Research

Current Research Projects

• Characterising Thick Ethical Concepts

  What makes a thick ethical concept thick? How are the descriptive and the evaluative component of a thick concept connected? Is it a semantic or a pragmatic connection? Let's provide empirical answers!

  Project Description

  As Williams (1985) suggested, two kinds of evaluative terms and concepts can be distinguished, namely thick and thin ones. Thin terms and concepts evaluate an object as, for instance, right, permissible, wrong, or blameworthy, yet they don’t explicate in which way the object is right or wrong. You might evaluate an instance of lying as wrong without conveying any information as to why you think so. You might think that people have a right to be told the truth, that you fear the person will make the wrong decisions based on the lie, that it ruins friendships etc. Thick ethical terms and concepts provide such information in addition to evaluating the object. Typical examples are “rude”, “courageous”, “trustworthy”, or “dogmatic”. Describing an agent as courageous makes a descriptive judgment that the agent is willing to take risks, and it evaluates taking risks positively – in contrast to being reckless.

  It is exactly the combination of evaluative and merely descriptive elements of thick concepts that has attracted attention in philosophy of language and metaethics, and thick concepts are often recruited to do important philosophical work. First, virtue ethics put particular emphasis on thick terms and concepts, as all virtue and vice terms are thick. It has often been argued that action-centred ethics such as deontology and utilitarianism provide a wrong and misguided approach to morality because they oversimplify the complexity of evaluative moral thought. Thin concepts only reflect a minor part of the way we morally think and talk and are, thus, ill-equipped to provide a psychologically realistic account to morality. Second, some metaethical cognitivists argue that moral judgment, such as “torturing others for fun is bad”, can have truth-values in the same way as “The earth moves around its own axis”. Whether the judgment is true or false is determined by the world. In addition to this moral realism, some philosophers have also defended moral objectivism, the claim that the truth-values of moral judgments are universal and not culture dependent. Thick ethical terms and concepts have an important function in arguments for moral objectivism, as they anchor values in non-evaluative descriptions. Third, thick ethical terms and concepts seem to challenge the fact-value distinction and, therefore, the principle that an ‘ought’ can refer be derived from an ‘is’. But this principle, also known as Hume’s law, is at the heart of normative-ethical debates about which rules we ought to follow, and it motivates normative logics as an independent research project. Fourth, thick-ethical concepts raise important metaethical questions about the connection of moral judgments and reasons to act.

  Because of this distinctive significance of thick ethical terms and concepts for a variety of philosophical debates, it comes as no surprise that much theoretical work has been done on understanding thick ethical terms and concepts. Two main questions guide the debate:

  1. Evaluation Question: How are thick terms and concepts related to evaluations they may be used to convey?
  2. Combination Question: How do thick terms and concepts hold together evaluation and non-evaluative description?

  Interestingly though, these questions cannot only be answered by armchair philosophy but call experimental philosophers as well as experimental linguistics to the table. The aim of this research project is to investigate what we can do with thick ethical terms and concepts in ordinary language in order to answer the more fundamental question about the nature of thick concepts themselves. The results of this project have immediate implications for metaethical debates about cognitivism and non-cognitivism, the possibility of moral disagreement, and many others.

• Thick Concepts and Reasons for Action

  Thick concepts communicate descriptive and evaluative content at the same time. But in doing so, do they provide reasons for action?

  Project Description

  Thick Concepts and Normative Reasons

  Terms and concepts which are a unity of evaluation and description are called thick terms and concepts (Eklund 2011, Väyrynen 2019), such as “rude”, “friendly”, “honest”, “manipulative”, “cruel”, or “compassionate”. Other evaluative terms and concepts which merely evaluate, such as “bad”, “good, “permissible”, “right”, “wrong”, are called thin ethical terms and concepts. Both types of concepts stand in opposition to descriptive terms and concepts whose function it is to merely provide information on what the world is like, yet not to evaluate, such as “round”, “rectangular”, “blonde”, “Australian”, etc.

