Graduate seminar, Spring 2014. Open to graduate students only.

Adam Elga (follow link for contact information, office hours). Meeting time: 10a-12:50p Thursdays. Location: Marx Hall 201

Description: Could it be rationally required to do something you are (rationally) quite confident is irrational? How should disagreement about how to respond to disagreement inform one's view of disagreement? Need it count against a belief or desire that it reflects the influence of irrelevant factors? Can causal decision theory be reconciled with compatibilism about free will?

We will investigate these and other questions from the point of view of Bayesian decision theory.

Readings: To access the readings (all available online, as linked below), you will need a userid (the userid is "guest") and a password (announced in class). If you would like to preview the readings, please email .

Note: In some cases, only a subsection of the linked reading is required. In those cases, the required range is listed after the reading itself. Also, the readings will be adjusted as the semester progresses. Within a couple of days of each seminar meeting, I will post updated readings for the subsequent meeting.

Visiting session leaders:

April 3: Arif Ahmed, Cambridge

April 10: Alan Hájek, ANU

April 17: John Collins, Columbia University

Schedule of topics and readings

Thu Feb 06: Level bridging principles and rational false beliefs about rationality

Homework assignment to be done before class. Read the following paper and exerpts:

Richard Foley. Fumerton's puzzle.

David Christensen. Rational reflection.. Read sections 1, 2, 6.

Now for a challenge. Let us say that we have a level-splitting situation if the following conditions are all true of a subject:

  1. The subject realizes that her only options are X and Y.
  2. The subject's degrees of belief are all perfectly rational.
  3. The subject is .9999999 confident that: it is rational for her to choose X and irrational for her to choose Y.
  4. It is irrational for her to choose X and rational for her to choose Y.

The challenge is to use the clock example to create the best candidate you can for a level-splitting situation. Write down and bring to class your answer. If you get stuck, don't panic:--just write down your best attempt, and/or an explanation of how you got stuck.

Optional background reading: Richard Fumerton, Reason and Morality Chapter 4: The concept of rational action (Ithaca: Cornell University Press, 1989). Read pp. 117-128.

Thu Feb 13: Disagreement about disagreement

Your assignment is to (1) read the indicated sections of these two papers:

David Christensen. Epistemic modesty defended. Read sections 0, 2, 3, 6.

Maria Lasonen-Aarnio. Higher-order evidence and the limits of defeat. Read sections 1, 4, 7.

...and (2) write and bring to class a few sentences (aim for no more than 200 words total) explaining the solution Christensen ends up favoring in the final section of his paper, and the solution Lasonen-Aarnio favors in the final section of her paper. No objections or critique are required: just give as compact and crisp a statement of each view as you can. Print out and bring your answers to class, as we will exchange them during one of the breaks.

Optional background: Adam Elga. How to disagree about how to disagree.

Extremely optional: Maria Lasonen-Aarnio. New Rational Reflection and Internalism about Rationality. A powerful and interesting critique of the level-bridging principles that we talked about on Feb 6. See especially section 2, up to and including the subsection entitled "Reals, fakes, and diminishing discriminative abilities". Please do not distribute this (forthcoming) paper without L-A's permission.

Thu Feb 20: Anti-luminosity and anti-anti-luminosity

"We must eliminate grades," headmaster Kolber [1] announced. "Giving grades ensures that we treat some students unjustly. For our school has a great many students, and their performance smoothly varies over the range from clearly passing to clearly failing. Consider the best-performing student who deserves to fail a class. We could never be in a position to see that this student deserves to fail, because his performance is so similar to that of the worst-performing student who deserves to pass. And we could never be in a position to see that this student deserves to pass, because he doesn't deserve to pass. So we could never be in a position to justly grade this student."

Your assignment is to use the exerpts below to assess headmaster Kolber's claim. Would it help the cause of justice if only real-valued numerical grades were given? Would it help if grades were given together with a numerical indication of how confident that the grader is that the grade is deserved?

