Yves SchemeilProfessor Emeritus of Political Science at Sciences Po Grenoble and at the Pacte laboratory, @YvesSchemeil
How would you define objectivity in the social sciences?
When we choose an object to change, how can we avoid sticking to the explanation that appeals most without being the truest? By converting the subject into an object, in other words, by "objectifying" it. Which means keeping it at a distance from yourself. The more you're seduced by a research topic, the more you have to be wary of its appeal. We must always ask ourselves whether, beyond personal satisfaction, studying it will really generate new knowledge.
It follows that, unlike objectivity, which is a stage reached at a given moment, objectification is a never-ending process that leads us to take nothing for granted. Especially not the most obvious causal relationships, those that first impose themselves on our minds. Rather than testable hypotheses, these are often unprovable postulates adopted before we even set to work.
Proponents of laws so general that they seem applicable to all cases are often content with reduced objectification. As soon as an observation seems to "prove" an all-purpose explanation, it's adopted. The global explanations I'm thinking of come from Marx, Bourdieu and Foucault. Translated into every language, put to every sauce, they are used everywhere without doubting their validity in contexts for which they were not conceived. Such artificial transposition, sometimes even purely deferential and totemic, escapes the canons of epistemology.
The lack of interest in objectification has serious effects on the quality of our research and its social utility. Firstly, we run the risk of misjudging the real origin of a phenomenon by stopping too early in the search for causes, arbitrarily deciding not to doubt our results while knowing full well that no-one will repeat our reasoning or work on the same data as we did.
Keeping at arm's length an object we have often chosen to study because we believe it to be close by means we can avoid being fooled by cause-and-effect relationships that we no longer test once our initial results confirm our conjectures. Why go further? Because we can't just assume the weight of the "usual suspects" - standard of living, level of education, family inheritance, etc. - and then go on to the next step. The "rich" who are also "dominant" only exist in library sociology. In real life, there are different forms of capital (to speak like James Coleman) and different spheres of injustice (to paraphrase Michael Walzer).
So we need to make sure we've left no stone unturned, rather than clinging to the first result that comes along because it suits us. We mustn't stop when we start to fear that, by continuing to search, we'll find one that disrupts our convictions.
Is researcher neutrality possible and desirable?
Scientists do not judge the subjects of their research, nor do they take sides with them. Weber didn't expect the same of "axiological neutrality" - a frequently heard expression that still has an effect, although nobody knows exactly what "axiological" means, or why the adjective is systematically attached to the noun.
The great Max never wrote that we could treat criminals or humanitarians with equal detachment. We don't defend an accused person regardless of the faults he or she is accused of, just out of respect for his or her rights! What we can assess is the coherence between the goals pursued and the means employed. This is why the addition of the word "axiological" is useful: taken from the Greek axios, it refers to the goals people set for their actions: terrorists aim to kill as many civilians as possible, while war medicine tries to save as many as it can - it's easy to see where our sympathies lie, but this inclination shouldn't prevent us from explaining the behaviour of unsympathetic people.
According to Weber, we should consider these individual goals without labeling them as "good" or "bad". This principled attitude is dictated by his "comprehensive sociology": he wants to know why and how we make our decisions, with the goals pursued upstream triggering our actions downstream. Many political scientists try their hand at this, brandishing Le savant et le politique at arm's length. Yet we are never in a position to judge an act without knowing the beliefs from which it stems. We can certainly praise resistance to the oppression of an illegitimate government, as in Syria or Myanmar. But that won't tell us why their leaders chose this option. Instead, we need to make sure we have a complete list of all their motives.
We must remain neutral on the goals pursued, not on the people hoping to achieve them. Weber's lesson is that the obstacle to an explanation of social phenomena is ethics (a word he always preferred to ethos).
How important are methods to you as a researcher?
A major place! I've used almost all of them, combining the qualitative and the quantitative. Determining what a "fact" is requires precise conceptualization, deep immersion and measurement using effective tools (which are constantly being invented). Starting with a research protocol and following it as far as you can is essential. Consolidating the rise to generality through case comparison is a must. I've always tried to be counter-intuitive and to test my conjectures using tools adapted to my objects. I have therefore combined analysis of primary sources done with great respect for texts, the use of figures, field surveys including data collection by selective or mass interviews and lasting ethnographic experience enabling me to control my ideas.
Let's take an example. After the massive influx of Syrian refugees to the Palestinian camps, followed by a serious explosion that killed and injured hundreds of people and left thousands homeless, and finally an interminable economic and financial crisis, how can we explain why the Lebanese political class remains in power? The usual explanations are only partly helpful: yes, class relations and domination are zero-sum games, and what some gain is lost by others; yes, the elites can't afford to leave the government, because they'd lose everything (including impunity). Enough to prove Marx and Bourdieu right, respectively.
And yet, class and place struggles alone cannot explain such a massive monopolization of resources as in Lebanon. To explain it, we need to turn to other paradigms, such as the rationalist neo-institutionalism of D. North or D. Acemoglu. Acemoglu: in a divided country, the fragmentation of support forces leaders to drain the resources their opponents would need to oust them from office. Leaders then monopolize them in order to redistribute them to their community, relatives or clientele, without counting on public services (here, cultural anthropology is invaluable). It's not enough to appropriate as many resources as possible, as predicted by Marx and Bourdieu: you also have to deprive the State and the other "pillars" of the system. Until a mechanism of progressive taxation or charitable foundations makes it possible to implement distributive justice, there can be no hope of change at the head of the country (at this point, we are witnessing the live entry of political theory, in particular that of justice - what opera buff Jean Leca has called the "entry of the duenna", always there to remind us of the norm).