  Thick terms have sparked the interest of philosophers, not only because of their hybrid structure. Another reason is their potential to guide our actions by providing us with reasons to act. As Williams said,

  “The way these notions are applied is determined by what the world is like (for instance, by how someone has behaved), and yet, at the same time, their application usually involves a certain valuation of the situation, of a person or actions. Moreover, they usually (though not necessarily directly) provide reasons for actions.” (Williams, 1985, p. 143 f.; own emphasis)

  The idea that thick concepts are action-guiding goes back to Bernard Williams, but many philosophers have followed his reasoning. While this is a quite plausible and seldomly disputed assumption, no empirical evidence has been offered in its support. In this paper, we investigate the use of thick concepts in ordinary language and how they relate to reasons for actions. More specifically, we investigate whether statements containing thick terms provide normative reasons for action. We need to understand a) whether there are circumstances in which thick concepts provide reasons for action, b) whether there are circumstances in which thick concepts do not provide reasons for action, and c) how these two classes of circumstances differ from one another.

  This is a joint project with Judith Martens.

• Thick Concepts in Moral Psychology

  Moral psychology is incomplete as long as we solely focus on thin ethical judgments. In this research project, I empirically investigate the role of thick ethical judgments in moral psychology.

  While experiments investigating people’s intuitions about moral cases differ along various lines, they are similar in one crucial respect: in all experiments, participants are asked questions or are supposed to indicate their agreement with a statement that all contain thin concepts, such as good, permissible, appropriate, bad, evil, impermissible, and the like. For instance, in several study by Greene and colleagues (2001, 2008), participants need to judge the vignette in terms of how appropriate or inappropriate the agent’s behaviour was. Follow-up studies in other languages, such as German (Heekeren and colleagues 2003, p. 1216; 2005, p. 889) or Italian (Ciaramelli and colleagues 2007), use the same test queries. Other researchers, such as Wiegmann and colleagues (Waldmann 2010) asked participants whether an agent should act or refrain from acting. Waldmann and Dieterich used the same wording (2007). In later studies by Greene and colleagues (2009), participants had to indicate whether it was morally acceptable for the agent to act in a certain way. Other thin term that frequently figures in experiments are “permissible” or “morally permissible” and “obligatory” (Cushman et al., 2006, pp. 1083-1084; Hauser et al., 2007, pp. 18-19). Young, Nichols and Saxe (2010) as well as Nichols (2007) asked participants how blameworthy the agent was. Alfano and colleagues (2012), but also Knobe and colleagues or Hindricks, Douven, and Singmann had a slightly different formulation and asked whether the agent deserved praise or blame for their actions. Haidt and Baron (1996) asked participants to rate the agent’s “goodness or virtue” on a scale from extremely bad” to “extremely good”. In other experiments, Haidt and colleagues used queries asking whether it was “Ok” or “wrong” for the agent to act in the way they did (Haidt 2001, p. 814; Haidt et al., 1993, p. 617; Schnall et al., 2008, pp. 1107-1108).

  All of these experiments test people’s moral intuitions related to thin ethical concepts, such as wrongness and goodness, (im)permissibility, (in)acceptability, blame- and praiseworthiness, etc. The methodologically approach to moral psychology so far has provided important empirical insights into how people make moral judgments. However, conclusions drawn from these insights have gone way too far and have oversimplified moral psychology.

  Thin moral psychology is not enough! Moral psychology has operated under the implicit assumption that thin concepts are sufficient to investigate people’s morality. There are several reasons for why this seems an acceptable and even a plausible assumption. On the one hand, action-focused ethics such as deontology and utilitarianism have been extremely influential in the philosophical tradition. In addition, deontology and utilitarianism often make very different predictions about what people should deem morally acceptable, especially in moral dilemmas Testing such intuitions is relatively easy and relies on a sparse and intuitively understandable vocabulary, namely thin concepts. So psychologists have good reasons to start off from deontological and utilitarian ethics and, thus, to focus on thin concepts.

  However, limiting empirical research to action-focused ethics ignores the complexity of the debate in moral philosophy and is likely to neglect an important aspect of moral psychology. Many normative-ethical schools have objected against deontology and utilitarianism that their conception of morality is oversimplified, psychologically implausible, or even misguided and fruitless. Other traditions such as Virtue Ethics, Discourse Ethics, but also Buddhist and Confucian ethics deny that focusing on actions is of any moral interest and can answer important practical questions. However, virtue ethical intuitions cannot be examined by using thin concepts. They require thick concepts, such as courageous, trustworthy, kind, self-less, and rude, nasty, or cruel instead. Given the importance of these alternative conceptions of morality, they deserve the same attention in empirical research to allow for an unbiased and fair examination of which normative theories seem to be in line with actual moral judgments.

