commonalities underlying tacit knowledge/knowing and explicit knowledge processes (Gourlay, 2004a; for a useful review of semiotics in this broad context, see Whitson, 1997).
one abstract concept is defined in terms of another equally abstract one. Since
“information” remains undefined a common-sense understanding is implicitly assumed which will not do: Stamper (1996) has shown that information has a variety of meanings, depending on the level of analysis.
Alvesson and K¨arreman (2001) concluded that knowledge is a loose, ambiguous, and rich concept that precludes reduction to simple sets of distinctions, a view Blackler endorsed (Blackler, 2002, p. 54). These views echo a conclusion reached half a century earlier. After much discussion, Dewey and Bentley concluded that knowledge is one of those “ ‘vague words’ one is at times compelled to use,” and “a ‘loose name’ ” because it has been used to refer to a great many often different things (Dewey &
Bentley, 1949, pp. 48, 78; see Ratner & Altman, 1964, for the discussion).
The situation has deteriorated since they wrote. In 1945 Bentley could note that whenever “knowledge” was used “living organisms are involved also” (Ratner & Altman, 1964, p. 459) but now “knowledge” is said to be embedded/embodied in “technology” (e.g., Teece, 2001, pp. 126–30; Argote and Darr, 2000, p. 53; Herschbach, 1995, pp. 31–2), in documents, repositories, organizational routines, practices and norms, (Davenport & Prusak, 1998, quoted in Alvesson & K¨arreman, 2001, pp. 998–9), and in the “physical structure of the workplace” (Argote & Ingram, 2000, p. 152). “Knowledge” is indeed a “tricky” concept (Tsoukas & Vladimirou, 2001, p. 975)!
An important distinction in the literature is between authors who treat knowledge as an object, and those who treat it as a process (Kakihara &
Sørensen, 2002; Sahdra & Thagard, 2003). Rarely, we also find authors who consider both aspects are important to its understanding (e.g., Cook & Brown, 1999). Viewing knowledge as an object is the dominant approach in management studies, and more widely, deriving as it does from the information processing paradigm (Kakihara & Sørensen, 2002, pp. 50–51, 53). We have already seen that Garavelli and his colleagues talk of knowledge being
“materialized into knowledge objects” (Garavelli et al. 2002, p. 270). Blackler (2002, pp. 48–54) described this as the “traditional” approach wherein knowledge is regarded as an entity in people’s minds, something they have, an idea neatly captured the phrase the “epistemology of possession”
(Cook & Brown, 1999, p. 382). This treats knowledge as abstract, as
“about” the world, and as something used in activities (Cook & Brown, 1999, pp. 382, 387–388). The similarity between knowledge in this sense, and explicit knowledge, is obvious, although Nonaka and his colleagues sought to distinguish them (Gourlay & Nurse, 2005). Kakihara and Sørensen called this a “representationistic” perspective since knowledge is assumed to be symbolic representations of reality held in various places (2002, p. 50). As Clancey noted, reviewing a parallel debate in cognitive science, “knowledge”
has been equated with representations and in turn with collections of symbols, such as words and word networks, and thus with stored descriptions (1997a, pp. 250–253; see also Bechtel, 1998).
The processual perspective is less easy to characterize, partly because it is newer, and partly because it includes several different ideas which to group together risks over-simplifying the picture. Kakihara and Sørensen suggested there are three “anti-representationistic” perspectives: knowledge as interpretation, knowledge as process, and knowledge as relationship (2002, pp. 51–4). It seems that these share much in common, and with Cook and Brown’s action oriented “epistemology of practice” (1999, p. 382) and Blackler’s emphasis on “knowing” linked with situated practices as distinct from “knowledge” (2002, pp. 51, 54–56). Gherardi and Nicolini (2000, pp. 330–
33) described a “social-material constructionist approach” to knowledge that is also clearly processual in emphasis. These all emphasize knowledge as an inseparable aspect of on-going situated activity, as distinct from something abstracted from and set apart from activity. Clancey’s (1997a, p. 254) metaphor of knowledge as “dynamically developed coordination processes,”
echoes both situated and dynamic systems perspectives. He also noted that organisms have been conceptualized as systems with a capacity to know (Clancey, 1997b p. 251) as Tsoukas & Vladimirou (2001, pp. 967–983) also proposed.
Of course, suggesting knowledge is, or can be regarded as, an object, (or a process), still begs the question: what is it that is an object or a process? Indeed, distinguishing object and process appears to achieve little more than reproduce the explicit/tacit distinction. This is obvious in so far as explicit knowledge entails representations and is often held to be objective (as well as an object), and tacit knowledge is often rendered as know-how. and largely concerns process (see, e.g., Sahdra & Thagard, 2003). The problem is not a new one, as Dewey pointed out in 1916, and again in the 1920s (Dewey, [1916], 1930). One solution, which Dewey favored at that time, is to reserve the word “knowledge” for the products of reflective, intellectual, processes. Holzner (1972, p. 9) ventured a complementary definition, writing of knowledge as: “the communicable mapping of some aspect of experience by an observer in terms of a symbolic system and frame of reference deemed relevant and appropriate.” As has been suggested above, we can avoid the usual assumption that “knowledge” has somehow been incorporated into the writing or the mapping by recognizing that documents are one of a variety of semiotic objects people use to orient their actions. As Holzner’s formulation indicates, the user of a “communicable mapping” must be capable of using the same symbolic system, and possess the same frame of reference, as the writer if they are to construct meanings similar to those of the writer, and thus orient their actions along similar lines.
This discussion is already blurring the distinction between “knowledge”
and “meaning” or “understanding,” perhaps inevitable as it makes
“knowledge” (if we wish to retain the word) a mentalistic concept—knowledge is only in the mind, and involves processes like understanding, learning, and so on (Wilson, 2002). Connectionists might accept this since the interconnections in the networks constitute knowledge, or perhaps dynamic representations
(Clark, 1997) in their models. Dynamic systems theorists would disagree for although they too agree that knowledge is emergent they see the organism- in-environment as the relevant system, not the organism-bound-by-its-skin, thus resurrecting an important methodological issue first raised in the 1940s (Bentley, 1954a). In their concern for ongoing activity, connectionists and dynamicists, along with situationists, seem to have overlooked the role of abstracted formalized representations in knowing and knowledge transfer (Clark, 1997; Bereiter, 2002). If, on the other hand, we regard these as artifacts that mediate the generation or construction of “knowledge,” as is implied above (see also Whitson, 1997; Gherardi & Nicolini, 2000; Wilson, 2002) then we can incorporate knowledge as communicable mappings into these models.
This further implies that talk of focusing on knowledge as process (e.g., Blackler, 2002; Clancey, 1997a, b) rather than knowledge as object is really beside the point. Instead we should follow Cook and Brown’s suggestion, to regard the “possession” and “practice” epistemologies as mutually interacting and their interplay as a “generative” phenomenon (1999, p. 383). But we need to move beyond metaphors to explore ways of conceptualizing these in a single framework that also points toward how we might study these processes. The following framework attempts to do this.