Abstract
This conceptual paper challenges assumptions about the primacy of leaders and leading over followers and following in the leadership process. Leadership cannot be holistically understood unless followership and leadership are researched as they are enacted—in tandem. To move toward a more complete understanding of leadership, we introduce a leadership system which involves a co-created leadership process, unfolds over time, and accounts for stability and change in partners’ expectations and behaviors. Drawing on role theory and implicit leadership (ILTs) and followership (IFTs) theories, we suggest how expectations of self and other influence leading and following behaviors and the leadership process. Personal learning acquired through the experience of leadership results in stability or change in one’s future expectations. Our framework explains how stability and change in each dyad partners’ leadership and followership expectations comes about through constructionist and constructivist mechanisms. These mechanisms occur across two timelines: The first is the microadjustments made to expectations and behaviors within a particular leadership occurrence. The second is the loop between personal learning and the expectations each member carries into future leadership occurrences and relationships. Practical implications arising from this new framework include considerations for leadership and followership development and a contribution to leadership forecasting.
A Google search for TED talks and leadership returns 13,300,000 results. The same search on followership finds a mere 22,300. The discrepancy is stark, despite our intuitive understanding that leading implies and requires following. This paper challenges assumptions about the primacy of leaders and leading over followers and following in the leadership process. Leadership cannot be holistically understood unless followership and leadership are researched—as they are enacted—in tandem. To move toward a more complete understanding, we introduce a leadership system which involves a co-created leadership process, unfolds over time, and accounts for stability and change in partners’ expectations and behaviors.
Taking Uhl-Bien and colleagues’ summary illustration of the leadership process (Uhl-Bien et al., 2014) as our starting point, we draw on role theory and schemas such as implicit leadership (ILTs) and followership (IFTs) theories to understand how expectations of self and other influence behaviors and the process. In so doing, we join together different forms of expectations from various research perspectives into one conceptual model that provides a cohesive and practical leadership framework. A leadership system that accounts for the interplay between member expectations, leading and following behaviors, and a dyadic interactional exchange process within a single leadership occurrence illustrates this framework. Personal learning acquired through this experience results in stability or change in one’s future expectations that reside both contextually (within the same relationship or when in the same role) and context-free (carried across relationships and roles).
In the following sections, we begin by situating our thinking in relation to key leadership and followership research directions. Next, we introduce role theory, implicit theories and constructionist and constructivist notions as theoretical underpinnings relevant to stability and change in expectations and behaviors. We then introduce our model in detail and discuss its theoretical and practical implications.
Leading and Following in the Leadership Process
Recent trends include followership as an important contributing factor in the leadership process (i.e., Oc & Bashshur, 2013). We join a stream of leadership conceptualization that recognizes leadership and followership as equally important (Blair & Bligh, 2018; Carsten et al., 2018; Howell & Shamir, 2005; Kelley, 1988. Rather than taking a unidirectional perspective of leaders influencing followers in which most current literature is entrenched (i.e., Gottfriedson & Aguinis, 2017; Knippenberg & Sitkin, 2013; Yukl, 2012), we understand leadership to be a relational (Dulebohn et al., 2012; Emirbayer, 1997) and emergent (Acton et al., 2019; Avolio et al., 2009) process. Accordingly, we envisage leadership as a process which is co-constructed through the interaction of leadership and followership. This definition aligns with critical evaluations of leadership processes such that leadership process is co-created through combined acts of leading and following (DeRue & Ashford, 2010; Hannah et al., 2014; Hollander, 2009).
Situating our conceptualizations
Key Leadership Literature which Situates Our Model.
Laying out the Landscape: Relevant Theoretical Bases
Before introducing our model, we provide an overview of its grounding in role theory and describe how we draw on implicit leadership and followership theories to explain generation of expectations. We will describe a constructionist perspective consistent with the co-creation of leadership and micro-adjustments to expectations and behaviors along with a constructivist perspective consistent with learning from experience which may either reinforce or effect changes to future expectations. We begin with expectations.
The Development of Expectations
Expectations represent one’s understanding of what following and leading should look like within the dyad. They arise from various sources including social and cultural norms (Turner, 2001) and organizational norms and standards (Carmeli & Schaubroeck, 2007) which are reflected in formal and informal structures and processes. Social norms, organizational context, and previous experiences also contribute to development of schema (Epitropaki et al., 2013) or mental models held by individuals. Schemas comprise part of an individual’s leader or follower identity and professional identity within a leadership dyad (Epitropaki et al. 2017; Lord et al. 2020; Weinberg, 2019) and shape how they behave in their role (Sluss and Ashforth (2007). Self-efficacy (i.e., Bandura, 1997) or assurance in one’s capability to meet expectations and to act in accordance with their schema is a further contribution to expectations. Moreover, expectations arise from repeated interactions between dyadic partners (Turner, 2001), a notion elemental to role theory.
