What Comes After Black Swans and Gray Rhinos? An Organizational Perspective of Disruption

Current Understanding of Disruption

The concept of disruption has long interested innovation and organization scholars. It first gained popularity with Christensen’s (1997) use of the “disruptive innovation” label to describe how a new product or service initially takes root in simple applications at the low end of a market and then moves upmarket, eventually displacing established solutions. Organizational scholars define disruption more broadly, moving beyond the innovation domain. For example, Kumaraswamy et al. (2018, p. 1025) define disruption as “fundamental changes that disturb or re-order the ways in which firms and their ecosystems operate.” Similarly, Foss (2020, p. 1323) defines disruption as “a situation in which a low-probability or even entirely unanticipated event emerges that has drastic impact and consequences at a systemic level, not just upsetting relations between firms and their stakeholders within a single industry but hitting at the level of the entire economy.” More recently, Roulet and Bothello (2022) conceptualize disruption as an outcome of an event chain—a series of temporally and causally connected events. Some related terms used in macro- and micro-organization studies include rare event (Lampel et al., 2009), extreme event and extreme context (Hannah et al., 2009), disruptive event (Morgeson, 2005), and event system (Morgeson et al., 2015).

Adopting an organizational perspective, we conceptualize disruption as a discontinuity in organizations’ external and internal environments (Hoffman & Ocasio, 2001; Morgeson et al., 2015). The external environment includes various elements in organizations’ outer surroundings, such as technological innovations, regulation, competition, and the natural environment, while the internal environment comprises resources and conditions within organizational boundaries, such as physical and administrative structures, employees, processes, and outcomes (Dess & Beard, 1984). The discontinuity in organizations’ external and internal environments manifests as a sudden stop or change in the regular state and activities associated with external (e.g., the natural environment) and internal elements (e.g., processes and outcomes; Morgeson et al., 2015). For example, in the aftermath of an earthquake of magnitude of 7.9 (external environment change), organizations’ communication and coordination systems may stop working (a sudden stop in internal processes). Thus, our conceptualization of disruption has two components: discontinuity in the organization’s external environment, represented by an external event, and discontinuity in its internal environment, represented by that event’s impact on the organization’s internal processes and outcomes.

To explain the first component of disruption (external events), we draw on EST, which describes the interplay between external environment and impact on organizations (Morgeson et al., 2015). In developing EST, Morgeson et al. (2015) began with Allport’s (1940, p. 418) broad definition of events as “happenings between explicitly denotable things.” They then refined this definition in three ways to define events from an organizational perspective. First, events are observable circumstances that may originate from the external or internal environment. Second, events have a specific beginning and end time and progress in a specific setting. Third, events represent an interaction between entities (Weick, 1990). In summary, “Events are external, bounded in time and space, and involve the intersection of different entities” (Morgeson et al., 2015, p. 520).

As EST posits, not all external events are equally salient to organizational information processing and response (Morgeson, 2005; Nigam & Ocasio, 2010). Morgeson et al. (2015) identify novelty, disruption, and criticality as three key event attributes determining salience. While we draw from their general EST, our interest is in the subclass of disruptive external events, which share these three attributes. For Morgeson et al. (2015), disruption is only one of several event attributes; we seek to unpack that specific aspect of their model. In simple terms, we drill down into, and elaborate on, their second proposition: “The more disruptive an event, the more likely it will change or create behaviors, features, and events” (p. 521). Further, we add to their model the crucial element of organizational response and propose a more nuanced relationship between external events and organizational routine change.

To explicate the second component of disruption (the impact of external events on organizations), we focus on organizational responses and draw from the organizational routines literature (Becker, 2005; Cohen & Bacdayan, 1994; Feldman, 2000; Feldman & Pentland, 2003; Nelson & Winter, 1982). Organizations may respond to external events in many ways, such as doing nothing (i.e., keeping the existing way of doing things), creating new behaviors (e.g., organizational routines) or features (e.g., organizational structure) (Morgeson et al., 2015). We are interested in organizational routine change because routines—regular and predictable patterns of interdependent actions used by individuals and groups— are sources of organizational stability and change (Becker, 2004; Feldman, 2000; Feldman & Pentland, 2003; Nelson & Winter, 1982).

Organizational routines are traditionally considered as sources of stability (Cohen & Bacdayan, 1994; Nelson & Winter, 1982), yet practice theory-based research into routine dynamics suggests that routines are also a source of organizational change when triggered by exogenous factors (Becker, 2004; Feldman, 2000). Exogenous triggers of routine change may originate from the external environment, such as a fiscal crisis or new technology (Feldman, 2000), represented by external events in our framework. In responding to external events, four properties of organizational routines may facilitate change (Becker, 2005; Feldman, 2000; Feldman & Pentland 2003): entities, interactions among entities, context, and the temporality (timing and sequencing) of routines. First, entities perform routines and change them when actions do not produce intended outcomes (Feldman, 2000). Second, routines involve interactions among entities within and across organizations (Kremser & Schreyögg, 2016). Third, organizational constituents perform routines in a specific context but look for relevant routines in the larger organizational context (Becker, 2004). Fourth, because routines as patterns of actions are temporal and processual, and actions repeated over time can change, there are variations in routines performed by entities at different times (Feldman, 2000).

