INTRODUCTION

Since its inception in the late 1960s, the field of international business (IB) has been a multi-disciplinary endeavor. Defined and characterized not by its adherence to an agreed set of theories and methods but by a common research focus on foreign direct investments (FDI) and the multinational enterprise (MNE), IB scholars have over the years eclectically seized upon theories, concepts, and methodologies from other fields (Liesch, Håkanson, McGaughey, Middleton, & Cretchley, 2011). The Point by Len Treviño and Jonathan Doh (2020) aligns well with this tradition. The authors borrow methods and concepts of discourse analysis – an approach not frequently used in IB – and apply them to one of the field’s central topics, the internationalization process of the firm. The result is a thought-provoking and stimulating paper, raising several intriguing questions. At the root of these are two interrelated issues that this Counterpoint seeks to address: The first relates to the delimitation of the topic and the empirical phenomena that fall within the realm of inherited internationalization process theory as formulated in the so-called Uppsala school. The second concerns the nature and theoretical underpinnings of discourse analysis, and how it might fruitfully enrich our understanding of the internationalization process of the firm.

Treviño and Doh (2020) offer as illustrations of their argument (1) the recent international activities of Bulova, following its 2008 acquisition by Japanese Citizen, and (2) Google’s 2018 decision to reenter the Chinese market with a search engine compatible with “the Chinese willingness [sic] to censor searches” (Treviño and Doh, 2020). The former is based on ‘personal communication’ (of undisclosed nature) with the CEO appointed by Citizen after the takeover of Bulova, the latter on two newspaper articles in the New York Times. While both constitute interesting examples of complex international strategies, neither of them relates to the internationalization process of the firm. The unhappy choice of illustrations is indicative of an increasingly common ambiguity and tension regarding both explananda and explanantia in the literature on the internationalization process of the firm.

This Point takes as its starting point the inherent time-dependence of the topic, and the historical perspective evident in the seminal Uppsala model, which for several decades provided the paradigm for research on the internationalization process of the firm. The historical perspective has important implications. First, the relevance and validity of the model is dependent on historical context. The changing nature of that context over the last 40 years has called for a series of elaborations and revisions (Vahlne, 2020). While both timely and justified, these have fundamentally changed the nature of the model, in the process undermining its usefulness and paradigmatic status. Although not yet widely recognized, this has created a ‘crisis’ in the Kuhnian denotation (Kuhn, 1970: 66–76). Other than the paradigm shifts and scientific revolutions that Thomas Kuhn discusses, it clearly does not affect the IB field in its entirety. However, as the following section outlines, the crisis makes evident the need for a paradigm shift in research on firm internationalization. Second, lest we are prepared to throw out the baby with the bathwater, the search for a novel framework should acknowledge the power and potential of the inherited historical approach (Buckley, 2009, 2016; Jones and Khanna, 2006), and the necessity to apply an historical perspective and methodology to the topic. The point by Treviño and Doh (2020) is a timely contribution to this search. Discourse analysis offers a potentially highly fruitful historical methodology (Foucault, 1969/2002) to the study of the internationalization process of the firm, and may well prove influential in the formulation of a new paradigmatic framework. This is the topic of the third section of the Counterpoint.

THE INTERNATIONALIZATION PROCESS MODEL

The Uppsala Paradigm

The dominant model of the internationalization process in international business (IB) research was developed at Uppsala University in the early 1970s (Carlson, 1974; Hörnell, Vahlne & Wiedersheim-Paul, 1973). The results received the attention of a wider audience through the seminal papers by Jan Johanson and Finn Widersheim-Paul (1975) and Johanson and Jan-Erik Vahlne (1977), outlining the parsimonious, and intuitively appealing theoretical model of the internationalization process of the firm now known as the Uppsala model. As argued elsewhere (Håkanson & Kappen, 2017), the Uppsala model over time became a ‘scientific paradigm’ in Kuhn’s (1970: 23 f.) sense:

… [A] paradigm… is at the start largely a promise of success discoverable in selected still incomplete examples. Normal science consists in the actualization of the promise, an actualization achieved by extending the knowledge of those facts that the paradigm displays as particularly revealing, by increasing the extent of the match between the those facts and the paradigm’s predictions, and by further articulating the paradigm itself.

