can u see a file attached

Hind Benbya and Ning Nan served as associate editors for the special issue. Huseyin Tanriverdi and Youngjin Yoo served as senior editors. William McKelveywas a SE for the special issue but was unable to participate in the writing of the introductory essay.DOI: 10.25300/MISQ/2020/13304MIS Quarterly Vol. 44 No. 1, pp. 1-17/March 20201 Benbya et al./Introduction: Complexity & IS ResearchIntroductionWhen we conduct a search on Google, it returns hundreds, ofthousands, results instantaneously. The results not onlyreflect the interests of the one who is doing the search, butalso the millions of internet users who created or clicked onhyperlinks of websites. As more users search, link, and clickwith similar keywords, the results will continue to changeaccording to user location and search time. A search for“Korean restaurants” in Munich, Germany, for example, givesdifferent results from a search in Cleveland, OH, USA. Con-ducting the same search a day or two later also producesdifferent results. A simple Google search result is an emer-gent property, a complex web of interactions among users,websites, topics, advertisers, and many other social or tech-nical entities. In short, our daily experience of using mundanedigital tools is a dynamic emergent outcome of complexsociotechnical systems.As early as 2010, the world-wide production of transistors hasexceeded that of rice, and is much cheaper (Lucas et al. 2012). Devices—large and small—powered by microprocessors andconnected by the internet are filling every inhabited corner ofthe earth. Some of these devices are not just passivelywaiting for commands; equipped with a powerful artificialintelligence engine, they often act on their own. We alreadysee autonomous vehicles on the streets interacting with trafficsignals that respond to changing traffic patterns, in the midstof human-controlled vehicles and pedestrians. Sprinklers areconnected to the weather service on the internet to control theamount of water on a lawn. The temperature of millions ofhouses is controlled by Nest connected to the Google HomeAssist service. Connected speakers recommend differentmusic playlists based on the time, location, and, of course,your preference. Social network services also enable everyuser as a potential content creator on the internet. Oncecreated, user-generated content can be liked, shared, andmashed with other content by other users, often creatingunpredictably complex forms of diffusions. Digital platformecosystems such as Uber and AirBnB connect millions ofusers and providers globally. More than 80% of movieswatched on Netflix are recommended by algorithms.2These examples illustrate truly astonishing advances from thehumble start of computers in organizations in the early 20thcentury. After merely a few decades, what once seemed to beglorified calculators have evolved into digital technologiesthat permeate our lives and work. These digital technologiesin turn foster new sociotechnical systems such as wikis, socialmedia, and platform ecosystems that are fundamentallychanging the way people work and live.Not every technological invention has such a transformationalimpact. What set apart digital technologies? At the heart ofdigital technologies is symbol-based computation. Bistrings(0s and 1s) provide a standard form of symbols to encodeinput, process, and output of a wide variety of tasks (Faulknerand Runde 2019). They reduce the design specificity of hard-ware for operationalizing the symbol-based computation. Furthermore, simplicity of bitstrings eases the effort to shrinkthe size, reduce the cost, and increase the processing power ofhardware. Symbol-based computation provides a generali-zable and applicable mechanism to unite the operations ofmatter and the abstract mental processes (Lovelace 1842). Itlays the foundation for digital technology to rapidly advancebeyond the function of a calculator. More importantly,symbol-based computation sets in motion the emergence ofcomplex sociotechnical systems.Emanating from symbol-based computation are a fewcomplexity-inducing characteristics of digital technologies.•Embedded: as described by the vision for symbol-basedcomputation (Lovelace 1842; Shannon 1993, Turning1950), digital capabilities are increasingly embedded inobjects that previously have pure material composition(Yoo et al. 2012). Digital capabilities can encode andautomate abstract cognitive processes for converting newinformation into adaptive changes of objects. They alsoenable objects to provide decision support to adaptivecognitive processes of social actors. •Connected: objects embedded with digital capabilitiesand users of such objects can be connected into webs ofsociotechnical relations (Sarker et al. 2019) becausesymbol-based computation homogenizes data (Yoo2010). When information is shared in the webs of socio-technical relations, abstract cognitive processes encodedin objects or possessed by social actors become mutuallydependent.•Editable: digital technologies are editable (Kallinikos etal. 2013; Yoo 2012) due to symbol-based computation. This editability allows increasingly diverse cognitiveprocesses to be introduced into the webs of socio-technical relations. Recurrent adaptation of diverse,connected, and mutually dependent objects and socialactors can amplify or diminish an initial change in asociotechnical system, producing outcomes that defysimple extrapolation from the initial change (Arthur2015; Holland 1995; Page 2010). Complexity, therefore,becomes a salient attribute of sociotechnical systems.2

See https://mobilesyrup.com/2017/08/22/80-percent-netflix-shows-discovered-recommendation/.2MIS Quarterly Vol. 44 No. 1/March 2020 Benbya et al./Introduction: Complexity & IS Research•Reprogrammable: through the separation of hardwareand software of symbol-based computation, digital tech-nology is reprogrammable (Yoo et al. 2010). The samehardware can perform different functions depending onthe software that runs on the device.•Communicable: digital technologies are communicableby following a set of agreed-upon protocols (Lyytinenand King, 2006; Yoo 2010). With the pervasive diffu-sion of digital technologies, they now form a globaldigital infrastructure (Tilson et al. 2010).•Identifiable: each and every device connected to thedigital infrastructure is uniquely identifiable through itsown unique address (Yoo 2010). The increasing digitalpenetration leads to a higher degree of identifiability,allowing for more granular manipulation levels of digitalobjects.•Associable: digital objects are associable through sharedtraits. The associability of distributed heterogeneousdevices and data allows one to identify emerging patternsacross different realms and geographies in a way that wassimply not possible in the past.Digital technologies not only give rise to complex sociotech-nical systems; they also distinguish sociotechnical systemsfrom other complex physical or social systems. While com-plexity in physical or social system is predominantly drivenby either material operations or human agency, complexity insociotechnical systems arises from the continuing andevolving entanglement of the social (human agency), thesymbolic (symbol-based computation in digital technologies),and the material (physical artifacts that house or interact withcomputing machines). The functions of digital technologiesand the roles of social actors are perpetually defined andredefined by each other (Faulkner and Runde 2019; Zittrain2006). This sociotechnical entanglement limits the generali-zability of complexity insights obtained from nondigitalsystems to complex digital systems. Furthermore, whilematerial operations or human agency either increase ordampen complexity in physical or social systems, digital tech-nologies can both mitigate and intensify complexity. This isbecause individuals and organizations engaged with complexsociotechnical systems often turn to digital technologies (e.g.,data analytics) for solutions to complex problems. Yet, theapplication of a solution can instigate a new round of digitallyenabled interactions that diminish the intended effect of thesolution. This dual effect of digital technologies on com-plexity can produce dynamic interaction patterns and out-comes that are qualitatively different from those in othercomplex systems. The distinct effects of digital technologies on complex socio-technical systems present an important opportunity for infor-mation systems (IS) researchers to extract novel insightsregarding the nature and relevance of digital technologies. ISresearchers can apply theories and methods from complexityscience to model observations that defy simple extrapolationfrom initial changes in a sociotechnical system. In this essay,we introduce key complexity theories such as emergence,coevolution, chaos, and scalable dynamics as the most likelyfoundation for IS researchers to rethink predictability, caus-ality, boundary, and durability of observations in the digitalworld. Subsequently, we explain how the centrality ofsymbol-based computation in IS research paves the way forIS-specific research themes to extend complexity science. The articles in this special issue are briefly described to illus-trate a few prominent themes such as IS development forrapidly changing requirements and using digital technologiesto steer or tame complexity.Complexity Science: Key Theoriesand MethodsComplexity science’s origins lie in 50 years of research intononlinear dynamics in natural sciences and spans a variety ofscholarly disciplines including biology (Kauffman 1993),chemistry (Prigogine and Stengers 1984), computer science(Holland 1995; Simon 1962), physics (Gell-Mann 1995), andeconomics (Arthur 1989). Developments across disciplinesover time resulted in a meta-theoretical framework withinwhich several theoretically consistent approaches andmethods can be integrated.Complexity science theories and methods combine differentepistemologies (i.e., positivism, interpretivism, and realism)to provide novel opportunities to question assumptions (e.g.,equilibrium, stability, etc.), manage tensions and paradoxes,and rethink the way we view many sociotechnical phenomenaat the center of our field. Their value is particularly promi-nent when the research community faces new phenomena andquestions that do not lend themselves well to the traditional,reductionist approaches.Complexity Drivers and StatesComplexity is an attribute of systems made up of large num-bers of diverse and interdependent agents3

