2 Core concepts
2.1 Representation, interpretation and communities of practice
Let us start with a thought experiment.
Where is the music?
The music is in the musical notation.
No, the music is in the mind of the composer.
No, the music is in the performance.
No, the music is in the hearing.
For ‘music’ read ‘knowledge’.
‘What is knowledge?’ is obviously a weighty question. In thinking about knowledge specifically in relation to ICT, we will introduce a key concept: representation. Representation is the way in which information is made manipulable and shareable. Representations may be designed to be written and read by people or computers, providing the basis for human-human, computer-computer and human-computer interaction. Their design involves formulating a language to describe some aspect of the world. To create a representation (computer-based or not), we must codify information; that is, translate it into the vocabulary and grammar of the particular language. Codification is therefore critical, as we will see in Section 2.2.
A further concept for clarifying distinctions between information technology and knowledge technology is interpretation. Representations embody information, but they are useless – indeed meaningless – unless someone or something (another computer) interprets them. We define interpretation here with an intentionally practical orientation: the process of assessing information (perceived via one or more representational medium) with respect to a goal (for example, to solve a problem; to judge someone's character; to gauge the tone of a meeting). Once information has been interpreted, you have knowledge for action, even if it is to decide that the information is irrelevant. A consequence of this view of knowledge is that information may be interpreted in many different ways – its significance depends on the reader. As an expert, you may be able to glance at a spreadsheet and immediately spot a statistical trend that a junior member of staff has missed; or you may glean implicit messages from reading a memo which a less experienced colleague would miss.
An implication of this view is that, while information technologies deliver data structured using different representations, knowledge technologies – if we can justify this term – will be distinguished by their support for interpreting those representations: for instance, by making it easier to access and understand context, or at least, by linking to people who can help supply missing context. A further implication is that, if knowledge is the contextualised interpretation of information, a computer can be said to ‘know’ something in a limited sense if it has the ability to reason about information; that is, if it can interpret information with a notion of ‘context’ and act appropriately on it in some way. We describe such systems in ‘Example: an “intelligent” email system’ in Section 4.3.
If context and interpretation are key to ascribing meaning to information, a radical constructivist implication is that, since each of us encounters the world through our own particular lens, in principle there is no such thing as ‘codified knowledge’ in an artefact, whether digital or paper.
A particular view of the relationship between knowledge and technologies has been set out. Do you agree with the idea that it is not possible for ‘knowledge’ to exist in a form that can be stored and embodied in objects and documents, or do you think this is rather an extreme view? Is it in fact the case that manufactured objects do contain the design knowledge that went into them, and that patents and books do contain knowledge? Communities of practice, for example, are a foundation for effective teamwork and the sharing of tacit knowledge.
If you answered ‘no’ to the first question and ‘yes’ to the second, does this mean that the knowledge embodied in an object is the same regardless of who interprets it? After all, we do not normally expect an object to change its properties unless someone modifies it. You may find it helpful to think about documents or objects in your own work context. Perhaps some objects are more ambiguous (that is, more likely to change their meaning) than others? But then, surely it depends on the people who are interpreting them?