Carlo Martini

Experts in science: a view from the trenches

In this paper I analyze four so-called “principles of expertise”; that is, good epistemic practices that are normatively motivated by the epistemological literature on expert judgment. I highlight some of the problems that the four principles of expertise run into, when we try to implement them in concrete contexts of application (e.g. in science committees). I suggest some possible alternatives and adjustments to the principles, arguing in general that the epistemology of expertise should be informed both by case studies and by the literature on the use of experts in science practice.

Experts and consensus in social science

When we evaluate the outcomes of investigative actions as justified or unjustified, good or bad, rational or irrational, we make, in a broad sense of the term, evaluative judgements about them. We look at operational accuracy as a desirable and evaluable quality of the outcomes and explore how the concepts of accuracy and precision, on the basis of insights borrowed from pragmatics and measurement theory, can be seen to do useful work in epistemology. Operational accuracy (but not metaphysical accuracy!) focuses on how a statement fits an explicit or implicit standard set by participants involved in a shared project. While truth can remain a thin semantic property of propositions, operational accuracy, as a quality of an outcome of inquiry and typically attached to a statement, a model, a diagram or a representation is an evaluation based on the the non-epistemic goals set by the goal of inquiry (which every inquiry has), and a substantial evaluative notion. The goals, often made explicit by relevant questions in a context of inquiry, act as a filter, with truths a reliable epistemic method has access to functioning as the input, and accurate representations as its output. Responsible inquiry seeks pragmatic equilibrium between what reliable knowledge on the one hand and degrees of accuracy required by the goal of inquiry.

Consensus Formation in Networked Groups

This chapter proposes a solution, based on the theory of social networks, to the problem of weight assignment in the Lehrer-Wagner model for consensus. The Lehrer-Wagner model of consensus is introduced, and the problem of weight assignment is outlined, together with a number of possible solutions previously suggested in the literature. The chapter argues that there is no one-size-fits-all solution to the problem of weight assignment in the Lehrer-Wagner model, and suggests an alternative solution, which is based on the idea of deriving weights from existing networks of relations in the group. This proposal, it is argued, is particularly useful for maximizing or limiting the influence of a network of relations on the consensual opinion resulting from the model.

Expertise and institutional design in economic committees

In this paper, I consider the problem of selecting and justifying a lost of principles of expertise as part of a methodology of expert judgment in economics. I argue that a methodology of expertise, trying to list and justify a number of principles of expertise, is in need of a theoretical background against which to select, evaluate, and weight each of the principles. I explain by means of case studies why problems arise for lack of such background, using the Bank of England’s Monetary Policy Committee and the Council of Economic Advisors as examples. I then make a proposal for a categorization in ‘types of committees’ and ‘types of expertise,’ and use it to evaluate some of the principles of expertise that have been previously suggested in the philosophical and economic literatures.

The role of experts in the methodology of economics

Is subjective expert judgment a source of evidence in economics? In this paper, I will argue that it is, on a par with other sources like modeling, statistics, experimental, etc. I will also argue that it is not derivative, that is, reducible to the previous ones. But what is exactly the role of experts in economics? The contribution to the current methodological debate that I propose not only takes the role of expertise in economics as indispensable, but also suggests a need for a methodology associated with that role. I defend an indispensability argument, and indirectly David Colanders plea for refocusing the methodological debate on the lost art of economics. Finally, I indicate where to look for guidelines in formulating a methodology of economic expertise.

Seeking Consensus in the Social Sciences

In this chapter I clarify some of the terminology used in debates over the value of scientific consensus. I distinguish between an investigative phase of science and a disseminating one and defend the first main thesis in this chapter: The meaningful question about consensus in science is not whether we should seek it or not, but rather what kind of consensus we should seek and what kind we should give up. The second issue I pursue in this chapter is how we should seek consensus in science. I will defend the second main thesis of this chapter: consensus should not be sought “in the science”, but rather “among the scientists”. In other words, subjective expertise is not a dispensable source of evidence in the social sciences, and one of the goals of methodologists should be to find the right place for expertise in the pursuit of scientific inquiry. The point of the second thesis is that there is no real contraposition between the value of subjective expertise, and the value of other (objective) methods, but both can and should contribute to the formation of scientific consensus.

Introduction: Experts and Consensus in Social Science

The ideals of science as objectivity and consensus are – unsurprisingly – not so easy to attain in scientific practice. Science is ultimately a product of individual scientists with their own personal backgrounds and experiences, and there is no unique methodology to de-personalize and objectify knowledge. Social scientists, in particular, use a variety of tools for their investigations: They gather evidence from different sources, under different conditions and with different instruments. They are both the locus where different sources of evidence aggregate and also a direct source of evidence that comes in as intuitions and background knowledge. Acknowledging this wide variety of sources of evidence and methods in social science, different kinds of methodologies for reaching consensus have been developed. What kind of consensus is indicative of good science? What are the rules for consensus formation? And, is there a normative aspect to the formation of scientific and policy making consensus? The contributions of this book focus on experts: those institutional figures that act as a liaison between science and policy makers, politicians, governments, and other public domains.

A puzzle about belief updating

In recent decades much literature has been produced on disagreement; the puzzling conclusion being that epistemic disagreement is, for the most part, either impossible (e.g. Aumann (Ann Stat 4(6):1236–1239, 1976)), or at least easily resolvable (e.g. Elga (Noûs 41(3):478–502, 2007)). In this paper I show that, under certain conditions, an equally puzzling result arises: that is, disagreement cannot be rationally resolved by belief updating. I suggest a solution to the puzzle which makes use of some of the principles of Hintikka’s Socratic epistemology.