that influence each3

These could range from molecules to individual human beings to organizedcollectives.MIS Quarterly Vol. 44 No. 1/March 20203 Benbya et al./Introduction: Complexity & IS ResearchFigure 1. States of Complex Systems (Benbya and McKelvey 2011) other in a nonlinear way and are constantly adapting to inter-nal or external tensions (Holland 1995). Because suchsystems are constantly evolving, they have a large degree ofunpredictability. They cannot, therefore, be understood bysimply examining the properties of a system’s components.Four key characteristics influence the level of complexity ina system: (1) diversity, (2) adaptiveness, (3) connectedness,and (4) mutual dependency among agents in the system (e.g.,Cilliers 1998; Holland 1995). The nonlinear interplay of theabove four characteristics coupled with increased tension inthe form of external or internal challenges and/or oppor-tunities drive the system from one state to another.A system can exist or fluctuate between three states orregions: stable at one extreme, chaos at the other, with an in-between state called the edge of chaos (Kauffman 1995;Lewin 1992). Figure 1 provides an illustration of the threestates.Specifically, in the stable state, the diversity, adaptiveness,connectedness,and mutual dependency of agents in the systemare all at low levels. Consequently, adaptive tensions are low(Page 2010) and complexity is benign (Tanriverdi and Lim2017). The system rapidly settles into a predictable andrepetitive cycle of behavior. In such stable systems, noveltyis rare. There is a tendency for stable systems to ossify.As the diversity, adaptiveness, connectedness, and mutualdependency levels of systems reach moderate levels, the com-plexity level increases (Page 2010). Systems with increasedlevels of complexity enter the so-called “edge of chaos” stateor a region of emergent complexity (Boisot and McKelvey2010). By staying in this intermediate state, these systemsnever quite settle into a stable equilibrium but never quite fallapart. They exhibit continuous change, adaptation, coevolu-tion and emergence (Kauffman 1993; Lewin 1992).Increasing levels of tensions, beyond a certain threshold,might result in chaos or extreme outcomes (e.g., catastrophes,crises, etc.) which exhibit fractals, power laws, and scalabledynamics. Chaotic systems never really settle down into anyobservable patterns. Since they are sensitive to initial condi-tions, they can amplify exponentially and have monumentalconsequences (Gleick 1987).Complexity TheoriesAs outlined above, many living systems (e.g., organisms,neural networks, ecosystems) on the edge of chaos appear toconstantly adapt and self-organize to create configurationsthat ensure compatibility with an ever-changing environment. This perpetual fluidity is regarded as the norm in systems onthe edge of chaos; it can lead to processes and outcomes asdiverse as phase transitions, catastrophic failures, and unpre-dictable outcomes (see Table 1). Complexity theories such asemergence, coevolution, chaos, and extremes, as well asscalable dynamics, offer an explanation of such processes andoutcomes.4MIS Quarterly Vol. 44 No. 1/March 2020 Benbya et al./Introduction: Complexity & IS ResearchTable 1. Processes and Outcomes of Complex SystemsComplexity TheoriesProcessesOutcomesEmergence•Disequilibrium situations: tensions,triggers and small events outside the norm •Positive feedback and bursts ofamplification•Phase-transitions•Self-organization •Unpredictable outcomes: new structures,patterns, and properties within a system (e.g.,distributed leadership emergence), a new levelof analysis (e.g., a network), or a collectivephenomenon (e.g., collective action)•Emergence can take two forms: composition orcompilationCoevolution•Interdependency and boundary-crossinginterrelationships•Multilevel dynamics•Bidirectional or two-way causality•Mutual influences•Reciprocal adaptations and changes over timeChaos•Sensitivity to initial conditions•Constrained trajectory (e.g., strangeattractor)•Time-dependency and irreversibledynamics •Catastrophic failures (e.g., systemic risk, cyber-security breaches)•Escalation of causes leading to disastroussocietal consequences (e.g., disrupting lives ona large scale)Scalable Dynamics•Instability and large variations•Single cause leading to a cascade ofinterconnected events•Self-similarity across scales•Positive or negative extreme outcomes•Fractal dynamics •Power laws EmergenceEmergence is a dynamic process of interactions amongheterogeneous agents that unfolds and evolves over time,resulting in various kinds of unexpected novel individual- andgroup-level configurations and/or broader social structures(Benbya and McKelvey 2016). Complexity and organizationscholars have theorized such a dynamic process for some time(Kozlowski et al. 2013; Plowman et al. 2007).Systems-wide changes in natural open systems revealed howunorganized entities in a given system, subjected to an exter-nally imposed tension, can engage in far-from-equilibriumdynamics. The entities can therefore self-organize into dis-tinct phase transitions leading to a new higher-level order(Prigogine and Stengers 1984).Social systems put under tension, through recession, crisis,organizational change, and so forth, can exhibit similar phasetransitions and emergent outcomes. As such, many socialscientists have made a direct mathematical parallel betweenphysical and social systems to deduce the process mech-anisms inherent in micro interaction dynamics that yield thehigher-level order and its emergent novel outcomes. Theyhave identified two forms of emergence: composition orcompilation (Kozlowski and Klein 2000). In compositionmodels, emergent processes allow individuals’ perceptions,feelings, and behaviors to become similar to one another. Compilation models, on the other hand, capture divergence. They characterize processes in which lower-level phenomenaare combined in complex and nonlinear ways to reflect unit-level phenomena that are not reducible to their constituentparts. The discovery of emergence involves either a post hocanalysis of time series data (e.g., system behavior) andconceptual tools that allow scholars to verify the existence ofemergence dynamics in systems, or an analytical mapping ofthe sequential phases of emergence dynamics (e.g., Plowmanet al. 2007).Interactions among sociotechnical entities yield manyemergent outcomes in information systems. Examplesinclude the collaborative creation of online order and tech-nology affordances (e.g., Nan and Lu 2014), IS alignment(Benbya et al. 2019), and new configurations among organi-zation, platform, and participant dimensions (Benbya andLeidner 2018). An emergence perspective offers a lens tounderstand many unpredictable sociotechnical phenomenathat span individual, group, organizational, and societal levelsin the context of widening digitalization.CoevolutionCoevolution refers to the simultaneous evolution of entitiesand their environments, whether these entities beingorganisms or organizations (McKelvey 2004). Ehrlich andMIS Quarterly Vol. 44 No. 1/March 20205 Benbya et al./Introduction: Complexity & IS ResearchRaven (1964) introduced the term coevolution to characterizethe mutual genetic evolution of butterflies, and associatedplant species. Such a process encompasses the twin notionsof interdependency and mutual adaptation, with the idea thatspecies or organizations evolve in relation to their environ-ments, while at the same time these environments evolve inrelation to them. In addition, to the above characteristics, coevolutionaryprocesses have three main properties. First, coevolutionaryphenomena are multilevel. They encompass at least two dif-ferent levels of analysis. Second, coevolutionary phenomenatake time to manifest. This implies that longitudinal designsare necessary to understand coevolutionary processes. Third,bidirectional causality or two-way relationships (e.g., Yan etal. 2019) are central to coevolutionary processes.In IS research, coevolution theory has been used to theorizethe codesign of organizations and information systems(Nissen and Jin 2007; Vidgen and Wang 2009), the alignmentof business and IT (Benbya et al. 2019; Benbya andMcKelvey 2006b; Tanriverdi, Rai, and Venkatraman 2010;Vessey and Ward 2013), coevolution of business strategywith the competitive landscape (Lee et al. 2010); and coevo-lution of platform architecture, governance, and environ-mental dynamics (Tiwana et al. 2010).ChaosChaos theory was initially developed with Lorenz’s (1963)work in response to an anomaly in atmospheric science. Chaotic systems are sensitive to initial conditions. Thissensitivity to initial conditions, called the “butterfly effect,”implies that even a slight change, analogous to a butterfly’swing-beat, can lead to radical consequences on a much largerscale.In addition to being unstable and sensitive to initial condi-tions, chaotic systems are deterministic because the system’strajectory is constrained. Such chaotic systems possess astrange attractor, a value or a set of values that system vari-ables