Although the existing perspective has provided valuable insights into disruption as a discontinuity in organizations’ external and internal environments, understanding of this phenomenon has been limited by two important research gaps: (1) how an organization makes sense of the external event; and (2) as an information-processing system comprising individuals and teams, how an organization reaches consensus on accepting change in response. This observation suggests the appropriateness of an information-processing lens to study disruption.

Disruption Through an Information-Processing Lens

Research suggests that external events prompt information processing and organizational response (Morgeson, 2005; Nigam & Ocasio, 2010). As information-processing systems, organizations assess their information needs, develop information-processing capabilities, and fit the former with the latter (Galbraith, 1974). However, amid the complexity and uncertainty that follow external events, organizations face difficulty in identifying what external information is salient and deciding how to use it to revise existing routines (Foss, 2020). For example, organizations may fail to accurately diagnose external events, correctly predict the impact on themselves, and properly evaluate the available response alternatives (Milliken, 1987). In particular, existing organizational processes may not produce solutions to the problems created by external events or enable new opportunities to respond thereto (Feldman, 2000). Thus, organizations need to obtain and analyze information on external events and match their information-processing capacities and needs.

However, in responding to external events, organizations face three major cognitive barriers (Becker, 2004; Chesbrough, 2010; Feldman, 2000). First, their limited information-processing capacities preclude processing the vast amount of complex and uncertain external information, which inhibits understanding of what is happening in the external environment and how to respond at any given time (Cyert & March, 1963; Galbraith, 1974; Gavetti & Levinthal, 2000). Second, even if organizations understand developments in the external environment, they may not make rational decisions because bounded rationality inhibits accurate interpretation of the cause–effect relationship between an external event and its impact on the organization (Cyert & March, 1963; Lant, 2002). Third, and more critically, because organizations are aggregations of individuals with multiple mental models, reaching common understanding on the interpretation of events and establishing a cause–effect relationship is difficult (Burke et al., 2006; Johnson-Laird, 1983; Kellermanns et al., 2008).

With respect to external events, shared mental models (Klimoski & Mohammed, 1994) can help organizations align the perspectives of individuals and teams, collectively identify potential challenges and opportunities, and develop effective responses (Cannon-Bowers et al., 1993). Prior research has conceptualized mental models at individual and collective levels. At the individual level, mental models are working schemas including descriptions, explanations, and observations of how things work, enabling individuals to organize, process, and store information quickly and flexibly (Cannon-Bowers et al., 1993). As representations of the external environment, mental models help individuals make sense of an event and generate effective responses (Burke et al., 2006; Johnson-Laird, 1983; Kellermanns et al., 2008).

Conceptualized at the collective level, mental models are representations of external events shared among various team members, or organizational constituents (Klimoski & Mohammed, 1994; Maynard & Gilson, 2014). Whereas prior research uses the concepts of shared mental models and team mental models interchangeably (Mohammed et al., 2010), we prefer the former in this article because we focus on the organization as the unit of analysis. Organizational constituents hold multiple mental models simultaneously (Cannon-Bowers et al., 1993; Rouse et al., 1992), and shared mental models are collective phenomena originating from individuals’ cognition (Kozlowski & Klein, 2000). Thus, we suggest that shared mental models represent a common understanding among organizational constituents of how the organization will operate in response to events in the external environment (Klimoski & Mohammed, 1994).

As a property of mental models, their sharedness ¹ has been a major discussion topic in organizational research (Mohammed et al., 2010). Sharedness is the extent of consistency between organizational constituents’ mental models, but does not represent having the same mental models (Cannon-Bowers et al., 1993; Mohammed et al., 2010). Mental models research has consistently indicated a link between mental models and behavioral outcomes, such as adaptation, coordination, communication, and routines (Burke et al., 2006; Mohammed et al., 2010; Waller et al., 2004). Hence, we posit that the SoMM is a moderating factor in the relationship between external events and organizational routine change.

