International alliances offer firms opportunities to draw upon knowledge and capabilities not currently controlled or available within their home country (OECD, 2000). Among other benefits, this can help firms share costs, enter new markets (Glaister and Buckley, 1996), supplement their capabilities (Inkpen and Dinur, 1998; Hitt et al., 2000; Lane et al., 2001), seek more radical innovations by integrating knowledge from different areas of science and technology (Lubatkin et al., 2001; Nummela, 2003), and create common platforms for products and services (Mowery et al., 1998; Caloghirou et al., 2003). However, international alliances also bring challenges not found within domestic alliances. Research has shown that differences in national culture can disrupt collaboration and learning between alliance partners (Lane and Beamish, 1990; Parkhe, 1991; Lyles and Salk, 1996; Hennart and Zeng, 2002). However, the conclusion that national culture differences alone disrupt knowledge sharing between partners recently has been questioned. Pothukuchi et al. (2002) suggested that the importance accorded to national culture differences on international alliance performance may be overstated because most studies have failed to consider or specify the influence of organizational culture differences as well. Whereas national culture relates primarily to deep-seated values, organizational culture relates primarily to shared beliefs in organizational practices and processes (Hofstede et al., 1990). Examining a large sample of This content downloaded from 139.182.75.138 on Thu, 25 Apr 2019 06: A model of cultural differences David G Sirmon and Peter I Lane 307 international joint ventures, Pothukuchi et al. (2002, p. 258) found that ‘the presumed negative effect from partner dissimilarity on IJV perfor- mance originates more from differences in organi- zational culture than from differences in national culture’. Pothukuchi et al.’s (2002) findings suggest that a more generalized effect may exist: partners’ cultural differences may have more influence on interna- tional alliance performance as those differences become more directly related to the alliance’s primary value-creating activities. This paper builds on that inference by developing a theoretical model of the performance effects of cultural differences in international alliances. International alliances are the focus of the paper because they ‘reside at the confluence of different cultures, including national, corporate and occupational’ (Salk and Shenkar, 2001, 163). We argue that an international alliance’s perfor- mance is driven by the alliance’s effectiveness in achieving its primary value-creating activities. Resource complementarity between alliance part- ners is often a necessary condition to optimize this value creation (Harrison et al., 2001). However, in order to share, combine and leverage complemen- tary resources, which include both tangible and intangible assets but not business processes or ‘the actions that firms engage in to accomplish some business purpose or objective’ (Ray et al., 2004, 24), the partners’ employees must interact effectively. Research suggests that national and organizational culture differences between these employees affect their interactions, but this study expands the consideration of cultural differences to include professional culture differences. Professional cul- tures form as people, who span individual organi- zations, share a set of norms, values and beliefs related to their occupation (Van Maanen and Barley, 1984; Jordan, 1990; Trice and Beyer, 1993). The essence of our argument is that cultural differences stemming from national, organizational and professional cultures inhibit international alliance partners’ employees’ ability to interact effectively. Additionally, we argue that professional culture differences are often the most relevant and Differences in Partners’ National Cultures P2 + P3 + Differences in Partners’ Organizational Cultures P4 Differences in Partners’ Professional Cultures P5 Related- PI Effectiveness of Performance Complementary Alliance’s = = of the Resources Value-creating International Activities Alliance a The thickness of the arrows indicates the relative strength of the relationships (Proposition 6). Figure 1 A model of cultural differences and international alliance performance. The thickness of the arrows indicates the relative strength of the relationships (Proposition 6). journal of International Business Studies This content downloaded from 139.182.75.138 on Thu, 25 Apr 2019 06:55:22 UTC All use subject to https://about.jstor.org/terms A model of cultural differences David G Sirmon and Peter J Lane 308 salient cultural differences that the interacting employees face, and thus professional culture differences are the most disruptive to the alliance’s effectiveness in achieving its primary value-creating activities. See Figure 1 The model and propositions developed in this paper make several contributions. First, we offer a more detailed theory of how national culture differences influence international alliance perfor- mance. We suggest that most of the influence of national culture is mediated by organizational and professional culture differences. Second, we pro- pose that the negative influence of cultural differ- ences increases as those differences become more relevant to the alliance’s value-creating activities. Third, although we have proposed a general model applicable to all types of international alliance, it is especially relevant to international R&D alliances (IRDA), which include both R&D joint ventures and joint R&D agreements (Hagedoorn and Narula, 1996). The likelihood of finding professional cul- ture differences between R&D personnel, and the increasing importance of IRDA to a wide range of companies (Narula, 1999; Narula and Hagedoorn, 1999; Hagedoorn, 2002; Caloghirou et al., 2003; Nakamura, 2003), suggest that firms engaging in them need to be as sensitive to professional culture differences in the laboratory as they are to national culture differences across the conference