Durham University Business School, ENG
Durham University Business School, ENG
The proposal that the metaphors associated with complexity theory can inform the business world is made by several writers (Wheatley, 1992; Stacey, 1996; McMaster, 1996, Merry, 1995), but is open to critique that the metaphors are not grounded in the field of study, but in other domains that may or may not be analogous. In previous articles, the authors (Fuller, 1998, 1999; Fuller and Moran, 1999) have illustrated the apparent analogies between complex adaptive systems and the world of small firms. However, because there is no grounding of these analogies in that domain, there is no evidence that complexity theory has validity in describing or explaining empirical observation. For example, a new firm starting up may be associated with the metaphor of emergence, but whether theories of emergence as developed in thermodynamic systems have any analogous properties with a business start-up is problematic. This article investigates how complexity theory can inform an understanding of small firms, which we posit as an example of socioeconomic systems, in a more rigorous and scientific way than metaphor. Our approach to this is to investigate the possibility of a methodology that is plausible in its relationship to small firms, and developed from the conceptions and literature of complexity. Methodology is about how we conceptualize, theorize, and abstract (e.g., Sayer, 1992): our modes of explanation, understanding, research design, and methods of analysis. In this article, a methodological position is developed, grounded in the literature of complexity theory and in substantive small business research literature. The methodology embodies philosophical principles, concepts, ontology, questions, methods, and ethics. The purpose of the article is to open discussion on these methodological aspects. The purpose of the methodology is for application to real-world problems and issues in small business.
Complexity is a science concerned with nonlinear dynamics and open, dissipative systems. Central to the enterprise would seem to be analogy, with dynamic modeling as a mode through which descriptions of dynamic behavior are made; for example dynamics as changes in patterns of relationships, in the emergence of events or conjunctions through time phases, and of the emergence of forms that display apparently different characteristics from their constituent parts. Approaches to modeling are varied, and include computer simulation (Hiebeler, 1994) and tracing of historic evolutionary paths (Gould, 1989). Models, too, provide a mode of explanation in terms of the results of unpredictable effects of multiple causal powers (e.g., codified as “rules” of behavior or inheritance). The effect on the researcher of such modeling is to create explanatory frames of reference that guide further abstraction and modeling.
In assimilating a systemic approach into a study of the social world, there is an explicit acceptance of what Cohen argues as the “insight that organisms are systems” (Cohen, 1998). For example, in a rubric to students, Axelrod (1998) suggests that a research goal is to “discover new principles about the dynamics of complex systems, especially complex adaptive systems which are typical of social processes.” Protagonists have assimilated the scientific metaphors. For example:
The evolution of dissipative social systems is chaotically driven and is sensitive to initial conditions. The structure is generated by symmetry breaking mechanisms and is consequently ontologically layered … These evolutionary properties establish the foundations for the historicity of the entities and the events under consideration. (Harvey and Reed, 1996: 306)
And, according to Byrne (1998), these systemic ideas transcend the limitations of the homeostatic systems model basic to Parsonian structuralfunctionalism. Complexity enables us to reflect the character of the social world as consisting of complex nested systems with a two-way system of determinant interrelationships among the levels. Also, it:
enables us to deal with both of the crucial problems identified for any sociological theory by Mouzelis (1995). It provides a way of relating the macro and the micro which is not inherently aggregative and reductionist and it provides a way of describing the relationship between agency and structure which takes account of Elias's assertion of the fifth dimension of reflexive human consciousness. (Byrne, 1998, Chapter 2)
Those searching for “science” in their research of society, including the domain of business, are attracted to complexity because of its scientific antecedents. Complexity studies provide the social scientist with many metaphors of dynamical systemic behavior. Are these metaphors analogous with social “systems”? Rosenhead (1998) and Fuller (1999) both critique the elevation of metaphors, grounded in nonanalogous phenomena, to the status of causal reasoning in social systems. The approach is open to a fallacy that metaphors are the same as reality.
The mistake here is directly to link metaphors of complexity with empirical experience. At issue is the extent to which patterns identified empirically, or modeled theoretically in the physical and natural sciences, provide ontological adequacy. Is it plausible to use metaphors of fitness, of attractors, of emergent properties, of rules and conditions, and to have adequate grounding of meaning in the business domain?
We suggest that these metaphors do not provide ontological adequacy per se, but have a role in informing the design of models or abstractions that may have such adequacy. From an evolutionary perspective, this kind of methodological positioning can legitimately be developed as potentially fallible, and from a scientific perspective it requires substantive reasoning or evidence for its claims. One issue arising from this is, therefore, how we test the adequacy of this work at a level of meaning. What is its instrumental reliability? We posit methods for this later in the article, attempting to find a starting point, with links to the empirical domain, for an investigation of the value of complexity science to the understanding of certain characteristics observed in small businesses.