tend toward over time but never quite reach (Lorenz1963). Sudden discontinuous shifts in chaotic systems drivethem from one attractor to another, leading thereby to catas-trophes and disastrous societal consequences.Chaos theory has been used to theorize social and organi-zational dynamics as nonlinear chaotic systems by virtue oftheir sensitivity to initial conditions. For example, McBride(2005) used concepts of chaos theory to study the dynamicinteractions between information systems and their hostorganizations. Guo et al. (2009) use chaos theory to developa framework to illustrate blog system dynamics arising frommicro (individual blog traffic dynamics) and macro (blogo-sphere structure) levels. Hung and Tu (2014) provide anempirical analysis of the applicability of chaos theory toexplain technological change processes. Tanriverdi and Lim(2017) theorize about IS-enabled complexity vigilancecapabilities for detecting whether a complex ecosystemapproaches the edge of chaos/discontinuity.Scalable Dynamics, Fractals,and Power LawsScalable dynamics refer to self-similarity of underlyingpatterns across different levels of analysis (Manderbrot et al.1983). This notion of self-similarity across scales has becomea core tenet of complexity science and has led to varioustheories to characterize how a single cause can scale up intopositive or negative extreme events and drive similar out-comes at multiple levels (for reviews, Adriani and McKelvey2006; Benbya and McKelvey 2011).The dimensionality of such self-similarity across scales can bemeasured using a mathematical mapping technique referredto as fractals. In other terms, fractals measure the “density”of a nonlinear data set, such as stock market behaviors or theshape of a coastline (Casti, 1994). When such measures aretaken at increasing orders of magnitude, each fractal dimen-sion is “self-similar” to the ones before and after it, meaningthat the underlying patterns are the same across levels ofanalysis. These relationships are always governed by a powerlaw (Cramer 1993).Fractal analysis has helped describe and explain differentchanges that occur within similar patterns at multiple scalesacross organizations, markets, and industries. For example,Farjoun and Levin (2001) use a fractal analysis to characterizeindustry dynamism over time and capture the rate, amplitude,and unpredictability of change.Methods Research on complex sociotechnical systems has used avariety of methods, some are well established while others arejust emerging. IS scholars have studied dynamics of complexsystems by using established research methods such as longi-tudinal qualitative case studies (e.g., Benbya and Leidner2018; Paul and McDaniel 2016), morphogenetic approaches(e.g., Njihia and Merali, 2013), statistical methods for longi-tudinal data analyses (e.g., Nan and Lu, 2014; Tanriverdi andDu 2020; Tanriverdi, Roumani, and Nwankpa 2019). How-ever, complex sociotechnical systems that operate far from6MIS Quarterly Vol. 44 No. 1/March 2020 Benbya et al./Introduction: Complexity & IS Researchequilibrium conditions also present challenges for some estab-lished research methods such as closed-form analyticalmodeling methods. As such, newer methods have emerged tostudy under nonequilibrium conditions, complex interactionsamong multiple variables, and multilevel causality. Thosenew methods include agent-based simulation, the qualitativecomparative analysis (QCA) method, and dynamic networkmodeling based on graph theories.Agent-based simulation utilizes symbol-based computation toprecisely express a theory about a complexity concept such asagents, interactions, and the environment involved in an emer-gent process. The computational expression can then be usedto simulate and test the theory in controlled and replicableways. This methodological approach was advanced by theSanta Fe Institute (SFI), a multidisciplinary research centercreated in the mid-1980s (Waldrop 1992). Applications ofsimulation methods include genetic algorithms (Holland1995), cellular automata (Krugman 1996), NK landscapemodels (Kauffman 1993), and a combination of severalapproaches found in agent-based models (Carley and Svoboda1996).The QCA method allows researchers to identify how multiplecausal attributes combine into distinct configurations toproduce an outcome of interest, and to assess the relativeimportance of each configuration to the same outcome (Raginand Rubinson 2009). It relies on the set-theoretic approachand Boolean algebra to conceptualize and analyze causalcomplexity described as “equifinality, conjunctural causation,and causal asymmetry” (Ragin 2000, p. 103). Scholars fromdifferent social science disciplines including the IS field haveadvocated the use of QCA to embrace causal complexity thatis typical of social or sociotechnical systems (see El Sawy etal. 2010; Fichman 2004; Misangyi et al. 2017; Park et al.2020). Dynamic network modeling focuses on interactions that arethe root cause of complexity in a phenomenon. Agents andtheir interactions are modeled as nodes and edges in a net-work. Dynamic network modeling enables researchers toidentify patterns of interactions among a population of agentsin a system. Scholars have been using tools like spatio-temporal network modeling to understand how new edges areformed (George et al. 2007; Taylor et al. 2010). For example,complex patterns of evolution in a digital platform ecosystemcan be modeled as a network of third-party complements andboundary resources (Um et al. 2015). Here, third-party com-plementary products are modeled as agents interacting withone another through shared boundary resources. Scholarshave used a similar approach to explore the relationshipbetween consumers and brands (Zhang et al. 2016) and tounderstand the emergent nature of social relationships usingrelational event network (Schecter et al. 2017). Anotherimportant tool based on dynamic network modeling is net-work community (Sekara et al. 2016). A network communityis set of densely connected nodes (Newman and Girvan2004). For example, scholars have used network communityto discover dynamic emerging patterns of routines (Pentlandet al. 2020).Implications of Complexityfor IS Research The increased levels of complexity in sociotechnical systemsin the context of widening digitalization creates numerousopportunities and challenges for IS research. Due to thedistinct effects of digital technologies on complex socio-technical systems, simply replicating middle-level theoriesand models for complex physical, biological, or socialsystems would not fully capture IS-specific complexity issues. A fruitful approach for IS researchers is to use complexityscience as a meta-theoretical lens to rethink a few funda-mental research challenges (see Table 2). In this section, wediscuss a few of the challenges as exemplified by the fol-lowing questions:•Under what conditions is prediction feasible in complex,sociotechnical systems?•What is the nature of causality in complex, socio-technical systems?•How can researchers circumscribe the boundaries of acomplex, sociotechnical system to study? •How durable is newly discovered knowledge in complex,sociotechnical systems?Limits to Prediction in Complex Sociotechnical SystemsPrediction of potential outcomes in a given sociotechnicalsystem is one of the perennial questions in IS literature. It hasbecome even more important with recent developments in bigdata and artificial intelligence (AI) technologies. However,complexity of sociotechnical systems present major chal-lenges for prediction. Interactions among a diverse set ofconnected, mutually dependent, and adaptive agents in asociotechnical system lead to the emergence of unexpectedoutcomes that defy the extrapolation techniques at the heart ofprediction models. Properties of complex sociotechnicalsystems, such as nonlinearity, self-organization, coevolution,MIS Quarterly Vol. 44 No. 1/March 20207 Benbya et al./Introduction: Complexity & IS ResearchTable 2. Implications of Complexity for IS ResearchIssueImplication for IS ResearchPrediction of behaviorsof complex systemsThere are limits to the prediction of behaviors of complex sociotechnical systems. System-levelproperties such as non-decomposability, nonlinearity, self-organization, and coevolutioninevitably lead to emergent, unpredictable system behaviors.Prediction efforts of IS research should focus not on the ability to foresee specific, well-definedsystem events in space and time (i.e., paths), but on the ability to anticipate the range ofpossible behaviors the system might adopt (i.e., patterns).Nature of causality incomplex systemsA linear view of causality between inputs and outputs of the complex sociotechnical system isinadequate. There are multiple causal mechanisms and different forms of causality in complexsociotechnical systems.Three distinguishing features of causality in complex sociotechnical systems are: (1) conjunc-tion, which means that outcomes rarely have a single cause but rather result from the inter-dependence of multiple conditions; (2) equifinality, which entails more than one pathway from aninput to an outcome; and (3) asymmetry, which implies that attributes found to be causallyrelated in one context may be unrelated or even inversely related in another context.Boundaries of complexsystemsIt is challenging to accurately circumscribe the boundaries of a complex sociotechnical systembecause complex systems are open systems.IS researchers can potentially address this challenge by building on Salthe’s (1985) three-levelspecification in which agents of a complex sociotechnical system are defined by their com-ponents, a focal level of action, and by their contexts (Koestler 1978; Salthe 1989, 1985).Durability of newknowledge claims incomplex systemsPatterns of causal relationships