• Developing New Tools to Detect Evaluative Language

  Evaluation comes in various form. The aim of this project is to develop corpus linguistic and experimental tools that allow us to detect evaluativity in language.

  Project Description 

  Philosophers assume that in addition to describing or representing the world, language can also evaluate it. When we say that the cake was delicious or that Jim’s behaviour yesterday was a disgrace, we evaluate the cake and Jim’s behaviour. In my research, I try to understand how exactly evaluation is conveyed in ordinary conversational contexts.

  In many of my other projects, I have used experimental studies to answer philosophical questions. However, when we want to understand how people use language, a less artificial and more natural context promises mor reliable results. The analysis of linguistic corpora can help us gain such reliable results. A linguistic corpus, roughly speaking, is a collection of spoken or written text. Technically, everything you consider worth investigating can constitute such a corpus, for instance the Stanford Encyclopedia of Philosophie, Google, the New York Times, or an online blog. Together with my collaborators Kevin Reuter and Lucien Baumgartner, we attempt to understand evaluative language by analysing such corpora.

  Before we can start analysing, we need to determine what we are actually searching for and how we believe to find it. So far we lack the means to identify and measure a wide range of evaluative concepts. In this project, we develop corpus-linguistic methods that allow us to distinguish thick adjectives like ‘courageous’ from descriptive adjectives like ‘tall’, and value-associated adjectives like ‘sunny’.

  We argue that the connective ‘and’ provide a useful means. We show that evaluative adjectives behave differently to descriptive adjectives when being combined with other adjectives through the connective ‘and’. In addition, the modifiers “truly” and “really” highlight the evaluative dimension of thick and thin concepts, allowing for them to be uniquely classified.

  You can watch a video of Kevin and Lucien explaining the tools we developed at the Corpus Fortnight 2020.

• Evaluative Language in the Legal Discourse

  To what extent does the legal discourse contain evaluative language and how is this evaluativity communicated?

  Project Description

   

  ---

• The Folk Concept of Lying

  I empirically investigate under what conditions an agent is judged to have lied to another person. What are the necesssary conditions for something to count as a lie?

  Project Description

  My research on lying is in full collaboration with Alex Wiegmann.

  Lying is an everyday moral phenomenon about which philosophers have written a lot. The moral status of lying has been intensively discussed, but also what it means to lie in the first place. Perhaps the most important criterion for an adequate definition of lying is that it fits with people’s understanding and use of this concept. In this light, it comes as a surprise that  researchers only recently started to empirically investigate the folk concept of lying. 

  One of the supposedly four necessary, and jointly sufficient conditions is the Statement condition, according to which lying requires the utterance of a statement that the speaker believes to be false at the level of what is said. This, however, excludes many relevant phenomena in which one might have the intuition that the speaker lied despite not having said anything false at the level of what is said. Rather, the agent lied by implying or presupposing something that he believed to be false. For instance, think about cases in which speakers intentionally leave out relevant information from a conversation: “Where were you last night?” – “I had a drink with a new colleague”. Imagine that what the speaker said is completely true. Going beyond what is said, such an answer further seems to imply that the speaker went to a bar and just had drinks (plus the typical social interaction one has while having a drink). However, if the speaker actually had champaign with his new affair in a hotel room, it seems that the speaker willingly deceived the addressee. But is such an utterance also a lie?

  ---

  Related Publications

  For an answer to this question, see: Wiegmann & Willemsen (2017)

  ---

  Drafts

  A preprint investigating whether questions can be lies, please see a pre-print of my paper with Emanuel Viebahn, Alex Wiegmann, and Neele Engelmann (the pre-registration of the experiment can be found here.
• Alternative Possibilities, Moral Responsibility, and Free Will

  I address questions such as whether the ability to act other one actually did is relevant for the attribution of moral responsibility and free will.

  Project Description 

  Is there an Action/Omission Asymmetry?

  The question of how moral responsibility and free will are undermined by the truth of determinism is one of the most extensively, but also most controversially discussed ones in philosophy. Many philosophers hav come to believe that the truth of determinism does not threaten moral responsibility and free will, as they reject the necessity to the ability to act other than one actually did.

