ENTERPRENEURSHIP RESEARCH ON NETWORK PROCESSES: A REVIEW AND WAYS FORWARD

Although entrepreneurship research on networks has studied issues pertaining to network content, governance and structure, we believe it requires a greater understanding of network processes. In this paper, we review how the entrepreneurship literature interprets and applies the concept of process to the study of networks. This allows us to identify areas for future investigation. Our work is also informed by social network theory and research on dyadic interactions in business networks. The paper concludes by presenting a theoretical framework for conceptualizing and studying the various processes associated with network development.

Introduction
In recent years, an interest in networks has permeated entrepreneurship research. A review of network research by Hoang and Antoncic (2003) demonstrates that the entrepreneurship literature emphasizes network content (the nature of relationships and the resource access they provide), network governance (how networks and resource flows are coordinated) and network structure (the patterns of relationships within the network). In their discussion, studies are categorized as either: (1) focusing on how networks impact the entrepreneurial process; or (2) focusing on how entrepreneurial processes impact network development. This categorization is consistent with Borgatti and Foster's (2003, p. 1000) observation that a "fundamental dimension distinguishing among network studies is whether the studies are about the causes of network structure or their consequences." In this article, we build on Hoang and Antoncic to examine how the entrepreneurship literature views networks and how it interprets and applies the concept of process to the study of networks. We also assess other approaches to conceptualizing and studying networks and use this comparison to identify a number of issues for entrepreneurship research. Finally, we develop a theoretical framework to capture the various processes associated with network development. The background to this research and our specific research objectives are outlined in the next section.

Background
A primary contribution of Hoang and Antoncic (2003) is their timely and rich critique of the network literature in entrepreneurship that culminates in directions for further inquiry on network process issues. This stimulated our thinking for the current research for three reasons. First, we note that in the relatively short period of time since Hoang and Antoncic, a number of studies focused on network processes have appeared; studies that could be examined with the sole purpose of understanding process-related issues. Related to this, we are aware of other, earlier, studies that escaped Hoang and Antoncic' s attention; studies that could also provide insight.

Second, we note that while Hoang and Antoncic (2003) imply the need for research that addresses the concept of process, it is not clear how they interpret this concept nor how process may have been defined in past entrepreneurship network literature. For example, while Hoang and Antoncic seem to consider process as involving general sequential activity, Van de Ven (1992) explains that a developmental sequence of events can be explained by four different abstract ideal theories. He also suggests that the developmental view of process is the least understood with researchers more likely to adopt other meanings of the concept. That is, where process is viewed as a logic to explain causation between variables or alternatively, variables are measured over time to capture change. This complexity around the "meaning of process" provides another opportunity for our study since we believe it is important to understand the various meanings used to guide the theoretical arguments and empirical investigations of networks in entrepreneurship.

Third, as a result of conducting our own research we are aware of the range of scholarly areas interested in network phenomena. Thus, while we agree with Hoang and Antoncic's (2003) approach to their review, we see the opportunity for an extension that would: (1) focus on process-related network literature in entrepreneurship, but also, (2) import knowledge from other approaches to help inform future entrepreneurship research.

Following from this, we have a number of specific objectives for this study. First, we extend Hoang and Antoncic (2003) by assessing other network process literature published since their review or not included in their original arguments. This allows us to develop a greater understanding of network processes and to identify areas worthy of further attention. Second, we combine the research from our review with that discussed by Hoang and Antoncic to assess which meaning(s) of process are applied by entrepreneurship scholars as they study networks. This provides insight as to how our understanding of network phenomena reflects the interpretation of process we use. Third, since networks have been studied outside entrepreneurship, scholars can likely benefit from being familiar with different perspectives on the topic (Berry et al., 2004; Zahra, 2007). Consequently, we compare two particular approaches with the entrepreneurship literature in order to identify issues relevant to future research. One has its roots in sociology and focuses on measuring the structure of networks. The other is commonly found in the industrial marketing literature, and emphasizes dyadic interactions within the network. As an outcome of addressing these objectives, we identify the need to more fully conceptualize network development processes. Accordingly, we advance a theoretical argument on this issue. This effort integrates multiple views of process, multiple levels of analysis, and multiple perspectives on network development. We now begin by reviewing the conclusions of Hoang and Antoncic and the meaning of "process."

Revisiting Hoang and Antoncic (2003) and the Meaning of Process Early entrepreneurship research focused on the characteristics of the single entrepreneur. Scholars then began to question: (1) why entrepreneurs were viewed in isolation, and (2) why the entrepreneurial process was separated from other social phenomena. This led to research examining "the causes and consequences of embeddedness in the entrepreneurial process" (Hoang & Antoncic, 2003, p. 167). In particular, Birley (1985) recognized that networks play a catalytic role in organizational emergence, and Aldrich and Zimmer (1986, p. 17) proposed a perspective "which views entrepreneurship as embedded in networks of continuing social relations." Since these studies, networks have been embraced as an instrument for investigating the creation and development of new ventures. This is largely because networks have been shown to improve entrepreneurial effectiveness by providing access to resources and competitive advantage without capital investment.

In a detailed review of network research in entrepreneurship, Hoang and Antoncic (2003) assess the then-extant literature and define two categories of research. The first positions the network as an independent variable by trying to understand how networks affect the entrepreneurial process and outcomes. The second positions the network as a dependent variable by focusing on how entrepreneurial processes influence network development. Contributions from each category are identified as they relate to understanding the content of network relationships, network governance, and network structures. From this base, a set of recommendations are offered. For example, Hoang and Antoncic identify the need to improve our understanding of networks as an independent variable by using longitudinal research to examine how the network shapes the opportunities being pursued or how different governance characteristics affect entrepreneurial outcomes. Recommendations for research where the network is positioned as the dependent variable include understanding the influence of the entrepreneur on the network and examining how interorganizational relationships are developed at the dyadic level. Underpinning all these ideas for research is a conceptualization of process that is characterized by change. For example, implicit in the argument for longitudinal research is the need to track change. Similarly, understanding how dyads are developed or how the network shapes an opportunity implies a need for process research that reflects change over time.

Beyond change, however, what is actually meant by the term "process"? As argued by Van de Ven (1992), scholars tend to adopt different meanings for this concept; meanings which then influence the questions, methods, and contributions of their research. In an effort to reduce confusion in the strategy literature, Van de Ven delineates three meanings of process: (1) when a process logic is used to explain a causal relationship between variables, (2) where concepts are operationalized as a process construct and measured to assess their change over time, and (3) where process is depicted or described using a developmental event sequence.

Within the latter meaning, Van de Ven and Poole (1995) outline four underlying theories of explanation. First, one might apply life cycle theory to describe a linear and prescribed sequence of events. Second, one might adopt a teleological approach by arguing that an end goal is obtained through a discontinuous and adaptive approach manifest in cooperation. Third, a dialectic view would argue that a discontinuous sequence is driven by ongoing conflict or contradiction that resolves itself by balancing power from opposing forces in the development of the entity in question. Fourth, the evolution-based argument would suggest that development is a function of competitive survival whereby change is environmentally influenced and proceeds through a continuous cycle of variation, selection, and retention.

Distinguishing the meanings of process does not however, suggest they are independent. Indeed, Van de Ven and Poole (1995) note it is logical to assume that theories can be combined. One example is Greiner's (1972) model of organizational development depicting a progression of stages for the entrepreneurial firm where each stage and development within it is triggered by conflict and synthesis. This model is argued by Van de Ven and Poole to reflect a combination of the life cycle and dialectic theories within the general meaning of process that refers to a developmental event sequence.

Having summarized Hoang and Antoncic's (2003) arguments and clarified how the concept of process might be interpreted, we turn to a discussion of the entrepreneurship literature pertaining to network processes.

Reviewing the Network Process Literature In this section, we first consider conceptual arguments and then turn to empirical studies. Our focus is on research published post-Hoang and Antoncic (2003) or not included in their original review. We then discuss which meaning(s) of process seem to be applied in the extant literature and how this influences our understanding of network phenomena. Finally, we examine how other areas of scholarly inquiry view networks and issues of process. To maintain focus in our review, we restrict our efforts to understanding the contributions of those studies focused on either how networks affect the entrepreneurial process and outcomes, or how the entrepreneurial process and outcomes influence the network. Accordingly, we do not address literature at the network or entrepreneurship interface specific to (for example) immigrant entrepreneurs, social capital, or internationalization. While such research is important, each represents a significant body of literature in its own right and as such, their examination is beyond the scope of this article.

Contributions to Our Understanding of Network Processes
Hoang and Antoncic (2003) specifically call for further research on network development processes and our search to extend their work identified six empirical articles focused on this topic. The other eleven articles pertain to how networks influence entrepreneurial processes. Beyond these seventeen studies, only one conceptual argument is identified (Hite & Hesterly, 2001). We note this because Hoang and Antoncic also identify a single theoretical contribution: Larson and Starr's (1993) model of organizational formation. This dearth of conceptual work appears to support the concerns of Busenitz, West, Shepherd, Nelson, and Zacharakis (2003) and Zahra (2007) that theory building is a challenge for entrepreneurship researchers. This is perhaps because process-variables are "... messy and difficult to capture" (Zahra, p. 448).

