2.4 Codification and formalisation continued
An important point is that the process of ‘objectifying’ knowledge brings with it a gradual change in the knowledge represented, because content and form are inextricably linked. McLuhan's famous quotation ‘the medium is the message’ highlights this phenomenon, but overstates the case a little. We can say that the medium shapes the message, as follows:
As we move from tacit, individual, pre-understanding to shared, formal, computer-based representation, we express our thoughts in an increasingly structured way, providing the computer with greater access to the content of the information. The intellectual effort required to transform knowledge representations from one state to another can lead to new insights, since the particular representation used forces us to make certain information explicit that was previously implicit. Typically, ‘information chunks’ have to be broken down into smaller units of particular classes, given names, classified and structured. Having to reason about these can clarify our thinking.
However, as we move from tacit towards finer-grained symbolic representations, we strip away details of the context(s) in which that knowledge was displayed and/or has meaning. It is usually difficult, and often impossible, to reverse the direction and recover tacit pre-understanding from symbolic representations. The ‘knowledge processes’ in Figure 2 should be understood as interpretative acts; that is, in a situation, from a perspective, for a purpose. The transformations bring about ontological changes (see Box 2.1) that unavoidably distort knowledge and ‘alienate’ it from the person possessing it in particular ways, effecting a gradual shift in definition of knowledge and expertise from an ability to a symbolically encoded fact. This critical standpoint is not intended to be ‘anti-technology’; rather, it is a principled basis on which to understand how technologies come to embody and perpetuate world views and associated value systems.
Different ways of codifying knowledge give us different languages in which to describe the world. If we take seriously the argument that the language we use to talk about the world constrains and shapes our understanding, then we can even talk about the possibility of an ‘ontological shift’ taking place – the set of distinctions we regard as important to make when describing the world changes, depending on the representation we are using and the reasons for using it.
Box 2.1 Ontology
In philosophy, an ontology refers to ‘being’ or ‘the nature of being in the world’. The term ontology has been appropriated by artificial intelligence research to mean a ‘reusable terminological scheme’; that is, a scheme for providing a rigorous description of the concepts and attributes, and their interrelationships, that are deemed relevant to describe a particular aspect of the world. Its precision means that it can serve as a ‘technical dictionary’ to ensure a common point of reference in a complex area. An ontology is an abstract knowledge model which does not need software to exist. However, a strength is that it can also be implemented as software to build a knowledge-based system. We return to ontologies in Section 4.3.