A number of authors—e.g., McKelvey (1998); Reed and Harvey (1992)—have noted the proximity of complexity to the epistemology of scientific realism (Aronson, Harré, and Way, 1994; Suppe, 1989), and in social sciences to critical realism (Bhaskar, 1978; Outhwaite, 1987; Sayer, 1992). Realism provides philosophical principles on which dynamical nonlinear characteristics can be understood. For example, the appearance of novel structures and patterns can be explained by a conception of contingent or latent powers inherent in the interrelationships, rather than by the external imposition of order. One epistemological implication is that causality is not identified from the observation of empirical regularities per se. Causality in a specific context may be traced by theory building using concrete, intensive methods (Harré, 1979), but does not carry the same construct of being generalizable that the notion of causality carries in social positivism. Complexity is itself a scientific ontology
which fits Bhaskar's philosophical framework: one which treats nature and society as if they were ontologically open and historically constituted; hierarchically structured, yet interactively complex; non-reductive and indeterminate, yet amenable to rational explanation; capable of seeing nature as a “self-organising” enterprise without succumbing to anthropomorphism or mystifying animism. (Reed and Harvey, 1992: 359)
Small firms may lend themselves particularly well to a complexity-based research paradigm, possibly more so than large corporations, since the latter may be “overcomplex” (“complicated”?) in the sense of Kauffman's notion of “complexity catastrophe” (see Kauffman, 1993, 1995). This is because of the tendencies toward excessive (imposed) order, centralization, overengineering etc., which can result in a reduction of the overall fitness of the system and a thwarting of the selectionist process. As McKelvey (1999) puts it, “internal complexity leads to complexity catastrophe but external complexity leads away from catastrophe,” thus pointing up the importance of decentralized, disaggregated structures, following the logic of autonomous but co-evolving “patches” (Kauffman, 1995), which is resonant with our understanding of how small firms behave. Organizational theorists have not been able to mount a convincing case so far that modern corporate organizations can be adequately studied from within the paradigm of complexity, apart from in a purely metaphorical sense. As Rosenhead points out in a critique of “complexity” management texts:
It hardly needs saying that there is no formally validated evidence demonstrating that the complexity theory-based prescriptions for management style, structure and process do produce the results claimed for them. (1998: 10)
A small firm, by contrast, is relatively simple as an entity, although with possibilities of complex behavior arising because of the influence of the human agent (usually one person, i.e., the owner-manager/entrepreneur), and the high degree of interaction with other firms/agents that can lead to the evolution of new forms of structure. Such structures may be perceived, for example, as networks or clustering. The small firm can thus be viewed as a (relatively) simple system and as part of a more dynamic, complex whole, where multiple agents and interactions take place, giving rise to phenomena such as “swarming” and other emergent behavior.
Empirically, populations of small firms resemble the characteristics that Holland ascribes to a complex adaptive system, that is,
[an] evolving perpetually novel world where there are many niches with no universal optimum of competitor, where innovation is a regular feature and equilibrium rare and temporary and where anticipations change the course of the system, even when they are not realised. (1995)
Evolutionary and ecological metaphors of emergence, fitness, and mimicry resonate with observations of the large number of smaller firms in the economy. Small businesses are not a homogeneous population. They vary considerably in size and sector activity, in their ownership, their location, the markets served, and so on. Each business is different. Each has its own “initial conditions,” and each incurs a number of “accidents” in its temporal path. Given that entrepreneurs are “innovative,” many businesses will operate with their own “rules,” as well as complying (more or less) to more general rules. Business strategies explicitly operationalize the metaphor of “niche specialization.”
Some of the features of businesses' domain are common or shared. They all interact with key economic stakeholders, such as banks and government agencies. Businesses operate in a regulated environment, providing at least some of the “rules” of behavior. The mimicry of doing business, i.e., copycat methods and the diffusion of information through benchmarking and best-practice guides, is ubiquitous. Swarming is commonplace, for example physically in business districts and clusters (e.g., Gillies et al., 1998), or in the use of particular technologies (e.g., North et al., 1991). And energy, in the form of cash and perhaps technological innovation, flows within the system, with those firms that do not maintain cashflow or adopt new ideas ceasing to operate.
In its assimilation into the small business domain, complexity theory may become trapped in its own metaphors, but there are at least four areas in which it can move beyond the metaphor as a surface description of observed behavior. These areas are interlinked, but conceptually different.
Take first the notion of the small firm, or some attribute of the small firm, as an adaptive agent; see, for example, Rydal, 1996; Casti, 1997. The notion of an adaptive agent is highly resonant with Schumpeterian notions of entrepreneurial innovation. Indeed, Schumpeter's work stimulated Nelson and Winter's (1982) contribution to evolutionary economics. The “adaptive” (entrepreneurial) actions—“the capacity of seeing things in a way which afterwards proves to be true, even though it cannot be established at the moment” (Schumpeter, 1934: 85)—appear reflexive, taking into account the existing perspectives and external stimulus (Lewis and Fuller, 1998). This reflexivity is perhaps more likely to be understood through the investigation of learning and social processes, rather than a two-dimensional, systemic concept of adaptation. The articulation of rule-like, reflexive behavior or the nature of the learning that gives rise to changes in reflexive responses has not yet been adequately codified. Adaptation is conceptualized herein as a reflexive process, one in which the adapter exercises agency.
