There is increasing recognition that the link between individual action and organizational processes and outcomes needs to be better understood (Devinney, 2013; Greve, 2013; Winter, 2013). Complexity science provides such a framework and thus offers the prospect for new methods that show promise in this direction (Allen et al., 2013). This Special Issue is an outgrowth of lively discussions on this topic during formal Organization and Management Theory Caucus sessions held at each of the last two Academy of Management meetings, the first in Boston in 2012 with a follow-up in Orlando in 2013. The articles selected highlight an emerging complexity-informed paradigm for organizational research that we are calling human interaction dynamics (HID).

The HID approach uses a complex systems theoretical framework to study the nature of fine-grained human interactions and emergent coarse-grained properties as well as the resultant organizing forms that emerge from these dynamics (Hazy & Ashley, 2011). Treating agent interactions as the unit of analysis, HID explores the unique and heterogeneous detail within the micro-states that occur during interactions—including the rules that govern these interactions, how they are enacted, and how they change. This heterogeneity makes the human interaction very different from the types of interactions normally studied in the natural sciences.

Like the natural sciences, however, coarse-grained structures and properties emerge from these interactions. These emergent forms, like firms, business strategies, or organizational capabilities, also interact across levels of analysis to entrain the same fine-grained interactions from which they are also emerging. The dynamics that provide influence across levels in both directions at the same time are nonlinear and not easily parsed. However, by defining macro-states of the complex system as recognizable “properties” at the coarse-grained level, i.e., coarse-grained properties (CGPs), less important details can sometimes be ignored, or at least set aside for analysis purposes. Using a probabilistic approach that takes into account the system’s effective or statistical complexity (Gell-Mann, 2002; Prokopenko et al., 2009), coarse-grained properties and their probable outcomes, such as sales growth or the potential for an innovative product launch, can be modeled and forecasted. HID seeks to use complex systems models and techniques to shed additional light on the efficacy of this process without ignoring the underlying detail of individual choice and action.

Academic studies in human interaction dynamics

The first academic paper by Schwandt and Szabla (this issue) brings sociology and social psychology into the complexity conversation to describe the mechanisms at work within HID. Schwandt and Szabla argue that Giddens’s (1984) theory of structuration and Stones (2005) extrapolation of that theory together connect the mainstream social sciences of sociology and social psychology to complexity research and in particular to the study of human interaction dynamics (HID).

Exploring this assertion, Schwandt and Szabla argue that Giddens’s notion of duality of structure is congruent with the complexity idea of circular influence (Haken, 2006) between fine-grain interactions and coarse-grain properties (Gell-Mann, 2002) in complex adaptive systems (Holland, 1975). The authors go on to argue that the Giddens’ constructs of signification, domination and legitimation reflect the institutional social pressure that shapes the possibility space that constrains individual action. This paper is a good example of the issues associated with defining and representing the structural complexity associated with organizing in general. The nonlinear nature of structural complexity becomes relevant as fine-grain interactions among autonomous agents and the integrative coarse-grain properties that emerge from their interactions becomes the objects of study in HID.

The second academic paper by Edwards and Baker explores possible mechanisms that enable individuals to interact in an effort to form themselves into like-minded collectives to further a coarse-grain objective that provides an anticipated social benefit. The authors explore the formation of nascent community networks among motivated individual agents. These individuals choose to make use of their autonomy to further their own agenda, but at the same time they also purposefully choose to come under the sway of the emerging collective coarse-grain properties that they themselves have helped to enable (Goldstein, 2011). Through this process, autonomous individuals seek to integrate with the activities of others to further their own specific interests. The process Edwards and Baker study is thus effectively the first stage of the process that forms the mechanisms of downward influence on individuals to further collective objectives.

