History in Process Organization Studies: What, why and how

The chapter highlights the versatile nature –, in an onto‐epistemological and methodological sense –, of historical studiess; and the opportunities a proper historical approach could offer for process scholars. We suggest that such aA historical approach is particularly valuable when studyingstudying processes located in the past or ; when studying relationships between processes and their outcomes. It also allows ; and whento considering other outcomes that were possible at the time and to , examineing why they did not materialize and what the consequences might have been if they had. We emphasize that the use of historical sources requires a very different type of philosophical stance towards reasoning compared to other social sciences and specific types of skills and a very different type of philosophical stance towards reasoning compared to other social sciences. Thus, wEspecially, when making interpretations of the motives and logics of past actors, historians engage in an active reasoning process that results in historical descriptions and conclusions. As for the skills, wA historical approach is valuable when studying processes located in the past; when studying relationships between processes and their outcomes; and when considering other outcomes that were possible at the time, examining why they did not materialize and what the consequences might have been if they had. We Muotoiltu: Taso 1, Riviväli: Kaksois Muotoiltu: Riviväli: Kaksois Muotoiltu: Fontti: Lihavoitu Muotoiltu: Taso 1, Riviväli: Kaksois Muotoiltu: Riviväli: Kaksois Muotoiltu: Tarkista oikeinkirjoitus ja kielioppi


Introduction
There is great potential for a historical approach to process organization studies. On a general level, both historians and process scholars study processes that unfold over time.
Yet, dDespite these similarities and the possible advantages of combining history and process analysis, the literature that actually does this is rather limited. Thus, in-depth analyses of organizational processes and even the broader notion of 'process' are almost entirely absent from the work of business historians (cf. but see e.g. Carter and McKinlay, 2013). Similarly, process management scholars in general and organization theorists in particular have only rarely used history, in the sense of looking at past events and their outcomes. And the few exceptional cases are generally based on published data and/or retrospective interviews rather than primary sources drawn from the archives of the organizations studied (see e.g. Vuori and Huy, 2015; or, for a rare exception of a process study based on archival sources, see Wright and Zammuto, 2013). This has also been highlighted by Kipping and Üsdiken (2014) in their review of history in organization and management theory. While finding that history -either as data or as a theoretical construct -was used more extensively than widely assumed, they pinpointed process theorizing as a Muotoiltu: Taso 1, Riviväli: Kaksois Muotoiltu: Riviväli: Kaksois 6 research, formal methods have tended to come from historical sociology, while for constructivism, they have drawn on the formal analysis of texts in literature and critical studies. Figure 1 summarizes the origins and main features of the four resulting combinations. Each of them will be discussed in some more detail in the following subsections. constructivism. Second, by highlighting the fact that there exists more than one alternative for conducting historical research we continue a tradition in qualitative inquiry to match problems with methods (Langley, 1999), evoking different sets of opportunities for theoretical and historical work in process studies. Third, although the approaches are to some extent incommensurable and offer competing world-views, they are also complementary even with some authors combining different methodologies.

Narrativist approaches
Narrativist approaches combine a constructivist or post-structuralist (Kellner, 1987) ontology that sees historical sources as a reality in their own right with an interpretative, non-formal -often implicit rather than explicit -methodology. Among business historians such approaches have remained relatively rare -despite a number of studies, positing themselves as exemplars, such as the work by Hansen (2007)

Structuralist approaches
In the context of historical process studies, structuralist approaches, sometimes framed as narratological methods and defined as 'the stucturalist study of narrative plots' (Macey, 2001: 263) are rare, especially in business history. In organization studies, structural research designs are more common (e.g. Islam, 2009;Thomas, Wilson and Leeds, 2013) bbutut by no means dominant. Structuralists are interested in objective and value-free structural analysis of texts based on an assumption of universal narratological principles.
Structuralist approaches include a variety of semantic and narratological methods, with (French) structuralism, formalism, and (critical) discourse analysis being the most prevalent. As this Handbook includes specific chapters on narrative methoda (Rantakari and Vaara, 2016 this volume) and discourse theory (Heracleous, 2016this volume) we focus here on the few exemplary historical studies that use structuralist and formalist techniques, and especially on the opportunities such approaches might provide.
There are a few historical studies using critical discourse analysis (e.g. Mueller et al., 2013; Thomas, Wilson and Leeds 2013), typically challenging and problematizing power structures and clashes of ideology (e.g. Cakmakci and Oba, 2007;Vaara, 2002). Yet process 12 structuralist approach is to identify formal semantic structures from specific historical texts in contrast to historical narratives that researchers have created. Reveley's (2010) formal analysis of Australian entrepreneur Jules Joubert's autobiography is a good example.
Overall, while structuralist analyses remain marginal compared to other approaches, the formal analytical techniques and corresponding theoretical logics make it possible to use materials unsuitable for causalist or (con)sequentalist analysis. Moreover, the increasing digitization of historical material allows automated coding of texts (Malec, 2010), which might further motivate the use of structuralist approaches to historical analysis.

