What Theory Is Not

Critical Analysis

In this text, the authors, observing that much of academic theory is poor, offer five examples of what does not “count” as theory.

The authors first note that there is a broad lack of agreement about what a theory is, making it complicated for authors to develop strong theories. For example, there is a “lack of agreement about whether a model and a theory can be distinguished” and “whether the strength of a theory depends on how interesting it is.” They explain that here, they will not offer a “magic” way to construct theories, and instead, will offer examples of how not to construct theories.

A list of references is not a theory, because no logic is presented. Often, when researchers are asked to provide theory, they simply add more citations rather than connecting the citations to a logical argument. This argument should explain why the theories led to an author’s predictions. Authors should “explicate which concepts and casual arguments are adopted from cited sources and how they are linked to the theory being developed or tested,” and logic from past work (instead of simply references to past work) must be included.

Data alone is also not theory. Data describes what patterns were observed; theory explains why those patterns were observed. Quotes from qualitative research “may get a bit closer to the underlying causal forces” but this is still data without theory. Data does not generate theory. Researchers generate theory from data.

Lists of variables are not theory, because theory must explain where these variables come from, or why they are connected to each other.

The authors argue that diagrams are not theory; they “can be a valuable part of a research paper but also, by themselves, rarely constitute theory.” Some diagrams are lists of categories, and these are not helpful. More helpful are diagrams that show causal relationships or logic flows. Time-based diagrams are also valuable. The authors note that “On occasion, diagrams can be a useful aid in building theory,” particularly for researchers who are not good writers. But diagrams themselves are “stage props rather than performance itself.” This is because they don’t explain why a connection will be observed; “some verbal explication is almost always necessary” because the underlying logic needs to be spelled out. Explanation is “difficult to replace by references to a diagram.”

Hypotheses are also not theory. They allow for a bridge between data and theory, but they don’t include logical arguments.

Strong theory answers the question “why.” It is about connections; it is an explanatory story that emphasizes causal relationships, sequencing, and timing. It examines processes and subprocesses, and “burrows” in and around social phenomenon. A good theory “explains, predicts, and delights.”

Research Value

The value of this work in informing my own research is that it:

• It identifies qualities of valuable theories as arguing a causal relationship between data and theory, answering the question “why,” and telling a story about relationships, sequence, processes, subprocesses, and timing
• It offers a provocation around the value of diagrams (one which I disagree with)