Second, the notion of the firm as being part of a wider system, “ecology,” or nexus of stakeholder relationships and actions (Fuller, 1997) is significant in theorizing the small firm. Small firms are not individual entities per se, but part of interrelated structures of relationships. The nature of these relationships is not well articulated in the literature. For example its representation in agency theory (Williamson, 1991) as a nexus of contracts does not adequately take account of qualitative or noneconomic factors. Small firms are theorized as operating in “networks” by a number of authors (e.g., Johannisson, 1987; Jarillo, 1988; Lorenzoni and Ornati, 1988; Larson, 1992; Castells, 1996). These studies stress the importance of both social and economic rationales for the relationships. However, the nature of the relations and “coupling” between small firms and their environment is not well enough understood to have yet produced plausible complex adaptive models. In the complexity literature, relationships between the individual agent and others are often definitionally implicit, yet crucial. For example, in the “Ant” rules—Coveney and Highfield, 1995: 250, if you find food, take it home and mark a trail; if you cross a trail and have no food, follow the trail to the food etc.—the crucial relationship between the ant colony, the behavior of individual ants and food is axiomatic, and the necessary relationship between ants and food for survival is implicit. Relationships are conceptualized herein as interdependent powers between firms, individuals, other agencies, and other objects or mechanisms.
Third, the notion of fitness, and the maintenance of fitness, are synonymous with “competitiveness,” but also with growth or survival. Life is short for most small firms and the rate of new firm formation alters in different conditions. Maintaining fitness in complex adaptive systems is said to be informed by what Holland calls “look ahead.” Lane and Maxfield (1995) address this with regard to strategy in organizations, arguing that only those “inside” the system can have any sense of prediction of strategies. The concept of fitness and emergence in alternative conditions is also to be found in the work of Fuller et al. (Fuller, 1999) on foresighting. Their approach uses the idea of structural coupling to simulate the emergence of typical new firms and innovation from scenarios of alternative (future) initial conditions.
In small business research, links between conditions and systemic fitness are largely empirical and judgmental, with little theoretical explanation. This leads to a critique of empirical discovery of regularities associated with “success” at any point in time. Most positivist research in the small business field makes claims with regard to the association of hypothesized factors and some form of success. There is no evidence that this has any predictive capability, nor any explanatory value. There have been some classic errors, such as Peters and Waterman (1988). From a complexity perspective, the reason that such empirical evidence is unreliable as a guide to behavior is that the systemic interdependencies or reflexive linkages between the firm and the environment are not adequately understood from an external perspective. More fundamentally, in open systems fitness is a highly dynamic and unpredictable state.
Fitness is conceptualized herein as a state of relative performance, which may be the result of reflexive adaptation. It may be articulated or described partly in terms of relationships, but is inherent to a firm within its context, i.e., it is relative.
Fourth, the causal concept of structural emergence through selforganization or autopoiesis provides a powerful methodological construct for the investigation of change in the small firm domain. The production of results from the Prigogine and Lefever experiments (Prigogine and Stengers, 1984) showed that nonlinearity occurs in a chemical reaction if a product catalyzes its own production, a feedback process known as autocatalysis. Prigogine introduced the term “dissipative structures” (the dissipation of introduced energy) to emphasize the origins of self-organization in far-from-equilibrium thermodynamic processes.
This idea of a system retaining energy through the formation of additional structure resonates with Anderson's ideas of “symmetry breaking” (Anderson, 1972). This implies that dynamical systems do not become ever more complex, in a “flat” sense of more features, although they do create new structures, new ontological levels. If the systems were entropic, then they would become more chaotic. Dissipative structures do not necessarily become more chaotic, but dissipate entropy to outside the system. According to Harvey and Reed (1996:306), sustainable dissipative systems:
Luhmann's work (e.g., 1986) is seminal in linking autopoiesis to social systems. Open systems are dynamic: energy flows within them and in and out. The precise circumstances that give rise to an ordering property are unique, unlikely to exist more than once. The existence of novel form creates novel conditions and vice versa. The authors' guide to theorizing, abstracting, or conceptualizing is a sense of what Allen (1997) calls an “evolutionary tree of successive structures.” In this context, the arrow of time is one way, not reversible—events cannot be undone, nor ever repeated exactly.
Such a central concept as autopoiesis we believe is significant in developing a methodology for researching small firms in a complexity paradigm. This is developed in the next section of the article through the idea of ontological layers. An example of linking the analytical ontological perspective of interrelationships with model-centered theory is in the work of Gillies et al. (1998). However, autopoiesis may also inform an understanding of other creative processes, for example innovation and generative relations (Lane, 1996).
Emergence is conceptualized herein as the concrete result of a reflexive or self-organized, creative or generative process, whose form may be empirically observed, or whose presence empirically sensed.
These four main concepts—reflexivity and learning, relationship with the environment and other agents, fitness and innovation, and autopoietic structural emergence—may perhaps be understood as interlinked facets of a process of contingent adaptation and survival in a population of small firms. The concepts inform a methodology with surface validity for investigating the dynamics of small firms. The claim for validity is that the dynamical characteristics that the concepts label in experimental fields of complexity have analogical or metaphorical resonance with observations in the small firm domain.