(Con)sequentialist approaches
(Con)sequentialist approaches combine a realist ontology with an interpretative methodology. They originated in the first wave of scientific historical research in the 19 th century and have remained prevalent, especially among business historians -albeit in an implicit rather than explicit fashion (De Jong et al., 2015;Kipping, Wadhwani and Bucheli, 2014). We refer to these approaches as (con)sequentialist drawing on Stinchcombe (2005) (Pajunen, 2005;2008). In these studies, Pajunen used event structure analysis and other qualitative comparative analysis (QCA) techniques to study turnaround processes in Finnish forest industry firms. These studies were among the first attempts to integrate history, process perspectives, and formal methodological techniques borrowed from historical and comparative sociology; and thus provide an example of how and why formal analytical methods might work in the context of historical process studies.
While the less formal methods help uncover historical 'truths' or anomalies and develop new, innovative theories, the formal methods of causalists allow scholars to identify patterns that are a/typical across a variety of cases and build more precise theoretical models. For process studies, causalist historical approaches potentially offer many benefits.
First, the basic character of these methods is relatively close to realist case studies as conducted and reported in management journals. However, causalist approaches 16 strengthen the basic case study methods by allowing increased transparency in terms of how sources are used and reported, providing more complete and more authentic sources, and knowledge of the whole life cycle of the particular processes in focus. Second, causalist methods make systematic comparisons across cases possible (e.g., Pajunen, 2004). Thise use of historical source material combined with process tracing techniques, fuzzy set analyses, and other qualitative comparative methods seems of particular value and warrants further consideration.

Why study organizational processes historically? Worlds of possibilities
As shown in the previous section, there is a small corpus of research located at the intersection of process organization studies and history, and it counts some prominent management scholars as authors and advocates. However, these studies have ultimately remained marginal. More can and should be done. Taking a historical approach to studying organizational processes, we would argue, offers a number of possibilities, which are more difficult, if not impossible to achieve through other approaches. In particular, historical research allows scholars to: (i) Examine processes located in the past by definition, such as imprinting or field creating events; (ii) Understand the relationship between processes and their outcomes by studying whole processes with a beginning and an end; We now briefly examine these alternatives, including their advantages and potential pitfalls.

Focusing on past processes
The first and fairly obvious advantage of a historical approach derives from its ability to study processes located in the past by definition, such as imprinting, field creating events, or path dependencies -i.e., in research where history probably has its most natural place.
These studies belong to what Kipping and Üsdiken (2014)  Studying such processes has problems of its own, owing to the difficulty of obtaining data that are akin to observations. Historical data tend to be less comprehensive and may only reflect the parts of the process that have been documented in writing, usually by certain actors for certain purposes. This is why, whenever possible, researchers have looked at processes that took place in the recent past often relying on retrospective interviews, which however have their own limitations due to selective memory (Golden, 1992). For those that did take place in a more distant past, it is important to properly contextualize them and avoid imposing concepts from the present, as was done, for example, in the above mentioneda study on the origins of the Paris Opera where King Louis XIV is characterized rather incongruously as a powerful 'stakeholder' (Johnson 2007). The historical literature has developed methods to address these biases in the available data that we will discuss below, making it possible to examine formative events that are located in a more distant past, where interviews are no longer an option.

Learning from complete processes with known outcomes
The second advantage of historical studies in comparison with contemporary studies of organizational processes is that a historical approach allows the examination of complete processes from their origins to their (final) outcomes. This is especially important for research that tries to make claims about the 'quality' of decisions and organizational actions, i.e. assessing whether or not the processes achieved the intended outcomes, or, more broadly, about the underlying causal structures of processes in terms of how they are unfolding and lead to certain outcomes, intended or not (e.g. Mintzberg and Waters, 198Kipping and Cailluet, 20102). It is also helpful for cases where changes are incremental, which helps explain why they were difficult to detect for actors at the time, while the outcome is only clear with hindsight.
This is a possibility not always considered in current process organization studies, because it requires both a long-term perspective and the benefit of hindsight. One example is Jacobides' (2007)  At the same time, it is important to note that a historical approach to process organization studies is not totally devoid of periodization problems: When does historical causality end??, i.e. when can the past no longer be made responsible for subsequent outcomes or of potential biases, often inherent in the historical sources themselves? Historical methodology has developed tools to address these issues that we discuss later.