Underlying Assumptions and Boundary Conditions

Before we present our theory, we articulate certain assumptions and boundary conditions employed in our theorization. First, we adopt a broad definition of disruption as a discontinuity in an organization’s environment: this requires the presence of an external event and an organization’s acceptance thereof, as indicated by organizational routine change. Second, as information-processing systems, organizations assess their information needs, develop information-processing capabilities, and pursue fit between the former and latter (Galbraith, 1974). However, as a collection of individuals and groups with at least partially conflicting interests (Eisenhardt & Zbaracki, 1992), organizations cannot easily process information and achieve common understanding of what the external event is and how it will impact the organization. Therefore, based on Mathieu et al. (2005), we assume that shared mental models contribute to organizations’ consensus building on the nature and outcomes of an external event. This assumption implies that shared mental models can influence the effect of external events on organizational routine change. Third, while organizations face various external and internal events, our theorization focuses only on the former. Internal events are also observable happenings but occur within organizational boundaries (Morgeson et al., 2015), so organizational constituents’ perceptions of those events are more complex due to their immediacy and emotional content.

For instance, consider a new CEO taking charge of an organization. Such executive changes can be emotionally and politically charged internal events because they may elicit uncertainty and insecurity among employees and may involve controversial or unpopular decisions by top-level executives (Smith & Grandey, 2022). Thus, to theorize on how internal events (e.g., a new CEO) impact organizational routine change, it may be necessary to consider observable yet covert organizational actions such as coalition building, lobbying, control of various organizational agendas, and manipulation and control of information channels (Eisenhardt & Zbaracki, 1992). Developing theory to explain how organizations respond to internal events is outside the scope of this study but could be the focus of future research. In concentrating on external events, we do not imply that internal events are less important, only harder to explicate in a single study.

Circumplex Model of an External Event

As stated earlier, EST proposes that “events are external, bounded in time and space, and involve the intersection of different entities” (Morgeson et al., 2015, p. 520). Thus, drawing from EST, we conceptualize the external event as a multidimensional construct to describe happenings in an organization’s external environment that vary in magnitude, involve interactions among entities, and are confined to a specific time and place. This conception combines four dimensions: magnitude, coupling (i.e., interaction), clustering (i.e., time), and space. We derive these dimensions by adapting Morgeson et al.’s (2015) three event components of strength, space, and time. More specifically, we isolate the magnitude dimension from event strength and the coupling, clustering, and space dimensions from event time and space. In deriving these external event dimensions, our approach is comparable to that of Roulet and Bothello (2022), who also drew on Morgeson et al.’s (2015) three event components to propose four aspects of a temporally and causally connected event chain: incremental change from one event to the next, interval between events, intersection of two event chains, and irregularity of events. Thus, our derivation approach is not only well-founded in core EST studies but also consistent with other recent contributions.

Guttman (1954, p. 325) introduced the circumplex model as a “system of variables which has a circular law of order.” There are three reasons why we consider this an appropriate model structure for external event dimensions. First, EST proposes that an external event influences entities across levels and time (Morgeson et al., 2015), suggesting that all dimensions are equally important and agnostic to hierarchical levels. Second, there is no empirical evidence indicating a hierarchical structure among the different dimensions (Foss, 2020; Hällgren et al., 2018; Hannah et al., 2009; Kumaraswamy et al., 2018). Finally, the circumplex model is ideally suited to describing both independent and interrelated dimensions of a concept (Russel, 1980).

We now describe the four dimensions of our EST-aligned external event construct to develop our circumplex model. For each relevant external event dimension, we provide stylized examples that communicate the essence of the dimension, acknowledging that those examples will never capture all aspects of reality. For ease of theoretical exposition, we present the external event dimensions as independent of one another—a necessary simplification in the first stage of theory development. However, the very construction of a circumplex model suggests that these dimensions may operate simultaneously: we will elaborate in the Discussion section.

The magnitude dimension denotes the size of an external event, which can range from small to large (Hällgren et al., 2018). The magnitude can be determined by objective assessment independent of an event’s effect on the focal organization. ² For example, the magnitude of an earthquake is measured by an objective number unaffected by one’s proximity to the epicenter or what the tremors feel like (US Geological Survey, 2023). Magnitude is an important dimension because organizations operate in a steady state until a small or large external event occurs (Brammer et al., 2020), and large events more likely require an exceptional organizational response (Morgeson et al., 2015).

Expanding on the earthquake example, larger earthquakes require bigger organizational responses than smaller ones. Consider two recent earthquakes: The first occurred in southeastern Türkiye on February 6, 2023 and measured 7.8 in magnitude, according to the USGS. This was categorized as a large earthquake and required responses from local and global organizations. The USGS Office of Communications and Publishing issued a press briefing to communicate the size of the earthquake to the world. The second occurred in The Geysers, CA on April 9, 2023, with a magnitude of 2.6. Because this earthquake was categorized as small, it did not require an exceptional organizational response.

The coupling dimension refers to the number of elements involved in an external event (e.g., the number of entities, technologies, transportation links, and communication channels) and the interactions among them. Increases in the number of and extent of interactions between elements may bring differences in the external event’s diffusion and in organizations’ response thereto (Kumaraswamy et al., 2018). Orton and Weick (1990) characterized the degree of interactivity in systems as coupling. Since modularity facilitates the management of complex systems by dividing functions and structuring interfaces (Pil & Cohen, 2006), interactivity pulls systems in the opposite direction of making them more complicated to manage. The more tightly coupled a system’s components, the more sensitive it is to subsystem malfunctions and other perturbations; at the other end of the spectrum, loosely coupled systems are more resilient and forgiving because their components retain some measure of independence (Orton & Weick, 1990).