If we consider Larson and Starr (1993) relative to Hite and Hesterly (2001), the former discusses how the entrepreneurial firm's network develops through three stages. Within each stage, there is a process of exploring, selecting, and using dyadic ties. This process is driven by the entrepreneur or firm, and actions are shaped by the actor's social context. Larson and Starr argue that over time, the network reflects increasing density, complexity, and interdependence of actors, leading to the creation of an organization. In comparison, Hite and Hesterly argue that networks change from being identity-based to more calculative and the network shifts from being dominated by socially embedded ties to having a balance of embedded and arm's-length ties. As the firm develops, the initially cohesive network is expected to shift to one that is sparse or loosely integrated, and characterized by structural holes. The network also shifts from being path-dependent (reliant on history and chance) to one that is more proactively or intentionally managed by the entrepreneur. Hite and Hesterly' s concluding argument is that new firms can benefit from networks that are cohesive (following Coleman, 1988) but also networks that emphasize structural holes (following Burt, 1992). As such, they suggest that one type of network will serve the firm at emergence and another will be more appropriate at early growth. The network and organization are seen to codevelop and are understood relative to the environmental influences or challenges of resource availability, access, and uncertainty.

Drawing these two conceptual arguments together, we see that Larson and Starr (1993) and Hite and Hesterly (2001) both agree that networks become more complex over time. Both also portray process as a developmental course of activity and base their arguments on a rational-action view whereby entrepreneurs create and manage their networks; networks that are adapted and aligned to gain resources. Interestingly, while both contributions focus on describing network development, the Larson and Starr model is one of "organizational formation" and Hite and Hesterly (p. 275) develop arguments to explain which kinds of networks are "more conducive to the success of new firms." Thus, the intent of both arguments can be placed in Category 1 where the network is an independent variable, although each can straddle both categories identified by Hoang and Antoncic (2003).

A notable point of difference in the two conceptualizations is that Larson and Starr (1993) emphasize the relational dimensions of dyads and argue that with time, the network becomes increasingly dense. In contrast, Hite and Hesterly (2001) emphasize the structure of the overall network and argue that network density and cohesion will decrease. We suggest that this "difference" might be more usefully considered as complementary, since the broader arguments provide a conceptual basis for understanding both dyadic relationships and the overall network and allows for the network to be viewed as either a dependent or independent variable (or both).

Turning to the empirical literature addressing network process issues, our review identified 17 articles either published since Hoang and Antoncic (2003) or not analyzed by them. Each of these articles can be placed within Hoang and Antoncic' s two categories for analysis. We begin with Table 1, which summarizes the empirical studies we identify with Hoang and Antoncic's Category 1 (where the network is positioned as an independent variable). (1) One group of studies emphasizes analysis at the level of the dyad. For example, Elfring and Hulsink's (2003) case research shows that a mix of strong and weak ties influences how the start-up discovers opportunities, secures resources, or obtains legitimacy. They also highlight the context of radical innovation, where strong ties are emphasized for securing resources and weak ties help obtain legitimacy. This adds to Hoang and Antoncic's discussion about when strong and weak ties impact entrepreneurial processes.

The various ethnographic studies by Jack also expand our understanding of tie strength and in particular, Jack, Drakopoulou Dodd, and Anderson (2004) demonstrate the need to move from the dichotomy of strong vs. weak ties toward a more multiplex perspective. That is, where ties are differentiated not only by intensity, but also the content of the relationship. This is extended by Jack (2005) who argues for an appreciation of information requirements, tie usefulness, and trust. She also notes the existence of dormant ties; ties which may awaken or be reactivated later in time. Interestingly, the various studies by Jack show little evidence of weak ties in venture development. Her results do, however, indicate that strong ties provide both opportunities and constraints to the new venture, thus supporting Elfring and Hulsink's (2003) observations on the risks of overembeddedness.

Consistent across all of Jack's work is her argument that a process of embedding is necessary. That is, rather than simply developing ties, the new venture must actively become part of and maintain its network, a view consistent with Larson and Starr (1993) and Hite and Hesterly (2001) and the general literature that argues networks are embedded in economic and social relationships (e.g., Granovetter, 1985). Jack and Anderson (2002) also emphasize the need to understand the broader environment in terms of its social and spatial context, since the entrepreneur draws from his/her local environment and contributes to it. Equally, they note the endogenous influence of the founder's motives on (for example) locational decisions. Jack (2005, p. 1253) also implies the need to examine relational issues at two levels: (1) the dyad; and (2) the broader network, observing that: "the nature of networks is about the links and bonds that form the foundations of the network and shape its structure."

Notable in Category 1 are the studies we found that are interested in the influence of network change on venture development; a focus not apparent in Hoang and Antoncic's (2003) review. For example, Butler and Hansen (1991) use longitudinal qualitative data to identify how social, business, or strategic networks are used through stages of venture development. They focus on the purposeful development of action sets and note these supplement the social role sets of the actor. While the role of strategic networks is confirmed in the final stage of network development ("ongoing business"), social networks play a significant role in all phases of organizational development. Important in this study is the impact of the environmental context in terms of industry influences and demand patterns.

Taking a slightly different approach, Lechner and Dowling (2003) and Lechner et al. (2006) focus not on the type of network (in the sense of social vs. business) but the benefits it offers in different stages of venture development. For example, Lechner and Dowling's case shows that social and reputational researchnetworks decrease in importance over time, while the importance of co-operative networks increases. These results are summarized in a staged model of organizational development for high growth firms as it relates to network composition. Then, Lechner et al.'s follow-up survey research provides insight as to which ties forming different types of networks, matter when. At start-up, for example, cooperative technology networks negatively impact performance, since they signal that the firm is not yet ready to exploit opportunities. In contrast, marketing information networks positively impact sales after start-up.

The work of Lechner et al. (2006) is notable in its efforts to include performance as an outcome variable, an area of weakness in the literature identified by Hoang and Antoncic (2003). Other research taking this approach includes Donckels and Lambrecht's (1995) study of the impact of network characteristics on small-business growth, Littunen's (2000) study of how characteristics influence the start-up vs. operational phases of growth, Havnes and Senneseth's (2001) panel study examining the impact of network size on performance at different points in time, and Watson's (2007) efforts to model the impact of certain advice network characteristics on survival, growth, and ROE. This suggests that although research linking networks to performance is needed, we do have some data as to how networknetwork characteristics must be understood and managed for the benefits they provide to the venture over time and at specific stages of development.

Generally common to the research placed in Category 1 is the use of cross-sectional data to compare variables across defined stages. While this is a practical approach to data collection, it highlights Hoang and Antoncic's (2003) concern regarding the methodologies used in network process research since cross-sectional studies do not capture the dynamics of change. Even those studies using historical data (Havnes & Senneseth, 2001; Littunen, 2000; Watson, 2007) only capture surface-level patterns of change because their focus is on testing causal relationships.

We now turn to Category 2, where networks are considered to be the outcome of an entrepreneurial process, i.e., a dependent variable. As seen in Table 2, we identify six new studies beyond Hoang and Antoncic (2003). (2) One of these (Greve & Salaff, 2003) offers a (rare) international comparison of how different phases of establishment impact the network. In an approach reminiscent of Butler and Hansen (1991), this study uses cross-sectional survey data from the United States, Italy, Sweden, and Norway to assess how advice networks differ across the development phases of motivation, planning, and establishment. Rather than finding network stability, Greve and Salaff show that the smallest networks are used in the first phase of development. These grow in the second phase, but then decrease and become more focused in the third phase.

Of the remaining five studies, three provide a retrospective analysis (Hite, 2005; Larson, 1991; Lorenzoni & Ornati, 1988) and two offer a longitudinal perspective to understanding networks (Schutjens & Stam, 2003; Steier & Greenwood, 2000). A notable pattern across Category 2 studies is therefore the use of methodological designs that provide time-sensitive insight and fine-grained data regarding network development issues--designs argued by Hoang and Antoncic (2003) to be lacking in the entrepreneurship literature.

The most recent example of a Category 2 type of investigation is from Hite (2005). She employs case research to understand how the entrepreneurial firm transforms network ties toward full relational embeddedness. Her results show that the shift is influenced by network entry, social leverage, and trust facilitation, and she describes a dynamic and strategic picture of transformation. Hite also discusses the disadvantages and risks of relationships. For example, continuous change in social components may increase the need for formal governance mechanisms. Her results suggest a switch in character within relationships rather than a switch between relationships. This is consistent with Lechner and Dowling (2003), who conclude that weak ties must be transformed into strong ties for value exploitation, although firms differ in their relational and combinative capabilities and absorptive capacity.

Like Larson and Starr (1993), Hite (2005) highlights the additive nature of ties whereby social relationships can develop toward business ties in a path-dependent pattern. However, Hite also suggests that firms can control social leveraging by being proactive. This is consistent with the early work of Lorenzoni and Ornati (1988), who use case data to portray the growth patterns of entrepreneurial networks as shifting from unplanned (loose) to planned (efficient) to structured (effective). Similar patterns are seen in Larson' s (1991) case research. She shows that network relationships are not formed by chance, but reflect predicted exchange patterns based on a company's changing needs in a context influenced by competitive forces. Of additional interest is her conclusion that relationships between network partners develop in parallel with the firm itself development occurs.

