By Kiran Johny

Abstract

Highlights

Complexity Science Based Design

This framework is based on complexity science. It considers factors like non-linearity, inter-relatedness, emergent property, etc. It promotes self-organization and designability as complementary pairs.

Enabling Transparency And Big Picture Thinking

This framework makes it easier for you to transparently see and navigate the big picture(The Forest), even if you had to attend to the contextual demands(The Tree).

Harmonizing and integrating the human cumulative intelligence

The framework stresses the need to harmonize the growing collective human intelligence of both practitioner and academia-developed models and bring it all accessible and affordable for the entrepreneurial actor at the time of his/her decision making and action.

Interdisciplinary

This framework acknowledges that real-world problems and their solutions are interdisciplinary in nature, and promotes interdisciplinary exploration.

Evolving to greater evolvability

This framework embodies a “Work In Progress” attitude. The framework itself accepts its possible weakness. Everything about it can be changed and designed to suit the idiosyncrasy of the user and the complexity of the context.

A Framework And A Methodology

ESO-Loop can be used both as a framework and a method. As a framework, ESO-Loop provides a complexity frame, a constrain inside which varieties and improvisations are possible. As a method, it can be used in a straightforward manner without any change.

Enabling Adaptiveness

Existing prescriptive models propose permanent structures that end up being inflexible. The framework’s design solution(part 2(ll) is based on constraints that are not permanent structures but an impermanent scaffold that is meant to enable emergence. It can be changed, designed, and suited for the person and context.

Learning and Transfer

ESO-Loop embodies and encapsulates various complexity-friendly functional dispositions into a single framework so that an ecology of ideas in its dynamics will transfer to the users, not just one or two ideas or a linear process.

Immunity against agency hijack

This framework has default immunity against agency hijacking and certainty merchants. People and ideas will compete for your attention and agency. This framework provides inbuilt dispositions and awareness against such an effort.

Part 1
Introduction to complexity, nature of entrepreneurship as a complex domain and the limitations of current thinking space.

Part 2

Building of a complexity-friendly framework.

Part 2(l)

The right worldview.

Part 2(ll)

Effectual Self-Organization

Part 2(lll)

Constraints Theory and Design

Part 3
Conclusion

Part 4
Citation and other details.

Part 1 Introduction

1. Introduction

Entrepreneurship is a complex (McKelvey 2004), heterogeneous, and multi-level phenomenon. It is affected by a large number of interconnected and interacting variables (Bygrave and Minniti, 2000; Giannetti and Simonov,2009), and depends on numerous contextual factors (Welter, 2011; Audretsch. et. al, 2012). This may include factors like; socioeconomic status (Giacomin et al. 2011), education (Millan et al. 2014), gender (Warnecke, 2013; Marlow and Martinez, 2018), time (Klyver et al., 2018; Bird and West, 1997), history (Wadhwani, 2016), location (Audretsch et al., 2012; Storey and Johnson, 2002), stakeholders (Dew and Sarasvathy, 2007), cognition (Dew, et al. 2015; Grégoire et al. 2011; Ward, 2004), intentions (Krueger et al., 2000; Fayolle and Liñán,2014; Krueger, 2017), skills (Lazear, 2004; Hsieh et al, 2017; Oosterbeek et al, 2010), opportunity recognition (Singh, 2001; Lumpkin and Lichtenstein, 2005), acquisition of resources (Leung et al. 2006; Martens et al. 2007), product development (Giardino, et al.,2014), marketing (Lam and Harker, 2015), etc, to list only a few.

One of the major implications of such complexity in entrepreneurship is the inherent uncertainty when it comes to decision-making (Knight, 1921 ). In complex domains like entrepreneurship, past behavior cases may not help us in predicting the future. To predict an event accurately, we need to have a stable system with adequate data. In the case of complex emergent systems, history never repeats. This is why when studying complex domains like entrepreneurship it is essential to use a complexity lens (Berger and Kuckertz, 2016).

In the following part of the chapter, I will discuss two of my core arguments.

1. What is complexity and Why entrepreneurship is called a complex decision domain: In this part, I will explore the nature of complexity by looking into scientific and also decision-making implications of complexity. I will then list down some of the core features of entrepreneurial complexity.

2. Problems with existing models: Although entrepreneurship is a complex field most of the existing prescriptive action models like lean-startup or business planning, etc are created and used without any consideration to the essence of complexity. In this part, I will list some of the core weaknesses of existing entrepreneurship models when dealing with complexity.

2. Complexity and entrepreneurial complexity

Definition

The definition of complexity is by itself context-dependent (Standish, 2001). Its usefulness depends mostly on the domain and specific problems in question. This further adds to the difficulty in providing a universal definition of complexity. To have a generic definition applicable to all contexts, it is necessary to have adequate flexibility. Bruce Edmonds proposes a meta-level and accommodative definition of complexity i.e: “That property of a language expression which makes it difficult to formulate its overall behavior, even when given almost complete information about its atomic components and their inter-relations” (Edmonds, 1995). According to him, this is a very general definition, which is intended to have different interpretations in different contexts.

Melanie Mitchell(2009) also agrees with the above observation by suggesting that “various(complex) systems are quite different”, but adds that, “viewed at an abstract level they have some intriguing properties in common”. They are; Complex collective behavior: It is the collective action of vast numbers of components that give rise to the complex, hard-to-predict, and changing patterns of behavior that fascinate us; Signaling and information processing: Complex systems produce and use information and signals from both their internal and external environments; and Adaptation: All these systems adapt—that is, change their behavior to improve their chances of survival or success—through learning or evolutionary processes. Combining this observation, she proposes her definition of a complex system, i.e. “a system in which large networks of components with no central control and simple rules of operation give rise to complex collective behavior, sophisticated information processing, and adaptation via learning or evolution” (Mitchell, 2009). For an explanation from the entrepreneurship field, Benyamin Lichtenstein(2000) suggested that four assumptions characterize complex systems. They are; Dynamics—Complex systems are dynamic and constantly changing; Irreducibility of elements: Elements are Irreducible due to the entwined nature of the elements; Interdependencies—The causality in complex systems cannot be described by linear models, as the causality is interdependent; Non-proportionality—Small inputs might have a large impact, whereas large inputs might hardly change the outcome.

Understanding by distinction

Apart from definition, One of the best ways to understand the nature of complexity is to find how it is different from other systems like e.g. simple and complicated. Complexity as a field of scientific inquiry was explained well by such a distinction made by American mathematician Warren Weaver in his paper called Science and Complexity(1948).