Role theory
Role theory explains how familiarity and ongoing routines allow members to develop expectations about, and anticipate, one another’s behavior (Goffman, 1990; Parsons & Shils, 1951). There exist two streams of thought (Biddle, 1986; Turner, 2001): Whereas a structuralist understanding (i.e., Busenbark et al., 2016; Shen & Cannella, 2002) sees roles as primarily derived from formal positions within organizations, interactionist role theory (Turner, 2001) describes roles as adapted and acted out through interactions between individuals. Although role theory suggests a tendency toward stability, it does allow for incremental change. Kovach (1986), for instance, applies a role theory lens to suggest that conceptualization of the leader or follower role may adapt, guided by personality and past experience. Performance of leading and following behaviors may also change in response to context such as the structure and culture within the dyad or organization. The possibility for more dramatic change—when one’s partner’s behavior surprises or when new learnings arise from the exchange of leadership and followership which comprise the leadership process—would, from a role theory perspective, call into question currently held roles within the dyad. Consistent experience of the “new” behavior would be reflected in changed role expectations. Role theory suggests that expectations arise out of repeated behavior. Implicit theories predict that expectations influence behavior. Both allow for change.
Implicit theories
Implicit leadership theories and implicit followership theories represent structures or prototypes which identify and describe characteristic abilities and traits (Epitropaki et al., 2013). Individuals rely on their implicit theories to understand and respond to others (Shondrick & Lord, 2010) and as a basis for their own behavior (Engle & Lord, 1997). ILTs reflect followers’ specification and rating of leaders’ abilities (Lord et al., 2020). Despite their focus on followership, IFTs largely reflect leaders’ ideal representation and understandings of followers with an emphasis on behaviors; whereas the relatively sparse literature on followership orientation tends to focus on follower’s understanding of following (i.e., Carsten et al., 2016, 2018). As we will discuss, implicit leadership and followership theories constitute important components of expectations in our model, but we describe how representative schemas are held by both parties regardless of status or role.
Constructionist foundation of the leadership process
Constructionism suggests co-creation and serves as the mechanism through which dyadic partners co-construct the leadership process. We define the leadership process as that which emerges from the interactional exchange between members. Role theory and implicit leadership and followership theories both allow for members’ contributions to this co-constructed approach. Accordingly, our conceptualization aligns with a constructionist (Uhl-Bien et al., 2014), interactionist approach to the leadership process (e.g., Shamir, 2007; Uhl-Bien, 2006) which understands following and leading as fluid engagements, the dynamics of which co-construct leadership. It is through these dynamics that dyad partners develop each unique leadership occurrence (Camargo-Borges & Rasera, 2013), suggesting that leadership is created and defined at the dyadic level through the exchange of relational interactions (Alvesson, 2019; Einola & Alvesson, 2021; Raelin, 2011).
Constructivist foundation of the leadership system
Constructivism is an individual process. It centers on learning through personal discovery (Clark, 2018) and using what was previously learned to create and apply mental models toward future situations (Amineh & Asl, 2015; Jonassen & Henning, 1999). This allows for experience to challenge and change previously held expectations or for learning from experience to reinforce expectations. Thus, a constructivist perspective accounts for both stability and change in expectations. This perspective allows for both a more immediate and a time-lagged reflection on learning from experiences to inform deliberate changes to one’s expectational inputs in preparation for and engagement with future leadership occurrences: a leadership cycle. We discuss stability and change in expectations and the formulation of this cycle following the introduction of our model.
Introduction of a New Model
Figure 1 illustrates our model which germinates from Uhl-Bien and colleagues’ (2014) illustration of the leadership process. Our model offers a substantive extension through the addition of expectations and the interactions between and mutual influence of expectations and behaviors; the emergence of the leadership process as an asymmetric exchange of endowing, acknowledging, accepting, and deferring; and identification of personal learning as a focal outcome. Time is included in the sequencing of and interplay between the elements of the model and through a recursive loop. Before walking through each element of the model, we overview the relationships between partners as expectations change or stabilize over time. We also highlight the duality of leading and following. Proposed leadership system.
Characteristics of the model
Relationships between dyad partners
For each leadership occurrence (Times 0 through 2) there are three possible relationships between the dyadic partners (P1 and P2). The first and most traditional is where partners operate within an organizational hierarchy and their behaviors match their position and roles and reinforce the leader and follower expectations of their partner. The second is likewise where partners operate within organizational hierarchy, but a member’s behaviors do not match their organizationally ascribed roles and may therefore challenge their partner’s expectations. Implicit theories of leadership and followership (e.g., Shondrick & Lord, 2010) suggest that each member holds expectations that leaders and followers should behave in certain ways. Even if each member behaves according to his or her role at the beginning of a leadership occurrence, it is possible that as the leadership occurrence unfolds, each member may bring forward what they deem to be contextually appropriate behavioral responses. This may include leaders contributing following behaviors and followers engaging in leading behaviors. This allows the opportunity for members’ expectations to expand and include a broader range of behaviors as leading and following. The third type of relationship is peer-to-peer, in which partners do not have different formal status but both hold leader and follower expectations, though these expectations are not prescribed to a particular role. For each dyadic relationship, whether hierarchical or peer, expectations will be reinforced and kept stable or challenged and changed at two points in our model.
