Jump to ContentJump to Main Navigation
Management ConsultancyBoundaries and Knowledge in Action$

Andrew Sturdy, Karen Handley, Timothy Clark, and Robin Fincham

Print publication date: 2009

Print ISBN-13: 9780199212644

Published to Oxford Scholarship Online: May 2009

DOI: 10.1093/acprof:oso/9780199212644.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 17 August 2018

Boundaries and knowledge flow

Boundaries and knowledge flow

(p.23) 2 Boundaries and knowledge flow
Management Consultancy

Andrew Sturdy (Contributor Webpage)

Karen Handley (Contributor Webpage)

Timothy Clark (Contributor Webpage)

Robin Fincham (Contributor Webpage)

Oxford University Press

Abstract and Keywords

Boundaries are shown to be a core element of social science in that they capture fundamental social processes of structuring and relationality. This is linked to a debate in the sociology of knowledge in the 1970s concerning the possibility and value of the knowledge of social ‘insiders and outsiders’. Following Merton, emphasis is placed on the idea of simultaneous and multiple insider — outsider statuses or identities. By drawing on recent organizational and learning literatures (e.g. Nooteboom, Wenger, Orlikowski, and Carlile), three related boundaries are then outlined — physical, cultural and political boundaries — and identified as necessary conditions for knowledge flow. Here, the importance for learning of cultural/cognitive distance between parties is introduced along with boundary contexts of particular importance to the study of consultancy — project working and liminality. The chapter concludes with a framework of boundary relations and dynamics which are drawn on in subsequent chapters.

Keywords:   boundaries, cognitive distance, liminality, insider knowledge, project based learning


Theoretically, the consultancy literature spans a range of perspectives, often mirroring those in the study of management idea adoption more generally (see Sturdy 2004 for a review). Given a rather prescriptive bias in much of the literature and/or a sense that consultancy is, in itself, a field of study, theoretical positions are often more or less implicit. Nevertheless, certain perspectives are clearly evident. In particular and as we have noted, much of the prescriptive literature reflects some of the assumptions of transaction cost economics. It also tends to adopt a view of knowledge as object‐like which is transferred, rather than translated (cf. Latour 1986). Less prescriptive approaches often draw on dominant perspectives in organization studies more generally, neo‐institutionalism and social embeddedness in particular (e.g. Armbrüster 2006; Kipping and Engwall 2002; Saint‐Martin 2000). Other perspectives are also evident, often in combination, such as psychodynamics, actor–network theory, dramaturgy, identity theory, and, to a lesser extent, labour process theory (Alvesson 2004; Bloomfield and Best 1992; Clark 1995; Jackall 1986; Sturdy 1997b).

Our own analysis of consultancy is informed by a range of broadly sociological interests and influences which, hopefully, shed light on our object and subject of enquiry. These include situated learning theory, dramaturgy, critical theory, and concerns with power, knowledge, and identity more generally, and each is more or less evident in the following chapters. Clearly, notwithstanding debates around paradigm (in)commensurability (Willmott 1993), this represents quite a mixture. Overall, however, our approach is a dynamic structural one. Following traditions of critical realism and structuration in particular, we see a (p.24) structural analysis, whereby social structures are not wholly determining, but simultaneously reproduced and transformed through social (inter)action (e.g. Giddens 1984; Whittington 1992) as one (not the only) useful way of making sense of the phenomena under consideration. As we hope to demonstrate, this view of structuring as ‘loose’, negotiated, and dynamic helps makes sense of broad and largely expected patterns of behaviour at the same time as their local reproduction and deviations from what one might expect from a more determinist approach.

Given such a perspective, it is, perhaps, unsurprising that we have a particular interest in boundaries for they can be seen as demarcations of social structure in action—social structuring (Santos and Eisenhardt 2005). But before going on to explore boundaries in more detail, it is important to note that the concept has a much broader resonance and relevance to our study. First, it is evident in everyday ‘insider–outsider’ discourses and experiences of consultancy and, as we shall see, in the context of knowledge flow, in terms of prescriptions centred on breaking down or spanning boundaries (Tushman and Scanlan 1981). More generally, boundaries have received renewed attention in recent years from the apparent prospect of their demise into ‘boundarylessness’ or their becoming increasingly fluid, elusive, shifting, and porous with the advent of post‐bureaucracy, post‐modernity, and globalization for example. Here, art, fashion, politics, management, markets, organizations, and geography are said to become less bounded and distinct. However, such epochalism, as Hernes and Paulsen point out, is countered by the fact that boundaries have long ‘been elusive and complex phenomena’ (2003: 8; also Marshall 2003). Instead, such developments are better seen in terms of our coming to understand things in more processual ways—a world of flux and flow rather than stability and order (e.g. knowledge as a process rather than an object). In addition, they reflect empirical changes in what (and, perhaps, how many) boundaries are felt as important—the organization or nation as no longer a core source of identity for example, but one of a number of shifting identities. In short, the confusion arises from seeing boundaries simply as things rather than as more or less conscious structuring processes.

We now explore boundaries in three specific contexts relevant to our study—social science, organizations, and knowledge flow with a particular emphasis on the latter. Here, we draw on a range of studies to develop a simple framing of interconnected physical, cultural/knowledge, and political boundaries and begin to develop a more contextual understanding (p.25) of these in terms of liminality and project working before developing this further in subsequent chapters in the specific context of consulting projects.

Boundaries, Social Science, and the Sociology of Knowledge

A useful starting point for our discussion of boundaries is Lamont and Molnar's review (2002) of the concept. For them, ‘boundaries are part of the classical conceptual toolkit of the social sciences’ because the idea ‘captures a fundamental social process, that of relationality’ (2002: 167, 169). For example, the authors outline a range of different fields of enquiry associated with concerns to create, maintain, contest, or dissolve institutionalized social differences such as class, gender, race, and territorial inequality—‘us and them’ or ‘insiders and outsiders’ (ibid. 168). But boundaries are not simply an analytical tool for social scientists. Lamont and Molnar (2002) draw a conceptual distinction between two types of interrelated and ‘equally real’ boundaries—symbolic and social.

