[MUSIC] We're going to tackle in the next ten minutes trust in science and trust in quantification, how these two things are related. But before we do that, we will also ask ourselves the question, is it so that everything which is told by science should in principle be trusted? In fact, we will be arguing that science is done by human and can err and can be mistake. But at the same time, the use of science for policy demands a relation of trust between those who produce a number and those who would use them. So starting with the first argument, whether we should always trust science, the answer is of course no. We should always be vigilant and circumspect with the evidence we are provided by science. The example I offer you here among many which could be used, either one of Karl Pearson, the father of statistics, you may have studied in your class in Statistics 101 key square and other statistics done by this famous statistician. But he was also a supporter of eugenics at the beginning of the 20th century. The idea was very popular in that period that the genes determine the quality of human being. So you had superior race and not superior race. Everybody at that time subscribed to various forms of the social Darwin, which eventually merged into eugenics. And I would say statistician had the dominant role in this story because they established laboratory and journal to study and promote eugenics. And Pearson was so convinced of eugenics that he thought that it was a waste of time to spend money to help poor people. So for him, all the social program were the waste of money. And the argument was if we help poor people to study or to eat, if we give them money, this will not improve their genes, and hence that will remain of an inferior quality. And hence, they will reproduce, and that will aggravate the problem. So the best things to improve mankind would be simply not to use any kind of social program. Okay, now a day, the problem of science are very much different. Because with industrialization of science and the incorporation of science within a system of profit, I would say, of growth and profit, very often, you have the same scientist playing as academic but also as a consultant. And the issue of conflict of interest is very clear, as you can see from the title of this article published on the New York Times. So depending on whether you are acting as an academician or a consultant, you may happen to behave in different ways. Worst than that, the penetration of corporate interest into academic life is so strong that academia active that enterprises, and corporate interest actively recruit academicians to their service. And more than that, they even control journals. Here we have an article which describe how two journals are, in a sense, controlled by pharma companies so that article which are favorable to the product of those companies can have a good reception on those two journals. And this article, which happen on the publicintegrity.org, which is an NGO, is that there are law companies, law firms, firms of lawyers, in other words, whose work is to intermediate between corporate interests and scientists and academia. There had been really many examples of pressure from corporate interest on medicine, for instance. And this is the most egregious, the one I'm showing you here. This is a paper just a few months ago published on the Journal of American Medical Association. Where they admit, in a sense, or declare that the sugar industry funded research on the effect of cholesterol on coronary heart disease to down the effect, distract attention to the possible negative effect of sugar. And here is a sentence which is taken from the article. So they even created a research program. But what they did was not much unlike what was done by the tobacco companies in the early '70s and '80s. They funded a lot of research on alternative cases of cancer. Which was good, in a sense, because what money was given for research. But these were done for the purpose of distracting people from the relation between smoking and cancer. And even here, statisticians played a role, because Fisher, another famous statistician, was convinced that the error of causality was from cancer to smoking, and not from smoking to cancer. This is also not a story you might want or wish to go and check on the internet. So for this reason, many people have expressed concern about this kind of contrast between the official image of science and the practice of science where scientists are often enrolled to defend corporate interest. And here you see Also, the idea that there is, in science, a certain tendency to side with the establishment and to go where money is to defend the status quo, in a sense. And it is the article of Macilwain, which is quoted on this slide. And then he said in this article, before Trump was elected, he said, well, if Trump will be elected eventually, scientists would need to look at their part in its downfall. In the downfall, in this case, of western democracy. Let's discuss a bit the case of economics. In economics, there is now a very interesting discussion going on would be to the reaction of the incapacity of economics to foresee the depression, the crisis starting in 2008. And this is Paul Romer, a very well-known macroeconomist. Recently, he has been appointed to be the chief economist of World Bank. And he has this theory of Mathiness, and he described mathiness has a tendency of many economy to use mathematics to bail, to hide ideology. And you can read here how it explains the meaning of mathiness. He says, this is an approach where room is led for slippage between statement in natural versus formal language and between statements with theoretical as opposed to empirical content. Yet again, a way to use mathematics in a rhetoricalization, a topic we have already touched. Even more strange effect of this relationship between science and corporate interest. It's that not only those with the deepest pocket can purchase much science and advertise this science the most. But the same science can become a currency, which the lobbyists, for instance, use to purchase authority and power and attention from the political establishment. This is a thesis made in two parallel books, get a clearer independent but parallel in their content. One American, one European, and in both books, the point is made that very often, the lobbyist capture the politician. Thanks to his capacity to dispose of result, to dispose of facts, to dispose of data. And this is because he control more science. In this way, you see that there is, in the idea of evidence-based policy, a warring symptom of a potential asymmetry. Someone has a capacity to mobilize a lot of science, and those who receive this don't have such a capacity. And if you want to read the same story from the point of view of the lobbyists, this is a book written by the lobbyist where the orders instruct the lobbyist as he should go about being regulated and hence having to produce science and support its own thesis and put doing this to recruit scientists. And hence, the suggestion is here, to recruit the best scientist, not just any scientist, and then to recruit them using a particular diplomacy. The slide says when you want to recruit one scientist, this activity requires a modicum of finesse. It must not be too blatant for the experts themselves must not recognize that they have lost their objectivity and freedom of action. I promised to say something on link between trust and quantification, and this is a book where the storyteller will explain, Theodor Porter, Trust In Numbers. And Theodor Porter contrast two groups of engineers. [FOREIGN] Des Ponts Et Chaussees in France and the US Army Corps of Engineers. And the story is that though the engineers in these two groups are equally good, equally clever, then you operate in different institutional context. In France, engineers are trusted, so they can qualify at ease using reasonable instrument of quantification. On the other hand, you ask army corp engineer to operate in a very adversarial system where the quantification is prescribed and the rule of quantification are also prescribed. And as a result, the quantification produced by those engineers are often unreasonable, because they have been done mechanically just in order to satisfy guidelines. Well, this concludes our short discussion of the relation between trust in quantification.
