4.5.1 Personal Knowledge as the Unit of Analysis

As discussed earlier, Robert Grant is critical of those who subscribe entirely to organizational knowledge and is one of the few who make a clear reference to the “unit of analysis,” arguing that this should be the “person” not the organization. In this model, the goal is not knowledge transfer, but knowledge integration although one could argue that this is merely splitting hairs. Note, however, that Grant does not appear to dispute the existence of organizational knowledge, only that it should not be the focus of knowledge activities at the expense of personal knowledge.

Although the KBVF has been interpreted as a theoretical position, Grant himself questions whether it adds up to a singular theory. Nonetheless, according to Grant, it is the knowledge constitutive of persons that is the valued asset, and this can only be leveraged by integration through teamwork. By (borderline) implication, Grant, somewhat ironically, reifies knowledge as the object of integration mechanisms: rules and directives, sequencing, routines, group problem-solving, and decision-making. There is also a strong implication of the commodification of knowledge: “(K)knowledge is the overwhelmingly important productive resource in terms of market value…” (2002 : 136). Yet, if one sets aside those particular references, Grant is also arguably casting knowledge as social action in his emphasis on (social) integration practices as key to managing knowledge.

4.5.2 Knowledge as Practice Contrasted with Knowledge as Possession

Whereas Grant invokes the discourse of possession, Scott Cook and John Seely Brown of the Xerox Palo Alto Research Center in California, in their peculiarly named “generative dance” theory, take issue with this whole idea. They argue that such a discourse, with its implication of knowledge as something that people have, is limiting because it emphasizes one type of knowledge (explicit) above the other (tacit), resulting in a growing literature with a habit of explicitly or implicitly labeling knowledge as a typology of one. While there may be some truth in the latter statement, subsequent discussions reveal an opposite analysis to the notion of personal knowledge emphasizing the explicit. While Cook and his coworker’s theory is interpreted as approaching “knowing” as social action, their focus is more at organizational level in contrast to Grant’s approach.

In opposing an epistemology of possession, Cook and his colleague argue for an approach that views knowledge as a tool of knowing, with knowing connected to how we interact with the world. In their thesis, it is the interaction between knowing and knowledge that leads to the creation of new knowledge and “ways of knowing.” Thus, the term “generative dance” is coined and which is the source of all innovation. Their prescription is that knowledge is both used and grounded in action. Consequently, more attention should be paid to what knowledge work is done—in other words, practice. In effectively emphasizing knowledge accomplished in social interaction, and the action orientation of knowledge, are they that far apart from Grant’s emphasis on social integration practices? Not really. Brown and his colleague do arguably soften their case by proposing that epistemologies of possession and practice, while completely separate, are not incompatible. This immediately begs the question of why one would want to introduce a novel theory while clinging on to the traditionalist view. Nonetheless, the notion of knowing as action done in social contexts with knowledge as the requisite tool is generally consistent with a long line of theoretical development stretching back to the early 1990s and on to the present day. What they do not do is offer a perspective on how those “social contexts” come into being. At the most common-sense interpretative level, this idea implicates both the KS thematic categories of context and identity as a minimum.

There are two other points of interest to draw from the “generative dance” work: their criticism of the traditional scientific view of knowledge as something that must be sought, found, and articulated—in other words, knowledge as object—and their strong disagreement with the proposal that tacit knowledge can be converted to explicit and vice versa. This is the conversion process that lies at the heart of the influential theory of the knowledge-creating firm. Yet despite their position on the conversion issue, Cook and his colleague group this theory along with a number of other works, describing them as provocative and insightful.

The “generative dance” framework can be compared and contrasted with that proposed by Gustavo Guzman, of Griffith University, Australia, whose theory is interpreted as focusing on personal knowledge. He investigates the nature of practical knowledge, finding two opposing positions on this question: one approaches practical knowledge as stored in people’s heads, while the other sees it as situated in practice. Similarly to Cook and his coworker’s treatment of the ideas of possession and practice Guzman reasons that the two positions are not mutually exclusive. His theory, which has the aim of clarifying the confusing diversity of opinion, underlines the synergies between the cognitive and practice dimensions of knowledge using a simple case to make his point: it is the users of rules (as one form of knowledge) who determine when and how to use those rules. Thus, he makes the connection between knowledge how and knowledge that. Perspectives on those two concepts and their relationship are returned to presently.

