|
Rationality in Action
The Basics
This is the offical course page for the specialization course "Rationality in Action", spring 2009, Section for Philosophy and Science studies, Roskilde University.
Students taking this course are adviced to keep attention to this page since it will provide all relevant information, updates and material. The page will be continuously under construction until February 1.
Instructor: Pelle G. Hansen
Time and location: Fridays 9:15 - 12:15 (10:15-11:15 tutorial session), Theory room PA 6. Except the first lecture which will be at the same time, but Wednesday, February 4.
Textbooks
- Choice Theory: A Very Short Introduction, Michael Allingham, Oxford University Press (2002)
- A Guide to Game Theory, Fiona Carmichael, Prentice Hall (2004).
Course Description (in Danish)
Målsætningen i dette kursus ville være i gennem 6 kursusgange, at udstyre de studerende med et grundlæggende kendskab til begrebet om instrumentelt rationel handling, herunder dets forcer og begrænsninger i praktiske såvel som teoretiske sammenhænge. Kurset vil tage udgangspunkt i en karakteristik af begrebet om instrumentel rationalitet sammenlignet med andre såkaldte `rationalitetsformer'. Grundlæggende elementer og problemstillinger i teorien om rationelle valg, ofte også kaldet beslutningsteori, gennemgås og teoriens begrænsninger afsøges i en diskussion af teoriens specielle antagelser og medførende paradokser. Herefter gennemgås de grundelementer der er nødvendige for at give de studerende en rudimentær forståelse af teorien om interaktive beslutninger; dvs. spilteori. Igen afsøges denne teoretiske tilgangs begrænsninger og forcer i en diskussion af teoriens specielle antagelser og medførende paradokser. Jeg vil naturligvis afholde kurset i en underholdende stemning, hvor de studerende som en del af undervisningen blandt andet skal forvente at deltage i fælles øvelser med henblik på at opleve og forstå teoretiske pointer i praksis. Videre vil der være et stærkt fokus på at disse pointer bliver diskuteret i sammenhæng med velkendte problemer inden for praktisk filosofi.
[foreløbig information om eksaminationsgrundlag]
Lecture no. 1: Rationality and Action
[slides], [handout #1], [score-sheet #1]
In this, the first lecture, we begin by looking at the general aim and content of the course. We then set out to discuss some foundational issues in the theory of action: what is it for something to be an 'action' in the first place; what is it to describe an action; how is explaining an action different from describing an action? In discussing these issues we touch upon themes such as the regress problem in explaining an action, basic actions, volitional vs. causal theories of action, actions as events, action and intentionality. Ultimately we arrive at an idea about actions as intentional phenomena. I will then argue that a close relationship obtains between intentional actions and instrumental rationality. This quite natural have us attending to the details of the idea of instrumentally rational actions and the roles played by pro-attitudes and beliefs in such actions. In particular, we will use your answers for this week's assignment to illustrate the points to be made in this connection.
If successfully, this argument provides the entry point for our interest in the theory of rational choice. However, instead of entering the theory of rational choice by means of blackboard and chalk, we do this by first playing a series of decision games. In this connection each of you will be handed a sum of course-money and score-sheets that you will have to administer for the rest of the course. Course-money will be used throughout the course as prize-money in the various exercises and games played. As you progress through the course they will, however, also turn out to have further purposes yet to be disclosed. After having played the decision games we then go through the basic ideas of rational choice theory on the blackboard.
Notice! For the first lecture you will have to write down no less than half an A4 page explaining why you chose to take this course and what you aim to achieve by it. Please also include your name, what your other master-subject is, titles of projects you have written and, if you like, also a picture of yourself. Consider also when writing it, that the assignment will be read by your fellow students and discussed in class. Finally, it has to be written in word, rtf or some other basic format usually accessible on PC and e-mailed to pgh@ruc.dk no later than Tuesday 9 a.m., 3rd of February.
Also, please read in
Choice Theory: A Very Short Introduction, Michael Allingham, Oxford University Press (2002)
Also, from late Monday morning the 2nd of February the following texts suggested for further reading is to be found for you to copy in my pigeonhole:
-
Davidson, Donald (1963) 'Actions, reasons and causes' in Donald Davidson (1980) Essays on Actions and Events, Clarendon Press, p. 3-20.
Finally, I ask you to consider and respect the cooperative nature of our common venture. That means:
-
Please respect that while classes begin at the time announced, they may go on a bit longer than originally planned, why rushing off to catch the train, packing bags while people are still discussing, and the like will not be tolerated. Instead, please leave during one of the breaks if you have other priorities to attend to.
Lecture 2: Characterizing Rational Choice
[slides], [Monty Hall Problem], [American conception of genius]
In this, the second lecture, we begin with a small written test on what we went through in the first lecture. After the test we summarize the points of the first lecture by means of Q&A on basis of the test - this will provide an opportunity for gaining as well as loosing course-money. Obviously, then, I expect everyone to have looked through the points made in the first lecture, such as:
1. To what objects may we ascribe the property of being rational?
2. What is a rational action?
3. Are actions always events?
4. What is a causal theory of action, and what problems are
associatiated with such a theory?
5. What is a volitonal theory of action, and what problems are
associatiated with such theories?
6. What is it for something to be intentional?
7. What is difference between describing and explaining an action?.. and more
We then begin to look into what characterizes rational choice. Please notice that what we go through in this lecture is on the one hand necessary for understanding most of what will go on in the rest of the course, and on the other hand, not particularly stimulating in every detail given our philosophical point of view.
