they have two options:

  • The first, to get the professor to identify his tacit knowledge and explain it, thereby making the tacit stuff "explicit". This is an example of transforming knowledge (the professor's tacit knowledge) to information (the professor's extended explanation).
  • Work themselves up to the same level as the professor, by learning the "missing steps" and making these part of their own mindset too - this may take some time ... This is an example of transforming information (the additional studying material) to knowledge (the students newly acquired knowledge).

Appendix C: Why computers are stupid

Computers are known to be extremely able information handlers. They can calculate faster than any human and a few can even beat Gary Kasparov in a chess match. However, computers are generally completely outwitted by humans if it comes down to adaptivity and creative solutions for problems. The reason for this is the very limited ability of computers to combine knowledge. Most software focuses on a very narrow, precisely defined problem field, where the designers have explored almost every possibility. Here the computers can excel and show off with their rapid computation abilities. Of course designers make their applications as flexible as possible. A word processor has no idea what the first word of a newly created document will be, but it can be sure that it will be made up using one of a limited set of character code pages used in the computer world. But what if it got fed Egyptian hieroglyphs? You could give it the Rosetta stone* too but it would not be able to use that to translate the hieroglyphs into Greek. Humans can (with difficulty).
The difference here is made by the background knowledge of humans, who are able to combine all kinds of knowledge that the computer does not have. Humans can make all kinds of combinations and do so continuously, sometimes arriving at very creative solutions and ideas. To get computers at the same level, they would have to be able to store amounts of information many orders of magnitude greater than they are now, greatly improve their indexing and searching abilities, significantly enhance their computation speed and expand their parallel processing capabilities a zillion times.
In short: computers will be able to beat humans in a few selected areas for quite some time, as they do now. However before they rival us on all fronts, some very major progress still need to be made. But who knows what will happen, see how far computing got in only half a century!

* The Rosetta stone is a famous object that shows a text in three languages: Egyptian hieroglyps, Greek and demotic. It allowed linguists to make the first decodings of hieroglyphs.