The authors identify two mechanisms that further this integration. First, by allowing for multiple interpretations of what is expected or hoped for at some future state, strategic ambiguity enables alliances to form even absent complete alignment of perspectives. Effectively, if the espoused strategy is sufficiently vague, it enables divergence so that each person can adapt it to their own interpretation thus avoiding conflict. One can reap the early benefits of organizing while retaining the option to choose differently later, perhaps by trying to further shape the strategy more towards one’s individual perspective at a point in the future. The group also benefits by including diverse points-of-view in an effort to discover a better way forward. Second, Edwards and Baker describe the gradual emergence (Goldstein, 2011) of interpretative dominance within the group with respect to the nature of the circumstances that the group is facing. As discourse increasingly converges toward a particular interpretation of the situation, more elements in the inference models used by individuals are shared across the collective and thus provide a common interpretation of events (Hazy & Backström, this issue). As a result, communication is improved and the activities of the collective become more integrated.

This paper is an example of the issues associated with the computational or algorithmic complexity of a system as it evolves over time (Richardson, 2010). In particular, it hints at the challenges associated with being inside such a system while attempting to gather the distributed information that is available locally and use it to predict the relationship between the current state of the ecosystem and its possible future states given the inescapable presence of entropy.

The third academic paper by Surie and Singh (this issue) explores how technology firms in developing economies expand into international knowledge markets. The authors describe their field research at fourteen biotech and software firms in India and what they learned about high velocity technology markets. (There are addition details with respect to the case studies in an Appendix.) In contrast to the other papers in this special issue, Surie and Singh treat firms as the agents comparable to the individuals in the other papers. In doing so, this paper is an example of the observation within the complexity literature that suggests that self-similar organizing mechanisms operate at higher levels of scale and at longer time horizons. The authors identify opportunity seeking and entrepreneurial aspirations as critical mechanisms for success in these markets. The former relates to the value discovering imperative within changing ecosystems while the latter relates to the value creating aspects of organizations.

In their analysis, Surie and Singh observe an evolution of organizational forms, including the emergence for a time of what they call the heteromorphic or H-Form, a period of quasi-stability as the organization makes use of multiple coexisting sub-structures such as contracts, alliances, and internal development to seek opportunities to exploit. When uncertainty subsides, these firms eventually position themselves within the global institutional hierarchy of their industry.

A typology for HID research

The perspectives and contributions of these three theoretical academic papers are summarized in Figure 1. Each of the three represents an exemplar study in a distinct level of HID research. At the most basic level, what we call the Identity-Level, analyses relate to the structure of a unity. This is exemplified by the Schwandt and Szabla (this issue) article. This level of analysis explores the challenges associated with clarifying and defining the nature of the system as the focal entity or “object”, the structure of its “unity” as a complex system and the characteristics of its “identity”. This challenge is analogous to the study in elementary particles in physics, or the internal workings of a living cell in biology. More discussion of the objects and relationships that make up each of the categories described here is available in the Technical Appendix to Hazy and Backström (Hazy, this issue—available as online supplementary information).

The unfolding of time adds an additional layer of complexity, what we call the Event-Level, to the study of organizations. This is because the presence of entropy ensures a level of uncertainty in predicting future states. This type of analysis is exemplified by Edwards and Baker (this issue). This level of analysis introduces the time dimension and explores the relationship between an entity’s current state and its trajectory toward its possible future states given the presence of uncertainty and entropy. This level of analysis explores HID with the realization that the emergence of coarse-grain properties necessarily require a longer time horizon than what is relevant in fine-grained interactions (Haken, 2006). It is therefore apparent that any event level analysis of the changing state of a complex system must always be considered in terms of both the dimensions of its current state, and the rate that each of these is changing as the system progresses toward some future state. This challenge is analogous to the

Fig. 1: A typology for HID research

The academic articles included in this issue represent exemplars of three categories of coarse-grained properties (CGPs) that characterize systems of human interaction dynamics. Each representational level reflects a duality (in italics) which must be navigated by practitioners, as described by Backström (this issue).

study of the unfolding of chemical reactions in microbiology as perceived from a position inside the system.

The third level of analysis exemplified in the Special Issue is represented by Surie and Singh (this issue). Analyses that we call Ecosystem-Level reflect the competitive dynamics within an ecosystem and the requisite strategic game playing needed for survival. Competitive conditions within an ecosystem can be explored using several analytical techniques such as dynamical systems or fitness landscapes (Kauffman, 1993). Agents in this space—in the Surie and Singh paper the authors consider the agents to be firms—change their internal configurations through mechanisms that are analogous to those in biological systems—like variation and recombination—to establish their fitness or performance at a high enough level to survive.

