Séverine LouvelProfessor of Sociology at Sciences Po Grenoble and at the Pacte laboratory, @LouvelSeverine
How would you define objectivity in the social sciences?
For me, objectivity means that researchers have to explain how they use a set of data to construct scientific knowledge, the validity of which can be discussed collectively.
The points common to the scientific approach here go far beyond the specificities of the ways in which research is carried out in a given discipline or social science speciality. Any production of knowledge is linked to a scientific position: both a theoretical position (inscribed in the history of the discipline, in a school of thought) and a methodological device implemented to confirm or invalidate previous analyses, propose hypotheses, etc. Academics sometimes emphasize the differences in positioning, depending on the discipline. However, there are also major differences between fields within the same discipline.
These different positions correspond to different approaches to objectivity in science, and different definitions of the validity of scientific analysis. We shouldn't be too hasty in opposing the so-called hard sciences (in which researchers can more easily agree on descriptions of the world they study and on ways of establishing scientific facts) to the social sciences (which produce incommensurable or incompatible analyses of social worlds). Such an opposition - which very schematically overlaps with a divide between "realist" and "constructivist" or "interpretive" approaches to science - is criticized in both the so-called hard sciences and the social sciences.
What's more, this opposition does not take into account the internal heterogeneity of the social sciences, nor the coexistence of different ways of defining objectivity in the "hard" sciences, depending on the theoretical and methodological apparatus of the research specialties. I'll illustrate this point with a joke I've heard hard-science colleagues tell at mainstream scientific events(more examples here): "A mathematician, a physicist and an engineer are traveling through Scotland and see a black sheep through the train window. "Aha," says the engineer, "I see Scottish sheep are black." "Hmm," says the physicist, "you mean some Scottish sheep are black." "No," said the mathematician, "all we know is that there's at least one sheep in Scotland, and at least one side of that sheep is black!"".
Whatever their discipline, researchers need to make their theoretical positions and methodological choices explicit: not only in scientific articles (read mainly by their peers), but also when they present their research publicly, and last but not least when they take a public stand.
Presenting one's frameworks of analysis, the conduct of research and the limits of the results obtained can be a challenge given the formats of mainstream communication, particularly in the media. And yet, particularly in the social sciences, presenting the relationship to objectivity associated with one's scientific approach is a useful way of distinguishing scientific analysis from other discourses on the social world (e.g., essays or testimonials). It also makes it possible to position oneself not as a spokesperson for the social sciences (a perilous undertaking if ever there was one), but as the producer of a discourse and analysis situated within them.
Is researcher neutrality possible and desirable?
It is not only possible for researchers to take a stance in social debates, to formulate reformulations, to draw attention to problems, but also to meet strong social expectations of science. Those who do so must be careful to distinguish between what they say as committed citizens and what they say on the basis of their scientific position. It should not be assumed that, in the social sciences, such positions would often have difficulty separating themselves from biases (linked, for example, to political opinions, socio-cultural prejudices, or even interests), whereas, in the other sciences, scientists would easily keep values at a safe distance from facts.
Conversely, the sociology of science provides numerous examples of research in the so-called "hard" sciences that respect the canons of scientific objectivity, but whose approach is nonetheless imbued with bias. Let me take an example from the biomedical sciences. For a long time, the study of parental influences on the development and health of young children focused almost exclusively on the mother. It's only recently that research has turned its attention to the role of fathers. Why this imbalance? Sociologists of science have shown that maternal influences, some of which have been scientifically demonstrated (in particular, the role of diet and smoking during pregnancy), have been widely explored in relation to a multitude of factors (among others, stress levels, "maternal care"), because of the supposedly preponderant role of the perinatal period (pregnancy and the first two years of life) on children's health, and the preponderant role of mothers during this period. Epidemiological studies have been carried out on mothers and young children, but no comparable data have been collected on fathers or older children. Mechanically, they have highlighted maternal influences, justifying further research on mothers, and making paternal influences even more invisible.
This example shows how initial biases lead to choices of methods that reinforce them, since there is no data to challenge them or qualify the scope of the results obtained. It also suggests that the strength of these biases stems from a certain scientific conformism that can be explained by the ways in which knowledge is validated (knowledge that confirms the existing state of the art tends to be more ultimately considered valid, as shown by studies on confirmation bias in research) as well as by the organization and funding of research (it is more difficult to obtain funding for "breakthrough" research considered to be riskier).