Symbolic boundaries are ‘conceptual distinctions made by social actors to categorize objects, people, practices, and even time and space’ (ibid. 168), but they are also inter‐subjective or experienced. They emotionally separate and unite people and are contested in struggles over reality, resources, and status. For example, the organizational boundary is important to the extent to which, or when, actors identify with the organization. Such a view is similar to Wenger's notion of boundaries as discontinuities defined by practice, made visible when trying to cross them (1998). Thus, boundaries are a way of expressing the constructions, labels, and experiences produced through a combination of perceiving identity (what something is), difference (from something else), and some intention (desire or thought) of reducing or maintaining that difference.1 For example, a wall becomes a boundary with the awareness that the land either side of it is different in some way and/or with a desire to retain/change that difference or move from one side to the other, with national or property borders for instance. An important point to note here is that a symbolic boundary might be experienced or felt without necessarily expressing it as assuming a particular form—it is only partially (p.26) discursive. This is evident in the more explicitly experiential notion of insiders and outsiders from the traditions of alienation or deviance—a sense of belonging or otherness (Becker 1963; Camus 1946).

Social boundaries, according to Lamont and Molnar, are akin to conventional social structures. They arise when ‘symbolic boundaries are widely agreed upon’ and take a constraining (and enabling) character. In other words, they are ‘objectified forms of social differences manifested in unequal access to and unequal distribution of resources and social opportunities’ (2002: 168). Here then, the organization boundary can be important regardless of whether actors identify with it or its other(s). This view is closer to the, somewhat problematic, Parsonian notion of boundary maintenance and systemic order and ordering. However, the concept of social boundary can be sustained in a more dynamic, socially constructed sense by regarding any order and disorder as the outcome of ongoing negotiation or social action—the continuous interplay between structure and action (Giddens 1984; see also Heracleous 2004; Hernes 2004; Nippert‐Eng 2003). Furthermore, dynamism is also condition and consequence of structural multiplicity and complexity in most social contexts (Whittington 1992). This point is illustrated in an essay by Merton (1972) on insiders and outsiders and the sociology of knowledge (see also Merton 1968: 338–54, 405–7). Given its relevance to our own concerns with boundary relations and knowledge, we briefly consider this study before exploring boundaries in the context of organizational knowledge specifically.

Merton was concerned with a specific political and epistemological debate within sociology at the time of writing about whether one can only claim legitimate knowledge of a cultural context from the inside (as a member) or from the outside, as a ‘stranger who moves on’—the ‘insider and outsider doctrines’ (1972: 32). However, the debate had a much broader sociological significance and, for our purposes, clearly has implications not only for an understanding of the knowledge of external management consultants and their clients, but also for our own position as ‘outsider’ researchers—‘observers as participants’ (see Chapter 3).

In short, Merton rejected the extreme versions of both doctrines in favour of the view that both insiders and outsiders have ‘distinctive assets and liabilities’ (1972: 33). For example, he cites Simmel's view of the outsider role and its epistemological assets which, incidentally, parallel contemporary claims of many consultants as well as academic researchers. (p.27)

He [sic] is freer, practically and theoretically … he surveys conditions with less prejudice … he is not tied down in his action by habit, piety, and precedent [but the objectivity of the stranger] does not simply involve passivity and detachment; it is a particular structure composed of distance and nearness, indifference and involvement.

(1950: 404–5, cited in Merton 1972; see also Marsh 2008; Smith 2008)

Merton goes on to point out that, given the highly socially differentiated nature of most societies (e.g. age, gender, race, and occupation), we are all both insiders and outsiders in a dynamic way, according to context. However, this multiplicity and dynamism was not acknowledged in the extreme doctrines which located people in terms of a single social category. As Merton stated:

This neglects the crucial fact of social structure that individuals have not a single status but a status set: a complement of variously interrelated statuses which interact to affect both their behavior and perspectives. The structural fact of status sets … introduces severe theoretical problems for total Insider (and Outsider) doctrines of social epistemology … [for] aggregates of individuals share some statuses and not others; or, to put this in context, that they typically confront one another simultaneously as Insiders and Outsiders. (ibid. 22)

The implication of this heterogeneity for Merton is that ‘the boundaries between Insiders and Outsiders’ are relatively permeable—‘with or without intent, the process of intellectual exchange takes place precisely because the conflicting groups are in interaction’ (pp. 37–8). He does not elaborate on this potential—as we will do—but points to the barrier presented to this process by the insider and outsider doctrines themselves which contribute to a form of groupthink—‘perspectives become self‐confirming as both Insiders and Outsiders tend to shut themselves off from ideas and information at odds with their own conceptions’ (p. 40). In other words, the view that either cultural insiders or outsiders have a superior knowledge impedes the very flow of knowledge between parties—that is, symbolic boundaries become social boundaries in constraining interaction. This is clearly significant in the context of our study given what we said in the previous chapter of the dominant view of consultants as outsider experts. If either clients or consultants are seen as having a monopoly on legitimate knowledge by virtue of their insider/outsider status, learning is precluded.

