Business and Management Review Vol. 2(12) pp. 07 – 12 February, 2013 Available online at http://www.businessjournalz.org/bmr ISSN: 2047 - 0398 Discourse Analysis of Interfirm Networks in Japan Hajime USHIMARU,Ph.D. Professor, School of Business Administration, Meiji University, Tokyo, Japan. E-mail: ushimaru@kisc.meiji.ac.jp Masato YOTSUMOTO Guest Researcher, Organization for the Strategic Coordination of Research and Intellectual Property, Meiji University, Tokyo, Japan. E-mail: miyabi-4@xa2.so-net.ne.jp ABSTRACT We have explored effective discourse in interfirm networks with different characteristics: Star Alliances and Poitan. Network analysis and discourse analysis were adopted as analytical methods of this study. Results showed three hypothetical findings: 1) When network patterns differ, differences also are apparent in the discourse used within the network. 2) Cooperative discourse is used where network uncertainty is low. On the other hand, where uncertainty is high, competitive discourse is used. 3) A firm can secure high rents by using different types of discourse suited to the characteristics of each network. Keywords: Coleman-type network, Status-type network, organizational discourse, critical discourse analysis, intertextual analysis 1. INTRODUCTION The purpose of this study is to explore effective discourse in inter firm networks that have different characteristics. For this reason, we will conduct discourse analysis of two inter firm networks with different characteristics: Star Alliance and Poitan. The research domain known as interfirm networks or interiorganizational networks (“interfirm networks” hereinafter) was born in the 1990s. One possible reason behind this development was the increase in the practical validity of looking at firms as being in cooperative, dependent relationships instead of in exclusive, competitive ones. A typical example is the increase in strategic alliances. A time had arrived in which it would be difficult for firms to maintain and grow their positions without cooperation, while still competing with each other, in order to respond to intensifying competition, rapid globalization, and acceleration of organizational learning. Also, the networks to which a firm should belong and how it should position itself within them became important to a firm’s survival and growth. An important point here is the fact that firms necessarily employ discourse when seeking, negotiating with, and maintaining relations with partner firms. Discourse refers in general to methods of communication, such as conversation and dialogue, narratives and stories, rhetoric and allegory, and symbols. Explanatory theories for interfirm networks include interorganization theory and social network analysis. Interorganization theory holds that networks form as each firm strives to demonstrate the greatest power (resource dependence theory), secure legitimacy (institutional organization theory), or minimize transaction costs (transaction cost theory). On the other hand, social network analysis ascertains quantitatively information on the network as a whole, the units making up the network, and the structural and relational properties among units. The concept of discourse can be considered to serve as a bridge between interorganization theory and social network theory. 7 Business and Management Review Vol. 2(12) pp. 07 – 12 February, 2013 Available online at http://www.businessjournalz.org/bmr ISSN: 2047 - 0398 2. BASIC CONCEPTS OF ORGANIZATIONAL DISCOURSE ANALYSIS Definitions of discourse vary among researchers. There is no fixed definition. However, by summarizing various definitions we can define it as being used to express or represent certain meanings under certain social conditions. In general, discourse is expressed in diverse textual forms (Philips & Hardy, 2002). Texts include spoken words, organizational stories and narratives, speech, dialogue, and conversation, as well as social practice on the Internet such as email, chat, and social networking service (SNS), as typified by advances in information and communications technology in recent years or the public-relations bulletins, press releases, internal organs, company histories, websites, and CSR reports issued by firms. Organizational discourse analysis (ODA) analyzes organizations from discursive perspective, considering the organization to be constructed socially (Putnum & Fairhurst, 2001). As used here “socially” means that the organization and its behavior do not exist in an objectively stable state but are habitual and unstable, constantly given meaning and transformed by social contexts. For this reason, organizational discourse analysis considers organizations and organizational behavior to be formed on the cultural grounding of the values and beliefs of society. Mainly there are two research approaches to discourse analysis: 1) critical discourse analysis, and 2) intertextual analysis. Critical discourse analysis refers to analysis of discourse as seen in the vertical systems in a social structure. Here the purpose is to investigate the authoritarian, dominant ideologies behind discourses. For this reason it attempts analysis from three dimensions. The first dimension is that of discourse as text, an attempt to analyze the authoritarian, dominant ideologies through analysis of textual vocabulary, grammar and structure. The second dimension is that of discursive practice, which analyzes subjects such as how text is created, how it moves from one state to another, and how it is analyzed, interpreted, and influenced. The third dimension is that of discourse as a social practice, which analyses the ideological relations and hegemonic relations demonstrated by and involved in the text. Such discourse as political practice establishes, maintains, and transforms power relations and the collective entities (i.e., classes, coalitions, communities, and groups) secured by such power relations (Fairclough, 