Finally, this example shows that the scientists who present their studies and draw recommendations for public action from them are not always aware that their research has been guided by certain biases, and that they are not in a position to talk about it to indicate its scope and limits. Are the social sciences more at ease with this exercise? Researchers in these fields learn to adopt a reflexive stance on their subjects as early as the doctorate. They are sometimes called upon to call to order their colleagues in the hard sciences, who are said to practice "science without conscience". Yet, as François Thoreau and Vinciane Despret remind us, reflexivity is not a virtue possessed by the social sciences; it's a skill that researchers forge through experience, but which is never fully acquired.
How important are methods to you as a researcher?
Methods are essential to my approach: it's a question of defining and using a methodological device for collecting and analyzing research data, and linking it to epistemological reflection (which addresses questions such as: what kind of knowledge about the social world do my empirical materials and my method of analysis enable me to produce? Symmetrically, what can I not conclude on this basis?)
Sociologists of science who have observed the work of scientists have noted a discrepancy between the "cooking" that goes on behind the scenes (in laboratories, researchers spend their days discussing protocols, data and analyses) and scientific articles, which tend to erase the rough edges of methodological discussions in order to present validated "results".
I readily identify with the first part of their observations. A very important part of my research activities consists precisely in reflecting on my protocols, data and analyses: whether it's the work I've done on my own, or the collaborative projects I've carried out with colleagues, or the work I've done mentoring young researchers.
Methods are also an essential component of my "lifelong learning" in research, and they accompany and even drive the evolution of my objects of study. Over the past 10 years, I have broadened the range of methods I use, adding scientometrics (the statistical analysis of scientific publications) and lexical statistics (the statistical analysis of the lexicon contained in documents) to in-depth semi-structured interviews and ethnographic observations. This broadening has enabled me to tackle new research questions: for example, I have developed work on the public circulation of scientific promises, thanks to a lexical analysis of Internet and media content relating to the "promising" field of epigenetics.
But discovering new methods also led me to take a different approach to scientific issues I was already working on. In particular, scientometrics provided me with new perspectives, complementary to those offered by interviews, on the structuring of scientific fields and the knowledge policies at work within them (the strategies deployed by scientists, institutions and funders to define the knowledge that counts and distribute material and symbolic resources).
As in other disciplines, some social science publications tend to gloss over methodological elements a little (too) quickly. I regret that, in some articles, the presentation of the corpus and the method of analysis is relegated to a footnote, or at best to a short box. Social science journals have a role to play here, and they tend, especially in the English-speaking world, to require authors to say more about the behind-the-scenes aspects of their work, describing their methods in greater detail.
This development is in line with what has been happening in a number of disciplines since the 2010s: prompted by a lively debate on the poor reproducibility of experimental results (in chemistry, biology, cognitive psychology -a cognitive science- or social psychology -a social science-...), journals have established "minimum reporting guidelines", in the form of checklists of information to be included in a methodological appendix, adapted to the methods used. This initiative is not particularly restrictive for scientists, but greatly facilitates discussion and criticism of published research.
Could you present an example of research, ideally from your own work, to illustrate the issues and tensions surrounding objectivity and neutrality in the social sciences?
I'm not going to talk about my own work here, but rather about my work evaluating scientific projects on behalf of various funding bodies (universities, national funding agencies, ministries). Given the importance of project-based research funding, this type of evaluation often occupies a significant place in researchers' agendas. I have sat on interdisciplinary committees (social sciences, biological and medical sciences and/or mathematics and computer science) that evaluate projects with a disciplinary focus and an interdisciplinary component. The job of these committees is to gather three or four internal and external expert opinions and synthesize them to give an overall opinion on the project and recommend (or not) its funding.
For me, these committees were a privileged place for observing how researchers from different disciplinary backgrounds agree on the scientific quality of a project. On committees with a very broad spectrum of interdisciplinarity, I found again the analyses of G. Mallard, M. Lamont and J. Guetzkow, concerning the difficulties that run through the work of interdisciplinary committees within the social sciences: in particular, the strong disparity between the judgment criteria of the members, the difficulty in knowing to what extent to trust experts who are not from one's own field to evaluate projects objectively (and not to seek to privilege one's own discipline). How, in spite of everything, do the members of these committees agree on what defines a good project? They adopt broad evaluation criteria (For example, is the methodology detailed? Are the research objectives clear?), compatible with a wide variety of scientific approaches. In this context, the issues and tensions relating to scientific objectivity, and the responses to them, do not seem to me to differ fundamentally between the social sciences and other disciplines.