The paradigmatic character of the model derives as much from its few basic assumptions as from its incompleteness and open-endedness. It provided the IB community with a shared reference for the fruitful ‘puzzle-solving’ characteristic of ‘normal science’ (Kuhn, 1970: 23–42), helping to define both the nature of the research questions (‘unresolved puzzles’) to which it applies, and the criteria for what counts as acceptable solutions (Kuhn, 1970: 184). In Mats Forsgren’s concise summary, the main issues of the model are “…pace, direction, and commitments of firms’ foreign operations…” (Forsgren, 2016: 1142, italics added). Although not spelled out, this can be taken to include mode of entry, the model’s main indicator of ‘commitment’.

The dynamic aspect of the model is significant. “The object to be explained was the internationalization process, not internationalization” (Vahlne, 2020: 241, italics added). As in all process research, history matters, and the order and sequencing of events are significant (van de Veen & Poole, 1995). Important questions include “who did what, where they did it, and how, and why they did the things they did” (Welch, Nummela, & Liesch, 2016: 783). Although often obscured in quantitative cross-sectional research, the process focus of the Uppsala model reflects its inherently historical research approach (Buckley, 2016), the object of which is to understand change: “History deals in events, not states; it investigates things that happen and not things that are…. The historian may well interest himself in the state of things…, but if he is to understand historically… he will have to concentrate on understanding change…“(Elton, 1969: 22).

In a recent retrospective, Vahlne (2020) outlines the evolution of the Uppsala model as articulated by himself and Johanson in a series of articles elaborating upon and reformulating the 1977 original. He describes how the model’s objective changed from explaining the nature of the internationalization process of the firm to explaining ‘evolution’ more generally. In the interest of simplicity and brevity – leaving out much detail and valuable elaborations published elsewhere – the changes can usefully be summarized by reference to three landmark articles in JIBS: the Johanson and Vahlne (1977) original version, the award-winning Johanson and Vahlne (2009) revision of the model, and the Vahlne and Johanson (2017) Point on its 40th anniversary.

The core of the Uppsala model is its explanation of how and why business firms expand into foreign markets. As in all models of organizational evolution (van de Veen & Poole, 1995), key elements include (1) the nature of the context to which it applies, and (2) the micro-foundations that drive organizational change, in this case internationalization events, and their timing and sequence. In the course of the model’s revisions, both elements have changed.

Context and Micro-foundations

A premise underlying the Uppsala model revisions is the perception that technological, institutional, and political developments of recent decades have fundamentally changed both the context of international business activities, and the managerial practices that guide firm behavior. Better knowledge and reduced uncertainty about market conditions abroad have diminished the liability of foreignness and uncertainty confronting firms entering foreign markets. At the same time, new business practices, especially the increasing importance of international mergers and acquisitions, have affected the nature of learning mechanisms and the handling of uncertainty encountered in the internationalization process. In the reformulations of the model in response these changes, its focus has gradually shifted.

In the first major iteration, Johanson and Vahlne (2009) invoked the network perspective, originally developed to analyze marketing and sales of industrial products (Johanson, 1966), and later extended to include cooperative arrangements more generally. The 2009 model places the challenges facing internationalizing firms in the more general context of growing firms having to enter new business networks or strengthen their position in existing ones. In principle, the challenges facing a start-up firm as it attempts to get a foothold in its domestic market are the same as those facing a firm attempting to enter a foreign market (Forsgren, 2016). The latter may be complicated by a residual ‘liability of foreignness’, but this is now a factor among many and of no special significance.

