Realism,Empiricism,and Fetishism in the Study of Entrepreneurship

There have been persistent attempts to portray arguments about the ontological nature of opportunities as a core issue for building theory about entrepreneurship. A “discovery” view of opportunities as “objective phenomena that are not known to all parties at all times” (Shane & Venkataraman, 2000, p. 220) is contrasted with a “creation” view of opportunities as “created endogenously by the actions of entrepreneurs that seek to exploit them” (Alvarez, Barney, & Anderson, 2013, p. 303). The unsettled debate between these perspectives has become taken for granted and morphed into a philosophical discourse around the meaning of “objective.” A recent critique suggests that opportunities are not ontologically objective, but “merely” epistemologically so due to the impossibility of the former claim to “make contact with the world somewhere, typically by means of a test, however indirect” (Alvarez, Barney, McBride, & Wuebker, 2014, p. 228).

As tempting as it is to indulge in such debate, we argue that this dialogue does not help scholars build falsifiable theory and, moreover, amounts to theoretical “fetishism” that is detached from empirical reality. We appreciate that behind the dialogue lies a motivation to make research on entrepreneurship and entrepreneurial opportunities more theoretically robust. Indeed, it is important for scholars to understand their own philosophical lens and be aware of its limitations. Realism calls for explanations of entrepreneurial phenomena, whether (a) entrepreneurially relevant phenomena ontologically exist before an entrepreneur creates a new idea or product, (b) an entrepreneur first creates an idea that later becomes ontologically real, or (c) elements of both occur. Needless to say, either of the processes that lead to (a) the discovery of ontologically existing phenomena that allow a new idea to become reality or (b) the creation of an ontologically new entity that did not exist until an entrepreneur created it needs to be understood via effective entrepreneurial research. In fact, to be considered effective and valid, entrepreneurship researchers’ methods should offer clear evidence of sensitivity to all three kinds of phenomena: discovery, creation, and/or both.

However, we question whether the adoption of a creation view makes any difference. Both the discovery and creation views identify opportunities as something that can be specified (referred to) and simply vary in their focus: The former view focuses on its final state, the latter view on the process through which this state comes about. But just as it is easy to challenge a discovery theorist on whether what one pursues is really an opportunity, so it is easy to challenge a creation theorist on whether someone who acts and interacts is really an entrepreneur. If entrepreneurship is defined as the pursuit of opportunities (Shane & Venkataraman, 2000), then knowing that something is an opportunity would make the person pursuing it an entrepreneur, and vice versa, knowing that someone is an entrepreneur would make what they are pursuing an opportunity. However, because we agree with Gartner (1988) that “who is an entrepreneur?” is the wrong question, so is the question of “what is an opportunity?” For entrepreneurship, a domain struggling for legitimacy and practitioner relevance (Shane, 2012), we feel that focusing on these quarrels makes it even more difficult for the field to build theory.

The problem we see with the case for a creation view is the rationalist nature of its logic. In contrast to an empiricist (one who focuses on the collection and analysis of data to build and test theory), a rationalist embarks on theory development as a creative writing exercise, focusing efforts on “deconstructing prior theory” instead of acknowledging “challenges from contradictory phenomena” (Suddaby, 2014, p. 408). As evidence, Alvarez et al. (2014) completely dismiss decades of qualitative and quantitative empirical research about entrepreneurship when saying “ . . . opportunities have objective metrics in support of their viability ex ante the entrepreneurial action, which, of course, is false” (p. 228). The authors continue with “[objective opportunities] could mean that many others have the same venture idea, but this is false as well” (p. 228). Can two (or more) individuals have the exact same idea to open a new restaurant when they see an opportunity that has a “Business for Sale” sign in the window? Are pharmaceutical companies not racing/competing for the same opportunity to develop a cancer-curing drug? Data from more than 15 years of empirical research on representative samples of nascent entrepreneurs identify that, when founders are asked why they formed their venture, the majority reply, “I saw an opportunity” (Davidsson & Gordon, 2012).