For example, consider the Spanish flu of 1918 and COVID-19, two different global pandemics that respectively emerged from America and China. During the Spanish flu pandemic, medical technology was less well developed than today, no vaccines were available, institutions such as World Health Organization did not exist, healthcare systems were not developed, and there was little effective knowledge sharing among countries and scientific communities. As a result, 50 million people died. By contrast, while the COVID-19 pandemic entails multiple elements, organizations worldwide are responding relatively well by sharing knowledge and ideas (Barton et al., 2020). Moreover, advances in medicine and biotechnology have enabled scientists to develop new treatments and vaccines more efficiently. Telecommunication technologies have allowed information sharing across countries and enabled a significant portion of the workforce to work remotely, thereby making the economic downturn less severe. While the effects of COVID-19 have been exacerbated by behavioral responses to the virus and by government policy interventions such as travel restrictions and closure of non-essential businesses (Baker et al., 2020), organizations with appropriate resources or capacity to direct them have been better able to respond to this pandemic.

Another insightful comparison is found in the financial world. On March 10, 2023, Silicon Valley Bank (SVB) failed after a bank run—an example of an event low in coupling. SVB was a lender to new startup technology companies, most of which were not listed and had already received startup money from outside investors. The main entities involved in the failure event were SVB itself, tech startups in Silicon Valley that deposited money in SVB, and the Federal Reserve (as the primary regulator). By contrast, the Lehman Brothers bankruptcy was an event high in coupling: in addition to depositors and the Federal Reserve, entities involved in the failure devised and traded various financial products such as derivatives, credit default swaps, and exotic mortgages, all requiring connections among different financial products and institutions around the globe.

The clustering dimension refers to whether the triggering external event is a single occurrence or a set of related, temporally proximate happenings (Roulet & Bothello, 2022). Some external events may be discrete and emerge periodically, while others may be more processual (Hällgren et al., 2018; Hannah et al., 2009; Kumaraswamy et al., 2018). We label the former as episodic and the latter as continuous: whether an external event falls in the former or latter category becomes apparent over time. Episodic and continuous external events can lead organizations to move away from a steady state and respond by operating at a new equilibrium.

A useful illustration of episodic external events is terrorist attacks—discrete events with a definite beginning and end. For example, the attack on the World Trade Center in New York City on September 11, 2001 was an episodic external event. Although limited in time (9/11/2001) and scope (New York City), it resulted in much larger organizational responses: across the United States, airports, public buildings, and private offices/facilities implemented or upgraded security systems. By contrast, one continuous external event is urbanization across the developing world, as human populations move from rural areas to cities. This ongoing process of growth and development of urban areas encompasses physical infrastructure, education, affordable housing, and other services to meet the growing needs of people concentrating in cities and towns. The World Bank (2023) forecasts that the world’s urban population will reach 6 billion by 2045. Thus, urbanization will require responses from organizations in numerous industries such as education, construction, and health services.

The space dimension involves the distance between two elements of an external event: its origin and the organizations impacted. An external event originates from organizations’ external surroundings, and its effects may manifest where the event occurred or elsewhere (Ahlstrom et al., 2020). The effects of an external event may diffuse via the micro, meso, macro, and mundo loci (Morgeson et al., 2015). For instance, an external event originating in the micro locus may affect organizations in the same locus and then diffuse to other loci. In the micro locus, it is often possible to identify an external event and observe its effects on organizations. Where the origin and effects of an external event are in the same locus, they are considered proximate. We use the term distance in a broad sense that includes time and physical, psychological, technological, and social distance (Hannah et al., 2009). An external event originating in the micro locus but causing effects in the meso, macro, or mundo loci may not be directly identified in the latter, even if its effects can be observed there.

To illustrate, consider the discovery of semiconductors as an external event. Semiconductors are materials whose conductivity lies between that of conductors (e.g., copper) and insulators (e.g., plastic). Since their discovery, semiconductors have enabled the development of wide-ranging technologies (Holbrook et al., 2000); driven advances in various industries, such as communications, computing, health care, military systems, and transportation; and facilitated the creation of new technologies, such as robotics and artificial intelligence. Thus, the technological distance between semiconductor discovery and electronics is proximate. By contrast, consider as an external event the invention of mini steel mills, which use an electric arc furnace to produce steel from recycled steel, scrap, and direct reduced iron. Mini steel mills had no impact beyond steel and steel-consuming industries; thus, the technological distance between their invention and other industries is not proximate.

