Vincent Tournierlecturer in political science at Sciences Po Grenoble and the Pacte laboratory

How would you define objectivity in the social sciences?

Objectivity can be defined quite simply by combining a negative and a positive component: in the negative, objectivity consists in distancing oneself from all forms of subjectivity, whether ideological, religious or artistic; in the positive, it implies basing one's conclusions on a coherent demonstration, based on clear hypotheses and verifiable evidence. This framework is a little crude, but it already makes it possible to rule out statements that do not fall within the realm of science. For example, inclusive writing is presented by its advocates as a tool for reducing gender inequality, but at the moment it's a militant proposition that, in my opinion, lacks solid hypotheses and verifiable data.

Does this make objectivity accessible to the social sciences? This is not certain, but we must at least distinguish two registers: the descriptive register and the explanatory register. In the first case, it is perfectly possible to be objective. There is no reason why phenomena such as socio-demographic trends, electoral results, public opinion, political party platforms, ideologies, public policies, the unfolding of crises, etc., cannot be accounted for objectively. Of course, there is always room for disagreement, as the instruments of knowledge and measurement are far from perfect, even when statistics are available (think of the divergent interpretations of delinquency). Some concepts are also more operational and consensual than others, and some phenomena are easier to document than others.

But it is above all in the explanatory register that objectivity is more delicate. The same phenomena can be seen in different ways. Even for ancient events, such as the French Revolution or the 1914 war, it is difficult to reach a consensus on the causes. The same applies to contemporary phenomena: is populism an anti-democratic movement or, on the contrary, does it express a desire to reintroduce democracy into political systems won over by elitism? Stanley Milgram may have produced highly original results from a spectacular experiment on submission to authority, but he also derived a debatable theory on the "agentic state".

Objectivity is even trickier when the issues are sensitive or divisive, on the one hand because passions come into play, and on the other because the range of interpretations overlaps with those available on the market of political ideas. Of course, a partisan interpretation is not necessarily a false one: after all, liberals who analyzed communism as a totalitarian movement were not wrong. Nevertheless, this overlapping of interpretations is a challenge for the social sciences. Durkheim had already remarked: " the role of sociology must consist in freeing us from all parties, not so much by opposing doctrines to doctrines, as by making minds contract a special attitude that science alone can give through direct contact with things ". In other words, sociology must strive to propose a distinct way of looking at things. To achieve this, reflection must focus on the analysis of causes. For example, the aim is not to justify parity in politics or gay marriage, but to understand why such reforms have been adopted. Such an objective implies restricting reflection to areas that are accessible to a minimum of documentation and for which it is possible to apply a scientific approach, i.e. based on clear and coherent hypotheses, and aimed at identifying causalities. 

Is researcher neutrality possible and desirable?

Neutrality is not always good in itself. In political science, there is a philosophical tradition that proposes normative reflection on the good and the just: the tradition bequeathed by Aristotle, Hobbes and Rousseau. This tradition, which has lost none of its legitimacy, aims less to respect epistemological neutrality than to seek answers to ethical debates, such as whether it is right to disobey the law or to give specific rights to minorities.

On the other hand, there's a type of neutrality that isn't very interesting: it's the neutrality that consists in producing a lukewarm, uninvolving discourse, or in not taking sides in an academic debate where antagonistic theses clash. Here, neutrality is more a matter of wait-and-see or personal comfort. Of course, it's not always easy to settle certain debates (for example, is immigration beneficial for the societies of origin and destination?) but, paradoxically, it's often the strongest theses that fuel scientific controversy, as in Samuel Huntington's "Clash of Civilizations", one of the most controversial books ever written.

But is neutrality possible? It all depends on what we mean by neutrality: neutrality in effects or in values? In the first case, neutrality is virtually impossible, because by definition, any public activity or intervention, even if strictly factual, has social effects. What's more, no one is immune to the risk of their work being co-opted: if a conservative party adopts a study showing that pornography has negative effects on young people, will it be deduced that the study lacks neutrality? That's the problem with using social sciences.

