Abstract
How do paradoxical tensions become salient in organizations over time? Ambidexterity and paradox studies have, thus far, primarily focused on how tensions inside organizations are managed after they have been rendered salient for actors. Using a longitudinal, embedded case study of four strategic business units within a media organization, we theorize the role of the top management team leader’s practices in enabling tensions to become salient for their respective lower-level managers when there are initial differences in how tensions are interpreted across levels. Our findings extend a dynamic equilibrium model of organizing by adding interpretive context as an enabling condition that shapes the emergence of salience through the provision of a constellation of cues that guide sensemaking. Informed by a practice-based perspective on paradox, we also contribute a conceptual model of leadership as practice, and outline the implications for ambidexterity studies.
Organizations are rife with tensions, but some tensions are strategically more important than others (Cameron & Quinn, 1988; Poole & Van de Ven, 1989). In particular, paradoxical tensions represent demands that are contradictory but must be pursued interdependently to sustain long-term organizational performance (Lewis, 2000; Smith & Lewis, 2011). Although a growing number of studies have begun to unpack how actors respond to paradoxical tensions once they become aware of them (Andriopoulos & Lewis, 2009; Jarzabkowski, Lê, & Van de Ven, 2013), we still know relatively little about the early stages of how paradoxical tensions become salient in the first place. Such an examination is important for theoretical and practical reasons. Theoretically, identifying changes between the early and later stages in dealing with paradox can provide important clues about the nature and construction of paradox in organizations (Smith & Lewis, 2011). Practically, the timely and effective recognition of paradoxical tensions is important for managers who seek to drive improvements in organizational performance (Lewis, Andriopoulos, & Smith, 2014; Smith, 2014). We know from prior work that actors are better placed to have an engaged response to paradox once they are made aware of the tension (Luscher & Lewis, 2008; Miron-Spektor, Gino, & Argote, 2011). Yet, the existence of tension in the environment is not enough to automatically trigger action: tensions can remain latent and go unperceived or unnoticed by organizational actors for years (Dutton & Dukerich, 1991; Gilbert, 2005). Our paper focuses on this issue, and specifically the role of the TMT leader in making paradoxical tensions salient to others.
Paradox theory adopts two underlying assumptions about the nature of paradox that are relevant in this regard (Lewis & Smith, 2014). First, paradoxical tensions are inherent in organizational life and exist ‘beyond the will or power of management’ (Clegg, Cunha, & Cunha, 2002, p. 484). Accordingly, paradoxical tensions emanate from underlying and interdependent contradictions in organizational systems such that organizational actors have no choice but to deal with them. Second, paradoxes are also cognitively and socially constructed as actors perceive the relationship between poles via paradoxical cognition (Smith & Tushman, 2005). As such, it is actors’ recognition of the paradoxical tensions that renders paradoxes salient (Lewis, 2000). A dynamic equilibrium model of organizing brings these two perspectives together, positing that system-level contradictions and sensemaking processes come together to render latent tensions salient when (a) there are changes in environmental conditions of plurality, scarcity and change, and (b) actors apply paradoxical cognition (Smith & Lewis, 2011). Yet, the model leaves open what intermediates the relationship between these two conditions. We know, for example, that rhetorical (Jarzabkowski et al., 2013) or discursive interactions (Abdallah, Denis, & Langley, 2011) can socially construct paradoxes through micro instantiations of tension that can have ripple effects elsewhere across the organization (Luscher & Lewis, 2008). Yet, actors may also lack a shared understanding of these mixed messages on occasion, raising questions about how salience emerges in these instances (Dutton & Dukerich, 1991; Dutton, Ashford, Lawrence, & Miner-Rubino, 2002). How can these positions be reconciled?
To examine this question, we adopt a practice-based perspective (Schatzki, 2002; Whittington, Molloy, Mayer, & Smith, 2006) focused on the leadership practices of individual top management team leaders (TMT leaders) within the embedded case study of MediaCo. We define the TMT leader as the (one) leader in charge of a strategic business unit (SBU); and we define SBUs as independent business units operating within a parent company. A practice theory approach conceptualizes paradoxes as being enmeshed within everyday activities, as actors construct and make sense of tensions (Lê & Bednarek, forthcoming). However, a focus on leadership practices recognizes that not all organizational actors have equal responsibility for issues (Carroll, Levy, & Richmond, 2008), and that the micro-activities of a leader can have more far-reaching effects, for example, through resource allocation, organizational design and product design across the organization (Smith, 2014). Our paper was inspired by observing this first-hand over the course of our ethnographic study. At the start of our study, paradoxical tensions between exploration and exploitation were embedded within the media sector at large but were only partially perceived by lower-level managers within MediaCo. Over a 24-month period, we observed how individual TMT leaders shifted these managers’ appreciation of paradoxical tensions through the activities they orchestrated within their respective SBUs. Our research was therefore motivated by asking: How does the TMT leader enable latent paradoxical tensions to become salient for lower-level managers through their leadership practices?
Our findings show that the TMT leader’s practices are central to the interactional dynamics that rendered paradox salient for lower-level managers. Here, we define salience as when an organizational actor appreciates the relationship between alternate poles as both contradictory as well as interrelated (Smith & Lewis, 2011; see also Lewis, 2000). Our findings also demonstrate that the TMT leader constructs ‘interpretive contexts’, which draw attention to cues that prime actors’ awareness of the contradiction and interrelatedness between poles. These ‘interpretive contexts’ represent repeated and converging cues that set the sensemaking in motion for lower-level managers, and culminate in salience (Dutton et al., 2002; Weber & Glynn, 2006). Our work provides an important contribution to paradox theory, by showing how the social construction of paradox is not only facilitated through the micro-instantiations of paradox, but also the context within which the language is situated over time. In this respect, the ‘interpretive context’ is neither the inherent system-level structures, nor the sensemaking about the system. Rather it is a constellation of cues that can become objectified in social processes over time and allow the TMT leader to bridge the gap for lower-level managers between system-level contradictions and their paradoxical cognition.
Nature and Temporal Dynamics of Paradoxical Tensions
Paradoxes are ‘contradictory yet interrelated elements that appear simultaneously and persist over time’ (Smith & Lewis 2011, p. 382). These contradictory elements are inherent in organizing systems, and include tensions between maximizing profits and improving social welfare (Margolis & Walsh, 2003), global integration and local adaptation (Marquis & Battilana, 2009) and exploration and exploitation (Farjoun, 2010; March, 1991), among others. An extensive body of work has examined the responses that actors deploy when confronted with paradoxical tensions (Poole & Van de Ven, 1989; Smith & Berg, 1987; Vince & Broussine, 1996). More recently, this work has been complemented by practice scholars who have taken growing interest in the earlier stages of social construction. Here, the iterative and dynamic micro-interactions between actors are theorized as consequential in rendering paradoxes salient, and facilitating ongoing attention to both poles. However, most of this work has concentrated on rhetorical practices that actors deploy to ‘work through’ ambiguities in the present (Luscher & Lewis, 2008), which has overshadowed the opportunity to also examine the larger, social processes in the organization which guide the pace and sequence of sensemaking taking place (Weick, Sutcliffe, & Obstfeld, 2005).