The central property of dynamical systems of symmetry breaking and the creation of novel ontological layers provides a theoretical dimension to investigate multiple layers of firm characteristics and dynamics. The firm may need to be understood to exist simultaneously on many layers, possibly unconnected, and each having different meaning and different characteristics. This is partly why it is so difficult to operationalize interdisciplinary research work: each discipline is concerned with different, epistemologically or ethically separated, ontology, not just different perspectives on the same phenomenon.
A challenge for small firm research is to define the relevant “ontological layers” of the small firm “domain” and how these may interrelate and possibly give rise to emergent behavior and structures. Clearly, some ontological layers are outwith the scope of small firm research, but are nevertheless important as influences on “micro states.” As McKelvey (1999) points out, modeling of complex adaptive systems is focused on how micro-state events (including human agents or firms) “self-organize into emergent aggregate structure.” The division of structures is important here as a means of maintaining emergent structure far from equilibrium (i.e., “negentropy”) and therefore a networked form of structure is potentially more stable and adaptive over time than one based on merging structures (i.e., a large corporation) (see for example Kelly, 1995; Castells, 1996). The latter requires large amounts of energy to sustain it and will be incapable of rapid change; whereas the former is dynamic, adaptive, and, because of the very nature of its structure, does not require large overall amounts of energy to sustain it (the energy inputs are in effect “localized” due to the independent actions of adaptive agents).
Figure 1 illustrates six theorized ontological layers, derived from the canon of research literature within the small firm domain, and the “boundaries” at each end. For small firms, the relevant layers are posited to range from “micro economies” to individual mental models and cognition (e.g., of the entrepreneur). The layers are intended to reflect key areas of research and debate in the small business field, i.e., networks/clusters (e.g., Chaston, 1996; Curran et al., 1992; Hansen, 1995; Johannisson, 1987, 1995); external relationships in the value chain (e.g., Hall and Andriani, 1998; Lewis and Fuller, 1998; Mitchell and Agle, 1995); business model/strategy/vision etc. (e.g., Atherton and Hannon, 1997; Gibb and Scott, 1986; Miller and Toulouse, 1986); internal resources/processes (e.g., Garnsey, 1998; Hendry et al., 1995); capabilities and motivations (e.g., Bellu and Sherman, 1995; Carsrud et al., 1989; Harrison and Leitch, 1994; Miner, 1997); individual cognitions etc. (e.g., Chell et al., 1991; Gatewood et al., 1995; McGaffey and Christy, 1975; Moran, 1998). Beyond the “top” boundary is where aggregations become superordinate structures such as the macro or global economy. Below the “bottom” boundary is where physiology, biochemistry, and so on down to the quantum level influence individual cognitions, mental models etc.
Figure 1. Posited ontological layers in the small firm domain
These are also ontologies but of different “domains,” albeit impinging on the small firm “domain,” which is about how ways of seeing, thinking, and so forth are manifested through successive ontological layers to result in micro-economies of small firms existing in interrelationship with each other. In effect, the diagram reflects how small firm “ecologies” are built up from particular “micro states,” including individual personal characteristics and attributes of human psychology, through successive emergent realities.
The question arising from the above is to what extent these ontologies (or “perspectives”) reflect real-world mechanisms with causal properties, and how they might be operationalized in real experiments or studies. There is an issue here about the “permeability” of the “layers” in terms of the tendency among researchers to stay within tightly prescribed disciplinary boundaries. This is particularly important in this context in exploring the interactions between layers or how emergent properties arise from the lower-level micro-state interactions. Focusing solely within one layer may result in a limited view of the overall phenomenon and of how the “reality” of one layer is due to behavior or events at the layer below reaching some critical threshold (or “phase transition”) sufficient to create new, emergent structure or form. Thus, while the “business model” within the small firm domain (layer 4) may be legitimately studied in its own right, only a partial understanding (in the widest sense) will be achieved if the forces and influences that give rise to it at lower ontological layers are ignored or “assumed away” as not being germane. However, from an existential perspective, it must be remembered that a phenomenological entity termed “a business” can only exist because of a particular nexus of human activities and relationships, influenced themselves by particular competencies, drives, cognitions, and sense-making mechanisms. This reinforces the importance of the “bottom-up” nature of complexity science (Epstein and Axtell, 1996).
An example of research reaching down through several layers is currently in progress by one of the authors (see Moran, 1998). This research originated in the personality profiling of owner-managers (level 6) and how these relate to “growth orientation” (level 5). This is now being developed through in-depth interviews to explore issues such as the future shape and direction of the business (level 3), and key external relationships and their impact on the business (level 2). For completeness, the internal processes and relationships should also be explored (level 4). Being able to make connections between findings from different “layers” for the same cohort of firms may enable the construction of systemic models reflecting the complex, dynamical nature of small firms arising from particular micro-state realities, which can be tested within the “model-based science” paradigm using simulations (see Casti, 1997).
From the above analysis—i.e., four significant complexity concepts and six small firm ontological levels that can be posited as having a hierarchical or nested relationship—a potential field of study emerges. Drawing on the previously discussed concepts of autopoesis and symmetry breaking, conceptually we would expect that the dynamics associated with complex adaptive systems would be related to the linking of hierarchical (emergent) ontological structures. Thus we can generate a plausible field of study by the simple cross-tabulation of these two sets of characteristics, shown in Figure 2. The range of research questions generated in this conceptual space requires further work. Some examples of substantive issues, still largely understood only in atheoretical (empirical or heuristic) terms in the domain of small firms, are given below (see Table 1).