If we now turn to values, the answer is more complex. Of course, no one can claim to be devoid of convictions or beliefs. Everyone is the product of a history and a path. Every researcher is the bearer of a conception of the Good that influences his or her choices, and even the objects of research or the way in which he or she approaches them. If Durkheim reflected on suicide and Weber on Protestantism, it's not by chance.

But does the existence of values invalidate all research? Not necessarily, because we moderns have developed and cultivated certain skills. We are shaped by multiple identities and affiliations, we know how to empathize with others and, above all, we have received an intellectual training that provides us with knowledge and fuels our reflexivity. We are capable of stepping out of our comfort zone, following reasoned arguments and respecting tried-and-tested methodologies - all principles that enable us to keep beliefs and misconceptions at bay. What's more, there's a scholarly community that offers opportunities for exchange and confrontation, which normally helps to rule out risky theories.

But these filters don't solve everything. After the fact, we discover that scientific discourse is often shaped by the times. What's more, neutrality is not automatic; it depends very much on the circumstances and issues of the moment. Durkheim and Weber may have pleaded for the neutrality of science, but they intervened publicly in support of their respective countries during the First World War. The social sciences are also deeply imbued with a desire for reform. Many researchers aspire to change society. Studies on inequality, discrimination and participatory democracy are often driven by the desire to influence public policy. This desire for action sometimes takes a more radical turn, as there is a revolutionary tradition, bequeathed by Marxism, which intends to use the social sciences as a tool to emancipate groups considered to be dominated.

This desire for reform has always been present to a greater or lesser extent in universities, but it may be asked whether it is not stronger today due to three factors: firstly, the growing links between the social sciences and public institutions, manifested in a variety of ways (expert reports, hearings, think tanks, funding, appointments to administrative authorities, etc.); secondly, the transformation of higher education in a globalized world, one of the effects of which has been to give students a greater role, particularly in the evaluation of teaching; and thirdly, the increase in resources available to the social sciences.Secondly, the transformation of higher education in a globalized world, one of the effects of which has been to give students a greater role, particularly in the evaluation of teaching; and thirdly, the increase in the means available to academics to express their opinions, with digital technologies adding to the traditional means of petition and press gallery.

How important are methods to you as a researcher?

It's a big one, since my publications as a researcher are largely based on quantitative surveys. Since my doctorate in political science, which led me to set up a survey of 1,500 high school students and their parents with the aim of studying the transmission of political opinions, I have carried out several surveys, but above all I have been led to exploit surveys carried out by other researchers, whether local, national or international (of the EVS or ISSP type). I have also conducted interviews, but mostly in the preparatory phases of my surveys. The only time I really used interviews was for a study on Les Guignols de l'info. I also did some discourse analysis, studying how extremist parties had perceived and deciphered the "Arab Spring".

I'm convinced that quantitative data are more satisfactory than qualitative data for initiating a process of objectification. Of course, techniques such as interviews and observation are invaluable, and I always advocate them to students, as they invite us to listen to people and visualize their behavior, which is all the more necessary as elites tend to lock themselves into their certainties. But the advantage of quantitative data (which can also be derived from interviews or observations) is that they exert a very strong constraint on the researcher, reinforcing his objectivity, even if this is not a guarantee. Quantification has at least three major merits that other data do not possess: it fixes orders of magnitude, it highlights correlations and it gives the possibility of controlling the main explanatory factors.

However, I would like to add two remarks. The first is that political science is far from being reduced to the exploitation of empirical data collected from individuals. It also requires an understanding of realities that are not always accessible to social science methods. The study of elections or public policies, for example, requires us to take into account the strategies of the players involved, which in turn requires reconstruction and interpretation.