in complex sociotechnical systems evolve over time. Thus, newknowledge discovered in one state of the system could be transient and inapplicable in anotherstate of the system. In making claims to new knowledge in studies of complex sociotechnical system, IS researchersshould report how frequently the system might be going through state changes and how durablethe newly discovered knowledge might be. In addition, if the study time frame involves anyphase transition of the complex system, researchers should report how the causal relationshipsmight differ qualitatively before, during, and after the phase transition.bifurcations, etc., lead inevitably to unpredictable states. Reductionist approaches that assume away some elements andinteractions in the complex system could make formal predic-tion models feasible to implement. Although some behaviorsof complex systems can be understood through formalmodels, those models cannot necessarily predict how a givensystem will evolve. Reductionist formal models also run therisk of generating biased, inaccurate predictions. This leadsto an important question: Under what conditions is it feasibleto make predictions in complex sociotechnical systems? Twoobservations can be made on this. First, predictions in complex sociotechnical systems requiresus to distinguish between patterns and path (Dooley and Vande Ven 2000). Path is the specific temporal trajectory, or setof points, that a system follows moment-by-moment; patternis the distinctive (often visual) temporal shape that emergeswhen one views the path over a long period of time, plottedin a particular manner. Linear systems are predictable in bothpath and patterns. Chaotic systems are predictable in patterns,but not path (Bohm 1957). Although accurate prediction ofa chaotic system’s path through a space of possible states isvery difficult because of sensitivity to initial conditions, clearoverall patterns are nevertheless observable because thesystem’s trajectory is constrained.Second, apart from deterministic chaotic systems that remainpredictable, what complexity science suggests is the inevit-ability of surprise (McDaniel 2004). Prediction becomes notthe ability to foresee specific, well-defined events in spaceand time (i.e., path) but, at best, the ability to anticipate therange of possible behaviors the system might adopt (i.e.,patterns). This then leads to the development of diverse con-figurations and states, or a portfolio of inter-related decision8MIS Quarterly Vol. 44 No. 1/March 2020 Benbya et al./Introduction: Complexity & IS Researchstrategies that can be employed, as future possibilities unfoldto become current realities. While predictability remainslimited given complexity, anticipation remains fluid with re-spect to changing conditions and tensions, thereby facilitatingadaptive action and survival (Boisot and McKelvey 2007).Nature of Causality in ComplexSociotechnical SystemsThe notion of causality is central to understanding the natureof interactions among heterogeneous elements in complexsociotechnical systems as the outcomes produced by suchinteractions (Sarker et al. 2019). Subsequently, the notion ofcausality is central to building theories for the IS discipline(Gregor and Hokorva 2011; Rivard 2014). Complexityscience offers significant ways to extend prior thinking on thenature of causality in complex sociotechnical systems. Conventional approaches dominant (explicitly or implicitly)in current models of thinking and analytical techniques, suchas generalized linear models, often treat causality as an unob-servable “black box,” and focus on discovering whether thereis a systematic relationship between inputs and outputs byrelying on additive, unifinal, and symmetric notions of caus-ality (Fiss 2007; Meyer et al. 2005; Mohr 1996). Such alinear view of causation remains limited and inadequate forexplaining increased nonlinear interactions in digitized envi-ronments with processes such as coevolution, emergence, andself-organization as they involve multiple causal mechanismsand different forms of causality.For instance, coevolutionary dynamics involve interactionsamong the components of a social system (such as a group, acommunity, or an organization), which in turn interacts withits environment. These multilevel interactions create orgenerate circular changes over time in one or several compo-nents of the system (Morgan 1923). In such contexts,unidirectional causation is not a good fit; rather, bidirectionalcausation, where the focus is on feedback dynamics in orderto promote the key links among the components, is necessary(e.g., Yan et al. 2019). Similarly, emergence, which spans multiple levels and leadsto novel emergents, can be also envisioned as a positive feed-back process starting with (1) bottom-up dynamic interactionamong lower level entities (i.e., individuals, teams, units)which—over time—yield phenomena that manifest at higher,collective levels, (upward causation) (Kozlowski et al. 2013),and (2) an emergent higher collective level that influences thecomponents’ behaviors on the lower level from which itsimultaneously emerges (downward causation) (Campbell1974; Kim 1992). Thus, complexity theory brings to lightnotions of multidirectional causality (e.g., upward, downward,and circular causality), uncertainty, and, hence, a sense of themultiplicity of possible outcomes.Building on this insight, IS scholars need to account for com-plex causality with its three distinguishing features: (1) con-junction, which means that outcomes rarely have a singlecause but rather result from the interdependence of multipleconditions; (2) equifinality, which entails more than onepathway to a given outcome; and (3) asymmetry, whichimplies that attributes found to be causally related in onecontext may be unrelated or even inversely related in another(Meyer et al. 1993).Boundaries of a Complex System andImplications for Multilevel IS ResearchBecause complex sociotechnical systems are open systems,we cannot accurately determine the boundaries of the system.In order to model a system precisely, we, therefore, have tomodel each and every interaction in the system, each andevery interaction with the environment—which is, of course,also complex—and each and every interaction in the historyof the system (Cilliers 2001). Since there are also relation-ships with the environment specifying clearly where a bound-ary could be, this is not obvious. Salthe’s (1985) three-levelspecification in which agents are defined by their compo-nents, a focal level of action, and their contexts (Koestler1978; Salthe 1989, 1993) helps to address this endeavor.According to this basic triadic specification, complex socio-technical systems can be best described at three adjacentlevels of interactions: (1) the level where we actually observeit, or where it can be meaningfully perceived (focal level);(2) its relations with the parts described at a lower level(usually, but not necessarily always, the next lower level); and(3) to take into account entities or processes at a higher level(also usually, but not always, the next higher level), in whichthe entities or processes observed at the focal level areembedded.Such a perspective, therefore, suggests that in order totheorize complex sociotechnical outcomes across levels it isnecessary to (1) articulate the emergent collective construct,a construct’s lower-level entities, and the constraints at thehigher level related to the role of a selective environment;(2) zoom in to consider both focal entities, interactions amonglower-level entities their internal structures and functions andto zoom out to consider both the focal entities and their exter-nal contexts; and (3) specify what kinds of top-down influ-ence or bottom-up process are potentially relevant and assessthe likelihood of interaction among the different kinds.MIS Quarterly Vol. 44 No. 1/March 20209 Benbya et al./Introduction: Complexity & IS ResearchDurability of New Knowledge Claims inComplex Sociotechnical SystemsAs a complex sociotechnical system dynamically coevolveswith changes in its environment and its constituent compo-nents, the pattern of causal relationships within the complexsystem can also evolve over time. This raises a questionabout the durability of new knowledge discovered aboutcausal relationships in a complex sociotechnical system. Complexity science conceives the development of newknowledge on the causal relationships as an evolutionaryprocess of proposing conjectures (blind variation) followed bythe refutation (selective elimination) of those conjectures thatare empirically falsified (Campbell 1974; Popper 1983). Thisimplies that over time, based on the cumulative research oncomplex systems, the collective view of the phenomenareflects the improved validity of the findings until provenotherwise, while the more invalid ones are abandoned. If thestates of a complex sociotechnical system change frequently,however, the newly discovered knowledge in one state of thesystem could be transient and not applicable to the next stateof the system. Thus, in making claims to new knowledge instudies of a complex, dynamically evolving sociotechnicalsystem, it is important for researchers to report how fre-quently the complex sociotechnical system might be goingthrough state changes and how durable the newly discoveredknowledge might be. If the findings about the complex socio-technical system are consistent over time, the new knowledgecould be valid and durable during that time period.In addition, if a complex sociotechnical system goes througha phase transition, the existing roles, structures, and causalrelationships in the system can dissipate and new ones canemerge, resulting in a qualitatively different set of causalrelationships (Tanriverdi and Lim 2017). Thus, it