Figure 2. Ontological levels tabulated with complexity dynamical concepts (numbers refer to Table 1
|1||In what sense do small firms co-evolve with one another/other stakeholders?|
|2||What is the result of this co-evolution?|
|3||To what extent do small firms aggregate and create self-sustaining systems (e.g., "clusters")? What evolutionary characteristics emerge within these higher-order systems?|
|4||Why do firms network? Why do these relationships continue or discontinue?|
|5||What are relevant boundaries to the firm?|
|6||What is the role of the owner-manager in the process of adaptation in a small firm?|
|7||How are firms deemed to be fitter or less fit over time?|
|8||Does a firm's fitness co-evolve with stakeholders?|
|9||What sense-making and schema-building strategies do owner-managers use to improve the positioning of the business and thereby increase the chances of survival?|
|10||What new concepts do owner-managers develop from their experiences?|
Dynamical processes can only be understood through a time dimension. This might be historical or “real time.” We suggest that there is little or no historic evidence available that has been gathered through the conceptual framework developed in Figure 2. This requires further investigation, but it is likely that a longitudinal study is required if an empirical grounding is sought.
The authors propose an iterative modeling/grounding approach to operationalize this research. They take the view that knowledge of the workings of any social system (of which the small firm is posited as being part) requires deep insight that is normally only available to its experienced actors. The common sense that such insights might generate may be shown ultimately to be “wrong,” but insights are, we suggest, closest to making sense of experienced dynamical processes at the relevant ontological layer. In such a case, the methodology demands the participation of system actors.
We further propose that in order to operationalize a methodology that takes account of dynamical properties, some form of simulation model is required. The construction of this model should be informed by grounded theory or propositions of salient features identified initially by inspection of the literature and by intensive (Harré, 1979) reasoning from empirical evidence. This in itself may require considerable fieldwork, or can draw on existing research.
Simulations may help to clarify interactions and emergent patterns that might be fed back to assist in strategic decision making and executive action. In other words, the research enterprise would not merely be a way of creating new knowledge and models, but of adding practical value to the small firm community (i.e., create a “fitter” ecology from an evolutionary perspective). This requires that results of simulations are validated through field testing over time. The schema for this is illustrated in Figure 3, a Mandala or loop of modeling and testing, implying a learning or theory-building process.
Figure 3. Generic research cycle
We therefore propose a method that iterates between everyday practice and analogous modeling. The method is guided by the concept shown in Figure 4, which places interpretation centrally, communicated through language and shared theory in practice between researchers and actors in the domain. Modeling provides an experimental form for scientific analysis (McKelvey, 1998); practice provides a grounding and testing of the emergent or evolving theories. In a sense this is a closely coupled microcosm of social theoretical evolution.
It is important to note from Figure 3 that the intermediate step of model building is required to “convert” observations and data into something that can be simulated in order to facilitate more in-depth understanding of the phenomenon. The simulation is thus only as good as the dynamical model from which it is derived. The loop is then closed by the testing of the outcomes of the simulation in relation to the real-world agents from which further observations/data would continue to be
Figure 4. Model of evolutionary theory building through modeling, insight, and practice
collected, and the dynamical model refined accordingly for further iterations of the simulation. Any such model would abstract features from the experience of the observers, but its linkage in this process could ground the features in the day-to-day experiences of actors, providing a sense of salience.
We propose to use this simulation with actors (i.e., owners of small firms) as a way of helping them understand and articulate their worlds (i.e., the systems of which they are a part). We conceptualize that such an action will lead to a reconceptualizing of the individual or shared “theory” of the system, which in turn may lead to new strategies or behavior. The degree of utility and resonance that the models have for these actors will act as a test of instrumental reliability. The models themselves can be independently tested for robustness, for example by the use of counterfactual tests (see example below).
This approach has ethical and practical issues associated with it. A particular perspective here emphasizes the utility of research within the human systems domain, particularly focusing on the researcher/client relationship. In small firm research there is the possibility of intervention to the benefit or detriment of the firm, particularly if the researcher appears to be a“credible” source. Being wholly detached/objective is difficult if the work involves working inside the small firm with the ownermanager and/or other members of the company. There is certainly scope for researcher and owner-manager (practitioner) to influence and learn from each other through a positive feedback cycle. This is resonant with Schon's (1991) notion of “reflective research,” where the researcher uses both observation and intervention to help the practitioner develop insight and capability (“reflection-in-action”). Of course, it is also important that the researcher recognizes when not to intervene and understands the importance of using experimentally valid methods within the research inquiry.
The challenge therefore is to develop a research methodology in the small firm domain that seeks to build productive relationships with owner-managers as clients/practitioners in order to acquire a deep(er) understanding of systemic processes, relationships, and dynamics in small firms. This understanding can then inform the building of improved models, which can lead via testing to better interventions and improved capability in the small firm domain and thus, potentially, “better” (i.e., fitter) small firms.