Weaver(1991) divided the problems of interest in science into three categories. The first category is called Problems of Simplicity. These are problems in which two things are related to each other. These problems are characterized by just a few variables and relations between them. The problems of simplicity were solved by science by developing experimental and analytical methods for handling problems in which one variable, say a gas pressure, depends primarily upon a second quantity, eg, the volume of the gas. These problems are characterized by the fact that the behavior of the first quantity can be described with some degree of accuracy by taking into account only its dependence upon the second variable and by ignoring the negligible influence of other factors. He calls the second category problems of Disorganized Complexity. According to him a problem of disorganized complexity “is a problem in which the number of variables is very large, and one in which each of the many variables has a behavior which is individually erratic, or perhaps totally unknown. However, despite the unknowns, the behavior of all the individual variables, the system as a whole possesses certain orderly and analyzable average properties. These are understood through taking averages over the large set of variables while ignoring or assuming very little interaction among the variables. Examples of such systems are ideal gas (as discussed in thermodynamics) and random coin tosses(as discussed in probability theory). In addition to “problems of simplicity” that are solvable with hard sciences and “disorganized complexity” that can be dealt with statistics and probability, Weaver introduced a third kind called problems of organized complexity. These are problems that involve a moderate to a large number of variables. But the key here is that, due to their strong nonlinear interactions, the variables cannot be meaningfully averaged. Weaver characterized these as “problems which involve dealing simultaneously with a sizable number of factors which are interrelated into an organic whole”, such as a biological or societal system. The third category “organized complexity” is the representation of the type of complex system we are talking about in this paper.

Another powerful model that gives a lot of clarity from distinction is a sense-making framework developed by Dave Snowden (Snowden, 2010) called Cynefin. It differentiates between three kinds of systems (Van Beurden et al, 2013): Ordered systems(simple/ complicated) in which cause and effect relationships are either clear or discoverable through analysis; Chaotic systems in which turbulence prevails and immediate stabilizing action is required; and Complex systems in which the only way to understand the system is to interact. In a complicated context, at least one right answer exists. “In a complex context, however, right answers can’t be ferreted out. It’s like the difference between, say, a Ferrari and the Brazilian rainforest. Ferraris are complicated machines, but an expert mechanic can take one apart and reassemble it without changing a thing. The car is static, and the whole is the sum of its parts. The rain-forest, on the other hand, is in constant flux—a species becomes extinct, weather patterns change, an agricultural project reroutes a water source—and the whole is far more than the sum of its parts” (Snowden and Boone, 2007).

Complexity and Entrepreneurship

It has to be noted that entrepreneurship has components of all decision types(ordered, complex and chaotic Or problems of simplicity, disorganized complexity, organized complexity). But the core aspect of venture creation comes mostly under Weaver’s third category, organized complexity; or in other words a complex domain. Even though scholars have widely recognized entrepreneurship as a complex and highly uncertain process (Lichtenstein et al, 2007; Brouwer, 2000), it is still popularly depicted as an individual activity (Drakopoulou, 2007) that takes place in a perfect market. In most cases, the role of the individual entrepreneur is exaggerated in such representation (Gaddefors and Anderson, 2017). Complex domains like entrepreneurship exhibit several features which cannot be taken into account by most currently existing reductionist models and perspectives (Berger and Kuckertz, 2016).

Following are some of the major characteristics of a complex system(or complex adaptive system) with entrepreneurship as an example. It is important to keep in mind that these characteristics are not separate from each other, and each requires all others for any attempt to explain it.

1. A large number of heterogeneous agents/components/elements:

Complex systems consist of a large number of elements that in themselves can be simple. When the number of elements is relatively small, the behavior of the elements can often be given a formal description in conventional terms. However, when the number becomes sufficiently large, conventional means not only become impractical, they also cease to assist in any understanding of the system (Cilliers, 2002). Further, these elements can be heterogeneous, in that they differ in their behavior, location, history, or other properties. Such heterogeneity, in turn, can have far-reaching impacts on the functioning of the system. For example, different species within an ecosystem will have been in the ecosystem for different lengths of time, possess different functional traits, and occupy different niches (Fisher, 2020).

Heterogeneity of entrepreneurship (Spilling, 2008) may manifest in many dimensions, which are by themselves complex; e.g. entrepreneur, cognition, skills, disposition, idea, customer, teams, investors, venture capitalists, government, bureaucracy, consultants, location, networks, resources, market, technology, law, policy, connection, trust, etc. The importance and inevitability of heterogeneity in entrepreneurship is evident from scholarly works focusing on various perspectives to approach heterogeneity (Davidsson, 2007; Breitenecker and Harms, 2010; Dufays and Huybrechts, 2016).

2. Non-Linear dynamics and interaction among elements(agents/components/variables):

According to Cilliers(2002), a large number of elements are necessary, but not sufficient. In order to constitute a complex system, the elements have to interact, and this interaction must be dynamic. A complex system changes with time. The interactions do not have to be physical; they can also be thought of as the transference of information. This is also evident from Herbert Simon’s definition of a complex system; ie. “one made up of a large number of parts that have many interactions” (Simon 1996). Secondly, the interactions are non-linear. A large system of linear elements can usually be collapsed into an equivalent system that is very much smaller. Non-linearity also guarantees that small causes can have large results, and vice versa. Further, even if specific agents may only interact with a few others, the impact of these interactions are propagated throughout the system. Through this interaction, agents strive to improve their fitness with the environment but the outcome of these attempts depends on the disposition and behaviors of other agents (Mitleton-Kelly, 2003).

In entrepreneurship, agents interact with others in a non-linear manner. For example, an entrepreneur at the same time has to interact with his team, investors, customers, partners, government, etc. He/She gets feedback from the interaction, which triggers more interaction. These interactions will not follow a certain format or order; in other words, it is non-linear. Further, studies have also found that successful entrepreneurship ecosystems outperform others due in part to more amplifying and less dampening of nonlinear interactions (Han et al,2021).

3. Feedback Loop:

Complex systems are characterized by feedback loops in which the product of the process is necessary for the process itself. A feedback loop is “a circular arrangement of causally connected elements, so that each element has an effect on the next, until the last “feeds back” the effect into the first element of the cycle” (Capra 1997, 56). Change in a variable results in either an amplification (positive feedback) or a dampening (negative feedback) of that change (Kastens et al., 2009). Both kinds(positive—enhancing, stimulating or negative—detracting, inhibiting) are necessary. The technical term for this aspect of a complex system is recurrency (Cilliers, 2002). Further, when the interactions of large numbers of components involve positive feedback loops, some behaviors self amplify, quickly crowding out others. Groups of components become locked into self-reinforcing feedback cycles that lead to predictable collective behavior (Anderson, 1999).

Many scholars also stress the non-linearity aspect of feedback loops, calling them non-linear feedback loops (Juarrero, 1999: 47-48). According to Ralph Stacey(1995) organizations are nonlinear network feedback systems and it, therefore, follows logically that the fundamental properties of such systems should apply to organizations. Examples from entrepreneurship include explicit feedback like customer feedback (Eisenmann et al., 2012), stakeholder feedback, and long-term positive or negative feedback. At the macro-level, entrepreneurship is impacted by monetary policy, social climate, etc, and will impact the same by looping back; Like that economic activity promotes entrepreneurship and innovation activities, and entrepreneurship enhances economic activity by looping back (Galindo and Méndez, 2014). Feedback is also discussed in entrepreneurship literature as the one aiding effective decision-making (Haynie and Shepherd, 2007).