is alsoimportant for researchers to assess if the study time frameinvolves any phase transition of the complex sociotechnicalsystem, and how the causal relationships might differ quali-tatively before, during, and after the phase transition. Forexample, Tanriverdi and Lim (2017) posit that qualitativelydifferent types of IS capabilities are relevant to firm perfor-mance before, during, and after a phase transition of a com-plex sociotechnical system.Background and the Contentsof the Special IssueThis special issue is an outcome of on-going dialogues amongIS scholars who have been interested in complexity. Initially,to bring together interested scholars and foster further interest,Tanriverdi, Nan, and Benbya organized research symposiumson managing in complex adaptive business systems. Theinaugural symposium was held at the University of Texas atAustin (Austin, TX) in 2013. The second symposium washeld at the University of British Columbia (Vancouver,Canada) in 2014. The third symposium was held at Mont-pellier Business School (Montpellier, France) in 2015. Then,Benbya, McKelvey, Nan, Tanriverdi, and Yoo organized aProfessional Development Workshop entitled “Complexity inInformation Systems and Digital Business” at the Academyof Management Meetings in Vancouver, Canada, in August2015. On behalf of those who were involved in thesesymposia and workshops, McKelvey, Tanriverdi and Yooproposed a special issue on complexity and subsequentlyagreed to serve as senior editors for the current special issue. In December 2015, Tanriverdi and Yoo organized a Pre-ICISPaper Development Workshop for prospective authors whowere interested in submitting their research to the specialissue, where 26 extended abstracts were submitted anddiscussed with editorial board members.The special issue received a total of 50 submissions. Fortysubmissions were sent out to review after the initial screening.In the next round, 22 submissions were invited for revisionand resubmission. In the third round, 10 submission wereinvited for further revisions and resubmission. In the fourthround, seven submissions were invited for a final revision. Inthe final round, five articles were accepted for publication inthe special issue. To recognize their involvement from thebeginning of the entire process of the special issue develop-ment, Benbya and Nan joined Tanriverdi and Yoo in writingthis introductory essay for the special issue.The five articles in this special issue illustrate how IS scholarsbuild on theoretical lenses and methodological tools of com-plexity science to study digitally induced complexity insociotechnical systems. They demonstrate the promise for ISresearchers to not only draw on but also extend complexityscience in digital worlds. In discussing a special issue article,we first introduce the broad research theme motivated by acomplex phenomenon and then discuss how the special issuearticle addressed some aspects of the phenomenon.Research Theme 1: Designing IS to Unknownor Rapidly Changing Requirements Although much IS research accounts for the degree of com-plexity inherent in IS development (ISD), it rests on theassumption that the ISD process can be rationally planned andcontrolled. Such an assumption, however, is not suitable forexplaining rapid and unexpected changes characterizing theincreasingly interconnected IS collectives found in contem-porary organizations. Nor is it sufficient if we are to under-10MIS Quarterly Vol. 44 No. 1/March 2020 Benbya et al./Introduction: Complexity & IS Researchstand the generativity and emergent properties with whichdigital infrastructures and platforms are inextricably inter-twined. This raises an important question for ISD in complexenvironments: How do we design systems to be evolvable soas to match the transient pace of changing environments andorganizational goals? A number of IS scholars over the pastfew years have increasingly begun to consider the distinctiveinsights offered by complexity science to guide the develop-ment of evolvable and agile systems (e.g., Benbya andMcKelvey 2006a, 2006b; Montealegre et al. 2014; Tiwana etal. 2010; Vidgen and Wang 2009).In their article, “The Dynamics of Drift in Digitized Pro-cesses,” Brian Pentland, Peng Liu, Waldemar Kremser, andThorvald Hærem explore unexpectedly changing require-ments in digitized business processes supported by digitaltechnologies. They observe that incremental endogenouschanges in a digitized business process can suddenly push theprocess to a state of self-organized criticality. Through asimulation, they show that endogenous changes in the processcan lead to nonlinear bursts of complexity, causing a trans-formative phase change in the process. After the burst ofcomplexity, they further find that the dominant pattern ofdigitized business processes looks much different from theinitial condition. Their finding raises significant questionsabout the way we design digital technologies to supportdigitized business processes. Their simulation results showthat systems with adaptive programming are prone to trans-formative phase changes while systems with deterministicprogramming are not. Although it is infeasible to predictemergent process requirements, authors argue that digitaltechnology can be designed and used to influence the likeli-hood and severity of transformative phase changes caused byemerging requirements in digitized business processes.Research Theme 2: Using Digital Technologiesto Steer a Complex Sociotechnical SystemThrough Phase TransitionsDigitally induced, evolutionary transformative phase transi-tions also take place in products, business models, and neworganizational forms. Some firms embed digital technologiesin physical products to transition from physical products todigitized products (Tarafdar and Tanriverdi, 2018; Yoo et al.2010). Some firms transition from a portfolio of digitizedproducts to a product platform over which they develop anoffer of a family of derivative digital products (Gawer 2014).The pervasiveness of digital technologies in products alsoleads to the development of new business models arounddigital platforms and ecosystems (Yoo et al. 2012). Somefirms want to transition from competing on products to com-peting on digital platform ecosystems (Parker et al. 2016).They seek to transition toward multi-sided digital platformecosystems that can solve common problems of very largenumbers of consumers and third-party content developers.Such transitions also emerge in IT-enabled organizationalforms, such as communities and markets, which can dynam-ically transform over time. Benbya et al. (2015), for example,theorize about how a market can transition into a community,and vice versa. Rather than conceptualizing these forms asalternative, stable structures of knowledge sharing andinnovation, they call for investigating movements andtransitions between them. These transition endeavors raise animportant question for IS research: How can firms use digitaltechnologies to deliberately steer a complex sociotechnicalsystem from a given state through phase transitions toward adesired new state? In their article, “Digitization and Phase Transitions in Plat-form Organizing Logics: Evidence from the Process Automa-tion Industry,” Johan Sandberg, Jonny Holmström, and KalleLyytinen address this question by conducting a longitudinalcase study of digitally induced transformative phase transi-tions in ASEA Brown Boweri (ABB). Specifically, theystudy how ABB started with an analog automation productplatform, infused it with digital technologies deepening itsdigital capacities over a 40-year period, and tried to steer ittoward an ecosystem-centered organizing logic. Sandbergand his colleagues use the constrained generating procedures(CGPs) notion of the complex adaptive systems theory toanalyze three mechanisms of phase transitions: interactionrules, design control, and stimuli-response variety. Thefindings suggest that firms can leverage digital technologiesin trying to deliberately steer complex sociotechnical systemsthrough phase transitions toward desired new states. How-ever, the outcomes of digitally induced phase transitions arenot easily foreseen as they are often mediated by unintended,emergent changes in CGPs.Research Theme 3: Understanding How Com-plex Institutional Structures Shape the Evolu-tion of Enterprise Information SystemsA major desire of organizations is to foresee how their digitalor nondigital interventions would affect the evolution of theirenterprise information systems and performance outcomes. However, due to the complexity of their institutional environ-ments, such outcomes are often emergent and infeasible toforesee. Nevertheless, complexity scholars and practitionerssuggest modeling and simulating such complex environmentsto gain insights into possible evolution patterns of enterpriseIS and performance outcomes.MIS Quarterly Vol. 44 No. 1/March 202011 Benbya et al./Introduction: Complexity & IS ResearchIn “The Evolution of Information Systems Architecture: AnAgent-Based Simulation Model,” Kazem Haki, Jannis Beese,Stephan Aier, and Robert Winter build a theory-informedagent-based simulation model to generate insights about theevolution of enterprise IS architecture and efficiency andflexibility outcomes in complex institutional contexts. Speci-fically, they model the complex institutional environments oforganizations by modeling three institutional forces (i.e.,normative, coercive, and mimetic forces), and creating dif-ferent combinations of the institutional forces. They alsomodel how a heterogeneous set of agents (e.g., individualsand organizational units) would interact with the institutionalforces in trying to complete a dynamically changing set oftasks in the environment. The dynamic changes in the taskenvironment influence how