1992). This kind of discourse can be seen to a large extent in interfirm networks. The meaning of the network world can be considered to be formed, accepted, maintained, and transformed from diverse positions in power relations. Intertextual analysis refers to analysis of discourse as seen in the horizontal systems in a social structure. It aims to investigate the membership behind discourse. For this reason, it analyzes how the texts of each constituent unit interact with each other, correct each other, and form shared texts, and the kind of influence that the shared texts thus formed have on each constituent unit. A text is seen to exist not on its own but within an interconnected relationship with other texts. While autonomous, participants in communication also are constituent members of the subjective community built at that point. If discourses between participants are linked horizontally and the words of each continue to generate interdiscursivity, then these texts are depicted as interrelated and multilayered. When attempting discourse analysis in interfirm networks, critical discourse analysis would be effective if vertical relationships are apparent, as in intragroup business transactions, while intertextual analysis would be effective in the case of flat relations, such as strategic alliances. 3. ANALYTICAL METHOD Investigating discourse in interfirm networks having different characteristics requires the availability of a variety of types of interfirm networks. However, since this is a pilot study we will use only the following two network types, in order to simplify the discussion: The first is what we will call a Coleman-type network, abstracted from Star Alliance in 8 Business and Management Review Vol. 2(12) pp. 07 – 12 February, 2013 Available online at http://www.businessjournalz.org/bmr ISSN: 2047 - 0398 the airline industry. The other is what we will call a status-type network, abstracted from Poitan, an interfirm loyalty-point alliance in Japan. The distinctive features of each of these networks are outlined below. 3.1 Star Alliance The world’s largest airline alliance, Star Alliance had 25 airlines as members (including regional airlines) as of May 2012. These included ANA. The Japanese mileage-program comparison site Mile de Tokutoku1 identifies the following points as characteristics of Star Alliance: It makes it possible for flyers to earn and use points nearly everywhere in the world. It makes it easy to earn points (with few fares not eligible for points). It has a strong network in Europe. It continues to grow. Star Alliance is characterized by a very high level of network density, with very little apparent difference in mileage services among member firms. Even new member airlines can form mileage-exchange relationships on largely equal terms with all member airlines. Star Alliance can be described as a Coleman-type network (Kogut, 2000), which refers to a network having close relations. 3.2 Poitan Wakabayashi and Katsumata (2011) have studied the characteristics of the Poitan network. Their study identified the following characteristics: The loyalty-point exchange network is irreversible and asymmetrical. For example, it includes relations in which it is possible to convert points from a to b but not from b to a. In other words, the point exchange network is a directed graph. According to data current as of March 4, 2010, its membership consisted of 155 firms in 16 industries. Eight of these firms were isolated, without having concluded alliances with any other firms. A look at the 147 firms other than the eight isolated ones shows that there were 665 interfirm alliances in existence. Thus 665 was the number of links in the network. Main network indicators: Its density was 0.031, the clustering coefficient of the network as a whole was 0.120, and the average path length was 2.814. The network power exponents show an indegree value of 2.097 and an out degree value of 2.007. Asymmetric alliances had formed, reflecting the expectations of each firm. There were disparities among industries between those with high tendencies to issue points and those with high tendencies to accept points. From these findings, it can be said that the point network includes distorted networks and disparities based on position, as industries and firms in important positions can make their own choices of partner firms and form relationships with them while firms in low positions are unable to form networks. Since this network has the characteristics of a status network as identified by Podolny (1993), we will call it a status-type network. 3.3 Data As discourses subject to analysis, we chose joint websites2, individual firms’ websites, various reports, and pamphlets. 4. ANALYSIS 4.1 Star Alliance We will start our discourse analysis with Star Alliance. Star Alliance Services GmbH manages the Star Alliance network. This network organization, based in Frankfurt, Germany, was formed as the world’s first airline alliance. Its membership, which has a strongly international flavor, consists of 25 airlines from 20 countries. What is their mission? According to the official Star Alliance website it is3: “Executing leadership in managing a portfolio of alliance products and services using an agreed process.” 