According to this view, knowledge is seen as a multidimensional concept that can be personal, situated, and socially constructed at the same time. In terms of our KS thematic categories, if knowledge is personal, situated, and socially constructed, then it is bound to context and implicates identity with the aspects of “interpretation” and “determining” invoking notions of trust and risk.

Perhaps a more important contribution of Guzman’s proposal is that he treats dimensions of practical knowledge as “fuzzy”—they are not clear-cut categories but rather form a continuum. His aim is to explicate the various “quadrants” of practical knowledge, linking these to learning strategies with the objective of providing organizational managers with the ability to select the most appropriate learning strategy to suit the particular practical knowledge formula. In other words by road-mapping the type of practical knowledge that is to be shared or transferred between persons, the most effective learning approach can be selected and applied. Guzman’s theory of practical knowledge has some weight and rational validity, but it is entirely based on a review of the literature and it is not clear if the theory has been empirically tested.

Both of these theories invoke the idea of “Communities of Practice,” an influential notion that will be addressed in due course.

4.5.3 Connecting the “Organizational” with the “Personal”: Social, Situated, and Constructed

A number of theories that follow the notion of “knowledge as social action” and which are categorized as having a joint focus on both personal and organizational knowledge include those by Spender (1996), Tsoukas and Vladimirou, Boisot, Jakubik, and Crane and Bontis. Max Boisot’s “I-Space” theory of knowledge creation, similarly to Guzman’s project, emphasizes the importance of social learning as the foundation of knowledge creation, claiming that people do not share knowledge, rather they share information, which becomes knowledge once internalized to the individual. In Boisot’s model, knowledge is highly personal and relies on shared repertoires between individuals to reach common understandings (thus broadening the focus to the “organization”), which thesis has resonance with J.R. Searle’s influential notion of shared rules in speech acts, published in the late 1960s. The model can therefore be interpreted as having a joint focus both on the personal and organizational aspects of knowledge.

Although using different terminology, this theory is also consistent with the earlier account of organizational knowledge proposed by Tsoukas and his colleague, for instance. A divergence is Boisot’s leaning toward a cognitive understanding of knowledge with his emphasis on the cognitive effort in the form of abstraction and codification that is required in putting one’s knowledge into words such that it may be shared with others. This raises some interesting questions around people’s short- and long-term memory abilities, cognitive demand, limits on attention, and so forth, although a discussion of these does not feature in the framework and perhaps deservedly so as these are substantial topics for scientific research stretching back more than 100 years in the field of psychology alone. That aside, Boisot’s model, perhaps more strongly than others, foregrounds the role of identity, trust and risk, and context in the idea of “shared repertoires” and “common understandings.”

In something of a departure, Maria Jakubik’s “transformative teleology framework,” where “teleology” relates to a phenomenon’s end or purpose, proposes that the individual and social are not separated into different levels of organizational life. Instead, and with echoes of Guzman, knowledge is seen as embedded in human action, interactions, and situated practice. In her investigation of the relationship between knowledge creation and learning, she draws on the social view of learning defined by Etienne Wenger: “…learning is not located in individual heads, but in the processes of co-participation and in experiences. Learning can be seen as a social act, as a process of practice” (2011: 384). Wenger, an independent consultant, researcher, and author and who is credited along with Jean Lave with introducing the notion of “Communities of Practice” during the 1990s, claims that identity is a key structuring element in how we know because knowing is an act of belonging, referring here to Communities of Practice. In his account of these Communities as Social Learning Systems, Wenger also arguably invokes the importance of context. Two of the KS categories are therefore implicated here. What Wenger and Jakubik do not do, in the same vein as that noted in reference to Cook and Brown (1999) earlier and as found in most if not all KM theories, is explain how these social contexts, identity, and so forth come into being.

A fascinating account of the structure of knowledge is proposed by H. Collins. However as this does not really address KM per se, although it is included in the taxonomy because of its perceived importance, the commentary on this is restricted to the following. He posits two types of knowledge-relevant human action: regular acts that are normally associated with rule-following or rule-establishing and that can largely function at the unconscious level. Secondly, there are behavior-specific acts that are not really “acts” but more understood as behaviors, which can be articulated and therefore explicitly learned. One way of interpreting this is to apply the tacit/explicit labels, but this is perhaps to dilute the insight that Collins offers—the action-orientated, constructive nature of knowledge. And the implication of both context and identity as influencing factors.

What is beginning to emerge is the idea of knowledge being both inherently socially enacted, and as singular to the individual: it represents each person’s unique ability to de-code, interpret and make sense of their world without which socially enacted knowledge action could not and would not exist. Moreover, to a larger or lesser extent, many of the theories discussed so far can be interpreted as invoking categories of context, identity, trust, and risk (implicated earlier in Chapter 3 as factors in KS). Next, we review a number of KM theories located on the “knowledge as social action” axis, most of which focus on organizational knowledge: Activity Theory, the phenomenological perspective, “knowing how” and “knowing that,” Communities of Practice, and Creative Abrasion.

4.5.4 Knowledge, Activity Theory, and Activity Systems

Frank Blackler is responsible for some of the most insightful work in the field of KM—and some of the least recognized or understood. Blackler’s Knowledge and the Theory of Organizations is based on a modified version of Activity Theory, and like several others (e.g., Boisot), he emphasizes the central role of social learning. Activity Theory was originally proposed by the psychologist L.S. Vygotsky, working in the aftermath of the Russian Revolution. This constitutes people’s material actions and communications processes as the focus of study of human activity. These ideas locate the context of actions within “activity systems.” As Blackler caustically summarizes, “…the theory of organizations as activity systems offers an antidote to simplistic interpretations of the nature of individual knowledge and action, and organizational cultures and competencies” (1993 : 882).

In proposing that people construct themselves in action as a historically evolving process, Blackler is advocating a constructivist, action-orientated approach to knowledge. As such, he may be one of the first in this domain to refer to the idea of “identity construction.” He is critical of the mainstream rational-cognitive approaches to KM in their reification of knowledge and assumption of the rationality of both organization and individual. Interestingly, Philip Selznick, of the University of California, in his Foundations of the Theory of Organization, published in 1948, formulates the organization as a cooperative system of rational action, which is at risk from the “indiscipline” or irrationality of its participants!

As touched on in Section 1.4, Blackler draws on the work of H. Collins speculating a shift away from knowledge as situated in bodies and routines (embodied, embedded) toward knowledge as situated in brains (embrained), dialogue (encultured), and symbols (encoded). In other words, a move from knowledge as objective, tangible, and routinizable, to knowledge as constituted in social action. Thus the unit of analysis should be the socially distributed activity system within which knowing is done. He further emphasizes the importance of language in such systems, referring to it as the “archetypal communal activity” without which action would not be possible. From this perspective, the five-image framework of knowledge that Blackler had earlier seemed to espouse, he now sees as lacking. Instead, Blackler argues that the lens of research, practice and analysis should be on the activity systems in which knowledge is socially done: knowledge as mediated (e.g., through language), situated (in context), provisional (constantly developing) and contested (the subject of power relations). A clear reference can be drawn from Blackler’s theory to the KS categories of context and identity, with the implication of trust and risk. Moreover, the connection between these “actions” and language is made explicit.

Blackler is an early advocate of what could be described as a postmodernist approach to KM. (A core tenet of postmodernism is that knowledge is constructed rather than discovered through, for instance, observation, and is the means of power). Blackler positions his theory as a critical reaction to mainstream conventions (the “rational-cognitive” approach), although unfortunately his work has been largely overlooked in the field. Nonetheless, the view of knowledge as socially actioned has many proponents as shown in the taxonomy, and many of which emphasize the pivotal role of communications: these are, for instance, landmarked in the late 1980s by Gaines, the late 1990s by Drucker and in the more recent work of theorists including Tsoukas, which is discussed next.