A key characteristic of prior paradox studies, for example, is that they have tended to focus on lower-level managers (Jarzabkowski & Sillince, 2007) in contexts where managers have no choice but to respond to decisions already enacted by leaders (Andriopoulos & Lewis, 2009; Jarzabkowski et al., 2013; Smith, 2014). For example, Luscher and Lewis’s (2008) study of middle managers took place several months after a comprehensive restructure had been carried out by the CEO, and overlooked the ongoing tactical efforts that the CEO may have undertaken to shape sensemaking. However, the authors suggested that the external facilitators in their action research method may have been vital to paradoxical inquiry because of their ‘viewpoint unencumbered by daily managerial responsibilities’ (ibid. p. 235), hinting that larger situational contexts may shape sensemaking. Smith’s (2014) study of TMT leaders across six SBUs in a technology company has shifted the focus to leaders who have the responsibility to decide over these processes, but the study is largely focused on the decision-making practices within these top management teams rather than between actors with access to different types of information.
This presents the research gap explored in this paper: if the strategic importance of paradoxes as an organizational phenomenon is to be realized, we need an understanding of not only the sensemaking practices that individuals use to ‘appraise the stimuli’ (Margolis & Walsh, 2003, p. 285), but also how certain cues come to the attention of actors to enable paradoxes to become salient in the first place. This is an important distinction, since the presence of inherent tensions and their social construction need not occur simultaneously. In a study of the New York Port Authority, for example, actors took over seven years to recognize that systemic homelessness in their local environment had important implications for the organization’s identity, thereby eliciting a response (Dutton & Dukerich, 1991). Other studies suggest that recognizing the relationship between poles may be delayed as actors initially perceive alternate poles as complementary or even unrelated (Mantere, Schildt, & Sillince, 2012). At issue here is a more nuanced understanding of how and when actors recognize paradoxical tensions through their cognition (Miron-Spektor et al., 2011; Smith & Tushman, 2005); and, therefore, how paradoxes become salient.
To focus our inquiry on the situated aspects of these processes, we draw on Weber and Glynn’s (2006) appreciation of sensemaking as embedded within social space and time. Rather than perceiving system-level contradictions as imposing cognitive constraints on the actors doing the sensemaking, Weber and Glynn (2006) argue that actors face a constellation of sensemaking cues and multiple corresponding roles and actions. Thus, it is for the local context (the one closest to the particular actors) to supply the cues that prime action and provide a guide to future actions. For example, actors use rhetoric to support claims that over time ‘can become taken for granted and objectified. Once objectified, these claims can function as the backing that grounds future argumentation’ (Harmon, Green, & Goodnight, 2015. p. 88). Building on this work, we define ‘interpretive contexts’ as the repeated and converging combinations of cues that are created by leaders to direct attention to particular issues that motivate sensemaking by lower-level maangers (see Weber & Glynn, 2006). These cues are formed through leadership practices in interaction with their associated artefacts (e.g. product development documents), which may then become diffused and institutionalized at the intrafirm and interfirm levels as they become formalized into social processes (Harmon et al., 2015). This is consistent with a practice-based approach in which everyday activities are inextricably intertwined with structuring processes that spur action, yet our study places more attention on how these interpretive contexts are created in their early stages. Thus, even though actions can be studied through the everyday, rhetorical practices of actors, a focus on ‘interpretive contexts’ allows certain actors (such as leaders) to instil contexts that reiterate certain cues beyond others (Hardy & Thomas, 2014).
The Role of Leaders in Exploration and Exploitation: Ambidexterity and Paradox Views
Our study focuses on the organizing paradox of exploration and exploitation (Smith, 2014). These represent two fundamentally different activities, whereby exploitation involves ‘refinement, efficiency, selection and implementation’ and exploration involves ‘search, variation, experimentation and innovation’ (March 1991, p. 71). These tensions constitute a strategic paradox since their long-term management are of specific importance to an organization’s goals (Smith, 2014).
Building on the interest in ambidexterity research, we take the practices of individual TMT leaders as our level of analysis. Ambidexterity scholars highlight the importance of TMT leaders in overseeing the structuring decisions over resource allocation, product design and organizational structure (Lavie, Stettner, & Tushman, 2010). Organizations differentiate between competing demands through temporal (Nickerson & Zenger, 2002; Siggelkow & Levinthal, 2003) or structural separation (O’Reilly & Tushman, 2004); and TMT leaders play an important role in enabling differentiation, but also support simultaneous integration across units and time periods (Smith & Tushman, 2005). O’Reilly and Tushman (2004) argue that top management teams act as the ‘corporate glue’, and set an integrative vision to motivate organizational actors. Jansen, Tempelaar, Van den Bosch and Volberda (2009) show the need for social integration across the senior team to coordinate structurally separated units (Jansen et al., 2009), while Gibson and Birkinshaw (2004) point to the importance of senior leaders across levels, especially in creating a supportive context in the business units where lower-level managers experience a culture of stretch, trust, discipline and support.
In developing this latter focus on how context enables ambidexterity, Zimmermann, Raisch and Birkinshaw (2015) show how relational initiatives shape context by overcoming political and trust-based tensions as organizations move from a one-sided (exploitation or exploration) to an ambidextrous (exploitation and exploration) charter. Their study provides a key advancement in our understanding of ambidexterity as it demonstrates the importance of lower-level managers in shaping an appropriate context. However, lower-level managers initiated these practices in their study after they experienced dissonance. Thus, the prior step of how the TMT leader helped lower-level managers to recognize paradoxical tensions in the first place was not addressed (Zimmerman et al., 2015).
Furthermore, the ambidexterity literature adopts a meta-theoretical lens in which leaders select or switch between alternative contingent approaches (Papachroni, Heracleous, Paroutis, 2015; Smith & Lewis, 2014). Taking a paradox perspective, then, has the potential to complement current ambidexterity studies with an understanding of the early stages of the process through which the TMT leader’s behaviours create the conditions that support synthesis between alternate poles. Initial work in this area shows how TMT leaders take dynamic decisions in order to manage strategic paradoxes within their own teams (Smith, 2014), but more studies are needed to understand how these practices extend across organizational levels (Papachroni, Heracleous, & Paroutis, 2016). Taken together, the above literature and our understanding of the gaps in relation to the nature of salience inform our research question, which we frame as follows: How does the TMT leader enable latent paradoxical tensions to become salient for lower-level managers through their leadership practices?
Method
Longitudinal case studies offer an ideal way to examine poorly understood phenomena (Birkinshaw, Brannen, & Tung, 2011; Eisenhardt & Graebner, 2007). Within this, we adopt a practice perspective that sensitizes our analysis to ‘a deeper understanding of micro-processes and of the interplay between culture and context in the collaboration and integration of activities’ (Birkinshaw, et al., 2011, p. 43).
Research setting
Our study is situated in the media sector, which is an appropriate setting for studying inherent exploration and exploitation tensions (Gilbert, 2005; O’Reilly & Tushman, 2004; Smith, Binns, & Tushman, 2010). These tensions are inherent in the sense that technological changes are transforming the traditional business model for newspaper publishing and television broadcast: traditional media businesses have depended on sales of the physical newspaper or a live broadcast schedule for revenue, since these technologies enable passive distribution of advertising and news content. However, in a digital environment where consumers can be hyper-selective about what they watch and read, the audience for passively distributed content has disappeared. While traditional media companies may seek to ignore digital distribution channels to preserve their established businesses, they also ultimately depend on them to reach customers as consumption patterns change. This presents a paradox between print and digital agendas: digital competes with print, but print needs digital to survive.