agents would interact with eachother and with the institutional forces and whether they wouldadopt standardized IT solutions of the corporation or developcustomized IT solutions locally to address the tasks. Haki andhis colleagues conduct simulations to understand how theenterprise IS architecture of the organization would evolveunder different levels and combinations of the institutionalforces and what kinds of efficiency and flexibility outcomescould be expected. This study illustrates how IS researchersuse the theoretical and methodological tools of complexityscience to help managers anticipate how complex institutionalenvironments of their organizations could shape the evolutionof their enterprise IS architectures and organizations’ perfor-mance outcomes.Research Theme 4: Taming Complexitywith AlgorithmsComplexity science argues that problems emerging out of thecomplexity of sociotechnical systems are “wicked” problemsthat cannot be “solved” but that they could be “tamed”(Tanriverdi, Rai, and Venkatraman 2010). A wicked problemhas a large number of diverse stakeholders who have differentobjectives, values, and priorities. Those stakeholders areconnected and mutually dependent. A wicked problememerging out of digitally induced complexity in digitalplatform ecosystems is a search problem. The searchbehavior of one stakeholder affects search outcomes of theother stakeholders. The roots of the wicked search problemare tangled. The wicked problem morphs into another formwith every attempt to address it. The challenge has no prece-dent. There is nothing to indicate that there is a right answerto the wicked search problem. While it is infeasible to solvesuch wicked problems, IS researchers turn to big data,machine learning, and AI algorithms to “tame” them.Onkar Malgonde, He Zhang, Balaji Padmanabhan, and MoezLimayem, in their article “Taming Complexity in SearchMatching: Two-Sided Recommender Systems on DigitalPlatforms,” develop a complexity theoretic recommenderalgorithm to address the wicked search matching problemarising out of digitally induced complexity in Internet-basededucational platforms. They view an Internet-based educa-tional platform as a complex adaptive business system(CABS) where multiple sides of the platform have differentand evolving objectives, preferences, and constraints. Theyargue that search matching is a wicked problem in suchCABS and that it cannot be tamed by traditional one-sidedrecommender algorithms. They build on complex adaptivesystems theory to develop a two-sided recommender algo-rithm for taming the complexity of the search matchingproblem by allowing agents to co-evolve and learn in the sys-tem. Using an agent-based simulation model, they show thatthe proposed recommendation algorithms tame the wickedsearch matching problem although it cannot fully solve it.Research Theme 5: IT-Enabled CompetitiveAdvantage in Complex CompetitiveEnvironmentsIT and competitive advantage remains a key topic of interestto the IS discipline. As the complexity of competitive envi-ronment increased, however, firms started to find it in-creasingly challenging to achieve their quest for competitiveadvantage. Rivals and new start-ups use digital technologiesin innovative new ways to make frequent and bold competi-tive moves to erode the advantages of the incumbents. Asincumbents attempt to renew their advantages, the perfor-mance rank orderings of firms in the industry keep fluctuatingrapidly, a phenomenon known as hypercompetition (Nan andTanriverdi 2017). As such, the traditional quest of sustainedcompetitive advantage ought to shift to a quest to create andrenew temporary competitive advantages in complex, hyper-competitive environments (Tanriverdi, Rai, and Venkatraman2010). Some scholars argue that firms can cope with suchenvironments by developing IT-enabled dynamic and impro-visational capabilities and IT-enabled agility (El Sawy andPavlou 2008; Pavlou and El Sawy 2010). Other scholarsargue that adaptation to a rapidly changing environment maynot be sufficient for firm survival and performance. Coevolu-tionary views of IS strategy have been proposed to betteraccount for the mutual influences of a firm’s IT-based stra-tegic actions in complex, dynamically changing environments(Benbya and McKelvey 2006a, 2006b).Concomittant with these trends, economists warn that somefirms, especially the leaders of digital platform ecosystems,enjoy rising monopoly power and persistently high monopolyprofits. There have also been reports of declining dynamismand competition in the U.S. economy (Shambaugh et al.2018). Since the era of personal computers in the 1980s,there has been a marked increase in concentration rates and a12MIS Quarterly Vol. 44 No. 1/March 2020 Benbya et al./Introduction: Complexity & IS Researchdecrease in competition in the United States (Eggertsson et al.2018). After the take-off of the Internet in the mid-1990s,concentration rates have been rising faster in high IT-intensive industries than in low IT-intensive industries(McAfee and Brynjolfsson 2008). The highest concentrationrates are seen in industries characterized by digital tech-nologies that have large returns to scale and network effects(Shambaugh et al. 2018). The gap between winners andlosers has also widened dramatically. These seemingly con-flicting trends raise important questions for IS strategyresearch: Does digitalization amplify hypercompetition anderode the IT-enabled sources of sustained competitiveadvantages? Alternatively, does digitalization attenuatehyper-competition by concentrating the ownership and controlof digitalized resources and market power in a few digitalgiants, thereby reducing rivalry? If both scenarios arefeasible, how can firms strategize to create competitiveadvantages with IT?YoungKi Park and Sunil Mithas, in their article “OrganizedComplexity of Digital Business Strategy: A ConfigurationalPerspective,” examine how firms configure their six keydigital and nondigital capabilities to achieve high performancein complex digital environments. They use a configurationalperspective and a fuzzy-set qualitative comparative analysis(fsQCA) method to study how different configurations of thecomplex, nonlinear relationships among the six digital andnondigital capabilities affect firm performance in differenteconomic sectors with varying digitization and environmentalturbulence levels. A key finding is that digital capabilityalone is neither necessary nor sufficient for high performancein any configuration. However, digital capability is animportant element of the overall capability configuration. Depending on the economic sector, digital capability playsdifferent roles in the overall capability configuration; forexample, no role at all, a counterproductive role, or a highcontributor to performance.ConclusionsDigitally induced complexity is pervasive in sociotechnicalsystems. Complexity presents fundamental challenges to ISresearch such as the difficulty in circumscribing the bound-aries of a complex system, the multilevel nature of thecomplex phenomena, the difficulty of causal inference, thelimited durability of new knowledge claims, and the limits topredictability. Nevertheless, complexity science offers theo-retical and methodological tools to address these challengesand turn them into opportunities. The five articles in thisspecial issue illustrate how IS researchers use the theoreticaland methodological tools of complexity science to studywicked problems arising out of digitally induced complexityin the digital world. Each of these special issue articlesrecognizes that the new complex phenomenon it focuses onwould not have been feasible to study with conventionaltheories and methods. By building on theories and methodsfrom complexity science, these studies were able to study thecomplex new phenomena and generate new insights andexplanations. However, these articles are not mere appli-cations of known complexity concepts in the IS context. They also leverage the uniqueness of the IS context togenerate new insights to contribute back to complexityscience. Specifically, these IS studies inform complexityscience how the digitally enabled hyper-connections, hyper-speed, and hyper-turbulence in sociotechnical systems createpreviously unprecedented levels of complexity and dynamismand pose fundamental challenges to individuals, organiza-tions, and society. In the natural and biological worldsstudied by complexity science, major evolutionary and trans-formative changes take millions of years to unfold. Incomparison, major evolutionary and transformative changestrigged by digitally induced complexity take place in a matterof years, if not months, days, and even hours in modern daysociotechnical systems. The articles in the special issue lever-aged the unique properties of digital technologies, digitizedprocesses, products, platforms, ecosystems, and businessmodels to study how and why these transformations takeplace. They also combined their theories and methods withthose of complexity science to generate new explanations asto how wicked problems created by digitally induced com-plexity could be tamed. The approaches developed by theseIS studies could potentially inform complexity studies in otherdisciples.AcknowledgmentsWe would like to acknowledge the valuable contributions of BillMcKelvey to this special issue. Bill provided comments andsuggestions along the process and has been a source of inspirationand influence in writing this essay. We also gratefully acknowledgetwo editors-in-chief. Paulo Goes helped us to sharpen the purposeand the goal of the special issue and approved our proposal. ArunRai patiently guided our editorial process, providing invaluablecomments and suggestions throughout the process and for thisparticular essay. We would also like