Turning to longitudinal work at the network level, Steier and Greenwood's (2000) case study suggests that to overcome the liability of newness and competitive pressures, the entrepreneur should develop a network that is diverse rather than uniform, and extensive rather than limited in size. Consistent with many of the other studies, Steier and Greenwood conclude that the network requires strong rather than weak ties. Like Jack (2005) and Hite (2005), they note the benefit of dormant ties, and redundant ties were found to be advantageous if the tie provided (for example) future potential.

The other longitudinal study is from Schutjens and Stam (2003), who extend Butler and Hansen (1991) with an examination of the major contacts of 313 new firms. They find that as the firm develops, upstream contacts (i.e., with suppliers) become more commercial, while downstream contacts (i.e., sales relationships) change from being business-focused to include both and social relationships. In clarifying which ties become more commercial and which become more social, Schutjens and Stam provide a bridge to Larson and Starr's (1993) discussion on tie complexity and whether the network changes to become more social or business-oriented. Although Schutjens and Stam find that firms become more selective with customer relationships over time, they provide little evidence of how and why changes take place, i.e., how interorganizational relationships and their resultant networks develop. Larson (1991) and Steier and Greenwood (2000) provide some insight into this issue, with the former suggesting that networks evolve in a goal-oriented rather than reactive manner. The latter provides a rich description of how a network develops by continuously constructing and reconstructing ties that become more multiplex and robust, with network reconfiguration characterized by pivotal years or punctuation points. An appreciation of "how and why" networks develop is also offered by Lorenzoni and Ornati (1988) and Hite (2005). This contrasts with the studies discussed by Hoang and Antoncic (2003), many of which use comparative data to test for differences in the networks developed by (for example) male vs. female entrepreneurs rather than developing rich insight to the businessdevelopment process per se.

Summarizing this, it seems that by extending Hoang and Antoncic (2003) to focus on inadvertently excluded or published since their review, we are able to identify a small group of studies that further our understanding of network development issues. Nevertheless, it is clear that, as noted by Hoang and Antoncic, our understanding of this topic is relatively limited. This conclusion is supported by the patterns of Tables 1 and 2, which show a clear emphasis on research where the network is the independent rather than dependent variable. Beyond the question of where research is focused, a further question arises. That is, in categorizing networks as either an independent or dependent variable, it might be suggested that entrepreneurship scholars view networks through a positivistic lens; a lens that is causal and explanatory in nature. Is this appropriate given the concept under discussion is that of "process"? To help understand this, we now assess how the concept of process appears to have been applied or interpreted in both: (1) the studies in our review; and (2) those reviewed by Hoang and Antoncic.

Interpretations of Process
As summarized in Table 1, all 22 studies in Category 1 seem to interpret process using Van de Ven's (1992) first definition. That is, process is a form of logic used to explain a cause-and-effect relationship. In contrast, the seven studies in Category 2 (Table 2) demonstrate a primary view of process that reflects either Van de Ven's second meaning (four studies interpret it as a concept that can be measured and tracked for change over time), or his third meaning (three studies depict process as reflecting a developmental sequence).

Taking this analysis further, we see in Tables 1 and 2 that eight studies use what Van de Ven and Poole (1995) refer to as a single-motor, i.e., a single interpretation of process to focus their work. Using Van de Ven and Poole's terminology, four studies incorporate a triple-motor perspective (Davidsson & Honig, 2003; Lechner & Dowling, 2003; Littunen, 2000). Most common, therefore, are investigations with a dual-motor (17 of the 29 studies). This pattern is consistent with Van de Ven and Poole's observations that most arguments related to organizational development and change are composites of two or more ideal-type motors. What is notable here, however, is that the composite approach integrates not only theories pertaining to process as a developmental sequence but also the other, separate meanings of process that focus on causation or measurement of variable change. Further, the rich perspective made possible by viewing process as a developmental sequence, argued by Van de Ven (1992) to be the least understood approach, is limited to a secondary role in most of the research reviewed here. This is particularly evident in Table 1.

If we focus on how the "developmental sequence" meaning of process is characterized in our literature, prescriptive life-cycle theories dominate. Beyond this emphasis, three points are noteworthy. First, while teleological views of process are somewhat in evidence (most notably with Larson, 1991, or Hara & Kanai, 1994), they tend to be used in a summary format to portray an unpredictable or divergent view of process that has been measured in other ways (see Elfring & Hulsink, 2003; Hite, 2005; Jack, 2005; Schutjens & Stam, 2003). Second, although Greiner (1972) depicts venture development as a dialectic process, only Hara and Kanai move in this direction. Even then, this is only when summarizing the nature of interactions during tie formation, and they do not take a dialectic perspective in their main discussion. Third, although most studies use the term "evolution" interchangeably with "process," only three examine networks in a manner that reflects Van de Ven's (1992) definition of evolutionary theory (Hite; Lechner & Dowling, 2003; Steier & Greenwood, 2000). Of these, only Steier and Greenwood view their entire study through an evolutionary lens by describing the network in terms of a progression of variation, retention, and selection.

Again, however, most of the investigations in Tables 1 and 2 seem to interpret process using a meaning other than "developmental sequence of events" and as a result, track change in either a causal or descriptive manner. While this might facilitate the conduct of a focused investigation with a relatively straightforward approach to data collection and analysis, it also leads to a somewhat predictive view of process. This is further reinforced by the fact that most of the developmental process arguments tend to apply life cycle theory, possibly to enable retrospective analysis. Life cycle theory is by its nature, however, prescriptive. This may explain the rather positivistic references to the network as either a dependent or independent variable, but we caution that one consequence of this approach is a lack of depth in understanding the "how and why" of network processes. It also implies that network processes are clear-cut, predictive, and involving a single entity. We suggest this misrepresents reality and believe that the studies offering the richest understanding of this are the empirical articles underpinned by teleological theory (Larson, 1991) and evolutionary theory (Steier & Greenwood, 2000). The former shows that network processes can be viewed as constructive rather than prescribed, and the latter highlights the need to accommodate multiple entities and multiple levels of analysis (in comparison to the single entity approach of the life cycle argument).

At this point, we have addressed our first two research objectives. Our third objective is guided by the suggestion that entrepreneurship research on network processes may benefit from being informed by different perspectives on the topic. Accordingly, we now review two other approaches to network research.

Other Views on Networks
By distinguishing between studies that position the network as either an independent or dependent variable, Hoang and Antoncic (2003) imply there are different ways to study network processes. To represent the school of thought that generally examines the impact of the network on the social group or organization (e.g., where the network is an variable), we discuss social network (SN) research. In contrast, the business network (BN) (3) approach emphasizes an understanding of the interactions that create dyadic relationships and consequently, the wider network. It therefore represents the second category: networks as a dependent variable.

The Social Network Perspective
The SN literature has a foundation in Simmel's (1955) arguments regarding the importance of understanding group composition in order to understand social life (Smith-Doerr & Powell, 2005). Later efforts emphasized network structure and over the years, concepts from SN research have been widely adopted in various literatures including entrepreneurship. Indeed, the studies summarized in Table 1 draw heavily on the SN literature. Within this perspective, the structural aspects of networks are generally emphasized with a reliance on mathematical models of the motion of change (Kilduff & Tsai, 2003). These models apply a variety of measures to assess tie configurations and identify similarities or differences across networks (Smith-Doerr & Powell). Theoretical arguments include Coleman's (1988) explanation of the importance of a cohesive network, Burt's (1992) argument for structural holes, and the extensive discussion of strong and weak ties (Granovetter, 1973).

Importantly, the SN literature provides a rich discussion of the concept of embeddedness (Granovetter, 1985; Uzzi, 1996) and argues that economic behavior is embedded in a social context or in a network of relationships. The SN research also considers political, cultural, economic, and technological development as exogenous influences on both individual and interorganizational levels of cooperation. In this perspective, a specific environment is understood to "constitute an opportunity structure containing a resource pool uniquely suited to organizational forms that adapt to it or help shape it" (Aldrich & Zimmer, 1986, p. 10). Arguments pertaining to this concept of a resource niche are found through the SN literature and are consistent with Aldrich (1999) in positioning the environment, be it the social context or others, as determinant. At the same time, SN research recognizes endogenous influences such as efforts by the focal firm to access resources by structuring relationships in an efficient manner. Koka, Madhavan, and Prescott (2006) refer to this as purposeful network action.

Of particular interest in the SN tradition are studies on the networks of individuals. Examples include research on the impact of social ties in job-seeking (Granovetter, 1995) and career advancement (Podolny & Baron, 1997). In such studies, the SN approach uses the formation and dissolution of ties (i.e., the appearance and "death" of nodes) to measure network change, and analysis is focused on structural characteristics such as network size, density, or the position of actors in the network. This enables an understanding of (for example) the potential for innovation or the identification of power bases for information control and brokerage (see Ahuja, 2000, for an empirical example).

Another feature of SN research is the common use of longitudinal studies of large datasets to examine networks (often interorganizational) at different points in time to identify aggregate tie patterns. For instance, Owen-Smith, Riccaboni, Pammolli, and Powell (2002) compare university-industry relations in the United States and Europe. They analyze 12 years of data consisting of 1,026 linkages and use network visualization methods and large scale network analysis techniques to identify different collaborative systems. Other work involving clusters and their patterns of interaction includes Powell, Koput, and Smith-Doerr (1996) while Seabright, Levinthal, and Fichman (1992) explore how the nature of attachment between organizations impacts the dissolution of auditor-client dyads. This study is relatively rare within the SN approach since it provides some insight as to the impetus for severing a tie. As with much SN research, however, it does not capture the actions and explanations underlying tie dissolution and network change, although Kim, Oh, and Swaminathan (2006) recently argued for research on implementing network change within existing relationships.

