Comments on Causality in Qualitative Inquiry

For some time I’ve been persuaded by the writing and speaking of Joe Maxwell and by the writing of the philosophers of science that he quotes, as they assert that causal relationships can indeed be identified through direct observation in certain circumstances, thus contradicting Hume. And this is not cause in general but the specification of causal mechanisms whose very identification and description helps explain the causal processes that are being claimed—how and why the causes work as they do. I think Maxwell succeeds in the arguments he makes.

The remaining papers in this collection all provide compelling examples of the clarification of causal relationships in social life and in educational settings that comes from information gained by firsthand participant observation and interviewing. By these means of information collection “preponderance of evidence” is amassed—usually at molecular rather than molar levels of social process—which makes it possible to warrant causal claims plausibly and show what the specific mechanisms of cause are that are at work in the particular setting that was studied.

Let me make just two more points. The first concerns the interpretive purposes of qualitative inquiry and their relation to causal analysis. A qualitative study does not simply report observed behavior accurately. Rather it describes social action—what people do with one another in the light of the meanings they attribute to each other’s actions. Qualitative description is thus necessarily interpretive—and the aim is to produce descriptions and analyses that have “hermeneutic validity”—describing what people in a local setting do in terms that make contact with the meaning perspectives within their actions make sense, from their points of view.

In one of my early essays on qualitative research in education (Erickson, 1986, p. 127). I pointed out that a hermeneutically valid description and analysis points to meaning perspectives as causal in social life. I used the example of Joan of Arc. What “caused” her death? We could say that the most proximal cause of her death was the heat generated through the combustion of oxygen and hydrocarbons, which in turn was caused by an English soldier placing a burning torch beneath the bundles of faggots on which Joan was standing while tied to a stake. There is a valid biochemical explanation of cause that can be made.

But why was Joan tied to the stake in the first place? Explaining that is not a matter of chemistry or physics but of history—and of local social action. Without the meaning perspectives within which her execution took place (a) the political interests of the English invaders in discrediting a charismatic leader of French military resistance to their invasion, and (b) the general beliefs in witchcraft and its appropriate penalties that provided a medium for discrediting the charismatic leader, something other than burning Joan at the stake might have happened—her decapitation (a merciful death), or her incarceration, or her release on the payment of ransom. Oxygen and hydrogen molecules were involved in Joan’s death, but they don’t explain her execution in terms of social action. Rather, the causal explanations that are appropriate in Joan’s case lie in the realm of human meanings and interests. (Here I can’t resist a pun in saying that in her situation there were human “interests at stake,” quite literally. The personal was indeed very political.) And hermeneutically valid identification of those meaning perspectives is in itself a causal explanation, as it elucidates specific meaning mechanisms that influence social action to proceed in certain directions rather than others.

My second point concerns our reasons for wanting to make causal claims. If we grant that qualitative inquiry can identify cause, why would qualitative researchers want to do that? Not for the reasons that some other social researchers do. Another way to say this is to say that the basic issue isn’t causal claims in themselves, but generalization of those claims from a local setting to others. In two recent essays (Erickson, 2009, 2011) I’ve repeated a line of argument I made in earlier writing (e.g., Erickson, 1986) that qualitative inquiry is not simply a first descriptive phase in a generalizing social science that basically involves the same procedures and aims of generalizing physical science. Rather qualitative inquiry has a different aim—the close description and analysis of a particular case at hand (including the discovery of causal processes at work within the case) rather than treating the case primarily as an instance of a general class of similar cases. The aim of the case study is not generalization beyond the case but the discovery of patterns and elucidation of processes within the case; if you will, a focus on internal generalization rather than on external generalization.

By contrast, in “hard science” social inquiry the purpose of discovering cause is to enable prediction and control beyond the case at hand. The case is considered as a sample that stands for the whole population. In educational research the aim of “hard science” inquiry is that kind of prediction and control—the discovery of research-based “best practices” that can be generally mandated, exported for “high fidelity” implementation, and enforced by surveillance. And lest we think this is just a predilection of contemporary policy researchers in the current audit culture of what is called the “new public management” (Barzelay, 2001), Donmoyer’s quote from Thorndike shows how a frightening thirst for prediction and control—based on the identification of general laws in a kind of social physics—drove the very first efforts at “scientific” research in education:

Near the start of the twentieth century, E. L. Thorndike (1910) wrote the following in the lead article of the inaugural issue of The Journal of Educational Psychology:

