INFORMING THE OTHER VIEWS REGARDING SOFT KNOWLEDGE

Earlier we came to the conclusion that the Distributed Cognition and boundary object approaches are primarily representationist. We also saw that in CoPs and with Common Ground dealing with soft knowledge on its own is problematic. This resulted in the suggestion to regard knowledge as a duality. Wenger's (1998) work in CoPs with his emphasis on the participation/reification duality provides us with a meta-framework to help clarify the Distributed Cognition and boundary object, Common Ground, and CoP approaches, adding further weight to the duality view.

Distributed Cognition

There is an apparent contradiction in Distributed Cognition. Hutchins and Klausen (1991) in applying DC to an airline cockpit refer to the pilots as members of a CoP, although in this case they are referring to the wider CoP of pilots in general. Hutchins (1995a) also emphasises the cultural aspect of the artefacts:

Although some of the representations are internal, they are still cultural in the sense that they are the residua of a process enacted by a Community of Practice rather than idiosyncratic inventions of their individual users (p. 130).

There are numerous other examples of the groups where DC has been applied that could very well be CoPs, for example, the medics examined by Cicourel (1990) and the radiology groups of Rogers and Ellis (1994) and Symon et al. (1996). The best example, however, is provided by Hutchins (1990, 1995a) with his description of the quartermasters, and indeed we find that the quartermasters themselves are described as a CoP by Lave and Wenger (1991).

Although, Hutchins and Klausen (1991) emphasise the cultural aspect and refer to a CoP, we have already noted that Hutchins' Distributed Cognition approach is primarily representationist.

It is perhaps not surprising that the groups described in some of the DC literature bear such a strong resemblance to CoPs because of the emphasis it places on some CoP characteristics. Hutchins (1995a) emphasises the fluid self-perpetuative nature of the community and the social distribution of the knowledge. There are experts or gurus regarding specific aspects, but as time passes, there is a sharing of knowledge through interaction. The social distribution of knowledge gradually changes—it is the "dynamic" part of the CoP. The social distribution of the knowledge also indicates that there is some overlap. Members do know some of the same things. Hutchins (1995a) emphasises the importance of this overlapping knowledge distribution in cooperative work as it makes the system robust and able to cope with difficulties. He shows this with the example of the navigation team by criticising those analysts who see knowledge distributed between cooperative participants in a mutually exclusive manner. In the navigation team, there is some clear overlap between the knowledge of the plotter, the bearing taker, and the bearing timer-recorder, leading to a robust system. Hutchins (1990) also shows that the artefacts support this distribution of knowledge among the members of the navigation team to make it robust should an individual component fail.

Hutchins' (1990, 1995a) representationist approach is perhaps best seen in the artefacts. Wenger (1998) also uses shared artefacts in the reification part of his duality. Wenger's use of shared artefacts in the form of boundary objects and the use of shared artefacts in DC should not be seen as dichotomous but rather as complementary. DC concentrates on externalised representations of knowledge and absolute meaning, whereas boundary objects have an interpretive flexibility. This shows that domain and soft knowledge are necessary in order to benefit from the knowledge embedded in the artefact, be it a document, a tool, or a story. An outsider will understand little, a newcomer with some domain knowledge would be able to understand the story at face value perhaps and what it means, but the old-timer will be able to see new possibilities and make different interpretations.

Robinson's (1997) description of procedures as artefacts has great importance for CoPs. Procedures embody the knowledge of previous practitioners; they provide a form of accountability and cover if something goes wrong. Just as important, they are a "way-in" for newcomers to the community. A newcomer will follow the procedure. Gradually, as the newcomer becomes proficient, (s)he will not have to think as explicitly about the steps to be taken. As (s)he becomes an old-timer, (s)he may develop the soft knowledge to omit some steps or even to conceive of improvements to the procedure.

Wenger's (1998) reification/participation approach helps us to understand the apparent contradiction of Distributed Cognition being a representationist approach but referring to CoPs. The difference lies in the emphasis. As we saw earlier, the use of artefacts in DC concentrates on the harder aspects of knowledge. Wenger's duality shows us that what is lacking in DC is a consideration of the participation. His emphasis on the duality is also a reflection of Cole (1996), who emphasised the dual nature of the artefact itself, thus moving away from the hard representational aspect. Cole's view of artefacts is that they are dual in nature. They are both ideal (conceptual) and material. Their material form is made up of the physical characteristics, but they are also ideal, that is, conceptual. This means that the material form is a result of changes that have been brought about by use. The changes might also have been brought about by participation in interactions with other artefacts, with actions and with human social worlds between which they mediate. This is supported by Cook and Seely Brown (1999) who, as explained earlier, emphasised the importance of interaction. Cole's view would suggest that cultural aspects are not embedded in the artefact in the "hard" sense, as suggested by DC, but are to be found in the "ideal" form of the artefact.

DC also provides us with a complementary view to LPP. Whereas LPP is concerned with the social structure of the community and how newcomers learn, DC is concerned with the process of how work gets done, thus providing us with a functional view of the community to complement the social view.

Boundary Objects

Wenger (1998) used Star's (1989) ideas of boundary objects to further describe the role of the artefacts in his CoPs. In Wenger's case, the artefacts as boundary objects crossed boundaries between different CoPs. He explains that the connections that boundary objects create between CoPs are reificative. By this he does not mean that they do not involve participation, but, being artefacts, they use forms of reification and thereby are able to bridge different forms of participation. They enable coordination between different communities without creating a direct link, or, as Wenger calls it, a bridge.

He also explains how a boundary object reflects the relations between the communities. He gives the example of a memo created for wide distribution. A memo appears as a self-contained artefact and one would assume that the memo tells its own story. However, Wenger (1998) observes that the memo gives rise to different meanings, and these are a function of the relations between the practices involved:

When a person reads the memo, what is really going on involves not merely a relation between the person and the memo but also a relation between communities of practice: those where the memo originated and those to which the person belongs (p. 108).

He points out that this means the problem of communication is one of both reification and participation, and therefore designing artefacts also means designing boundary objects. Designing a boundary object (which is reification in the form of an artefact) means designing for participation, thus moving the emphasis from the harder aspects to the softer.

Common Ground

When we were exploring the notion of soft knowledge earlier in this chapter we observed that there are some of the softer aspects of knowledge to be found in Common Ground. Wenger's (1998) emphasis on the duality adds weight to the view of knowledge as a soft/hard duality and enables us to see that there are both soft and hard aspects in Common Ground. The softer aspects are in the knowledge that the members of a community have built up and with which they can understand each other. The harder aspects of knowledge are in the utterance itself. The softer aspects allow the members to understand the correct meaning. The conch shell in the example given earlier shows that an artefact can provide a shared basis for Common Ground and the Common Ground can determine the interpretation of an artefact within a community.