table. Effectiveness Of Alliances’ Value-Creating Activities Because international alliances differ in their primary value-creating activities, the specification of a metric to measure performance is problematic (Geringer and Hebert, 1989). For example, the primary value-creating activity of a marketing- based alliance may involve the delivery of a strong co-branded messages, whereas for R&D-based alliances it is often the development of innovations or ‘scientific or technical knowledge and the application of that knowledge to the creation of new and improved products or processes’ (Hage- doorn, 2002, 477). Although these activities may affect profitability, it is often difficult to detect such effects. Therefore we suggest that the effectiveness with which the alliance achieves its primary value- creating activities is a precursor of alliance perfor- mance, but that such primary value-creating activ- ities are specific to the purposes of specific international alliances. The literature suggests that the pooling of complementary resources – those that are not identical, yet are mutually supportive – is often necessary for alliances to optimize their primary value-creating activities (Lane and Lubatkin, 1998; Das and Teng, 2000; Harrison et al., 2001). Com- plementary resources allow an alliance to pursue economies of scope and growth opportunities, while decreasing risk (Lane and Lubatkin, 1998; Das and Teng, 2000; Harrison et al., 2001). For example, Lu and Beamish (2001) found that small and medium-sized enterprises experienced better performance when they internationalized by part- nering with firms that possessed local knowledge relevant to the newly entered region. The local knowledge of the partner was a complementary resource for the internationalizing firm. Further, Vidal-Suarez (2002) found that the stock market awarded Spanish firms a 0.2% abnormal return the day they announced global alliances with partners that held complementary resources. However, the presence of complementary resources alone does not always lead to positive results for international alliances. Caloghirou et al. (2003) found no support for the expected positive relationship between complementary resources and international research alliance performance, as measured by the partners’ perceptions of success. Further, Johnson et al. (1996) found that comple- mentary resources led US firms to trust their Japanese partners more but not vice versa. Similarly, Sarkar et al. (2001) found that the presence of complementary resources did not positively affect trust or bilateral information exchange between partners. Further, these authors found that com- plementary resources had a positive effect on project outcomes, but not on strategic outcomes. And, most interestingly, Sarkar et al. (2001) found that partner compatibility was critical in determin- ing the effects of complementary resources on alliance performance. Model misspecification could be the cause of such inconsistent results. First, not all the resources of the partners are relevant to a given alliance’s primary value-creating activities (Das and Teng, 2000). Thus the benefits of pooling complementary resources should positively affect alliance perfor- mance only when those resources are related to the primary value-creating activities of the alliance. For example, marketing alliances might be concerned with the complementarity of delivery systems, compensation practices, and brand or firm reputa- tion, whereas R&D alliances would be concerned with the complementarity of research facilities and research orientations (Caloghirou et al., 2003). Journal of International Business Studies This content downloaded from 139.182.75.138 on Thu, 25 Apr 2019 06:55:22 UTC All use subject to https://about.jstor.org/terms A model of cultural differences David G Sirmon and Peter J Lane 309 Second, consideration of the compatibility of the partners’ employees who actually interact in the primary value-creating activities of the alliance may be underdeveloped. Madhok and Tallman (1998) argued that many alliances fail to achieve their goals in part because the partners underestimate the difficulty of working together. One common cause of this shortcoming is a failure to attend to the differences in how each firm approaches the processes involved in the primary value-creating activities of the alliance. Thus, as the implication of Pothukuchi et al.’s (2002) finding suggests, increased specificity in the relational compatibility of partners’ employees should lead to increased understanding of international alliance success. More formally, the preceding arguments lead to the following proposition: P1: The complementarity of partners’ resources positively affects alliance performance only when those resources are related to the primary value- creating activities of the international alliance, and when the employees involved in those activities interact effectively. The interaction of the employees involved in the primary value-creating activity of the alliance is likely to be driven by sensemaking. Sensemaking is the process of placing stimuli into categories in order ‘to comprehend, understand, explain, attri- bute, extrapolate, and predict’ (Starbuck and Milli- ken, 1988, 51). Because sensemaking is shaped by personal experiences, the interacting employees can differ in their categorization and linking of stimuli. In the words of Weick (1995, 62): What I say and single out and conclude are determined by who socialized me and how I was socialized, as well as the audience I anticipate will audit the conclusions I will reach. Thus, assuming that the appropriate resources are made available, understanding why some interna- tional alliances fail to create value effectively requires the exploration of systematic differences in the partners’ employees’ socialization. Three sources of such socialization are explored in this paper: one’s nation, one’s organization and one’s profession. We address each in sequence. National culture As stated previously, national culture refers to deeply set values that