The effect of this perspectival change is to subsume the internationalization process of the firm into a more general model of firm evolution. In the new conceptualization, the original questions regarding where, how, when, and how rapidly firms internationalize are not addressed. Experiential learning is still important, but that learning now concerns primarily the understanding and building of trust with business partners, and is substantially the same as encountered in domestic networks. Since experimental learning takes time, greenfield commitment decisions will still tend to be incremental, but given the growing importance of mergers and acquisitions, incremental foreign expansion cannot generally be assumed. In partial consequence, Johanson and Vahlne (2009:11) state that “the correlation between the order in which a company enters foreign markets and psychic distance has weakened”, but precisely what this means is unclear. Originally and in the bulk of the literature, psychic distance was a country level construct, a summary measure of the ease or difficulty faced by managers on average when trying to obtain and interpret information about a foreign market. It now “applies at the level of the decision-maker” (Johanson & Vahlne, 2009: 11). Although this is not explicitly stated, the argument rests on the reasonable assumption that over time individual psychic distance perceptions have become increasingly heterogeneous due to more common and more varied exposure to foreign countries (through expatriation, student exchange programs, and vacations in exotic destinations, for example). In consequence, randomly idiosyncratic personal experiences of individual managers determine the order in which firms enter foreign markets. In short, the network conceptualization of 2009 neither provides conclusions nor invites hypotheses about the speed, form, and geographical patterns in firm internationalization.

At the time of the publication of the initial Uppsala model (Johanson & Vahlne, 1977), theoretical assumptions and perspectives in economics dominated IB research (Liesch et al., 2011). One of the reasons for its impact was the adoption of a managerial perspective and the behavioral assumptions of bounded rationality, uncertainty avoidance and satisficing (Simon, 1947; Cyert & March, 1963), as well as an emphasis on the importance of incremental experiential learning. In combination with the contextual premises, these assumptions are the core aspects of the internationalization process model, and determine its expected outcomes. Surprisingly, however, the driving forces behind firm growth are given only scant attention: “In the model, it is assumed that the firm strives to increase its long-term profit, which is assumed to be equivalent to growth…” (Johanson & Vahlne, 1977: 27). The authors invoke Edith Penrose’s (1959) seminal work in relation to the concept of ‘experiential knowledge’, but her discussion of the factors driving firm growth appears not to have had any influence.

The absence of a clear conceptualization of the determinants of firm growth, in combination with the de-emphasis of the importance of country differences and the ‘liability of foreignness’, drastically undermines the Uppsala model’s relevance for the internationalization process of the firm. In apparent recognition of this, the latest iteration (Vahlne & Johanson, 2017) abandons this focus altogether. It has now become a general model of ‘evolution’, applicable to most everything that goes on in firms, from learning and capability development, over resource deployment to strategy formulation and implementation. At its core is now the behavioral assumption that “decision-makers weigh the gains of exploiting an opportunity against the risk of losing the resources at hand… [and] the decision-making mechanism … whereby managers compare expected benefits and potential risks” (Vahlne & Johanson, 2020: 8), essentially a reiteration of behavioral models of managerial decision-making, dating back to Herbert Simon’s (1947) ‘administrative man’.

DISCOURSE ANALYSIS AND HISTORICAL METHOD

The term ‘discourse analysis’ does not unambiguously designate a well-defined scientific perspective or method but includes a whole range of approaches (Keller, 2013). This is not surprising since it has found application in many disciplines, and, as Treviño and Doh (2020) note, ‘discourse’ is a broad concept, in its widest sense referring to “the practice of talking and writing.” In research on the internationalization process of the firm, three categories of analysis of surviving documents and other artifacts of discourse, (including recordings and transcripts of interviews, for example) can be identified.

The first corresponds to that traditionally employed in historical research and analysis, where textual and symbolic artifacts provide historical evidence, based on which the occurrence and timing of events can be traced, analyzed, and interpreted. Not coincidentally, the appearance of written records in a civilization marks the demarcation between history and prehistory. Important methodological concerns include the determination of the authenticity and meaning of surviving textual documents, questions regarding their truthfulness – can the messages be taken at face value, were they misinformed or possibly deliberately misleading, and so on. They also include the problem of survival bias – extant evidence tends systematically to ignore certain phenomena in favor of others, most commonly and conspicuously in IB research, unsuccessful firms and strategies in favor of successful ones (Denrell, 2005). The early research of the Uppsala group (Hörnell, Vahlne, & Wiedersheim-Paul, 1973) clearly fall into this category. It used primary and secondary documentary evidence to trace the evolution of the international activities of Swedish multinationals through the appointment of agents and establishment of wholly or partially owned foreign subsidiaries. Like in much of the subsequent literature on ‘born globals’ or ‘international ventures’, little or no attention was paid neither to companies that did not internationalize, or had tried but failed to establish a foreign presence, nor to the resulting bias in the sample of observations that formed the basis for the analysis. This explains the original model’s seemingly (and much criticized) ‘deterministic’ character; its aim was to explain observed patterns of internationalization in successful and growing firms. Later modifications and clarifications were not based on empirical observations but on theoretical considerations.