If we shift our focus to consider levels of analysis, Ibarra, Kilduff, and Tsai (2005) argue that little attempt has been made to link individuals and their networks with larger network systems. However, Liebeskind, Oliver, Zucker, and Brewer (1996) study the biotechnology industry to understand how individual and firm-level networks impact organizational learning and flexibility. Oliver and Liebeskind (1997) then draw on this data to argue, like Ibarra et al., that networks must be understood at both the individual and organizational level, as well as within and across organizations. Some SN studies also consider both dyadic ties and the broader network. For example, in addition to Seabright et al. (1992), Powell, White, Koput, and Owen-Smith (2005) examine how the formation, dissolution, and reestablishment of ties by 482 firms over a 12-year period shaped the network structure of the biotechnology industry. Uzzi (1999) combines ethnographic research (used to understand the difference between embedded and arm' s-length ties) with a survey of 2,300 U.S. firms to examine how embeddedness can influence which firms access financial capital and at what cost. In doing so, he assesses: (1) the dyadic ties between the entrepreneur and the loan managers at a bank, and (2) the ego network of direct ties between a firm and all its banks. This study is similar to other works by Uzzi (e.g., Uzzi, 1996) and his approach is somewhat unique in SN research because he examines tie quallty and how network configuration influences a firm's ability to perform. In crossing different levels of analysis, these studies represent arguments in the SN literature regarding the need to develop a multilevel understanding of interorganizational networks (Contractor, Wasserman, & Faust, 2006; Hagedoorn, 2006).

Overall, the SN literature generally emphasizes the identification and measurement of tie and network characteristics to understand the influence of structural change. This means that if we apply Van de Ven (1992), a common interpretation of process in this type of research is one that examines how variables change over time. The SN literature also tends to view process as a logic to explain causation. Importantly, even the stream of research that connects the dyad with the network tends to focus on structural analysis with a positivist lens. For example, although Uzzi (1996) considers how ties become embedded in the apparel industry, his primary interest is to assess the impact of ties on economic performance. Further, his general approach to research considers only one actor in the dyad and is cross-sectional rather than longitudinal. This leads us to an alternative approach to understanding networks: the business network perspective.

The Business Network Perspective

A particular characteristic of BN research is that it accounts for both actors in a dyad and investigates how and why relationships change over time. This perspective argues that a change in the dyad results from: (1) actors learning about how to utilize new combinations of resources, (2) the contrasting perceptions of actors in relationships, and (3) actors continually looking for opportunities to improve their position towards important partners (Hakansson & Snehota, 1995). Following from this, BN research suggests that network development is cumulative in that relationships are continually established, maintained, developed, and broken to provide satisfactory economic return or to create a position in the network. This implies change is driven by factors endogenous to the firm. The BN approach also argues that neither a hierarchy nor a single central actor is in charge of organizing the network per se. Instead, networks are seen as multiplex adaptive systems, where actors are simultaneously involved in ongoing network management (Ritter, Wilkinson, & Johnston, 2004). Thus, change is also endogenous to the network. This is reflected in Freytag and Ritter's (2005, p. 644) statement that it is not a question of managing a network but managing in networks and thus, it is "more appropriate to talk about networking, influencing and interacting, i.e., processes instead of outcomes."

At the most macro level, the BN perspective argues that exogenous influences such as economic conditions or technological advancement will be transformed into or combined with endogenous factors such as confrontation between actors. Thus, changes originate in the dyad (Hallnen, Salmi, & Havila, 1999) in a manner that can be positive or negative (Ritter, 1999), and any change in one part of the network will produce change throughout the whole network. Any dyad causing network change will also receive and transmit change (Hakansson & Snehota, 1995; Hallnen et al.; Hertz, 1996); change that can be proactive or reactive as the nature of any relationship shifts. This is exemplified in Hertz' s (1998) longitudinal case research on how change in one relationship explains sequential consecutive change in others.

This focus on connected change allows for BN researchers to study transformation within networks. They do so by investigating the concept of "interaction" between parties, where relationship development is conceptuallzed as interaction (rather than action) between independent firms or actors (Ford & Hakansson, 2006). Relationship development and transformation are therefore believed to be reciprocal and dependent on the expectations of both parties regarding their future interactions (Hakansson & Snehota, 2006). Further, relationship development increases each actor's knowledge and helps them create reallstic expectations of one another (Seines & Sallis, 2003). In this sense, the network is understood to coevolve with the relationships that form it, and experiences from one relationship are transferred to another in the network (Hakansson, Havila, & Pedersen, 1999). This highlights the interplay between dyads and the overall network.

As Johanson and Mattsson (1994, p. 325) note, research in the BN tradition "emphasizes dynamic, individual and interconnected exchange relationships within systems that contain interdependencies of both a complementary and a substitutive nature." Accordingly, BN research takes the position that the network structure is never stable. That is, "it is a structure with inherent dynamic features, characterized by a continuous organizing process" (Hakansson & Snehota, 1995, p. 271). Even if network patterns appear static, the BN perspective recognizes that existing relationships can change their content and strength. That is, change occurs within relationships.

BN research also regards the network as being comprised of different types of relationships. At one level, it recognizes they may be positive or negative and allows for both cooperation and competition. Going deeper, Hertz (1996) distinguishes between one-way, passive, infrequent,

or temporary relationships and argues that in order to have interactions, a certain degree of frequency, intensity, and stability must exist. This is connected to the fact that the BN perspective is not restricted to the present, but takes into account the past and future of relationships. As part of understanding network history for example, BN research acknowledges the concept of "sleeping ties," referring to existing but dormant relationships that can be reactivated (Johanson & Mattsson, 1992).

To summarize, the BN perspective focuses on understanding how to establish, build, and maintain or change relationships to create a position within a network. This signals the connection between various levels of the network. Further, the BN approach is focused on how relationships change and why change occurs (unlike SN research). Thus, compared with the methodologies prevalent in SN studies, those in the BN tradition are generally more case-based and interpretivist in nature. While analysis of network structure is not paramount, an understanding of all potential relationships is considered relevant (including their history and role), and the focus of analysis is on the interaction between actors. Accordingly, if we apply Van de Ven (1992), BN researchers tend to assess how variables change over time but do so using "process as development" theories that portray interaction as being: (1) purposeful and adaptive but not necessarily sequential, (2) characterized by opposing forces that can lead to the status quo or change, or (3) involving a course of action characterized by continuous variation, selection, and retention. Thus, they are teleological, dialectic, or evolutionary in nature, or possibly a combination of these, rather than based on life cycle theory. The BN perspective also discusses the multi-directionality of change, not often considered by SN research.

Informing the Entrepreneurship Literature

Up to this point, we have reviewed the entrepreneurship network literature for recent contributions and interpretations of the term "process." We have also discussed two other approaches to network research. We now integrate this information to identify potentially fruitful areas of inquiry for entrepreneurship scholars. To facilitate this, Table 3 summarizes what is currently understood from the entrepreneurship, SN, and BN literatures.

We begin by assessing the most common level of analysis in network research. Relatively little research in entrepreneurship provides a link between multiple levels in the network. This is in spite of an appreciation in the SN and particularly, BN literatures that it is important to understand how the dyad and network are interrelated. Of interest here for example, is understanding how observed changes within or across specific relationships impacts the entire network and in turn, how change at the network level influences identified relationships. This could be as simple as applying the SN perspective to consider how the addition or deletion of ties changes the network. It could also involve using the BN approach to explore how different dyads perceive trust and mutuality and how this benefits the broader network. Alternatively, one could examine how information sharing across the network allows for learning and other cognitive benefits at the level of the dyad.

As regards the type of network studied, the entrepreneurship literature tends to focus on networks with the horizon or boundary defined by the focal firm. This differs markedly from the BN perspective, which views networks as borderless and the SN approach which considers networks to have clear membership boundaries. Both the BN and SN approaches also tend to examine the broader network, including all potential relationships connecting all the actors. At the same time, the lens applied by each view is different. For example, SN scholars tend to be interested in the wider system of ties and their various characteristics. Consequently, the full network is relevant. In contrast, BN researchers tend to study the interplay within and between dyads as they relate to the broader context of an inter-organizational network and the external influences that shape it. Thus, the fluidity of a borderless network is appropriate as is viewing all relationships as part of the same network. This differs from some entrepreneurship research where functionally different networks are considered (e.g., Lechner et al., 2006). Entrepreneurship research also tends to focus on network processes as they pertain to a particular venture and consequently, the defined egonet is most relevant. We suggest, however, that all three perspectives may be useful in entrepreneurship research. For example, to understand a venture's position in the network, it would be appropriate to adopt the SN approach of examining (for example) centrallty in the full network rather than the egonet. Similarly, analysis of the egonet or full network often identifies dyads of particular interest and brings forth questions regarding the endogenous or exogenous influences on the entrepreneurial network. The theoretical arguments and methodological approaches to examine such issues can be found in BN research.

