Journal of Knowledge Management Practice, Vol. 7, No. 2, June 2006

Organisation Design And Tacit Knowledge Transferal: An Examination Of Three IT Firms

Peter Busch, Macquarie University

ABSTRACT:

Tacit knowledge is a contemporary area of research that is being explored for its ability to aid in developing a firm’s knowledge capital. At the same time little work has yet been undertaken on the structure of companies and other factors such as human networks or usage of information technology that may impact on how likely tacit knowledge can be transferred from one individual or group of workers to the next. An empirical study in three IT firms of varying sizes, type and structure has been conducted. The methodology included tacit knowledge testing in individuals, Formal Concept Analysis to interpret tacit knowledge results and Social Network Analysis to map communication flows between staff. Initial research indicates there is likely to be a positive correlation with organizations whose structure and work design promote face-to-face contact and have employees sharing close physical proximity, being more successful at tacit knowledge sharing.

Keywords: knowledge management, tacit knowledge, social networks, information technology, formal concept analysis


1.         Introduction


Much has been written in the last few years on the phenomenon of tacit knowledge.  Western post-industrialist society is coming to terms with the role knowledge plays in organisational wealth creation. Now that much secondary industry is taking place offshore, the advantages occidental firms must exploit are those relating to capitalizing on their employee’s knowledge assets. The intellectual capital that staff contribute to a firm help make the company competitive in a global arena.  Unfortunately little research has been conducted on firms with regard to tacit knowledge creation or diffusion. With respect to tacit knowledge specifically, empirical research if actually conducted, tends to be at the level of the individual.  On the other hand organisation theory with regard to tacit knowledge management is inclined to be principally descriptive. Minimal investigation has taken place with regard to either organizational design, or factors affecting intra-organisational knowledge flows with regard to tacit knowledge from an empirical standpoint. This paper begins by presenting a background to the tacit knowledge under study. Soft knowledge is typically diffused through human networks and so subsequent discussion explores the formality of networks and the strength of relationships between individuals. Measuring tacit knowledge flows requires testing for such knowledge in the first instance and understanding the parameters of knowledge flows in the second. After a brief theoretical discussion of organizational types, three companies are introduced in this paper. An analysis of the three firms with regard to their organizational category provides a basis for a discussion of the likelihood of tacit knowledge flows.

2.         Background To The Study

Evidence indicates high performance workplaces in the US and Australia are far more likely to benefit through the nurturing of intellectual capital, chief among which is the effective utilization of tacit knowledge (Curtain, 1998). Exploitation and transferal of tacit knowledge is important for the growth of a firm’s intellectual capital. Little research to date has explored the role of communication patterns at an intra-organizational level to understand the tacit knowledge flows in firms or the impact organization type has on these flows. Even the term tacit knowledge is a contentious one. Some argue that tacit knowledge is precisely that, once articulated it can no longer be considered tacit. Others acknowledge such claims, but take a more flexible interpretation of tacit knowledge whereby it could be said to construe knowledge not actually codified but potentially codifiable.

Grounded theory (Glaser and Strauss 1967) was conducted on sixty four papers where authors defined tacit knowledge. Grounded theories are induced from the data rather than preceding them (Cutcliffe, 2000; Partington, 2000). Through the examination of documents using Atlas™ qualitative software, a number of themes began to emerge. Polanyi (1967) defined tacit knowledge as non-verbalized, non-verbalizable, intuitive and unarticulated. In the course of grounded theory, a number of themes began to emerge (Dampney et al, 2002). It was discovered that tacit knowledge was a type of managerial knowledge, it was individualistic, it was contextual and at the same time usually organizationally or workplace based. Two major subsets of tacit knowledge were discovered, illustrating that it was either inarticulable, meaning a true form of tacit knowledge, whereas another subset could be codified over time. It was this latter articulable form of tacit knowledge that became the subject of the research. The outcome of conducting grounded theory on tacit knowledge literature led to a definition in this research of articulable implicit managerial information technology related know how, in other words an individual form of ‘workplace smarts’.

Having established the knowledge type under study, the second aim of the research was to examine soft knowledge flows. There are a number of factors that can either enhance or detract from the tacit knowledge diffusion process. These include the strength of relationships between staff; whether the establishment insists upon making use of information technology to communicate internally; the strength of human networks within the firm and also the structure of the company itself.

3.         Knowledge And Human Networks As A Means Of Knowledge Diffusion

Software may be appropriate for information transfer, but individuals are generally considered appropriate for ‘knowledge’ transfer (Jacob and Ebrahimpur, 2001).Much information management to date has meant the management of information technology (Anand et al, 1998), where large investments in IT were considered to be sufficient for knowledge transferal (Clarke and Rollo, 2001). What is now clear is that IT is not necessarily the savior of knowledge management, and typically not that of tacit knowledge management (Johannessen et al, 2001). Granted tacit knowledge is difficult to diffuse technologically (Haldin-Herrgard, 2000), some evidence would suggest that for knowledge to be transferred, it needs to be codified (Asheim and Dunford, 1997).  If we conclude that tacit knowledge is both socially embedded (Keane and Allison, 1999; Lado and Zhang, 1998), and contextually based (Busch et al, 2003), tacit knowledge is not likely to be effectively transferred. In other words, the marginal cost of transmitting tacit knowledge rises with distance (Audretsch, 1998), which explains at a macro level at least, the conglomeration of industry based on access to tacit knowledge (Dahlstrand, 1999; Keane and Allison, 1999). Silicon Valley provides a superlative example of this phenomenon (Shariq, 1999). It has also been noted that tacit knowledge is best transmitted through face-to-face interaction and repeated contact (Audretsch, 1998). Goods are called non-rival when they can be consumed by many agents simultaneously,  and the “non-rival” nature of tacit knowledge (Arrow, 1962) means that the benefits of this knowledge type are able to spread to other knowledge domains. This would tend to suggest why tacit knowledge possesses attributes corresponding to low environmental support for its distribution (Sternberg, 1995), as the value of the knowledge to one individual is also likely to benefit another. That is to say, because of the low environmental support, we keep such knowledge to ourselves. This last point is not what organisations necessarily desire, at least at an intra-organisational level, let alone the fact diffusion is difficult at an inter-organisational level (Rycroft and Kash, 1999).