4. Emergence or emergent property:

Complex systems show emergent property. Emergence is a systemic process through which properties and or structures come into being that are unexpected, given the known attributes of component agents and environmental forces (Lichtenstein and McKelvey, 2011). Emergent properties refer to a characteristic that is found across the system but which individual parts of the system do not themselves hold. Other words, if a thing can have properties or capabilities that are not possessed by its parts, such properties are called emergent properties (Elder-Vass, 2014). The emergent higher-order behavior cannot simply be derived by aggregating behavior at the level of the elements (Irreducibility, Beckage, 2013). This is because the whole is more than the sum of its parts when it comes to complexity. Another feature of emergent property is non-proportionality (Lichtenstein and Mendenhall, 2002). The effect of input is not proportional to the strength of that input. Due to the non-proportionality, small inputs might have a large impact, whereas large inputs might hardly change the outcome.

Entrepreneurship scholars for long had identified the importance of emergence as a frame for understanding entrepreneurial complexity (Katz and Gartner 1988; Gartner, 1993; Lichtenstein et al, 2006; Fuller et al, 2008). An emergent property is manifested in entrepreneurship in many different ways. At the basic level entrepreneurial venture or a startup itself is an example of emergent property of a large number of factors like intention, cognition, various skills, various stakeholders, location, experts, policy, networks, technology, etc. Further, emergence is manifested at the macro level when innovations breed other innovations(e.g. internet, google by democratizing information access), technologies that result in new disruptive responses, the emergence of new markets and new business models(e.g. app store, android), and collaborative activities(social media, collaboration apps), etc.

The implication of emergent property is that an opportunity(product, organization, market) can come into reality in the absence of deliberate planning. Users often adapt products to support tasks that designers never intended (Johnson, 2006). E.g. Instagram was started as an HTML5 supported location-based service; Facebook was started as an app to compare two people’s pictures and the rate which one was more attractive.

5. Self-Organisation:

Self-organization refers to the emergence of stable patterns through autonomous and self-reinforcing dynamics at the micro-level (Kauffman, 1995; Ska˚r, 2003; Anzola et al., 2017). It is a bottom-up process where complex organization emerges at multiple levels from the interaction of lower-level entities. The final product is the result of nonlinear interactions rather than planning and design; and it is not known a priori (De Roo, 2016). This can be contrasted with the top-down approaches where planning precedes implementation, and the desired final system is known by design. In self-organization, there is no hierarchy of command and control and there is no planning or managing, but there is a constant re-organizing to find the best fit with the environment. The system is continually self-organizing through the process of emergence and feedback. It changes the relationships between the distributed elements of the system under influence of both the external environment and the history of the system (Cilliers, 2002). Further, emergence happens naturally from the interactions within a complex system; they do not need to be imposed top-down in a centralized way. For example, termites are known to construct the highest structures on the planet relative to the size of the builders. Yet there is no chief executive among termites, no architect termite, and no blueprint. Each termite acts locally, following only a few simple shared rules of behavior within a context of other termites also acting locally. The termite mound emerges from this process of self-organization (Plsek and Greenhalgh, 2001).

We can find self-organization as an important character of the market, business, and entrepreneurship. For example, we can see the venture as a self-organized system that emerged to satisfy customer demands. It can be the result of self-organization like Effectuation (Sarasvathy, 2001), Entrepreneurial bricolage (Baker and Nelson, 2005), User innovation (Hippel, 1988), etc. Further, according to Peltoniemi & Vuori(2004), even “the formation of a business ecosystem is a process where participants are gathered voluntarily and without external or internal leader. Goals are set in local interactions, where companies negotiate and create a new order.”

6. Fuzzy boundaries and Nested-embedded complexity:

In a complicated system like a car or algorithm, boundaries are fixed and well defined. Complex systems are systems typically characterized by fuzzy boundaries. Membership of a complex system can change, and agents can simultaneously be members of several systems (Plsek and Greenhalgh, 2001). E.g. investor could be a family member or she might invest in other companies, or join as a startup team member.

Complex systems are also embedded/nested in other systems which are constantly interacting with one another. For example, Our planet is a complex system, as is our body, the organizations that we create, or our social and economic systems (Héraud et al. 2019). When it comes to the economic system, buyers and sellers form markets and markets form economies, whereas the buyers and sellers themselves can be people or organizations formed by people (Noell, 2007). All of these are complex systems by themselves. Because of this embedded nature, it is virtually impossible to accurately mark the boundaries of one system from other systems. Further, Since each agent and each system is nested within other systems, all evolving together and interacting, we cannot fully understand any of the agents or systems without reference to the others (Plsek and Greenhalgh, 2001).

An entrepreneur is a system nested/embedded itself with other complex systems like startup team, customers, market, investors, family, suppliers, partners, government, country, and society itself. In fields of study like entrepreneurship, business, organization, etc, scholars have been exploring the nested and embedded nature of multiple complex co-evolving systems (Holm,1995; Baum and Singh, 1994; Dieleman and Sachs, 2008; Ng, 2011). Scholars have also explored the contextually embedded nature of entrepreneurial actors (Baker and Welter, 2018; Autio et al, 2014; Lounsbury et al, 2019), suggesting the entangled and embedded nature of entrepreneurship to the context in which it is situated.

7. Dynamics of evolution: Evolution, Adaptation, Co-evolution, Path Dependence:

Complex systems follow the dynamics of evolution. The behavior of complex systems evolves from the interaction of agents at a local level without external direction or the presence of internal control (Kernick, 2006). This should not be viewed as local Darwinian processes and optimization alone (Pendleton and Brown, 2018), but also as dynamic exchanges between components and the system as a whole, impacting other complex systems. Following are some of the ways in which the dynamics of evolution manifests in complex systems. First of all, adaptation forms one of the fundamental features of evolution. All complex systems adapt—that is, change their behavior to improve their chances of survival or success—through learning or evolutionary processes (Melanie Mitchell, 2009). They can (re)organize their internal structure without the intervention of an external agent (Paul Cilliers, Emergence, March 2000). Secondly, In nature and complex systems, evolution is manifested in the form of co-evolution (Ehrlich and Raven, 1964). When organisms adapts in a reciprocal way, it will lead to Coevolution (Thompson and Cunningham, 2002). It refers to the simultaneous evolution of entities and their environments, whether these entities are organisms or organizations (Baum & Singh, 1994). It encompasses the twin notions of interdependency and mutual adaptation, with the idea that species or organizations evolve in relation to their environments, while at the same time these environments evolve in relation to them (Porter, 2006). Adding to this we can take Kauffman’s(1993, 237) observation that, “all ‘evolution’ is really coevolution”. According to him, “the true and stunning success of biology reflects the fact that organisms do not merely evolve, they coevolve both with other organisms and with a changing abiotic environment”. He argues that coevolution is at the root of self-organizing behavior, constant change in systems, the production of novel macro structures, and non-linearities.