MediaCo was a leading national media company, and operated over 100 daily, Sunday and bi-weekly newspapers as well as a subscription TV service. Thus, it was deeply embedded within the traditional media business model. The company was structured as four SBUs: PrintSBU, BroadcastSBU, MagazineSBU and MarketplaceSBU. Each SBU had their own general manager, who we define as a TMT leader since they reported directly to the CEO of MediaCo, the parent company. We define managers below the TMT leaders as ‘lower-level managers’.
In the late 2000s, MediaCo’s board committed to a long-term strategy to move the organization towards embracing both a traditional print and a digital future, leading the Chairman to announce this new strategy to shareholders:
At [MediaCo], where we’re both a video programmer as well as a newspaper publisher, the rewards of getting this right are enormous. We’ve spent billions of dollars developing unique sports, news and general entertainment programming. We have a library as rich as anyone in this world. Our job now is to bring this content profitably into the broadband world – to marry our video to our publishing assets, and to garner our fair share – hopefully more than our fair share – of the advertising dollars that will come from successfully converging these media. [emphasis added]
Although the board had embraced this shift, MediaCo’s long-standing and much respected CEO had remained sceptical. Having built his career as a print journalist, the CEO continued to organize the company’s strategy around delivering high-quality print journalism and restricted investment into the digital agenda. As revenues from the print mastheads continued to decline, the CEO was eventually sacked by the board and replaced by a new CEO with a background in digital broadcasting. This study commenced soon after the appointment of the new CEO and following a subsequent series of hires to the top management team in which the ‘old guard’ was replaced by TMT leaders charged with delivering the company’s digital transformation.
Data collection
Qualitative data was collected during a 24-month period, in which one year was captured retrospectively through interviews and archival materials and one year was in real time. Real-time data collection began 6 months after the appointment of the new CEO. Data collection commenced within PrintSBU, which was MediaCo’s largest SBU and responsible for 70% of MediaCo’s overall revenues. Although each masthead newspaper had a freely accessible website, none earned revenue through digital subscriptions at the commencement of the study.
When we started collecting data, PrintSBU had appointed a new TMT leader, Chris. As our understanding of the PrintSBU context deepened and we built trust, data collection was extended to three other SBUs within MediaCo. These were smaller operations by revenue and number of employees but were tasked with the same company strategy. MagazineSBU, led by Sophie (TMT leader), was most similar to PrintSBU in that it operated a portfolio of glossy lifestyle magazines, though none earned digital subscription revenue at the start of our study. BroadcastSBU, led by Lev (TMT leader), operated a subscription television network and had no digital subscription revenue from streamed content. Finally, MarketplaceSBU, led by Mark (TMT leader), was a small retail business unit, which operated affiliated services such as events marketing and loyalty programmes within MediaCo. Table 1 summarizes the case context.
Exploitation and Exploration Tensions across the Four MediaCo SBUs.
To avoid biases from a single data source, a range of field methods were used to triangulate findings, including observations, interviews and archival documents (Eisenhardt, 1989; Yin, 1994).
Observations
We observed 36 meetings across the case site, including TMT leaders’ meetings, SBU strategy workshops, and regular team meetings within each SBU. These meetings lasted between 1 and 4 hours, and allowed us to observe interactions within the TMT leadership, as well as by and between individual TMT leaders and lower-level managers as tensions emerged. Prolonged engagement with the site enabled a deeper understanding of the context (Lincoln & Guba, 1985), which we complemented with interviews to aid our interpretation of interactions. Because of the sensitive nature of issues discussed, not all meetings were recorded (Miller, Cardinal, & Glick, 1997) but extensive notes were taken in real time and written up within 24 hours of the meetings (Miles & Snow, 1978).
Interviews
A total of 57 open-ended interviews with 41 distinct informants were conducted. This included 16 serial interviews. Since our paper focuses on how the TMT leader interacted with lower-level managers within their respective SBUs, interviews commenced with the TMT leader and proceeded to lower-level managers. Interviews lasted between 30 and 90 minutes, and were fully recorded and transcribed. Respondents were asked to comment on initiatives related to both the print and digital business, including contradictions, tensions and ambiguities related to these demands (Miles & Huberman, 1994). This supported our understanding of the purpose behind leadership efforts, as well as how these were perceived by lower-level managers. Interviews were conducted in serial with key respondents and spaced with 3- to 4-month gaps to enable sufficient time for new reflections and interactions to emerge from the case environment (Jarzabkowski, Lê, & Feldman, 2011).
Archival materials
Finally, to enable further triangulation and increase reliability, we also collected 1,544 pages of archival documents across the four businesses (Eisenhardt, 1989; Jick, 1979). This included internal email communications, strategy documents and PowerPoint presentations. These documents were important as the TMT leader was often short of time and therefore corresponded with lower-level managers remotely through emails, or edits to PowerPoints (Kaplan, 2010). Analysis of this archival material aided understanding of context, strategy and outcomes. A summary of the data corpus is recorded in Table 2.
Data Collected.
Data analysis
We used three distinct stages in order to systematically move from raw data to theoretical insight (Gioia, Corley, & Hamilton, 2013). In the first stage, we developed a rich chronology of the case site over the 24-month period of our study, covering both the real-time and the retrospective data collection periods (Langley, 1999). This data was rendered through thick description and provided insight about the business context, leadership behaviours and outcomes from those behaviours within each respective business unit (Geertz, 1973; Yin, 1994). We then shared these case studies with key informants to validate veracity and enhance robustness (Lincoln & Guba, 1985).
In the second stage, we focused on specific issues in which tensions between exploring and exploiting were present for each TMT leader, being issues of resource allocation, organizational design and product design (Smith, 2014). Here, we examined how tensions and contradictions surfaced in meeting transcripts by using Andriopoulos and Lewis’s (2009) approach of coding for language indicators in respondents’ own words, such as: ‘yet’, ‘but’, ‘problem’, ‘alternative’, ‘tension’, ‘would like…but’, and ‘should…but’. This generated 56 instances of tensions, from which we then identified a subset of 34 instances specifically related to exploration and exploitation.
Using these tensions, we analysed the interview transcripts to code for and identify how respondents interacted with each other to escalate or accentuate recognition of a pole. Initially, we concentrated on the practices of the TMT leader, working between the data and the paradox literature. Early coding on ‘integrating’ and ‘differentiating’ (Andriopoulos & Lewis, 2009; Lawrence & Lorsch, 1967) was later discarded as greater concentration was placed on the early stages, when paradoxical tensions were not yet front of mind for lower-level managers. This yielded codes of diversifying, devaluing and multi-tasking practices related to supporting the alternate pole. These codes emerged as we moved from one case study to the next, focusing on patterns between cases as well as noting any differences.
