Tsang and Ellsaesser 2011: How Contrastive Explanation Facilitates Theory Building
Rationale and background: This paper is in response to the Academy of Management Review call for papers on new kinds of organizational theorizing. It appears in the special issue section called the “Forum for New Theory Development” in issue 2011 36:2. Here Tsang and Ellsaesser present the theory of contrastive explanation as a tool to refine theory development.
Objectives: Tsang and Ellsaesser introduce the theory of contrastive explanation into management theorizing and present it as a tool for framing appropriate research questions that can be investigated and answered with solid theory. They present eight heuristics for applying contrastive explanations.
Theoretical background: It is unclear whether or not contrastive explanation is Tsang and Ellsaesser’s own invention. Most of the article gives this impression, but the conclusion says, “Discussions of contrastive explanation, mostly in philosophy, have focused on evaluating its potential for explicating the nature of explanation relative to some existing theories, such as the deductive nomological model of Hempel (1965).” This seems to imply that this is a general technique in logic that they are introducing to management theorizing.
Methodology: This methodology guide (for refining research questions) describes the concept (contrastive explanation) and then provides guidelines on how apply it, with concrete examples given in the context of developing research questions for corporate philanthropy.
Key findings: A contrastive question is one that is presented in the form: “Why P and not Q?” That is, why is P true rather than Q? Such questions have allomorphs, which are identical versions of the question that differ only in what aspect of the question is being investigated. P is called the fact, and not-Q is called the foil, that is being contrasted with P. For such a question to be answered, there must be a difference condition; that is, P and not-Q should have some common causes, but must also have some cause(s) that differ between the two results.
Tsang and Ellsaesser use the development of transaction cost economics as an example where contrastive explanations have been implicitly or explicitly used to refine the theory’s research questions. They also use organizational ecology as a counterexample where the lack of well-formed contrastive questions resulted in very murky research questions.
Tsang and Ellsaesser present eight heuristics for developing good contrastive research questions:
1. “For a contrastive question of the form ‘Why P rather than Q?’ both P and the negation of Q are true statements.”
2. “The fact and the foil of a contrastive question have similar causal histories.”
3. “It is preferable to have a fact and a foil at the same level of analysis.”
4. “Make a contrast explicit.”
5. “Work out a fact and a foil that are interesting to the audience and whose differences are not already fully explained by existing theories.”
6. “Generate allomorphs by shifting emphasis to different words or phrases in the statement that describes a phenomenon.”
7. “Generate contrastive questions by analyzing the contrastive focus of an allomorph.”
8. “Use the difference condition to compare the causal histories of a fact and a foil, and identify relevant factors that explain the related contrast.”
Key implications: Tsang and Ellsaesser noted three key contributions of their approach:
1. Using contrastive explanation is extremely useful in theory creation, otherwise the theoretician or other researchers might be confused as to what aspect of the research question is really being investigated.
2. Contrastive explanation helps develop broader theories, that is, theories that explain a broader range of phenomena, because the explanations being contrasted can be tuned to be broader.
3. Contrastive explanation helps develop deeper theories, that is, theories that explain a phenomenon more thoroughly, because the fundamental causes are explained more thoroughly. This corresponds closely to critical realism’s search for real mechanisms; Tsang is in fact a critical realist, as noted in another article of his that I have summarized.
Comments: I find Tsang and Ellsaesser’s contrastive explanation approach an intriguing and hopefully valuable technique for framing research questions in a way that would help to generate interesting and meaningful theories. I also appreciate their critical realist undercurrent—this way of framing questions can help sound out horizontal and vertical explanations of real mechanisms. I hope to apply this in my own theory development.