4.5.5 The Phenomenological Perspective

Haridimos Tsoukas’ Phenomenological View of Tacit Knowledge quite clearly has a focus on personal knowledge, with its ideas grounded in those theorized by Michael Polanyi. As discussed in Section 1.4, far from theorizing about KM, Polanyi’s 1962 work, Personal Knowledge: Towards a Post-Critical Philosophy, is in fact an argument for turning away from the traditional view and practice of the exact sciences and its pursuit of objective knowledge and scientific detachment. Instead, he posits the importance of the scientist—the “knower”—in the act of discovery and validation of scientific knowledge. Drawing on Polanyi’s reasoning, Tsoukas criticizes the contemporary movement toward the decontextualization of knowledge. In his phenomenological approach to tacit knowledge he argues that explicit knowledge cannot exist without the tacit. This personal co-efficient factor suggests that all knowing—which implies the actual and potential to know residing in the individual’s head—is rooted to some context and action. It is important to note here that Tsoukas’ theory is centered on personal knowledge as an action-orientated phenomenon as opposed to a global understanding of knowledge (e.g., written knowledge). Resonance with many earlier works is clear. In this, context emerges as essential to knowing action.

Tsoukas is a particular critic of Nonaka’s theory of the knowledge-creating firm, devoting a substantial part of his paper to its unraveling. In his criticism, Tsoukas is not alone.

4.5.6 “Know(ing) How” and “Know(ing) That” and Communities of Practice

In his conceptualization of knowledge creation (and scathing criticism of “Nonaka’s theory”), Stephen Gourlay stands aside from the mainstream’s general acceptance of the tacit and explicit components of knowledge. Instead, he offers an explanation based on the concepts of “know how” and “know that.” This sounds familiar: recall, for instance, Paul Duguid’s use of similar terms, although there are some differences notably in Duguid’s preference for “knowing how/that” (see Section 1.4) and Guzman’s theory of practical knowledge. In Gourlay’s thesis, “know how” is rooted in and characteristic of everyday life, some of which is not capable of articulation, whereas “know that” emerges from the processes of reflecting and theorizing. A further distinction is that Gourlay views both of these concepts as behaviors. He reasons that in order to manage knowledge one must do so indirectly by managing behavior. This, he warns, raises a new panoply of issues and the potential to destroy that which management seeks to control (see Section 2.6 for a discussion around ethics and KM). Again, an interpretive view implicates the four KS thematic categories of context, identity, trust, and risk: that is, “know how” rooted in everyday life (context), “know that” as an outcome of reflection, and theorizing (identity), which collectively with the notion of behavior gives rise to the potential for trust and risk.

Like Blackler, Gourlay’s reasoning is consistent with postmodernist thinking and has particular relevance to the discussions around Implicit Learning in Chapter 7. Note also the synergies with the ideas proposed by, for instance, Tsoukas, through the emphasis on knowing as behavior, connecting knowledge—knowing something—to action. That said, his conceptualization of “know how” and “know that” is no doubt the source of some confusion when the attempt is made to compare this with Paul Duguid’s conceptualization, for instance. It is to Duguid and his colleague’s work, arguably more influential than that of Gourlay, that we now turn.

Communities of Practice (CoPs) lie at the heart of John Seely Brown and Paul Duguid’s architecture for organizational knowledge. In this, knowledge is mostly collective with successful CoPs being informal in nature, resonant of Etienne Wenger’s claim for the primacy of the informal over the formal. Ironically their architecture is largely dominated by themes of command and control. For instance, they emphasize the “organizing” work of firms, which contrasts with Robert Grant’s view that the organization exists to foster and coordinate an environment in which individuals can integrate their personal knowledge with that of others. Brown and Duguid view the organization as the means of knowledge generation and see organizational knowledge as more important than personal knowledge—the sum is greater than its parts, which once again recalls Vincent Barabba and colleagues’ systemic view of the organization. Their Architecture for Organizational Knowledge is largely about interfirm communications and the establishment of human, technological, and process-based conduits through which a team’s knowledge can be shared among other teams. For example, they propose that “translators,” people who are attached to more than one team group, are important transports for knowledge between groups in a kind of cross-fertilization process.