The research paradigm suggested here might be interpreted either as a whole methodology, or as a method. From a methodological perspective, the model could help to position and make coherent discrete research activities. As a method, interactive modeling between researcher and small business owner (as decision maker) is relatively novel.
Three examples of the authors' current research are given below, showing how the metaphors of complexity contribute to the post hoc interpretation of present findings (rather than the framing of the original research).
The longitudinal research with owner-managers described above (Moran, 1998) is intended to explore research questions as detailed in Figure 2 (particularly 6, 7, 9, 10) concerned with the interaction of individual agent (owner-manager) and the business “system.” The research conducted to date has focused on developing insight into the “psychology” of a cohort of owner-managers and linking this to an independent (quasi-performance) measure of “growth orientation” (GO). Thus, relationships can be explored between personal characteristics and orientation toward the business (ontological levels 5 and 6) in such a way that “rules” for adaptation and learning linked to the fitness of the business entity (system) may be derived and tested further. The research is currently entering a “grounding” phase in which actual performance and development of the businesses can be related to the assessments of the individual ownermanagers from the initial study. This will help to ascertain the degree of predictive validity of the previous measures and deepen our understanding of the processes of change and the influence of the individual agent on them. This move takes the research to ontological level 3, with a continuing linkage through to levels 5 and 6.
A study of small firm stakeholder relationships (Lewis and Fuller, 1998) grounded a typology of relationships, through a qualitative analysis of indepth interviews with about 40 small firms. Some five separate approaches to relationships were identified, which can be used to categorize individual firms in the sample. This work provides insights into the nature of the firms' responses to changes in the stakeholder environment, in particular to new uses of information and communications technology. As such, it helps to identify reflexivity, which can be conceived as agency (causing change) in a dynamical system. Further ethnographic studies were also carried out to discover whether an owner-manager's perspective or relationship style was carried through in the whole business. Conceptually, this links level 2 with level 6 in Figure 2, which may itself present a healthy critique for the ontology per se.
In the development of foresight among groups of businesspeople and their stakeholders, it is common to develop scenarios of future possible worlds and to extrapolate from these the nature of business opportunities and innovations. The process involves explicit “soft” modeling of the landscape, i.e., making assumptions about the interconnections between different actors and the relative strengths of forces and relationships. From the process of describing and constructing these mental landscapes, the actors intuitively create possible strategies and rationale for these. The soft models can be subject to some counterfactual examination of “what ifs” (Fuller, 1999; Carrier et al., 1999).
Each of these examples informs an interpretation from a complexity perspective, but none employs the complete methodology in the sense outlined here. However, these research activities have between them many of the methodological characteristics. For example:
The complexity perspective is important here in introducing a theoretical framework concerning the behavior of agent-based systems that are open, dynamic, evolving, and sufficiently complex to be capable of “emergent” behavior. This framework directs attention to particular aspects of the phenomenon and provides a language for describing what is observed. This language is in terms of dynamical systems and seems to fit intuitively with what we know about small businesses (e.g., they are many, varied, interconnected, and subject to rapid change, including growth, decline, or “death”). The involvement of the human agent (i.e., the ownermanager) entails a concern with “reflexivity” (i.e., conscious intention can be a significant factor in the making of “adaptive moves” and these are not wholly dependent on environmental stimuli). In principle, models can be developed that take account of reflexivity in explaining how particular developments and outcomes occur.
The way in which this research can be developed to further the methodology is of considerable interest to the authors. The aim would not be to describe the “whole” small business system, but to focus on understanding the dynamics of adaptation, learning, and change at both the individual and business level, and how they interact to produce particular outcomes. The role of adaptive agents (owner-managers) is critical here, as are their connections (relationships) with other adaptive agents within a networked “community.”
The selection of salient modeling features from the process of grounding attributes such as personality types, typologies of rule such as reflexive behavior, actor descriptions, and soft models of the relevant landscapes provide a rich basis for abstraction and modeling and the possibility of scientific approaches to theory testing.
The practical output from this research could be particular “sensemaking” tools that could be used by owner-managers themselves or their advisers to understand their situation better and improve their ability to make better adaptive moves. There might also be the opportunity afforded by the building of dynamical models to explore alternative “trajectories” of business growth/development at particular critical junctures in order to aid decision making. The opportunity to test out models/processes via simulation studies might also be explored.
The authors gratefully acknowledge the feedback on an earlier draft of this article from participants in the EIASM Workshop on Complexity and Organisation in Brussels, Belgium, June 25-26, 1999.
Allen, P. M. (1997) Cities and Regions as Self-Organising Systems: Models of Complexity, Reading: Gordon and Breach.
Anderson, P. W. (1972) “More is different: broken symmetry and the nature of the hierar- chical structure of science,” Science, 177: 393-6.
Aronson, J. L., Harré, R., and Way, E. C. (1994) Realism Rescued, London: Duckworth.
Atherton, A. and Hannon, P. (1997) “Strategic awareness and the process of innovation,” Journal of Enterprising Culture, 5 (2).
Axelrod, R. (1998) http:\www-personal.umich.edu axe_793_Syllabus.html.
Bellu, R. and Sherman, H. (1995) “Predicting firm success from task motivation and attri- butional style,” Entrepreneurship and Regional Development, 7: 349-63.