to thank all members of theeditorial board and reviewers. Finally, we would like to thank allauthors who submitted their precious work to our special issue. Thesymposiums and workshops of the special issue and the materials inthis introductory essay are, in part, supported by research grantsfrom McCombs School of Business, Sauder School of Business,Montpellier Business School, the National Science Foundation(Grant # 1447670), Social Sciences and Humanities ResearchCouncil of Canada, the Ministry of Education of the Republic ofKorea and the National Research Foundation of Korea (NRF-2018S1A3A2075114). Any opinions, findings, and conclusions orrecommendations expressed in this material are those of theauthor(s) and do not necessarily reflect the views of the fundingorganizations.MIS Quarterly Vol. 44 No. 1/March 202013 Benbya et al./Introduction: Complexity & IS ResearchReferencesAndriani, P., and McKelvey, B. 2006. “From Gaussian to ParetianThinking: Causes and Implications of Power Laws in Organi-zations,” Organization Science (20:6), pp. 1053-1071.Arthur, B. 1989. “Competing Technologies, Increasing Returns andLock-in by Historical Events,” Economic Journal (99), pp.106-131.Benbya, H., and Leidner, D. 2018. “How Allianz UK Used an IdeaManagement Platform to Harness Employee Innovation,” MISQuarterly Executive (17:2), pp. 141-157.Benbya, H., Leidner, D., and Preston, D. 2019. “Research Curationon Information Systems Alignment,” MIS Quarterly (https://www.misqresearchcurations.org/blog/2019/3/14/information-systems-alignment).Benbya, H., and McKelvey B. 2006a. “Toward a ComplexityTheory of Information Systems Development,” InformationTechnology and People Journal (19:1), pp. 12-34.Benbya, H., and McKelvey B. 2006b. “Using Coevolutionary andComplexity Theories to Improve IS Alignment: A MultilevelApproach,” Journal of Information Technology (21:4), pp.284-298.Benbya, H., and McKelvey, B. 2011. “Using Power-Law Scienceto Enhance Knowledge for Practical Relevance,” in Best PaperProceedings of the Academy of Management Conference, G.Atinc (ed.), San Antonio, TX, pp. 1-6.Benbya, H., and McKelvey, B. 2016. “Advancing Our Under-standing of Emergent Phenomena: A Multidisciplinary Reviewand Research Directions,” in Academy of Management AnnualMeetings Proceedings, J. Humphreys (ed.), Anaheim, CA, pp.1-30.Benbya, H., Menon, T., and Belbaly, N. 2019. Explaining Transi-tions between Communal and Market-Based Knowledge Sharing: Collective Action to Resolve Uncertainty,” in Proceedings of the21st

Americas Conference on Information Systems, Puerto Rico.Bohm, D. 1957. Causality and Chance in Modern Physics, Abingdon-on-Thames, UK: Routledge & Kegan Paul Ltd.Boisot, M., and McKelvey, B. 2007. “Integrating Modernist andPostmodernist Perspectives on Organizations: A ComplexityScience Bridge,” Academy of Management Review (35:3), pp.415-433.Boisot, M., and McKelvey, B. 2010. “Integrating Modernist andPostmodernist Perspectives on Organizations: a ComplexityScience Bridge,” The Academy of Management Review (35:3),pp. 415-433. Campbell, D. T. 1974. “Evolutionary Epistemology,” in ThePhilosophy of Karl Popper: The Library of Living PhilosophersP. A. Schilpp (ed.), La Salle, IL: Open Court PublishingCompany, pp. 413-463. [Reprinted in G. Radnitzky and W. W.Bartley, III (eds.), Evolutionary Epistemology, Rationality, andthe Sociology of Knowledge, La Salle, IL: Open CourtPublishing Company, 1987, pp. 47!89.]Carley, K., and Svoboda, M. 1996. “Modeling OrganizationalAdaptation as a Simulated Annealing Process,” SociologicalMethods and Research 25(1), pp. 138-168.Casti, J. 1994. Complexification: Explaining a Paradoxical WorldThrough the Science of Surprise, New York: HarperCollins.Cilliers, P. 1998. Complexity and Postmodernism. UnderstandingComplex Systems, London: Routledge.Cilliers, P. 2001. “Boundaries, Hierarchies and Networks in Com-plex Systems,” International Journal of Innovation Management(5:2), pp. 135-147.Cramer, F. 1993. Chaos and Order: The Complex Structure ofLiving Things (trans. D. L. Loewus), New York: VCH.Dooley, K., and Van de Ven, A. 1999. “Explaining ComplexOrganizational Dynamics,” Organization Science (10:3), pp.358-372.Eggertsson, G. B., Robbins, J. A., and Wold, E. G. 2018. “Kaldorand Piketty’s Facts: The Rise of Monopoly Power in the UnitedStates,” National Bureau of Economic Research Working PaperSeries (No. 24287).Ehrlich, P., and Raven, P. 1964. “Butterflies and Plants: A Studyin Coevolution,” Evolution (18:4), pp. 586-608El Sawy, O., and Pavlou, P. 2008. “IT-Enabled Business Capa-bilities for Turbulent Environments,” MIS Quarterly Executive(7:3), pp. 139-150.El Sawy, O. Malhotra, A. Park, Y., and Pavlou, P. 2010. “Seekingthe Configurations of Digital Ecodynamics: It Takes Three toTango,” Information Systems Research (21: 4), pp. 835-848.Faulkner, P., and Runde, J. 2019. “Theorizing the Digital Object,” MIS Quarterly (43:4), pp. 1279-1302.Farjoun, M., and Levin, M. 2001. “A Fractal Approach to IndustryDynamism,” Organization Studies (32:6), pp. 825-851.Fichman, R. 2004. “Going Beyond the Dominant Paradigm forInformation Technology Innovation Research: EmergingConcepts and Methods,” Journal of the Association for Infor-mation Systems (5:8), pp. 314-355.Fiss, P. C. 2007. “Towards a Set-Theoretic Approach for StudyingOrganizational Configurations,” Academy of ManagementReview (32), pp. 1180-1198.Gawer, A. 2014. “Bridging Differing Perspectives onTechnological Platforms: Toward an Integrative Framework.”Research Policy (43:7), pp. 1239-1249.Gell-Mann, M. 1995. “What Is Complexity?,” Complexity (1:1),pp. 16-19. George, B., Kim, S., and Shekhar, S. 2007. “Spatio-TemporalNetwork Databases and Routing Algorithms: A Summary ofResults International Symposium on Spatial and TemporalDatabases,” in Advances in Spatial and Temporal Databases , D.Papadias, D. Zhang, and G. Kollios (eds.), Berlin: Springer, pp.460-477.Gleick, J. 1987. Chaos: Making a New Science, New York: Penguin Books.Gregor, S., and Hovorka, D. 2011. “Causality: The Elephant in theRoom in the Information Systems Epistemology,” in Proceedingsof the 19th

European Conference on Information Systems, V. K.Tuuainen. M. Rossi, and J. Nandakumar (eds.), Helsinki.Guo, X., Vogel, D., Zhou, Z., Zhang, X., and Chen, H. 2009. “Chaos Theory as a Lens for Interpreting Blogging,” Journal ofManagement Information Systems (26:1), pp. 101-127.Haki, K., Beese, J., Aier, S., and Winter, R. 2020. “The Evolutionof Information Systems Architecture: An Agent-Based Simula-tion Model,” MIS Quarterly (44:1), pp. 155-184.14MIS Quarterly Vol. 44 No. 1/March 2020 Benbya et al./Introduction: Complexity & IS ResearchHolland, J. H. 1988. “The Global Economy as an AdaptiveSystem,” in The Economy as an Evolving Complex System, P. W. Anderson, K. J. Arrow, and D. Pines (eds.), Reading, MA: Addison-Wesley, 117-124.Holland, J. H. 1995. Hidden Order: How Adaptation BuildsComplexity, Reading, MA: Addison-Wesley.Hung, S., and Tu, M. 2014. “Is Small Actually Big: The Chaos ofTechnological Change,” Research Policy (43:7), pp.1227-1238.Kallinikos, J., Aaltonen, A., and Marton, A. 2013. “The Ambiva-lent Ontology of Digital Artifacts,” MIS Quarterly (37:2), pp.357-370.Kauffman, S. A. 1993. The Origins of Order: Self-Organizationand Selection in Evolution, New York: Oxford University Press.Kauffman, S. A. 1995. At Home in the Universe, Oxford, UK:Oxford University Press.Kim, J. 1992. “Downward Causation in Emergentism and Non-reductive Physicalism.” in Emergence or Reduction? Prospectsfor Nonreductive Physicalism, A. Beckermann, H. Flohr, and J.Kim (eds.), Berlin: De Gruyter, pp. 119-138.Koestler, A. 1978. Janus: A Summing Up, New York: RandomHouse.Kozlowski, S. W. J., Chao, G. T., Grand, J. A., Braun, M. T., andKuljanin, G. 2013. “Advancing Multilevel Research Design: Capturing the Dynamics of Emergence,” OrganizationalResearch Methods (16:4), pp. 581-615.Kozlowski, S. W. J., and Klein, K. J. 2000. “A MultilevelApproach to Theory and Research in Organizations: Contextual,Temporal, and Emergent Responses,” in Multilevel Theory,Research, and Methods in Organizations: Foundations, Exten-sions, and New Directions, K. J. Klein and S. W. J. Kozlowski(eds.), San Francisco: Jossey-Bass, pp. 3-90.Krugman, P. 1996. The Self-Organizing Economy, Hoboken, NJ: Blackwell Publishers.Lee, C. H., Venkatraman, N., Tanriverdi, H., and Iyer, B. 2010.