Originally introduced in the early 1990s, Communities of Practice theory has proved immensely influential but frequently misused according to Duguid writing in the mid-2000s. Emphasizing the inherently social substance of CoP theory, he conjures the CoP as the stage on which the shared “knowing how” of the community is on display, not just that of its individual members. His explication of CoP theory is largely concerned with dispelling the “myth” that tacit knowledge can or should be converted to the explicit. Equating “knowing how” to tacit knowledge as the “art of the knowing,” which he importantly constructs as the action of “learning to be,” and “knowing that” to explicit knowledge as the action of “learning about,” he reasons that to attempt to transform the former into the latter will likely lead to a transformation of learning to be into learning about. The profound importance of this lies in the simple analogy: “learning to be” is the art of knowing; of becoming a member of a Community of Practice, for instance; or of assimilating an art from a mentor over time. “Learning about” is equivalent to reading from books. To use Duguid’s reasoning, one does not become an accountant by reading an accountancy textbook. One must learn from a master.

Duguid’s ideas have considerable significance for the ideas developed here and in subsequent parts of the book. To complete this section, we make a brief visit to the intriguingly named Theory of Creative Abrasion.

4.5.7 Creative Abrasion

Dorothy Leonard and her colleague, writing in 2002 , are also interested in people sharing knowledge in group work. Their Theory of Creative Abrasion frames the different backgrounds, skills, experiences, and understood social norms that individuals have as the factors that generate the melting pot of innovation. In this implied chaotic environment, people will challenge each other leading to an abrasion of different ideas, which in turn give rise to new ones. Their explanation of this outcome is simple: people will search for novel solutions to problems when they are exposed to ideas and perspectives which “challenge” the majority’s prevailing wisdom. Moreover, it is the tacit dimension of an individual’s knowledge, learned through practice, not capable of articulation, which formulates them as valuable contributors to group work and innovation. From a pragmatist’s perspective, this thesis does rely on the majority group within any given organizational community being, in the first instance, open to novel concept: this cannot be simply assumed. But it does have an interesting connection to Chris Argyris’ notion of Double Loop Learning (DLL), which we will come to shortly.

As a basic idea, this is not entirely divorced from the concepts of social learning advocated elsewhere. It sits on the organization end of the organization versus personal knowledge continuum through its focus on the organization—teams of workers—as the unit of analysis. From their perspective, the individual is problematic in the tendency to hoard tacit knowledge and reluctance to share knowledge through fear of failure or of looking foolish. A further point to draw is their acknowledgment of the need for a cross-disciplinary approach to understanding knowledge and innovation: “(C)clearly, many different fields of inquiry are relevant, including ones as diverse as design, cognitive psychology, group dynamics and information technology” (Leonard and Sensiper, 2002: 495). It sounds complicated, but this theory does have echoes of Peter Drucker’s visionary piece on the future organization comprising teams of specialists dedicated to specific projects. It also has resonance with the KS categories of interest—context (teamwork), identity (team membership), trust, and risk.

An earlier piece of work by Leonard and coworker Susaan Straus takes a more general, higher-level viewpoint arguing that innovation, of necessity, requires a mixture of different styles of human thinking. Accordingly, innovation will not emerge among teams of like-minded individuals. Wenger (2000) makes a similar point. Innovation needs conflict. Consequently, organizations need the means of determining thinking styles and preferences, and for that, psychometric tests are recommended. While their reasoning is logical, the implication that individuals be assessed for thinking style and appointed based on this assessment inevitably raises some ethical concerns. Leonard and Strauss’ ideas can be interpreted as orienting to “knowledge as social action” through their emphasis on social interaction, with consequences for knowledge as an action category.

We now turn to the “knowledge as object” end of the continuum, noting how this end of the field is largely dominated by one theory.

4.6.1 “The Theory” and the Theorist as Bricoleur

The driving force behind the theory, according to Nonaka, writing in 1994 , is a reaction against the then dominant conventional conceptualization of the organization as a static, passive, “input–process–output” system. According to his radical (at the time) reasoning, building an action-orientated understanding of the organization means attending to two key aspects of the organization itself: its interaction with the environment in which it operates (and presumably those in which it would like to operate) and the ways and means by which it creates and communicates knowledge and information. These are ranked as the most important features of the organization. In effect, what Nonaka aims to do is to shift the emphasis from how the organization processes information and knowledge to how it creates it. That is surely a worthy ambition and appropriate to the emergence of the “knowledge society.”