This leads us to consider the issue of network management. The entrepreneurship literature generally portrays network development as being controlled by the focal firm or entrepreneur as per Weick (1979). The SN literature also sees network formation as calculative but often reflects Aldrich's (1999) view that the driver of change is the context rather than the entrepreneur. An alternative view (typical of the BN literature) is that the entrepreneur both engages in purposeful action and is externally controlled (Johannisson, 1988; Koka et al., 2006). That is, rather than behaving as a reactive economic actor (as per Aldrich & Zimmer, 1986), the entrepreneur enacts their organizing context (network) to managing within the broader set of environmental influences. As noted in the BN literature, this involves managing in a network rather than management of a network. We suggest that by adopting the BN approach, entrepreneurship scholars might expand their understanding of the role of the entrepreneur relative to the environment in different contexts and over time.

Related to the discussion on network management is the question of how the different approaches view exogenous and endogenous influences. Generally speaking, the entrepreneurship literature tends to emphasize the endogenous influence of the firm or entrepreneur on the network and the SN literature studies the exogenous influences of the environment. Further, the concept of embeddedness is present in both the entrepreneurship and SN literatures, with the BN perspective noting that since the firm is dependent on other organizations in a network, it needs to be in constant interaction with this context. We suggest however that embeddedness needs to be understood beyond the social contexts influence on the focal firm, entrepreneur, or specific tie. Rather, both these levels are influenced by a broader set of environmental factors (Hagedoorn, 2006). This view is consistent with the BN literature and suggests that it would be appropriate for entrepreneurship scholars to distinguish between: (1) the focal firm and entrepreneur that are endogenous to the network and broader system, (2) the network or social context that is exogenous to the focal firm or entrepreneur and also endogenous to the broader system, and (3) the macro environment or context that is exogenous to the network, focal firm, and entrepreneur (e.g., sector influences, market, economic, or legislated conditions, technological or cultural change). Taking this approach would provide specific parameters for study and the potential to compare multiple levels of analysis.

Turning from the network to characteristics of the ties that form it, we see in Table 3 that the SN approach tends to emphasize tie existence or tie strength, and recognizes that ties can be both social and economic. In contrast, entrepreneurship research has begun to take a broader outlook by incorporating (for example) tie usefulness, trust, and information requirements. Certain entrepreneurship scholars also note the existence of dormant ties, but to this point, the array of tie characteristics is not yet well explored in this literature. This, however, is where the BN perspective is useful since the richness of tie characteristics is perhaps best captured by this approach. For example, BN researchers recognize that relational content can and will change in intensity, length, and depth, and can include contradicting (positive and negative) dimensions. Interestingly, while the SN literature views constraint as a negative concept, the BN perspective views it as offering a positive trigger for tie development or change. The notion of "reciprocity" from the BN perspective is also relevant since it is not self-evident that a relationship means the same thing to both partners. As such, it would be appropriate for entrepreneurship research interested in tie and network development or the influence of ties on firm growth, to understand both actors in a dyad and the process of change within dyads.

In terms of measuring change at the level of the dyad or network, Table 2 suggests that empirical efforts to track how a network develops are relatively rare in entrepreneurship. This is in spite of the foundation arguments offered by Larson and Starr (1993) and Hite and Hesterly (2001). We suggest this provides an opportunity to integrate these two complementary yet competing theoretical views on network process phenomena. Doing so would also offer an understanding of change across levels given Larson and Starr's focus on the dyad while Hite and Hesterly emphasize the network. Further, in spite of recognizing that a network is comprised of component ties, the entrepreneurship literature has only just begun to investigate how relationships are developed and transformed. More specifically, the entrepreneurship literature lacks a rich understanding of when, how, and why ties shift from weak to strong, social to economic, or short-term to long-term (or vice versa). It is the BN approach that could be helpful here, in terms of following development and change within relationships to identify how actors adapt and learn over time, or how changes in dyads affect the network and vice versa. For those interested in assessing specific tie characteristics and their impact on the organization, Uzzi's (1996) approach to integrating depth interviews with either an egonet or full network analysis could be helpful. We also suggest it is important to extend beyond descriptions of "how" and "when" the network or ties change to more fully understand "who" drives the change. Longitudinal case research in the BN style would be useful here, but so too could the SN approach to structural analysis if actor power and position were examined as part of measuring network change over time.

Finally, some entrepreneurship network research has begun to identify what kinds of ties are needed at different stages of firm development. Other research examines the influence of network characteristics on entrepreneurial outcomes. This approach to studying how the network affects performance is consistent with the SN literature, as is discussion on the risks of "overembeddedness." Lacking, however, is rich investigation along the lines of Hite (2005), including macro level research that helps us understand either the general effects of ties on the network or at a more micro level, ties on other ties. That is, are such effects useful and constructive or are they deleterious? These types of questions would be best addressed with the BN approach, which has a tradition of studying the positive and negative aspects of relational change in a time-sensitive manner. We also suggest it is important to understand both the connection between network interactions (i.e., ties), network structure, and performance outcomes, and the dynamics of how these change over time. Entrepreneurship research in this area would be aided by integrating the SN approach by first assessing how interactions lead to network structure and then linking structural changes in (for example) network density or actor centrality to organizational performance. This approach accommodates both the causes of the network and the effects. As such, it takes the holistic view that entrepreneurial and network processes are intertwined with entrepreneurial and network outcomes.

Discussion

In reflecting on the issues raised in the previous section, we suggest that future entrepreneurship research on networks would benefit from: (1) applying multiple theoretical perspectives regarding process, (2) integrating the SN and BN approaches to investigate both the macro level of network structure and the micro level of dyadic interactions, and (3) shifting away from the emphasis on networks as an independent variable to studying them as a dependent variable or perhaps more appropriately stated, a developmental outcome. We expand on this later.

If we refer to the interpretation of process summarized in Table 3, it appears that entrepreneurship researchers have been strongly influenced by the language and tenor of the SN approach. Accordingly, they tend to set research questions necessitating the development of hypotheses where process is viewed as a logic to explain causation or, requiring variable change to be measured. In contrast, the BN approach generally depicts a developmental sequence of events. There is, however, evidence from Tables 1 and 2 of a dual-motor perspective in entrepreneurship and we suggest this hybrid view has the potential to accommodate the complexity of process. What concerns us however is that the entrepreneurship literature lacks the richness that is offered when a study is grounded in teleological, dialectic, or evolutionary theory. We suggest this is a result of not assessing the wider context and forces within which relationships are initiated, developed, and transformed over time, as is more common in BN research. Entrepreneurship research also tends to take a fairly clinical and positivistic approach to understanding network process (similar to SN research) and lacks the realism of the interpretive approach common to BN studies. This emphasis on the structuralist approach is perhaps not surprising if we consider the entrepreneurship field to be relatively young and seeking legitimacy. Similar arguments are made by Borgatti and Foster (2003) as regards network research in general.

One consequence of this is that we tend to view networks through a lens of progression. Reality suggests, however, that relationships and networks involve both progression (forward growth and advancement) and regression (backward movement and deterioration). Another influence on network development is randomness. Unpredictable incidents may occur exogenous to the network in the form of (for example) regulation encouraging (or prohibiting) a certain alliance. Another random incident may be (for example) one firm going bankrupt thus destabilizing a tie and consequently, the network. As a result, it is important to allow for such occurrences when investigating, interpreting, and depicting network process. This might best be captured in a spiral of development that incorporates progression, regression, increases and decreases in network size, as well as change within relationships. Again, this requires theory that recognizes and incorporates teleological, dialectic, or evolutionary arguments.

Turning to the lens through which we view network research, early arguments suggest that theory should include both the structure of the network and the interactions between actors (Burt, 1992; Coleman, 1988; Granovetter, 1985). In spite of this, most studies in entrepreneurship consider either aggregate network patterns (structure) or the ties (interactions) forming the network. To some extent, this dichotomy is captured by the SN and BN literatures. That is, the former focuses on the construction of an efficient and effective network and thus allows for research on the structural dimensions of networks and their impact on firm growth and other outcomes. In contrast, the focus of the BN approach is on understanding development and change within relationships as well as between and across relationships. In some ways therefore, the SN and BN perspectives offer opposing views for network scholars. If, however, we consider these approaches integratively, they offer a useful "bifocal" lens for the entrepreneurship researcher interested in issues of network process. Stated most simply, adopting the BN approach can offer a deep understanding of specific relationships, particularly in terms of assessing interactions and change within a tie, while a macrolevel understanding of structural change and influence in the network can be aided by the SN literature.

Finally, if we are to conceptualize and study networks as a developmental outcome in entrepreneurship, what is the theoretical foundation for moving forward? We agree with Hoang and Antoncic (2003) that entrepreneurship research could benefit from examining Larson and Starr (1993) in a longitudinal study. As we note previously, however, that model is complemented by the more recent efforts of Hite and Hesterly (2001). Accordingly, a first step would be to combine these two arguments into a framework that incorporates both the dyad and network as units of analysis. Additional steps include augmenting this framework with: (1) insight from the empirical findings on network processes, and (2) an understanding of the different ways the meaning of process has been (and might be) applied to network research. Furthermore, a theoretical foundation should apply a combination of the SN and BN perspectives since as noted by Hoang and Antoncic, we need to be able to explain not only the effect of networks but how and why they form and may be managed over time.