The second point is that, in addition to my work as a researcher, I'm also a teacher. In teaching, the question of methods and neutrality also arises, but in a different way. Teaching is not an exact science, especially in an IEP where general culture and topical debates play an important role, as evidenced by the Grand Oral test.

With the exception of certain courses, teachers can't claim to be experts: they rarely teach on the subject of their thesis. A course is not a research article; its aim is to transmit information, but also to stimulate reflection and open up new horizons. A course is also an encounter between people with different expectations and outlooks. A group is made up of heterogeneous elements whose interactions never produce the same effects. Each session is unique, rarely repeatable. Teaching involves a certain amount of imperfection and trial-and-error, a degree of randomness; it can take unplanned paths, be fuelled by dramatic events or provoke emotions.

When I was a student, I used to wait (often in vain) for teachers to share their own conclusions, not to ape them but to help me situate myself. Reflection is nourished by confrontation. In higher education, the teacher is not a machine, and even less a simple distributor of words. He or she must also know how to propose counter-arguments, to take the opposite view from commonly accepted ideas. If higher education ceases to be a place where dogmas and taboos are challenged, where will it be? For their part, students must learn to be disturbed. They need to accept being confronted with divergent opinions all the more because, later on, they will have to deal with discord and conflict. So we mustn't try to protect them, and even less to format them.

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?

It's a question I'm constantly asking myself, as someone who chose, some ten years ago, to work on Islam and Muslims in France, both in my teaching and in my research. Is it possible to take a cold approach to such a hot topic? If by "coldly" we mean an analysis devoid of emotion, the answer is no because, like many people, I've been deeply affected by the tragedies that have struck our country since 2012. I'm also very interested in the tensions surrounding Islam, which are immense and growing, which reinforces my view that this subject is one of the great challenges of our time.

When I started working on these issues, I had the feeling that the social and political sciences were not very concerned with them, or that they were too moralistic. For example, I found that the importance of religion among young Muslims tended to be downplayed or relativized. This is what I tried to show using local or national French data. After the 2015 attacks, we saw academics adopt very committed positions. The subject became very divisive, particularly on the question of secularism and blasphemy, with a tangle of sensitive debates on immigration, minorities, gender equality, anti-Semitism.

For me as a teacher, this complicates the task, but also makes it more stimulating. In my specialized course on "Islam and the Muslim in contemporary France", I'm walking on eggshells as passions run high. How do I strike the right balance between avoiding offence and raising issues, even if it means presenting arguments that students don't like to hear? Should we avoid sensitive points or tackle them head-on, knowing that personal safety is now at risk on this subject?

Another difficulty is whether it's possible to sort out the relevant analyses from the less relevant ones. Am I myself capable of doing this? On what grounds are my analyses, or at least those that seem convincing to me, more neutral than those with which I disagree? I obviously don't have the answer, but it seems to me that the basic rules of the scientific process make it possible to establish hierarchies. Olivier Roy's highly successful thesis on the " Islamization of radicalism " is certainly very stimulating, but its flaw is that it is not based on empirical demonstration.

Similarly, on the highly sensitive issue of Islamophobia, we are struck by the lack of factual arguments, or by the presence of erroneous interpretations, as Philippe d'Iribarne has so aptly analyzed. I myself have been able to show, with the help of European Values Surveys (EVS), that France has the lowest proportion of people who refuse to have Muslim neighbors. The lack of evidence is even more glaring with the thesis of "state Islamophobia", taken up again recently by political scientist Jean-François Bayart. The question is why some academics are so receptive to such questionable analyses. Perhaps it's an illustration of Thomas Kuhn's analysis of the workings of scholarly communities and their resistance to anomalies. The social sciences, whose members are mainly drawn from the educated and urbanized sectors of society, tend to act as spokesmen for certain values. They adopt reading grids that make them resistant to discordant elements. Herein lies one of the challenges facing contemporary social sciences: how can we avoid confining our thinking? How can we guarantee the plurality and diversity of analyses, while respecting the minimal rules of scientific demonstration?