are common to the members of a nation (Hofstede, 1991; Hill, 1997). It is a system of shared norms, values, and priorities that, taken together, constitute a ‘design for living’ for a people (Hill, 1997, 67). Importantly, national culture is learned, and provides meaning to ‘how things ought to be’ and ‘how things ought to be done’ for individuals in a country (Berger and Luckmann, 1967; Terpstra and David, 1991). These shared beliefs are acquired early in life through a person’s primary socializing in families, in schools and at play (Berger and Luckmann, 1967; Terpstra and David, 1991). Further, the influence of national culture is strong and long lasting. For example, Hofstede (1991) found that national culture explains 50% of the differences in managers’ attitudes, beliefs, and values, and Laurent (1983) found that managers of multinational organiza- tions retain many of their original national values despite routinely working in culturally diverse situations. No nation is expected to have a completely homogeneous national culture, but national differ- ences are clearly seen in economic and political systems (Albert, 1991; Thurow, 1993), educational systems (Calori et al., 1997), and other institutions (DiMaggio and Powell, 1983; Hall, 1986; Clegg and Redding, 1990; Sorge and Maurice, 1990). Thus, national culture differences between alliance part- ners can challenge the development of successful relationships. These challenges stem partially from the lack of shared norms or values (Park and Ungson, 1997). This lack of common understanding may under- mine the partners’ interpretation of each other’s strategic intent, which is crucial in global markets and partnerships (Hitt et al., 1995). Further, a lack of shared norms and values may reduce effective communication (Rao and Schmidt, 1998), trust (Aulakh et al., 1996; Doney et al., 1998) and knowledge sharing in joint ventures (Parkhe, 1991; Mohr and Spekman, 1994; Lyles and Salk, 1996). These problems, in turn, have been found to lead to lower alliance performance (Lane et al., 2001). In one noteworthy study, Barkema and Vermeu- len (1997) examined the influence of differences in partners’ national cultures on international alliance performance using Hofstede’s (1980, 1991) dimen- sions of national culture. They found that partner differences in two of the dimensions (uncertainty avoidance and long-term orientation) had a strong negative r A model of cultural differences David G Sirmon and Peter J Lane 310 long-term orientation could represent differences in how partners perceive and adapt to opportu- nities and threats in their environment (Schneider, 1991; Schneider and De Meyer, 1991), and thus may be more difficult to resolve than differences along the other three dimensions, which represent attitudes towards personnel. However, other evidence suggests that differences in national culture can be beneficial. Because managers tend to be more aware of the potential challenges when working with foreign partners, they may be more willing to spend effort on avoiding misunderstandings in international alli- ances than they would in domestic alliances (Very et al., 1996). In such cases, differences in national culture can lead to high-level communication and a more sustained collaboration (Shenkar and Zeira, 1992; Park and Ungson, 1997). Thus, in some cases, increased national culture differences can lead to higher international alliance performance (Morosi- ni et al., 1998). These contradictory results suggest that differ- ences in national culture have a more complex relationship with international alliance perfor- mance than has been previously proposed (Salk and Shenkar, 2001). We argue that this complex relationship includes the influence of national culture on both organizational and professional cultures (Terpstra and David, 1991, chapter 1), which then in turn influence international alliance performance by allowing for more effective sharing, combining and leveraging of the alliance’s com- plementary resources. Because the beliefs and norms of national culture are learned early in life, and endure despite subsequent socialization by organizations and occupations, national variation within organiza- tional and professional cultures is likely. First, differences in national culture systematically influ- ence organizational cultures through the firm’s administrative heritage – the historical manage- ment practices that have been used by organiza- tions within a nation (Bartlett and Ghoshal, 1989; Calori et al., 1997). A nation’s commonly shared administrative practices provide constraints on the actions of firms within that nation. However, national culture does not determine a monolithic organizational culture for all firms within a coun- try; instead these firms develop organizational cultures that vary around certain broad assump- tions (Bartlett and Ghoshal, 1989; Calori et al., 1997). Therefore, when international alliance part- ners collaborate, differences in their national cultures can manifest themselves as differences in the partners’ organizational cultures. Second, theory suggests and evidence confirms that national culture is associated with attitudes that affect professional activities. Extending Hof- stede’s (1980) arguments, Shane (1992, 1993) found that nations with low power distance, weak uncertainty avoidance, and high individualism had higher rates of innovation. Additionally, Kedia et al. (1992) found that R&D laboratory productiv- ity was positively associated with low power distance and high masculinity in both academic and industrial research settings. Lastly, Shane et al. (1995) found that the preferred behaviors and strategies for championing innovations within a firm varied considerably with national culture. In short, professionals in the same functional domain but from different countries can systematically vary in their attitudes toward and implementation of the preferred solutions for the problems that those in their occupation face. Therefore national culture differences lead to differences in professional