  However, while many philosophers reject the Principle of Alternate Possibilities for actions, they still believe it to be essential when it comes to moral responsibility for omissions. The fact that an agent could not have saved a drowning child because, had he tried, he would have been hindered by a strong, indurmountabale current, seems to render the agent blameless. He might be to blame for not trying to save the child or for being the kind of person who doesn’t feel the strong urge to help, but he is not morally responsible for the child’s death.

  Various arguments have been presented as to why there is an asymmetry between actions and omissions with respect to alternate possibilites. And also, there is much debate about the conditions under which the agent is relevantly unable to act other than he actually did. One predominant and extremely influential tool used in the debate are philosophical thought experiments about hypothetical cases. And philosophers using these cases often make claims such as “in this case, we all have the clear intuition that alternate possibilities are irrelevant” or that “here, the lack of alternate possibilities undermines responsibility”.

  In this research project, I investigate whether these empirical predictions actually hold. To do so, I sift through the libraries written in the debate about determinism and free will and moral responsbility. I identify the most commonly used cases and compile the empirical predictions as to what intuitions “we all have” about them. I then methodologically revise these cases to make them suitable for systematic experiments. I then conduct controlled experiments on laypeople’s intuitions about these cases and evaluate the extent to which they confirm or question philosophical theories.

  Publications

  For a pre-print of my latest empirical research on alternative possibilities click here. The paper will appear in the 2019 volume of Oxford Studies in Experimental Philosophy.

  ---

  Project Description 

  Are we causally connected to situations which we could not have avoided?

  Many philosophers believe that in order to be morally responsible for something, you need to be causally responsible for this thing. For instance, if I were to hold you morally responsible for the broken window in my bathroom, I would have to demonstrate that you are causally responsible for the broken window. If you can prove that someone else broke it, you are off the hook. This supposedly necessary connection between moral and causal responsibility is supposed to hold for both actions and omissions alike.

  While this claim is not much debated when it comes to actions, metaphysicians do have some trouble to explain the causal relevance of omissions. How could you cause anything by not doing anything? I have argued elsewhere (Willemsen & Reuter 2016; Willemsen 2016; Stephan, Willemsen & Gerstenberg 2017; as well as Willemsen 2019) that the folk believe omissions to be causally relevant and that people are causally responsible for the consequences of omissions. So it might seem as if Necessity is empirically supported and indeed all moral judgments are supported by a causal judgment about the agent.

  Interestingly, the causal (ir)relevance of omissions plays a key role in some versions of the Action/Omission Asymmetry Thesis. Carolina Sartorio recently argued that both actions and omissions can be causes of an outcome, and agents can be held morally responsible for both actions and omissions — but only as long as the outcome could have been avoided. In Frankfurt-style scenarios, so when an agent could not have avoided a negative outcome, she can only be held responsible for her actions, yet not her omissions. The reason, so Sartorio argues, lies in the causal connection of the agent with the outcome. The consequences of an agent’s unavoidable action are still causally connected to the agent. For omissions, this causal connection is missing. Assuming Necessity to be true, this means that agent’s cannot be causally and thus not morally responsible for the unavoidable consequences of an omission.

  Building on my most recent paper on Alternative Possibilities, I test the explanatory power of this suggestion. Is it possible, that moral differences between actions and omissions can be explained by a difference in their causal status? We will find out!

  [Spoiler: Pre-tests suggest that there are differences in omissions’ and actions’ causal status. However, this differences is merely a difference in the degree to which they are judged causes of the outcome. At no point is it denied that omissions whose consequences couldn’t have been avoided are causally relevant.]

• Experimental Moral Philosophy

  I try to identify the factors relevant for the attribution of moral responsibility.

  Project Description

  I am interested in the factors that affect moral responsibility attribution. I start off from a rather classical philosophical stance on moral responsibility which philosophers have defined by means of necessary and sufficient conditions.

  For instance, it has been argued that causal responsibility is a necessary pre-condition for moral responsibility — if you haven’t caused it, you’re not to blame for it. In my research on omissions, I have worked quite intensively on whether causal responsibility plays a role in the attribution of blame for omissions.

  In contrast, it has been argued that several factors are irrelevant for moral responsibility, such as the agent’s social status.  However, empirical research indicates that this is not the case and social status and hierarchies do matter for the attribution of moral responsibility.