This leads us to the theoretical arguments in Figure 1 where the network is positioned as a developmental outcome of a new venture's entrepreneurial process. In Part A, we ask the question "what develops?" and integrate the arguments of Larson and Starr (1993) with Hite and Hesterly (2001). Here, we see that both the new venture and the network develop in a predictive manner that follows a life cycle approach and indeed, the firm and network codevelop (Hite & Hesterly). Of note, network development moves through a life cycle of: (1) variation (new ties emerging blindly or intentionally), (2) selection (ties contributing something and therefore being selected), and then (3) retention (the embedding and transformation of ties discussed by Hite, 2005). This process of tie variation, selection, and retention parallels the arguments of Larson and Starr, and occurs within and through the organizational stages discussed by Hite and Hesterly. Of note, however, we refer to these as "states" in Figure 1, allowing for the possibility of both progression and regression. This is consistent with Hite's finding of the "lack of evolution and presence of de-evolution" (p. 138). We also note that while we portray the V-S-R phases as sequential, they occur simultaneously in practice (Aldrich, 1999) and are influenced by the phases of organizational development.

[FIGURE 1 OMITTED]

Moving to the question of "how and why does the network develop?" (Parts B and C), all of the activities in Part A are influenced by the focal firm or entrepreneur purposefully respecifying goals where s/he initiates change in ties (and therefore the network) to either accommodate or enact the external environment (Hite, 2005; Larson, 1991; Schutjens & Stam, 2003; Steier & Greenwood, 2000). Although the environment may limit entrepreneurial action (Van de Ven & Poole, 1995) and unexpected events may occur (Larson), this strategic adaptation influences tie selection and retention and results in the development of management knowledge. It also reflects the constructive process that is teleology (Part B).

In parallel with this, organizational development efforts will create interactions between multiple entities (e.g., two actors in a dyad or a number of actors in a network) or between the network and the environment. Over time, network members will come and go (Greve & Salaff, 2003; Lorenzoni & Ornati, 1988) and may cause a dialectic opposition between thesis (current path) and antithesis (new entrant) that results in synthesis (Part C). Further, the network will adjust to the cultural and social context as well as market conditions. Another form of dialectic resolution may occur in the interactions between actors to overcome (for example) conflict in goals or implementation strategies. Overall, we see the new venture conform with or deviate from the environment in terms of (for example) the social or economic context in which it operates. Like Part B, this dialectic perspective is constructive rather than prescriptive. Together, Parts B and C extend our view of network development from "what happens" vis a vis the life cycle argument to "how and why it happens" from a teleological and dialectic perspective.

Finally, we ask "what occurs over time?" (Part D). Since the cycle repeats itself during an organizing episode, the evolutionary motor is evidenced with the passage of time through the variation, selection, and retention of certain network configurations over others. Here, the immediate influence of organizing is driven through a life cycle motor as noted by Larson and Starr (1993). It is, however, also influenced by a teleological motor of participant' s choices of adaptations and a dialectic motor of interaction and synthesis. For example, selection activities seek to align the network with the environment over time, but these are punctuated at points where the focal firm shifts direction as a result of purposeful enactment (e.g., entering a new market). This is similar to what is described in Steier and Greenwood (2000) as part of the overall development of the network. Thus, over the longer run, short-term actions contribute to an evolutionary process and what we see is an overall path of development that incorporates change with stability. This is consistent with arguments in the BN perspective and also Aldrich's (1999) view that evolutionary theory comprises a metatheory. That is, it borrows selectively from the other (related) process theories.

In relation to Figure 1, we argue that network change occurs over space and time for the new venture. This implies that several motors can come into play with no single one offering a complete and sufficient explanation of process (Van de Ven & Poole, 1995). We therefore accommodate multiple motors and levels of analysis. The life cycle and evolutionary motors (common to SN research) pertain to the network as a whole, while the teleology and dialectic motors (found in BN studies) pertain to the actions of actors within the network and their dyads. Thus, we connect the macro (SN) and micro (BN) levels that are pertinent to network studies. Since there has been a tendency in the entrepreneurship literature to rely primarily on life cycle theory and to a lesser extent, teleological arguments (both of which focus on single rather than multiple entities), our inclusion of dialectic and evolutionary theory helps balance the overall argument. It also allows us to better appreciate how and why the network develops. The notion of balance is further reflected in the observation that if a theoretical model is overly prescriptive (i.e., it follows the life cycle argument in particular, but also the evolutionary view), it does not allow for sufficient innovation. Similarly, an overemphasis on teleology or the dialectic argument would create too much variety. Consequently, in light of the BN perspective's stance regarding the inherent balance between stability and change in the network, our model accommodates both predictive and constructive views of process.

Inherent in this model is the argument that the essence of entrepreneurship is the entrepreneur (Bygrave, 1993) and human volition. Similar arguments are made by Stevenson and Harmeling (1990) so while we allow for the influence of the environment, change can be made with conscious intent (Weick, 1979). Thus, a strategic adaptation perspective is employed, consistent with Larson and Starr (1993) and Hite and Hesterly (2001). The model also allows for both exogenous and endogenous influences in network development. Although Weick (1979) puts decision makers at the center of organizational development, Aldrich (1999) generally argues that the entrepreneurial process takes on meaning only in the context of the broader social context (e.g., the network) and that environmental selection procedures are determinant. Our model attempts to reconcile these two classic arguments in a manner that allows for both to co-exist. That is, we recognize that the tie coevolves with its social context (the network), and the organization and the network coevolve. Further, both the organization and network coevolve with the environment.

Conclusion

With this article, we have attempted to contribute to the entrepreneurship literature in four ways. First, we extend one aspect of Hoang and Antoncic (2003) to offer a summary of the current state of entrepreneurship knowledge specific to network process issues. Second, we assess this literature to understand which meaning(s) of process are applied by entrepreneurship scholars and to identify how this influences our understanding of network phenomena. A particularly notable finding is the continued dearth of studies focused on the processes associated with network development. Third, we examine how the social and business network literatures view networks and then use this to generate a number of considerations for research. Finally, we offer a theoretical foundation for conceptualizing and studying networks as a developmental outcome. In doing so, we integrate multiple perspectives, levels of analysis, and views of process.

In moving forward we acknowledge certain limitations in our work. As explained early in the paper, we confine our discussion to issues of network process. We also note that our review of studies outside entrepreneurship is limited to the social and business network literatures. We consider these to be particularly informative to our research, but acknowledge that both literatures are richer than what we could capture here. Further, we limit our conceptualization to one that describes process as a developmental event sequence. This, however, complements the extant literature where attention has focused on capturing network processes through causal research or studies measuring variable change.

As presented, Figure 1 incorporates a set of questions to guide researchers. These are reflected in the four components of the model: (Part A) "what develops?", (Parts B and C) "how and why does the network develop?", and (Part D) "what occurs over time?" In combination or as a collective, these components offer a purposefully broad and integrative overview of the ways in which network development can be understood. Consequently, testing it within a single research study would present a considerable undertaking. Indeed, it is not our intent to position this model as one which is testable in its entirety. Rather, we position it as a means to an end. That is, we feel there is clear opportunity for developing a range of more precise models that fall within the umbrella of our general conceptualization. For example, some researchers will be interested in how patterns of tie variation, selection and retention change through the stage of network and new venture codevelopment. Other researchers may investigate why such patterns emerge by studying the degree and nature of purposeful enactment or dialectic tension (or both) as perceived by multiple actors over time. Yet others may focus on understanding how the entrepreneur's actions influence tie creation or dissolution and thus, how the network develops in a teleological manner. Researchers might also assess the evolutionary development of a network by deconstructing it to identify (for example) specific exogenous influences that created a disruption or change in the network. Overall, we suggest that future research should examine parts of our "general" model. They should also treat such parts as pieces of an emerging puzzle whereby adding one piece at a time helps to reveal the nature of the bigger process in question.

Turning to managerial implications, our theoretical arguments suggest that while the entrepreneur engages in purposeful action, these actions are influenced by forces external to the venture or to the network. Further, while the entrepreneur's horizon might be egocentric, their venture operates within a broader system of ties. Following from these points, the entrepreneur should be aware that they are involved with managing in a network rather than management of a network. This requires the entrepreneur to build insight as to the complexity of tie interactions since ties will differ in (for example) their intensity, reciprocity, or impact over time. Similarly, entrepreneurs require an understanding of the overall pattern of the network they operate in. They will also benefit from an appreciation that the network is a dynamic system where ongoing change occurs at different levels: in dyads, across several actors, and within the broader environment.

REFERENCES

Ahuja, G. (2000). Collaboration networks, structural holes, and innovation: A longitudinal study. Administrative Science Quarterly, 45(3), 425-455.

Aldrich, H. (1999). Organizations evolving. London: Sage Publications.

Aldrich, H. & Zimmer, C. (1986). Entrepreneurship through social networks. In D.L. Sexton & R.W. Smilor (Eds.), The art and science of entrepreneurship (pp. 3-23). Cambridge, MA: Ballinger Publishing Company.

Baum, J.A.C., Calabrese, T., & Silverman, B.S. (2000). Don't go it alone: Alliance network composition and startups' performance in Canadian biotechnology. Strategic Management Journal, 21, 267-294.

Berry, F., Brower, R.S., Choi, S.O., Goa, W.X., Jang, H., Kwon, M., et al. (2004). Three traditions of network research: What the public management research agenda can learn from other research communities. Public Administration Review, 64(5), 539-552.