Stability and change in expectations
The first point for stability or change is at the interplay between behaviors and expectations (labels a1 through e2 in Figure 1). These align with implicit theories—the matching of experience against models held in memory and therefore existing at the intra-individual level. Expectations in role theory are also individually held. But these arise out of the leading and following behaviors which begin the leadership occurrence. The second point for stability or change in expectations occurs when personal learning resulting from the leadership process influences future expectations (labels h1 and h2 in Figure 1). Members’ expectations about the leading and following behaviors they and their partner will engage comes in part from learning how these have played out for them in past iterations of leadership. ILTs reflect the effects of memory of general information that accumulates with experience or learning such that the comparison and matching of currently experienced behaviors to prototypical categories reinforces the identification of individuals as a leader or a follower. On the other hand, a mismatch arising from unexpected behaviors may advance the learning of a new cognitive category (Shondrick and Lord, 2010), and is likely to produce changed future expectations.
Event systems theory (Morgeson et al., 2015) and anchoring events (Ballinger & Rockmann, 2010) provide means through which role theory’s familiarity may be interrupted, thereby disrupting members’ capacity to anticipate the other’s behavior. They offer complementary explanatory mechanisms for change that originate from exposure to and reflection on unanticipated occurrences which disrupt expectations. Both are consistent with the understanding of memory in ILTs. They are also consistent with constructivism and thus aligned with personal learning.
Anchoring events occur in a dyadic relationship when an exchange fails to produce an expected resolution and has a sufficiently powerful emotional impact. A memory is created both of the event and of how the event felt, thereby changing the rules or script used to evaluate their current and future partners’ behavior. From an interactionist role theory perspective, such anchoring events may shape the enactment and adaptation of leading and following. Event systems theory by contrast does not specify individual psychological processes. Events may occur within the dyad or more broadly thus allowing for learning arising from external (to the dyad) sources. Novelty, strength, and criticality of an event will determine the degree to which it has the power to disrupt learned expectations and adapt different behaviors in the current or future relationships.
Our model illustrates stability and change in expectations. Members begin the leadership cycle at Time 0 anticipating how their partner specifically and leaders and followers generally should and will behave. However, their behaviors influenced by those expectations may not align with their partner’s expectations. If their partner’s behavior surprises them by conflicting with their initial expectations, a member may immediately adjust their expectations and corresponding behavior; the resulting leadership and followership exchange makes each leadership occurrence unique (Abbott, 1998).
Partners may of course interpret their own and others’ behaviors, the leadership process, and its outcome differently. However, it is an individual’s response to their partner’s behaviors (and learning from their experience of leadership) which influences stability or change in expectations. Notably, our model is not concerned with dyadic partners’ agreement as to the interpretation of behaviors or outcomes. We would however suggest that shared implicit theories or orientations could produce responses that reinforce both partners’ expectations, whereas differences would likely challenge those expectations (Carsten et al., 2016). For example, if partners share a proactive followership orientation, leading behaviors which include expecting their partner to identify problems and solutions will likely be met with following behaviors such as suggesting solutions. On the other hand, mismatches in expectations may produce more frequent instances of behaviors deemed surprising or challenging to their partner; effectively producing learning opportunities effecting a shift in expectations carried forward to future leadership occurrences.
Duality: Leading and following by both partners
Leading and following behaviors are equally substantive and, as alluded to earlier, may occur outside of a fixed or formal hierarchical understanding (Einola & Alvesson, 2021). Indeed, leadership has been acknowledged as being most effective when individuals are able to engage in both leading and following behaviors (Peters & Haslam, 2018), such as when managers demonstrate humility (Owens & Hekman, 2016) or defer to experienced staff members (Neeley & Reiche, 2022); or when followers exercise courage (Chaleff, 2009). This helpfully shifts us away from considerations of leader–follower relationships that are closely associated with hierarchical patterns such as manager–subordinate. A leader in one occurrence may be a follower in another and either partner may engage in leading or following behaviors at various phases of their working relationship across leadership cycles. Even in highly structured circumstances both leaders and followers may engage in leading and following behaviors. This duality (e.g., Blair & Bligh, 2018; Graen & Uhl-Bien, 1995; Tourish, 2014) includes instances when middle managers both lead their team members and follow their senior manager’s direction (Jaser, 2020). Duality, with either partner engaging in leading or following behaviors, may occur within dyads as well as in peer dyads and dyads consisting of manager/staff member. In workplace relationships, the purpose of the leading and following exchange in each leadership occurrence is to contribute to meeting organizational goals. While one member may more often engage in leading behavior, and their partner more often engage in following (as when leaders’ and followers’ behaviors match their organizational status), it may at times better serve this purpose for this to not always be the case. A more junior staff member with special technical expertise, for instance, may provide direction to their manager who follows their staff member’s advice. The best leaders know when to lead and when to follow; furthermore, the best followers know when to follow and when to lead.