Please print out a hard copy of your answers and bring them to class. Please write no more than 400 words, maximum.

Timothy Williamson. Knowledge and its limits Chapter 4. Read sections 4.1 - 4.5.

Jonathan Vogel. Luminosity and indiscriminability. Read from p. 549 through "...then the luminist can make good her criticism that the ALA is a sorites." (p. 557)

Selim Berker. Luminosity regained. Read from the beginning of section 4 (p. 10) through " misplaced? I think not" (p. 14).

[1] Headmaster Kolber's argument is inspired by Adam Kolber's Smooth and bumpy law and also Ted Sider's Hell and vagueness though neither work is assigned reading for this class.

Thu Feb 27: Decision theory without luminosity

Your assignment is to (1) read:

John Hawthorne. Knowledge and evidence. Read section 3.

Timothy Williamson. "Reply to Hawthorne": Replies to commentators. Read subsection 3 of the "Reply to Hawthorne" section (pp. 479-484).

...and (2) print out and bring to class your less-than-200-word answer to the following question: Why does Williamson think that good cognitive habits "are too loose and contingent on the accidents of human psychology to provide a systematic decision theory" (p. 483) ?

Optional: John Hawthorne and Amia Srinivasan. Disagreement Without Transparency.

Thu Mar 06: No class

Makeup class on Thu May 8.

Thu Mar 13: Individuation of preferences

Broome, J. (1991), Weighing Goods: Equality, Uncertainty, and Time (Blackwell Publishers Ltd.) Chapter 5: Expected utility and rationality. Read from "I shall begin, then, by looking..." (p. 100) through "... That is likely to be discovered only by debate." (p. 105)

Pettit, Philip (1991). Decision Theory and Folk Psychology. In Foundations of Decision Theory, eds. Michael Bacharach and Susan Hurley. Oxford: Basil Blackwell Ltd. Read the Apple example from page 163. (You need only read the second paragraph of section 5.2.)

What would Broome's reply be to the to the claim that Pettit's apple example shows that it can be rational to have non-transitive preferences? Is the reply plausible in this case? Print out your answers and bring them to class -- 200 words maximum, total.

Optional background reading: Broome, J. (1993), Can a Humean be moderate?, in G. Frey and C. Morris (eds.), Value, Welfare and Morality (Cambridge University Press), 279-86.

Thu Mar 27: Causal decision theory

We will discuss:

David Lewis. Causal decision theory.

Your assignment is to read section 2 through the bottom of p. 7, section 3, section 4 through " stands refuted" (p. 9), and section 5 through " act rationally is to realise some U-maximal option." (top of p. 12). Also print out and bring to class your answer to the following question:

Pretend that all you care about is whether you eat a turkey sandwich in the next ten minutes. Describe a dependency hypothesis that you place significant credence in, and one that you place very little credence in. Make your descriptions brief (fewer than 200 words total), but instructive enough so that by reading them, an outsider would get a good understanding of what it is to be a dependency hypothesis.

Optional question to think about: in the definition of a dependency hypothesis for an agent as "a maximally specific proposition about how the things he cares about do and do not depend causally on his present actions" (p. 11), what would go wrong (if anything) if the phrase "the things he cares about" were replaced by "things"?

Optional reading:

Andy Egan. Some counterexamples to causal decision theory. Philosophical Review 116 (1):93-114 (2007)

Thu Apr 03: Guest leader: Arif Ahmed

Topic: Causal decision theory, compatibilism, and counterfactuals

  1. Read:

    Arif Ahmed. Causal decision theory: a counterexample. Philosophical Review. Read sections 1-2.

    Arif Ahmed. Causal decision theory and the fixity of the past.. Read from the beginning up to footnote 4 on p. 296.

  2. Print out and bring to class your less-than-200-word answer to this (challenging!) question (courtesy of Arif Ahmed): Could defenders of causal decision theory get around the objections by formulating the theory either (a) in non-counterfactual terms or (b) in terms of counterfactuals that conformed to some non-standard semantics?