In a third stage, we expanded our coding as it became evident that lower-level managers used meetings with their TMT leaders to raise attention to the dominant pole. Thus, rather than only promoting the alternate pole, we realized that the TMT leader also gave attention to the dominant pole. Here, we were guided by examining transcripts with two questions in mind (Jarzabkowski et al., 2011). The first question – ‘How does this TMT leader’s practice support the lower-level managers’ interpretive understanding of the dominant pole?’ – helped us to understand recognition of exploitation. The second question – ‘How does this TMT leader’s practice support lower-level managers’ interpretive understanding of the alternate pole?’ – allowed us to focus on recognition of exploration. Examining TMT and lower-level managers’ interactions with this in mind enabled us to develop a second set of TMT practices supporting the dominant pole, being consolidating, supporting and prioritizing.
As we examined how and when leaders shifted between practices over the duration of our study, we realized that these practices emphasised three distinct types of cues – what we collectively call the interpretive context. We defined the constellation of cues related to incentives and rewards as the instrumental context, which involved things leaders said and did around strategic plan documents, key performance indicators and subscription targets. TMT leaders used these cues to prime lower-level managers’ actions in relation to performance expectations. We defined the constellation of cues around roles and task formation as the relational context, which included things leaders said and did around job descriptions, product design plans and organizational charts. TMT leaders used these to prime lower-level managers’ actions towards performing particular roles or tasks. Finally, we defined the constellation of cues related to resource constraints as the temporal context, which included things leaders said and did around project planning deadlines, launch events and schedules. TMT leaders used these cues to prime the priorizations of actions. In what follows, we structure our findings by presenting data on how the TMT leader’s practices constructed the interpretive context to draw attention to the contradictory yet interdependent relationship between poles (salience), and then examine how the contexts are dynamically related.
Findings
The goal of our study was to understand how the TMT leader enabled inherent though latent paradoxical tensions to become salient for lower-level managers. Although MediaCo was composed of four SBUs, we observed significant similarity among the TMT leaders at the head of each SBU. While this was initially surprising, we account for this based on the fact that each was embedded within the same organizational context and responded to similar timelines within the strategic plan, as agreed by MediaCo’s CEO.
We first present the process story through PrintSBU, but demonstrate the richness of our data with illustrative examples from the other case settings and in the supporting Table 3. We show how the TMT leader’s practices constructed an interpretive context by supplying cues that primed sensemaking through three related contexts. In the second part of our findings we draw on data from across our cases to highlight the relationship between the contexts, showing how leaders maintained salience by building interpretive linkages between contexts and over time.
Representative Data of Practices Enabling Recognition of Paradoxical Salience.
TMT leader’s practices shape the interpretive context for salience
Leader’s practices shaping the instrumental context
At the start of our study, PrintSBU’s print newsroom was the primary process around which daily workflow and projects were organized by lower-level managers. Editors and journalists planned their work around the layout of the physical newspaper, with all stories filed by 5 pm daily to reach the ink printers. To the extent digitally produced content featured in the day-to-day incentives of managers, it was as a ‘training’ ground for new journalists. Thus, goals and incentives were framed within the context of supporting exploitation of the print business model only. As one PrintSBU executive reflected:
Digital is very much seen as the training wheels for the main game. Anything to do with promotion, awards, and getting an editor’s job…is still based on whether you have cut your teeth in the print business.
Chris challenged managers to recognize the alternate pole by priming sensemaking through the use of a strategic plan document, which was continuously resurfaced as a sensemaking tool in meetings. We describe the provisioning of cues related to incentives as constructing the instrumental context. We describe the set of TMT practices differentiating from existing incentives and interests as diversifying because they diverged from existing rewards (i.e. support exploration), and consolidating as practices that reinforced existing rewards systems and interests (i.e. supported exploitation).
For example, shortly after being appointed, Chris organized a strategy workshop to discuss digital plans under MediaCo’s annual strategic plan document. Chris asked lower-level managers to describe their current responsibilities, so that they would articulate underlying assumptions about their existing roles. Chris then directed lower-level managers to respond to the following statement contained within the strategic plan document: ‘We are moving from a reading to a viewing led world. The reality is that people are not reading the newspaper anymore.’
This statement from the document highlighted contradiction between the newspaper business and new digital projects. Even as managers recognized contradiction, they failed to recognize interdependence and the implications of digital for their own priorities. As one marketing manager commented in the meeting:
We have a full pipeline of things that we need to do over the next 6 months…but we can’t even do those yet, so I don’t know how we’re going to be able fit in more projects before the end of the year.
This cast print and digital as a trade-off choice. Chris responded to this interpretation of the strategy document by giving sense to the synergistic relationship between print and digital agendas. For example, rather than promote print or digital exclusively, Chris described the future goal of PrintSBU, as articulated in the document, as ‘entertainment’ and ‘media’. This compassed both priorities:
We have to wake up and realize we’re not a news business; we’re a media and entertainment business. The sooner we realized that, the sooner we are able to lift our game.
Chris deployed the annual strategy plan document, then, to spur managers to make sense of print and digital agendas in the context of their own annual performance templates. Provoked by this cue, one manager retorted the following to Chris in a team meeting:
But the bigger question is what are consumers willing to pay for when they are getting a lot of this stuff online for free already. And how do we ensure we are not disintermediated [sic] by retailers? I can tell you that [clients] are both happy to support ads in paper (although reducing) but pay next to nothing for content marketing and communication about their digital sales channel. That’s true… [But] You are the only content person in the business with a brain big enough to solve that quandary, and who can directly influence her content output. But I’m happy to sit in a room and whiteboard this stuff with you.
Chris’s response acknowledged the contradictions articulated (‘That’s true’), but forced managers to confront these tensions for themselves (‘You are the only person…with a brain big enough to solve that quandary’). Therefore in moving between the diversifying and consolidating practices, Chris recognized business-as-usual priorities but put pressure on lower-level managers to address both tensions in their goals and incentives (instrumental context), rather than giving them either/or orders.
Diversifying and consolidating practices also primed the instrumental context in the other SBUs. At MagazineSBU, Sophie organized a number of guest lectures with digital-savvy speakers for her staff to evaluate digital thinking. She justified this diversifying practice as follows:
We need a way for our people to access creativity outside the existing business to drive new sources of competitive advantage, and expose internal employees to new ways of thinking about digital product and content innovation … We’re so used to doing things in a particular way that we keep doing them, not because they are the right thing to do, but because it’s what people are used to.
Managers were then directed to incorporate ideas from the digital speakers in their key performance indicators. The guest lectures therefore served as a sensemaking cue, which Sophie leveraged to both confront contradictory agendas among lower-level managers, and enmesh digital within existing performance frameworks. At BroadcastSBU, the instrumental context was primed by Lev through a series of strategic reviews that forced lower-level managers to appraise the business’s market position. BroadcastSBU provided a broadcast television service but was losing market share to digitally streamed content. Lev spurred lower-level managers to brainstorm new ideas through the strategic review. This forced them to confront problems within their own performance metrics and goal templates. As Lev cajoled in one meeting:
There is a lot of white space in the mobile TV market [for us to do exploration]. [Our main competitor] is now entering this market so unless we do something aggressively or we will all be out of a job.
This diversifying practice dovetailed with a consolidating practice, which forced managers to incorporate new ideas within their existing workplans.
Leader’s practices shaping the relational context
In PrintSBU, the brainstorming and strategy workshops resulted in a product design document outlining the functional specifications for the new digital pay wall, and the roles of various lower-level managers therein. This document was a sensemaking cue that allowed Chris to construct the relational context, being managers’ individual roles and their daily tasks. However, debates over what to include within the product design document primed sensemaking as Chris shifted attention between alternate poles. Devaluing practices attributed prestige to roles pursuing the digital agenda (i.e. supporting exploration), whereas supporting practices validated the social worth of actors engaged in traditional print journalism (i.e. supporting exploitation). Deploying both practices not only highlighted contradiction between alternate poles but also forced managers to consider how digital was interdependent with print now and in the future.
For example, a number of PrintSBU managers had sought to discredit the product design for the digital pay wall, arguing that it was a fleeting idea that lacked political weight. One manager framed the idea of the pay wall in an interview as follows:
We’ve got to stop this ‘sample of one’ approach which is ‘I think this is brilliant, therefore it’s brilliant’. The [digital] idea never gets properly formed and it gets killed… Too much of what we see is a thought bubble from somewhere in the business, or some executive’s pet project. It’s not part of our core business, and unless it relates to our core products it’s not going to get supported.
Chris rebuffed this suggestion, devaluing this managers’ opinion and emphasizing the CEO’s personal commitment to the pay wall design. For example, in a strategy meeting to plan technical upgrades being developed under the product design document, Chris contemplated the opportunity for the CEO to showcase the social worth he gave to the digital initiative by reiterating his financial commitment through an announcement:
It could be worth thinking about the number that we can promote. I don’t know what it is but there is something quite powerful about [the MediaCo CEO] saying he is willing to invest, say $100m over the next 3 years for [the digital pay wall]. How that gets broken down is t.b.c. but it makes it clear what funding is up for grabs.
This was designed to prime lower-level managers’ appreciation of digital. However, at the same time, Chris’ deployed supporting practices to keep print journalists engaged by reifying the company’s commitment to its traditional roots. For example, within the product design plan, a functional specification needed to be agreed about what media content would be published on the paywall. Rather than devalue the print agenda, Chris supported their opinions for the design. As Chris reflected in one meeting in relation to wireframe designs that over-emphasized digital content:
It depends on the audience a bit. I worry that [the design] is a bit out there for the majority of the business and for some executives. I can just see [the print executives] rolling their eyes. So it probably needs a really simple articulation of what this means in practice. What are the investment requirements and what will I get back for that – in words of one syllable… We can then get [the print executives] feedback.
By using the wireframe design as a sensemaking tool to engage print executives, Chris illustrated his intention to reconcile tensions between print and digital managers through constructive engagement and amendments to the document. An important issue in these subsequent negotiations was how print journalists would be employed under the pay wall model, since their written product would now be used twice (in print and digital). William, a sceptical print manger, pushed for a pay rise for print journalists who contributed to digital. Rather than reject this proposal, Chris supported this by interpreting it within the relational context. Chris was willing to concede further budget support for print journalists (exploitation) on the condition that print journalists were willing to be more proactively engaged in producing digital articles (exploration), as the following exchange in a meeting illustrates:
We need to realize that our talent is still predominantly print talent, so we still need to recognize that we will own commentary and news in sport. That’s what we’re good at and that is why people read the [masthead newspaper] OK, so what I think I’m hearing is that we like [the digital innovation] but I need to be more specific around what we need to do [to integrate it into print, and reward journalists]. Yes, we need something we can take back to the [journalists] and which they can live with. [PrintSBU] is always going to have [anchor journalist] so we need to dial that up so they can go along with the new stuff.
This facilitated both/and sensemaking as the digital product document was used as a cue to prime attention to both the contradiction yet interrelatedness between both poles. In BroadcastSBU, Lev also primed sensemaking about paradoxical tensions through wireframe designs for the digital broadcast product. A specific issue was how to deal with customers’ privacy for streamed digital content. Digital managers supported streaming video content on mobile devices because it allowed BroadcastSBU to collect personalized information about consumers’ behaviours – which was commercially valuable for the digital team. However, rather than accommodate this exploration, traditional broadcast managers identified reasons why privacy was a major technical barrier for the digital project progressing. Perceiving this debate as a tension over whose opinions counted the most (that is, relational context), Lev pursued devaluing and supporting practices to highlight contradiction and interdependence. As he reflected:
The best way to get around these problems [between the business and technical managers] is to just show that customers really want this, and then you can make the case. So you don’t want to get into a tech discussion about whether or not this is possible [to fix privacy] because that will not get you anywhere. In the end, the business needs to be led by the customers, not by the tech.
Lev supported traditional managers by noting that their concerns were not prima facie invalid. However, he also devalued commitment to existing processes without further research, requesting additional information from customers on their needs. This enabled both/and solutions to emerge: lower-level managers enabled digital content on mobile devices, but only after customers gave permission to collect private information.
Leader’s practices shaping the temporal context
As the deadline for the PrintSBU pay wall launch approached, lower-level managers experienced tension around how to prioritize their time in relation to existing responsibilities. These project deadlines were socially constructed cues supplied by TMT leaders to prime managers’ interpretive understanding of their priorities – what we term temporal context. In constructing these project plans, TMT leaders deployed prioritizing practices to focus time allocation on print agendas (supporting exploitation), and multi-tasking practices to divert focus to digital agendas (supporting exploration). This forced lower-level managers to address the contradiction yet interrelatedness between both poles.
For example, within PrintSBU, one of the marketing teams wrestled with how to meet their existing campaign commitments for the printed newspapers as well as Chris’ deadlines to launch advertising for the digital pay wall. Chris used the deadline to prioritize the team on meeting its print commitment, while also multi-tasking by reinforcing its need to satisfy its digital programme. The contradictory yet interdependent relationship between these poles was recognized in the following:
Why can’t we kill two birds with one stone? …It doesn’t make sense for us to go to market with print copy, and then have to update it in 6 months. We should just merge this into the same process and bring them both out at the same time.
Had Chris removed the project deadline, managers may have been able to indefinitely postpone synergistic solutions. However, Chris’ reiteration of the deadline (the temporal context) served as an important sensemaking tool to facilitate attention to both poles.
TMT leader’s practices shaping the relationships within the interpretive context
The above practices are a revealing representation of how the TMT leader’s practices deployed a combination of cues to shape lower-level managers’ sensemaking, highlighting not only the contradictory but, importantly, the growing interdependent relationship between poles. While we introduced each in isolation, the data showed that leaders emphasized the linkages between related contexts in order to maintain salience beyond its initial instantiation, such as in a strategy plan or a project deadline. Complementing the PrintSBU findings above, we now draw on data from the other settings to show how TMT leaders constructed the relationships between related contexts.
Relationship between instrumental and relational contexts
In each of the SBUs, the TMT leader’s sensemaking cues in constructing the instrumental context were interdependent with the relational context. For example, in directing lower-level managers to make sense of their performance goals (instrumental context), TMT leaders also directed managers to formulate product design plans that formalized their roles between print and digital agendas (relational context). To illustrate this in detail, we draw on an example from within MagazineSBU.