cultures. Specifically, differences in national culture will increase the differences of partners’ employees’ professional cultures. These arguments lead to the following two propositions: P2: Differences in international alliance partners’ national culture increase those partners’ organi- zational culture differences. P3: Differences in international alliance partners’ national culture increase the differences of partners’ employees’ professional cultures. Organizational culture Consensus has not been reached with regard to the definition of organizational culture. A review by Verbeke et al. (1998) found 54 different definitions of organizational culture between 1960 and 1993. However, these definitions of organizational cul- ture revolve around shared group meaning (Hof- stede et al, 1990; Golden, 1992; Ostroff and Kinicki, 2001). For this study we follow Smircich’s, (1983, 344) definition of organizational culture, which is: social or normative glue that holds an organization together … it expresses the values or social ideals and the beliefs that organization member come to share. Thus organizational culture forms a type of social control that identifies appropriate behaviors and attitudes for organization members to display (O’Reilly and Chatman, 1996). Further, organizational journal of International Business Studies This content downloaded from 139.182.75.138 on Thu, 25 Apr 2019 06:55:22 UTC All use subject to https://about.jstor.org/terms A model of cultural differences David G Sirmon and Peter J Lane 311 culture provides more proximal cues for organiza- tion members’ behavior than does national culture because it provides members with an organiza- tional identity, and facilitates collective commit- ments (Smircich, 1983). However, it does not completely supersede or displace the influence of one’s national culture (Hofstede et al., 1990). There has been less research on organizational culture and international alliance performance than on national culture and international alliance performance. This is surprising, as people from the same national background differentiate their work practices as a result of the influence of hetero- geneous organizational socialization (Hofstede et al., 1990). However, the results of those studies strongly suggest that differences in partners’ orga- nizational culture matter greatly to the outcomes of international alliances. Based on a study of alliances between Japanese and Western firms, Brown et al. (1988) concluded that differences in partners’ organizational cultures can have a significant negative influence on IJV performance. Such differences can even lead to the dissolution of alliances as partners are forced to divert attention and energy to develop interaction routines aimed at overcoming these differences (Park and Ungson, 1997). Pothukuchi et al. (2002) found that organizational culture dissimilarity negatively affected international partners’ satisfac- tion with the relationship. Additionally, studies of domestic alliances show a similar relationship between shared organizational culture and performance outcomes. In their study of US R&D alliances, Lane and Lubatkin (1998) argued that multiple manifestations of organiza- tional culture would affect inter-organizational learning. These authors found that partner similar- ity in research centralization and research compen- sation practices affected learning outcomes positively. Further, organizational culture has been found to be important to the success of mergers and acquisitions. Weber et al. (1996) found that dissimilar organizational cultures between acquirer and target decreased top man- agers’ cooperation and increased negative attitudes toward the merger. Overall, similarity of partners’ organizational culture increases partner learning, satisfaction and effectiveness of interactions, whereas differences in organizational culture decrease these positive out- comes. In short, decreased learning, satisfaction and effectiveness of interactions are expected to inhibit the business processes used to share, combine and leverage resources such as knowledge, relationships and physical assets. Thus partners with dissimilar organizational cultures will be less likely to effectively achieve the alliance’s primary value-creating activities, even when the necessary resources are present: P4: Differences in the organizational cultures of international alliance partners negatively moder- ate the relationship between related complemen- tary resources and effectiveness of the alliance’s value-creating activities. Professional culture Professional culture is another important type of culture that can affect international alliances, but it has received even less attention from scholars than organizational culture. A professional culture exists when a group of people who are employed in a functionally similar occupation share a set of norms, values and beliefs related to that occupa- tion. Professional cultures develop through the socialization that individuals receive during their occupational education and training (Van Maanen and Barley, 1984; Jordan, 1990). This initial socia- lization is then reinforced through their profes- sional experiences and interactions that lead to a broad understanding of how their occupation should be conducted (Brown and Duguid, 1991; Lave and Wenger, 1991). The strength and pervasiveness of occupational socialization that leads to professional cultures should not be understated. Trice and Beyer (1993, 178) argue that individuals’ occupations are ‘[t]he most highly organized, distinctive, and pervasive sources of subcultures in work organizations’. Members of a professional culture share a ‘con- sciousness of kind’ that not only encompasses functional understanding, but also includes the sharing of emotional demands and a social identity related to the occupation (Salaman, 1974; Van Maanen and Barley, 1984). Therefore most func- tional areas in an organization (e.g. accounting, marketing, engineering, operations, sales and science-based groups) are expected to have distinct professional cultures; however, we