Pierre Bréchonemeritus professor of political science at Sciences Po Grenoble and at the Pacte laboratory

How would you define objectivity in political science?

The social sciences must try to understand and explain what societies are. These sciences, sometimes referred to as soft sciences, are in any case plural. Each discipline develops a particular view of society. Economics seeks to explain the social through economic phenomena, while the political scientist understands them through political structures, actions and ideologies. Just as the sociologist understands a phenomenon through all the social relations that help produce it. The label of sociologist should be reserved for empirical approaches, in search of evidence or at least relevant explanatory clues. Edgar Morin, for example, is an impressive and highly cultured thinker. Today, he is developing an ideal of savoir-vivre and bien-vivre ensemble, in a society whose dynamism is born of the contradictions and conflicts that take place within it. We can be captivated by this thinking, but it no longer properly belongs in the realm of sociology. Even if several of Edgar Morin's earlier writings have become classics of sociology, such as La rumeur d'Orléans, of ethnography, such as La métamorphose de Plozévet, or of social anthropology, such as L'Homme et la Mort.

Objectivity is an objective - difficult to achieve - that consists in analyzing social facts using a methodology that is as appropriate as possible for understanding and explaining phenomena. Objectivity cannot be achieved without recourse to methodologies that can be replicated by others, and which may be quantitative or qualitative. Adapted methodologies are not exactly the same from one discipline to another, but can be classified into four main types: observation (participatory or not) by ethnologists; surveys (quantitative or qualitative) by political scientists and sociologists; trace analysis, the preferred tool of archaeologists and historians; and laboratory experimentation, favoured by psychologists.

Certain methodologies make it easier to achieve objectivity, when they do not require subjective interpretation by the researcher. For example, when I use a qualitative methodology, such as non-directive interviews, I have to interpret what is said, not just what is said explicitly, but also what is not said, what is implicit. There is then a risk of over-interpretation, which may be linked to my own frame of mind and values.

When I conduct a quantitative questionnaire survey, there's less risk of subjective slippage. If I adopt good business practice in formulating questions, they must not encourage a particular type of response. The answers obtained may be biased, but they have a certain objectivity, as they are collected from representative samples of the population under similar conditions. These responses can be cross-referenced with many other variables to uncover correlations that show the majority logics of opinion. And I can eventually look beyond the correlations to see if an explanatory causality can be inferred. This causality is often not completely proven, but only plausible.

Thus, in all surveys on xenophobia, a sensitive subject, we observe that the level of education is the most explanatory "heavy" variable (i.e. of a socio-economic type) when we carry out a regression analysis "all other things being equal". This tells us with certainty that one is much more "likely" to be xenophobic at the bottom of the knowledge ladder than at the top. But of course, this is not the only variable influencing the level of xenophobia. People who are part of a left-wing culture and have a low level of education may well not be xenophobic, their political values compensating in some way for their handicap in terms of knowledge and openness to the world.

Is researcher neutrality possible and desirable?

The social sciences are not designed to say what is right. They are not a political and social philosophy. As mentioned above, objectivity is an elusive goal, but a necessary one if we are to truly understand and explain. All too often, in the social sciences, analysis veers towards hagiography or denunciation. In fact, denunciation is now more common than incensing, due to the evolution of our culture, which is much less conformist than it used to be. This means that researchers have frames of thought that are dependent on those of their society, and that they must sometimes be wary of. This is particularly important in relation to the media, which broadcast a universe of thought marked by media zooms: the exceptional becomes the significant fact, often forgetting past history where similar phenomena have already taken place. The occurrence of a few attacks is seen as proof that society as a whole, or a particular group, is moving towards violence. Surveys, however, do not necessarily confirm this type of evolution in the population as a whole.

Some researchers claim to have a critical and denunciatory eye, which sometimes leads them to try and prove the validity of their political values through questionable social science analyses. The researcher then becomes the one who wants to give the right diagnosis on the state of society, and above all on what's going wrong, to then explain what needs to be changed.

But even when you want to respect the neutrality of your objective viewpoint and leave it to others to define the necessary change and its implementation, this posture is not easy to maintain. For example, when a researcher is interviewed by a journalist, the pressure is on to depart from neutrality. The journalist, like his audience, is interested in what's going to happen. They'd like to be able to anticipate the future and try to get the researcher to tell them what developments are coming. They are also looking for solutions to problems, transforming the researcher into an expert at the service of public policy implementation. 