[A] complete science of psychology would tell every fact about everyone’s intellect and character and behavior, would tell the cause of every change in human nature, would tell the result which every educational force—every act of every person that changed any other or the agent himself [or herself]—would have. It would aid us to use human beings for the world’s welfare with the same surety of result that we now have when we use falling bodies or chemical elements. In proportion as we get such a science we shall become masters [and mistresses] of our own souls as we are now masters of heat and light. Progress toward such a science is being made [p. 6, emphasis added] (Donmoyer 2012, p. 663)

“ . . we shall become masters of our own souls as we are now masters of heat and light.” So that was the ultimate aim of “scientific psychology”—research that leads, by the discovery of general laws and careful empirical monitoring, to total prediction, which in turn supports total control. Wow! If I had much hair left it would be standing on end. Fortunately the gods have punished such hubris. The Promethean aims for social inquiry as stated by Thorndike and others repeatedly fail, and that is one reason why qualitative research remains in business.

The reason that generalizing social science fails to deliver in discovering uniform laws of social process is that local circumstances differ from one place to the next in small ways that sometimes make a big difference for the particular setting as a system—as a specific local ecology. Settings that on the surface may seem to be similar—cities, formal organizations (business firms, hospitals, schools), households—on the detailed inspection that is provided by the participant observation and interviewing that qualitative researchers do, can be seen to differ both in the local arrangements of material particulars and in the meaning perspectives taken by local social actors. This is why the aphorism attributed to former Speaker of the United States House of Representatives, Tip O’Neil, holds that “all politics is local.” And it is also the reason that Mark Twain is said to have said, “History doesn’t repeat itself, but sometimes it rhymes.” Another way to say this is that while certain causal processes may be at work in a local setting, the specific causal mechanisms in operation there may manifest differently in another setting, depending on the local social and cultural ecology of each. Thus “what works” is a matter of local specificity, not something that necessarily generalizes from one setting to the next. In actuality “best practices” do not travel well, although they may sometimes rhyme.

School administrators and teachers—and the parents of school children—know this. They are aware that instruction doesn’t happen in quite the same ways in Ms. Jones’s first-grade classroom as it does in Ms. Smith’s first-grade classroom right across the hall in the same school building. There is a differing local social and cultural ecology constituting the learning environments in the two rooms. And savvy parents with a sense of entitlement will fight to get their kids placed in Ms. Smith’s classroom rather than in Ms. Jones’s.

An argument along similar lines has been elegantly stated recently by a Danish professor of urban planning, Bent Flyvbjerg (2001, 2006). He says that although policy makers think they want general knowledge that will support prediction and control, what they actually need to inform policy making is detailed particular knowledge of the decision situation at hand; knowledge that will enable them to make wise decisions. He notes that Aristotle pointed this out in the Nichomachean Ethics (Aristotle, 1934) as he distinguished between general knowledge (e.g., in mathematics—what we would call “scientific” knowledge), which he called episteme, and the knowledge of the craftsman, which he called techne, and the knowledge that inform the practical choices made by rulers, which he called phronesis. The latter, Aristotle claimed, was a virtue because it led to wise choices of appropriate actions that actually benefited people in society. In translation into Latin, phronesis was called prudentia. Together with justice, temperance, and courage, prudence was considered to be one of the “cardinal virtues”—and it was seen as being the first among them.

Good case study, Flyvbjerg claims, is what supports phronesis. He uses as a key example the planning of new parking arrangements and pedestrian walkways in the center of the Danish city of Aalborg. It is an understanding of local particulars that will enable an appropriate, prudent plan for this. You don’t plan well for Aalborg by looking at Minneapolis, Tokyo, Rio de Janeiro, and London and summing across them to discover parking and pedestrian accommodation solutions in general. You do right for Aalborg by paying close attention to Aalborg itself.

It follows that in educational research, the purpose of using qualitative methods to discover cause is to discover how specific causal mechanisms are operating in specific local situations. This can inform prudent judgments by educators. Educational research that tries to be social physics repeatedly fails to deliver in its attempts to identify “best practices” that will “work” universally. This is because local circumstances differ—we don’t teach people in general, we teach specific people in particular circumstances. Thus educational practices that are situationally appropriate need to be built locally, chosen and tried out in phronetically informed local social action.

The papers collected here make a strong case for this. Cause can indeed be discovered by using qualitative methods, and it is cause that manifests in particular ways in a local situation. Discovering the particulars of what local social actors are doing and what the meaning perspectives are that inform and frame their doings can support decision-making that is prudential. Even Joan of Arc could have benefited from the usable knowledge produced by qualitative causal analysis in a case study of the local circumstances in which she found herself: