Literature Review

This section reviews some key themes in the literature concerned with the role of IT in knowledge management initiatives.

Technologies and knowledge management

Over the last decade and more a plethora of technologies have emerged that have been associated with knowledge management; and specifically with the articulation, storage, transfer, creation, and retrieval of knowledge. Though not exhaustive, Table 5.1 highlights the most notable IT artifacts and platforms associated with knowledge management projects (Alavi and Leidner, 1999). Drawing on Zack (1999a) I distinguish between two types of ITs that have been associated with knowledge management projects: integrative and interactive applications. Much of the academic literature and practitioner accounts have focused on integrative applications which take the form of structured databases that allow employees to store and retrieve information on past projects. They also comprise expert finders, electronic bulletin boards through to best practice reports and working papers (Butler and Murphy, 2007; Zack, 1999b; Alavi and Leidner, 2001; Chua, 2004). Interactive applications take the form of email, desk-top conferencing, and discussion forums allowing for interactions with other staff and the garnering of their views and experiences regardless of physical location (Moffett et al., 2003; Leidner, 2000; Alavi and Leidner, 1999).

Table 5.1 Key technologies associated with knowledge management

With regard to interactive applications, groupware and intranet platforms have dominated the academic and practitioner accounts to date (they are also the focus of much of this chapter). Most recently, interactive applications have come to include E2.0 platforms. E2.0 comprises a number of technologies that are typically associated with Web 2.0. The key difference being E2.0 technologies are put in place to support collaboration and knowledge working within the organization. Key technologies include wikis, blogs, social networking and instant messaging, the ability to link out to other pages, and the categorization of data by users through tagging (McAfee, 2006). Tags are furthered through extensions, which is a form of automated tagging through pattern matching algorithms (e.g. Amazon). Such technologies are attracting considerable current attention amongst practitioners and are claimed to offer new ways to document, distribute, and retrieve knowledge within organizations (McAfee, 2006). This is a theme that I will return to towards the end of the chapter.

Understanding knowledge and technology

Within the academic literature, two contrasting epistemological approaches have underpinned accounts of knowledge management: the content and relational perspectives (Scarbrough and Burrell, 1997; Tsoukas, 1996). This chapter adopts the latter perspective. From a content perspective, knowledge is defined as being a predicative truth as it prescribes what to do (Nonaka and Takeuchi, 1994; Galliers and Newell, 2000). Knowledge is viewed as being able to be codified and stored in repositories, so that knowledge can be shared, built upon and retained regardless of employee turnover (Wasko and Faraj, 2000). This perspective is dominant in the information systems literature (see Sambamurthy and Subramani, 2005); and much of this literature has a practitioner orientation and has focused on collecting, distributing, reusing, and measuring existing codified knowledge (Lam and Chua, 2005; Bansler and Havn, 2004; Cohen, 1998; Knock and McQueen, 1998). Knowledge is considered to be an economic asset that can be codified, stored, and exchanged between individuals within a firm (Bohm, 1994; Shin et al., 2001; Pan and Scarbrough, 1999; Currie and Kerrin, 2004).

Relational writers are critical of this dominant view of knowledge (Lave, 1988; Blackler et al., 1997), and suggest that instead of treating knowledge as being a largely cerebral and tradable entity, knowledge should be viewed as being relative, provisional, and primarily context-bound (Scarbrough, 2008; Barley, 1996; Orr, 1990; Blackler et al., 1993). Critiques of much of the knowledge management literature suggest that knowledge is always embedded and as such can only be shared actively through social groups. Rather than knowledge, relational writers argue that the focus of enquiry should be on the process of knowing and the capability to act (Schultze, 2000; Blackler, 1995; Brown and Duguid, 1998). Furthermore, many are critical of the content approach for its weak empirical base and prescriptive standpoint (Pan and Scarbrough, 1999). They suggest that exchanging knowledge as if it were an economic asset via IT does not relate to the actual experience of the use of knowledge management applications within specific contexts (Schultze, 2000; Galliers and Newell, 2000). Within the literature, this focus on sharing explicit knowledge through IT systems has led to the terms ‘information’ and ‘knowledge’ often being used interchangeably and uncritically by many authors (Sambamurthy and Subramani, 2005).

Technology-centric

One pervasive theme in the literature has been the dominance of techno-centric accounts, as is exemplified by Sambamurthy and Subramani (2005: 2) who suggest that knowledge management involves ‘developing searchable document repositories to support the digital capture, storage, retrieval, and distribution of an organization’s explicitly documented knowledge.’ Indeed, there has been a preoccupation with the problem of finding the location of knowledge (Sambamurthy and Subramani, 2005; Massey et al., 2002; Benbya, 2006; Gray and Durcikova, 2006). Such a position assumes knowledge can be codified and stored. Thompson and Walsham (2004) claim this has been at the expense of the more ‘contextual elements of knowing.’ Here they include a willingness to document and share information, an attention to detail, and the establishment of software tools and methodologies to support such changes. Similarly, based on his case studies of Technology Research Inc. and Buckman Laboratories, Zack (1999a) argues that the technology focus is a major obstacle to engendering an organizational climate that values and encourages co-operation, trust, and innovation (Butler and Murphy, 2007; Butler et al., 2007). Knox et al. (2008: 290) are also critical of such techno-centric accounts, arguing that managers know that ‘their claims to expertise rest on their ability to perform the reconnections between the systems outputs and other more contingent “upstream” practices, situations and possibilities’. Thus, they argue that those that adopt a techno-centric perspective do not understand or appreciate the ways and the efforts to which the enactments of some people allow for such a perspective to ‘work.’