Main Effects: How External Events Impact Routines

We posit that the magnitude of an external event influences organizational routine change because individual cognitive biases in processing information result in greater attention being directed toward large external events. First, an organization faces numerous external events, of which some are more salient and command more attention at any given time (Morgeson et al., 2015). Large external events are severe occurrences in organizations’ external environment that are expected to result in significant organizational distress (Mithani, 2020; Zellmer-Bruhn, 2003). We contend that organizations’ information processing is prone to salience or vividness bias, such that decision-making is based on the most recognizable and large occurrences, rather than all occurrences in the external environment (Tversky & Kahneman, 1974). Thus, organizations are more likely to change their routines in the aftermath of a large external event.

Second, large external events change organizational attention structures, which govern the allocation of time, effort, and attentional focus in organizational decision-making (March & Olsen, 1976). Large external events attract the organization’s attention and increase its commitment to follow through on decisions regarding how, when, and for how long to respond (Hannah et al., 2009; Staw, 1997). Further, organizational decision-makers are motivated to devote more effort toward making sense of a large external event and taking relevant actions in times of significant distress (Feldman & Pentland, 2003), facilitating the cause–effect relationship between a large external event and the decision to change organizational routines. For example, organizations located in a region prone to big hurricanes will more likely change organizational routines to methodically assess meteorological information and ensure quick reactions to an approaching Category Four hurricane. Thus:

Proposition 1:

The magnitude of an external event is positively related to organizational routine change, such that events of greater (vs. lesser) magnitude result in more (vs. less) organizational routine change.

We also suggest that as the degree of interactivity among external event elements increases, organizations are less likely to change their routines—for two main reasons. First, when an external event has many interacting elements with different values, its effects depend on the states of all elements combined (Csaszar, 2018). Where numerous system elements interact, they create a complex system because changing one element may also necessitate changes to other elements (von Bertalanffy, 1950). Yet, as Garud et al. (2011, p. 740) state,

organizations are often unable to deal with such complexities because they have been designed to reduce or suppress them. For instance, organizations may adopt a “boxes within boxes” approach (March & Simon, 1958) that reduces interactions and locks people into “thought worlds” (Dougherty, 1992).

As interactions become more complex, it is less likely that organizations can respond effectively to changes in their environment (Levinthal, 1997). Moreover, poor understanding of interactions between elements of an external event can make it difficult to discern which are most relevant to organizational routine change. Thus, with rising complexity of the many elements of an external event and their interactions, it becomes more difficult for a boundedly rational organization to accurately recognize relevant cues and, hence, make rational decisions on what organizational routines to change (Burke et al., 2006; Cyert & March, 1963).

Second, the interactions among elements of an external event likely lead to interdependencies governed by a huge set of rules or heuristics (Cohen et al., 1996). The increased information complexity and ambiguity from a large number of interacting elements can lead to organizations developing ambiguous and inconsistent rules on how to change routines to diagnose and solve complex problems created by an external event (Saavedra et al., 1993). Thus, when an external event has many interacting elements, it is more difficult for organizations to establish accurate cause–effect relationships, ascertain the temporal precedence of these event elements, and assess which are non-spurious (Duncan, 1972; Ocasio, 1997). For example, the COVID-19 event is characterized by many unknown, interacting factors such as information overload and decision paralysis (Foss, 2020) that complicate accurate assessment of the complex situation and formulation of an appropriate response. Owing to cognitive limits on organizations’ processing of complex information and the interdependencies among external event elements, an inappropriate response to one external event element is more likely to create an incorrect solution, resulting in a snowball effect that hinders routine change. Thus:

Proposition 2:

The interactivity among external event elements (coupling dimension) is negatively related to organizational routine change, such that events with greater (vs. lesser) interactivity result in less (vs. more) organizational routine change.

We next posit that the temporal flow of external events (episodic or continuous) is related to organizational routine change because time-related information-processing behaviors such as time awareness, task prioritization, and task scheduling influence what organizational routines to change and when (Conte et al., 1995). By definition, an episodic external event is transitory in nature and, hence, requires information processing of one-time event—limited in time or scope—and only the current impact of the event (Cohen & Bacdayan, 1994). Organizations are stability-seeking entities with a natural tendency to revert to familiar behavioral patterns once a threat has passed (Nelson & Winter, 1982)—a “punctuated equilibrium” perspective (Romanelli & Tushman, 1994). Since episodic external events are discrete occurrences with a clear beginning and end, organizational constituents may feel time-pressured to process large amounts of information. The time pressure can inhibit establishing unambiguous cause–effect relationships between an external event and appropriate changes to organizational routines (Waller et al., 2004).