Further, according to Cilliers(2002), Complex systems have a history. Not only do they evolve through time, but their past is co-responsible for their present behavior. This is equivalent to ideas like path dependence (Liebovitz and Margolis, 1995; lock-in’s(Arthur, 1989), or Imprinting (Levinthal, 2003; Johnson, 2007, etc.

An evolutionary perspective is indispensable from the complexity worldview of entrepreneurship. According to Cordes et al(2008), “firms are culturally variable and evolve new cultural forms as time passes. This evolution is partly driven by entrepreneurs and other business leaders in entrepreneurial roles (Penrose, 1959; Langlois, 1998), partly by the decisions made by rank-and-file members, partly by the firm’s competitive success or failure (Alchian, 1950), and partly by cultural evolution in the larger society within which firms are embedded”. Acknowledging this centrality, evolutionary dynamics in entrepreneurship is one of the most widely explored topics among scholars. This includes meta-level exploration of the evolution of the field itself (Landström, 2020; Carlsson et al, 2013) to various other dimensions of evolutionary dynamics (Schaltegger et al. 2016; Pacheco et al., 2014; Tiwana et al., 2010; Inkpen et al., 2004; Lewin et al, 1999; Jones, 2001; Grodal et al., 2015).
.

8. Attractor behavior:

In a complex emergent system, an attractor is a set of states towards which a system will naturally gravitate and remain cycling through. They are islands of stability in a sea of chaos. Complex systems are inherently uncertain, but, they usually settle down into one of a number of possible steady states. These steady states are called “attractor basins”.
As a universal feature of complex adaptive systems, the dynamics of attractors are active in entrepreneurship too. Complexity theory views organizations(or startups, new ventures, etc) as “complex adaptive systems” that coevolve with the environment through the self-organizing behavior of agents navigating “fitness landscapes” (Kauffman, 1995) of market opportunities and competitive dynamics (Coleman, 1999). Such self-organizing systems typically evolve towards a state of equilibrium, or an attractor state. Once there, the further dynamics and evolution of the system are likely constrained to remain near the attractor. This also means changing external and internal “attractors” influence the process of adaptation by agents (Coleman, 1999; Stacey, 1996). Since attractors are the most stable and robust elements in these systems, they are more feasible targets for foresight than the several variants that they configure and effectuate (Kuhmonen, 2017).

The word ‘attractor’ has the connotation that it attracts the system to a certain state. However, it is not a ‘thing’ out there, but an expression of the extent to which a system can change, indicated by its resilience against disturbances (Gerrits, 2012; Marion, 1999). Further, it is a model representation of a system’s behavior. It is not a force of attraction or a goal-oriented presence in the system; it simply depicts where the system is headed based on its rules of motion (Dolan et al, 2000)

There are 3 major types of attractors (Gerrits, 2012). When a system always returns to exactly the same state when under pressure, it can be expressed by a fixed point attractor (Otter, 2000). A periodic or torus attractor describes the alternation of systems between a limited number of states (Mackenzie, 2005). The third category is highlighted by the inability of a system to return to any of its previous states, in which–each new state is ever so slightly different from the previous one. This category is called the strange attractor (Byrne, 2002). It can remain in this current state because of existing feedback loops that maintain a particular situation. It limits the dynamics of the system but also makes it impossible to predict exactly where the system is going to be.

The strange attractor is very important in domains like entrepreneurship, business, sociology, etc, because many of the dynamics in these systems behave as if they were guided by these strange attractors. They don’t go back to the old state, or jump between states, but are guided by the dynamics of strange attractors. New market opportunities are a good example of attractors. These attractors “pull” entrepreneurs to innovate within existing firms or found new enterprises (Miles et al., 1998).

9. Miscellaneous:

Apart from the above 8, the following are also core aspects of complex systems.

Complex systems have a history and history matters; According to Cilliers(2002), Complex systems have a history. Not only do they evolve through time, but their past is co-responsible for their present behavior. This is equivalent to ideas like path dependence (Liebovitz and Margolis, 1995); lock-ins (Arthur, 1989), or Imprinting (Levinthal, 2003; Johnson, 2007), etc.

Complex systems are open systems; They exchange energy or information with their environment—and operate at conditions far from equilibrium. It is also important to note that Self-organization only occurs in such open systems that import energy from the outside (Prigogine and Lefever, 1974). As a result of this, it is often difficult to define the border of a complex system. Further, closed systems are usually merely complicated (Cilliers, 2002). Eg. A venture started in India will be affected by dynamics of US or Chinese markets, Oil price, technology emergence, a crisis like COVID, etc. All these factors are not going to affect the functioning of a complicated system like a PC, car, or machine.

Complex systems evolve in the adjacent possible (Kauffman, 1996); a zone towards which change and evolution is more likely because of the current disposition of the system. The concept of “adjacent possible” was introduced by Stuart Kauffman(1996; 2000) in evolutionary biology and complex adaptive systems to explain how biological evolution can be seen as exploration and actualization of what is adjacent possible. It can be defined as “the set of possibilities available to individuals, communities, institutions, organisms, productive processes, etc., at a given point in time during their evolution” (Loreto, 2015). The concept of the “adjacent possible” is useful for understanding how entrepreneurial adjacent possibilities emerge, and how the new adjacent possible will lead to yet newer adjacent possibilities.

Complex adaptive systems like entrepreneurship display power-law distributions; Power laws and outliers are pervasively present across many types of variables that are central to most entrepreneurial activities (Crawford et al, 2015). It is also clear in the success and impact of entrepreneurial ventures. For example, 95% of all U.S. businesses are small–employing 20 people or fewer, and more than 60% of all new jobs are created by a mere .03% of all entrepreneurial start-ups. Further, most successful companies of our time, Facebook, Google, Apple, Amazon, Walmart, etc, are extreme outliers (Aguinis and Joo, 2015).

Finally, Complex systems may show behaviors like the Mathew-effect(those who begin with advantage accumulate more advantage over time and vice-versa for those who begin with disadvantage), Network-effect( increases in product value with increases in the number of users), Preferential attachment(more connected a node is, the more likely it is to receive new links). For E.g. Initial success improves access to deal flow in venture capital, which leads to accumulated advantage (Nanda,2020). These effects may explain the success of already successful or second-time entrepreneurs, etc.

3. Implication of complexity in decision making

Domains like entrepreneurship exemplify all of the previously listed characteristics of a complex system. Following are some of the characteristics of decision-making in real-world complexity.