Domain specific knowledge

From a relational perspective, knowledge is viewed as circulating easily when people work within a similar domain of practice, or have experience of working with other knowledge domains. This provides for a shared sense of what practice is and what the standards for judgment are (Brown and Duguid, 1998). Zack (1999a) argues that due to the shared understanding of practice, integrative applications that allow access to data stored centrally are sufficient to support work between employees from the same domain of work, though others still advocate more interactive applications (Boland and Tenkasi, 1995; McDermott, 1999). More recently Butler and Murphy (2007) argue that while technologies can assist knowledge creation and sharing within domains, they advocate the establishment of careful rules as to who can contribute posts to specific parts of the knowledge management system. This they claim ensures that only those who are expert are able to contribute, and reduces possibilities for misunderstandings between staff from different domains.

Orlikowski (2002) argued that the relative success that Kappa achieved with their knowledge management initiative was in part attributable to their ongoing reinforcement of a strong identity. This identity formation, Orlikowski (2002: 267) argued, allowed people from across the organization, regardless of their group or location ‘to internalize and identify with a common way of thinking about and engaging in their product development work. This facilitates the communication and coordination of hundreds of product developers across time (19 time zones) and space (15 geographic locations).’ Orlikowski (2002) goes on to explain that it also allowed people across the organization to share a common language. Importantly, such identity formation and the development of shared language were between staff from a similar knowledge domain, and consequently they shared aspects of their practice with each other.

Working between domains of knowledge

Sharing knowledge between employees with different professional backgrounds is viewed as more complex (Newell et al., 2000; Brown and Duguid, 1998; Ruhleder, 1995). Schultze and Boland’s (2000a) study of US Company, a manufacturer of building materials, found that the introduction of KnowMor, a knowledge management application that bridged the different functions and locations, led to changing temporal and spatial work arrangements. They provide an example of the Notes administrators being required to perform not only technical activities, as previously, but also to write accounts of their actions into shared databases. Work practice changed in that they would now point users to these entries rather than deal with requests themselves. Schultze and Boland (2000a) explain that the new work practices required them to perform a dual role of acting (in terms of the technical change) and accounting self (in terms of making explicit their technical work).

A further challenge relates to the limited understandings that can be developed when reliance is placed on IT to share knowledge between communities. Brown’s (1998) insightful study of the use of the Internet to support knowledge working found that a reliance on technology as a means of transferring knowledge is insufficient. Instead he contended that abstractions recorded and shared on the Internet need to be considered as being inseparable from their own historical and social locations of practice. McDermott (1999) provided a detailed analysis of why a reliance on abstractions is problematic. He cited the case of a diverse group of systems designers to illustrate what he termed the difficulties of thinking ‘outside an expert’s own territory.’ He ascertained that, rather than needing each other’s documentation stored on a common database, the system designers needed to understand the logic that other designers used in practice, such as the rationale behind the combination of specific software, hardware, and service plans. Similarly, Schultze’s (2000) case study of experts across US Company discovered that each group had their own conventions. She explained that the IT professionals emphasized the importance of documentation, while the competitive intelligence analysts emphasized secrecy and selective dissemination (Ruhleder, 1995; Hislop et al., 2000; Hayes, 2000). Currie and Kerrin’s (2004:25) case study highlighted how the antagonistic relations between the sales and marketing teams were exacerbated following the introduction of the intranet as ‘Sales and Marketing did not share the necessary language and context to effectively exchange knowledge.’

Thus, due to the difficulties of working between communities, many suggest that attention needs to be given to the nature of IT support. Butler and Murphy (2007) argue that knowledge management systems need to be designed so that they only allow people with relevant experience and insights to contribute to or review specific parts of them (Markus et al., 2002). When there is no shared history of working together, Zack (1999a) argues that integrative applications are unsuitable. Instead he recommends the use of interactive applications that support ad hoc collaboration (such as discussion databases and video conferencing).

Establishing social capital and social networks

Based on their study of an enterprise resource planning (ERP) implementation, Newell et al. (2004: 54) argue that a lack of social capital between organizational members from different knowledge domains made it impossible to ‘expose and explore the different thought worlds.’ Newell et al. (2004) thus argued that social capital was crucial for organizational knowledge to be accessed via an ERP system and integrated effectively (Huysman and Wulf, 2004). Similarly, Olivera’s (2000) study of a consulting company found that users placed great effort on interacting with people rather than relying on computer based systems such as the intranet. Others have suggested working between communities requires the creation of new facilitating roles. Zack (1999b) emphasizes the importance of appointing facilitators charged with encouraging, interpreting and evaluating participation in cross-community electronic forums. Storck and Hill (2000) found that Xerox unit managers undertook facilitating roles so as to encourage knowledge sharing, such as presentations describing their initiatives, questioning participants about what they had learned, and electronically circulating what they considered to be key developments. Facilitators in Xerox looked for mutual interests, sought to mask differences in hierarchical status, encourage commitment, and reduce uncertainty (Schultze and Boland, 2000a; Storck and Hill, 2000).