Another important consideration is that knowledge exists as part of a holistic system, it “is not a thing in, and of, itself, but is rather a bundle or network of various elements: bodies, machines, communications technologies and materials of all sorts. …..   No one has ever observed a fact, a theory or a machine that could survive outside of the networks that gave birth to them. Still more fragile than termites, facts and machines can travel along extended galleries, but they cannot survive one minute in this famous and mythical ‘out-thereness' so vaunted by philosophers of science” (Latour, 1987; pp. 248). Networks form the vital infrastructure needed for knowledge and particularly tacit knowledge transfer, in the social network sense of the term they represent a distinct set of nodes or actors linked together by a set of edges or relations (Wasserman and Faust, 1994). Networks form social conduits (Ansell, 1997) with a person’s position in the network determining how effective knowledge transfer is likely to be. Traditionally one means of diffusing tacit knowledge in the workplace, reliant also upon one’s role in the network, has been through ‘war stories’, where employees discussed ways they achieved technical success. It is interesting to note management at XeroxÔ had tried to prevent this, and then decided it was for the best (Brown, 2000). An alternative approach is ‘storking’ or popping one’s head up from cubicle dividers (Leonard and Sensiper, 1998).

While some employees are unaware of their own tacit knowledge, colleagues will nevertheless seek out workers whom they feel will provide the expertise they require. The fundamental proposition of social capital theory is that network ties provide access to resources. One of the central themes in the literature is that social capital constitutes a valuable source of information benefits (i.e., "who you know" affects "what you know"). … information is important in providing a basis for action but is costly to gather. However, social relations, often established for other purposes, constitute information channels that reduce the amount of time and investment required to gather information (Nahapiet and Ghoshal, 1998). Other than counting the movements of highly skilled persons from one section of the organisation to the next (Stevens, 1996), we are forced to examine the networks of relationships between individuals in an organisation to determine knowledge and particularly tacit knowledge flows for evidence suggests this must necessarily be so (Krackhardt and Hansen, 1993).

4.         Relationship Formality In The Knowledge Transfer Process

Having established that the means of tacit knowledge diffusion are through networks, what roles do the attributes of the networks play in the diffusion process? First there is the importance and frequency of contact with other individuals in the network, which in turn encompasses the strength of the relationship.  At the same time, it has been noted that the frequency of communication between individuals is likely to be far less important than the type (meaning importance) of the person involved in the relationship (Lee, 1994: in Bennett and Gabriel, 1999). Furthermore, knowledge networks may be considered hard and soft as well as long and short.  Generally speaking softer networks or contacts between people tend to be shorter, generate softer facts, require fewer resources and leave room for negotiation (Latour, 1987). In other words within a community of practice where people know one another, and the work practices and contexts are familiar, hard networks become less common.  There is greater discretion between individuals in terms of the interpretation of knowledge, and in turn greater likelihood for the transferral of tacit knowledge. Indeed in certain professional communities of practice, such as the accounting discipline, a high degree of discretion has been noted (French, 2000), nevertheless professional bodies do typically act in a self-regulating manner. The transferal of harder ‘facts’ is noted to be implicit in ‘harder’ networks which utilise specific means of articulation to transfer information, for example the use of forms, metrics and so on (Latour, 1987).

The second major attribute relates to the type of meeting that takes place. If tacit knowledge is only able to be passed in personal settings, what type of meeting is more appropriate?  McAdam and McCreedy (1998) had conducted a survey of 97 UK firms showing that in actual fact a majority (81%) of tacit knowledge was elicited through informal meetings, whereas a substantially lesser amount had been acquired through formal discussion (44%).  Similarly the study had indicated that lessons learned on the job were largely passed onto colleagues through discussion groups (54%) rather than through presentations (7%). What of the type of interaction at the micro-organizational level? For example Daft and Lengel (1984) noted a five-step knowledge transfer richness continuum being comprised of: 1. face-to-face interaction; 2. the telephone; 3. written personal communication; 4. written formal communication; and lastly 5. numeric formal communication.  The face-to-face medium was considered the richest means for transferring knowledge, with the telephone less rich due to the absence of visual cues. Written forms of communication were considered less rich again, nevertheless personal forms of communication such as emails could be considered to exhibit more richness than formal communication which may be addressed to more than one individual (Koskinen and Vanharanta, 2002). At the time Daft and Lengel published their paper, the internet did not exist in the form we know of today. For example video conferencing with its inclusion of facial cues and body language would certainly permit the transfer of tacit knowledge. Nevertheless we should not ignore the fact a knowledge continuum exists depending on the type of technology utilized. Having examined the technology enabling knowledge transfer, what of the strength of ties between personnel and their impact on knowledge transmission?

5.         The Impact Of Strong And Weak Ties

In addition to links (relationships) between nodes (individuals) forming a means by which soft knowledge can be transferred, Granovetter (1973) had noticed the strength of ties between people influenced the likelihood of knowledge transfer. “Strong ties, identified by high-trust, lengthy timeframes and close relationships, are ideal for the sharing of tacit, complex knowledge. Weak ties, on the other hand, limit the exchange of knowledge and even information” (Fernie et al, 2003).

At the same time Hansen (1999) had found that weak interunit ties will impede the flow of tacit knowledge, but did aid in forcing units themselves to seek information from other units. Strong ties within a group often meant the group satisfied their demands for knowledge within that group, and provided the infrastructure for repeated contact necessary for soft knowledge transfer (Polanyi, 1967). It has also been observed that strong ties do increase the likelihood of the receiver of tacit knowledge possessing what Daft and Lengel (1984) label absorptive capacity. In other words “two actors that are strongly tied tend to have developed a relationship-specific heuristic for processing noncodified knowledge between them” (Hansen, 1999). Whereas codified knowledge is transferred easily through either strong or weak ties, evidence would seem to suggest tacit knowledge can truly only be passed through strong ties. The disadvantage in the latter case being that redundant knowledge is likely to be transferred as well (Hansen, 1999).