A walkthrough of the model
Expectations and behaviors as inputs
Member expectations represent the leftmost constructs in our model. A constructivist perspective suggests that even at Time 0 members hold expectations of leaders and of followers learned from previous experiences. Both members bring expectations of workplace behavior from previous work relationships with others and/or with this partner. Both will hold some degree of followership (and leadership) orientations and identities. Those in a hierarchical relationship will additionally have the expectation that they and their partner will engage leading and following behaviors consistent with their formal status and will hold IFTs or ILTs aligned to these positions. Peers will bring expectations of workplace behavior from previous work relationships with others and with this partner. They too will hold followership (and leadership) orientations and identities but will not have yet activated IFTs or ILTs in relation to the current leadership occurrence.
As the model unfolds and we move into Time 1, our understanding of expectations draws on, and is consistent with, implicit theories and followership orientation (Epitropaki et al., 2013 and Lord et al., 2020 provide reviews). Here, leading and following behaviors serve as a secondary input, and the interplays between expectations and behaviors are crucial.
The expectations and behaviors of partners in a formal hierarchical relationship whose behaviors match their position and role will reinforce each other and the partner’ leader and follower identity consistent with their status. For those in a hierarchical relationship whose behaviors challenge expectations, there exists the possibility that the range of one’s own and other’s expectations of leaders and or followers will be enlarged. This may be reflected in exhibition of a broader range of leading and following behaviors. It might go so far as to reverse the leader and follower roles (with dyadic partners behaving as peers, ignoring organizational status) or the leader engaging in following behavior and the follower engaging in leading behavior while maintaining hierarchical status. In peer dyads either, or both, partners will enact leading or following behaviors reinforcing or challenging their own and their partner’s expectations because no established leader and follower roles exist at this point in the leadership occurrence.
One’s behavior is influenced by self-expectations
The arrows labeled a1 and a2 on Figure 1 represent the influences of member expectations of themselves on their leading behaviors: P1’s self-expectations represented as path a1: leader self-expectations (Shondrick & Lord, 2010) and P2 self-expectations on their own following behaviors, as represented by path a2: follower self-expectations (Knoll et al., 2017). Illustrative examples: If I expect that followers do as they are told, my leading behavior will be to provide direction and my following behavior will be to follow direction. However, if I expect that followers challenge leaders my following behavior will include voicing concerns or making suggestions.
One’s behavior is influenced by others’ expectations
Leader and follower expectations have been shown to relate to the work performance and behavior of others (Jackson & Johnson, 2012). The arrow labeled b1 in our model represents P1 expectation of followers predicting follower behaviors. This is consistent with the Pygmalion effect: that what one person, not necessarily a manager or a teacher although this is how the relationship is often operationalized, expects of another, becomes a self-fulfilling prophecy (Rosenthal, 2010). Studies show Pygmalion (Knoll et al. 2017; Lord et al., 1999; Veestraeten et al. 2020; Whiteley et al., 2012) or Golem (Leung & Sy, 2018) effects between managers and staff members. P2 expectations of leaders may influence leader behaviors (b2 in our model), for example in instances of cross-cultural leadership (Bildstein et al., 2012). Illustrative examples: If my partner’s expectations of leaders include listening before a decision is made, I will be more consultative than I would be with a partner who expects that leaders make decisions without discussion. I may speak up more assertively if my partner expects followers to be actively involved in planning and decision making and keep my ideas to myself if the expectation is of acquiescence.
One’s behavior is influenced by one’s expectations of others
Consistent with role theory and theorized in the identity literature (Lord et al., 1999; Sluss & Ashforth, 2007), expectations of another influence one’s own behavior (van Quanquebeke et al. 2011).
Leaders who hold high expectations of followers have been shown to exhibit leading behaviors that are conducive to the follower’s efforts (Tierney & Farmer, 2004). The arrows labeled points c1 and c2 in Figure 1 respectively illustrate the influence of P1 expectations of followers on their own leading behaviors (c1) and P2 expectations of leaders on their own following behaviors (c2). Illustrative examples: My expectation that a leader will want to be familiar with project details means that I will provide regular updates although I may not need my partner’s involvement. My expectation that a follower needs direction informs me to provide detailed explanations of what needs to be done.
Expectations of others are influenced by their behavior
Labels d1 and d2 in Figure 1 represent inverse influence whereby one’s observations of another’s behaviors will drive the individual’s expectations of that other person as suggested by recognition-based categorization of leaders and followers (Lord, Foti, & Phillips, 1982; Shondrick & Lord, 2010). In other words, we learn what to expect from experience. We propose that expectations of followers will be altered based on experience of following behaviors (d1). Leaders with more supervisory experience rely less on positive IFTs (Sy, 2010) in relation to their followers (Whiteley et al., 2012). Likewise, expectations of leaders will shift (d2) as a result of observation of enacted leading behaviors (Veestraeten et al., 2020). Illustrative example: I might initially expect followers to be acquiescent. However, I work with partners who behave proactively, and I will change my expectations of followers to include more assertive behavior. I may expect leaders to be directive but if my experience is with a more consultative leader I will adjust my behavior to provide feedback and ideas.