  The aim of my research is to work out the relevant factors for the attribution of moral responsiblity, and to discuss this everyday practice in light with the normative philosophical definitions.

• Omissions

  Things we don't do play an important role in our ordinary life. We are blamed for failing to do things, punished and often considered causally responsible for them. Why and how is that?

  Project Description

  The moral status of omissions

  Omissions broadly speaking are things that don‘t exist. They are your failure to water my plants, the accident you luckily never got involved in or the pony your parents never allowed you to have. And typically, when we highlight the absence of an action, an event, or a physical object, we do so in order to do something else. We blame people who fail to water our plants, and very often, we ascribe causal relevance to omissions, such as when we say that the lack of rain caused the drought, my not hitting the brakes caused the accident etc.

  But omissions are weird little biests. What exactly are we talking about when we say that my notting hitting the brakes caused the accident? What is the ontological status of omissions? The fact that it is not clear what kind if thing the absence of an action is creates further problems. According to many philosophers, causation is a relation in the world, holding between events. However, is my not doing a certain thing an event? And how can something that didn’t happen cause anything in the world?

  These and other questions motivate my work on omissions. In particular, I am interested in the relevance of omissions of actions for moral issues. It is typically assumed that moral responsibility presupposes that the agent has caused what she is blamed for. And this assumption has some great prima facie plausibility, considering hoe easy it is to defend oneself by saying “But I didn’t do it! I wasn’t even there. Here, this guy caused it, blame him!”. But if omissions are denied causal efficacy, it seems that we cannot blame agents for the consequences of their omissions. According to such a view, a mother who fails to feed her child cannot be held responsible for the child’s death, as she didn’t caused her child to die. This view, of course, strikingly contradicts our ordinary moral judgments.

  In my research on omissions, I investigate how people think about omissions in both ontological terms as well as in terms of their causal relevance.

  ---

  Related Publications

  Willemsen 2019: Omissions and their moral relevance. Assessing causal and moral responsibility for the things we failed to do.

  Stephan, Willemsen, Gerstenberg 2017: Marbles in Inaction. 

  Willemsen 2016: Omissions and expectations. A new approach to the things we failed to do.

  Willemsen & Reuter 2016: Is there really an omission effect?

• Counterfactual Reasoning and Causal Responsibility

  How to laypeople make causal judgments and what are the factors influencing them? What are the philosophical implications?

  Project Description

  Do the folk believe omissions to be causally relevant?

  My research on causal judgments is closely connected to my interest in moral responsibility attribution. It is widely held that in order to be morally responsible for an outcome, an agent must have caused that outcome. If she didn’t cause it, it seems absurd to ascribe moral responsibility for it.

  It has recently been argued that causal judgments are sensitive to norms and expectations. For instance, whether an outcome was produced by an agent violating a norm will affect the causal evaluation of the agent’s causal relevance. Also unexpected actions and omissions are considered to be more causal. Some researchers have suggested that this effect can be explained by the underlying cognitive process that leads to causal judgments. Causal judgments are made on the basis of counterfactual simulations in the sense of “What would have happened, had the agent acted differently?”. Those considerations are, however, heavily dependend on norms on expectations.

  In my research, I investigate what factors actually affect causal judgments and why that is. I have published three papers on causal judgments, putting emphasis on causation by omissions.

  Talks

  Click here to watch a talk on my project with Justin Sytsma, Roland Bluhm, and Kevin Reuter on Causal Attribution and Corpus Analysis.

  Other Related Publications

  Willemsen 2019: Omissions and their moral relevance. Assessing causal and moral responsibility for the things we failed to do.

  Stephan, Willemsen, Gerstenberg 2017: Marbles in Inaction. 

  Willemsen 2016: Omissions and expectations. A new approach to the things we failed to do.

  Willemsen & Reuter 2016: Is there really an omission effect?

  ---

  Project Description

  Moral responsibility, causation, and the Entailment Claim

  Just recently, Justin Sytsma, Kevin Reuter, and myself started a research project on the concept of causation. Most experimental research on causation investigate causal cognition, that is the mechanisms underlying causal judgments and the factors that can affect this mechanism. In this project, we examine the folk concept of causation or, more precisely, what people wish to express when the use the lemma ’cause’.

  ---