We shall return to these issues at various points in our analysis, but for the moment, we shall simply draw some contemporary parallels with Merton's position. For example, one need not take such a strong and (p.28) traditional structural perspective in order to relate to Merton's image of multiple positions. For example, status sets can be readily compared to identity sets. Indeed, many recent observers have described work organizations in a similar way, particularly with regard to different, and sometimes conflicting, social systems (Whittington 1992), social identities (Parker 1995; Trice and Beyer 1993), or, from a post‐structural perspective, fragmented and ‘overdetermined’ selves (Laclau and Mouffe 1985). Furthermore, at an empirical level, in the current late modern context of seeing bureaucratic distinctions between (paid) work and non‐work and temporally stable senses of self as less viable, such heterogeneity and dynamism should be unsurprising (Hochschild 1997; Webb 2004). However, within organization studies, the proclivity to give primacy to a single social category—in this case organizational employee—persists. For example, in a review of studies of inter‐organizational relationships, Marchington and Vincent bemoan the neglect of what they call inter‐personal (as well as institutional) factors in arguing that

there has typically been a tendency to treat organizations as homogeneous and cohesive agents … as the principal (and often sole) level of analysis, so ignoring influences both beyond and within the organization. (2004: 1030–1)

As we saw in the last chapter, this tendency is also reflected in accounts of knowledge flow in client–consultancy relations where the organizational form of the insider–outsider relation is dominant and the consultant's (organizational) outsider status is seen as the basis of his or her expertise and its legitimacy in relation to the client. The limitations or partiality of this view became evident in our discussion of more complex boundaries in consulting projects and this is supported by specific studies of boundaries within organizational studies more generally, to which we now turn.

Boundaries and Organizations

We have noted how boundaries have received growing attention across disciplines. This is also evident in organization studies in the form of economic, strategic, and organizational approaches (Human Relations 2004; Long Range Planning 2004). Indeed, more generally, boundaries and boundary setting can be seen as ‘intrinsic to the very process of organizing’ (Hernes 2004: 10). More specifically, however, boundaries are a particular contemporary concern in relation to three broad areas: variations of the ‘make or buy’ decision inherent in transaction cost economics (p.29) (TCE); the flow of knowledge within and between organizations in terms of resource‐based and learning/knowledge‐intensive organizations; and changes in the structure of organizations and relations between work and ‘everyday life’ (see Hancock and Tyler 2009; Hernes and Paulsen 2003 for a review).

In an assessment of theoretical perspectives, Santos and Eisenhardt are critical of the dominance of a legal or contractual view of organizations and TCE in particular, which emphasize concerns with efficiency (2005). They point to a range of other perspectives and construct a typology of boundary forms on this basis. Here, the contractual view is complemented by boundaries associated with organizational competence (or knowledge), identity (cognition/emotion), and power, in the rather limited sense of a sphere of organizational influence. Similarly, but from a slightly broader focus, Hernes (2004) uses Lefebvre's work (1991) on space to develop a framework comprising ‘physical’ boundaries in terms of ‘physical’ structures such as electronic communications as well as formal rules; ‘social’ boundaries of identity and belonging; and ‘mental’ boundaries in the sense of ideas and concepts which are important to particular groups.

Such typologies are useful in representing (and, indeed, drawing) boundaries which are seen to have analytical relevance. But they also present problems. For example, they were devised for different purposes and so are not directly comparable. Neither locates particular boundaries within the classic social/symbolic distinction. In fact, Hernes' terms may even confuse the issue in that his social boundaries appear similar to Lamont and Molnar's ‘symbolic’ view (2002), while his mental boundaries appear akin to their notion of ‘social’ boundaries or structures. Furthermore, in each case, the different boundaries are clearly closely related or overlapping. This is, in fact, partially addressed in both studies. For example, the physical boundary also has social and mental characteristics (Lefebvre 1991). Similarly, but more importantly, in Santos and Eisenhardt's case (2005), power is inherent in each of the other forms of boundary—competence, identity, and contract. Indeed, it is also a necessary effect of the act of drawing a boundary itself—who and what is included and excluded (Hacking 1999). Despite such problems, as the saying goes, ‘you have to draw the line (boundary) somewhere!’ We have attempted this by adapting the models of Hernes (2004) and Santos and Eisenhardt (2005) for our own specific purpose and context into a ‘loose’ framework of interrelated physical, cultural, and political boundaries (see Table 2.1).

Before turning to this development in detail through a consideration of boundaries and knowledge flow, it is important to highlight boundary (p.30)

Table 2.1. Classifications of primary boundaries (without consideration of process)

Santos and Eisenhardt (2005)

{Hernes (2004) (after Lefebvre 1991)

This study

Contractuala/legal (efficiency)

Physical (e.g. electronic) and formal rules



Social—identity/bonding (emotional)

Culturalb (cognitive/emotional)

Power.(sphere of influence)

Political (interests, rules, contractsa, exclusion)


Mental (central ideas and concepts in group)b

a/b Highlight related phenomena in different rows.

processes. As noted above, in order to avoid overly deterministic analysis and an undue emphasis on order, our approach follows that of Hernes (2004) and others (e.g. Hacking 1999; Nippert‐Eng 2003) in their attention to the role of actors in negotiating, reproducing, challenging, and deconstructing boundaries and the enabling and constraining consequences of this ‘boundary work’ (see Chapters 6 and 7 in particular). Here, in contrast to our own and others' basic classifications of boundaries and, in keeping with Merton's view above, diversity and complexity are key features. Boundaries become composite (i.e. multiple sets of varying strength, substance, and form) and ‘are constantly subject to construction and reconstruction … [but this] does not prevent some boundaries from being relatively stable’ in a given historical context (Hernes 2004: 10). In other words, the social and/or symbolic nature of boundaries is specific to context.

Boundaries and Knowledge Flow

As we have seen, all social interaction can be seen as an apparent movement of knowledge across boundaries, between simultaneous insiders and outsiders. However, attention tends to focus on what is deemed by particular actors as special knowledge or particularly salient boundaries. Thus, in the context of technological ‘development’ and cultural change, for example, anthropologists have long been concerned with studying the ‘spread’ of knowledge between disparate groups of people. Likewise, Christian missionaries seeking to spread the word of God found that even this relatively ‘dogmatic’ form of knowledge was adapted or translated as (p.31) it ‘travelled’ (Huczynski 1993). Similarly, there are numerous classic historical cases of technical innovations or discoveries which failed to travel across geographical (and other) boundaries or over time (see Rogers 1995).