Self-expectations are influenced by other’s behaviors
Others’ behavior can affect self-expectations, as illustrated by the arrow heads labeled e1 and e2 in Figure 1. These labels also represent inverse influence whereby the experience of another’s behavior has an emotional or reflective impact consistent with consequences identified in the abusive supervision (Tepper, 2007) and women’s representation in STEM (Avolio et al., 2020) literatures. Illustrative example: A partners’ poor performance may cause a leader to consider the performance in part a result of their leadership. They may alter their expectations of themselves as being a less effective leader than they thought (thence adjusting their leading behaviors, perhaps by providing additional direction or support to the follower). A follower who is treated poorly may adjust their self-expectations by reducing their feelings of confidence, alternatively positive leading behavior may build their partner’s expectations that they are more capable than they originally thought.
Arrows b1-e1 and b2-e2 in our model are consistent with Lord and colleagues’ (2020) assertion that ILTs provide behavioral templates and that such double interacts reflect reciprocal behavioral interactions between leaders and followers. We suggest that these aspects of our model illustrate almost simultaneous micro-adjustments which are made by dyad members in response to the other’s behaviors and the reinforcement or changes these cause to expectations (i.e., Shondrick & Lord, 2010). Role theory suggests this too with the repeated experience of behaviors building expectations. The enactment of leading and following behaviors and the leadership process (labeled point f in Figure 1) is central to our model. We have noted that either party may engage in leading and following behaviors, moreover, partners may engage in an alternating sequence of leading and following behaviors in quick succession. For instance, a member who initially creates a plan (and thereby exhibits leading behavior) may rapidly and successively recognize their partner’s expertise and listen to incorporate their advice before pivoting back to make a decision (a leading behavior) or to defer the decision to the other (a following behavior). Throughout the same series of exchanges their partner would be receptive to the member’s planning efforts. They may then engage in leading behaviors by providing technical expertise before accepting their partner’s final decision (a following behavior) or alternatively taking the lead on new initiatives. This will particularly be the case in peer-to-peer engagements as these have less rigidly imposed (or a complete lack of) roles, but it could also occur in hierarchical dyads where the formal leader engages in leading behaviors and/or the formal follower engages in following behaviors.
These interactions between individuals directly stem from expectations of leaders and followers that members bring with them to the leadership occurrence. When the dyad reflects the formal hierarchy with leader and follower behaviors and expectations matching partners’ organizational roles P2 (assuming P2 is the member in the subordinate position) will swiftly endow leadership. In dyads whose expectations and behaviors do not match their organizational roles the same dyadic partners are likely to engage in more traditional or peer relationships in different leadership instances. Even in hierarchical organizations managers and staff members do not necessarily, or always, lead and follow in manners consistent with their hierarchal relationship (Larsson & Nielsen, 2021). In peer dyads the interplay of leading and following behaviors and expectations will be resolved when leadership is endowed on and accepted by one party—the member who within this particular leadership occurrence serves as the leader. The enactment of leading and following behaviors continues until one member endows leadership upon the other and leader and follower roles are thus settled in this leadership occurrence.
Leadership process: Endowing, acknowledging, accepting and deferring
We identify four aspects of followership and leadership. The asymmetric exchange of followership (endowing, acknowledging, accepting, and deferring) and leadership (accepting and acknowledging) comprises the emergent leadership process (labeled f in Figure 1). The process begins with one member endowing leadership on their dyadic partner by being willing, in this leadership occurrence, to follow. In another leadership instance, the same member may be endowed leadership by their partner and so, for that exchange, lead. Endowing leadership begins the exchange of following and leading behaviors. For example: the follower may ask “What would you like me to do?” or rather than immediately making a suggestion or providing information, the follower may wait for the leader to request that support. For the exchange to continue, leadership is accepted by the leader who acknowledges their partner as following. The leader may, for example, express appreciation and support, describe what they would like the follower to do, or provide feedback (DeRue et al., 2011; Gottfriedson & Aguinis, 2017; Yukl, 2012).
If the behaviors associated with acknowledgment are accepted by the follower, the follower then defers to their partner. Deference is recognized as fundamental to followership (Uhl-Bien & Pillai, 2007). However, deferring, for example, by doing things the way the leader has directed (Carsten et al. 2010) or accepting and applying feedback (Benson et al., 2016) will not happen unless the follower has first endowed and acknowledged leadership.
Endowing leadership is powerful, as is deferring to another’s leadership. Dyad partners wielding deferential power therefore have a great deal of influence in the leadership process. The members of the dyad then acknowledge each other’s leadership and followership until the exchange ceases. Examples might include leadership in encouraging innovative thinking or defining tasks and performance (DeRue, 2011; Yukl, 2012) and followership in offering input, feedback or advice or assuming responsibility (Benson et al., 2016; Carsten et al. 2010).
