Causal Inference Lab

MoL Project

Hypothetical Reasoning: its development and impact

Welcome to the website of the ILLC Master of Logic coordinated project, Hypothetical Reasoning: Its development and impact, happenning from 6 January to 3 February, 2020, and coordinated by Dean McHugh as part of the University of Amsterdam's Causal Inference Lab.



Course information is available on datanose at[88056].

Class schedule

Week 1

  • Wednedsay 8 January, 15:00–17:00, ILLC F0.25 (Science Park 107), Moral responsibility I: Causation
  • Friday 10 January, 13:00–14:00, Startup Village (Science Park 608), Moral responsibility II: Game-theoretic approaches to moral responsibility

Week 2

  • Tuesday 14 January 15:00–17:00, Science Park 904, G2.13, Causation in the law
  • Wednesday 15 January 15:00–17:00, ILLC F0.25, Causation
  • Friday 17 January 13:00–15:00, Startup Village, The development of hypothetical reasoning

Week 3

  • Tuesday 21 January 13:00–15:00, Science Park G2.13, Causation & Discrimination Law
  • Wednesday 22 January 15:00–17:00, ILLC F0.25. Presentations: Arie, Billy, Nachiket, Tianwei
  • Friday 24 January 13:00–15:00, Science Park 904, A1.06. Presentations: Flavia, Gerson, Grace, Laura

Week 4

  • Tuesday 18 January 15:00–17:00, Science Park G3.05, Discussion/supervision meetings
  • Wednesday 29 January, 15:00–17:00, ILLC F0.25 Supervision meetings
  • Friday 31 January, 13:00–15:00, Startup Village, Project wrap up


Students will be assessed by class participation, a talk given in the third week, and a two-page conference abstract due on 18:00 Friday January 31, 2020.


Schedule of student presentations

  • Wednesday 22 January 15:00-17:00 (ILLC F0.25)
    • 15:00-15:30 Arie
    • 15:30-16:00 Billy
    • 16:00-16:30 Nachiket
    • 16:30-17:00 Tianwei
  • Friday 24 January 13:00-15:00 (Science Park 904, A1.06)
    • 13:00-13:30 Flavia
    • 13:30-14:00 Gerson
    • 14:00-14:30 Grace
    • 14:30-15:00 Laura

Lecture slides


Project description

Many issues we face involve not only what actually happened, but spill over into hypothetical scenarios, asking what would have happened had certain things been different. This kind of reasoning, called hypothetical reasoning, is essential in many areas of life: 

  • Planning (“If we take the train, we’ll be there earlier.”)
  • Determining causation (“If the sensors had been working, the plane would not have crashed”)
  • Drafting policy (“If we build social housing, poverty levels will decline”)
  • Attributions of moral responsibility, praise and blame (“The accident is your fault: if you had been more aware, the accident would not have happened“)
  • Law (“If the employee were male, her salary would be higher: so the employer unlawfully discriminated on the basis of sex”).

We will investigate the development and moral implications of hypothetical reasoning. The course will consist of four lecutres, each on a different topic, and individual supervision for students to develop their own project.


Lecture 1. Moral responsibility

Attributions of moral responsibility involve an essential hypothetical component, considering what would be different had the agent acted differently. Perhaps, then, by understanding the logic of hypothetical reasoning, we can better understand what it means to be morally responsible.

This involves tackling the following questions, among others.

  • Can one be morally responsible for an outcome without causing it?
  • Even if causation is necessary for moral responsibility, what else is required?
  • Can moral responsibility be attributed to groups (such as companies and governments), or is all moral responsibility ultimately an aggregate of individual responsibility? (Consider Ed Thurlow’s 18th Century remark, “Did you ever expect a corporation to have a conscience, when it has no soul to damn, and no body to be kicked?”)
  • Can one be responsible for overdetermined events? That is, is the following a sound argument: "I'm not responsible for the outcome, because even if had not done what I did, someone else would have done it anyway, and the same outcome would have occurred"?

To examine the last question more concretely, have a look at the following argument, taken from the website of the ABP (the Dutch pension fund for employees in education and the government), which currently invests employee pension contributions in the fossil fuel industry:

Would it not be better for ABP to get out of fossil fuels altogether?
Some people believe that selling all of our fossil fuel investments would send a powerful signal that we were taking climate change seriously. In practice, however, the effect would be minimal. Fossil energy companies would remain financially attractive investments and other investors would simply take our place. We might give a powerful signal but we would lose our influence at the same time. We believe we can do more on the climate change front by continuing to invest in these companies and encouraging them to become more sustainable.” [link to document]

Required reading


Lecture 2. Causation in the law

In this part we examine issues arising from the moral implications of causal reasoning. We ask two questions in particular. Firstly, is causation as an all-or-nothing notion (like innocence versus guilt), or is better modelled as gradable (like punishment)? Secondly, can causal judgements ever be divorced from moral considerations? If not, to what extent can courts be expected to decide matters of causation in an impartial way? And if not, to what extent can legislatures legitimately write laws invoking causal concepts, expecting courts to interpret them?

Required reading


Lecture 3. Causation in the law, a case study: discrimination law

In this part we put analyses of dependence to the test with a particular case study: discrimination law. There is an intuitive parallel between dependence and discrimination: to be discriminated against on the basis of a given protected trait is to be treated unfavourably because of that trait; in other words, for one's maltreatment to depend on how one has that trait. But what precise test should one apply to determine when an action was carried out “because”/“in virtue”/“on account” of a given trait? Does proof of discrimination require the existence of an actual comparator (one differing in the protected trait but the same in all other relevant respects)? What facts from the actual world can one hold fixed when imagining the person differing in the protected respect? 

For instance, suppose an employer fires any employee, regardless of their sex, who requires more time off than their contract provides. Does that employer discriminate because of sex when they fire an employee who requires extra time off because of complications from pregnancy? When imagining that person being a different sex, one cannot fix the fact that they are pregnant, but should one fix the fact that they require extra time off?

Required reading/listening