April 26, 2019, by Brigitte Nerlich

Comments on qualitative methods in the humanities and social sciences

I while ago Chris Toumey (Toumey@mailbox.sc.edu) wrote a guest post for this blog, announcing his new book on nanotech and the humanities. A reader had a question that Chris didn’t have time to answer at the time. This post is an answer to that question.


Regarding my book on nanotechnology, Nanotech and the Humanities, a friendly astrophysicist asked whether persons in Science & Technology Studies [STS] did not have research methods. The short answer is yes, we do have research methods, but ours tend to be very different from those of hypothesis-testing experiments in the natural sciences. While persons in STS sometimes conduct hypothesis-testing experiments, our research is grounded more in the humanities. In fact I appreciate the relationship in German education and scholarship between the Geisteswissenschaften and the Naturwissenschaften. Both are considered equally valid, even though they have different methods for discovering and developing knowledge. With that in mind, here is my appreciation of the research methods of the humanities, including my home discipline of cultural anthropology.

A historical perspective

First, we often assemble a historical perspective. This addresses three questions. How does a historical situation affect a form of policy, behavior, or understanding? This can include a good biography which shows how unique historical circumstances shape a person’s values and actions. If some persons in the past were very influential for instigating changes, especially in science and technology, we want to know why and how the contemporaneous conditions in which they lived contributed to shaping their contributions to change.

A historical perspective also explains how and why things change.

The third benefit of a historical perspective is to separate the inevitable from the contingent. Often when we are accustomed to the consequences of a past event, we might think that those consequences were inevitable. But a good historical perspective often shows that they were not.

In my work on nanotechnology, my favorite example of this is Cyrus Mody’s social history of the invention and adoption of the STM and the AFM. There the author describes certain special conditions at IBM which enabled Binnig and Rohrer to invent the STM. Furthermore, IBM and Bell Labs once had a corporate culture in which they supported numerous post-docs, and the post-docs had considerable intellectual freedom. They became the principal advocates for adopting and using that instrument. Subsequently, after the AFM was invented, a small number of scientists in California began to use the AFM in ambient conditions, which was previously considered bad science. Corporate culture has changed since the 1980’s, so that one might see that the invention and the adaptation of the STM and the AFM benefited from unique conditions. Those events were not inevitable. They were contingent. And if someone believes that the invention of the STM was inevitable, they should become familiar with the earlier invention of the topographiner, and its unfortunate demise.

Or consider the case of Picasso and other cubists who were inspired by Metzger, Cezanne, and others. The earlier artists hoped to use two-dimensional paintings to present multiple perspectives which were not limited to the three classical parameters of length, width, and depth. Picasso, Braque, and others learned from them, and then invented cubism to realize those hopes. One might say that there would not have been twentieth-century cubism without its late-nineteenth-century predecessors.

Comparative case studies

Secondly, we try to make good use of comparative case studies. These sometimes involve quantitative methods, but not necessarily, because many comparisons amount to apples-and-oranges. In nanotechnology, for example, one might say that physicists see nanotech one way, and chemists see it a different way. And within chemistry, scientists in organic chemistry may think differently from scientists in inorganic chemistry.

There are also comparisons between national scientific cultures. Consider Medicine and Culture, by Lynne Payer. This author was a medical journalist, originally from the States, who worked in Europe. When she had a certain medical condition, she discovered that the diagnosis was more or less the same in France, West Germany, the U.K., and the States. But the therapeutic strategies differed among the four nations. This was not because one country was more scientific than another, but because the four medical cultures had different values and different philosophies of medicine and science.

An ethnographic approach

Also in qualitative research methods, we often make good use of ethnographic practices, e.g., face-to-face interviews, focus groups, and participant observation. In the latter practice, the ethnographer participates in the life of the people that she or he is studying, and so has a close-up view of their behavior and the values which guide their behavior. This is extremely valuable for understanding other cultures. Ethnographic work sometimes uses quantitative analysis, but not necessarily.

Thick description

The fourth foundation stone of qualitative methods is called thick description. Events, behavior, values and other considerations are described in great detail so as to capture and present all the nuance and context which helps one understand what is going on. When thick description is done well, it reads like a good novel. Except that it is non-fiction. This is too much for some readers to appreciate: it lacks the spare logic of problem-hypothesis-experiment-conclusion. But when it is necessary to understand human events, behavior, and the values which guide them, thick description is the way to capture the nuances and context which connect those phenomena of human behavior.

The work of Clifford Geertz is the gold standard for thick description, but one can find this approach in many parts of the humanities. In my own case, I explored the question of whether Richard Feynman’s 1959 talk, “There’s Plenty of Room at the Bottom”, was the origin of nanotechnology. My paper required a great deal of nuance and interpretation, and I could not answer that question without writing thick description.

Mixing methods

I should also say that qualitative research methods can sometimes articulate with quantitative research methods. In the social sciences, there are historical, comparative, and ethnographic methods, plus thick description, which can help the quantitative social sciences, e.g., sociology, political science, and economics, to formulate testable hypotheses about human behavior. When the two approaches are synthesized, this is social science at its best.

And finally

Finally, it should be mentioned that not all science is quantitative hypothesis-testing.

Stephen Jay Gould’s 1986 paper, “Evolution and the Triumph of Homology”, explains that one ordinarily imagines that “science” consists of hypothesis-testing experiments, based on the well-established understanding of the Scientific Method. But Gould then reminds the reader that there are the historical sciences, e.g., paleontology and historical geology, co-existing with experimental sciences. The historical sciences have different research methods, and different standards. Those standards, according to Gould, are anatomical homologies [especially the forelimb homology of mammal skeletons], seriation [i.e., patterns in which organisms undergo observable changes, usually from simple to complex], and vestigial structures.

This last standard indicates that anatomical parts which are anomalous in the present, e.g., the human appendix or the panda’s “thumb”, are evidence that conditions in the past were different. And these standards of the historical sciences do not necessarily answer to quantitative hypothesis-testing because the scientists cannot go back into the past to manipulate variables. It is satisfying indeed when hypothesis-testing science contributes to the historical sciences, but that is not their primary method.

Considering that our friend who asked about research methods in the humanities is an astrophysicist, I would like to mention that some of the work of astrophysics may seem more like the historical sciences than the experimental sciences: there are limits to which astrophysicists can manipulate an independent variable in outer space for the purpose of discovering how that affects a dependent variable. If there is an anomaly in the orbit of a planet or asteroid, then this can lead to a hypothesis that there is another body which causes the anomaly. And that hypothesis can be tested. But when the age or distance of a star is derived from its position on the red shift, then this is a form of seriation, or a style of reasoning very close to seriation.

Because there are limits to which astrophysicists can manipulate an independent variable in outer space for the purpose of discovering how that affects a dependent variable, it must be recognized that dating a star by using its red-shift position is much closer to the historical sciences than to the experimental sciences.

I hope this helps.

Posted in Social sciencesociology