TiLPS

The Tilburg Center for Logic, Ethics, and Philosophy of Science studies knowledge, reasoning, and value in all their forms.

tilburg university

Research Seminar in Epistemology and Philosophy of Science

This research seminar is a forum for all members of TiLPS with research interests in epistemology or philosophy of science to present and discuss their work. Sometimes we also have a guest speaker or discuss a recently published article. Master students and Research Master students may take this research seminar for credit. Please contact Matteo Colombo for more information. Papers for discussion and some background reading will be available from this website at least one week in advance.


Upcoming Seminars



Recent seminars

Wednesday, 2 September 2015, 16:45 - 18:00

Room: DZ 119

Speaker: Thomas Boyer-Kassem (Tilburg University)

Title: Explaining scientific collaboration: On the epistemic efficiency of groups in a competitive environment
(joint work with Cyrille Imbert)

Abstract: More and more scientific articles are now produced collaboratively, by larger and larger teams. There is little doubt that this trend can be explained by combinations of various factors, which can be epistemic or non-epistemic. Our claim is that, to explain scientific collaboration, it can be sufficient to invoke small differences in a collaboration’s efficiency to pass the various steps of a research project. Our rationale builds on a formal model proposed by Boyer-Kassem and Imbert (forthcoming), and goes further: whereas this model gives normative statements, we use them here to provide an explanation of collaboration, in a functional sense.

Wednesday, 9 September 2015, 16:45 - 18:00

Room: DZ 119

Speaker: Alan Thomas (Tilburg University)

Title: Stags, Hares and Knowledge: A Genealogy of the Knowledge System as a ‘Mutual Assurance Game’

Abstract: This paper considers whether the analogy between two 'cooperative ventures for mutual advantage' – a market economy and the knowledge system – offers any explanatory insights for social epistemology. It is argued that it does in two ways: analysing the mechanisms of social cooperation and the kinds of goods produced suggests that the knowledge system is correctly modelled as a mutual assurance game. It primarily exhibits economies of scale and produces a 'steep' public good, namely, knowledge. This, in turn, has the consequence that the concept that this social practice embeds – knowledge – ought to receive a genealogical explanation. It is argued that this form of explanation is compatible with the concept having no interesting analysis.

Wednesday, 16 September 2015, 16:45 - 18:00

Room: DZ 119

Speaker: Lee Elkin (MCMP Munich)

Title: Confirmation Theory with Imprecise Probabilities

Abstract: Some Bayesians have argued that belief states are better represented by sets of probabilities or imprecise probabilities (IP) instead of sharp or precise probabilities. However, it is not immediately obvious in the IP account when new evidence confirms or disconfirms a hypothesis if an individual’s prior is imprecise. In this paper, I propose what appears to be a plausible theory of confirmation with imprecise probabilities, which turns out to be similar conceptually and structurally to traditional Bayesian confirmation theory. I then go on to describe a few unique instances of confirmation and disconfirmation that arise due to the flexible nature of imprecise belief states. Unfortunately, the naïve IP confirmation theory fails to accommodate the unique instances, revealing its shortcomings. In the end, I briefly remark on what the concerns suggest for the future of an IP confirmation theory.

Wednesday, 7 October 2015, 16:45 - 18:00

Room: DZ 119

Speaker: Silvia Ivani  (Tilburg University)

Title: What is Fertility?

Abstract: Philosophers traditionally include fertility among the cognitive values of good scientific theories. Unfortunately, philosophers do not share a clear definition of such a value. Thomas Kuhn, for example, claims that a fertile theory discloses new phenomena. Ernan McMullin, instead, maintains that a fertile theory provides imaginative resources to deal with anomalies. In this talk, I try to clarify the notion of fertility. First, I provide a list of definitions of this value. Second, I analyse how these definitions apply to a specific research, i.e., evolutionary psychology. Finally, I argue for the need to make a distinction between two different kinds of fertility.