Bhaskar, R. (1978) A Realist Theory of Science, Brighton: Harvester Press.
Byrne, D. (1998) Complexity Theory and the Social Sciences, London: Routledge.
Carrier, C., Cossette, P., and Verstraete, T. (1999) “Experimental implementation of a new creative method to support futurology by small business in a strategic management framework,” ICSB162 in G. Capaldo and M. Rafio (eds) Innovation and Economic Development: the Role of Entrepreneurship and SMEs, Proceedings of the 44th ICSB World Conference, Edzioni Scientifiche Italiane, July.
Carsrud, A., Olm, K., and Thomas, J. (1989) “Predicting entrepreneurial success: effects of multi-dimensional achievement motivation, levels of ownership and co-operative rela- tionships,” Entrepreneurship and Regional Development, 1: 237-44.
Castells, M. (1996) The Rise of the Network Society, Oxford: Blackwell.
Casti, J. (1997) Would-be Worlds, Chichester: John Wiley.
Chaston, I. (1996) “Critical events and process gaps in the Danish Technological Institute SME Structured Networking Model,” International Small Business Journal, 14 (3).
Chell, E., Haworth, J. and Brearley, S. (1991) The Entrepreneurial Personality: Concepts, Categories and Cases, London: Routledge.
Cohen, J. (1998) Verbatim from plenary lecture at Organisations as Complex Evolving Systems Conference, University of Warwick, December.
Coveney, P. and Highfield, R. (1995) Frontiers of Complexity, London: Faber and Faber
Curran, J., Jarvis, R., Blackburn, R., and Black, S. (1992) “Networks and small firms: con- structs, methodological strategies and some findings,” International Small Business Journal, 11 (2): 13-25.
Epstein, J. and Axtell, R. (1996) Growing Artificial Societies: Social Science from the Bottom Up, Cambridge, MA: MIT Press.
Fuller, E. (1997) “Management education of growing businesses— the next 10 years,” paper presented to EIASM Research in Entrepreneurship Conference XI, Mannheim, Germany, November.
Fuller, E. (1998) “Complexity metaphors and the process of small business foresighting,” paper presented to European Institute for Advanced Studies in Management (EIASM) Workshop on Complexity and Organisation, Brussels, June 8-9.
Fuller, E. (1999) “Complexity metaphors and the process of small business foresighting,” in
M. Lissack and H. Guntz (eds) Managing the Complex, Boston. MA: Quorum Books.
Fuller, E. and Moran, P. (1999) “Small firms as complex adaptive systems: a review,” proceedings of the 44th International Council of Small Business World Conference, Faculty of Engineering, P.le Tecchio 80, Naples, June.
Garnsey, E. (1998) “A theory of the early growth of the firm,” Industrial and Corporate Change, 7 (3).
Gatewood, E., Shaver, K., and Gartner, W. (1995) “A longitudinal study of cognitive factors influ- encing start-up and success of venture creation,” Journal of Business Venturing, 10: 371-91.
Gibb, A. and Scott, M. (1986) “Understanding small business growth and development,” in M. Scott *et al. *(eds) Small Firms: Growth and Development, Aldershot: Gower.
Gillies, J. M., Allen, P. M., and Fan, I.S. (1998) “Self-organising supply chain networks: the Italian garment industry,” Proceedings of the Organisations as Complex Evolving Systems Conference, University of Warwick, December: 245-59.
Gould, S. J. (1989) Wonderful Life, the Burgess Shale and the Nature of History, Harmondsworth: Penguin.
Hall, R. and Andriani, P. (1998) “Developing and managing strategic partnerships,” Purchasing and Supply Chain Management Journal.
Hansen, E. (1995) “Entrepreneurial networks and new organisation growth,” Entrepreneurship Theory and Practice, 19 (4, Summer): 7-20.
Harré, R. (1979) Social Being, Oxford: Blackwell.
Harrison, R. and Leitch, C. (1994) “Entrepreneurship and leadership,” Entrepreneurship and Regional Development, 6: 111-25.
Harvey, D. L. and Reed, M. (1996) “Social science as the study of complex systems,” in L.D. Kiel and E. Elliot (eds) Chaos Theory in the Social Sciences, Ann Arbor, MI: University of Michigan Press, 295-323.
Hendry, C., Arthur, M., and Jones, A. (1995) Strategy through People: Adaptation and Learning in the Small and Medium Enterprise, London: Routledge.
Hiebeler, D. (1994) “The swarm simulation system and individual-based modeling,” Working Paper 94-12-065, Santa Fe Institute.
Holland, J. (1995) “Innovation, risk and lever points,” Complexity and Strategy, London: Santa Fe Institute and Praxis Group.
Jarillo, J. (1988) “On strategic networks,” Strategic Management Journal, 9: 31-41.
Johannisson, B. (1987) “Anarchists and organizers: entrepreneurs in a network perspective,” International Studies of Management and Organization, 17 (1): 49-63.
Johannisson, B. (1995) “Paradigms and entrepreneurial networks— some methodological challenges,” Entrepreneurship and Regional Development, 7 (3): 215-32.