6.         Methodology

6.1.      Tacit Knowledge Testing

Much tacit knowledge literature, particularly within the field of knowledge management tends to be descriptive. If one is to conduct empirical research within organizations, there are few approaches that are viable. Arguably most experimental tacit knowledge research has arisen out of the Yale based psychology group under the directorship of Professor Robert Sternberg. Sternberg’s research is based upon two major approaches. One is that of the ‘critical incident technique’ and the other is that of the ‘simulation approach’. The former method involves interviewing personnel within the subject domain and eliciting information in relation to workplace tasks that were performed particularly well and those tasks performed poorly. At the same time this technique may also ask personnel whom they feel of their colleagues is proficient in the workplace and alternatively who of their colleagues are generally less productive or successful. Through interviewing both subsets of personnel, a distillation process with the information gathered can ultimately pinpoint both the techniques and personality types that make most successful use of their tacit knowledge. The latter approach, involves the observation of individuals undertaking tasks.  An example here might be an ‘in-basket test’ where employees are given a range of tasks to perform that appear in their ‘in-baskets’. The delegation of responsibility for certain tasks based on what is in their in-basket, is an example of employees making use of their workplace tacit knowledge.

Sternberg’s approach has evolved over time to incorporate a workplace-oriented means of assessment (Wagner and Sternberg, 1991a; 1991b). One typical scheme is along the following lines. “The measurement instruments used consist of a set of work - related situations, each with between 5 and 20 response items.  Each situation poses a problem for the participant to solve, and the participant indicates how he or she would solve the problem by rating the various response items” (Sternberg et al, 1995; pp. 918). Sternberg’s group acknowledges they are testing “practical know-how that rarely is expressed openly or taught directly”, in other words a form of ‘management knowledge’.

6.2.      Expert Identification

Because of the soft nature of tacit knowledge, there are few ways to establish face validity. Perhaps the least subjective means of doing so is to compare results from a tacit knowledge test with similar results from an expert sample. This necessitates the identification of an expert sample. One approach to identifying an expert group is to ask personnel within an organization to identify colleagues they feel are particularly proficient at what they do.

The tacit knowledge scenarios are given to all participants who firstly provide their scenario responses. At the same time personnel can be asked whom they consider to be experts. The scores of the identified experts are averaged and used as a basis. The closer the response of an individual to the expert group’s results, the greater the amount of tacit knowledge held by the individual. The approach taken by the Yale group has been to process their data using statistics. This method tends to require large sample populations and relatively lengthy periods of assessment time (e.g. 2 hours), which partly explains why much testing takes place on undergraduate students or other captive control groups, such as the military.

6.3.      Current Work

This work differs from Sternberg’s approach in a number of ways. First of all the empirical research has taken place in the IT domain. Secondly the study has been conducted on employees in Australian firms, not undergraduate student populations. Third and most importantly the working relationships between participants have been recorded to determine the likelihood of tacit knowledge flows. This approach means that the chosen sample populations are not random; nor are they necessarily large; captive; or anonymous. Because of such significant differences, alternative means of data interpretation to the statistical approaches used by Sternberg are necessary. One viable alternative approach to data interpretation is that of Formal Concept Analysis (FCA). Formal Concept Analysis, based on Wille’s work at Darmstadt (Ganter and Wille, 1999) as its name suggests is based on the philosophical notion of concepts. Each ‘concept’ is comprised of an intension and extension.  The former represents the attributes/properties or features of the object.  The latter concept relates to the set of objects, which  ‘meet’ or fulfill the attributes in the intension component (Fischer, 1993). Each concept thus forms a pair of mutually fitting sets (A,B), where A in this instance represents the set of objects, whilst B represents the set of attributes, corresponding to the former’s set of objects. The terms used in German, given the Darmstadt origins of FCA research, are those of: (G)egenstande for the objects,  (M)erkmalswerte for the attributes, and (I) representing the relationship between them.  We refer to the combination of (G,M,I) as a formal context. A binary K(ontext) (in German) may be formally expressed thus:

K := (G, M, I)

A multivalued context may be expressed as a quadtuple:

K := (G, M, W, I)  and I Í G ´ M ´ W

Where the relationship I is a subset of the combined components of Objects (G), Attributes (M) and merkmalsWerte (W) (Attribute-values). It is through the act of ‘conceptual scaling’ that the multivalued contexts are converted to single valued ones (Ganter and Wille, 1999). The act of converting a many valued context into a single valued one is in fact converting multivalued data into binary format.  Either a value exists at the intersection of a particular object and attribute, or it does not. The result of the binary ‘conversion’ process in turn produces a crosstable. The crosstable is then used to create a lattice which graphically represents the binary data. Readers are encouraged to turn to Busch and Richards (2004b) for further description of FCA.

After undertaking grounded theory on the tacit knowledge literature as well as conducting a dozen interviews with IT personnel, it was felt a number of basic ‘themes’ of tacit knowledge in the IT workplace could be handled with a test bank of 16 scenarios, each with between 6 to 13 solutions. For each one of the solutions there existed two seven-point Likert scales (Extremely Bad, Very Bad, Bad, Neither Good nor Bad, Good, Very Good, Extremely Good).  Two scales per scenario were presented because both an ethical and a realistic value was required as a means of working out how much variation there would be between what a person should (ethically) be doing, as opposed to what they would actually (realistically) do. It is worth noting a similar approach had been adopted by Wagner and Sternberg (1991a; 1991b). An example of a tacit knowledge scenario with a Likert scale answer arrangement may be seen in figure 1 below.