Traditionally, MagazineSBU managers were incentivized based on their ability to meet local, print-based targets for subscriptions and newsagency sales. This stemmed from a proud print tradition, which centred on producing glossy magazines. As MagazineSBU faced growing pressure through the strategic plan to embrace the digital agenda, managers became unclear on how their performance would be measured, given concerns that they would lose print subscriptions by sharing content online. To address this, Sophie reduced managers’ print subscription targets in exchange for an incentive scheme that recognized high page impressions for the lifestyle content viewed on the new PrintSBU pay wall. This subscription target became an important sensemaking tool to direct lower-level managers’ appreciation of joint print and digital objectives. As she stated:
The benefit of us being in the [MediaCo] business is that we get greater reach for our content. So we need to [introduce this new target] for the [MediaCo] business, but it is also a great win for all of us. We get better exposure for our content so that is great advertising for our [printed magazines]; but we are also getting ahead of our audiences by accessing [the digital product development tools]….so we will lead on new product ideas as well.
Here, constructing an instrumental context also elicited cues about the relational context. Managers were now expected to change their roles from being just journalists to also incorporate digital products (‘new product ideas’) in their roles. At the same time that the instrumental context shaped the relational context, the reverse was also true. As managers implemented a new content development plan for the digital product in their new roles (relational context), issues arose around whether measuring high page impressions was a fair performance metric (instrumental context). MagazineSBU managers had limited control over how their content appeared on PrintSBU’s pay wall. This was the main lever for controlling page impressions, and therefore put the page impression targets at risk. As one manager noted: ‘We could just become a service function to [PrintSBU], which diminishes the real value of what we do. … Unless we get rewarded, you’ll never get quality there [on the digital pay wall] because we have no skin in the game.’ This spurred Sophie to renegotiate the basis for the incentive scheme with lower-level managers. Rather than having to meet fixed targets for the year, lower-level managers were given a relative target, whereby they had to improve on each quarter’s performance. This enabled lower-level managers to make sense of their print and digital agendas simultaneously without a direct trade-off decision between them. The same manager later framed the interdependence between print and digital under the new digital targets as follows:
We need click bait [from page impressions on the pay wall landing page], but it is not all about click bait. We also need to balance journalistic integrity and quality. If [a client] says ‘why are you putting up stories about [a competitor]’ I have to say ‘well we’re independent, that’s what we do’ … So our new targets mean we can get a balance between growing online [through the pay wall] but not being a slave to online.
Thus, as the relational context became apparent, it had interdependent and synergistic links to interpreting the instrumental context and reinforced lower-level managers’ experience of paradoxical tensions.
Relationship between relational and temporal contexts
TMT leaders’ cues also enacted interdependence between relational and temporal contexts. In PrintSBU, the deadline for the pay wall launch (temporal context) was interconnected with negotiations over the design specifications for the paywall and managers’ different roles (relational context). The decision of when to launch remained in Chris’s discretion, as he wrestled with whether the quality of the digital product was sufficiently exploratory to launch in market. As Chris described it:
There is no point in us going to market until we get this right. The thing that keeps me awake at night is that we’re putting lipstick on a pig. We need to make sure we execute well [in terms of exploration]. The colour, the type set, the tone, the voice – all this matters.
In BroadcastSBU, Lev performed a similar role in moving between the relational and temporal context. BroadcastSBU had specialist expertise in sports broadcast, which was constrained by a series of rights negotiations with sporting associations. For example, in a football league, rights to broadcast content were divided and sold based on device (television vs mobile), time (live vs on demand) and regional jurisdiction. Although BroadcastSBU executives were initially happy to provide content for the PrintSBU pay wall, internal disagreement emerged as the timing of sports rights were reconciled with the launch of the digital product. One manager described the tension as follows:
The content creates confusion around the product you want to deliver because there are restrictions about the content you can deliver [based on legal rights]… But if you were starting with a garage, you could think ‘what content could I get from that position?’ So we have all these great insights around what people want and what is really interesting, but then the lawyers end up deciding based on when [the rights negotiations] come on stream.
Here, ‘garage’ was a reference to start-up companies incubated in Silicon Valley garages, and indicated what the company could pursue under exploration. Allocating roles to lower-level managers based on product-related tasks (relational context) was interdependent with what sports rights existed within the timeframe for the launch (temporal context). The lack of synergy projected exploration and exploitation as a trade-off choice. Lev’s multi-tasking practices intermediated these tensions, by identifying a new proposal that enabled both issues to be reconciled synergistically. Specifically, Lev proposed a new pricing schema that allowed very basic content to be delivered online in the short term to meet BroadcastSBU’s digital agenda. As the more exciting or explorative content rights came online later (for example, through the acquisition of new sports rights), Lev proposed that the price of the product could be increased. Lev described this decision as follows:
What I proposed is a soft launch where we go to market with something in [the next 6 months] but then we re-price [the content] over time as we get content rights to [other sporting leagues]. What I don’t want to do is sit on our hands and do nothing because then nothing will happen. … We just needed to get the process started, and we can then refine things later.
The pricing proposal illustrated how the TMT leaders continuously orchestrated a balance between relational and temporal cues, by shifting attention between the project deadlines (temporal context) and the quality of the product and actors’ roles therein (relational context).
Lack of a relationship between contexts
At the same time that TMT leaders constructed a constellation of sensemaking cues that linked each interpretive context to render paradoxical tensions salient, the failure to construct adequate cues led to a breakdown of salience as lower-level managers de-prioritized attention to a pole. This emerged in MarketplaceSBU as Mark’s initial efforts to create an interpretive context broke down as the basis for the instrumental context disappeared.
MarketplaceSBU operated as a supporting business to PrintSBU, providing products and services to support commercialization of the printed newspaper. Initially, Mark organized a strategy workshop to brainstorm new ideas for the business unit. However, since the business unit had been given an open remit, he lacked a strategic plan against which to measure or prioritize exploration. This created excessive variety, which Mark described as follows:
There are a million things we could do; but we need to focus on those that either enhance what we already do, or undermine our core business. So, the limiting factor on a lot of these ideas is noise. I have a problem if my managers get distracted with a lot of these ideas.
‘Noise’, here, was used as a metaphor for explorative ideas. A number of meetings were organized between the CEO and Mark to vet these ideas and prioritize them into projects for MarketplaceSBU. However, this process became increasingly postponed as the CEO’s attention became taken up by the launch of the digital pay wall. Thus, Mark was unable to agree with the CEO how to link the strategic plan to MarketplaceSBU’s business unit objectives (instrumental context), and therefore a corresponding product plan (relational context). As a result, efforts to resolve exploration were indefinitely abandoned and MarketplaceSBU continued to focus on its exploitation business.
Discussion
Our study was motivated by appreciating that the inherent nature of paradoxical tensions does not automatically trigger salience for organizational actors. We therefore examined a level of the organization for whom paradoxical tensions are strategically important (TMT leaders) to understand how the TMT leader enables latent paradoxical tensions to become salient for lower-level managers through their leadership practices. We draw together our findings into a conceptual process model, outlined in Figure 1, to show how the TMT leader’s practices prime sensemaking by orchestrating attention to a a repeated and converging constellation of cues. The relationship between the conceptual model and our specific cases is summarized in Table 4.
A Model of How Interpretive Context Renders Paradoxical Tensions Salient.