It is worth acknowledging the possibility that the leadership process will be interrupted and require additional behavioral and expectation interplay before it can proceed. Illustrative example: a leader expects leaders to be consultative but a follower expects leaders to be directive and followers to comply. In this instance, the follower could initiate the leadership process and endow leadership by waiting for instruction from the leader. Meanwhile, the leader accepts leadership and acknowledges the follower by anticipating their involvement in the discussion. Despite the pair processing leadership and followership, it is clear in this instance that expectations are not met through behaviors: the leader does not issue direction or the follower does not speak up. The leadership process cannot proceed unless at least one member adjusts their expectations or behaviors so that the exchange that makes up the leadership process is reinstated.
Personal learning outcomes
We now turn to the outcome of interest to our model (labeled as point g), specifically personal learning (Lankau & Scandura, 2002). Personal learning has two dimensions. The first is relational job learning which is context specific and applies to the current relationship or to the same role—leader or follower—in the leadership process, with another partner. The second dimension, adaptability, is context-free and can be applied across different types of relationships. The outcome of personal learning therefore fits with both a constructionist approach to leadership and followership and is consistent with the constructivist understanding that individuals learn through reflecting on past experience.
Personal learning arises from the co-construction of the leadership process. These learnings will either reinforce or disrupt partners’ future expectations of themselves and their dyadic partners thus influencing stability or change (labeled h1 and h2 in Figure 1). If the personal learning is sufficiently novel as to comprise an anchoring event (Ballinger & Rockmann, 2010) for one of the members, this member’s expectations may expand to allow broader range of behaviors to be included in their implicit theories. This is an individualized experience for each member, highlighting the intra-individual nature of constructivism: For one person, the learning may be an anchoring event; another person may experience the same learning but to them it does not constitute an event sufficient to shift future expectations.
We previously suggested that expectations do not necessarily influence behaviors positively. Neither does our model include the assumption highlighted by Alvesson (2019) inherent in many theoretical models that a culmination of positive behaviors predicts positive leadership outcomes. Proactive behaviors are generally considered beneficial (i.e., Belschak & Den Hartog, 2010); however, if one’s partner’s proactive leading or following behaviors are unexpected this would challenge one’s expectations. When expectations are challenged but adjustments are not made, the enactment of leading and following behaviors does not progress to a leadership process. Through a constructivist lens, however, experiencing a lack of leadership will still result in personal learning which will influence future expectations.
Outputs influence processes beyond the current dyad
Personal learning occurs not only within the same leadership instance but across concurrent and subsequent instances involving the same dyadic partners. Moreover, it also influences members’ approaches to all future leader–follower relationships in which they become involved (as depicted in Figure 2). Chain of leadership process learning outcomes influencing expectations.
Discussion
Our purpose has been to uncover new ways to think about leadership by introducing the concept of a co-constructed dyadic leadership occurrence within a leadership system that unfolds over time. Building on role theory and implicit leadership and followership theories, our posited framework explains how stability and change in each dyad partners’ leadership and followership expectations (which subsequently influence changes in behaviors) comes about through constructionist (co-created) and constructivist (experientially learned) mechanisms. These mechanisms occur across two timelines: The first is the micro-adjustments made to expectations and behaviors within a particular leadership occurrence contributing directly to the co-construction of leadership process. Until the point of dyadic exchange (endowing, acknowledging, accepting, and deferring), leadership does not yet exist: merely precursor expectations and leading and following behaviors. Thus, our constructionist consideration shifts the unit of analysis of leadership to the exchanges themselves. The second timeline is the constructivist loop between personal learning and its effects on the expectations each member carries into future leadership occurrences and relationships.
The predominant theoretical implication that stems from our conceptualization is the emergent relational and co-constructed nature of leadership in duality with and inseparable from followership. This understanding brings together role theory and implicit leadership and followership theories to understand how expectations, and hence behaviors, change and develop in the leadership/followership system. The mechanisms for change and stability are explained through a constructivist, learning perspective. We consider this, then suggest relevant research questions before focusing on two practical implications of our model: leadership and followership development and a contribution to leadership forecasting.
Leadership emergence
Leadership is more than the sum of its parts. As our model highlights, the leadership process is not merely the summation of members’ expectations, behaviors, and interactions. Rather it is the exchange of leadership and followership unique to each leadership occurrence and informed by both partners’ learnings from experiencing the leadership process. It is emergent. A recent review found no empirical studies of dyadic relational leadership in the workplace (Kim et al., 2020); this dearth of research does not, however, infer that emergent and relational leadership is unimportant—it is that emergent phenomena are difficult to grasp.
Our model sheds light on one reason why leadership emergence may be such an elusive construct. Namely, that it emerges out of both co-created constructionism (requiring dyadic cooperation) and reflective constructionism as each member takes away relational and adaptive lessons from having experienced the leadership process. In so doing, our model allows for an emergent, relational approach to leadership in which the interplay of one’s expectations and behaviors, the exhibition of leading and following behaviors by both parties, and the exchange of followership and leadership are all dynamic element of the leadership system. Exploring leadership through ideas grounded in sociology such as taking a perspective in favor of unfolding dynamic relationships, rather than a traditional focus on more fixed entities operating in a linear process (Abbott, 1988; Emirbayer, 1997) could be useful to expand our understanding and augment the system our model illustrates.