Wednesday, 21 October 2015, 16:45 - 18:00

Room: DZ 119

Speaker: Colin Elliot  (Tilburg University)

Title: E.T. Jaynes' solution to the problem of Countable Additivity

Abstract: The axiom of Countable Additivity regulates how we can distribute probabilities amongst a countable infinity of mutually exclusive events. It is routinely used in mathematical probability. If we interpret probabilities as degrees of belief, however, it seems that this technical axiom is dictating what we should believe! This has troubled many a philosopher, and remains an open problem to this day: should Countable Additivity be an axiom of probability? Should we drop the axiom if we use probability in philosophy? In this talk, I examine the solution proposed by E.T. Jaynes to this philosophical problem.

Wednesday, 28 October 2015, 16:45 - 18:00

Room: DZ 119

Speaker: Naftali Weinberger (Tilburg University) 

Title: What is the Problem of Extrapolation?

Abstract: It is well known that a policy that works in one area will not necessarily be effective in another. This is known as the problem of extrapolation. It is surprisingly difficult to give a precise characterization of this problem and to specify how it relates to causal and statistical inference. In this talk, I show how extrapolation problems may be represented using contemporary causal modeling techniques and make precise the sense in which extrapolation goes beyond standard causal inference.

Wednesday, 11 November 2014, 16:45 - 18:00

Room: DZ 119

Speaker: Filip Buekens (Tilburg University)

Title: Friction and Harmony in the Realm of Taste

Abstract: I explore issues about taste in the context of game theory and the cooperative view of communication. The game theoretical template for a public dispute over matters of taste is Battle of the Sexes. Speech act theory allows that speakers play different language games simultaneously. In a dispute in matters of taste speakers and their intended audience play two games: the game of informing others and the alignment of preferences game. The different games select different propositional contents. This fits well with the cooperative view of human communication developed by Tomasello (2008). I conclude with a critical assessment of contextualist and assessment relativist approaches to faultless disagreement.

Wednesday, 18 November 2015, 16:45 - 18:00

Room: DZ 119

Speaker: Jan Sprenger (Tilburg University)

Title: Conditional Probability, Counterfactual Reasoning and the Ramsey Test

Abstract: What does it mean to have a degree of belief in E given H? Frank P. Ramsey famously argued for the following test: let's add H hypothetically to our background knowledge and argue on that basis about E. I argue that only Ramsey's interpretation can make sense of statistical reasoning which is essentially counterfactual. This argument has substantial implications. First, it solves a variant of the Problem of Old Evidence. Second and more importantly, it shows that objective chance is no central concept in scientific modeling. Therefore, we need not invoke the Principal Principle for coordinating statistical probabilities and degrees of belief.

Wednesday,  25 November 2015, 16:45 - 18:00

Room: DZ 119

Speaker: Alfred Archer (Tilburg University)

Title: Obligation and Supererogation for Virtue Ethicists

Abstract: A prominent objection that has been raised against Virtue Ethics as an independent view of normative ethics is that it is insufficiently action guiding. In recent years, several virtue ethicists have sought to respond to this problem by providing accounts of obligation and supererogation in virtue ethical terms. I will argue that all of these extant accounts face important problems that give us good reason to look for an alternative account. I will then provide such an account and show how it avoids the objections facing the existing views.

Wednesday, 9 December 2015, 16:45 - 18:00

Room: DZ 119

Soeaker: Bart Engelen (Tilburg University)

Title: The Definition and Role of Self-Interest and Preference in Economic Models

Abstract: Models in positive economics generally aim to provide truthful predictions and explanations of significant (aspects of) economic phenomena. While the concepts of ‘self-interest maximization’ and ‘preference satisfaction’ figure prominently in micro-economic models of people’s behavior and interaction, they suffer from a remarkable lack of conceptual clarity and rigor. In this paper, I critically analyze Daniel Hausman’s book Preferences, Value, Choice, and Welfare (2012), which is by far the most extensive and promising attempt at addressing this problem. Showing both the value and the problems of his views, I aim to outline the challenges positive economists face and the possible strategies they can adopt. In my view, Hausman’s definition of preferences as ‘total comparative evaluations’ risks turning it into an passe-partout notion, reducing its predictive and explanatory power. In my view, a more narrow understanding preferences as ‘overall comparative evaluations’ and hence as one of the many factors that influence people’s behavior and interaction – next to e.g. urges, commitments, duties and norms – avoids these problems.