In the current context of ‘knowing capitalism’ and ‘knowledge‐intensive’ organizations and societies, management, as well as technological, innovation and best practice lie at its rhetorical core. Here, aside from the issue of protecting intellectual property, huge emphasis is placed on breaking down national, sectoral, departmental, and, especially, organizational boundaries to allow knowledge to flow so as to be ‘exploited’ or used to ‘explore’ and create new knowledge (March 1991). There are two key assumptions here. The first is that the knowledge in question is unambiguously positive in its outcomes—pro‐innovation bias (cf. Rogers 1995). This is rarely challenged in organizational research, even that which focuses on knowledge and learning (cf. Contu et al. 2003; O'Neil et al. 1998; Semadeni 2001). Second, and of particular relevance to our concerns, is the view that boundaries are synonymous with barriers—that they are dysfunctional.

In fact and following our earlier discussion of Merton (1972), boundaries are a necessary condition for knowledge flow and learning—a means of communication (Lamont and Molnar 2002: 177). Even at the common‐sense level of boundaries in knowledge flow, where else would new knowledge come from? But this is only part of the story, of course. The notion of knowledge as socially embedded—how it is either rooted in specific contexts or constituted by those contexts—suggests that boundaries are important. In the first, more structural and static view, the embedded nature of knowledge, its ‘stickiness’, makes it difficult to tease out and travel to new contexts, but it can be done (e.g. Szulanski 2003). In the second, more action‐oriented and dynamic view (e.g. Orlikowski 2002), knowledge is not in context but made by the context (as words constitute a sentence) so knowledge cannot be transferred or moved. Rather, new contexts (and knowledge) are constructed through interaction or practice and knowledge is translated or, in a metaphorical sense, it flows (also Czarniawska and Joerges 1996; Sahlin‐Andersson and Engwall 2002b). Practice is structured by boundaries, which are themselves simultaneously experienced, reproduced, negotiated, and/or transformed through it. Thus, boundaries become both a condition and/or barrier to learning according to context/s. They serve to represent the shifting contours of embeddedness.

These issues and perspectives lie at the heart of what has become a huge field of study—inter‐organizational learning or knowledge transfer— (p.32) although studies ‘seldom explicitly take the nature of boundaries into consideration’ (Easterby‐Smith et al. 2008: 685). Again, notwithstanding concerns with context or the situational specificity of knowledge,2 emphasis is typically placed on generating universal approaches to ‘freeing up’ knowledge on the assumption that this is a good thing whether in terms of spanning boundaries or removing them (re‐/dis‐embedding). In setting out what are considered key boundaries and issues in the field, we now briefly explore this literature selectively, with a particular emphasis on work which connects with our own concerns with knowledge flow in consultancy (e.g. Nooteboom 2004; Orlikowski 2002; Szulanski 2003; Wenger 1998). In particular, we discuss physical, cultural, and political boundaries which approximately correspond to those identified by Carlile (2004) in his study of managing knowledge across boundaries. We point to some of the interrelationships between these boundary classifications and, importantly, to the notions of shades of grey or relativity and temporality in boundary phenomena, through the concepts of optimum cognitive distance and liminality.

Physical Boundaries

It is self‐evident that knowledge flow is enabled or constrained by physical arrangements which allow or present a barrier to interaction and communication (Szulanski 2003). For example, the classic form of training through ‘sitting next to Nellie’ is now conceived more in terms of ‘operational proximity’ (e.g. Tagliaventi and Mattarelli 2006). The physical boundary is not, of course, restricted to such co‐presence or face‐to‐face contact, but relates to various forms of information and communication technology as well as architecture and ergonomics (e.g. Duffy 1997; Edenius and Yakhlef 2007) which both facilitate, but also shape, the nature and form of interaction and knowledge flow—in effect, boundary objects (Star 1989). Similarly, it might be extended to broader conditions which affect the traditional sociometric dimensions of relationships in terms of the frequency, duration, stability, and direction of interactions (Inkpen and Tsang 2005; Scott 2000).

This physical dimension compares to what Carlile (2004) describes as a ‘syntactic’ boundary which he associates with basic knowledge, or information ‘transfer’ in that, on its own, it is insufficient to facilitate (p.33) other knowledge processes such as those based in practice or the co‐production of knowledge (Scarbrough et al. 2004). In addition, and as with all boundaries, attention should be given to power—in this case, those who are physically included/excluded from interaction (Ebers and Grandori 1997; Lave and Wenger 1991). This is most evident in the case of project working, including consultancy, where close interaction can create a new boundary for knowledge flow—to and from project team members' respective departments or organizations (Tempest and Starkey 2004). It is also important to highlight that physical boundaries are not simply an objective or object‐like phenomenon. Space and objects should also be seen as social and emotional—cultural; understood by actors in different ways for example (Lefebvre 1991; Wilson et al. 2008). This means that ‘operational proximity’ by no means guarantees knowledge flow, in a co‐located project team for example, but it can help generate socio‐emotional identification and dis‐identification, to which we now turn.

Cultural (Cognitive/Emotional) Boundaries

In keeping with the breadth of the concept of culture, cultural boundaries are complex phenomena.3 In the context of knowledge flow however, two key dimensions are evident—cognitive and emotional (see Sturdy 2004 for a discussion of their intimate interrelationship). Crudely, this concerns boundaries of understanding and motivation—how people cannot or will not understand each other. In more sophisticated terms, it can be seen through the notions of optimum cognitive distance (Nooteboom et al. 2007) and what Wenger (1998) describes as ‘economies of meaning’—the sense of ownership or identity that individuals attach to knowledge. The following shows how this relates to the two features which we described earlier as the ‘burden of otherness’ in consultancy—alien knowledge and outsider status.