argue that – in fact – they are not organizational subcultures. Organizational subcultures develop within a firm and are based on any number of different types of shared experience, such as educational, regional, and ethnic backgrounds (Arnold, 1970), similar problems (Cohen, 1955), or occupations (Van Maanen and Barley, 1984; Trice and Beyer, 1993). journal of International Business Studies This content downloaded from 139.182.75.138 on Thu, 25 Apr 2019 06:55:22 UTC All use subject to https://about.jstor.org/terms A model of cultural differences David G Sirmon and Peter I Lane 312 However, Cohen (1972) describes a subculture as a solution to a group’s need to belong to a larger membership, while also possessing some level of independence from that membership. Professional cultures, as utilized in this paper, develop outside any single organization. The initial source of socialization for a professional culture is the educational system where such groups of profes- sionals are trained. For example, Terpstra and David (1991, 12) point out: ‘One could argue that professional schools spend more time imparting professional culture than teaching technical skills or knowledge.’ Because professional cultures do not fit within any single organizational culture, we do not consider them an organizational subculture. Instead, professional cultures cut across organiza- tional boundaries. When international alliance partners require employees from different professional cultures to interface in the primary value-creating activity of the alliance, the results are expected to be dis- appointing. Overall, these individuals lack a com- mon basis from which to interact effectively because their occupational socialization and result- ing professional cultures are distinct. First, indivi- duals from separate professional cultures lack a shared set of basic knowledge because their occupa- tional socialization involved different content material, which is reinforced by different profes- sional experiences. Second, these individuals often lack experience communicating with an auditing audience outside their professional culture. Thus communication between individuals from separate professional cultures is impaired. Both of these factors inhibit the finding of common ground from which the relationship can develop and produce value (Lane and Lubatkin, 1998). In such cases, energy must be diverted from the value-creating activity of the alliance to the development of basic routines that help establish a base of shared knowledge in order to communicate adequately. Developing such routines requires time, which leads to increased expenses and could lead to increased frustration (Park and Ungson, 1997). Even if these two obstacles can be adequately overcome, individuals from different professional cultures may still have deeply ingrained preferences in their approach to solving problems (Brown and Duguid, 1991; Lave and Wenger, 1991). These differences may be difficult to overcome, as the employees may reflect the ‘not-invented-here’ syndrome, which is the resistance to the utilization of knowledge created elsewhere (Michailova and Husted, 2003). Further, attempts to compromise in the approach taken in problem-solving is likely to lead to less desirable outcomes. For example, if a compromise is reached, and members from both professional cultures abandon their preferred pro- blem-solving approach, both effectively eliminate a significant amount of their valuable tacit knowl- edge. Likewise, if either member abandons their preferred problem-solving approach, the alliance has effectively lost the expertise of one half of its contributing members. The challenges discussed above inhibit the effec- tive interaction of individuals from different profes- sional cultures within an international alliance. This then decreases the likelihood that the alliance’s pooled complementary resources will be shared, combined and leveraged in a manner that effectively achieves the alliance’s primary value-creating activities. In order to illustrate the proposed relation- ship, we present a detailed example of two profes- sional cultures often involved in IRDAs. IRDAs are an important form of collaboration between international partners, and are increasing in number (for reviews see Narula and Hagedoorn, 1999; Hagedoorn, 2002). By R&D we imply standard research and development activity devoted to increasing scientific or technical knowledge and the application of that knowledge to the creation of new and improved products and processes. (Hagedoorn, 2002, 477) Although not perfect, we stipulate that an IRDA’s primary value-creating activity is found in new innovations and learning. Increases in a firm’s rate of innovation highlight the short-term benefits of an IRDA, whereas learning captures the potential long-term gains of the relationship (Sarkar et al., 2001), even when the actual relationship is short, as is often the case (Narula and Hagedoorn, 1999). R&D personnel in an IRDA are drawn predomi- nantly from two professional groups – scientists and engineers. Although these two groups are often treated as interchangeable, research on the sociol- ogy of knowledge and cognitive models of innova- tion has shown that they represent two distinct professional cultures with different norms, beliefs, values and priorities (De Solla Price, 1982, 1986; Mikhailov et al, 1984; Nightingale, 1998; Balmer and Sharp, 1993). Specifically, the professional cultures of scientists and engineers differ in the types of knowledge they emphasize, in the ways in which knowledge develops, and in the types of innovation process they utilize. Table 1 highlights these differences. journal of International Business Studies This content downloaded from 139.182.75.138 on Thu, 25 Apr 2019 06:55:22 UTC All use subject to https://about.jstor.org/terms A model of cultural differences David G Sirmon and Peter J Lane 313 Table 1 Engineering vs scientific professional cultures Type of professional Locus of critical knowledge Flow of critical knowledge Innovation process culture characteristics