Individualism versus collectivism

Orlikowski’s (2002) case of Kappa found that staff staved off a lot of self interest by consciously making sure that they allowed all the different groups opportunities to engage in discussions, allocated work across functional and geographical boundaries, and reassigned staff to work across the world in different groups. This, Orlikowski (2002: 269) claims, provides for the ‘crossing of temporal, geographic, technical, and political boundaries, because it provides for the distribution, and then the integration, of ideas and experiences.’

Storck and Hill’s (2000) study of Xerox found that in addition to technology, the relaxing of centralized control had allowed participants to identify with, and be committed to, their community rather than to Xerox as a whole. They concluded that within communities, this engendered a climate that allowed for openness, trust, and commitment (Huysman, 2004; Pan and Scarbrough, 1999; Malhotra and Galletta, 2003). Newell et al.’s (2004) case study found that long established self interests meant the integration of dispersed organizational knowledge necessary for implementing the ERP system was never achieved. Overall, they argued that where people and teams have not worked together before considerable effort and resources need to be invested so as to encourage sociability and develop an understanding and familiarity with each other. They argue that such bonding is essential if people are to share knowledge for the public good as well as for their own self interests. Without doing this, Newell et al. (2004: 56) claim it will lead to ‘a mechanistic pooling of knowledge that will not produce the knowledge integration leading to creativity and innovation that are needed in large-scale IT projects.’ Orlikowski (2002) claims the emphasis that Kappa placed on the ongoing education of their employees with regard to their software development expertise, and career mentoring, encourages knowledge sharing (Wasko and Faraj, 2005; Zack, 1999a; Quinn et al., 1996).

In addition, several writers have suggested that reward structures can also encourage knowledge sharing. Orlikowski’s (1993) study of a US consulting company argued that both the competitive promotion and financial reward structure led to Notes remaining largely unused by consultants. However, she found that the technologists and senior consultants who were not subject to the competitive culture did use Notes more formatively (McDermott and O’Dell, 2001; Hislop, 2003; Pan and Scarbrough, 1999; Zack, 1999b). Currie and Kerrin’s (2004) case study of a global pharmaceutical firm who were in the process of downsizing also found that sales staff horded rather than shared knowledge with graduate trainees as they feared the trainees may become their replacements. Currie and Kerrin’s (2004: 25) case also highlighted that when there is already animosity between different groups in an organization ‘the intranet hardened existing cultural cleavages in the organization’ rather than manufactured a sense of community.’

Visibility and control

A final important but often neglected theme to emerge in the literature relates to the inseparability of knowledge management initiatives from issues of power and politics (Easterby-Smith et al., 2007). Hislop et al. (2000) found that in a UK nuclear medicine company external expertise and information were utilized by groups as a political resource to reinforce and support their particular visions for change. They also described how groups supporting the interests of senior managers received the authoritative support and financial resources to bring about change. Similarly, Newell et al.’s (2000) study found that the introduction of communication forums reinforced the ‘powerful centrifugal forces’ in their firm.

Making work and views explicit on IT is also implicated in issues of power and control. Schultze and Boland’s (2000b) study of US Company indicated how the IT consultants would visibly record their work in the databases as a means to protect themselves from any blame that may subsequently be leveled by their clients or other consultants. Schultze and Boland’s case further indicated how the visibility that KnowMor provided could unintentionally have a negative effect on relationships within US Company. For example, one analyst felt that his role had inadvertently been threatened due to the way another employee had harnessed the visibility KnowMor provided. KnowMor had allowed this latter employee to be aware of and widely circulate some details about a merger that the analyst should have been aware of. The analyst resented the implication that he was not doing his job well and responded to the issue by revoking the employee’s access to KnowMor. McDermott (1999) argued that, due to the pervasiveness of politics in knowledge working, one key challenge is not to homogenize views but to encourage diverse collaboration and an appreciation of the context specific insights of members from other communities. McKinley (2000) supports this view and warns against IT being used in knowledge management projects for surveillance and control activities. In a later study, McKinlay (2002: 81) argued that recording one’s knowledge was sometimes viewed by staff in this case as ‘being asked to give myself away.’

Thus, studies have pointed to the importance of attending to the link between power and making explicit a staff member’s knowledge, hierarchical power relations, surveillance possibilities, and a competitive culture (Knights, 2008; Knights et al., 1993). Indeed, Currie and Kerrin (2004) argue that due to the overtly technological orientation, coupled with the broader issues of culture and power, the possibilities of IT to mediate knowledge sharing are limited.