Wednesday, 16 December 2015, 16:45 - 18:00

Room: DZ 119

Speaker: Matteo Colombo (Tilburg University)

Title: Sleeping Beauty goes to the lab: The psychology of self-locating evidence

Abstract: The Sleeping Beauty Problem has attracted enormous attention in philosophy and economics. It raises questions of unsuspected theoretical relevance for the foundations of probability and decision theory, and beyond. With this project, we aim to test the descriptive adequacy of the standard halfer and thirder accounts of the sleeping beauty problem. Our results show that naive reasoning displays an intricate pattern, which does not fit either the halfer or the thirder accounts. We suggest that that self-locating evidence is acknowledged as relevant but its quantitative impact is largely discounted.

Wednesday 13 January 2016, 16:45 - 18:00

Room: DZ 119

Speaker: Luo Dong (University of the Chinese Academy of Sciences, Beijing)

Title: Invariance: Reality or Objectivity?

Abstract: Invariance has great heuristic power in physics. Invariance can be interpreted either as a posteriori feature of statements on the world or a precondition for an objective description of the world. Both ‘invariance as a criterion for reality’ and ‘invariance means objectivity’ have supports among philosophers and scientist. By investigating the role of invariance in process of constructing theories, we argue that assertions about reality are assumed before proceeding invariance, and invariance itself cannot sever as a criterion for reality. The invariance criterion can sever as a precondition for constituting objectivity in the method of investigating the world by structures.

Wednesday 20 January 2016, 16:45 - 18:00

Room: DZ 119

Speaker: Raoul Gervais (Tilburg University)

Title: Selective experiments in classical genetics

Abstract: Experiments are often conceived as tests for the presence or absence of causal relations between two or more variables. Two forms of experiments are generally recognized: intervention experiments, where we manipulate the putative cause A and measure or observe the effect B, and the more passive observational studies, where we measure or observe A and B without manipulating either. However, a detailed analysis of some experiments in classical genetics suggests that some experiments resist classification as either intervention experiments or observational studies. In particular, Mendel's experiments on pea plants, though clearly tests for causal relations, do not involve direct manipulation. Instead, the manipulation of the putative cause for the distribution of phenotypical traits in filial generations depends on a preceding selection of the paternal and maternal plants. We therefore propose to expand the existing taxonomies of scientific experiments to include a new category: selective experiments.

Wednesday 3 February 2016, 16:45 - 18:00

Room: DZ 119

Speaker: Jun Lai (Tilburg University)

Title and abstract: t.b.a.

Wednesday 10 February 2016, 16:45 - 18:00

Room: DZ 119

Speaker: Michal Sikorski (Tilburg University)

Title: Conditionals and Causality

Abstract: In my presentation I will discuss the semantics for both indicative and subjunctive conditionals and the relation between true conditionals and true causal claims. On the basis of both, I will try to draw two lessons for theorizing about causality. The first lesson will concern probabilistic relevance as one of the minimal conditions for causation. The second lesson will concern the type/token distinction among causal relations.

Wednesday 9 March 2016, 16:45 - 18:00

Room: DZ 119

Speaker: Naftali Weinberger (Tilburg University)

Title: Measuring Discrimination

Abstract: To show that a company illegitimately discriminated against black individuals in its hiring of applicants, it is not sufficient to demonstrate that being black caused certain applicants not to get the position. Here’s why. In a country where black individuals are routinely discriminated against, it is plausible that as a result of this discrimination black applicants will be less qualified for certain positions. Yet it is not wrong to discriminate among applicants based on their qualifications. The conclusion I draw from this is that showing that someone was discriminated against on the basis of trait T involves more than simply showing that her having T caused her to be treated differently. Additionally, one must establish how her having T influenced her treatment. In my talk, I show how it is possible to establish such facts using causal models, and address two problems related to causal models for discrimination. One problem is that it is unclear whether fixed properties such as race and sex can count as genuine causal variables, since counterfactual questions about how an individual would have been had they been a different sex or race are not well defined. A second problem is that discrimination may be an ongoing process, which makes it hard to understand the time relations of the causal variables.