A noticeable feature of the theory is the number of other theories and philosophical works that it draws on from across a wide range of disciplines (subsequent discussions address several of these): for instance, JR Searle’s “Speech Acts” (philosophy of language), John Anderson’s “Adaptive Control of Thought” architecture of cognition (cognitive psychology), Polanyi’s philosophy of knowledge, Chris Argyris’s “DLL” theory (organizational behavior), John Seely Brown and Paul Duguid’s (version of) “Communities of Practice” (knowledge management), A.E. Scheflen’s work in “Interaction Rhythms” (communications behavior), and introduced in 1998, the concept of “Ba” derived from the works of two Japanese philosophers. Now, there is nothing wrong in drawing on the works of others, and indeed the majority of academic writing does precisely this. But this particular treatment suggests something of the theorist as “bricoleur.”

At the heart of the theory is the SECI model that explains knowledge creation as the outcome of a dynamic interaction between subjectivities and objectivities. Accordingly, new knowledge is created in a spiral of interaction between the processes of socialization, externalization (explicit knowledge), combination, and internalization (tacit knowledge). Despite its emphasis on the socialization principle (indeed, Nonaka’s theory is one of the very few to make a direct reference to a theory of language, namely, JR Searle’s Speech Acts), the theory is positioned at the “knowledge as object” end of the axis because of its central tenet that tacit knowledge can and should be converted to explicit knowledge in the fashion proposed by the SECI model. It is for this “conversion” process that it is most heavily criticized. The product of this theory is a view of knowledge reified as an object to be managed and converted. According to David Snowden, Nonaka and his colleagues conceptualize knowledge creation as the output of a process that passes knowledge from the individual onto the group, and from the group to the organization, and thence to wherever else it needs to or can be transferred to. This, he claims, is quite obviously a reified view of knowledge as the object of management that at some point in the SECI process can be rendered explicit. A further question is this: in a theory that sets out to explain and predict organizational knowledge creation, is it not essentially wedded to “process”?

The inherent irony is that while Nonaka and his coworkers emphasize the importance of tacit knowledge, implying its possession of greater value than the explicit variety, the mandate to convert tacit to explicit knowledge signals a consequential devaluation of the former—it is only useful if transformed to the explicit. As Charles Despres and his colleague somewhat dramatically phrase it: “(T)the root idea is that knowledge becomes only useful when it goes into the forge of social interaction” (2002: 93). A dissenting voice to this interpretation is Snowden, who claims that the SECI model does not mandate that all knowledge in people’s heads and conversations should or could be made explicit. However in this, Snowden stands out from a majority of scholars critical of the SECI model. Note that in later versions of the theory (see, for instance, Nonaka and Toyama, 2007) an explicit reference is made to an “interaction” between explicit and tacit knowledge, but the conversion mandate remains at its core, thus arguably diluting the reference to “interaction.”

The idea that new knowledge can be generated in this dynamic interaction between two types of knowledge representing the subjective and the objective, and which process is fertilized and promoted in the socially engineered environment of “Ba,” has been the target of considerable criticism.

4.6.2 Shuffling Ideas

Critics include Charles Despres and his coworker who condemn theories including Nonaka’s for being too prescriptive and, with their ideas based around “transformations and dynamics,” imply these represent nothing more than the shuffling of ideas back and forth in structures and systems. According to Mark Thompson and his coworker, the very notion of knowledge transformation leads to a contradiction because “…the meaning of any objective ‘knowledge’ will always remain the subjective product of the person in whose mind this is constituted, always relationally defined, and therefore does not transfer easily to others in a form which may be operationalized to the benefit of the organization” (2004 : 726: italics in original).

In this way, they invoke and promote the importance of context, reasoning that attempts to “transfer” knowledge are unlikely to lead to organizational success. Support for this perspective can be seen in Ilkka Virtanen’s review of the literature, which reports that most Information and Communications Technologies (ICT) focused KM initiatives aim to convert tacit to explicit knowledge—and most fail.

