Clark (1996) uses the notion of Common Ground and proposes it as being essential to communication. He describes Common Ground as "the sum of [two people's] mutual, common, or joint knowledge, beliefs, and suppositions" (p. 93). The example he gives is picking up a conch shell on a beach—he is aware that he is looking at the conch shell. He is then joined by his young son who also looks at the conch shell, and it becomes part of his son's awareness of his general situation, and therefore part of their Common Ground. The situation is the shared basis for their Common Ground. As they both look at the shell, they each have evidence that there is a conch shell between them, but they do not have evidence of the age of the shell. Clark calls this quality of evidence: excellent evidence that there is a conch shell, but poor evidence of its age; therefore, it is highly likely that the conch shell is part of their Common Ground but unlikely that its age is. He observes that people, in interaction with others, are constantly and unknowingly evaluating shared bases for Common Ground for their quality. The participants do not know they are doing this, they are applying their soft knowledge.

He takes the idea of Common Ground further and applies this to membership of cultural communities. For example, when meeting somebody new and getting to know him or her, one might find out that the other is a classical music enthusiast. If they are both enthusiasts they will find that they have a common stock of knowledge and language. There are different levels of cultural communities. Clark (1996) offers the example of a New Zealand ophthalmologist—this person would be a member of the communities of English speakers, of New Zealanders, and of ophthalmologists. What one would infer would depend on one's membership of these communities. One would expect that the person would tacitly know the use of English with its grammar and vocabulary. One would also assume that (s)he would know basic New Zealand geography and history, but one would only have a small part of that knowledge in common with the person. Likewise, one would expect the ophthalmologist to have detailed knowledge of eyes and their diseases. A non-ophthalmologist would only know the types of information the ophthalmologist would know. The ophthalmologist would know most of the particulars. Community members share a system of beliefs, practices, skills, know-how, and conventions and also have terms that have no meaning to an outsider. The ophthalmologists are an example of an occupational cultural community. Clark also gives examples of people with shared hobbies, religions, education, nationality, and employment.

A cultural community is more than a collection of people—they must have Common Ground. There are outsiders and insiders. Common Ground, according to Clark (1996), provides a means of differentiating—it shows what each knows about the community:

Inside information of a community is particular information that members of the community mutually assume is possessed by members of the community.

Outside information of a community is types of information that outsiders assume is inside information for that community (p. 101).

Common Ground is something that insiders of a community take for granted within each other—it includes "facts, beliefs, and assumptions about objects, norms of behaviour, conventions, procedures, skills, and even ineffable experiences" (Clark, 1996, p. 112), and it provides a means of defining a community.

In addition to common, shared knowledge, expertise, procedures, and beliefs, Common Ground can also take the form of a common language and a shared background. For example, the common language can take the form of language (national community), dialect (regional community), jargon, technical terminology, and shoptalk (occupational community). As an example of the shared background, Clark (1996) offers the example of a downhill skier—members of this community all know the experiences shared by each other, for example, the feeling of wind in the face, the different snow textures beneath the skis, and how to respond to the different surfaces. These are experiences that non-members cannot understand until they have themselves experienced them. Clark notes that more than facts (as in an encyclopaedia) are necessary, as facts cannot record or represent personal experience. He points out that community members do not just have shared knowledge but know-how, as in a community of accomplished pianists who take it for granted that each can play certain scales, play in certain modes, produce certain keyboard effects, and know what is and is not possible with regard to the piano.

Clark's (1996) focus was primarily on language, and it draws attention to soft knowledge in that he points out that the Common Ground is necessary in the interpretation of language—a phrase might have one meaning in one community and a different meaning in another community. The meaning depends on one's membership in the community. This provides a step forward in the exploration of soft knowledge, but a larger unit of analysis is needed to explore this in organisations. We need, therefore, to explore communities in the organisational context. In particular, we will focus on a particular type of organisational community that provides an environment for learning: the Community of Practice (Lave & Wenger, 1991).