Engineering Internal: technology Controlled: economic Problem driven communities property rights Strict timelines Market criteria Scientific External: science Open: few property rights Curiosity driven communities Lax timelines Curiosity criteria Scientific vs engineering professional cultures: knowledge types These professional cultures differ in the types of knowledge with which they are primarily con- cerned. The scientific culture emphasizes knowl- edge concerning general theories about relationships associated with physical, biological, and social phenomena, as well as the methods for testing them (Burgelman et al., 1996). This defini- tion encompasses both basic scientific research into fundamental relationships in nature and applied scientific research into fundamental relationships within and between classes of technology. By contrast, the engineering culture emphasizes tech- nology, or the theoretical and practical knowledge, skills and artifacts that can be used to develop products and services as well as their production and delivery systems. (Burgelman et al., 1996, 2) Although organizational researchers often treat science and technology as interchangeable (e.g. Porter, 1985; Latour, 1987), they are in fact fundamentally different, especially in their rela- tionship to marketable products and services. Scientific knowledge can be used to create new scientific knowledge and new technological knowl- edge, whereas technological knowledge is used to create new technological knowledge, processes, products or services. In other words, technological knowledge focuses on market needs and trajec- tories, whereas scientific knowledge focuses on more fundamental issues. Scientific vs engineering professional cultures: knowledge development Scientific and engineering cultures also differ in the processes by which they typically create knowledge (Mikhailov et al., 1984, 111-120). Scientific knowl- edge is usually developed via networks of scientists in different organizations, industries, and types of institutions, bound by a common interest in a broadly defined problem (Kuhn, 1970; Crane, 1972; Latour, 1987; Tushman and Rosenkopf, 1992). Members of scientific networks are not typically rewarded for the economic value of their research; instead they are rewarded for their ‘contribution’ to the shared understanding of the phenomena (Mulkay, 1977; de Solla Price, 1982, 1986). The uncompensated ‘contribution’ of their research creates a reciprocal right to draw on the pool of knowledge ‘donated’ by other network members (Hagstrom, 1982). Such reciprocal rights are critical to the progress of science, as few scientists can build a stream of research independent of others. The shared, open and accessible nature of scientific knowledge fits with its focus on pursuing funda- mental questions. By contrast, the engineering culture tends to create technological knowledge via technological networks composed of members located in one organization or in economically linked organiza- tions. These individuals’ rewards are often based on the economic value of the technological knowledge they create (Mulkay, 1977; Barnes and Edge, 1982). They may draw on external sources of expertise, but the knowledge created by technological networks is developed within economic organizations and is controlled by them (Tushman and Rosenkopf, 1992). The proprietary nature of technological knowledge fits with its focus on market needs and trajectories. Scientific vs engineering professional cultures: innovation process The professional culture of scientists tends to follow loose timetables, primarily utilizes abstract infor- journal of International Business Studies This content downloaded from 139.182.75.138 on Thu, 25 Apr 2019 06:55:22 UTC All use subject to https://about.jstor.org/terms A model of cultural differences David G Sirmon and Peter J Lane 314 mation, and values unexpected findings. New scientific knowledge often changes incrementally. New theories and discoveries absorb the knowledge that has preceded them, and are positioned relative to the state of the art. Radical new scientific perspectives emerge less frequently, but they typi- cally result in a re-conceptualization of prior findings, rather than their outright disregard (Kuhn, 1970). However, the professional culture of scientists is not based solely on a body of knowledge. Instead, members are socialized to follow agreed methods for gathering and communicating evidence via individual training (ie education) and the publica- tion process (e.g. peer review) (Gilbert and Mulkay, 1984; Latour, 1987). However, these agreed stan- dards change over time, and thus the process of scientific knowledge creation is dynamic (Master- man, 1970; Gilbert and Mulkay, 1984). The net effect of this dynamic process is that scientific knowledge advances through the accumulation of repeated ‘branchings’ and ‘spiralings’ (de Solla Price, 1986). By contrast, technology development processes are typically driven by pressures from markets (Clark, 1987; Balmer and Sharp, 1993; Adner and Levinthal, 2001), and begin with ideas of what is needed to satisfy those pressures. Rewards for quick responses to such market pressures (e.g. first mover advantage) often lead firms to perform local searches for solutions. In other words, technologi- cal knowledge is often produced by extending what is already well known by the firm. Thus technolo- gical knowledge is essentially about exploiting and adapting what is known in order to achieve what is desired (Rip, 1992; Garud and Rappa, 1994). When a solution to a technological need is not apparent, the firm works backwards from its preconceived ends and evaluates potential starting points (solu- tions) until an optimal one is found (Nightingale, 1998). However, in either case, social, political, and organizational dynamics play a large role in determining which technological options are pur- sued (Tushman and Rosenkopf, 1992). Further, current technological knowledge often makes prior technological knowledge obsolete. In compara- tively few cycles of innovation, much