As well as being criticized for its focus on the organization (at the expense of the individual), coupled with the reification and decontextualization of knowledge, the theory is censured for its ambiguity on the one hand and explicit distinction between knowledge and learning on the other. For instance, many would disagree with his proposal that learning is a static and limited concept (as noted earlier in Section 2.3.3) compared with the more dynamic and conceptually broader idea of knowledge creation. As discussed in Section 3.6, knowledge is elsewhere posited as a subset of learning.

Stephen Gourlay, perhaps one of the most outspoken critics of the theory, is certainly not alone in devoting much of his work on Conceptualizing Knowledge Creation to dissecting the theory, seeking to demonstrate the shifting sands and miscalculations on which is it is based. He specifically points to the “cracks in the engine” of the SECI model, arguing that this is more a theory of managerial decision-making than knowledge creation. On Nonaka’s evidence in support of the validity of the theory, Gourlay comments that this in reality represents nothing more than what managers believe to be the source of new ideas.

4.6.3 Misinterpretation and Misrepresentation

Perhaps most problematic is the suggestion that Nonaka and his colleagues have misinterpreted and misrepresented the works of Michael Polanyi in what Haridimos Tsoukas refers to as the “great misunderstanding” (see also Section 1.4 for additional discussions around this issue). As the core foundation of the theory draws on Polanyi’s work on the nature of knowledge this is a significant issue. The primary sticking point is over the nature of tacit knowledge. Polanyi proposes that all knowledge is personal, involves judgment, and that all knowledge contains a tacit element, which is often difficult if not impossible to articulate. According to Tsoukas, this establishes the contextual and action-orientated nature of knowing. However, Nonaka’s interpretation is slightly—but crucially—different: “Polanyi classified knowledge into two categories. ‘Explicit’ or codified knowledge refers to knowledge that is transmittable in formal, systematic language. On the other hand, ‘tacit’ knowledge has a personal quality which makes it hard to formalize and communicate” (Nonaka, 1994: 16, italics added).

In his analysis of Nonaka and his various coworkers’ interpretation of the works of Polanyi, Tsoukas is critical of their emphasis on reducing practical knowledge (viz., tacit knowledge) to “precisely definable content.” In other words, the aim of collecting knowledge located in people’s heads and “translating” this to explicit knowledge will achieve little more than reduce the value of that knowledge by converting what is known to that which is representable. This is precisely the argument that Paul Duguid and others imply in their view that ‘knowing that’ does not produce ‘knowing how’. So, while the theory draws on Polanyi’s ideas, the interpretation and representation of it—particularly the nature of tacit knowledge and its potential for conversion to explicit knowledge—is allegedly misrepresentative.

A related issue is highlighted by Kenneth Grant, following an earlier piece of work in which Grant and his coworker conclude that Polanyi’s theory concerning the tacit/explicit question has been largely misunderstood. Grant hypothesizes that many researchers in the KM field who refer to Polanyi have not actually read the source material, but are instead relying on others’ interpretation. This combined with a less than critical approach to the work of Nonaka and his colleagues, suggests Grant, is one of the underpinning reasons for KM’s lack of success.

As mentioned at the outset, the theory makes reference to the works of two other leading and influential theorists: Chris Argyris’s DLL in Organizations and John Anderson’s Architecture of Cognition (ACT theory). In both instances, it can be shown that Nonaka appears to misrepresent these works in how they are used to support his ideas. Argyris’ theory of DLL is essentially concerned with the need for senior managers to continuously challenge organizational policies, procedures, visions, objectives—as people’s internal theories of action—and so on in order to ensure progress and development. If not, firms will remain in a status quo and will eventually be overcome by others. According to Nonaka, firms find it difficult to implement DLL themselves, but, fortunately, it is “built in” to his theory. In reality, the only potential point of synergy between this theory and the theory of DLL is Nonaka’s parole that managers should challenge what employees know. However, Argyris is primarily talking about manager’s own internal theories of action based on organizational dictates, while Nonaka is talking about third parties challenging knowledge in other people’s heads—and how the latter interpret organizational dictates, for example. Arguably, then, DLL is not “built in” to the theory of the knowledge-creating firm.