* Real-world decisions involve uncertainty (Knight,1921): This will limit the scope for identifying a rationally optimum decision and allow for a range of competing proposals and preferences to claim credibility (Lyles, 1981). Uncertainty can be viewed as part of the real world and it affects individuals at various levels (Folta, 2007; Platt and Huettel, 2008). The role of uncertainty and its various dimensions of it in entrepreneurship is also a widely observed phenomenon (Sarasvathy and Kotha, 2001; Bylund, and McCaffrey, 2017).

* Real-world decisions are affected or constrained by Human Bounded Rationality: In a complex and uncertain world, humans make decisions under the constraints of limited knowledge, resources, and time. Herbert Simon (1957) thus argued that humans have bounded rationality. He suggested that the complexity of the environment and humans’ limited cognitive system make maximization impossible in most real-life decision-making situations.

* Real-world decision contexts thus are mostly based on satisficing logic (Simon, 1947; Brown,2004): We rarely make a decision after possessing all the facts, and as a corollary, In practice, we rarely make a decision based on all the facts in our possession. This is Simon’s (March and Simon, 1958) notion of ‘satisficing’, which refers to the less than optimal choices due to decision makers’ bounded rationality or limited cognitive capacities and lack of time and energy. This involves using of simple rules than complicated models, otherwise known as heuristics (Gigerenzer, 2011).

* Real-world decisions are context-dependent (Trueblood et al, 2013; Hutchins and Klausen, 1996). As demonstrated by Herbert Simon’s(1990) scissors analogy, real-world decisions must include cognition and context(environment). In other words, we can say, real-world cognition and decision-making are; ecological (Gibson, 1979); embodied, embedded, enactive (Varela et al., 1992), and extended (Clark and Chalmers, 1998) outside of our brain. This means it is essentially impossible to understand(or effect) real-world decisions by cognitive reductionism.

* Real-world decision-making in complex domains can be understood in different ways. Following are 3 of them (Campbell, 1988). Firstly, as a subjective psychological experience. Secondly as an interaction between task and person characteristics (March and Simon, 1958), in that, tasks can be said to be more or less complex relative to the capabilities of the individuals who perform the task. Thirdly, as a function of objective domain or task characteristics(climate change, Entrepreneurship).

* Real-world decision contexts don’t have boundaries. Categories like Simple, complicated, complex, chaotic (Snowden and Boone, 2007), even though discussed and studied separately, may appear part of a real-world decision context. This means decision-makers deliberately or intuitively engage in continuous sense-making effort (Kurtz and Snowden, 2003) to identify the decision context they are in. For example, the core aspect of entrepreneurship is complex and uncertain. Even though this is the case, there are complicated jobs to be done like a lawyer’s job, accounting, etc. Simple jobs like switching on the lights, cleaning the floor, etc.

* Real-world decision-makers use ambidextrous methods. Entrepreneurs use both causal and effectual approaches in a variety of combinations. Both causation and effectuation are integral parts of human reasoning that can occur simultaneously, overlapping and intertwining over different contexts of decisions and actions (Sarasvathy, 2009). According to Yang and Gabrielsson(2017), entrepreneurs use ambidexterity in methods. i.e. the capability to employ both exploration and exploitation methods, or simultaneously use the effectual process to explore and create a new market, and the causal process to exploit the existing market. Further,

* Real-world problems don’t care about disciplinary boundaries(Bendix, 2020). This asserts that interdisciplinary working is needed in order to explore the ‘real world’ problems (Dalrymple and Miller, 2006). This awareness may help the decision-maker to expand the search space. Real-world entrepreneurship is contextual and according to Welter’s (2011), “a contextualized view on entrepreneurship asks for an interdisciplinary perspective, as the solution cannot be to develop an overarching theory of entrepreneurship in all contexts, but rather working with disciplines like anthropology, sociology, and others, which possess some of the tools and concepts entrepreneurship scholars need to explore the variety, depths, and richness of contexts”. Scholars have also acknowledged this importance in entrepreneurship education (Penaluna and Penaluna, 2009; Janssen et al, 2007)

* Real-world predictions about the future cannot be accurate even with all the knowledge about the past. Out past and history is important because of the path dependence, but it can only suggest insights about the current disposition, not an accurate future state. This can be explained by the butterfly effect which occurs when a very small change in one part of a complex adaptive system may initially go unrecognized resulting in a massive disruption, surprise, or turbulence. The results may be impossible to predict (Bennet and Bennet, 2008). This is because we can’t get the accurate and massive amount of necessary data in its dynamics. This can also be seen as an implication of unavoidable uncertainty in the initial condition (Boffetta et al, 2002).

* Real-world decisions may involve trade-offs between exploit(efficiency) and explore(evolvability). Many decisions in the lives of animals and humans require a fine balance between the exploration of different options and the exploitation of their rewards (Mehlhorn et al, 2015). Exploitation maximizes rewards in the near term, while the information obtained during the exploration can later be used to maximize rewards in the long term (Barack and Gold, 2016). Too much exploitation could promote rigidity and optimization in a local niche, and too much exploration may result in becoming a situation like “a jack of all trades, but master of none”. Here also the concept of satisficing is important.

* Real-world decisions may have unintended-consequence (Merton, 1936), particularly in domains of many interacting agents. E.g. The Hawthorne effect (Sedgwick and Greenwood, 2015; Wickström and Bendix, 2000), Cobra Effect (Warczak Jr, 2020), etc. The unintended consequences are the result of complexity, and the inability of decision-makers to understand and anticipate the emergent realities.

* Real-world decisions may directly affect other people in the network. Further, people in one’s network affect his/her decision-making and are in turn affected by decisions of other agents (Christakis and Fowler, 2007; Edelenbos and Klijn, 2007; Sadovykh et al, 2015; Ingold and Leifeld, 2016).

* Real-world complex contexts may not have clear measures of performance or success. It may be absent, conflicting, or vague. Who can say person x made the better fatherly decision for his child than person y. It is totally relative. Even the output, i.e. the child’s success may or may not have anything to do with the parenting decision in a long run. This is not to say that there is no measurable performance in the real world; there are. In domains where performance can be objectively measured, such measures drive success, but when performance can’t be measured, it is possible that networks, people or society drive success” (Barabási, 2018). Success is thus a collective endeavor, independent from individual performance. Besides performance, success requires building trust and reputation in order to shape others’ perceptions regarding one’s achievements (Day, 2019).

* Real-world decisions are not just individual-driven but distributed (Rapley, 2008; Schneeweiss,2012). Decision-making in the real world is distributed across various stakeholders(eg. Investors, co-founders, family), institutions(government agencies, partner companies), artifacts(software, notebook, Mobile), etc. A real-world example is patient and doctor decision making, in which both patient and doctor have a legitimate investment in the treatment decision; hence both declare treatment preferences, the rationale for such preference, etc while trying to build consensus on the appropriate treatment to apply (Charles et al, 1997, 1999). In real-world contexts like entrepreneurship, there are multiple stakeholders with diverse and conflicting beliefs, preferences, and goals. This idea is enacted in the “Crazy Quilt Principle” of effectuation (Sarasvathy, 2009), where partnerships determine, to a great extent, which product or market the company will eventually end up entering or creating.