Birley, S. (1985). The role of networks in the entrepreneurial process. Journal of Business Venturing, 1, 107-117.

Borgatti, S. & Foster, P.C. (2003). The network paradigm in organizational research---A review and typology. Journal of Management, 29(6), 991-1013.

Brtiderl, J. & Preisendorfer, P. (1998). Network support and the success of newly founded businesses. Small Business Economics, 10, 213-225.

Burt, R. (1992). Structural holes. Cambridge, MA: Harvard University Press.

Busenitz, L., West, G.P., Shepherd, D., Nelson, T., & Zacharakis, A. (2003). Entrepreneurship research in emergence: Past trends and future direction. Journal of Management, 29(3), 285-308.

Granovetter, M. (1995). Getting a job: A study of contacts and careers (2nd ed.). Chicago: University of Chicago Press.

Granovetter, M.S. (1973). The strength of weak ties. American Journal of Sociology, 78, 1360-1380.

Granovetter, M.S. (1985). Economic action and social structure: The problem with embeddedness. American Journal of Sociology, 91, 481-510.

Greiner, L. (1972). Evolution and revolution as organizations grow. Harvard Business Review, 50(4), 37-46.

Greve, A. (1995). Networks and entrepreneurship--An analysis of social relations, occupational background and use of contacts during the establishment process. Scandinavian Journal of Management, 11(1), 1-24.

Greve, A. & Salaff, J.W. (2003). Social networks and entrepreneurship. Entrepreneurship Theory and Practice, 28(1), 1-22.

Hagedoorn, J. (2006). Understanding the cross-level embeddedness of interfirm partnership formation. Academy of Management Review, 31(3), 670-680.

Hakansson, H., Havila, V., & Pedersen, A.-C. (1999). Learning in networks. Industrial Marketing Management, 28, 443-452.

Hakansson, H. & Snehota, I. (Eds.) (I 995). Developing relationships in business networks. London: Routledge.

Hakansson, H. & Snehota, I. (2006). No business is an island: The network concept of business strategy. Scandinavian Journal of Management, 22, 256-270.

Halinen, A., Salmi, A., & Havila, V. (1999). From dyadic change to changing business networks: An analytical framework. Journal of Management Studies, 36(6), 779-794.

Hansen, E.L. (1995). Entrepreneurial networks and new organizational growth. Entrepreneurship Theory and Practice, 19(4), 7-19.

Hara, G. & Kanai, T. (1994). Entrepreneurial networks across oceans to promote international strategy alliances for small businesses. Journal of Business Venturing, 9, 489-507.

Havnes, P.-A. & Senneseth, K. (2001). A panel study of firm growth among SMEs in networks. Small Business Economics, 16, 293-302.

Hertz, S. (1996). Drifting closer and drifting away in networks: Gradual changes in interdependencies of networks. In D. lacobucci (Ed.), Networks in marketing (pp. 179-204). Thousand Oaks, CA: Sage Publications.

Hertz, S. (1998). Domino effects in international networks. Journal of Business-to-Business Marketing, 5(3), 3-32.

Hite, J.M. (2005). Evolutionary processes and paths of relationally embedded network ties in emerging entrepreneurial firms. Entrepreneurship Theory & Practice, 29(1), 113-144.

Hite, J.M. & Hesterly, W.S. (2001). The evolution of firm networks: From emergence to early growth of the firm. Strategic Management Journal, 22(3), 275-286.

Hoang, H. & Antoncic, B. (2003). Network-based research in entrepreneurship: A critical review. Journal of Business Venturing, 18(2), 165-187.

Human, S.E. & Provan, K.G. (1997). An emergent theory of structure and outcomes in small-finn strategic manufacturing networks. Academy of Management Journal, 40(2), 368-403.

Ibarra, H. (1992). Homophily and differential returns: Sex differences in network structures and access in an advertising firm. Administrative Science Quarterly, 37(3), 422-447.

Ibarra, H., Kilduff, M., & Tsai, W. (2005). Zooming in and out: Connecting individuals and collectivities at the frontiers of organizational network research. Organization Science, 16(4), 359-371.

Jack, S.L. (2005). The role, use and activation of strong and weak network ties: A qualitative analysis. Journal of Management Studies, 42(6), 1231-1259.

Jack, S.L. & Anderson, A.R. (2002). The effects of embeddedness on the entrepreneurial process. Journal of Business Venturing, 17, 467-487.

Jack, S.L., Drakopoulou Dodd, S., & Anderson, A.R. (2004). Social structures and entrepreneurial networks: The strength of strong ties. International Journal of Entrepreneurship and Innovation, 5(2), 107-120.

Johannisson, B. (1988). Business formation--A network approach. Scandinavian Journal of Management, 4(3/4), 83-99.

Johanson, J. & Mattsson, L.-G. (1992). Network positions and strategic action: An analytical framework. In B. Axelsson & G. Easton (Eds.), Industrial networks: A new view of reality (pp. 205-217). London: Routledge.

Johanson, J. & Mattsson, L.-G. (1994). The market-as-networks tradition in Sweden. In G. Laurent, G.L. Lilien, & B. Pras (Eds.), Research traditions in marketing (pp. 321-342). Boston: Kluwer Academic Publishers.

Kilduff, M. & Tsai, W. (2003). Social networks and organizations. London: Sage Publications.

Kim, T.-Y., Oh, H., & Swaminathan, A. (2006). Framing inter-organizational network change--A network inertia perspective. Academy of Management Review, 31(3), 704-720.

Koka, B.R., Madhavan, R., & Prescott, J.E. (2006). The evolution of firm networks: Environmental effects on patterns of network change. Academy of Management Review, 31(3), 721-737.

Larson, A. (1991). Partner networks: Leveraging external ties to improve entrepreneurial performance. Journal of Business Venturing, 6, 173-188.

Larson, A.L. & Starr, J.A. (1993). A network model of organization formation. Entrepreneurship Theory & Practice, 17(2), 5-15.

Lechner, C. & Dowling, M. (2003). Firm networks: External relationships as sources for the growth and competitiveness of entrepreneurial firms. Entrepreneurship & Regional Development, 15, 1-26.

Lechner, C., Dowling, M., & Welpe, I. (2006). Firm networks and firm development: The role of the relational mix. Journal of Business Venturing, 21(4), 514-540.

Liebeskind, J.P., Oliver, A.L., Zucker, L., & Brewer, M. (1996). Social networks, learning, and flexibility: Sourcing scientific knowledge in new biotechnology firms. Organization Science, 7(4), 428-443.

Littunen, H. (2000). Networks and local environmental characteristics in the survival of new firms. Small Business Economics, 15(1), 59-71.

Lorenzoni, G. & Ornati, O.A. (1988). Constellations of firms and new ventures. Journal of Business Venturing, 3, 41-57.

Oliver, A.L. & Liebeskind, J.P. (1997). Three levels of networking for sourcing intellectual capital in biotechnology: Implications for studying inter-organizational networks. International Studies of Management and Organizations, 27(4), 76-103.

Owen-Smith, J., Riccaboni, M., Pammolli, F., & Powell, W.W. (2002). A comparison of U.S. and European university-industry relations in the life sciences. Management Science, 48(1), 24-43.

Podolny, J.M. & Baron, J.N. (1997). Resources and relationships: Social networks and mobility in the workplace. American Sociological Review, 62(5), 673-693.

Powell, W.W., Koput, K.W., & Smith-Doerr, L. (1996). Inter-organizational collaboration and the locus of innovation: Networks of learning in biotechnology. Administrative Science Quarterly, 41(1), 116-145.

Powell, W.W., White, D.R., Koput, K.W., & Owen-Smith, J. (2005). Network dynamics and field evolution: The growth of inter-organizational collaboration in the life sciences. American Journal of Sociology, 110(4), 1132-1205.

Ritter, T. (1999). The networking company: Antecedents for coping with relationships and networks effectively. Industrial Marketing Management, 28(5), 497-506.

Ritter, T., Wilkinson, I.F., & Johnston, W.J. (2004). Managing in complex business networks. Industrial Marketing Management, 33, 175-183.

Schutjens, V. & Stam, E. (2003). The evolution and nature of young firm networks: A longitudinal perspective. Small Business Economics, 21, 115-134.

Seabright, M.A., Levinthal, D.A., & Fichman, M. (1992). Role of individual attachments in the dissolution of interorganizational relationships. Academy of Management Journal, 35(1), 122-160.

Selnes, F. & Sallis, J. (2003). Promoting relationship learning. Journal of Marketing, 67, 80-95.

Simmel, G. (1955). Conflict and the web of group-affiliations. New York: The Free Press.

Smith-Doerr, L. & Powell, W.W. (2005). Networks and economic life. In N.J. Smelser & R. Swedberg (Eds.), Handbook of economic sociology (pp. 379-402). Princeton, NJ: Princeton University Press.

Steier, L. & Greenwood, R. (2000). Entrepreneurship and the evolution of angel financial networks. Organization Studies, 21(1), 163-192.

Stevenson, H. & Harmeling, S. (1990). Entrepreneurial management's need for a more chaotic theory. Journal of Business Venturing, 5(1), 1-14.

Uzzi, B. (1996). The sources and consequences of embeddedness for the economic performance of organizations: The network effect. American Sociological Review, 61, 674-698.

Uzzi, B. (1997). Social structure and competition in interfirm networks--The paradox of embeddedness. Administrative Science Quarterly, 42, 35-67.