Discourse analysis in a more narrow sense, with analytical focus on the structure and content of written texts and oral accounts are not common in IB research. One group of studies has focused on the analysis of academic texts and textbooks (Adler & Bartholomew, 1992; Jack & Westwood, 2006; Fougère & Moulettes, 2012). More recently, Balogun and colleagues (Balogun, Jarzabkowski, & Vaara, 2011; Balogun, Fahy, & Vaara, 2019) fruitfully applied discourse analysis to research headquarters–subsidiary relationships. A related, relatively large literature has focused on the role of language in international business (Brannen, Piekkari & Tietze, 2014; Kuznetsov & Kuznetsova, 2014; Luo & Shenkar, 2006; Vaara, Tienari, Piekkari, & Säntti, 2005). Others have, like Treviño and Doh, used discourse analysis methodology to analyze strategy formulation more generally (Gopinath & Prasad, 2013; McKenna, 2011).

Of immediate relevance to the present topic are analyses of managers’ and entrepreneurs’ narratives of their internationalization experiences (Haley & Boje, 2014; McGaughey, 2007; Middleton, Liesch, & Steen, 2011), providing valuable insights regarding the perceptions and understandings that guide their decisions and internationalization strategies. The approach can be seen as a response to repeated calls for a strengthening of our understanding of the ‘micro-foundations’ of strategic management theory in general (Felin & Foss, 2005; Foss & Pedersen, 2016; Simon, 1985) and the internationalization process of the firm in particular (Coviello, Kano, & Liesch, 2017). In this vein, Vahlne and Johanson (2020: 7) suggest that “… research on cognitive and emotional processes can shed light on the phenomenon of internationalization, and firm evolution in general.” However, the fruitfulness of extending the model to the ‘mille-micro level’ of individual decision-makers has still to be demonstrated.

A third, as yet largely untried approach would be to anchor micro-foundational assumptions at the level of collectives, such as those of firms or industries. Here, historical discourse analysis in the form developed by Michel Foucault (2002) provides valuable methodological insights, 1969 seminally summarized in his L’Archélogie du Savoir (The Archaeology of Knowledge). Foucault’s primary interest is to investigate the formation of ‘epistemes’, the basic knowledge structures that define scientific and other disciplines at specific periods of time, and through discourse analysis to identify the rule-systems by which these emerge, consolidate, and change. He refers to this methodology as ‘archeology’, a task that treats discourses not as representations, “…but as practices that systematically form the objects of which they speak” (Foucault, 2002: 49). Archeological discourse analysis differs from the traditional employment of texts in historical analysis as a means to identify and explain the occurrence of particular historical events. Its aim is to understand the process of knowledge creation in a specific societal context, an approach that parallels social constructivist approaches to the sociology of knowledge (Berger & Luckmann, 1966: Holzner, 1968; Keller, 2011). Foucault applied his methodology to discourses in a wide range of areas, including the history of madness, medicine, sexuality, and the way societies discipline and exact punishment, but there is in principle no reason why it could not be applied also to discourse in business firms regarding the formulation and implementation of internationalization strategies. The main obstacle would seem to be the availability of and access to relevant textual or audio-visual data (Keller, 2013: 93–105). However, as demonstrated in the seminal works of business historians such as Alfred Chandler (1962, 1977) and Robert Freeland (2001), the task is not an impossible one (although it often requires a fortuitous combination of resources, in terms of time and sponsorship).