of the earlier technological knowledge has been discarded. Thus, unlike the development of science-based knowl- edge, the technological knowledge development process tends to follow set timetables, utilizes concrete information, views unexpected findings as useless or even troubling, and develops in short linear chains (Allen, 1984, 1997; Mikhailov et al., 1984, 111-120; de Solla Price, 1986). Scientific vs engineering professional cultures: an example Because the professional cultures of scientists and engineers represent such different sets of beliefs, norms and priorities, they can create severe com- munication problems and conflict between the groups. For example, Balmer and Sharp (1993) suggested that the differences between science and engineering led to the 1980s conflict in the UK over which research council would ‘shepherd’ biotechnology research. The Medical Research Council (MRC) had a scientific culture favoring broadly defined, science-driven research programs and ‘curiosity-led funding’. The engineering culture of the Science and Engineering Research Council (SERC), on the other hand, placed more emphasis on industry relevance and problem-solving. It routinely shifted priorities and resources in accor- dance with new developments, and was thus able to quickly set up the Biotechnology Directorate that funded research, meeting the industry’s needs. The MRC objected, and the debate that followed resulted in an inquiry by the House of Lords. During this inquiry, the MRC condemned the SERC’s biotechnology program as ‘opportunistic’ and ‘freewheeling’, interested only in narrowly focused ‘industrially relevant projects’ and in funding too much second-rate science. The SERC, in turn, called its Biotechnology Directorate ‘one of the jewels in our crown’. In total, this example demonstrates the potential for significant conflict between members of differ- ent professional cultures. However, the potential for conflict does not reside only in the interactions of science and engineering professional cultures. Instead, differences between any professional cul- tures represented in the value-creating activity of an international alliance can have significant negative ramifications. Specifically, we argue that differences in the professional culture of the employees who interact in an international alliance will inhibit the effective sharing, combining and leveraging of complementary resources. Formally: P5: Differences in the professional cultures of international alliance partners’ employees nega- tively moderate the relationship between related complementary resources and effectiveness of the alliance’s value-creating activities. Journal of International Business Studies This content downloaded from 139.182.75.138 on Thu, 25 Apr 2019 06:55:22 UTC All use subject to https://about.jstor.org/terms A model of cultural differences David G Sirmon and Peter I Lane 315 Thus far, we have discussed separately the effects of cultural differences in international alliances as they originate from national, organizational and professional cultures. We began with the contra- dictory results of the effects of national culture differences on international alliance performance, and then proceeded to discuss both organizational and professional cultures as they pertain to inter- national alliance performance. However, as Salk and Shenkar (2001) showed, international joint venture team members utilize several different social identities (national, organizational and occu- pational) as sensemaking mechanisms. Thus indi- viduals in international alliances rely on multiple sources of socialization to understand and organize their interactions and stimuli. However, their relative effect is left unspecified. In a study of national subcultures, Lenartowicz and Roth (2001) found that by identifying sub- cultures, which more accurately captured differ- ences in the Brazilian population, they were able to accurately predict differences in business perfor- mance based on those subcultures. In other words, by specifying and understanding a more proximal and salient culture to which people belong, one can predict behaviors more accurately. This logic helps explain Pothukuchi et al.’s (2002) finding that organizational culture differences have more influ- ence on alliance performance than national culture differences because organizational culture see- mingly is more proximal and salient to the behaviors of individuals that drive business perfor- mance in international alliances. Continuing this logic, we propose that profes- sional culture is often the most proximal and salient influence on the behavior of individuals interacting in the value-creating activities of an international alliance. Therefore differences between individuals’ professional cultures will be the most disruptive. As a result, the negative effects stemming from differ- ences in the professional cultures of partners’ employees will be greater than those effects stem- ming from difference between either the partners’ organizational or national cultures. Formally: P6: Differences in the professional cultures of the employees involved in the alliance’s primary value-creating activities will have more influence on the relationship between related complemen- tary resources and the effectiveness of the alliance’s value-creating activities than differ- ences in the partners’ organizational or national cultures. Discussion And Conclusion We have introduced a detailed model of cultural differences and international alliance performance that differs from prior research in several ways. First, we have proposed that the performance of international alliances is influenced by the effec- tiveness with which an alliance achieves its primary value-creating activities. Often a necessary condi- tion for such effectiveness is the presence of complementary resources that are related to those activities. However, the presence of complementary resources alone is insufficient. Additionally, the partners’ employees must be able to interact effectively in order to share, combine and leverage