By contrast, continuous external events are ongoing and require information processing of past and current actions to determine the best next action (Cohen & Bacdayan, 1994). During these events, organizations have a series of opportunities to address the discrepancy between time-related behaviors and appropriate organizational routines because the event’s effects are processual and cumulative. This increases the likelihood of identifying an unambiguous cause–effect relationship between the external event and required change to organizational routines (Hällgren et al., 2018). Unlike episodic external events, a series of external events involve multiple exposures to the sources and effects of past external events, allowing organizations to rely on memory of past actions to determine what the best next actions are and how they should be executed (Cohen & Bacdayan, 1994). Thus:

Proposition 3:

The continuousness of an external event (clustering dimension) is positively related to organizational routine change, such that more continuous (vs. episodic) events results in more (vs. less) organizational routine change.

We propose that with increasing separation (e.g., in time or physical, psychological, or technological distance) from the external event origin, it is more difficult for the focal organization to change routines—for two reasons. First, although organizations attend to and process information from their external environment (Lant, 2002), information-processing limitations restrict what events and responses they can consider at any given time (March & Shapira, 1992; Ocasio, 1997). The physical and temporal characteristics of external events influence organizations’ focus of attention (Ocasio, 1997), and they are less likely to attend to external events considered distant and with uncertain outcomes (Ocasio, 2011). By contrast, when organizations face a physically or temporally immediate threat, they are likely to engage more intently (McKean, 1994). For example, a Category Four hurricane only requires immediate situation assessment and response (in terms of starting evacuations) in locations close to the eye.

Second, physical and temporal distance between external event origin and its effects on an organization makes it more difficult to recognize the cue and ascribe meaning to it (Burke et al., 2006; Feldman, 2000), leading to ambiguity in the perceived cause–effect relationship. Organizations apply three criteria to establish accurate cause–effect relationships that inform rational decisions: association between the external event and organizational response, temporal precedence of the external event, and non-spuriousness (Duncan, 1972; Ocasio, 1997). We suggest that when there is physical and temporal distance between the event and its effects, organizations cannot observe the external event directly and concurrently experience its effects. This makes it harder for organizations to associate the external event with its effects (which may result from various external events) and to determine whether the event is a momentary glitch or signals deeper change. The same external event will be differently perceived and interpreted by two entities at different physical and temporal distances from it. With increasing separation between the external event and its effects, it becomes more difficult for organizations to change routines because their attention to distant external events is limited and establishing a clear cause–effect relationship is challenging. Thus:

Proposition 4:

The degree of separation between an external event and its effects on the focal organization (space dimension) is negatively related to organizational routine change, such that greater (vs. lesser) separation results in less (vs. more) organizational routine change.

Moderating Effects: How the SoMM Moderates the Impact of External Events on Routines

Proposition 1 suggests that the magnitude of an external event is positively related to organizational routine change. We now posit that this relationship is moderated by the SoMM: the greater the SoMM, the more positive the event–change relationship. As noted above, mental models include the knowledge, beliefs, assumptions, and/or perceptions shared by organizational constituents, and play a key role in determining organizational perception and actions (Klimoski & Mohammed, 1994; Maynard & Gilson, 2014). Information consistent with an existing mental model is more easily absorbed than information outside that model (Houghton et al., 2000). Consistent with this logic, as the SoMM increases, organizational constituents will more likely converge on accepting that an external event requires organizational response (Klimoski & Mohammed, 1994; Maynard & Gilson, 2014). Thus, it is reasonable to suggest that greater SoMM strengthens the relationship between external events and organizational routine change by (i) amplifying the vividness or salience biases that associate large external events with organizational distress and (ii) inducing convergence on a course of action to change organizational routines. Furthermore, because shared mental models facilitate identifying the relevance of a certain situation (Burke et al., 2006), rising SoMM leads to a more obvious relationship between a large external event and organizational routine change. Thus:

Proposition 5:

The SoMM strengthens the positive relationship between the magnitude of an external event and organizational routine change, such that the effect becomes more positive as the SoMM increases.

Proposition 2 posits a negative relationship between the degree of interactivity among external event elements (the coupling dimension) and organizational routine change. We now predict that the SoMM will attenuate this negative effect. More specifically, we propose that higher SoMM weakens the negative relationship by (i) allowing organizations to deal with the complexities created by interactions among many elements and (ii) reducing inconsistency in decisions that results from interdependent elements that vary in the number and strength (Garud et al., 2011; Orton & Weick, 1990; von Bertalanffy, 1950). As the SoMM increases, organizational constituents are likely to process information more quickly and less likely to be distracted by irrelevant information (Houghton et al., 2000). This increases the likelihood of them developing consistent mental models for the external event elements and their interactions leading to organizational routine change. For example, when an external event involves a large number of elements, different organizational constituents interpret individual elements and their interactions in divergent ways, reflecting their specific mental models; the greater the range of mental models, the more divergent are the causal narratives. Conversely, as the SoMM increases, the flexibility and implementation of organizational decisions also increase (Walsh & Fahey, 1986), facilitating development of unambiguous cause–effect relationships. We posit that as the SoMM rises, the organization becomes more likely to process complexity more efficiently to reach consensus and establish accurate cause–effect relationships among loosely coupled system elements, ultimately attenuating the negative effect of external event element interactivity on organizational routines. Thus:

Proposition 6:

The SoMM weakens the negative relationship between interactivity among external elements (coupling dimension) and organizational routine change, such that the effect is less negative as the SoMM increases.