It is at this stage, that the role of FCA comes into play. Each of the respondents to the tacit knowledge inventory or questionnaire were treated as objects and their answers as attributes. A crosstable could be constructed and data interpreted from there. The advantage in using the FCA approach is that finer granularity in data interpretation was possible. The disadvantage was that the process of checking the results of each crosstable for respondent’s answers was laborious. Formal Concept Analysis thus acted as the questionnaire/inventory data analysis technique, instead of more traditional statistical approaches to data interpretation.

Figure 1:  Illustrating Scenario 3, Answer 2 Of The IS Articulable Tacit Knowledge Inventory

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 

6.4.      Measurement Of Flow

The second goal of the research was to examine information flows amongst staff members of firms. Arguably the most appropriate means of examining the nature and strengths of relationships between individuals is that of Social Network Analysis (SNA). Ties between individuals constitute a fundamental principle in SNA (Scott, 1991). Pivotal in Social Network Analysis has been the work of Granovetter (1973). Such analysis has a number of underpinnings. These include the assumption that relations among actors or people are considered as channels or thoroughfares of resources. Secondly, that the interaction among actors is directly constrained or aided by the structure of the relationships themselves. Furthermore, that relations taking place between the actors determine all economic, political and social structures (Wasserman and Faust, 1994). As a result of this we consider the presence of cliques (groups comprised of at least 3 people) or groupings of individuals to aid in tacit knowledge flow.

Returning to the concept of hardness/softness and strength/weaknesses of networks, our research instrument for testing tacit knowledge in individuals also included an SNA section, which asked for workplace relationship data. Questions included how often personnel saw each other [hourly; daily; weekly; monthly; bi-monthly], how important the individual was in relation to them self [have to see; very important; moderately important; can get by without seeing; try not to see], and the type of meeting they had with another individual [formal organisational meeting; informal organisational meeting; see one another outside of work; have lunch/morning/afternoon tea together; phone one another; fax one another; ‘bump into’ one another in the workplace].

A relationship between individuals where staff had chosen for example [hourly; have to see; formal organisational meeting] could be considered a strong or hard tie.  Similarly a combination of [bi-monthly; try not to see; bump into one another in the workplace] could be considered to comprise a weaker tie between personnel.  From a tacit knowledge transferal point of view, there is strong evidence (Jacob and Ebrahimpur, 2001) that tacit knowledge would more likely be transferred in a face-to-face setting.  Whereas a weak tie that is to say one with poor knowledge transfer richness (Daft and Lengel, 1984), could comprise an information link that included [phone; fax]. We must understand however that knowledge flows are complex and it would be a mistake to simply classify flows as only weak or strong, hard or soft.  People can communicate on a number of levels and through many mediums.

7.         Organisation Types

Understanding the impact of knowledge transfer in the workplace requires contemplation of organization types. Mintzberg (1991 a, b, c, d, e, 1983) had classified firms into five categories, being those of entrepreneurial, machine, diversified, professional and finally innovative. Lam (2000) extended this classification, introducing the J(apanese)-form company. The question remains what impact are organization types likely to have on tacit knowledge transferal?

The entrepreneurial firm (Mintzberg, 1991b) tends to have few staff, labor loosely divided amongst those staff with management structure in turn being flat and minimal. The CEO (Chief Executive Officer) is typically the driving force of such an organisation, and given the small staff numbers this is not surprising. To some degree the structure of such organisations are characterized by the simpler nature of the products they produce, permitting strong leadership at the top. In contrast firms producing more complex goods or services must necessarily rely on a larger number of staff that bring with them a wider range of skills necessary for the production of such goods. The entrepreneurial firm can be expected to make large-scale use of tacit knowledge as a means of transferring soft knowledge from one staff member to the next, given the smaller number of staff and the closeness with which the staff operate.

The machine organisation (Mintzberg, 1991c, 1983) is exemplified by routine work, typically simple in nature, which means in turn that workflow processes can become highly standardized. Communication in such organisations tends to be formalized, with a well-formed middle management structure enabling such communication flows. Machine organisations rely heavily on encoded, that is to say, explicit collective forms of knowledge. The role of tacit knowledge is minimized as knowledge is formalized quickly (Lam 2000).

The diversified or divisionalized company (Mintzberg, 1991d, 1983) represents a modern structure with subdivisions within the firm having a central administrative headquarters. Such a structure has often come about because organisations have grown over time, with the end result being that certain divisions ultimately begin to specialize in particular fields. It is possible that tacit knowledge transference may take place within the individual divisions, however the central headquarters structure of such an organisation would likely diminish the importance placed on tacit knowledge, in favor of structured articulated knowledge, at least in the case of western organisations (Nonaka et al, 1996).

The professional bureaucracy (Mintzberg, 1991a, 1983) carries out complex work by way of its (usually) tertiary educated specialists. The output of the firm is heavily based on knowledge work although the products and services do tend to be relatively standardized. Lam (2000) notes that professional bureaucracies derive much of their capability from embrained knowledge, that is to say individual explicit knowledge. Because professionals typically work independently, they are often reluctant to share their tacit knowledge, one more extreme such example of this problem being noted in law firms by Terrett (1998) who notes that lawyers, partly due to their competitive nature, are often reluctant to work with one another or in teams. The implicit feeling being that knowledge is power, particularly the tacit component. Consequently “the lack of a shared perspective and the formal demarcation of job boundaries inhibit the transfer of non-routine tacit knowledge in the day-to-day work. Moreover the power and status of ‘authorized experts’ inhibits interaction and the sharing of knowledge with ‘non – experts’ ” (Lam, 2000; pp. 495).

The innovative firm (Mintzberg, 1991e) or operating adhocracy (Mintzberg, 1983) is often characterized by high technology firms, or companies that make one-off specialised products.  Teams of people tend to form around particular projects.  From this point of view the innovative firm could be said to be the opposite of the professional bureaucracy. Because of the specialised work that teams of creative personnel within the organisation undertake, much tacit knowledge sharing takes place.  “The knowledge base of an operating adhocracy is diverse, varied and ‘organic’. A large part of the knowledge in use is organic, i.e. tacit knowledge generated through interaction, trial-and-error and experimentation. …. The frequent re-structuring and shifting of individuals between project teams means that tacit knowledge may not be fully and adequately articulated before an individual moves on” (Lam, 2000; pp. 497).