When questioned about democratic values or the ideals of an authoritarian political system, it's difficult to remain neutral, because this neutrality, which enables us to understand and explain, is not easily accepted by the journalist, who expects us to take a clear stand! Sometimes you have to have the courage to disappoint them, and your own friends too, who don't understand the schizophrenia between the researcher's social science analyses and his commitments as a citizen.

There are few fields in which the social sciences can truly predict the future. They can predict changes in values due to generational turnover. But they cannot predict period effects linked to unforeseen events, and they have no control over trend reversals, some of which are probably improbable but not impossible. 

How important are methods to you as a researcher?

It's already clear that I attach a great deal of importance to social science methods. I taught them throughout my 41 years at Sciences po Grenoble. Since the 1970s, with a whole team of colleagues, we have favored a practical approach to methods, rather than a grand epistemological discourse.

We focused on learning quantitative and qualitative survey methods, having students conduct qualitative interviews and try their hand at sociological analysis. We also developed quantitative questionnaires, administered them to representative samples of the Grenoble conurbation, and produced survey reports with the students, analyzing the results and their rationale. Sometimes there was a client, a media outlet or a local authority that could, beyond the publication of the results, bring about a change in public action. We summarized this practical teaching in a manual.

What fascinates me is to observe, through surveys in all European countries, differences in values among populations, which remain very strong, despite the construction of Europe. And obviously, it's not enough to show that values remain different, we have to try and understand where the differences come from. Values are evolving rather slowly, with all countries moving in the same direction, towards greater individualization, i.e. individual autonomy. The pace of change is not the same, and there are major differences between traditional countries where individualization remains very low, and more modern countries where it is much more in demand. The religious matrix of cultures is probably an important factor in explaining these differences in values. But so is the level of economic development.

This type of survey shows that comparisons are very important for understanding and explaining how societies are structured and transformed. We can only fully understand a country by comparing it to others. Any analysis that is too hexagonal runs the risk of leading to erroneous interpretations. We sometimes hear it said that the French don't much like political parties, but they are very active in associations. When we look at the results of comparative European surveys, we see that while the French are indeed not very politicized, they are not characterized by a high level of associative participation either. In both areas, it is in Germany and the Scandinavian countries that we find a high level of political and social involvement. This level is low in Southern and Eastern Europe.

Could you present an example of research, ideally from your own work, to illustrate the issues and tensions surrounding objectivity and neutrality in SHS/social sciences/political science?

I've already mentioned several examples of the tensions surrounding objectivity in the social sciences. Let me take one more example, that of the Eurobarometer surveys carried out by the European Commission. They were set up in 1974, on the initiative of a "militant civil servant" for Europe, Jacques-René Rabier, ex-director of information at the EEC. From the start of his career in European institutions, he had been convinced that Europe could not be built by elites alone. From his point of view, it was becoming increasingly necessary to measure opinions and values about Europe, so that the state of European feelings in the population could be known by all, and so that political and social players could seize on the results to adapt an action strategy. It was also necessary to be able to measure the effect of the policies followed by Brussels, and the initiators deemed it necessary to carry out two surveys a year with a certain number of "trend" questions, i.e. identical from one wave to the next, in order to measure evolutions.

Created by the driving forces behind the construction of Europe, the aim was to measure objectively, hence the development of numerous survey questions to study the different facets of people's relationship with Europe. At the time, several international specialists in questionnaire surveys were called in to help steer the survey (Ronald Inglehart, Jean Stoetzel, Elisabeth Noelle-Neumann...). Some drift occurred, particularly in the 1980s and 1990s, when anti-European sentiment grew in many European countries. Some European leaders may have been tempted to suppress the Eurobarometer so that it would no longer show bad figures (a strategy based on breaking the thermometer to stop measuring the patient's temperature). The questions were sometimes changed, breaking up the series and making it impossible to see trends. Above all, the questions were multiplied, somewhat obsessively, in the hope of tracking down European sentiment where it might otherwise have been masked. 

All in all, the tensions surrounding the survey have been numerous, but the tool has been maintained, and the countries taken into account are now very numerous, not only the members of the Union but also a number of candidate countries. Despite all the ups and downs, and sometimes questionable interpretations of the results by pro- or anti-Europeans, an enormous comparative database has been built up, which in the final analysis reflects the European panorama fairly well, with all its tensions and fractures.