* Real-world decision problems are interactive and dynamic. In a complex domain, decision-makers(entrepreneurs or managers) are not confronted with problems that are independent of each other, but with dynamic situations that consist of complex systems of changing problems that interact with each other. These problems are labeled by Ackoff as messes (Ackoff 1978; Bennet et al, 2008).

* Real-world decisions involve an active role for emotions (Hermalin and Isen, 2008). Emotion is an intrinsic part of human decision-making in the real-world context. Research reveals that emotions constitute potent, pervasive, predictable, sometimes harmful, and sometimes beneficial drivers of decision making(Lerner et al. 2015). Many scholars now view emotions as one of the dominant drivers of most meaningful decisions in life. This is well observed in entrepreneurship also (Foo, 2011; Cardon et al, 2005; Arpiainen et al, 2013).

* Real-world decisions are path-dependent (Koch et al, 2009; Bindler and Hjalmarsson, 2019). Path dependence means that where we go next depends not only on where we are now but also upon where we have been; Thus, “History Matters” (Liebowitz and Margolis, 1999). Our decisions of the past and present will influence the future decision landscape, opening up some choices, and constraining others. In the same way, It is also influenced and framed by the path-dependence of other agents.

* Real-world decisions and success are influenced by timing. The right timing is one of the most important aspects of decision making in complexity and it is true in the entrepreneurship context too (Wadhwani et al, 2020). According to Bill Gross(2008), the single biggest reason why startups succeed is timing. Since complex domains exhibit dynamics of interaction and evolution (& co-evolution) the opportunity structure changes all the time. We must act when the evolutionary potential of other systems around us is primed perfectly for a change. For E.g. Facebook was started when a basic infrastructure was already evolved enough to support it, like the internet, connectivity(user base), previous models that showed the way(Orkut, myspace, etc.).

* Real-world decisions may be impulsive. Most human decisions in everyday life are not based on elaborate planning and research. Even in the case of important life decisions, people may rely more on emotions and impulsiveness. Simple heuristics are the most preferred decision-making device human beings use, and they mostly work (Gigerenzer, 2008). Studies have also shown that a significant part of entrepreneurial behaviors may also occur without ex-ante reasoning (Wiklund, 2019). According to Lerner et al(2018), non-deliberative impulse-driven behavioral logics can also be the basis for business venturing.

* Real-world decisions may be based on intuition. Researchers have discovered and acknowledged that decision-makers in real-world decision contexts rely heavily on intuition (e.g., Klein, 2015; Gigerenzer and Murray, 2015; Kruglanski and Gigerenzer, 2011). Further, according to William Duggan(2013), decision-makers may also use different types of intuition, namely; ordinary(every day), expert(fast judgment based on experience and practice), and strategic(a flash of insight that works in new situations).

* Real-world decisions are based on selective perception and use of information (Dearborn and Simon, 1958; Walsh, 1988; Beyer et al, 1997). There is always far more information in our objective environments than we can perceive or attend to. Thus, perceptions must be strongly guided by anticipations. As March and Simon (1958) noted, under conditions of complexity and imperfect information, decisions are made on the basis of selective perception and identification with sub-goals.

* Real-world decisions may have heterogeneous reasons. Bringing all of it together, it is absolutely clear that real-world decisions are massively complex and it can have a lot of different dimensions. This makes it difficult to pinpoint specific causes because you can only see the dimensions that you are looking at. Thus it can be, cognitive biases (Stanovich & West, 2008), past experience (Juliusson et al, 2005), escalation of commitment (Sleesman et al, 2012), age and individual differences (Bruin et al, 2007), belief in personal relevance (Acevedo, & Krueger, 2004), etc. To conclude, an economics model-driven view of decisions historically assumed the existence of a straightforward, linear thinking, rational decision-maker (Brzezicka and Wiśniewski, 2014). But in real-world complex contexts, decisions are often not what they appear to be. A decision presented as a technical decision may in fact be tactical, political and strategic, or egocentric, emotional and personal, or any or all of them, or more.

4. Implication of complexity in the development of expertise

In the previous chapters, I have demonstrated why entrepreneurship is a complex decision domain by examining some of the features of entrepreneurial complexity. This chapter is about the expertise dimension of complexity with reference to entrepreneurship.

Expertise under complexity

According to Kahnemann & Klein(2009), two conditions must be satisfied for a judgment to be recognized as coming from real expertise(expert intuition); First, the environment must provide adequately valid cues to the nature of the situation. The second one is that people must have an opportunity to learn the relevant cues. The determination of whether intuitive judgments can be trusted thus requires an examination of the environment in which the judgment is made and of the opportunity that the judge has had to learn the regularities of that environment. A crucial conclusion is that skilled intuitions will only develop in an environment of sufficient regularity, which provides valid cues to the situation. Thus, If an environment provides valid cues, prolonged practice opportunity with rapid and unequivocal feedback,—skill and expert intuition will eventually develop in individuals of sufficient talent. Such an environment also helps true experts develop skills to know when they don’t know something, something which non-experts certainly do not know (Kruger and Dunning, 1999).

Other scholars (before Kahnemann & Klein, 2009) have also tried to bring out this case. For example, James Shanteau(1992) used the classification of Type 1 Type 2 domains to analyze the complexity and learnability of various environments. Hogarth’s(2015) concept of the kind and wicked learning environments is another model representing a similar idea. He described wicked as situations in which feedback in the form of outcomes of actions or observations is poor, misleading, or even missing. By contrast, in learning environments that are kind, feedback links outcomes directly to the appropriate actions or judgments and is both accurate and plentiful.

Expertise in entrepreneurship.

First of all, any effort to understand expertise in a complex domain like entrepreneurship is by itself complex. Every single issue in question may include a heterogeneous mix of problems and domains(e.g. simple, complicated, complex, and chaotic parts, OR/AND Physics, Biology, Chemistry, Economics, Psychology, etc.), all of which require different frame-sets for understanding.

Secondly, scholars may argue that entrepreneurship is a low validity domain (Kahneman and Klein, 2009) where genuine expertise is not possible. To have genuine expertise to develop, the domains must be of high validity. i.e. “Skilled intuitions will only develop in an environment of sufficient regularity, which provides valid cues to the situation” (Kahneman and Klein, 2009). This feature was also previously spotted by a review by Shanteau(1992) which confirmed the importance of predictable environments and opportunities to learn them, in order to develop real expertise. To Kahneman and Klein(2009) prolonged practice and feedback that is both rapid and unequivocal are necessary conditions for expertise, provided by predictable environments.