Is this an objective metric? No doubt, it is probably a more subjective view. In essence, however, the above arguments suggest to practicing entrepreneurs who founded a business based on a recognized opportunity that they are wrong. This logic borders on theoretical “fetishism,” which Suddaby (2014) identifies as “self-absorbed” and “detached from the empirical world” (p. 34). With arguments such as these, it should be no surprise that much of our organizational research is viewed as irrelevant to practitioners (McKelvey, 2006). Indeed, if theory, as Suddaby (2014) suggests, “signals the values upon which knowledge is built” (p. 407), then the creation view of entrepreneurship theory signals that the domain does not value the empirical reality of entrepreneurs.

Is there a way to draw a meaningful demarcation between opportunities and entrepreneurs? After all, opportunities comprise elements—customers, suppliers, production, money, information, and other resources—that are both tangible and separate from the entrepreneur. Let’s consider the simple analogical example of a building. Before it is built, it exists only as a blueprint—in the mind of the builder or formally drawn, but, once built, it is separate from the builder and made of objective materials. A pile of bricks is not a building; it needs to be perceived as having a function. Similarly, a collection of people and resources is not an opportunity; they have to be arranged so as to function. The builder does not create the materials; she or he simply creates the functional relationships among them. In a like manner, an entrepreneur does not create the materials of which the opportunity is made; instead, she or he creates the functional relationships among them (e.g., production and consumption). Therefore, both buildings and opportunities can be viewed as emergent structures—the former physical, the latter social. These structures are emergent because they have ontologically real components, possess epistemologically real functional relationships, and require real human actions and interactions to come to fruition.

This interplay between components and relationships can be readily seen in the example of the iPhone, which can easily trigger arguments about whether Apple created people’s needs for the iPhone or simply discovered an already existing opportunity. Reality before the iPhone was invented and manufactured was that cellphone towers already existed; people already were using cellphones; people could connect with their friends, stores, restaurants, hotels, transport services, and so on, via the Internet and had access to all the other information available on the Internet; the app-making software was available before the iPhone was produced. In fact, all the technology Apple used to create the iPhone already existed in the real world before the iPhone came to market in June 2007. The story was the same with the steam engine, Ford’s Model T, Wright brothers’s airplane, the ice-cream cone, the first computer, and on and on. Apple’s success laid in making the existing ingredients work together in the most compelling way, by building strong functional relationships among them. In this sense, there is no opportunity without the relationships and, equally, no opportunity without the ingredients.

As well, these structures are emergent because the efficacy of the final product can be directly linked to the quality of the materials, the strength of the functional relationships, and the skill with and order in which the actions are completed. Consequently, for both theory and practice, arguing about the ontological nature of opportunities serves no substantive purpose—what matters are the realities that entrepreneurs face and deal with in creating their new ventures. As Dimov (2011) argues, developing theory about opportunities should focus on their substantive markers, namely articulated ideas, purposeful actions, and market relationships.

We urge an end to the current fetishism in entrepreneurship theory. Instead of creating untenable assumptions that dichotomize scholarly inquiry, we propose that the field embark on a more formal approach to building theory, one that can use the strengths of both discovery and creation assumptions, as well as variance and process perspectives, to build computational models that generate emergent outcomes in silico. In complement to mathematical proofs and formal logic—which are often based on faulty assumptions about human behavior (Read, 1990)—Adner, Pólos, Ryall, and Sorenson (2009) explain that agent-based computational simulations can be used to build more precise and falsifiable theory in a manner that is logically consistent and epistemologically agnostic, and does not require the successful emergence of “real” new ventures. In contrast to other formal methods, simulation models permit the use of repeated Monte Carlo experiments—where only one parameter of the model is changed to demonstrate cause-effect relationships. These experiments can demonstrate how micro-level interactions among heterogeneous agents might be sufficient to generate empirically observed macro-level outcomes of interest. Here, opportunities could be represented by either resources available in the environment that can be recombined (a discovery view) or emerge as hierarchies that result from agent interaction (a creation view). In simulation models such as these, transparent theoretical assumptions can satisfy the rationalists, validation of initial parameter inputs and emergent outcomes with extant data (e.g., Crawford, Aguinis, Lichtenstein, Davidsson, & McKelvey, 2015) can satisfy the empiricists, all while avoiding the fetishism that discounts the empirical reality of practicing entrepreneurs.