9 Business and Management Review Vol. 2(12) pp. 07 – 12 February, 2013 Available online at http://www.businessjournalz.org/bmr ISSN: 2047 - 0398 Star Alliance member airlines fly to more destinations than any other airline alliance in the world – which means easier travel and quicker connections. The main goal has always been to make your travel experience smoother. To achieve this, Star Alliance member airlines are located closer together in airports and connections teams are installed for faster transfers. Common airport facilities, coordinating schedules and a range of new technologies are also frequently introduced. As if in concert with the discourse expressed in the form taken by this mission, member airline Lufthansa, to give one example, summarizes its own strengths using the following discourse4: “A broad-ranging network: Lufthansa now flies to 198 cities in 82 countries around the world (according to the winter 2011 schedule). Among these destinations, the fullness of its routes to Eastern Europe and Commonwealth of Independent States nations (including Ukraine, Belarus, and Kazakhstan) is unmatched. It also continues expanding its routes to Western Africa and the Middle East. With Lufthansa you can schedule your travel the way you want it. Furthermore, Star Alliance, of which Lufthansa is a member, offers a global network connecting airports in 1196 cities in 186 countries around the world.” It also emphasizes the possibility of smooth connections using this network. For example, it points out that since flights from Japan to Frankfurt or Munich, located in the center of Europe, arrive in the afternoon it is possible to make seamless connections to cities across Europe. The official website for Japanese users contains a discourse stating that the network makes it possible to arrive via a flight from Japan on the same day in more than 100 cities in Europe alone and extolling the convenience of both the Frankfurt and Munich airports, where it takes only 30 – 45 minutes to change flights. Continental Airlines, which merged with United Airlines at the end of last year, joined Star Alliance in December 2009 as well. At that time, then-Continental CEO Larry Kellner stated, “Our membership in Star Alliance positions us to deliver a broader network to our customers, and to achieve better business results and a stronger future for my co-workers, our customers and communities as a result of the benefits from participating in the world’s largest airline alliance.” A press release issued October 27, 20095 includes three discourses on “Global Reach,” “Worldwide Recognition,” and “Seamless Travel.” These three discourses express clearly the reciprocal relationship realized through membership in Star Alliance. This shows the resulting strengthening of the network and convenience to customers. Similarly, a look at the official websites of Star Alliance member airlines shows one case after another of discourses describing the expansion of route networks, seamless connections, and reciprocal relationships among mileage programs. Lastly, we would like to note how the special Star Alliance website for the Japan region demonstrates the characteristics of this airline alliance network. It does so in a “relay column”6, a series of columns provided by member airlines. In this column, staff members from member airlines discuss the discourse of Star Alliance network services through introductions to the services that each airline provides itself, while also pledging a high level of services from the network as a whole. The discourse of this column is highly interesting in the way that it provides a glimpse of the differences among member airlines while also extolling the way their services provide the same level of quality. In this way, intertextual analysis of the Star Alliance network shows a very high abundance of discourses expressing reciprocal relationships, demonstrating that an organic network whose members maintain similar points of view is being generated even while each member company emphasizes its own individuality. 4.2 Poitan Next, we will conduct discourse analysis based on the official Poitan website. As is clear from its self-definition as a “loyalty point program portal site,” Poitan does not actively attempt to form a network. It merely provides a place for 10 Business and Management Review Vol. 2(12) pp. 07 – 12 February, 2013 Available online at http://www.businessjournalz.org/bmr ISSN: 2047 - 0398 networking (the website). However, we noticed that a massive network had formed by systematically linking the point-exchange services of the point systems deployed by each company. But what are the discourses of the network formation seen through Poitan? They are the routes by which points can be exchanged and the exchange rates, which indicate the ratios at which points can be exchanged for each other. Looking at the discourses of these routes and exchange rates using the case of JAL as an example makes the following two points clear. First, this network is very highly irreversible. That is, while JAL points can be converted to those of 11 firms, the number of firms whose points can be converted to JAL points is sixty-two. Of the latter group, JAL mileage cannot be converted directly to points for 52 firms. Next, as seen in the point exchange rates discussed above this network is strongly asymmetrical. While points can be converted from JAL points at a ratio of 1:1, exchanging points to JAL points entails a heavier cost, with an exchange rate of 1:2. Here we can use critical discourse analysis to identify the presence of power in this interrelation. In particular, the predominance of airline mileage point programs in the network, including those of JAL as seen in the above example along with ANA, American Airlines, and others, stands out. Also, a network of firms’ loyalty point systems as seen in the case of Poitan requires only that member firms be connected in a point exchange system. As such, in this case one does not see the kind of union based on overall philosophy or shared objectives as seen in an airline alliance. For this reason, the way of speaking in the discourses of Poitan gives a highly inorganic impression, as it sticks to extremely functional and rational subjects such as exchange rates and ratios. In discourse analysis, it is only such inorganic routes and exchange-rate figures that indicate intertextuality. For this reason, the reciprocal relationship here is a weak one, and instead one uncovers a structure made up of the dominance of firms that have predominant point systems over other firms. 