those resources. Thus factors that affect these interactions are critical in understanding interna- tional alliances’ outcomes. Following the literature, we focused on cultural differences as factors that affect such interactions. Building on the findings of Pothukuchi et al. (2002), we argued that the more salient the cultural differences are to the value-creating activities of an alliance, the more disruptive those differences will be on the alliance’s value-creating activities. Thus differences in organizational culture and profes- sional culture – not just differences in national culture – need to be assessed. We believe that our model offers valuable new insights for international alliances. We focused on international alliances because they are increasing in both number and competitive importance. Using an example of IRDA, we demonstrated our model. Building on research from the sociology of knowledge, we showed that R&D personnel’s professional cultures could be either science-based or engineering-based. Moreover, because these cultures are most relevant to the individuals actually interacting in an IRDA, differences between these professional cultures will be the most disruptive to the alliance’s value-creating activities. However, we also feel that our model presents implications for other contexts in which culture differences are believed to matter. This model is relevant to international mergers and consortia, and also to collaboration within large multinational firms, where different organiza- tional cultures exist in separate divisions: differ- ences in the divisions’ organizational cultures may be most salient. Further, the ideas developed in this study also have implications for domestic alliances and mergers. They suggest that the alignment of professional cultures matters more than the align- ment of organizational cultures. There is indirect Journal of International Business Studies This content downloaded from 139.182.75.138 on Thu, 25 Apr 2019 06:55:22 UTC All use subject to https://about.jstor.org/terms A model of cultural differences David G Sirmon and Peter j Lane 316 evidence of this phenomenon in Lane and Lubat- kin’s, (1998) study of R&D alliances between US pharmaceutical and biotechnology firms. They compared the formalization, centralization, and compensation practices of the partners, which were artifacts of their organizational cultures. They found that differences in research centralization and research compensation practices were nega- tively related to R&D alliance performance (ie learning), but differences in management centrali- zation and upper management formalization were not. These findings are in keeping with our contention that differences in beliefs and norms matter more as they become more relevant to the value-creating activities of the alliance. However, the generalizability of the model is limited to those situations where value is created via the sharing, combing and leveraging of com- plementary resources. Thus, the model does not apply as well to inter-firm relationships that focus on cost-cutting or the linking of similar resources. Clearly, an important next step in this line of research is the empirical testing of the model developed in this paper. Several well-developed measures of national culture and organizational culture can be used. For example, the study by Pothukuchi et al. (2002) used both Hofstede’s (1980) measures of national culture values and his measures of organizational culture (Hofstede, 1991). Professional culture could be assessed using various different measures, such as formality and length of professional training, degree of cross- functional experience of unit members, and the number of memberships that individuals have of occupational associations. Each of these measures of a professional culture would need to be adjusted for industry norms. The purpose of this paper was to deductively develop a model based on prior academic research. It has therefore avoided normative propositions. While this may make our ideas seem overly deterministic, and may suggest that key contingen- cies are beyond managerial control, this is not the case. Our model suggests that managers need to carefully evaluate both the complementarity of potential partners’ resources and the fit of the professional cultures of the employees who are likely to work together in the alliance. Without the former, there is little reason for an alliance. With- out the latter, the alliance may fail to fulfill the potential promised by the complementary resources that the partners possess. Most importantly, the addition of professional culture to the model provides managers with a mechanism, not apparent before, for addressing cultural conflict. Organizational cultures evolve over years and decades; national cultures evolve over decades and centuries. However, professional cultures may be represented by a relatively small group, which makes them more amenable to managerial intervention. Managers may influence their professional cultures by hiring and appropri- ately rewarding members who develop a level of proficiency with other professional cultures. In conclusion, international alliances are an increasingly important tool for expanding a firm’s competitive advantage. Because they face chal- lenges from several sources of cultural difference, effective collaboration depends on eliminating the most disruptive sources of cultural difference. In doing so, international alliances increase the like- lihood that their pooled complementary resources will be shared, combined and leveraged in the primary value-creating activities of the alliance, leading to increased performance. Acknowledgements We thank the participants of the Governing Knowl- edge Processes Conference as part of the Learning, Incentives, and Knowledge (LINK) Program, Copenha- gen School of Business, September 2001, for their helpful comments on earlier drafts of this paper. We also thank Nicolai juul Foss and Torben Pedersen, Departmental Editors, and four JIBS reviewers for their insights, critiques and suggestion