Cross-case Comparisons of the Role of Interpretive Context across Latent and Salient Paradoxical Tensions. Note: (a) focuses on the contradiction and (b) on the interdependence between poles.
We defined three related contexts – instrumental, relational and temporal – across our four case settings and show in Table 4 how interpretation of (a) the contradiction and (b) the interrelatedness between poles, changed over time. In Figure 1, we summarize the relationship between these contexts and their corresponding effect on allowing paradoxical tensions to become salient, which we now discuss.
First, the two intersecting boxes on the left of Figure 1 (labelled at 1) depict paradoxical tensions that are inherent in the environment. Initially, the contradictory and interrelated relationship between the dominant and alternate poles is latent because it is not yet recognized by organizational actors. In our study, these paradoxical tensions emanated from the interplay between print and digital business models: digital competed with print for revenue, but print needed digital for distribution and digital needed print for sources of high-quality journalistic content. The dotted line around each box indicates that these tensions existed in the environment, but organizational actors lacked a shared understanding of their paradoxical nature.
Second, as the TMT leader identifies these tensions to be of strategic importance, they employ particular bundles of practices (labelled at 2) in relation to artefacts such as plans and targets that shape the interpretive contexts for lower-level managers through the provision of cues. This is marked by the vertical arrows in Figure 1. These cues prime lower-level managers’ action by focusing their attention on organizational issues in which paradoxical tensions are embedded. Selection of practices is situated, rather than designed, as the TMT leader responds to lower-level managers’ perceived appreciation of paradoxical tensions. Specifically, as the TMT leader perceives awareness of only one pole, they deploy practices in relation to cues to draw attention the alternate pole. For example, the TMT leader uses strategic plans, subscription targets and performance templates to construct the instrumental context, and but shift their practice in relation to these cues between consolidating – which supports existing incentives(dominant pole) – and diversifying – which diverges from existing incentives. The TMT leader also constructs relational contexts by drawing attention to product design plans and role descriptions. Here, supporting practices within this context encourage existing patterns of conduct (dominant pole), whereas devaluing promotes conduct that supports the alternate pole. Finally, temporal contexts are composed of cues in relation to resource constraints, such as project planning deadlines. Prioritizing reinforces the dominant pole whereas multi-tasking encourages divergence towards the alternate pole.
Third, the TMT leader’s practices therefore create the conditions that induce lower-level managers to appreciate both the contradiction, but also the interrelatedness, between poles – what we define as salience. The gradual but converging constellation of cues is depicted by the horizontal arrows moving across Figure 1 from the left (where paradox is latent) to the right (where the paradox is salient). During this process of the paradoxical tensions becoming salient for lower-level managers, the emphasis the TMT leader places on each dimension within the interpretive context changes (labelled at 3). In our study, the TMT leader first transitioned from instrumental to relational and then from relational to temporal contexts. However, instrumental and relational contexts are interdependent: as plans are enacted in everyday practices, they also change aspects of the incentives. Furthermore, relational and temporal contexts are interdependent: as the TMT leader links the plans to resource constraints, they also render closure to a strategic episode. These relationships are reciprocal, rather than linear. Thus, relational context also influences the instrumental context, for example as commitments to the digital pay wall within MagazineSBU led to a change in their incentive structures. Similarly, the temporal context can also influence the relational context, as deadlines shift to accommodate further changes to the product design document. The reciprocal nature of these relationships is important, since it is the convergence of cues across multiple related contexts that enhances actors’ complex understanding of the paradoxical nature of tensions, and allows these paradoxical tensions to become salient. This culminates in the horizontal arrows depicted at the right-hand side of the figure (labelled at 4). In this respect, salience arises not only from a single instantiation of tension, but also from its diffusion and repetition over time.
Taken together, our conceptual framework makes three theoretical contributions. First, we contribute to paradox theory by highlighting the importance of ‘interpretive contexts’ in enabling organizational actors to appreciate salience. We define salience as when an organizational actor appreciates the relationship between alternate poles as both contradictory as well as interrelated. However, in extant practice-based studies of paradox, the focus of salience has been cognitive alone through rhetoric or discourse (Abdallah et al., 2011; Bednarek, Paroutis, & Sillince, 2014, 2016; Jarzabkowski & Lê, 2016; Jarzabkowski & Sillince, 2007). However, our findings highlight the structural underpinnings of paradoxical cognition, by showing how the TMT leader ‘sets the scene’ for lower-level managers by guiding them to attend to particular stimuli at particular points in time. This extends previous studies that apply a practice perspective to paradox (Jarzabkowski & Lê, 2016) by illustrating a diachronic rather than synchronic process for action formation. Individual enactments of paradox are necessary, but may not be sufficient, to enable salience since lower-level managers may need repeated sensemaking encounters in order to appreciate the complex relationship between poles. Our conceptual framework therefore seeks to preserve the ontological differentiation between structure and action, which is sometimes lost in practice-based approaches that render the action alone as the ‘smallest unit of analysis’ (Herepath, 2014; Reckwitz, 2002).
From a paradox perspective, focusing on rhetoric within a single level may overstate the consequentiality of ‘in the moment’ activities for the organization. Jarzabkowski and Lê (2016) construed the social construction of paradox, as enacted through humour, as entwined with and inseparable from the response paths that they set in motion within the organization. Elsewhere, speech acts, such as the ‘discourses of transcendence’ (Abdallah et al., 2011) or synergy rhetoric (Jarzabkowski & Sillince, 2007) are not only constitutive of the social context but are also its most consequential elements. Yet, Abdullah et al. (2011) acknowledge that rhetoric may ‘be enhanced when it is embedded in extant institutional ideas’ (p. 345). Our findings show that not all ‘in the moment’ activities are equally consequential. Rather it is actors’ converging understanding as they experience paradoxical tensions through cues supplied across three related contexts that enable paradoxes to become salient. This insight gives further specificity to a dynamic equilibrium model of organizing (Smith & Lewis, 2011), which identifies two enabling conditions rendering salience: (a) environmental conditions of plurality, scarcity and change; and (b) paradoxical cognition. We add a third condition, namely (c) interpretive contexts that render the contradiction and interdependence between poles simultaneously, as well as over time.
This leads to our second contribution, being our theoretical appreciation of the role of leadership as a practice in enabling exploration and exploitation within organizations. Extant approaches in the ambidexterity literature rely on competency models of leadership that attend to the specific traits and behavioural attributes of individuals (Alexiev, Jansen, Van den Bosch, & Volberda, 2010; Jansen, George, Van den Bosch, & Volberda, 2008). This ‘methodological individualism’ (Chia & Holt, 2006, p. 638) depicts the role of leaders through a linear, causal model. For example, leaders make either/or structural choices (O’Reilly & Tushman, 2013) and exhibit specific behaviours that optimize organizational performance based on contextual contingencies (Lubatkin, Simsek, Ling, & Veiga, 2006; Simsek, Veiga, Lubatkin, & Dino, 2005). These approaches focus on what leaders accomplish for themselves but say little about how they influence others.