Finally the J-Form organization (Lam, 2000) combines the structure of a bureaucracy with the flexibility of an adhocracy.  As noted by Nonaka, many Japanese firms are efficient bureaucratic structures and permit individual teams a great deal of creative leeway in designing products. The strong corporate culture is important for providing a sense of company togetherness.

There does appear to be evidence to suggest that the above organisational types can have an impact on tacit knowledge transferal (Lam, 2000). Whereas the machine bureaucracy deals with controlled articulated forms of knowledge, meaning that tacit knowledge is minimized, the professional bureaucracy by its nature tends to promote boundaries of knowledge, and particularly tacit knowledge on the basis of its use by individuals. Operating adhocracies fair a little better in terms of tacit knowledge generation, but the changing nature of the organisation means such knowledge is not easily pooled. The J-Form firm is considered to be most conducive to tacit knowledge transferal.

8.         The Case Studies

Again most tacit knowledge research particularly within the knowledge management domain tends to be descriptive. The little research that is empirical is often purely psychological in nature, and very little has been published on tacit knowledge as it exists or flows in actual firms. To ground this research three IT firms have been studied, which shall be referred to as X, Y and Z. All three were based in Sydney, Australia.

8.1.      Organisation X

Organisation X with 108 participants was multicultural, with 53 of the 108 staff speaking at least one language other than English. The gender breakdown of our sample population was roughly 60:40 male:female. The proportions of each of the genders by age and ACS [see  Terrett (1998) - there are 5 levels, ranging from 1 (graduate) through to 5 (Chief Information Officer)]. levels were quite even.  In other words, the gender roles were not clearly delineated along clichéd lines of females as clerical and males as professionals. Admittedly there were slightly larger numbers of males proportionately in the senior ACS levels. Our staff were independently working IT professionals of predominantly middle management level. The staff were roughly 35 to 50 years of age. They had typically 10 to 15 years IT experience. They had not necessarily been with the present organisation for a long time, most less than 4 years. They tended to be tertiary educated, a very small proportion even held doctorates. They were largely permanent as opposed to contract staff. Around 30% of our sample was identified as being expert.  The experts appeared to have lower academic qualifications than the remaining respondents.

8.2.      Organisation Y

Organisation Y, an IT management consultancy, had 7 respondents participate in our study. The firm had a staff profile that was senior in terms of age (50-55 years) and IT experience (generally >20 years). The staff were largely Anglo-Celtic Australian, six were male, one was female, and all were tertiary educated. The structure of the firm was flat. Five out of the seven participants were identified by their peers as being experts.

8.3.      Organisation Z

Organisation Z with 16 respondents was substantially more multicultural than Y, with effectively 50% of the staff being non-Anglo-Celtic Australian. The experts tended to be managerial and/or ‘front office’ positions, conversely novices were in more technical roles. There appeared to be a strong positive correlation between being considered an expert and having a number of subordinates. At the same time there was only a weak positive correlation between IT experience and being considered an expert. There was a strong positive correlation with expertise and length of tenure. Experts had generally been present for more than 5 years; the opposite was true of novices. Experts were not particularly well qualified from a technical computing qualification point of view, however they were consistently better formally educated on the whole than the novices as a group.

9.         Organisations And Knowledge Types

Placing Companies X, Y and Z into Mintzberg’s organizational types, we can come up with the following table (Table 1). Organisation X as a whole was a diversified company, however the IT branch (our focus of study) operated as a combination of a machine bureaucracy and a professional bureaucracy. Some of the work conducted is routine such as the help desk role, however much of the work they do is specialized and involves creating new software for insurance applications. The staffing profile seems to match this. It is comprised of tertiary educated professionals with a large number at the ACS 3 level performing a variety of specialized IT tasks.  On a national level the IT group is spread out in effectively each state, but the core of the IT group nevertheless remains in Sydney. The IT group in Organisation Z is in fact very similar to the IT group in Organisation X, except on a much smaller scale. Organisation Y could be said to comprise either an operating adhocracy or a professional bureaucracy.

 
 

 

 

 

 

 

 

 

 


Such a classification difference depends on the type of work being undertaken by the firm. Generally speaking it is likely that this Organisation is a professional bureaucracy. The profile of the firm’s employees fits the tertiary educated professionals category, and whilst the knowledge work conducted caters to each client individually, is nevertheless relatively standardized.

10.       Communication Patterns

This paper is unable to provide an in-depth discussion on the application of Social Network Analysis or Formal Concept Analysis in this study, for that the reader is referred to our other works (Busch and Richards, 2004a; Busch and Richards, 2004b; Busch et al, 2003; Richards and Busch, 2003). Nevertheless after careful examination of the data through the use of these two techniques, a number of interesting points were revealed.

10.1.    Organisation X

 
An overall impression of the social networks in the Organisation can be seen in Figure 1. Squares  nodes represent male actors, circles females. Black represents permanent staff, red represent contractors. The letter(s) E or ENE as a prefix or suffix to a number represent experts. One can see there are three major groupings of personnel, with a few individuals acting as bridges between groups of actors, for example actor 2257 (lower right hand side). Edges or lines between the nodes or actors vary in thickness according to the amount of contact between staff. Thick red lines indicate hourly contact; the thinnest lines indicate bi-monthly contact.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 1: Organisation X - All Staff

 

Examining the extensive quantitative SNA data, there were noted to be many (54) cliques in our Organisation X sample. Some of these cliques (around 6) could be construed to be power cliques insofar as they were comprised almost exclusively of experts, but very few novices. In other words there was likely to be a great deal of productivity with regard to tacit knowledge flowing between personnel. On the other hand there were around 3 cliques that did not include any expert personnel. Thus these cliques could be said to be tacit knowledge poor, indeed a minority of staff appeared to have no direct access to tacit knowledge expertise. One could conclude the only way they would achieve a state of tacit knowledge enlightenment was through self-experience. On a positive note most novices appeared through examination using Social Network Analysis to have access to at least one expert.