Ecosystem-level analyses study adaptation in changing ecosystems using techniques such as game theory, dynamical systems and the notion of attractors. This approach is relevant to business strategy and is analogous to the study of changing ecosystems and their relationship to epigenetics in biology, the area of research that informed the work of Stuart Kauffman (1993) and others. The challenges faced by firms that previously had been protected within local boundaries but are now entering the global ecosystem is analogous to that of individualized organisms within changing ecosystems as they struggle to establish, defend and stabilize their respective niches within a hierarchical structure.

A practitioner perspective

In the next article Backström (this issue) takes the point of view of the manager or leader to explore how the HID paradigm informs practice as individuals seek to influence organizing outcomes among their colleagues. The author argues that from the individual’s perspective, the nonlinearities inherent in complexity dynamics are reflected in perceived dualities as individuals seek to act.

Conceptual challenges related to understanding one’s identity in the context of others (in the framework shown in Figure 1, this would be identity level analysis) can be thought of as the duality between one’s felt sense of individual autonomy and what this implies versus the perceived expectations that arise from others in the institution and what these imply. Individuals are constantly challenged to act autonomously with self-interest but do so in the context of a socially constructed self, and at the same time act as an integrated member of the collective social-identity that is being continuously constructed during ongoing social interactions (see the classic paper by Buckley, this issue).

Backström goes on to show how the ways in which distributed information that is available through interactions with others is accessed by individuals (the event-level of analysis in Figure 1) can promote either convergent or divergent trajectories within HID. Confronting this duality within organizational life, and doing so effectively are also important considerations for leaders and managers. When discovering value is the objective, divergence and exploration of differences might be the prime objective. However, when one is creating value though organizing, a certain level of convergence might be necessary to fully realize the potential of the opportunity that has been discovered. (These concerns reflect the ecosystem level of analysis in Figure 1.) Backström offers some HID-informed practical guidelines for succeeding in today’s highly complex, increasingly fast-paced, and constantly evolving organizations.

Complexity and philosophy

The academic and practitioner papers are followed by the Complexity and Philosophy paper that offers the editors’ perspective with respect to the contribution of HID to the field of complexity. The paper situates the contributions in this issue into a larger philosophical framework. It argues for a holistic integration of the social science into the natural sciences though a greater reliance on many different fields of mathematics including the abstract field of category theory (Mac Lane, 1998). The potential contributions of mathematics extend well beyond the analytical tools commonly used to support statistical inference. Other areas of mathematics might also offer opportunities to build abstract models that can be used for deductive inference and therefore to predict outcomes in the manner of the natural sciences.

The paper describes the philosophical foundations of HID and explores the linkages from the sciences of complexity (Holland, 1975, Kauffman, 1993, Gell-Mann, 2002) to the social science (Anderson, 1999; Levinthal, 1997; McKelvey, 1997; Allen, Maguire, & McKelvey, 2011). It goes on to offer definitions and assumptions and to suggest future research directions. This paper includes a Technical Appendix (available in the online edition only) that is intended to begin the process of formalizing conceptual and analytical thinking about HID into a theoretical model.

A classic paper reflecting on an earlier classic paper

David Schwandt (this issue) provides an introduction to a paper by complexity pioneer Walter Buckley that reaches back to celebrate the early contributions of George Herbert Mead who brought insights from anthropology into the field of management. This classic paper about an earlier classic paper reminds us of the debt we owe to those who came before. The paper is a reminder that although we continue to learn more about human interactions, many of the most intractable issues that confound us have also confounded those who have gone before us. Each generation has the advantage of new science and new techniques that our predecessors could not have imagined. It is our responsibility to take what we are given by the previous generation and use it to move forward as far as possible, before we once again pass the baton to those who follow.

We remain hopeful about the promise of the approach described in the Special Issue. We believe that the papers we have included are a solid step forward in our collective effort to fully realize its potential. We hope that you agree.