    If you are completely stumped by the above question you may if you wish instead write down a very specific, detailed question about some aspect of the assigned reading that you did not understand.

Optional background reading:

David Lewis. Are we free to break the laws?

David Lewis. Counterfactual dependence and time's arrow.

Thu Apr 10: Guest leader: Alan Hajek

Session title: "All Values Great and Small"

Alan Hájek and Harris Nover. Perplexing expectations. Mind 115 (459):703 - 720 (2006).

Read sections 1, 3, and 6, and print out and bring to class your less-than-200-word answer to the following prompt:

Orthodox decision theory does not constraint the value of the Pasadena game. Propose and motivate an additional rule for ranking games, a rule that allows the Pasadena game to be compared to at least one other game. (Your rule might take the form: "Infinite gamble A is to be preferred to infinite gamble B whenever [your condition goes here].")

Thu Apr 17: Guest leader: John Collins

Note special meeting room this week only: 301 Marx Hall.

Topic: Causal decision theory and unstable decisions

Assignment: Print out and bring to class (rather than submitting by email) your under-200-word answer to the following prompt (courtesy of John Collins): Read Egan's "Smoking Lesion" and "Murder Lesion" examples (pp. 93-97), and compare them to Richter's "Modified Death Case" (pp. 395-396). Describe the relevant structure of these examples. In what significant ways, if any, do the Egan cases differ from Richter's example?

Andy Egan. Some counterexamples to causal decision theory. Philosophical Review 116 (1):93-114 (2007)

Reed Richter. Rationality revisited

Thu Apr 24: The relationship between decision theory and game theory

  1. Read:

    John Collins and Achille C. Varzi. Unsharpenable Vagueness. Philosophical Topics 28 (1):1-10 (2000). Read section 1.

    Brian Weatherson. Games, Beliefs and Credences. Read from the last paragraph of p. 6 through end of p. 7.

  2. Arrange to eat an extremely delicious meal with an appropriate interlocutor. Each selection above describes a paradox. Over the meal, spend much of the time discussing how those paradoxes should be resolved. (I encourage you to choose as your interlocutor someone in the seminar, but this is optional. If you choose someone outside of the seminar, you will need to explain the paradoxes to them while you eat.)

  3. Extract one observation or suggestion or critical point from the discussion. Write it down (200 words max), print it out and bring it to class. If your interlocutor is from the seminar, the two of you may (but don't have to) hand in a jointly-written assignment.

Optional background reading:

Philip Pettit and Robert Sugden. The backward induction paradox.

Robert Stalnaker. Belief revision in games: forward and backward induction.

Robert Stalnaker. Extensive and strategic forms: Games and models for games.

Thu May 01: Permissiveness and contingency

  1. Read

    Millgram, Elijah. Practical Induction. Chapter 2. Read from "I wish to become an automobile salesman..." (p. 16) through "...not a friend, but a fair-weather friend." (p. 18)

    Egan, Greg. Axiomatic. In Axiomatic (Harper Prism, December 1997). Read pp. 93-94.

  2. Choose from the above selections the case that you think is most interesting. In that case, should the subject count the fact that his desire or belief resulted from taking a pill as reason to give up that desire or belief? Why or why not? Bring to class your less-than-200-word-total response.

Optional background reading:

Shoemaker, S. Desiring at will (and at pill): a reply to Millgram. Read pages 26-29.

White Roger White. Epistemic permissiveness. Philosophical Perspectives, 19, Epistemology, 2005.

Miriam Schoenfield. Permission to Believe.

Thu May 08: Decision making under indeterminacy

(Makeup class during reading period.)

Miriam Schoenfield. A Puzzle about Imprecise Value. Read sections 1-3.

Optional background reading:

Caspar Hare. Take the sugar.

J. Robert G. Williams. Decision making under indeterminacy.

Susanna Rinard. A Supervaluationist Decision Theory for Imprecise Credences.

Adam Elga | Princeton University