Thirdly, deliberate practice may not work in a complex domain like entrepreneurship. Saraswathy(2009) defines an expert as someone who has attained a high level of performance in the domain as a result of years of experience and deliberate practice (Ericsson et al., 1993). But in recent scholarly works, it has been observed that deliberate practice may not guarantee better performance in extremely complex domains. A 2014 meta-analysis (Macnamara et al, 2014) has shown that deliberate practice only explained 26% of the variance in performance for games, 21% for music, 18% for sports, 4% for education, and less than 1% for professions. This further demonstrates a low connection between deliberate practice and performance in complex unstructured domains.

Fourthly, expertise in complex social domains is distributed (Edwards, 2010). It is not necessary that an entrepreneur must be an expert in finance, accounting, programming, law, etc. Such expertise is distributed(and or extended) across various individuals(lawyer, doctor) institutions(law enforcement, companies) and artifacts(tools, software). etc. It is not even necessary that the entrepreneur know the entrepreneurial core activities. He can still win in-case he is in the right high network place(e.g. Harvard, Stanford, etc.), get good people to mentor and work(e.g. Facebook case of Sean Parker, Peter Thiel), get access to specialized institutions(e.g. YC in the case of Dropbox), have a rich family to support, etc. He can also fail despite all of this(see next).

Fifthly, complex domains like entrepreneurship are subjected to various complexity laws like power laws, Mathew effects, reputation effects, ecosystem-embedded-preferential-attachment, etc. This invalidates success as a metric of expertise. Core events in complex systems like entrepreneurship never repeat in originality(strange attractor effect), feedback is delayed, and since complex systems are governed by power laws, small things(e.g. Harvard dorm Facebook) can result in huge success, and resource rich interventions can fail(google plus). A tangent is that the emergent property of a system may not be the result of the expertise of a particular agent or agents, but because of the dynamics of the whole system co-evolving with the ecosystem as a whole. This may prevent us from establishing any valid causal relationship between expertise and performance in a domain like entrepreneurship. Thus in complexity, high performance may not guarantee success, in that, the success of an individual does not depend uniquely on the quality of performance (Barabási, 2018).

Sixthly, If anyone becomes successful in such a system she(he) acquires what I call ecosystem embedded accumulated advantage(e.g. PayPal mafia), an entrepreneurial Mathew effect which makes it easy for the agent to be successful in any substituent pursuits. This makes it difficult to identify the causation behind any success. That also makes it difficult to compare a successful entrepreneurial agent with new startup founders.

Finally, all of the above issues lead to the idea that entrepreneurship, at least the core aspect of it involves “Complex indeterminate causation” (Hoffman et al., 2011) which represents a situation in which rational individual decision-making is hindered by an inability to discern a clear chain of causality. “In real-world settings, the evidence for causation is typically too ambiguous to permit valid reasoning, so [rationality] is not a generally useful standard” (Hoffman et al., 2011, p. 419).

5. Problems with current dominant entrepreneurship models

Adjoined with the technology entrepreneurship boom and the glamorization of entrepreneurship as an ideal career choice, there has been a growing interest among various actors in the entrepreneurial ecosystem to provide/get valid guidance in entrepreneurial action. This is also because many have grown increasingly skeptical of the usefulness of traditional methods like business plans. Both scholars and practitioners have responded to this demand by suggesting a variety of heuristics (Manimala, 1992) and prescriptive models (Mansoori, 2017).

Entrepreneurial heuristics includes thumb rules guiding the decisions involved in the start-up and management of a new venture. E.g. Product-market fit, Do Things that Don’t Scale, 1000 true fans(or 100), Mom Test, Jobs to be done, etc.

Prescriptive methods or models are principles of thought and action that guide the theoretical and practical aspects of human action (Mansoori 2017). Some examples are; Business Planning (Sahlman, 1997, Delmar and Shane, 2003), Contingency planning (Honig, 2004, Marc Gruber 2017), Discovery-driven planning (McGrath and MacMillan, 1995), Probe-and-Learn approach (Lynn et al., 1996), Lean start-up approach (Blank, 2013; Ries, 2011), Theory-Based View (Felin et al, 2020), Disciplined Entrepreneurship (Sull, 2004), Design thinking (Brown, 2008; Plattner, 2013), Effectual entrepreneurship (Sarasvathy, 2001), etc.

Weakness of existing prescriptive models.

Prescriptive methods of entrepreneurship have been criticized for their lack of rigor and relevance by various scholars (Garbuio et al., 2018; Frank and Landström 2016, Arend et al. 2015, Wolf and Rosenberg 2012; Yordanova, 2022; York and Danes, 2014). Scholars have also shown that entrepreneurship is a naturally complex system, and why it is necessary to study entrepreneurial phenomena by applying the frame of complexity science. This is a major challenge because complex systems are affected by the emergent property, evolutionary dynamics, non-linearity, etc. In complex systems like entrepreneurship past behavior may not predict its future behavior. A small change in any of the initial condition variables can result in a huge difference in the end result.

I argue that most prescriptive models are reductionist in origin and most are ignoring (or ignorant) of complexity by omission or prescription. Most of the models listed above are originated or validated by using traditional linear scientific methods, transforming the results into prescriptive insights for entrepreneurs, suggesting how should they act. Further, it has been found that regression analysis is rated as the most fundamental method in entrepreneurship research scholars should be familiar with, but is inappropriate for studying complex systems (Berger and Kuckertz, 2016). This is especially important when scholars have exposed various types of complexity-effects that cannot be understood by traditional methods. For e.g. Crawford et al(2015) have identified complexity effects in entrepreneurship in the form of power laws, suggesting the use of complexity informed methods.

Another source is anecdotal and idiosyncratic experiences of ecosystem actors like entrepreneurs or institutions like VCs or accelerators. A widely used example that fits the category is the lean startup (Ries 2011), which is often criticized for its explicit demonstration of survivorship bias(which is a bias of focusing on a successful sample and claiming it to be representative of the entire group).

Following are some of the significant weaknesses of existing methods/models when considering the nature of entrepreneurial complexity.

1. Ignoring context

Entrepreneurship is still a highly decontextualized research field, which is also reflected in the popular prescriptive models we use, and also one-size-fits-all advice received to entrepreneurs. Most of the dominant perspectives still assume that entrepreneurship is the same all over the world, regardless of historical, cultural, institutional, social, and spatial contexts. This is also clear from the growing homogeneity in mentorship, accelerator, and entrepreneurship education programs. This has been described as a McDonaldization of entrepreneurship education (Hytti, 2018), a highly standardized menu of activities is served up for student consumption, such as competitions and mini-company creation. According to Baker and Welter(2018), the messiness and complexity of entrepreneurship derive from contextual questions like; Who is driving it?; A community? A family? An individual?. What are the different types of social networks invoked?; friends, family, community, various stakeholders, etc. These are embedded within, and in turn, affect regulatory, and normative contexts at the community, regional, national, and broader levels. Thus entrepreneurship can only be understood within its historical, temporal, institutional, spatial, and social contexts, as these contexts provide individuals with opportunities and set boundaries for their actions. Most prescriptive models promote a model-centric view of the world and are not flexible or intelligent enough to capture crucial contextual insights.