First we go over the basic axioms characterizing 'reasonable choice'; i.e. Sen's alpha property and the Condorcet condition. Then we define a preference relation and discusses what kind of 'thing' a preference relation is and what it explains. Having done this we follow Allingham in introducing Samuelson's revealed preference condition so as to characterize rational choice by means of transitive preferences which is equivalent to saying that choice is rational if and only if it can be explained by a preference ordering. We then show that if choice is rational in this sense it may be represented as maximizing ordinal utilities. However, please take some care in noticing that such utility is derived from choice and not vice versa, as well as that such utility assigment does not provide for interpersonal comparisons.
We then go over some basic rules of probability theory in order to define the notions of a gamble, a degenerate gamble and a compund gamble. With this in hand we turn to rational choice in the context of uncertainty. We first notice that while we cannot talk about rational preferences in the context of certainty, this indeed seems to be possible in the context of uncertainty. We then make precise our intuitions in terms of the substitution condition and the Achimedian condition and define any preference ordering over gambles satisfying these conditions as rational. With the aim of gaining a nice way to analyze decision problems we set out to develop our understanding of expected utility. Where Allingham elegantly skips how expected utilities are assigned to degenerate gambles we pause to give it some consideration. Also, the contrast between cardinal and ordinal utility is emphasized, before concluding that preferences over gambles are rational if and only if they may be represented by an expected utility function.
Also, please read in
Choice Theory: A Very Short Introduction, Michael Allingham, Oxford University Press (2002)
Chapter 3: Racing and roulette, 28-49
Chapter 4: Gambling and insurance, 50-65.
Lecture 3: Game theory
[slides], [summmary of two first lectures], [Handout: Rousseau's Stag-hunt]
In this, the third lecture, we begin by summarizing what we've learned during the first two lectures. In particular, we go over the way utility functions are constructed once again to make sure that misunderstandings do not arise when working with the theory of games. Also we discuss the connection or lack thereoff between consequentialism in moral philosophy and utility theory and the extent to which moral philosophers could benefit from studying decision theory more closely.
We then do some introductory exercises beginning with an auction for a Café au Lait. Then we play 'Rock, Paper, Scissors' and 'pick a hand'.
After the exercises we ask the question "What is game theory?". A brief historical overview is provided and the choice of Fiona Carmichael's book is motivated. Then we go through the elements constituting the notion of a game: interdependent decision or strategic uncertainty, players, actions, preferences, and solutions. To make sure we got the idea we do some basic exercises in identifying these elements in some games - 'pick a hand', 'Rousseau's stag-hunt' and some situation you experienced, among others.
After a break we next go over the different categories and representations of, and distinctions in between, games: static games and dynamic games; bimatrix representations and extensive representations; repeated games; constant-sum, mixed motive and coordination games; and, finally, non-cooperative and cooperative game theory. As a philosophical perspective we pick out and briefly look at the kind of social contract theories given by Hobbes, Rawls, and Binmore. Again to make sure we got the basics right we do some basic exercises having us construct game representations of 'pick a hand' and 'Rousseau's stag-hunt'.
Notice: You are given the assignment of trying to construct a game representation of the game-like situation you reported to have experienced. This assignment is due Thursday 26. February, 9.a.m. Please mail it to pgh@ruc.dk.
Finally, we do an exercise replicating Thomas C. Schelling's famous focal-point experiment and discuss how it is that people are often able to coordinate in this experiment.
Notice: Everyone agreed to send Casper their results from their introductory exercises during the weekend.
Read:
A Guide to Game Theory, Fiona Carmichael, Prentice Hall (2004).
Lecture 4: Prisoner's dilemma games and dynamic games
[slides]
Read:
A Guide to Game Theory, Fiona Carmichael, Prentice Hall (2004)
Also, from late Monday morning the 23rd of February the following texts suggested for further reading will be available:
-
Kavka, Gregory S. (1999) 'Hobbes's War of All against All' in Christopher W. Morris (ed.) The Social Contract Theorists: critical essays on Hobbes, Locke and Rousseau, Rowman & Littlefield, p. 1-22.
-
Osborne, Martin J. (2004) 'Studying Nash equilibrium experimentally' & 'Experimental evidence on the Prisoner's dilemma' both in An Introduction to Game Theory, Oxford: Oxford University Press, 24-26, 28.
Lecture 5: Expected Utility Theory & Mixed Strategies & Evolution
[slides]
In this lecture we begin by looking at some case examples where the concepts and techniques we have discussed so far gives new insights to philosophical doctrines. We then look at expected utility theory in the context of risky choices and train our ability to calculate expected utility for decisions involving risk. In particular, we look at some of the issues expected utility raise in relation to choosing between possible ways for organizing society and distributing social goods.
After a break we go through a series of games together and discusses their solutions according to the different solution concepts we have already seen.
Finally we dive into the subject of mixed strategies as a stepping stone to a discussion of the equilibrium selection problem. After explicating the nature of this problem and its consequences for rational choice, we look at the motivation for taking up evolutionary approaches to social interaction and discusses these approaches in the context of social and moral institutions, respectively.
Notice: I suggest that those of you who find it difficult to follow every step of Section 5.3.2 read it over without struggling too much and get from it what they can. I will not expect of you to understand the details, but wish that you notice the critical points about expected utility theory raised. So don't despair.
Notice: I have suggested that you all hand in your working-hypothesis for the essay by mail Thursday morning 9 a.m. at the latest. This is not obligatory, but to those that do it will be able to get comments on it Friday after the lecture.
Read:
A Guide to Game Theory, Fiona Carmichael, Prentice Hall (2004)
Lecture 6: Philosophical topics
Read:
A Guide to Game Theory, Fiona Carmichael, Prentice Hall (2004)
Notice: You are only expected to browse over Chapter 8.
Examination
|