Identity and behaviors
Our focus on leadership emergence suggests that leader identity may be less important than behavior.Whether or not an individual identifies as a leader (follower), it is their leading (following) behaviors, and not their identification, which directly contribute to the leadership process. Our model suggests that holding a leader identity may be less impactful than having a partner who endows leadership. The model does not exclude the possibility of either partner holding leader and follower identities. It could be useful to raise questions as to whether members who hold a stronger leadership identity are more likely than those who hold a weaker leader identity to enact leading behaviors. Similarly, in dyads which reflect organizational hierarchy does the degree to which a subordinate holds a leader identity and a formal leader holds a follower identity effect the stability of their expectations over repeated exchanges? And do leader and follower identities shift as a result of experience which challenges expectations and provokes learning?
Influence of expectations beyond the dyad
Our model demonstrates that enactment of leading and following behaviors ultimately influence one another as the dyad engages in leadership and followership exchange. This offers some support to the assumption that leading behaviors motivate follower response (i.e., Breevaart et al., 2014) or more crudely that leaders motivate followers. By allowing for the notion that followers influence leaders, the model highlights gaps in leaders’ and followers’ understandings of each other from the individually siloed ILTs and IFTs and followership orientation literatures as a result of each perspective focusing on differing research emphases. ILTs specify leaders’ abilities (not behaviors) as identified by followers (not leaders). We know little about leaders’ ILTs. IFTs specify followers’ behaviors drawn from leaders’ ideal representations and understandings of followers. In contrast, followership orientation research explores followers’ understanding of the followership role. By accounting for the occurrence of leadership and followership in tandem, our model offers expansion on these under-considered phenomena. Researchers and practitioners are therefore urged to consider that both parties bring different ILTs and IFTs to their relationships (Engle & Lord, 1997).
Dyadic leadership research often touts the benefits of member agreement as a condition underlying the most effective forms of performance outcomes (Dulebohn et al., 2012; Martin et al., 2016). Yet, we cannot determine the level of agreement among dyadic members until we account for both parties’ cognitive schema. Our model provides a new pathway through which we may begin to better understand the development of expectations. Ultimately, whether future research focuses on the dyad as a unit of analysis or on the individual members of the dyad, it will be important to collect data from both members’ perspectives given the interactions and interplay of elements outlined in our model and our understanding of leadership as emergent (Kim et al., 2020).
The chain of effects in Figure 2 suggests a web of dyadic interactions. Each partner progresses into future phases of their shared relationship while concurrently participating in relationships with other partners. Repeated patterns and experiences within one relationship may be influential beyond that relationship by creating cognitive schema which remain in individuals’ internal knowledge structure across repeated partnerships (Medvedeff & Lord, 2007). Stability and change arising from expectations and behaviors in the leadership occurrence or system may contribute to organizational stability or to organizational change.
Contextual considerations
Our dyadic focus means that external contextual influences such as organizational or broader social norms and expectations are outside the scope of our model, though we recognize that they will have an influence. For our model to retain practical relevance, it is important to acknowledge real-world constraints associated with either hierarchical or social structures within which each leadership occurrence and relationship is bound. These would take the form of exogenous contributors to expectations that each party initially holds: elements such as societal expectations of leadership and followership, organizational culture, each member’s previous experiences of leadership, and their previous interactions with this partner. Such structural, reporting, or social constraints will play an influential role in shaping the initial form of the occurrence. For example, expectations rising from differences in societal culture have been theorized to influence employees’ behavior by shaping their implicit followership theories (Uhrbach et al., 2021). Even the way in which one processes learning from experience will be bound by such normative structures (Evaristo, 2007). Accordingly, future research is needed to consider how context, such as gender and varying levels of cultural influence may drive how each element of the leadership system plays out.
Leadership and followership development
Calls for leadership and followership development are not new (e.g., Carsten, 2017; Collinson & Tourish, 2015; Day & Lui, 2019). Yet, our conceptualization highlights the need for development that takes both a leadership and followership focused perspective. Communication plays a key role. For the dyad to work effectively, leaders and followers must communicate the degree to which the other meets one’s expectations. Our model provides a framework encompassing the different expectations held of oneself and one’s partner. Better understanding of these expectations and development to enhance the skill and adaptiveness with which behaviors are enacted would allow a more nuanced and specific approach to the development of self and other awareness. This has implications for formal and informal feedback including performance review and professional development, which should include opportunities to comment on met, unmet, and reconsideration of expectations. It also provides useful specificity to calls for more holistic leadership and followership development (e.g., Collinson & Tourish, 2015). In peer-to-peer relationships, for example, consideration of leaders’ expectations of other leaders could include peer assessment of others’ leadership capability or performance which could provide a foundation for behavioral anchoring as the leader enacts leading behaviors within her or his current leadership dyad. Similarly, but perhaps more radically, peer assessment of followership capability or performance might also be a helpful contributor to followership understanding and skill development. Furthermore we suggest that leadership and followership development should be integrated and available to all participants. A specific context where this would have significant personal and organizational impact is the MBA where a predominantly leader-focused perspective is typically taken (Hay & Hodgkinson, 2006; Lacerenza et al., 2017). Assuming that many MBA students undertake their studies to move into management roles or support recent promotion (Benjamin & O’Reilly, 2011) this would be a particularly opportune time for followership as well as leadership development to extend their own and their appreciation of others’ contributions in the workplace.