Optimum Cognitive Distance

Cultural boundaries lie at the heart of what Carlile (2004) terms semantic boundaries or conflicting languages or meanings such as different knowledge domains. A key concept here is that of the ‘cognitive distance’ between parties (Bogenrieder and Nooteboom 2004). Here, wholly shared knowledge bases (i.e. too little cognitive distance) implies that (p.34) there is no boundary and therefore no potential for learning, while too great a distance presents a barrier to shared understanding as any sense of resonance is lacking. In other words, it is not simply a question of knowledge differences being both a strength and a burden, as is evident in the dominant view of knowledge flow in consultancy. Rather, the situation is more nuanced—some ‘otherness’ is essential for learning, but not too much. This optimum balance of newness and resonance is a long‐standing theme (e.g. Simmel 1950), in rhetoric for example, and has come to be associated with ‘absorptive capacity’ in the sense of having a ‘stock of prior‐related knowledge’ as a prerequisite for using ‘outside sources of knowledge’ (Szulanski 2003: 29).

However, others point not only to the importance of a balance in cultural (i.e. cognitive) boundaries, but also to variations based on the type of knowledge or process in question (e.g. Hansen 1999; Holmqvist 2003). In particular, a distinction is drawn between the exploration, or the development of new knowledge and ideas, and the use or exploitation of existing knowledge (March 1991). Here, the weak ties and alien knowledge associated with the traditional ‘consultant as outsider’ view potentially facilitate innovation or ‘exploration’ (as well as allowing the exchange of explicit or simple knowledge). But, this does not simply bring a simultaneous dilemma or ‘burden’ of knowledge transfer problems as the conventional view suggests. Rather, the limitations are more specific in terms of ‘otherness’ hindering the exchange of more embedded/tacit/complex knowledge and exploiting existing knowledge. For these processes, less cognitive distance (i.e. more cultural closeness) is needed (also Hagedoorn and Duysters 2002; Sorenson et al. 2006).

Thus, we are now able to construct a more developed view of learning potential through boundaries, moving away from the simple tension of the simultaneous strength and burden of otherness towards the notion of an optimum level of cognitive distance and one which itself varies depending on the type of knowledge/process involved (see Table 2.2).

Table 2.2. Cognitive distance and knowledge processes

Level of ‘otherness’

Knowledge relationship potential

Relatively high cognitive distance

• Exploration (and exchange of explicit or simple knowledge)

Relatively low cognitive distance

• Exploitation (and exchange of both explicit/simple and embedded/tacit knowledge)

Adapted from Hansen (1999), Holmqvist (2003), Nooteboom (2004).

(p.35) Personal Relationships—‘Redundant Knowledge’ and Shared Characteristics as Bridges

Further complexity can be added to the above view by considering the role claimed for personal relationships in overcoming knowledge boundaries or sub‐optimal cognitive distance. Here, shared understandings or weak semantic boundaries in one domain may serve as a resource or learning bridge in other knowledge domains by virtue of helping establish an emotional connection or ‘intimacy’ in personal relationships (Szulanski 2003). Inkpen and Tsang (2005), for example, identify actors' shared norms as a key ‘relational’ element in the social capital which aids knowledge flow in inter‐organizational networks. These shared characteristics refer not only to common or ‘redundant’ knowledge (i.e. low cognitive closeness) (Nonaka 1994), but also to broader, albeit related, social similarities such as those which might arise from common social and cultural backgrounds (e.g. education, gender, class, ethnicity, and lifestyles) or develop over time such as from joint working in project teams. For example, Zucker (1986) refers to ‘characteristic‐based’ trust which may develop ‘freely’ and help overcome or lessen other boundaries such as contrasting knowledge domains and, even, conflicting interests. Clearly, in the context of consultancy, this translates directly to the importance of personal relations, shared social characteristics, and the instrumental tactics of consultants (‘relationship managers’) in seeking to establish close personal relations with clients, particularly those at, or likely to achieve, senior positions (Sturdy et al. 2006).

The Repulsion and Attraction of Outsider Knowledge

The potential for shared social characteristics and understandings to facilitate the flow of (not too) new knowledge is clearly more than a purely cognitive issue. Following long traditions in social identity theory and inter‐group behaviour, it relates to the emotions of belonging and in/out‐group identification. Here, crudely speaking, the value and knowledge of the in‐group are elevated while those of the out‐group/s are denigrated and blocked (see Paulsen 2003 for a discussion of these issues). In organizational contexts, this has been referred to as the ‘not invented here’ (NIH) syndrome (Katz and Allen 1982) of valuing only that which is associated with the in‐group and recoiling from outsider‐sourced knowledge. Thus, a cognitive boundary (different understandings or knowledge) can become an emotional barrier and vice versa. This can be particularly acute in the context of external competition or perceived (p.36) power inequalities such as the neo‐imperialism of particular countries (Chanlat 1996), multinationals' joint ventures with small firms (Child and Rodrigues 1996) and, as we shall see, management consultants. Here, there can be an inclination to reject knowledge associated not just with outsiders, but with powerful ones especially.

Some caution is needed here however. In particular, how can we explain the attractiveness of knowledge specifically on the basis of its association with outsiders, including powerful ones? This was the focus of Menon and Pfeffer (2003) who cited the examples of a preference for ‘Japanese’ manufacturing ideas in the USA in the 1980s and for the ideas of external consultants over insider knowledge. They highlight how concerns in the literature with knowledge transfer detract from the related process of knowledge valuation. Clearly, outsider knowledge can be seen as valuable through its association with economically successful users such as economies or firms—the ‘dominance effect’ (Smith and Meiksins 1995). But Menon and Pfeffer's argument adds to this. In particular, ‘while outsiders face social, physical, and legal obstacles that inhibit knowledge transfer’ (2003: 498) (i.e. the ‘burden of otherness’), they have two advantages.

First, especially in conditions of intra‐organization competitiveness, insider knowledge represents competition for status and advancement to all but those associated with it, while outsider‐sourced knowledge is less easily dismissed. Indeed, it can be used as a political tool for legitimation. Thus, ‘although organizational boundaries promote identification, they also demarcate an arena within which competition for promotion, status, and salaries occurs’ (2003: 498). In other words, and as we have already noted, in most organizational contexts, multiple or hybrid boundaries and identities prevail. These can vary in significance and strength according to time, place, and individual—once again, we are all insiders and outsiders simultaneously.

The second advantage noted for outsiders is that internal knowledge is relatively accessible and therefore assessable for flaws and does not have the same scarcity or uniqueness value of outsider knowledge, like fine art. As a result, while internal knowledge may be (cognitively and culturally) easier to transfer (as well as often being cheaper), this is hindered by undervaluing it in comparison to knowledge from extra‐organizational competitors and other external sources. Furthermore, although Menon and Pfeffer do not discuss this, the phenomenon is strengthened by the branding activities and pricing strategies of outside actors such as consulting firms.

(p.37) Overall then, and contrary to the dominant view of outsider knowledge being attractive for its newness, but difficult to adopt for the same reason, we find that not only is an optimum cognitive distance necessary for learning, but that in certain contexts at least, the attractiveness of outsider knowledge is based on its relative political legitimacy within internal boundaries, its scarcity, and the relative un‐testability or opaqueness of its economic value. This brings us to the importance of political boundaries and highlights the very loose distinctions between our three conceptual boundaries.

Political Boundaries—Beyond ‘Knowledge at Stake’

Communication or contact, optimal knowledge boundaries, close personal relations, and shared characteristics by no means guarantee knowledge flow. Political relations are crucial. These relate to Carlile's (2004) third key boundary after syntactic and semantic—the ‘pragmatic’ boundary in the adoption of new knowledge. Here, knowledge needs not only to be communicated (syntactic boundaries) and translated (semantic boundaries), but also to be transformed into something else. For Carlile, the key issue seems to be the existential (i.e. cognitive/emotional) threat posed to recipients' prior knowledge—their knowledge ‘at stake’—or what Szulanski (2003) refers to as overcoming the motivation to unlearn. This relates to the motivation we have seen linked to group identity for example (cf. Brown et al. 2005). However and once again, there is an implicit assumption of a single or primary source of identity or, at least, that the knowledge in question is tied to one important identity. Even if this is the case, it is entirely possible to understand, and even promote, an idea or concept and maintain an attachment to contradictory ideas and concepts (e.g. Festinger 1957; cf. Whittle 2005). In other words, knowledge is not always at stake.

More importantly, the approach reflects a somewhat limited view of politics (as well as motivation). Other studies of knowledge flow have pointed to a broader, more material notion of power and interests (e.g. Orlikowski 2002) such as the dependency relations involved in joint ventures or the subsidiaries of multinationals for example. Here, new practices (and their associated knowledge bases) may be imposed on subordinate units, with failure to adopt them being penalized (Child and Rodrigues 1996; Kostova and Roth 2002). Similarly, at the level of the individual employee, there may be little choice but to adopt and adapt a new practice, whether or not any prior knowledge ‘at stake’ is discarded. For (p.38) example, critiques of cultural theories of the flow of management ideas across geographical boundaries suggest that the employment relationship, in the form of the dependence of labour on capital, can significantly counter the hindering effects of any culture clash presented by a new knowledge (Sturdy 2001; Wilkinson 1996). Thus, knowledge flow is not primarily a question of establishing shared interests, as many commentators suggest (Inkpen and Tsang 2005), at least not with all parties concerned. Rather, the form of adoption is shaped by power relations varying between commitment and, in the most dependent/subordinate cases, behavioural compliance for example (Child and Rodrigues 1996). In this way, the legal/contractual form of the organizational boundary, as owners of subsidiaries or joint ventures and labour, is as much a political boundary as one associated simply with efficiency (cf. Santos and Eisenhardt 2005).

Other Boundaries

We have given primacy to what we describe as physical, cultural, and political dynamic boundaries. It is important to note that doing so is effectively an assertion that these are important phenomena and perspectives, given our focus of study and theoretical position. For, as we have seen, boundaries are concerned with structuring and, in their ‘symbolic’ form at least, are almost infinite in number. Orlikowski (2002), for example, sought to identify and classify numerous boundaries that the employees in her study actually felt routinely ‘shaped and challenged their everyday work’ (2002: 255). We share an interest in consciously experienced or articulated boundaries, but given our more structural view, we are also concerned with ‘social’ boundaries which may have become so institutionalized as to become less visible to actors, except in their effects, the employment relationship for example.

But our prioritization is also pragmatic in that an aim of this chapter was to introduce some of the key concepts which inform our subsequent analysis. Accordingly, two often cited boundaries are important to mention here, albeit briefly. The first is personal boundaries such as those which might be associated with personality or personal styles which can create barriers to communication and meaning construction. To a large extent, given our sociological position, they can also be seen as cultural boundaries in that individual differences are socially mediated. Nevertheless, individual dynamics, processes, and styles are important as will become evident in our empirical analysis. For example, when examining (p.39) the role of humour in boundary dynamics, it is clear that some individuals are more adept than others and that this is not solely a consequence of shared identities or common or ‘redundant knowledge’.

The second boundary, which is especially common in studies of knowledge flow, is the knowledge boundary. We have already seen this as part of cultural boundaries, in the sense of contrasting knowledge bases of actors. But more commonly, it has a different meaning—demarcations between types of knowledge (notably tacit and explicit, simple and complex), which are revealed when seeking to convert one to the other (Nonaka and Takeuchi 1995; Wenger 1998). Clearly, and as we have already shown in relation to the idea of optimum cognitive distance, forms of knowledge are important to its flow. Even at a common‐sense level, the idea of ‘acquiring’ knowledge associated with, say, a precise, mathematical technique seems very different from engaging with a ‘quality culture’ for example (Lillrank 1995). However, this depends crucially on context. Moreover, following others (e.g. Blackler 1995), and given our focus on the conditions and nature of knowledge in flow, we are sceptical of views of knowledge as having an objective and object‐like quality, even if such a discourse is often difficult to avoid in practice. At the same time of course, we are conscious that our choice of particular boundaries and what is meant by them effectively ‘valorizes some point of view and silences another’ (Bowker and Star 1999: 5).

Boundary Contexts—Liminality and Projects

Another area which we have not explored in any detail is that of generic prescriptions of managing or, in practice, spanning boundaries for knowledge flow. There are numerous such accounts, many of which are helpful in pointing to likely areas of importance in understanding knowledge flow (e.g. see Anand et al. 2002; Hargadon 1998; Lahti and Beyerlein 2000; Tushman and Scanlan 1981). For example, practices such as joint working; networking; the use of boundary spanners, and objects; facilitating communities of practice; the development of trust and shared values; inter‐personal styles and many others are presented as at least partial solutions to knowledge ‘stickiness’ or production. However, as Orlikowski (2002) points out, such boundary‐spanning activities are double‐edged and can have unintended ‘negative’ consequences:

… sharing identity becomes organizational groupthink, interacting face to face leads to burn out, aligning effort discourages improvisation, learning by doing is (p.40) lost through [staff] turnover, and supporting participation is immobilizing because of conflicts and time delays. (2002: 257)

In effect, context is all important. The idea that there are universal, ‘checklist’, solutions to these issues is not tenable. This is not to say that prescriptive accounts have no analytical or pragmatic value, for many tend to be based on similar broad (e.g. western, ‘knowledge‐intensive’) contexts. With this in mind, we shall now briefly explore literature which relates to two related contexts of particular empirical relevance to our study—liminality and project working. Indeed, liminality is also of conceptual or general relevance for it draws our attention to the relativity or shades of grey of boundary relations as well as to that of temporality and dynamism.

Liminality is an anthropological term referring to a social space that is ‘betwixt and between the original positions arrayed by law, custom, convention and ceremony’ (Turner 1977: 95, 1987). In other words, it refers to a space that is between boundaries, often in a dynamic sense of being in transition such as that between childhood and adulthood. Given recent attention to the idea that boundaries are becoming more fluid, some have argued that liminality is particularly apposite or common in late modernity—constantly betwixt and between (Barley and Kunda 2004; Czarniawska and Mazza 2003; Sennett 2006). However, this is questionable as we suggested earlier in relation to boundaries in general. What is important about the concept for our purposes is that it highlights how boundaries are not always clear cut—insider or outsider—but can be graduated and dynamic in the sense of moving between seemingly bounded states. This is in keeping with Wenger's (1998) notion of ‘boundaries of practice’ which

… are not simple lines of demarcation between inside and outside, but form a complex social landscape of boundaries and peripheries that open and close various forms of participation. (www.ewenger.com)

Liminality is also of particular relevance for its claimed experiential consequences and creative/learning potential. Finding oneself in a liminal space is seen as an uncomfortable and potentially disturbing experience. This is because the relatively settled identities, routines, and rules disappear. At the same time, this means that liminality may also be a creative, liberating, productive, and even desirable place by virtue of its location beyond ‘normal’ practices (Garsten 1999). In particular, the usual physical, cultural, and political boundaries, which may impede learning and (p.41) knowledge flow, are suspended. At the same time, however, this potential can be exaggerated, especially if liminality is becoming more habitual, for new identities, routines, and norms are opened up—new structures and boundaries between those within and those beyond the liminal space (Sturdy et al. 2006).

In organizational contexts, liminality has received growing research attention. Here, it is typically the regular, traditional routines and identifications of the formal organization which are suspended. Examples of those who dwell in liminal organizational spaces include temporary employees, not quite part of one organization or another (Garsten 1999), professionals who identify with neither their organization nor their occupational group (Zabusky and Barley 1997), and those engaged in inter‐organizational networks and joint ventures for example. In addition, and as we noted in Chapter 1, the focus of our study—project working, including inter‐organizational projects such as consultancy—has been identified as an important liminal space (Czarniawska and Mazza 2003) particularly with regard to its potential for knowledge creation and flow across boundaries (Clegg et al. 2004; Tempest and Starkey 2004). Indeed, in the case of projects more generally, a sub‐field of study, project‐based learning (PBL), has emerged to which we now briefly turn.

Projects are an important and growing organizational form in many industries and typically share certain characteristics—a finite duration (ranging from a few weeks to several years), a specific task, and the engagement of project members with differentiated expertise. These characteristics are often positioned as creating organizational advantages not only in terms of flexibility, but also in terms of the potential for learning. First, by bringing together people with different experience they are seen to draw on the ‘strength of weak ties’ and support tacit learning (Granovetter 1973; Schindler and Eppler 2003). Second, by working together towards a typically explicit goal, project members are deemed to learn in a way which is more difficult to achieve in functionally structured arrangements where relationships and knowledge are often more segregated or bounded—through learning‐by‐absorption and learning‐by‐reflection (Ford and Randolph 1992; Scarbrough et al. 2004). Third, and echoing the arguments of Menon and Pfeffer (2003) about outsider knowledge noted earlier, the relatively transient nature of projects can mean that the knowledge they produce poses a lesser threat to vested organizational interests than that arising from individual departments (Sydow et al. 2004).