This organisation employed a number of contract staff (25%) that were felt by their peers to be tacit knowledge savvy. There was some indication through SNA that were contractors to leave the organisation, tacit knowledge bottlenecking would likely occur, as contractors leaving the firm would take their soft knowledge with them. Whilst relations between colleagues in the organization were generally more than cordial, some bottlenecking could be seen to be occurring with certain work groups tending to avoid one another. If we couple this with a more or less widespread use of electronic mediums for data transfer then we conceivably have a situation where tacit knowledge transfer is limited within certain cliques and potentially uncompromisingly transferred within others. Note that Staff were not asked in this study if they communicated by way of teleconference. Rather the only electronic forms of communication asked for were telephone, email or fax. While recent evidence (Cavusgil et al, 2003) would suggest that firm size does not have any effect on the strength of relationships between individuals, it was interesting to note that in this, the largest of our firms, quite some degree of codification of knowledge and transfer through electronic mediums took place. Such usage of technology and codification of knowledge tends unfortunately to weaken the opportunity for tacit knowledge transfer (Jacob and Ebrahimpur, 2001; Haldin-Herrgard, 2000; Asheim and Dunford, 1997). We see some indication that interaction and mentoring patterns within the firm could be improved.

10.2.    Organisation Y

 
Figure 2 provides an illustration of the staff connectivity in Organisation Y. The small ‘cottage industry’ size of this firm permitted higher densities of communication patterns with fewer actors acting as communication bridges.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2: All staff – Organisation Y

 

The lack of any electronic means of information communication would likely greatly benefit the probability of tacit knowledge transfer. The fact that all staff met face to face regularly (at least daily, if not hourly) would indicate a high confidence in the tacit knowledge transfer process. The evidence that the staff appear to have no major areas of conflict in terms of not wishing to see colleagues would tend to indicate a fairly harmonious working relationship. Finally the presence of only one clique indicated all actors were interconnected without any intermediary.

In many ways the dynamics of Organisation Y have proved ideal with regard to the intra-organizational transfer of tacit knowledge. The ‘boutique’ size of the firm and the dense social interaction patterns have combined to form a harmonious working environment where all members are well interconnected and transfer their knowledge on a solidly face-to-face basis. The advantages of soft knowledge transfer in this firm are compounded by the general lack of electronic means of interaction indicating that wherever possible the employees see passing knowledge in a personal setting as preferable. The frequency with which meetings are conducted in this corporation indicates that any misunderstandings, which might occur because of a possible loss of tacit knowledge transfer, can easily be remedied given the almost hourly contact that takes place between the staff. A further advantage for the company is the physical layout of the offices and buildings (all on the same floor of the same building), which illustrates a firm that is able to make maximum use of its intra-organizational communication.

10.3.    Organisation Z

Figure 3 illustrates the small but highly interconnected nature of Organisation Z. Examining quantitative SNA data from this firm, a large number of cliques existed (13), which was surprising given the small size of the firm. It would also appear that network densities reflected a weak positive correlation with the tacitness quotient of the individuals. That is to say those people who we would expect to have higher readings of tacit knowledge (our experts) had lower network densities. Whilst experts may be connected to greater numbers of people, the strength of their ties tended to be weaker (e.g. they see people less often) because there are greater demands on their time. Concurrently, these people tended to be connected with greater numbers of individuals.

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


Figure 3: Organisation Z – All Staff

 

Because of the relatively diverse ethnic nature of Organisation Z, there did not appear to be any significant cliquing on the basis of language other than English. Organisation Z tended to be more hierarchical than Y, but less so than X. While phone conversations were limited, email was used extensively.

Organisation Z may not have had quite the staff diversity of Organisation X; however it would appear the CIO in the case of the former acted as the focus for tacit knowledge diffusion. The CIO in comparison was nowhere near as central in Organisation X. Firm Y on the other hand operated in a far less hierarchical way with knowledge pooled readily amongst the staff. It would appear that the potential for soft knowledge to flow is high in this firm given the fairly intimate size of its workgroup and their location within the same building. Factors, which may affect the soft knowledge flow for this firm, include for example the number of cliques. Unlike our small organisation (Y) with only one clique, this small-medium sized firm displays 13 cliques. There is some utilization of electronic means of communication, the extent of this is not significant; nevertheless it could have a negative bearing on tacit knowledge transfer.

11.       Discussion

Tacit knowledge can be gained through personal experience. Nevertheless should it need to be transferred from one individual to the next then this must necessarily take place through verbal and visual interaction. Whilst we cannot statistically generalize from a sample of just three firms, there does appear from our interpretivistic experience and from the literature, to be some evidence that the larger the organisation, the more a firm is likely to be inclined toward electronic forms of data transfer. There would also appear to be a strong positive correlation between organisational size and diminishing ease with which tacit knowledge is being transferred.

Although space limitations prevent a detailed study here of the Formal Concept Analysis and Social Network Analysis results, this research nevertheless indicates there is not a strong positive correlation between technically qualified or formally qualified personnel and their tacit knowledge richness. Nor for that matter is there a strong positive correlation between increasing age and tacit knowledge richness.

Based on these observations, from a tacit knowledge transfer effectiveness point of view the optimal firm would be that of Y, an operating adhocracy/professional bureaucracy (Mintzberg, 1991a; 1983). Factors that were helpful included the following; dense communication patterns insofar as daily meetings involve all staff, that is to say a network of strong ties. No staff were strongly avoided. Meetings in this firm were largely informal; to be precise the ties were soft. Staff occupied the same floor of the building. The organisation revolves around a management consultancy, which is naturally communication intensive. Staff were senior management types, a role that is strongly communication based. The firm is largely mono-cultural meaning cultural distances are minimal. The staff held each other largely in high esteem (given that the majority were selected as experts). In view of these factors, it is highly unlikely that tacit knowledge bottlenecking could be said to be taking place. It is highly probable that similar sized firms with analogous communication patterns would also be effective in transferring their tacit knowledge. With regard to tacit knowledge diffusion likelihood, the firm best placed for its transfer is clearly our small one (Y), but the next best would have to be our medium one (Z) followed by X.  There does appear to be some indication, that the larger the firm, the less likely tacit knowledge is to be successfully transferred from one individual to the next.