Proposition 3 suggests that a continuous (vs. episodic) external event will more likely lead to organizational routine change. We expect that this effect is strengthened by higher SoMM. As the SoMM increases, mental models become more consistent over time (Kellermanns et al., 2008). Continuous external events require uninterrupted information processing of past and current actions to determine the best next action (Cohen & Bacdayan, 1994). Therefore, greater SoMM increases the likelihood of organizations making consistent decisions following an assessment of each external event, in turn raising the likelihood of organizational routine change. By contrast, episodic external events are transitory in nature and each requires organizational action in response (Cohen & Bacdayan, 1994). Thus, we propose that for episodic external events, higher SoMM can hinder organizational routine change. For example, a transitory external event may later be found to be a temporary glitch, such that no organizational routine change was required following its occurrence. Thus:

Proposition 7:

The SoMM strengthens the positive relationship between the continuousness of an external event (clustering dimension) and organizational routine change, such that the effect is more positive as the SoMM increases.

Proposition 4 posits that the greater the degree of separation between external event origin and the effects on the focal organization, the less likely is organizational routine change. We now propose that the SoMM reduces this negative impact. Our logic is straightforward: greater SoMM reduces the likelihood that the degree of separation give rise to discordant explanations of the link between external event origin and effects. The degree of separation between origin and effects creates discrepancy in perceptions and interpretation of the same external event among different individuals and teams, making it harder for organizational constituents to develop consensus on the need for (and direction of) routine change. For example, greater distance between external event origin and effects leads to divergent constructions of the causal linkage, reflecting individuals’ respective mental models. The greater the range of mental models, the more divergent are the causal narratives; conversely, the more consistent the mental models (i.e., the greater the SoMM), the more consistent those narratives become. When different individuals and teams have consistent causal narratives on the distance between origin and effects, it is easier to achieve consensus on whether the external event is mere noise or signals fundamental change. Correspondingly, ceteris paribus, it is also easier to agree on how to change routines to adapt to the new environment. By contrast, when causal narratives are discordant, heightened state, effect, and response uncertainty (Milliken, 1987) inhibit change. Sharedness improves the consistency of causal narratives in the organization, thus influencing the relationship between origin–effects separation and organizational routine change. Thus:

Proposition 8:

The SoMM weakens the negative relationship between the degree of separation (space dimension) and organizational routine change, such that the effect is less negative as the SoMM increases.

Theoretical Contributions

Overall, we offer a comprehensive framework of disruption by examining its two components—external event and organizational routine change—and the relationship between them. Our study thus makes three primary contributions to contemporary theorizing on organizational responses to external events. First, we develop a circumplex model of external event dimensions that posits the disruptive effect of each on the focal organization. We thereby extend Morgeson et al.’s (2015) EST by drawing out the specific attributes of disruptive events—a relevant subclass of external events—and how those attributes impact organizations. Second, we model the relationship between event attributes in the circumplex model and organizational routine change. While routines are widely accepted as simultaneous sources of organizational change and stability (Feldman, 2000; Feldman & Pentland, 2003; Pentland et al., 2011), the field is yet to address how external event dimensions interact with organizational elements to promote or prevent routine change. Third, we offer a systems framework (Ashmos & Huber, 1987; Morgeson et al., 2015) of that relationship, using the information-processing lens to highlight how the SoMM moderates the impact of external event dimensions. The literature on disruptive events has not previously linked external event dimensions and the difficulty or ease of understanding and responding thereto. By viewing the external event and organizational routine change as related sources and consequences of disruption, we suggest that it is crucial to model how external event dimensions differentially influence change to routines.

Practical Implications

The circumplex model of external events promises to help policymakers, emergency responders, and managers of for-profit firms in preparing for and responding to these events. By understanding the position of an external event on the circumplex model, organizations can avoid conflict on an event’s underlying characteristics and develop more effective communication and responses. For example, we described the SVB’s failure as a low-magnitude external event with low coupling. Because SVB was not highly connected with global financial markets, its failure did not require the Federal Reserve and Treasury Department to take drastic system-wide measures: they simply announced that all SVB clients would be protected and able to access their money. By contrast, the Lehman Brothers bankruptcy was a high-magnitude event with high coupling. Ceteris paribus, had regulators recognized these two dimensions (which made Lehman Brothers a systemically important financial institution), they could have rescued the firm and probably averted the global financial crisis (Stewart & Eavis, 2014).