Forecasting
Perhaps the greatest practical implication of this research lies in its instrumentality for forecasting leadership patterns. Our framework stands in contrast to unidirectional strategies which remain common in organizational practice. By considering the mutual influence leader and follower expectations and behaviors on both members of the dyad, our focus on the enactment of leading and following behaviors, and the emergence of a co-created leadership process our model sets the stage for forecasting emergent leadership patterns. We briefly suggest three possible approaches which may prove fruitful.
The web of leadership systems depicted in Figure 2, which implies that changes in expectations have implications beyond the immediate dyad, suggests the potential for social network analysis to study and predict stability and change in leadership across organizations and relationships consistent with Emirbayer’s (1997) approach. It would be helpful for future research to ask questions regarding the degree to which influences from one dyadic relationship have a spillover effect into future leader and follower occurrences. At a broader level, this perspective could be useful in answering questions such as how leadership networks interact with each other and how leadership networks interact within and across imposed organizational structures. Our model’s chains and web of leadership processes also fits with notions that multiple complexity leadership processes (at different stages of development) occur across the organization (Uhl-Bien & Arena, 2018) simultaneously. Our model could therefore serve to introduce followership into complexity leadership theory which states that it is those who are not formally designated leaders who experiment and generate new understandings and ways of doing things in complex organizations (Uhl-Bien & Arena, 2017, 2018) but does not as yet consider followership. Further, our model provides a framework to consider questions as organizations develop over time. For example, it has the potential to consider questions of entrepreneurial and agile followership, in addition to leadership, as organizations evolve from startup to become more established.
Crucial to our model is its explicit assumption that partners are not restricted to or bound by their initial expectations. Freedom to change expectations, behaviors and thus the relationship and broader organization (the context within which the leadership process unfolds) is possible. A Bayesian (Andraszewicz et al., 2015; Van de Schoot et al., 2017; Zyphur & Oswald, 2013) approach, which incorporates incremental learnings into analysis, is likely to be of value. Similarly, Qualitative Comparative Analysis (QCA) enables researchers to identify the relevant, necessary, and sufficient conditions for an outcome to be reached (Ragin, 1987) The combination, including the absence, of conditions in the outcome and the possibility of more than one set of conditions leading to the same outcome (Ragin, 1987; Fiss, 2007) makes it potentially valuable for analyzing leading and following as we have modeled them rather than statistical methods that rely on linear algebra (i.e., Misangyi et al. 2017). Our model together with the analytical approach provided by QCA could be used to develop tools for forecasting multilinear leadership emergence from the interaction of co-constructed dyadic leadership. Researchers and practitioners interested in applying our model for predictive purposes will be further aided by recent advances in machine learning (i.e., Lee et al., 2020) which make accessible the tools necessary to develop time-series forecasts building on the patterns outlined in our model.
Conclusion
Leadership process has become an increasingly common approach to the study of leadership and followership in groups and organizations. Yet, the field lacks a unifying framework that explains where and how leadership process fits into the greater system of leadership. To address this gap, we have presented a followership-inclusive model that construes expectations and ensuing behaviors as inputs into an interactional exchange that we call the leadership process. This co-constructed process continues as members intertwine in a leadership dance of respectively endowing, acknowledging, accepting, and deferring through exchanges that adapt as the two continue to interact. Co-creation of leadership is complemented by individual learning outcomes as suggested by a time-lapsed recursive loop in our model that allows us to bring together constructionist notions of the co-creation of leadership process along with an extended intra-individual constructivist consideration through which experience of leadership process and its outcomes influence one’s expectations as they enter forthcoming leadership relationships. The lens offered by our framework sets the stage for several reimaginings of leadership and followership through which both leadership scholars and practitioners may build novel conceptualizations. It is our sincere hope that this model can reshape future considerations of leader-follower dynamics such that we are able to better understand and predict stability and change over time as members navigate an ongoing series of dyadic leadership relationships.
Footnotes
Acknowledgment
We are most grateful to the three anonymous reviewers and the Special Issue Editor whose thoughtful and constructive comments challenged us and significantly improved the paper. We also thank Dr Chia-Yen (Chad) Chiu, Professor Cheri Ostroff, and Professor Michelle Tuckey for their comments on an earlier version of this manuscript.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Ruth Sims’ PhD research is supported by an Australian Government Research Training Program Scholarship.