Kauffman, S. (1993) The Origins of Order: Self-Organisation and Selection in Evolution, New York and Oxford: Oxford University Press.
Kauffman, S. (1995) At Home in the Universe: The Search for Laws of Self-Organization and Complexity, Harmondsworth: Penguin.
Kelly, K. (1995) Out of Control: The New Biology of Machines, London: Fourth Estate.
Lane, D. (1996) “Strategy under Complexity: Fostering Generative Relationships,” Long Range Planning, 29 (2): 215-31.
Lane, D. and Maxfield, R. (1995) “Foresight, complexity and strategy,” working paper 95-12-10, Santa Fe Institute.
Larson, D. (1992) “Network dyads in entrepreneurial settings: a study of the governance of exchange relationships,” Administrative Science Quarterly, 37: 76.
Lewis, J. and Fuller, E. (1998) “IT perspectives— owner-manager'ss views on IT and small business relationships,” Celebrating the Small Business, proceedings of 21st National Small Firms Policy and Research Conference, Institute of Small Business Affairs, Durham: 912-29.
Lorenzoni, G. and Ornati, O. (1988) “Constellations of firms and new ventures,” Journal of Business Venturing, 3: 41-57.
Luhmann, N. (1986) “The autopoesis of social systems,” in F. Geyer and J. van de Zouwan (eds) Socio Cybernetic Paradoxes, London: Sage.
McGaffey, T. and Christy, R. (1975) “Information processing capability as a predictor of entre- preneurial effectiveness,” Academy of Management Journal, 18 (4).
McKelvey, B. (1998) “Toward a model-centred strategy science: more experiments, less his- tory,” in A. Heene and R. Sanchez (eds) Research in Competence-Based Management, proceedings of Fourth International Conference on Competence-Based Management, Oslo.
McKelvey, B. (1999) “Self-organization, complexity, catastrophe and microstate models at the edge of chaos,” in J. Baum and B. McKelvey (eds) Variations in Organization Science: in Honor of Donald T. Campbell, Thousand Oaks, CA: Sage.
McMaster, M. D. (1996) The Intelligence Advantage: Organizing for Complexity, Newton, MA: Butterworth Heinemann.
Merry, U. (1995) Coping With Uncertainty: Insights From the New Sciences of Chaos, Self- organization, and Complexity, Westport, CT: Praeger.
Miller, D. and Toulouse, J.-M. (1986) “Chief executive personality and corporate strategy and structure in small firms,” Management Science, 32 (11): 1389-409.
Miner, J. (1997) “Psychological typology and relationship to entrepreneurial success,” Entrepreneurship and Regional Development, 9: 319-34.
Mitchell, R. K. and Agle, B. R. (1995) “Toward a theory of stakeholder identification: defin- ing the principle of who and what really counts,” working paper, University of Victoria.
Moran, P. (1998) “Personality characteristics and growth-orientation of the small business owner-manager,” International Small Business Journal, 16 (3): 17-38. Mouzelis, N. (1995) Sociological Theory: What Went Wrong?, London: Routledge.
Nelson, R. R. and Winter, S.G. (1982) An Evolutionary Theory of Economic Change, Cambridge, MA and London: Belknap Press.
North, D., Leigh, R., and Smallbone, D. (1991) “A comparison of surviving and non-surviv- ing small and medium-sized manufacturing firms in London during the 1980s,” Working Paper No. 1, ESRC Small Business Research Initiative, Middlesex University Project.
Outhwaite, W. (1987) New Philosophies of Social Science: Realism, Hermeneutics and Critical Theory, London: Macmillan.
Peters, T. J. and Waterman, R. H. (1988) In Search of Excellence: Lessons From America'ss Best-Run Companies, New York: Warner Books.
Prigogine, I. and. Stengers, I. (1984) Order Out of Chaos: Man'ss New Dialogue With Nature, New York: Bantam.
Reed, M. and Harvey, D. L. (1992) “The new science and the old: complexity and realism in the social sciences,” Journal for the Theory of Social Behaviour, 22: 356--79.
Rosenhead, J. (1998) “Complexity theory and management practice,” Working Paper Series, LSEOR 98.25, London School of Economics.
Rydal, M. (1996) “Entrepreneurs: self-confident agents busting self-confirming equilibria,” Working Paper 96-08-068, Santa Fe Institute.
Sayer, A. (1992) Method in Social Science: A Realist Approach, 2nd edn, London: Routledge and Kegan Paul.
Schon, D. (1991) The Reflective Practitioner: How Professionals Think in Action, Aldershot: Arena.
Schumpeter, J. A. (1934) The Theory of Economic Development, Cambridge: Cambridge University Press.
Stacey, R. D. (1996) Complexity and Creativity in Organizations, San Francisco: Berrett- Koehler.
Suppe, F. (1989) The Semantic Conception of Theories and Scientific Realism, Urbana- Champaign, IL: University of Illinois Press.
Wheatley, M. J. (1992) Leadership and the New Science: Learning About Organization From an Orderly Universe, San Francisco: Berrett-Koehler.
Williamson, O. (1991) “Comparative economic organization: the analysis of discrete struc- tural alternatives,” Administrative Science Quarterly, 36: 269-96.