Uzzi, B. (1999). Embeddedness in the making of financial capital: How social relations and networks benefit finns seeking financing. American Sociological Review, 64(4), 481-505.

Van de Ven, A. (1992). Suggestions for studying strategy process: A research note. Strategic Management Journal, 13(Special Issue), 169-188.

Van de Ven, A. & Poole, M.S. (1995). Explaining development and change in organizations. Academy of Management Review, 20(3), 510-540.

Watson, J. (2007). Modeling the relationship between networking and firm performance. Journal of Business Venturing, 22(6), 852-874.

Weick, K.E. (1979). The social psychology of organizing (2nd ed.). Reading, MA: Addison Wesley.

Zahra, S.A. (2007). Contextualizing theory building in entrepreneurship research. Journal of Business Venturing, 22, 443-452.

Zhao, L. & Aram, J.D. (1995). Networking and growth of young technology-intensive ventures in China. Journal of Business Venturing, 10, 349-370.

Zimmer, C. & Aldrich, H. (1987). Resource mobilization through ethnic networks: Kinship and friendship ties of shopkeepers in England. Sociological Perspectives, 30, 422-445.

(1.) Tables 1 and 2 include a summary of both: (1) the studies identified in the current research, and (2) those discussed in Hoang and Antoncic (2003). This facilitates the later discussion on how process is interpreted in the extant literature but the current section is focused on the literature not included in Hoang and Antoncic.

(2.) In reviewing the articles discussed by Hoang and Antoncic (2003), it became evident that certain studies presented by them as Category 2 were better placed in Category 1 (e.g., Birley, 1985; Greve, 1995). These were re-classified accordingly. As a result, only Hara and Kanai (1994) remain in Category 2. Further, those studies discussing network process in a tangential manner or outside the context of entrepreneurship were excluded from our review (e.g., Gimeno, Folta, Cooper, & Woo, 1997; Ibarra, 1992). We also excluded conference proceedings unless they have since appeared in a journal (e.g., Davidsson & Honig, 2003).

(3.) Much of the literature in this tradition has been contributed by the Industrial Marketing and Purchasing (IMP) group. They refer to it as the Business Network approach.

Susanna Slotte-Kock is a doctoral student in the Department of Management and Organization, Swedish School of Economics and Business Administration (Hanken), Finland.

Nicole Coviello is Professor of Marketing in the School of Business and Economics, Wilfrid Laurier University, Canada.

The authors thank the Evald and Hilda Nissi Foundation of Finland for their financial support. They also are grateful to various anonymous reviewers for their helpful advice.

NEW PLAYING FIELDS: THE INTERNET IS STILL RAPIDLY EVOLVING AND ENTERPRENEURS CREATIVE WAYS TO KEEP UP

WE'RE LIVING through the mother of all digital convergences. Broadband internet access is becoming ubiquitous, new high-speed wireless networks are coming online, old media are going digital, and new media are being born every day. And the software, hardware and services to support it all are astonishingly cheap.

We're entering a golden age for entrepreneurs, says James Behrens, CEO and co-founder of Orb Networks. "The big companies have not fully recognized how disruptive this is," says Behrens, whose company delivers TV through any device that can stream video. "They're all trying to put up tollgates, but the internet is open."

Orb is among thousands of startups riding this wild wave of advanced internet and mobile services. Here are 10 rapidly evolving communications technologies that offer serious entrepreneurial opportunities--and in most cases, will also help you run your business better.

IPTV:
With more than half of U.S. users accessing the internet via broadband, and broadband delivery speeds climbing, the stage is set for TV over the net. MPEG Nation is one of many new ventures helping other businesses mine this opportunity. For an $8.95 one-time fee, customers can upload a video file up to 1GB and have MPEG Nation convert it to popular streaming formats for ongoing display. The service benefits from the convergence of ever-cheaper storage, high-speed networking and the flexibility to scale those resources on a just-in-time basis, says president Scott Wolf, 23. It was spun off just last year from Digital Silo, which stores and distributes home videos on the web. Together, the two firms have already attracted more than 10,000 customers and are approaching a million-dollar annual run rate, zeroing in on profitability.

Most organizations just use MPEG Nation to display videos on their websites. But others are finding ways to make businesses in new media. For example, one company targets the Indian marketplace in the U.S. by selling subscriptions to videos of Indian performers in concert. Internet TV is rewriting the business model for media. Says Wolf, "You can sell your indie film or build your own TV channel that covers just car mechanics." What are you waiting for?

WIKIS:
Think of a wiki as a collaborative document on the web. Wiki server software lets participants freely edit web page content with any browser. Ward Cunningham whipped up the first wiki by himself in 1995, and an estimated 1 million wikis are now collaborating away. They deliver a constantly refreshed view of a group's collective wisdom on a given topic without a confusing flood of e-mail. Wikis are quietly cropping up in businesses that tap them for internal projects and have spun offa cottage industry of software and service providers, such as lotSpot and Socialtext.

BLOGGING:
Blogs are a recent example of how relatively simple web technology can create a huge ripple. We've seen the blogosphere produce huge public mood swings during elections, and some firms have found them an excellent way to communicate with customers, partners and suppliers. They're great for project management and tech support, and a few of the most popular bloggers are even charging for blog access. Blogs depend heavily on ...

RSS:
Share nicely, as the day-care teachers say. That's what RSS does--it lets you instantly swap links of updated information across the web. You can tap RSS readers or sites (such as personalized home pages on Google or Yahoo!) to customize your news access. RSS also lets you spread your own message--your blog or website can use it to distribute information about your company. Down the road, some firms will even sell ads along with their RSS feeds--if it can be done without annoying recipients. One new medium RSS begat is ...

PODCASTING:
Audio downloading isn't new, but podcasting makes it simple. Your grateful public uses free podcasting software such as Apple's iTunes or Juice Receiver to download your MP3 voice or music files, which they listen to on digital music players or PCs at their leisure. A fine way to distribute customer information or training, it's also spawning homegrown radio stations.

GPS PHONES:
GPS-equipped phones aren't just for keeping track of teens and getting driving directions anymore. They can help you cut sales and supply costs, too. Right now, you probably don't know exactly where all your sales and delivery people--or the company vehicles they're using--are. But their cell phones do. And with the right phones--like some Nextel models--and the right software like Xora's--you can get powerful new options for dispatching, time and route tracking, staff and delivery scheduling, and emergency response. Down the road, GPS phones will be used to report mobile equipment diagnostics or just to let colleagues know when you're running late for a meeting.

INSTANT MESSAGING:
Like so many new forms of media. IM first achieved superstar status among consumers. But now the world's largest tech company couldn't run without it. "There are about 3.5 million instant messages a day within IBM," reports David Marshak, Big Blue's senior product manager for real-time and activity-centric collaboration "But the important number is zero--I get zero voice mails from co-workers."

"You use IM for problems you want to solve right now. and the entrepreneurial world is in no way slower than the corporate world," Marshak notes. Additionally, IM suppliers are now tailoring their products for small business and bundling an array of communications alternatives, one of which is ...

UBIQUITOUS TELEPHONY:
Here's the deal: Talk for free to anyone anywhere with Skype and an internet connection. Or for a couple of pennies a minute, you can call anyone with a conventional phone via the SkypeOut service. Voice quality is very good, and you can get every telephony feature conceived by man, from local phone numbers to answering services. What's not to like?

HANDS-0N VIDEO:
Apple created a sensation last fall when it started podcasting TV shows to their new video iPod. But now, both live and prerecorded programs are appearing on the most popular electronic device in history--our beloved cell phones. Only time will tell when the tiny screen is ready for full-blown TV. But there's certainly no shortage of big-ticket deals being cut or new TV-friendly devices being released.

Digital Video Broadcast Handheld technology, which is optimized for streaming video to energy-efficient handhelds like Nokia's new bigger-screen phones, appears ready for prime time. Video-enabled phones will be arriving in the tens of millions in the next couple of years, predicts Sam Leinhardt, CEO and co-founder of Penthera Technologies, which creates software for DVB-H networks. New technology always brings new opportunities, says Leinhardt, but trying to predict them "is like asking me 15 years ago, How can the internet be used?"

NO CHILD LEFT OFF THE INTERNET:
Maybe you laughed when you heard the MIT Media Lab was creating a $100 laptop for schoolchildren, but MIT has governments worldwide intrigued by its "laptop in every home" idea and will begin manufacturing them soon. Media Lab prototypes sport innovative displays, free Linux software that runs in flash memory, and peer-to-peer Wi-Fi networks. What if tens of millions of previously disenfranchised children wind up with inter net access? What a market that would be. "Nobody's ever installed computers this densely," notes Michael Bove, director of MIT's Consumer Electronics Lab. "Get them out in sufficient numbers, and the business case will be pretty obvious."

WALK THIS WAY:
What Sony once did for radios, it now wants to do for phones. Push a button on the side of the Sony Ericsson W800i Walkman Phone, and it will play your MP3s. Push twice to listen to MP3 players. There are the usual radio tuning features to navigate to your favorite radio station, and a 512MB Memory Stick will bolster W800i's internal 34 MB storage. Flip it Sideways to get a great 2-megapixel camera with shutter button and 4x digital zoom that feels like a stand-alone digital camera. Oh, and W800i makes phone calls, too.--Steve Cooper