12.       Conclusion

It is acknowledged that this is a limited study with three Australian IT firms. Tacit knowledge is a difficult area of research as is organizational knowledge diffusion. There are many factors that can affect both of these variables. Nevertheless we can gain some understanding of tacit knowledge flows through tools such as Social Network Analysis. While face to face contact between employees will prove infinitely more conducive to tacit knowledge transferal over say faxes, there are a myriad of possibilities for employees to transfer their soft knowledge. One such example would include meeting outside of formal gatherings and engaging in conversation around the office or even in official breaks that occur throughout the day. The role of experts is also important in tacit knowledge transferal, for not only do experts contribute the knowledge they already hold, but being in pivotal roles they also bring in new knowledge to the firm through numerous other ongoing contacts. Organisation structure can also play a role in helping a firm maximize its internal knowledge capital. Of Mintzberg’s categories, it is likely that the entrepreneurial firm is apt to be most successful in promoting soft knowledge transfer; somewhat similarly the firm most successful in its tacit knowledge transferal in this paper was an operating adhocracy. By no means is this paper intended to serve as the penultimate statement on the topic of intra-organizational tacit knowledge diffusion, rather it should serve as a focus point for further debate.

11.       References

Anand, V.,  Manz, C.,   Glick, W.,  (1998)  “An organizational memory approach to information management”  Academy of Management: The Academy of Management Review  October  23(4)  :769-809.

Ansell, C.,  (1997)  “Symbolic networks: The realignment of the french working class 1887-1894”  The American Journal of Sociology  103(2)  September  :359(32)  (electronic).

Arrow, K.,  (1962)  “Economic welfare and the allocation of resources for invention”  in Nelson, R.,  (ed.)  The Rate and Direction of Inventive Activity  Princeton NJ Princeton University Press.

Asheim, B.,  Dunford, M.,  (1997)  “Regional futures”  Regional Studies  31(5)  July  :445(11).

Audretsch, D.,  (1998)  “Agglomeration and the location of innovative activity”  Oxford Review of Economic Activity  14(2)  Summer  :18(12)  (electronic).

Bennett, R.,  Gabriel, H.,  (1999)  “Organisational factors and knowledge management within large marketing departments: An empirical study”  Journal of Knowledge Management  3(3)  :212-225.

Brown, J.,  (2000)  “Growing up digital: The future impact of the World Wide Web”  Change  32(2)  March  :11.

Busch P.,  Richards, D.,  (2004a)  “Applying social network analysis to knowledge management”  In B. Trezzini, P. Lambe and S. Hawamdeh (eds.) , The Proceedings of the First International Conference on Knowledge Management (ICKM'04), 13-15 December 2004, Singapore, 383. ISBN 981-256-149-8.

Busch, P.,  Richards, D.,  (2004b)  “Modelling tacit knowledge via questionnaire data”   Proceedings of the Second International Conference on Formal Concept Analysis (ICFCA 04), February 23-26, 2004, Sydney, Australia.

Busch, P.,  Richards, D.,  Dampney, C.,  (2003)  “The graphical interpretation of plausible tacit knowledge flows” Australian Symposium on Information Visualisation (InVis.au)  Adelaide   South Australia  Feb 3rd and 4th (electronic).

Cavusgil, S., Calantone, R., Zhao Y.,  (2003)  “Tacit knowledge transfer and firm innovation capability”  Journal of Business & Industrial Marketing  18(1)  :6-21.

Clarke, T.,  Rollo, C.,  (2001)  “Corporate initiatives in knowledge management”  Education + Training 43(4/5)  :206-214.

Curtain, R.,  (1998)  “The workplace of the future: Insights from future scenarios and today’s high performance workplaces”  Australian Bulletin of Labour  24(4)  December  :279-294.

Cutcliffe, J.,  (2000)  “Methodological issues in grounded theory”  Journal of advanced nursing  31(6)  June  :1476-1484

Daft, R.,  Lengel, R.,  (1984)  “Information richness: A new approach to managerial behavior and organization design”  Research in Organizational Behavior  Vol  6  :191-233.

Dahlstrand, A.,  (1999)  “Technology based SME’s in the Goteborg region: Their origin and interaction with universities and large firms”  Regional Studies  33(4)  June  :379(1).

Dampney, C.,  Busch, P.,  Richards, D.,  (2002)  “The Meaning of Tacit Knowledge”  Australian Journal of Information Systems  December 2002  ISSN 1039-7841  pp. 3-13 

Fernie, S., Green, S., Weller, S., Newcombe, R.,  (2003)  “Knowledge sharing: context, confusion and controversy”  International Journal of Project Management  21(3)  April  :177-187.

Fischer, W.,  (1993)  "Formal concept analysis as a research tool in mathematics education"  Hiroshima journal of mathematics education  Vol 1  March  :1-35.

French, S.,  (2000)  Re-scaling the economic geography of knowledge and information: Constructing life assurance markets”  Geoforum  Vol 31  :101-119.

Ganter, R.,  Wille, R.,  (1999)  Formal concept analysis: Mathematical foundations  Springer-Verlag  Berlin  Germany.

Glaser, B., Strauss, A.,  (1967)  The Discovery of Grounded Theory: Strategies for Qualitative Research Aldine Chicago  U.S.A.

Granovetter, M.,  (1973)  “The strength of weak ties”  American Journal of Sociology  78(6)  :1360-1380.

Haldin-Herrgard, T.,  (2000)  “Difficulties in diffusion of tacit knowledge in organisations”  Journal of Intellectual Capital  1(4)  :357-365.

Hansen, M.,  (1999)  "The search-transfer problem: the role of weak ties in sharing knowledge across organizational subunits"  Administrative Science Quarterly  44(1) 82-111.