A recent article nicely entitled "Bourdieu in Beirut" failed to prove the contrary. By showing that education is no guarantee of upward social mobility (driven by an entirely different kind of motor, the piston, the gift/counter-gift), the article concludes that there are no distinct "fields" - only one: the community field. It therefore concludes that it is impossible to use cultural resources to build effective strategies for overturning political domination.
The unbearable inertia of the political system is further exacerbated by heavy factors. The proportion of Lebanese residents to members of the diaspora, and the ratio of Lebanese to refugees, are out of all proportion to comparable percentages in Europe. The emergence of a true middle class at the end of the civil war was, for a time, an illusion, seemingly confirming the theory that there can be no capitalism without a petty-bourgeoisie, and no true democracy without a sincere electorate drawn from the middle categories of the population. But this was without counting on the "civilizing force of hypocrisy", for there is not a single person in the country who does not benefit from widespread clientelism. So, instead of forcing the system to open up, the transformations in Lebanon's social composition have led it to withdraw further into itself.
Could you present an example of research, ideally from your own work, to illustrate the issues and tensions surrounding objectivity and neutrality in political science?
I was seriously mistaken once, at the start of the Lebanese war in 1975. My extensive research (5,000 people interviewed, countless ethnographic surveys, long experience in the field) had convinced me that the conflict would be short-lived!
How could I have gone so far astray? The answer is simple: on the one hand, I didn't want the war to last, as I had too many links with Lebanese society; on the other, I knew much more about the so-called Christian parties than the so-called Muslim parties. Aggravatingly, I was convinced of the opposite, having been one of the few to investigate the latter. Well-informed about the military capabilities of the Maronite forces, which seemed impressive to me, while I unknowingly harbored a prejudice about the strategic incompetence of the Palestinians, born of an inability to investigate their militias, I neglected the latter. While they were training Lebanese Muslim civilians in the use of weapons, I only had reliable information on the balance of power thanks to the investigations carried out by two Master's students into the training camps of the Phalangist militias. If I had pursued my research instead of communicating provisional results to the public think-tanks in France and the United States that urged me to do so, I would soon have discovered the root cause of the war, which would determine its duration.
The cause was demographic: in the 1960s, the country experienced a brutal rural exodus that gave rise to the Beirut suburbs. The fertility of the Maronites in the mountains plummeted to ensure their children's education rather than destining them for the army or the police, religious orders or emigration. The number of Shiites leaving unproductive farming areas held steady: they remained faithful to the ideal of a large family. Eventually, the Christian majority had to disappear: that's why the 1975 war was launched in anticipation - still winnable for the Christians that year, it would inevitably have been lost twenty years later.
So much for Lebanon, where I spent a long time. As for other areas of research - Syria, where I lived, and Iraq, which I visited on several occasions - I had a much better idea of what would happen there. When, in 1976, Hafez el-Assad was passed off as the saviour of Lebanon, I predicted on the contrary that Syria would plunder it for years to come in order to make up for its lack of internal growth, itself knowingly maintained so that the newly rich would not turn against the regime. In 2011, at a time when Bashar al-Assad's days were numbered in specialized academic circles, I declared that he would win the war, a certainty born of an insider's knowledge of the upward social mobility that had enabled the less privileged, whatever their denomination, to marginalize the large urban families of Sunni notables who had until then dominated. When Saddam Hussein was given the all-clear in 1991, I told the media that he would survive the invasion of his country for at least ten years. I did so on the basis of the Baath's irrigation of Iraqi society, an explanation that came to me by comparison with the de-Sovietization underway in Europe and Russia, but also thanks to a better knowledge of inter-community relations.
The difference in explanatory performance between Lebanon, the country of my origins where I went to high school, and its two large neighbors where I had no roots, is due to a single cause: I had enough doubts about the Baathists and enough distance from their system of government to deepen my analyses of their resilience, my proximity to both their societies sparing me any misinterpretation of events. In Lebanon, on the other hand, my reading capacity was limited by an excess of knowledge, a touch of optimism and a lack of neutrality. Since then, I've developed a long-term reflection on foresight and the differences between extrapolation, forecasting, prediction and scenarization that it would take too long to detail here - the mere enumeration of these notions shows that deducing the future from what we think we know about the present and the past is far from simple.
After starting my scientific adventure with errors of judgment, how have I been able to avoid them ever since? Everywhere I went, I was assigned the social science epistemology courses that nobody wanted. I deduced that we should draw inspiration from physics to be rigorous, while adapting our approach to the singularity of our objects. Astrophysical research rarely involves falling in love with its object - an exoplanet, a black hole or antimatter, for example. This doesn't mean we don't need poetry to name the celestial bodies and particles we study (my favorite metaphor is "cosmic inflation"), but the risks of being wrong about the existence of an invisible source of gamma rays are not increased by the need to "liberate" them from gravitation.
Unfortunately, objectification and the quest for neutrality take time. Pressed to explain to large audiences what we don't yet understand ourselves, we fall back on what was taught to us by people to whom we owe our vocation, who themselves trained in the social sciences decades ago. That's why we're so fond of quoting eminent personalities and their writings from over a century ago. Naming the dead for want of working on the supposed discoveries of the living exempts us from both objectification and neutrality, because ghosts are always right.