However, counter‐arguments can be put forward. For example, it can be argued that a neutral (liminal) project status in the organization can (p.42) diminish the legitimacy and credibility of knowledge produced in the project team. At the same time, even if the project members are able to discard or suspend their other (e.g. departmental or organizational) identities and learn within the apparent liminality of the project team, a new boundary can develop around the team (Tempest and Starkey 2004). Furthermore, project working has its own structures and norms which may not always facilitate learning or innovation (Keegan and Turner 2002). For example, the demands of the immediate task—of doing—may take priority over or inhibit reflection and deeper understanding (Sweller 1988; Sydow et al. 2004; also Christensen and Klyver 2006). Similarly, some have argued that ‘the one‐off and non‐recurring nature of project activities’ provide limited scope for drawing out any generalized principles (Hobday 2000) which can be systematically applied and tested in new projects (Gann and Salter 2000). Finally, and echoing the view that is dominant in studies of knowledge flow in consultancy, there is an assumption that bringing together diverse knowledge bases is beneficial for learning. But as we have seen with regard to the notion of optimum cognitive distance, new ideas may be too different and challenging to foster acceptance or even understanding (Bogenrieder and Nooteboom 2004; Chinn and Brewer 1993).

Thus, and as in the case of universalistic prescriptions more generally, these polar arguments suggest that projects are by no means a panacea for encouraging knowledge flow or generation among members, let alone beyond project boundaries. Rather, and once again, knowledge flow is seen to be more complex and contingent. At the same time, however, there are common structural conditions to project working including that of management consultancy projects. We shall examine these along with the specific conditions of particular projects in the following chapters.


This chapter explored the nature of boundaries, particularly in the context of knowledge flow. We have seen how boundaries imply a dynamic, structural view, concerned with relationality and lie at the heart of social scientific analysis and, increasingly, organizational studies. Here, we introduced the distinction and relationship between social and symbolic boundaries and the importance of recognizing multiplicity, complexity, and dynamism such that, following Merton, we typically ‘confront (p.43) one another simultaneously as Insiders and Outsiders’ (1972: 22) through multiple status or identity sets. This was linked to those organizational studies which depart from tradition and do not assume a single or primary boundary or identity associated with the contractual limits of the organization. Rather, by developing the analysis of Santos and Eisenhardt (2005) and Hernes (2004), we constructed a typology of three core and intimately related boundary types or characteristics—physical, cultural, and political—and associated processes of negotiation, transformation, and reproduction.

In the context of knowledge flow, these processes can be seen as central to the different conceptions of knowledge as embedded in, or by, contexts (e.g. Orlikowski 2002; Szulanski 2003). This idea was then developed further by exploring physical, cultural, and political boundaries in relation to studies of knowledge (e.g. Carlile 2004). Here, attention was drawn to the simultaneous social, cognitive, and emotional character of cultural boundaries—cannot learn, will not learn—and a more materialist notion of political boundaries whereby knowledge can flow regardless of cultural clashes, through behavioural compliance for example. Related to this, we discussed Menon and Pfeffer's study (2003) which revealed how concerns with the problems of ‘transferring’ knowledge because it is alien sometimes mask its legitimatory value and attraction, especially in contexts of intra‐organizational competition—an outside source can attract as well as repel. In addition, and importantly, a more nuanced

Table 2.3. Summary of a composite and dynamic approach to boundaries and knowledge flow

Dynamism—social‐symbolic boundary interplay as enabling and constraining


• Operational proximity; technologies; architecture; (boundary) objects; sociometrics

Cultural (cognitive/emotional)

• Multiple knowledge domains and identity sets; optimum cognitive distance for knowledge types/processes; ‘redundant knowledge’ and personal characteristics; belonging (‘economies of meaning’; NIH) and attraction of outsider knowledge


• Knowledge at stake; structured interests (e.g. contractual/dependency relations); inclusion/exclusion

Generic context—for example, multiplicity and gradations of insider/outsider relations or identity sets, including liminality

(p.44) and contingent view of boundaries than that of the ‘black and white’, insider–outsider notion was developed, first, through a recognition of the context‐specific nature of boundaries and PBL and therefore, the problematic nature of a universalist, ‘checklist’ approach to managing them. Second, we introduced the notion of a graduated boundary or optimum cognitive distance for knowledge flow (Nooteboom 2004) according to particular knowledge types (e.g. tacit/explicit) and knowledge processes (exploration/exploitation). This countered the simplistic view of knowledge being either outside/new or inside/already known. Third, and of particular relevance to consultancy projects, we explored the concept of liminality—where actors dwell between boundaries, as neither insider nor outsider, but in transition—and of constructing new boundaries and dynamics, such as that between the project team and others.

The themes and concepts explored in this chapter are summarized in Table 2.3. This provides conceptual reference points for much of our subsequent discussion and analysis of boundaries in context. We shall draw selectively on the generic notions of cultural, physical, and political boundaries as well as those of cognitive distance and liminality as a way of organizing our understanding of consultancy projects in action.

For example, following a brief account of our research methods and descriptions of the case studies in the next chapter, our analysis begins with an attempt to illustrate the complexity and dynamism of multiple and simultaneous boundaries through an overview of knowledge flow potential in the consultancy projects which formed the focus of our research.


(1) This connects to notions of identity such as Gouldner's (1957) ‘manifest’ identity such as that of elder, as opposed to the ‘latent’ identity of cosmopolitan/local and to the experienced, ‘emic’, occupational identity outlined by Zabusky and Barley (1997) as opposed to an ascribed or ‘etic’ identity.

(2) Clearly, not only boundaries are context specific, but so are understandings of them. For example, an organizational boundary has different meanings in different cultural contexts (see Meyer and Lu 2004).

(3) For example, although relevant in a wider context, we shall not explore the issue of cultural variations in cognitive and learning styles (e.g. Bhagat et al. 2002; Warner 1991).