Jacob, M.,  Ebrahimpur, G.,  (2001)  “Experience vs. expertise: The role of implicit understandings of knowledge in determining the nature of knowledge transfer in two companies”  Journal of Intellectual Capital  2(1)  :74-78.

Johannessen, J.,  Olaisen, J.,  Olsen, B.,  (2001)  “Mismanagement of tacit knowledge: The importance of tacit knowledge, the danger of information technology and what to do about it”  International Journal of Information Management  Vol 21  :3-20.

Keane, J.,  Allison, J.,  (1999)  “The intersection of the learning region and local and regional economic development: Analysing the role of higher education”  Regional Studies  33(9)  December  :896.

Koskinen, K, Vanharanta, H.,  (2002)  “The role of tacit knowledge in innovation processes of small technology companies”  International Journal of Production Economics  80(1)  November  :57-64.

Krackhardt, D.,  Hanson, J.,  (1993)  “Informal networks: The company behind the chart”  Harvard Business Review  July-August  :104-111.

Lado, A.,  Zhang, M.,  (1998)  “Expert systems, knowledge management and utilisation, and sustained competitive advantage: A resource based model”  Journal of Management  24(4)  July-August  :489(2).

Lam, A.,  (2000)  “Tacit knowledge, organizational learning and societal institutions: An integrated framework”  Organization Studies  21(3)  :487-513.

Latour, B.,  (1987)  Science in Action: How to Follow Scientists and Engineers through Society  Open University Press  Milton Keynes.

Leonard, D., Sensiper, S.,  (1998)  "The role of tacit knowledge in group innovation"   California Management Review Berkeley  Spring  40(3) (electronic).

McAdam, R.,  McCready, S.,  (1998)  "The emerging trend of knowledge management within organisations: A critical assessment of both theory and practice"  1998 Annual Conference of the British Academy of Management  University of Nottingham (electronic).

Mintzberg, H.,  (1991a)  “The professional organisation”  in The strategy process: Concepts, contexts, cases  2nd. Ed.  Prentice Hall  Englewood Cliffs  New Jersey  U.S.A.  :704-717.

Mintzberg, H.,  (1991b)  “The entrepreneurial organisation”  in The strategy process: Concepts, contexts, cases  2nd. Ed.  Prentice Hall  Englewood Cliffs  New Jersey  U.S.A.  :604-613.

Mintzberg, H.,  (1991c)  “The machine organisation”  in The strategy process: Concepts, contexts, cases  2nd. Ed.  Prentice Hall  Englewood Cliffs  New Jersey  U.S.A.  :630-646.

Mintzberg, H.,  (1991d)  “The diversified organisation”  in The strategy process: Concepts, contexts, cases  2nd. Ed.  Prentice Hall  Englewood Cliffs  New Jersey  U.S.A.  :666-677.

Mintzberg, H.,  (1991e)  “The innovative organisation”  in The strategy process: Concepts, contexts, cases  2nd. Ed.  Prentice Hall  Englewood Cliffs  New Jersey  U.S.A.  :731-746.

Mintzberg, H., (1983) Structures in fives: Designing effective organisations  Prentice Hall Inc. Englewood Cliffs  New Jersey  U.S.A.

Nahapiet, J.,  Ghoshal, S.,  (1998)  “Social capital, intellectual capital, and the organisational advantage”  Academy of Management Review  23(2)  April  :242(25).

Nonaka, I.,  Takeuchi, H.,  Umemoto, K., (1996)  “A theory of organisational knowledge creation” International Journal of Technology Management  11(7/8)  :833-845.

Partington, D.,  (2000)  “Building grounded theories of management action”  British journal of management  11(2)  :91-102

Polanyi, M.,  (1967)  The Tacit Dimension  Routledge & Kegan Paul  London  U.K.

Richards, D.,  Busch, P.,  (2003)  “Acquiring and applying contextualised tacit knowledge” Journal of Information and Knowledge Management  2(2) June 2003  :179-190  ISSN 0219-6492.

Rycroft, R.,  Kash, D.,  (1999)  “Innovation policy for complex technologies”  Issues in Science and Technology  16(1)  Fall  :73.

Scott, J.,  (1991)  Social Network Analysis: A handbook  Sage Publications  London  U.K.

Shariq, S.,  (1999)  “How does knowledge transform as it is transferred?  Speculations on the possibility of a cognitive theory of knowledgescapes”  Journal of Knowledge Management  3(4)  :243-251.

Sternberg, R.,  (1995)  “Theory and management of tacit knowledge as a part of practical intelligence”  Zeitschrift für Psychologie  203(4)  :319-334.

Sternberg, R.,  Wagner, R.,  Williams, W.,  Horvath, J.,  (1995)  “Testing common sense”  American Psychologist  50(11)  November  :912-927.

Stevens, C.,  (1996)  “The knowledge-driven economy”  OECD Observer  June-July  n200  :6(5).

Terrett, A.,  (1998)  “Knowledge management and the law firm”  Journal of Knowledge Management  2(1)  September  :67-76.

Wagner, R.,  Sternberg, R.,  (1991a)  TKIM: The Common Sense Manager: Tacit Knowledge Inventory for Managers: Test Booklet  The Psychological Corporation  Harcourt Brace Jovanovich  San Antonio  U.S.A.

Wagner, R.,  Sternberg, R.,  (1991b)  TKIM: The Common Sense Manager: Tacit Knowledge Inventory for Managers: User Manual  The Psychological Corporation  Harcourt Brace Jovanovich  San Antonio  U.S.A.

Wasserman, S.,  Faust, K.,  (1994)  Social Network Analysis: Methods and Applications  Cambridge University Press  Cambridge U.K.


Contact the Author:

Dr. Peter Busch, Lecturer, Department of Computing, Macquarie University, North Ryde 2109, N.S.W., Australia.
Tel: +61 2 9850 9520; Fax: +61 2 9850 9551; Email: busch@ics.mq.edu.au