Wednesday 16 March 2016, 16:45 - 18:00

Room: DZ 119

Speaker: Thomas Boyer-Kassem (Tilburg University)

Title: Scientific expertise, risk assessment, and majority voting

Abstract: Scientists are often asked to advise political institutions on pressing risk-related questions, like climate change or the authorization of medical drugs. Given that deliberation will often not eliminate all disagreements between scientists, how should their risk assessments be aggregated? I argue that this problem is distinct from two familiar and well-studied problems in the literature: judgment aggregation and probability aggregation. I introduce a novel decision-theoretic model where risk assessments are compared with acceptability thresholds. Majority voting is then defended by means of robustness considerations.

Wednesday 20 April 2016, 16:45 - 18:00

Room: DZ 005

Speaker: Frans van Doorne

Title: Reconstructing and Modeling Habermas’ conceptual Frame of Social Interaction. Towards an integrative conceptual foundation of social sciences.

Abstract: Habermas’s theory of society has given important incentives to the analysis of the relation between scientific and everyday problemsolving in several social-scientific disciplines. I intend to show how the structuring conceptual network of his theory can be reconstructed with the use of mathematical means, and how it connects the different conceptual levels of his theory. The performance-oriented models thereof can be operationalized, being tested and made measurable in  social-scientific disciplines, thus assuring their social character and fostering interdisciplinary  research.

Wednesday 11 May 2016, 13:30 - 15:00

Room: DZ 119

Speaker: Mattia Andreoletti (SEMM European School of Molecular Medicine, Milan)

Title: The Replicability Crisis

Abstract: In the last few years, the scientific community has been dealing with the so-called replicability crisis: many published results often disappear when someone else tries to replicate the study. This crisis has reached the public attention and has jeopardized the reliability of biomedical sciences, generating a general mistrust in scientific research. In the meantime, a 'science of science', known as meta-research, has flourished promising to overcome such a problem, discriminating good science from bad science and investigating the true nature of scientific knowledge. Despite philosophers discussed the same problems for centuries, so far they offered little guidance for meta-researchers. In this talk, I will present the replicability crisis of biomedical sciences trying to figure out if and how philosophy of science can be relevant for meta-research.

Wednesday 11 May 2016, 16:45 - 18:00

Room: DZ 119

Speaker: Silvia Ivani (Tilburg University)

Title: What We (Should) Talk About When We Talk About Fruitfulness. An Update.

Abstract: Thomas Kuhn (1977) suggested a list of five desirable values that scientists should take into account in theory choice. That list included accuracy, consistency, scope, simplicity, and fruitfulness. Since then, several philosophers have discussed the meaning and role of these values. Surprisingly, little attention has been paid to fruitfulness. In this talk, I suggest an analysis of this value. I will focus on the methods and tests employed by a research program. Moreover, I make use of a specific case study, i.e., the adaptationist program, to show how this approach improves the understanding and assessment of fruitfulness.

Wednesday 18 May 2016, 16:45 - 18:00

Room: DZ 119

Speaker: Colin Elliot (Tilburg University)

Title: Inference with non-standard probabilities

Abstract: Suppose we are out observing swans and that we update, using Bayes' theorem, our opinion on the hypothesis H = 'all swans are white'. One non-white swan is enough to prove H false. If we model this situation using the usual axioms of probability, as we observe a growing number of white swans, we can be forced to become increasingly sure that H is true. This may seem just fine; but some authors (K.T. Kelly in particular) worry that we are taking an easy way out of the problem of induction: after all, we get this convergence towards H exclusively thanks to the axioms we are using, which also make it impossible to properly express the sceptical position that, however many white swans we may see, the next one might be black. I model this simple inference first in the usual way, then abandoning the contentious axiom (countable additivity), and finally using non-standard probabilities, to explore the expressive power and epistemological consequences these different axiomatisations of probability carry.

Mail a friend

What is the answer for one plus three