Identifying alternative process outcomes
Relatedly, studying the history of whole and complete processes allows researchers to understand which elements were crucial to achieve that particular outcome rather than other possible outcomes, some of which, at the outset might have seemed more plausible. In an in-depth archive-based case study, for instance, Kipping (19920027)  determining what evidence is preserved. In these cases, retrospective interviews do not provide much help either, because those who might have supported alternative solutions will perhaps 'remember' having supported the winning outcome.
In sum, a historical perspective has much to offer to the study of organizational processes and can redress some of the shortcomings and biases prevalent in research focusing purely on the present or the very recent past. However, studying the past is made difficult by the availability and nature of documentary evidence and requires deliberate research strategies and a careful, systematic analysis of the historical sources, as we see next.

How to do historical process organization studies? Research strategies
While the reasons for the lack of interest in process organization studies among business historians are mainly related to the predominant research questions being asked, the Kipping, Wadhwani and Bucheli, 2014). We attempt to offer some elements of guidance here, to help the process researcher better grasp the historical approach.

The nature of historical sources -and their critical analysis
Muotoiltu: Taso 1, Riviväli: Kaksois

24
The key difference for those wanting to study past processes using historical methods is the need to rely on historical sources. This implies the use of 'fragments' (Ginzburg, 1989) or 'relics' (Bryant, 1994) of the past in contrast to 'evidence' as defined by most other social science scholars. Historical sources are not constructed, controlled or manipulated by the researcher. Instead, they 'are, for the most of the part, natural or authentic elements of past social worlds […] from potsherds and pollen deposits to trash heaps and bureaucratic inventories' (Bryant, 1994: 8). As such, they 'represent either the meaningful creations or by-products of social activity […] as residual "traces" or "objectifications" of past human actions and existential conditions' and 'bring us into rather immediate or direct contact with our subjects, "their worlds as experienced"' (ibid.). This means they tend to be highly authentic, yet seldom fully comprehensive (see also Lipartito 2014).
Professional historians have dealt with this by critically scrutinizing each relic of past organizational activity at the moment they are read in order to determine their reliability (Grigg, 1991). No information is treated as a trustworthy sign of the past before its origin, the information it contains, and its relation to other information is scrutinized (e.g. Howell and Prevenier 2001). But trustworthiness in itself is not sufficient: As one of the founders of the micro-historical genre, Carlo Ginzburg (1989: 123), has stated, 'though reality may seem to be opaque, there are privileged zones -signs, clues -which allow us to penetrate it'. Thus, historians typically look at a variety of sources to find 'clues', on which to build their reasoning and verbalization of the processes studied. This sounds easier than it is.
High authenticity means that sources do not cover issues systematically: While the relevant 'signs' and 'clues' might be abundant in some historical sources, they are sparse or entirely absent in others. Moreover, these sources do not have a narrative structure that would generate easily usable insights or that could be manipulated by 'asking the right questions'.
Consequently, working with historical material requires the researcher to embed the historical relics found in their historical context; compare the documents with other sources carrying the same or similar information; and engage in logico-theoretical thinking processes, where the researcher actively works with different historical interpretations and counterfactual reasoning (Bloch, 195392;Kuzminski, 1979). Below, we briefly discuss two of the most pertinent methods used by historians in this respect.

The core of historical methodology: Triangulation and hermeneutics
In addition to source criticism, historians have developed other tools to increase the sources as well as interviews, which not only increased the validity of their findings but also helped them uncover the influence of broader technological and societal changes on decision-making at the firm level -ultimately leading to a conceptualization involving cognitive and path-dependent factors. Quite tellingly, when reconstructing the criteria behind these strategic decisions they placed significant trust in the confidential information obtained from the various company archives.
Hermeneutics moves beyond comparing and combining sources and/or methods to examine the historical context. For historians this generally involves relating each document to other primary sources from the same time period in an iterative process. This helps to better understand the opinions and intentions of the various authors (Howell and Prevenier 2001) or, as some have put it, to reconstruct the 'voice' of that particular document (Decker 2013).
Iteration occurs through re-reading and re-interpreting documents in light of newly discovered documents until interpretations remain stable even if new documents are added to the context -a process generally referred to as 'hermeneutic circle' and facilitated by the use of heterogeneous sources. While the secondary literature can serve as a starting point for this process, the researcher needs to be prepared to question and possibly reject extant interpretations in light of newly found primary sources, which might also shed new light on known documents.
A good example, where management scholars have used this approach, albeit without explicitly characterizing it as such, is the study by Munir and Phillips (2005) on the process of transforming photography into an everyday activity from its highly specialized, almost professional origins, by examining what they refer to as 'the birth of the Kodak moment'. In trying to identify the strategies that the company used to achieve this goal, they rely on primary documents from its archives but stress that these texts were 'not meaningful individually' and only became meaningful by applying a method that 'links texts to discourses and locates both within a particular social and historical context'.