The theory also draws on a 1983 version of Anderson’s ACT Model, however the later 1996 version published by Anderson suffices for the purpose here. In this, declarative memory (which Nonaka equates to explicit knowledge) is a schema like structure encoding a small bundle of knowledge, whereas procedural memory (equated to tacit knowledge) is applied automatically, is tied to context and can often not be articulated. Anderson proposes that complex cognitions are the result of interactions between declarative and procedural knowledge: “(A)all that there is to intelligence is the simple accrual and tuning of many small units of knowledge that in total produce complex cognition. The whole is no more than the sum of its parts, but it has a lot of parts” (1996: 356). He also claims that production rules that embody procedural knowledge can create declarative structures, but not the other way around. (This has echoes in Paul Duguid’s argument that knowing how cannot be derived from knowing that.) Note the use of the term “interactions” to describe the relationship between these two forms of knowledge/memory.

Interestingly, according to one interpretation of Anderson’s model, skill compilation leads to new skill acquisition, making knowledge a subset of learning (consistent with Krathwohl’s account of Bloom’s Taxonomy), whereas Nonaka’s theory constitutes learning as a subset of knowledge. Nonaka also clearly states that in Anderson’s theory declarative knowledge is transformed into procedural knowledge: “(T)the idea of ‘knowledge conversion’ may be traced from Anderson’s ACT model…” (1994: 18). This is not a strictly accurate account of Anderson’s own version of affairs in its introduction of the notion of “transformation” (viz., conversion), which is entirely absent from Anderson’s.

Note that both DLL and ACT are absent from the SECI model in versions subsequent to 1994.

4.6.4 More Troubling Observations on the Theory of the Knowledge-Creating Firm

Analogous with Brown and Duguid’s and Wenger’s conceptualization of Communities of Practice, the theory describes the “informal community” as the location of emerging knowledge. He goes on to suggest that these need to be integrated into the formal hierarchical structure of the organization. This implies a transformation of the informal to the formal. Would this not lead to a loss of the values and benefits afforded through informally organized work groups? Moreover, the scholar Charles Ehin, of Westminster College Utah, in his framework for “unmanaging” knowledge workers, claims that the formalization of communities of practice within a top-down hierarchical structure does not work. This mirrors a view previously promoted by Wenger.

Expanding on this theme of “communities,” a key claim made is that the theory is cross-cultural. This is difficult to substantiate particularly in the light of the nature of the theory. Recall that the theory is originally developed based on observations of Japanese firms. As is widely theorized elsewhere, knowledge is inextricably bound to culture (as context). Following H. Collins’ The Structure of Knowledge, culture undeniably varies from organization to organization, from community to community, from nation to nation, and so forth and so will the following of rules (see Section 3.5 for a discussion of this issue). The “one-size-fits-all” approach does not sit well with this conceptualization.

To complete this critical review, the theory’s stance on the temporal/cognitive nature of knowledge raises some interesting questions. The theory proposes that tacit knowledge refers to future events (with the inference that it only can refer to the future), while explicit knowledge deals with the past. Further, only tacit knowledge comprises cognitive elements. There is no evidence for this claim, and arguably, even if one takes the view that tacit knowledge is internalized, comprising skills, difficult to articulate, and the “more than we can tell” element of knowledge, why can tacit knowledge not refer to the past and present as well as the future? Additionally, if, as Polanyi argues, all knowledge contains a tacit element, then it is not logical to propose that explicit knowledge is bereft of cognitive components nor that it can only refer to events in the past.

With respect to the KS thematic categories of context, identity, trust, and risk, this theory arguably offers an ambiguous view: for instance, context from one direction is valued (e.g., social interaction, “Ba”), but then devalued from another (culture is irrelevant and can be “crossed,” the conversion model). Despite these difficulties, the theory continues to dominate the KM theoretical landscape with numerous other theorists following in its wake in one fashion or another, mainly through an uncritical acceptance and adoption of the tacit/explicit explanation at its heart.