2. Lacks Evolvability

Evolution is a fundamental feature of nature, biology, and human social life. In comparison to biological evolution, the human social system is capable of adapting to change at a faster rate. It is particularly significant in the case of entrepreneurship in that entrepreneurs often act as catalysts of social evolution and also are affected by the smallest of changes in the environment. One of the major sources of weakness in existing entrepreneurship models is the lack of evolvability. Once proposed and written down the model never changes. Even with this vulnerability, most ideas and models in the entrepreneurship domain are self-aggrandizing and self-perpetuating. This often works against the agent’s evolutionary potential because it gives a false sense of comfort and certainty, blinding the agent from ever realizing the true emergent nature of entrepreneurial complexity.

3. Fixation Errors and Design Fixation

According to De Keyser and Woods (1990), many critical real-world human problem-solving situations take place in dynamic event-drivers environments, where the evidence comes over time and situations can change rapidly. In these situations, people must amass and integrate, uncertain, incomplete, and changing evidence. A major source of human error in dynamic domains seems to be a failure to revise situation assessment as new evidence comes in. I argue that most of the commonly suggested methods like a business plan and lean-startup are highly vulnerable to what Keyser and Woods call Fixation errors. Prescriptive models are especially vulnerable to another type of fixation called design fixation (Jansson and Smith,1991), which refers to the designer’s inability to consider multiple strategies to formulate and solve a design need. It is the direct result of knowledge representation, with human knowledge argued to be organized categorically. These categories are defined by prototypes that exemplify the category (Condoor & LaVoie, 2007). Since linear approaches like Lean startup and business model canvas etc. are based on structured and step-by-step processes, they may add to the rigidity and hence may act as counterproductive in dynamic high-stress environments. This rigidity may lead to design fixation (Garbuio et al., 2018).

4. Structure and Issue of Agency

Agency is described as an individual’s ability at any given point in time, to act independently in order to change the internal or external environment (Bandura, 2001; Campbell, 2009; Hitlin & Elder, 2007). Scholars have long observed that the structure of society that includes other agents and artifacts will have influence over the individual agency. Thus it can be observed that once designed and introduced into the interactional scene by humans,—-texts, artifacts, and objects of any kind make sense and have an agency on their own (Caronia and Mortari, 2015; Bruno Latour, 1996). I argue that most of the popular prescriptive models have agency of their own, and most of them are designed to highjack entrepreneurial agency by providing a model-centric view of the world. This can result in developing commitment to one-size-fits-all models that may result in the agent’s evolutionary, learning, adaptive potential being seriously compromised. Further, In the descriptive sense, most of the prescriptive models assume that entrepreneurs possess excessive powers of agency (Kitching, J., & Rouse, J.2020). This is evident from the inherent assumption that regular entrepreneurs are using exactly what the model prescribes, almost like an algorithmic precision. In the practical sense, the models take control of the entrepreneur’s agency by prescribing things to do and think. It is suggestive of what to see, what to do, what not to do, etc.

5. Local optima, Bounded rationality, and Local search

Research has found that bounded rationality challenges new ventures differently than it does to established firms and that entrepreneurs appear to systematically satisfice (Simon, 1955) prematurely across many decisions (Cohen et al 2019). Most of the prescriptive models assume(by omission or prescription) that the entrepreneurs are able to review all information and make a rational decision in their best interest. Research on human bounded rationality suggests these assumptions are wrong. Another dimension of this issue is that entrepreneurs generally start their venture with a preference to ‘local search’, whereby they primarily explore opportunities that fit with their existing disposition and knowledge. This leads to the identification of local optima rather than global optima (Stuart & Podolny, 1996; Keinz & Prügl, 2010; Rosenkopf & Nerkar, 2001). An empirical study done by Shane(2000) has also shown that entrepreneurs tend to identify opportunities that were either already known to them in the past or are closely related to their existing stock of knowledge. This phenomenon is also known as local search bias (Stuart & Podolny, 1996; Rosenkopf & Nerkar, 2001). The local search behavior is of inevitable importance but the problem is that focusing on existing knowledge and expertise alone can Impede the entrepreneur from exploring distant solutions. I argue that most of the existing entrepreneurship models reinforce and augment bounded rationality and local search behavior. They are optimized for local search behavior in which they may make entrepreneurs more susceptible to being blinded by their own bounded rationality. For a design to be sustainable it must acknowledge that the end-users are fallible humans with bounded rationality. It must take into account cognitive limitations, bias, and dispositions. Models must anticipate the potential issues that could arise from bounded rationality and compensate for them with built-in design solutions.

6. In exclusion To others by omission or prescription.

In complexity and evolution, heterogeneity is an asset. The more the relevant elements of variation present(requisite variety ) in a system, the more resilient and evolvable a system will become. Despite that, most models are proposed as solutions for the problems of the existing one. It is often pitched as better than the previous one. E.g. Plan Vs Lean startup, effectuation Vs causation, Theory-based view vs Lean. Using the antecedent as a frame to marginally improve the model is also the standard practice. While this is the reality, studies have shown that real entrepreneurs don’t behave according to the dictum of any prescriptive models. Fisher(2012) for example investigated behavior underlying the venture founding process using effectuation, bricolage, and causation has found that entrepreneurs often use elements of all the 3 theoretical models and that there were no cases in which only the behaviors associated with causation were responsible for the development of the venture. Mansoori and Lackéus(2019) in their analysis stressed the possibility of using multiple methods complementing each other, which is a standard practice in weather prediction(ensemble models). Sarasvathy (2001) also asserted the point that both causation and effectuation are integral parts of human reasoning that can occur simultaneously, overlapping and intertwining over different contexts of decisions and actions” (p. 245). Real-world decisions in a complex adaptive system cannot be understood with linear thinking or binaries. Since it is impossible to know various emergent decision contexts beforehand, it is always ideal to design solutions for such variables and contingencies. That means not excluding ideas, models and theories, etc by prescription or omission.

6. Agenda for the next part

In this part I have discussed:

*The nature of complexity and demonstrated why entrepreneurship is called a complex domain.

*on the features of the nature of decision making and expertise in complex domains like entrepreneurship.

*The weakness of existing thinking, particularly the dominant prescriptive models used in the domain of entrepreneurship.

Taking all of this into consideration, in the following part, I am introducing a complexity-based design solution(a framework) that embodies dispositions like ecological perspective, self-organization, evolvability, diversity, adaptability, meta-cognition, learnability, agency, praxis, feedback, etc.

Part 2 Building Of A Complexity Friendly Framework

Read further@ EsoLoop-framework.github.io

ESOLoop: An Entrepreneurship Self-Organization Framework for a Complex, Dynamic and Interconnected world