“Complementarity-Based Hypercompetition in the SoftwareIndustry: Theory and Empirical Test, 1990-2002,” StrategicManagement Journal (31:13), pp. 1431-1456.Lewin, R. 1992. Complexity: Life at the Edge of Chaos, Chicago: University of Chicago Press.Lorenz, E. 1963. “Deterministic Non-Periodic Flow,” Journal ofAtmospheric Science (20), pp. 130-141.Lovelace, A. A. 1842. “Sketch of the Analytical Engine Inventedby Charles Babbage, by L. F. Menabrea, Officer of the MilitaryEngineers, with Notes upon the Memoir by the Translator,”Taylor’s Scientific Memoirs, 3, pp. 666-731 (retrieved fromhttp://www.fourmilab.ch/babbage/sketch.html).Lucas, P., Ballay, J., and McManus, M. 2012. Trillions: Thrivingin the Emerging Information Ecology, Hoboken, NJ: John Wiley& Sons.Lyytinen, K., and King, J. L. 2006. “Standard Making: A CriticalResearch Frontier for Information Systems Research,” MISQuarterly (30:Special Issue), pp. 405-411.Malgonde, O., Zhang, H., Padmanabhan, B., and Limayem, M. 2020. “Taming Complexity in Search Matching: Two-SidedRecommender Systems on Digital Platforms,” MIS Quarterly(44:1), pp. 49-84.Manderbrot, B., Benoit, B., and Wheeler, J. A. 1983. “The FractalGeometry of Nature,” American Journal of Physics (51:3), pp.286-287.McBride, N. K. 2005. “Chaos Theory as a Model for InterpretingInformation Systems in Organizations,” Information SystemsJournal (15:3), pp. 233-254.McAfee, A., and Brynjolfsson, E. 2008. “Investing in the IT ThatMakes a Competitive Difference,” Harvard Business Review(86:7-8), pp. 98-107.McDaniel, R. R. 2004. “Chaos and Complexity in a BioterrorismFuture,” in Advances in Health Care Management (Vol. 5), J. D.Blair, M. D. Fottler, and A. C. Zapanta (eds.), Stamford, CT: JAIPress, pp. 119-139.McKelvey, B. 2004. “Toward a Complexity Science of Entre-preneurship,” Journal of Business Venturing (19:3), pp. 313-341.Meyer, A. D., Gaba, V., and Colwell, K. A. 2005. “Organizing Farfrom Equilibrium: Non-linear Change in Organizational Fields,”Organization Science (16:5), pp. 456-473. Meyer, A. D., Tsui, A. S., and Hinings, C. R. 1993. “Configu-rational Approaches to Organizational Analysis,” Academy ofManagement Journal (36:6), pp. 1175-1195.Misangyi, V. F., Greckhamer, T., Furnaari, S., Fiss, P., Crilly, D.,and Aguilera, R. 2017. “Embracing Causal Complexity: TheEmergence of a Neo-Configurational Perspective,” Journal ofManagement (43:1), pp. 255-282.Mohr, L. B. 1996. The Causes of Human Behavior: Implicationsfor Theory and Method in the Social Sciences, Ann Arbor, MK: University of Michigan Press. Montealegre, R., Hovorka, D.,and Germonprez, M. 2014. “ACoevolutionary View of Information Services Development: Lessons from the U.S. National Oceanic and AtmosphericAdministration,” Journal of the Association for InformationSystems (15:9), pp. 577-613. Morgan, C. L. 1923. Emergent Evolution: The Gifford Lectures,Delivered in the University of St. Andrews in the Year 1922, NewYork: Henry Holt & Company.Nan, N., and Lu, Y. 2014. “Harnessing the Power of Self-Organization in an Online Community During OrganizationalCrisis,” MIS Quarterly (38:4), pp. 1135-1157.Nan, N., and Tanriverdi, H. 2017. “Unifying the Role of IT inHyperturbulence and Competitive Advantage Via a MultilevelPerspective of IS Strategy,” MIS Quarterly (41:3), pp. 935-958.Newman, M. E., and Girvan, M. 2004. “Finding and EvaluatingCommunity Structure in Networks,” Physical Review E (69:2),pp. 026113-1–026113-15.Nissen, M. E., and Jin, Y. 2007. “Coevolution and Co-design ofAgile Organizations and Information Systems through Agent-Based Modeling,” in Agile Information Systems: Conceptua-lization, Construction, and Management, K. C. DeSouza (ed.),Amsterdam: Elsevier, pp. 266-284.Njihia, J. M., and Merali, Y. 2013. “The Broader Context forICT4D Projects: A Morphogenetic Analysis,” MIS Quarterly(37:3), pp. 881-905.Page, S. 2010. Diversity and Complexity, Princeton, NJ: PrincetonUniversity Press.Park, Y., Fiss, P. C., and El Sawy, O. A. 2020. “Theorizing theMultiplicity of Digital Phenomena: The Ecology of Configu-MIS Quarterly Vol. 44 No. 1/March 202015 Benbya et al./Introduction: Complexity & IS Researchrations, Causal Recipes, and Guidelines for Applying QCA,” MISQuarterly , forthcoming.Park, Y., and Mithas, S. 2020. “Organized Complexity of DigitalBusiness Strategy: A Configurational Perspective,” MISQuarterly (44:1), pp. 85-127.Parker, G. P., Van Alstyne, M. W., Choudary, S. P. 2016. PlatformRevolution: How Networked Markets Are Transforming theEconomy—And How to Make Them Work for You, New York: Norton.Paul, D. L., and McDaniel, R. R. 2016. “Influences on Telecon-sultation Project Utilization Rates: The Role of DominantLogic,” BMC Medical Informatics and Decision Making (16),Article 155.Pavlou, P. A., and El Sawy, O. A. 2010. “The ‘Third Hand’: IT-Enabled Competitive Advantage in Turbulence throughImprovisational Capabilities,” Information Systems Research(21:3), pp. 443-471.Pentland, B. T., Liu, Peng, Kremser, W., and Hærem, T. 2020. “The Dynamics of Drift in Digitized Processes,” MIS Quarterly(44:1), pp. 19-47.Pentland, B., Recker, J., Ryan Wolf, J., Wyner, G. 2020. “BringingContext Inside Process Research with Digital Trace Data,”Journal of the Association for Information Systems, forthcoming.Plowman, D. A., Baker, L. T., Beck, T. E., Kulkarni, M., Solansky,S. T., and Travis, D. V. 2007. “Radical Change Accidentally: The Emergence and Amplification of Small Change,” Academyof Management Journal (50:3), pp. 515-43.Popper, K. R. 1983. “Realism and the Aim of Science” (Volume 1of Postscript to the Logic of Scientific Discovery), W. W. BartleyIII (ed.), London: Hutchinson and Company.Prigogine, I., and Stengers, I. 1984. Order Out of Chaos: Man’sNew Dialogue with Nature, New York: Bantam.Ragin, C. C. 2000. Fuzzy-Set Social Science, Chicago: Universityof Chicago Press.Ragin, C. C., and Rubinson C. 2009. “The Distinctiveness of Com-parative Research,” in The SAGE Handbook of ComparativePolitics, T. Landman and N. Robinson (eds.), London: SAGE.Rivard, S. 2014. “Editor’s Comments: The Ions of TheoryConstruction,” MIS Quarterly (38:2), pp. iii-xiii.Salthe, S. 1985. Evolving hierarchical Systems, New York: Columbia University Press.Salthe, S. 1989. “Self-Organization of/in Hierarchically StructuredSystems,” Systems Research and Behavioral Science (6), pp.199-208.Sandberg, J., Holmström, J., and Lyytinen, K. 2020. “Digitizationand Phase Transitions in Platform Organizing Logics: Evidencefrom the Process Automation Industry,” MIS Quarterly (44:1),pp. 129-153.Sarker, S. Chatterjee, S. Xiao, X., and Elbanna, A. 2019. “TheSociotechnical Axis of Cohesion for the IS Discipline: ItsHistorical Legacy and Its Continued Relevance,” MIS Quarterly(43:3), pp. 695-719. Schecter, A., Pilny, A., Leung, A., Poole, M. S., and Contractor, N. 2018. “Step by Step: Capturing the Dynamics of Work TeamProcess through Relational Event Sequences,” Journal ofOrganizational Behavior (39:9), pp. 1163-1181.Shannon, C. E. 1993. Collected Papers, N. J. A. Sloane and A. D.Wyner (eds.), New York: IEEE Press.Shambaugh, J., Nunn, R., Breitwieser, A., and Liu, P. 2018. “TheState of Competition and Dynamism: Facts About Concen-tration, Start-Ups, and Related Policies,” Brookings Institution,Washington, DC.Simon, H. 1962. “The Architecture of Complexity,” Proceedingsof the American Philosophical Society (106:6), pp. 467-482.Tanriverdi, H., and Du, K. 2020. “Corporate Strategy andInformation Technology Control Effectiveness,” MIS Quarterly,forthcoming.Tanriverdi, H., and Lim, Y. 2017. “How to Survive and Thrive inComplex, Hypercompetitive, and Disruptive Ecosystems? TheRoles of IS-Enabled Capabilities,” in Proceedings of the 38thInternational Conference on Information Systems, Seoul.Tanriverdi, H., Rai, A., and Venkatraman, N. 2010. “ResearchCommentary—Reframing the Dominant Quests of InformationSystems Strategy Research for Complex Adaptive BusinessSystems,” Information Systems Research (21:4), pp. 822-834.Tanriverdi, H., Roumani, Y., and Nwankpa, J. 2010. “StructuralComplexity and Data Breach Risk,” in Proceedings of the 40thInternational Conference on Information Systems, Munich,Germany.Taylor, G. W., Fergus, R., LeCun, Y., and Bregler, C. 2010. “Convolutional Learning of Spatio-temporal Features,” inComputer Vision—ECCV 2010, K. Daniilidis, P. Maragos,N. Paragios, (eds.), Berlin: Springer, pp. 140-153.Tarafdar, M., and Tanriverdi, H. 2018. “Impact of the InformationTechnology Unit on Information Technology-Embedded ProductInnovation,” Journal of the Association for Information Systems(19:8), pp. 716-751.Tilson, D., Lyytinen, K., and Sørensen, C. 2010. “Research Com-mentary—Digital Infrastructures: The Missing IS ResearchAgenda,” Information Systems Research (21:4), pp. 748-759.Tiwana, A., Konsynski, B., and Bush, A. A. 2010. “PlatformEvolution Coevolution of Platform Architecture, Governance,and Environmental Dynamics,” Information Systems Research(21:4), pp. 675-687.Turning, A. M. 1950. “Computing Machinery and Intelligence.”Minds and Machines (59:236), pp. 433-60.Um, S., Yoo, Y., and Wattal, S. 2015. “The Evolution of DigitalEcosystems: A Case of WordPress from 2004 to 2014,” inProceedings of the 36th

International Conference on InformationSystems, Fort Worth, TX.Vessey, I., and Ward, K. 2013. “The Dynamics of Sustainable ISAlignment: The Case for IS Adaptivity,” Journal of theAssociation for Information Systems (14:6), pp. 283-311.Vidgen, R., and Wang, X. 2009. “Coevolving Systems and theOrganization of Agile Software Development,” InformationSystems Research (20:3), pp.329-354.Waldrop, M. 1992. Complexity: The Emerging Science at the Edgeof Order and Chaos, New York: Simon & Shuster.Yan, K., Leidner, D., Benbya, H., and Zou, W. 2019, “TheInterplay between Social Capital and Knowledge Contribution inOnline User Communities,” Decision Support Systems (12:7), pp.113-131.16MIS Quarterly Vol. 44 No. 1/March 2020 Benbya et al./Introduction: Complexity & IS ResearchYoo, Y. 2010. “Computing in Everyday Life: A Call for Researchon Experiential Computing,” MIS Quarterly (34:2), pp. 213-231.Yoo, Y. 2012. “Digital Materiality and the Emergence of an Evolu-tionary Science of the Artificial,” in Materiality and Organizing: Social Interaction in a Technological World, P. M. Leonardi, B.A. Nardi, and J. Kallinikos (eds.), New York: Oxford UniversityPress, pp. 134-154.Yoo, Y., Boland, R. J., Lyytinen, K., and Majchrzak, A. 2012. “Organizing for Innovation in the Digitized World,” Organi-zation Science (23:5), pp. 1398-1408.Yoo, Y., Henfridsson, O., and Lyytinen, K. 2010. “The NewOrganizing Logic of Digital Innovation: An Agenda for Infor-mation Systems Research,” Information Systems Research (21:5),pp. 724-735.Zhang, K., Bhattacharyya, S., and Ram, S. 2016. “Large-ScaleNetwork Analysis for Online Social Brand Advertising,” MISQuarterly (40:4), pp. 849-868.Zittrain, J. 2006. “The Generative Internet,” Harvard Law Review(119), pp. 1974-2040.MIS Quarterly Vol. 44 No. 1/March 202017 Copyright

ofMIS Quarterly isthe property ofMIS Quarterly anditscontent maynotbe

copied

oremailed tomultiple sitesorposted toalistserv without thecopyright holder's

express

writtenpermission. However,usersmayprint, download, oremail articles for

individual

use.