We show how leadership as practice is a relational activity, which emerges as the TMT leader and lower-level managers give and make sense of their environmental context (Carroll et al., 2008). This extends paradox research by highlighting relationality as a key mechanism motivating the dualism at the heart of paradox theory (Lê & Bednarek, forthcoming; Suedfeld, Tetlock, & Streufert, 1992). Lê and Bednarek (forthcoming) argue that practice-based studies extend paradox theory by showing how actors’ responses to tension ‘feed off’ or are mutually constitutive of each other. This focus on the ‘between-ness’ of phenomena showcases how practices are linked to a wider nexus that have a ripple effect beyond localized activities (Clegg et al., 2002). This contrasts with other approaches that attend more closely to individuals’ cognition in relation to paradox. Paradoxical cognition activates dualism as actors apply individualized cognitive processes (Smith & Tushman, 2005). Even a dynamic decision-making model theorizes context as it is experienced by individual leaders (Smith, 2014). By showing the critical role that the TMT leader plays for lower-level managers – because they set up cues that allow appropriate interpretive contexts to emerge – we therefore position leadership as a more integral and prolific part of rendering paradoxes salient and enabling management responses. Paradoxical leadership, then, uses relationality to link the TMT leader’s understanding of environmental conditions back to lower-level managers’ understanding. Thus, we theorize paradoxical leadership as a ‘non-individualized phenomenon’ (Schatzki, 2005), in which leaders create the structural conditions for salience for lower-level managers by creating interpretive contexts.
Moreover, by showing how leadership as practice constructs a dynamic, reciprocal relationship between related contexts, we show how leaders allow paradoxical tensions to become salient over time. When leaders apply prioritizing and diversifying practices in constructing a temporal context their practices do not ‘end’ the social action but rather trigger further interdependent practices within the relational context. These practices, in turn, trigger practices in either the instrumental or temporal context, thus fostering a dynamic relationship between contexts. Leaders’ practices therefore facilitate interwoven communications with lower-level managers across related contexts, as opposed to designating action within separate, isolated events (Denis, Langley, & Cazale, 1996; Schatzki, 2006). When leaders perceive too much emphasis on one pole they support the alternate pole. Thus, leaders seek to move the action forward by doing ‘whatever it takes’ to create the conditions (and supply cues) that enable strategic paradoxes to be recognized across levels within the strategic business unit. When they fail to supply adequate cues, there is a breakdown in actors’ appreciation of paradox. This complements but adds to other patterns of communication recognized in the literature, such as finding novel synergies (Luscher & Lewis, 2008; Rothenberg, 1979; Takeuchi & Osono, 2008) and consistent inconsistency (Smith, 2014). These studies tend to focus on how leaders shape individual behaviours, whereas our study shows the intermediating role of interpretive contexts in empowering individuals to make their own appraisal of the paradoxical tensions rather than seeking to influence actors’ behaviour directly through one-to-one interactions and sparring sessions (see, for example, Luscher & Lewis 2008).
Finally, we contribute to the ambidexterity literature by showing how leaders foster attention to exploration and exploitation simultaneously within a single strategic business unit. Even though ambidexterity scholars differ in their emphasis between structural and contextual solutions to competing demands, both approaches envisage switching between either exploration or exploitation activities at different times depending on the situation (Papachroni et al., 2015, 2016). Applying a paradox lens, we show how leaders can manage tensions for other actors by orchestrating the interpretive interplay between the poles. This enables synergies rather than trade-off choices, as the TMT leader and lower-level managers co-construct interpretive compromises to both poles (Lewis & Smith, 2014). This complements and extends the focus of leadership in ambidexterity studies from what leaders do for themselves to include what they do for others (Gibson & Birkinshaw, 2004). We show that the TMT leader does not only rely on formal authority (Gilbert, 2005; Raisch & Birkinshaw, 2008) but also deploy relationality, through practice, to come to a negotiated understanding of local context. Thus, whereas contingent solutions to competing tensions promise balancing tensions through context–solution fit (Papachroni et al., 2015), the implications of our findings is that leadership to balance exploration and exploitation may be a more contested, fluctuating, and interpretive endeavour. Furthermore, by focusing this contribution on the early stages when there are interpretive differences between the TMT leaders and lower-level managers, we complement the study by Zimmerman and his colleagues (2015) on ambidexterity emergence. Their study, located in alliance formations across organizational boundaries, envisaged ‘ambidextrous charters’ as interpretive tools bridging political and trust-based tensions. We extend these findings to an intra-organizational context, showing how instrumental, relational and temporal contexts serve a similar interpretive role within business units. By maintaining interrelatedness between contexts, we offer an alternative, additional solution to vacillation between poles (Boumgarden, Nickerson, & Zenger, 2012) situated in TMT leaders and lower-level managers’ ‘in the moment’ activities to reconcile the structural and structuring poles.
Limitations and Conclusion
Notwithstanding these findings, our study has a number of limitations. First, case study approaches are limited with respect to the generalizability of findings. While we have made every effort to conduct comparisons across settings, future research could extend these insights to include cross-case comparisons between different organizational contexts (Andriopoulos & Lewis, 2010). For example, future research might consider boundary conditions around the construction of interpretive contexts, and how these might vary based on differences in environmental or organizational context. Here, legitimacy (Tost, Gino, & Larrick, 2013) and organizational identity (Gotsi, Andriopoulos, Lewis, & Ingram, 2010; O’Reilly & Tushman, 2013; Schultz & Hernes, 2013) have received recent attention and may be important constructs in moderating actors’ ability to host competing and interdependent tensions simultaneously.
Second, this study focuses on individual TMT-level practices in relation to lower-level managers. Future studies may examine the role of other actors such as frontline managers or external facilitators as they interact with managers, or otherwise TMT leaders operating within a team (Jay, 2013; Luscher & Lewis, 2008). We see these as promising areas of future research, especially as organizations extend beyond organizational boundaries to access new sources of exploration, as envisaged by new models for open innovation (Whittington, Cailluet, & Yakis-Douglas, 2011). Furthermore, other levels of the organization may be more important in studying different types of paradoxes. For example, although we show that the TMT leader was important for an organizing paradox through the construction of the interpretive context, these dynamics may be different, for example, in belonging and learning paradoxes (Jarzabkowski et al., 2013; Knight & Paroutis forthcoming). For example, in the Dutton & Dukerich (1991) study of a belonging paradox, the TMT leader’s practices were directed towards linking lower-level managers’ attention to paradoxical tensions through claims about the organization’s identity, rather than its organizing design and process. These remain important research questions if we are to understand how paradoxical tensions are introduced to enable long-term performance.
To conclude, organizations face multiple tensions in today’s competitive world. The pressure to innovate highlights the specific importance of explore – exploit tensions, but in this paper we have tried to be more sensitive to the early stages of paradox recognition in lower-level managers, since the recognition of both paradoxical tensions cannot be taken for granted. By showing the role that the TMT leader plays in these early stages we have sought to bring greater clarity to the nature of salience, advancing a practice-based perspective on the role of leaders in facilitating salience in others.
Footnotes
Acknowledgements
We would like to thank the guest editors, especially Wendy Smith, as well as the three anonymous reviewers for their constructive feedback, which was critical to this paper’s development. We are also grateful to colleagues for their helpful comments, particularly Jane Lê, Michael Tushman, Suresh Cuganesan and Rebecca Bednarek, who provided input at different stages.