Research projects and interests




Brief overview of my research


Selected projects:

1. Reasons and motivations underlying decisions
2. Logic in the social sciences
3. Social choice theory and judgment aggregation


Brief overview of my research

My research interests are as diverse and interdisciplinary as my academic background. Most of my work relates in one way or another to human agents and decisions, but the perspective varies considerably: philosophical or formal, normative or positive, individual or collective, foundational or more applied. I am currently developing a new focus on reasons and motivations, while also continuing to work on the aggregation of judgments. I pursue these two projects in a joint effort with Christian List. I am also working on logic, welfare economics, mathematical models of terrorism prevention, and foundational questions in game theory, probability theory and statistics.

back to top

Selected projects

1. Reasons and motivations underlying decisions

This recently started project aims to offer a revision of the standard notion of a rational agent as it is extensively used in the social sciences. Consider two statements about rationality:

"A rational agent acts on the basis of reasons."
"A rational agent acts on the basis of beliefs and desires."

Both statements appear to be correct; indeed I believe they are both correct. At the same time, these statements point towards two different paradigms. Simple attempts to reconcile these paradigms - such as to identify reasons with beliefs or desires - are inappropriate, as I believe. While a large philosophical literature is in line with the reason-based paradigm, standard modelling practice in the social sciences follows the second (`Humean') paradigm. Indeed, standard modelling practice is based on expected utility theory, which captures beliefs and desires via subjective probabilities and 'utilities', whereas motivating reasons do not explicitly enter the picture, so that their role - if there is any - has remained obscure.

In a joint effort with Christian List, I aim to give reasons their proper place in rational choice theory, and to explore philosophical questions in ethics (possibly non-consequentialist) and political philosophy (notably deliberative democracy, reason-based consensus, rights, and Amartya Sen's Liberal Paradox). Without going into details here, the proposed new reason-based theory of rational choice neither abolishes beliefs and desires in favour of reasons, nor simply adds reasons as a third explanatory component of rational agency. Rather, motivating reasons are viewed as more fundamental than both beliefs and desires, namely as their explanation. The resulting picture is that of an agent who is fundamentally driven by reasons; these lead him or her to adopt certain beliefs and desires, which in turn lead to behaviour.

For an introduction to the project, see my paper A reason-based theory of rational choice with Christian List. This paper addresses reasons underlying desires, but not yet reasons underlying beliefs (epistemic reasons).

back to top

2. Logic in the social sciences

Social sciences deal with human agents. But how should the social scientist model these agents, their internal deliberation, and deliberation between agents? Among the different ways in which one might think of human agents, the logic- and language-based way arguably comes closest to a human's state of mind, reasoning and deliberation. Beliefs, convictions, desires, hopes, needs, moral judgments and wishes are all attitudes which are held (or can often be thought of as being held) towards propositions. Whether a semantic or syntactic approach to propositions is more useful is context-dependent; which of the two approaches is metaphysically more adequate is an intriguing question for me.

The logic-based approach can be used fruitfully to study change in psychological states. I am particularly interested in phenomena such as changes in awareness, imagination, and qualitative understanding, which are not representable within classical rational choice theory. I am less persuaded by some common approaches to bounded rationality, which usually take the classical expected-utility paradigm as their starting point and introduce certain deviations from it (such as non-additivity of beliefs) without modelling the source of these deviations (such as limits or changes in conceptualisation or imagination).

back to top

3. Social choice theory and judgment aggregation

Social choice theory aims to offer ways of reaching collective decisions of various sorts in the face of disagreements within the group. The theory is of interest for several kinds of groups, such as governments, expert panels, juries, or even the world's entire population. If a single individual has different 'selves' - for instance, a selfish and a loving self - then social choice theory can be applied to that individual's internal decision-making process.

I am working on several aspects of social choice, most notably on the aggregation of judgments (much in collaboration with Christian List), but also on the Condorcet Jury Theorem (and its significance for epistemic democracy), the aggregation of probabilistic beliefs or other forms of beliefs, and other topics.

What is judgment aggregation about? While the more traditional branch of social choice theory aims to merge individual preferences over given alternatives, the more recent area of judgment aggregation aims to merge individual (yes/no-)judgments over logically interrelated propositions, such as the propositions

'Carbon dioxide emissions exceed threshold x',
'If these emissions exceed threshold x, then global warming will accelerate',
'Global warming will accelerate'.

A basic observation is that proposition-by-proposition majority voting may result in logically inconsistent collective judgment sets. This is where the problem starts: the problem of finding suitable aggregation procedures which avoid such collective inconsistencies and are yet 'democratic'. Among the most prominent proposals are the premise-based, conclusion-based, distance-based and priority-based procedures.

Some central problems addressed in my research with Christian List are:

(i) How can one model the aggregation of real-world judgments which are usually not expressible in the language of classical logic?
(ii) How can one assign liberal rights or expert rights to particular individuals or minorities?
(iii) How can one prevent strategic manipulation by voters and agenda setters?
(iv) When is it possible to vote on a proposition-by-proposition basis, and what can be done if this is impossible?

This research also resulted in many theorems, including direct analogues of Arrow's Theorem and Sen's Impossibility of a Paretian Liberal.

back to top