28
While the application of source criticism, triangulation and hermeneutics allows scholars to draw reliable and valid results from historical sources, this still leaves open the question of how to generalize from the insights gained -an area where the gap between the objectives of historians and management scholars seems rather wide, but might not be quite as unbridgeable as it looks.

Going beyond the 'good story': Validation and theorization
Thus, validity for historians is first and foremost achieved through the process of identifying and critically analyzing (primary) sources and is confirmed through triangulation and the use of the hermeneutic circle. But the process does not stop with what could be considered internal validity. Not unlike for other social sciences, there is also some external validation. As Morck and Yeung (2011: 54) succinctly note, 'To sustain credibility, a good historical narrative must connect the "dots" of all relevant historical events with causal links'. But this causality, as they also note (ibid.), is less constraining than in neoclassical economics and, we would add, in the vast majority of management scholarship, since 'history is more amenable to the concept of free will' -even if historians also recognize the influence of impersonal forces in addition to individual decisions.
This leaves, however, the already noted issue of how to learn from historical cases, especially if these are single cases (March, Sproull and Tamuz, 1991). In terms of generating theoretical insights, a historical research strategy can be problematic. The cause-effect relationships identified tend to be complex, reflecting the 'reality' of organizational processes. Thus, when business historians explain failure they typically find from five to ten This does not mean that historians necessarily eschew theory and theoretical reasoning completely as has been noted by David, Sine and Haveman (2013: 361) who point out that 'historians begin with some theory, which guides them in their selection of evidence to scrutinize. They then move between their reading of the evidence and their understanding of theory, updating theory, and updating their search for evidence to investigate new ideas as they arise.' Theoretical pre-understanding, this suggests, is essential for historical researchers to create any meaningful interpretation of the historical 'relics' they find in the archives.
Another way to understand the way results emerge is to see the work of the historian as continuous theorizing (Murphey, 1973), i.e., in some sense as a comparative method in which the comparative judgment occurs between the theoretical knowledge and the historical interpretation (von Leyden, 1984;Walker, 1980). As the archival sources lack a 30 coherent structure and/or narrative form, each piece of information must be woven into a process model in which information and events have links over time and in time (Griffin, 1993). It should be noted though that this kind of theorizing happens largely independently of the literature that motivates the researcher to go to the archive. By contrast in the other social sciences, theoretical motivations are more explicit. For the theorizing process to make sense for both business historians and management scholars it has to be highly iterative with the simultaneous reading of theories and archival research stimulating new ways to read the literature and the archival sources (Shanahan 1989).

Conclusion
We started by examining the reasons why process organization studies and historical research (as historians understand it) have remained largely isolated streams of research.
One major reason for the lack of cross-fertilization is that, on average, management scholars have only a very vague understanding of the requirements and advantages of the historical method. This is due to the lack of proper methods books, articles, and PhD courses on these methods. The situation has improved somewhat after the first calls for more history in the 1990s by the likes of Zald there are also camps that favor narrative analysis, formal structural analyses of texts, comparative methods, and process tracing techniques. As we have stressed, however, the use of historical sources requires specific types of skills and a very different type of philosophical stance towards reasoning compared to other social sciences. Especially, when making interpretations of the motives and logics of past actors, historians engage in an active reasoning process that results in historical descriptions and conclusions. The researcher's role is elemental not only as an observer but more importantly as a mechanism kind of mediator between the relics of the past and contemporary scholarly discussions.
Second, we have offered reasons for using a historical approach and we have suggested ways of conducting such research. A historical approach is especially valuable when studying processes located in the past 'by definition' (such as imprinting or field creating events and ); when studying relationships between processes and their outcomes. It also to make historical processes more accessible to process researchers. We have therefore highlighted the importance of source criticism, triangulation, hermeneutic treatment and constant iterative theorizing, and hermeneutic treatment of materials in the making of historical analyses of processes. Perhaps more importantly, our text hopefully functions as a reminder of the innovative historical research that has already been conducted in the field of process organization studies and offers some alternative ways forward.