A useful way to structure the analysis would be to apply Burkart Holzner’s (1968) concept of ‘epistemic communities’ to the evolution of a shared managerial understanding within firms as to the nature of the challenges associated with internationalization and the most suitable way to expand activities to foreign countries. As outlined elsewhere (Håkanson, 2007, 2010), epistemic communities can be characterized by their members’ shared mastery of the theory, codes and tools of a common practice. In internationalizing firms, ‘theory’ refers to the shared knowledge infrastructure of explicit cognitive schemata and unarticulated beliefs on which decisions regarding international expansion are made, ‘codes’ to the common vocabulary, whether prescribed in company manuals or in spontaneous vocabulary, in which these decisions are discussed, and ‘tools’ to the physical artefacts, such as computerized information systems, employed in the process.

Archeological discourse analysis based on surviving texts and other evidence, possibly complemented by interview data, could provide insights into the internationalization processes of firms not obtainable in the approaches traditionally employed. This includes the identification of contingencies. The formation of shared beliefs regarding the costs and benefits of different internationalization strategies is not the result of simple relations between causes and effects, but arises as an outcome of complex relationships between people and events. Since these people and the environments in which they operate differ between industries, countries, and over time, a new paradigm should incorporate different outcomes and viable internationalization logics (Håkanson & Kappen, 2017), and the contingencies under which they are likely to arise. Moreover, as Foucault emphasizes, discourse boundaries are not predetermined and static. They cannot be assumed to be limited to individual firms but are most likely common to larger communities, in the form of ‘industry recipes’ (Spender, 1989), or ‘dominant logics’ (Prahalad & Bettis, 1986).

An important limitation of a discourse-based approach to the study of internationalization processes is that the availability of textual evidence, in the form of minutes-of-meetings, internal memos and analyses, annual reports, press releases, written contracts, and other legal documents, for example, is skewed towards larger firms. Nevertheless, initial analysis of such documentary evidence in large firms may (as in the case of the Uppsala model) prove fruitful in generating novel concepts (‘a new language’), with which internationalization processes can be described and analyzed. These concepts would also inform the study of smaller firms, where written records are often scarce, and data need to be collected by other means. Such analysis should include questions regarding the rules that govern the formation of internationalization discourses, the positions and sources of legitimacy of the speakers, the construction of arguments, and their relationships to other discourses, and so on. A significant challenge is to identify also the things that are not said, either because they are unsayable (perhaps even ‘unthinkable’) in the context of the discourse, or because they need not be expressed, since they are part of a common tacit understanding, and therefore self-evident.

SUMMARY AND CONCLUSION

The aim of this Counterpoint is to explore the merits and challenges raised by Treviño and Doh’s (2020) invocation of discourse analysis as a methodology to elucidate the internationalization process of the firm. The proposition is a timely one, since the dominant Uppsala school paradigm, which for four decades so fruitfully informed research on firm internationalization, has now run its course. Drawing on the recent arguments advanced by Jan Johanson and Jan-Erik Vahlne, the originators of the Uppsala school, the Counterpoint explicates how and why the study of the internationalization process of the firm is in need of a paradigm shift. It outlines a possible avenue through which discourse analysis may be instrumental in the development of a new one. The suggested approach differs from the one taken by Treviño and Doh, who seem to view discourse analysis merely as a complementary approach to inherited theory, and whose research questions are, if at all, only marginally related to the internationalization process of the firm.

The argument advanced here is a more radical one. It suggests that by combining the methods of discourse analysis with the concepts and theories in the sociology of knowledge (Berger & Luckmann, 1966; Holzner, 1968; Keller, 2011) a new conceptual basis – perhaps at the end a ‘new paradigm’ – can be developed for the study of the internationalization process of the firm. The space of a short Counterpoint does not allow elaboration of details of such an endeavor, merely hint at some of its possible characteristics. The full-fledged argument will need to be elaborated elsewhere, but meanwhile the sketch outlined here may hopefully encourage the paradigm shift now needed, and provide some ideas as to its possible direction.