5. CONCLUSIONS In this study, we conducted an analysis of variance in discourse as seen in the Coleman-type interfirm network of Star Alliance and the status-type interfirm network of Poitan. Our results can be summarized in the following three hypothetical findings. We have seen that a Coleman-type network such as Star Alliance employs reciprocal discourse and involves cooperative relations. A cooperative interfirm text (i.e., interfirm culture) can be said to have been formed. On the other hand, in a status-type network such as Poitan unilateral discourse is deployed, and we have seen in this case structures and power relationships in which many firms are linked in dependence relationships to some firms that have high status. These facts present us with the following hypothetical finding. Hypothetical finding 1: When network patterns differ, differences also are apparent in the discourse used within the network. ANA, a member of the Star Alliance group, also is a member of the loyalty point group. Here ANA employs unilateral discourse. This differs from the reciprocal discourse seen in Star Alliance. One possible reason for this is the fact that the size of the network can be predicted (i.e., there is low uncertainty) in an industry such as the airline industry in which Star Alliance is active, where the number of firms is limited, and network density is high too. In such a case, relations of trust are fostered and one can see cases of choosing to use cooperative discourse. On the other hand, in an industry such as Poi tan’s in which it is relatively easy for firms to become members or withdraw from membership, it is difficult to predict the size of the network (i.e., uncertainty is high). Also, network density cannot be said to be high either in such a case. It is easy for “backstabbing” activity to take place in this case, and one can see cases of choosing discourse that attempts to outmaneuver the competition through use of competitive discourse. These facts present us with the following hypothetical finding. 11 Business and Management Review Vol. 2(12) pp. 07 – 12 February, 2013 Available online at http://www.businessjournalz.org/bmr ISSN: 2047 - 0398 Hypothetical finding 2: Cooperative discourse is used where network uncertainty is low. On the other hand, where uncertainty is high, competitive discourse is used. In addition, while in a Coleman-type network each firm can secure only average rents; in a status-type network a firm with higher status can secure higher rents. ANA belongs to both networks, and it used a different type of discourse in each. These facts present us with the following hypothetical finding. Hypothetical finding 3: A firm can secure high rents by using different types of discourse suited to the characteristics of each network. This study employs the differing research approaches of network analysis and discourse analysis. Regarding discourse analysis in particular, the critical discourse analysis and intertextual analysis employed in this study are analytical methods based on postmodernism, while network analysis employs a research approach that can be categorized under functionalism. For this reason, analysis fusing these approaches simultaneously may be open to criticism. On this point, in this study we employed network analysis as a means of seeking out a place for discourse analysis and identifying its characteristics. In this way, it is possible to conduct a study using both postmodernism and a functionalistic approach by employing different research approaches at different times in the analytical process. While postmodernism focuses on a variety of phenomena abstracted from the range of functionalism and it is meaningful in that it makes clear the various problems inherent to functionalism and provides new analytical perspectives, at the same time it also is impossible to discuss all phenomena using postmodernism alone. This is why in this study we were able to identify a new course of action in network analysis through seeking an organic unity between both approaches. This can be described as a major distinguishing feature of this study. In addition, in this study we identified only two of the numerous types into which interfirm networks can be categorized and elucidated their differences through discourse analysis. For this reason, there is a need to consider other types of networks as well in the future. Also, while in this study we looked at websites as the subjects of discourse analysis, we probably would be able to avoid falling into a situation of one-sided analysis by also using various other forms of discourse in a supplementary fashion. These are issues to address in the future. NOTES 1 http://www.mile-tokutoku.com/alliance/alliance_SA.htm; accurate as of May 21, 2012. 2 The Star Alliance website: http://www.staralliance.com/ja/ The Poitan website: http://www.poitan.net/ 3 http://www.staralliance.com/ja/about/organisation/ 4 http://www.lufthansa.com/online/portal/lh/jp/nonav/local?nodeid=1905573&l=ja 5 https://www.united.com/web/ja-JP/micro/company/press/2009/1027_02.aspx?Mobile=1 &SID=36F356451E084739B39C23674ABD8171& 6 http://www.staralliance.jp/pr/column/ REFERENCES Fairclough, N. (1992) Discourse and Social Change, Polity Press. Kogut, B. (2000) The Network as Knowledge: Generative Rules and the Emergence of Structure, Strategic Management Journal, 21, pp.405-425. Phillips, N. and Hardy,C. (2002) Discourse Analysis. Thousand Oaks, Sage. Podolny, J. M. (1993) A Status-based Model of Market Competition, American Journal of Sociology, 98(4), pp.829-872. Putnam, L. and Fairhurst,G. (2001) Discourse Analysis in Organizations: Issues and Concerns, Jablin,F.and Putnam,L. (eds), The New Handbook of Organizational Communication, Thousand Oaks, Sage. pp.78–136. Wakabayashi, T. and Katsumata, S. (2011) Which Factor Matters on the Formation of the Strategic Alliance Network: Industry, Firm, or Network? University of Tokyo, MMRC, Discussion Paper Series, No.366. 12