Methodological Implications

Although many of our theoretical arguments have some grounding in empirical research, future studies should empirically test our propositions. Testing both the main and moderating effects we have proposed will provide a more integrative view of disruption involving external event dimensions, the SoMM, and their effects on organizational routines. In event-oriented studies, there are several issues with data collection and empirical analyses to address. Morgeson et al. (2015) suggest that the appropriate empirical approach to testing an event system model is based on two considerations: whether the study focuses on analyzing events or reactions thereto. Our model includes both analysis of the external event and an organization’s direct and contingent response. Hence, to empirically test our model, data should be collected from both external and internal sources. For example, to identify and describe external events, researchers will need to collect environment-level data from sources external to the organization, including newspapers, online news websites, and social media websites; to assess the direct and contingent impact of external events on organizations, researchers will need to collect organization-level data from internal sources through interviews, observations, and questionnaires. Regarding statistical analysis, hierarchical linear modeling is an appropriate technique to analyze multilevel data obtained from interdependent sources.

Operationalizing some constructs, such as change in organizational routines, may prove challenging. However, the framework lends itself well to testing. For instance, the external event dimensions (magnitude, coupling, clustering, and space) could be measured with Likert-type scales on a questionnaire completed by representatives (e.g., team leaders, managers) of various organizations. Specifically, the space dimension could be measured by asking managers to assess organizational constituents’ perceptions of the proximity of an external event in terms of physical (Euclidian) distance and time (days, months). The same measurement approach would work for the other external event dimensions (Sarta et al., 2021). Furthermore, the SoMM can be measured via three methods: (i) elicitation (e.g., similarity ratings, concept maps, questionnaires, and rating scales); (ii) structure representation (e.g., card sorting); and (iii) representation of emergence (e.g., percentage overlap between groups and within-group inter-rater reliability; Mohammed et al., 2000). Finally, regarding change in organizational routines, researchers should measure not only which specific actions are performed but also how (Suchman, 1983). Thus, organizational routine change can be studied using qualitative approaches like interviews, observation, and ethnographic interaction (Aroles & McLean, 2016).

Limitations and Future Research

Aiming for tractable theorizing, we made a few boundary choices. First, we did not differentiate positive and negative external events—i.e., whether an external event is perceived as beneficial or detrimental to the focal organization. Future research may add further nuance by investigating if and how organizational behavior (response) varies for positive and negative external events. Second, we focused on external events as an exogenous phenomenon, treating internal events as outside scope. Two considerations dictated this choice: first, external events have more obvious system-wide impact. Second, internal events need more complex modeling, involving attributes such as emotionality and political charge, which potentially distort information-processing considerations. Future research may build on our study by focusing on internal (to the organization) events and theorize whether and how the effects of internal events on organizational routine change differ from those of external events.

Third, another important simplifying choice was to treat the four dimensions of external events as independent, which allowed us to theorize their direct effects with greater clarity and depth. However, our use of the circumplex model acknowledges that the dimensions may operate simultaneously. ³ For example, a low-magnitude event that is proximate to the focal organization (i.e., low separation) could be more impactful than a high-magnitude event that is more distant (i.e., high separation). Similarly, a continuous series of low-magnitude events may build cumulative impact that remains unnoticed by the organization until a fateful tipping point. Furthermore, a given external event may be high in magnitude and coupling, while another may be high in magnitude but low in coupling.

A useful illustrative comparison can be made between the 2011 tsunami in Japan and the 2023 earthquake in southeastern Türkiye. The Japanese tsunami began with an earthquake (9 in magnitude) off the northeast coast—a manufacturing hub with a large nuclear power plant and companies highly connected with global supply networks. Although the physical effect of the tsunami was not directly felt around the world, the high coupling of affected Japanese companies with global supply networks meant that a significant drop in industrial production impacted the economies of Western countries. By contrast, the recent 7.8 magnitude earthquake in Türkiye was also high in magnitude but lower in coupling. Southeastern Türkiye does not house global suppliers and is not well connected with other parts of the world. While both earthquakes were considered high-magnitude by the USGS, they differed in degree of coupling because of differences in the number of and interactions among external event elements. We can extend these examples to different combinations of external event dimensions. Thus, the current model leaves room for future development of theory on how the four external event dimensions interact to produce more complex effects.

Our study also suggests that measurement of the SoMM should take into account the emergent and dynamic nature of disruption. This might involve comparing past and future states of the world, perceptions of convergence or divergence between mental models, and organizational constituents’ subjective confidence in their own (vs. others’) mental models. Given these nuances, recent scales for measuring mental models (see, e.g., van Rensburg et al., 2022) show promise for empirical study of the SoMM in disruptive environments.

Conclusion

Overall, our circumplex model of external events and our information-processing view of their impact on organizational routines update understanding of how external changes produce internal changes. As managers and policymakers seek to sustain competitive advantage and maintain economic and societal stability in an environment with constant disruption, these theoretical contributions can lead to new organizational strategies for agility and adaptation.