G.E. Moore and the Supervenience
of Normative/Evaluative
Truth
on Purely
Descriptive, Naturalistic (PDN) Truth
Two Forms Of
Ethical Naturalism:
(1) MEANING NATURALISM: Normative/evaluative terms can be defined
using only purely
descriptive, naturalistic (PDN) terms.
This is the form of ethical naturalism that
(2) METAPHYSICAL
NATURALISM:
A PDN statement can imply a
normative/evaluative statement. This is
the form of ethical naturalism that
Note that Meaning Naturalism
implies Metaphysical Naturalism, but Metaphysical Naturalism does not imply
Meaning Naturalism.
Note also that there is no
agreement that what is called the ‘Naturalistic Fallacy” is really a fallacy! That will be one of the questions that we
will critically evaluate in this course.
According to Moore, what kind
of definition of “good” does not exist?
A
definition that gives "the nature of that object or idea” (508).
Consider a proposed PDN
definition of "good"—for example:
(PDN Def.) Good for X = what
X desires to desire (512).
Ask the following
question: Is what X desires to desire
what X desires to desire? This is a
closed question.
Now ask the following
question: Is what X desires to desire
good for X?
If (PDN Def.) were a
definition of "good for X", this would also be a closed
question. But it is not. It is an open
question, because we can wonder whether it is true.
Moore claims that any
substitution of a PDN predicate for "good" will produce an open
question, thus there can be no PDN definition of "good".
THE MAIN PROBLEM WITH
MOORE’S ARGUMENT
He tries to draw a metaphysical conclusion from an
epistemological argument.
What is the epistemological conclusion of his argument?
What is the metaphysical conclusion that he draws from the
argument?
A Different Argument for the
Metaphysical Conclusion
The analogy
with "yellow".
“This lemon is yellow” is not
a definition of “yellow”. It is a predication of yellow to something. There is no PDN definition of “yellow”.
Similarly, Moore thinks that
some things are good, but there is no PDN definition of “good”.
THE SUPERVENIENCE OF THE NONNATURAL ON THE NATURAL
What is the metaphysical idea
that Moore uses to explain the analogy? Supervenience.
Moore holds that because normative/evaluative
properties or truths are NONNATURAL, there are no PDN necessary and sufficient
conditions for normative/evaluative properties or truths. But there is a relation between the two. The NONNATURAL normative/evaluative
properties or truths SUPERVENE on PDN properties and truths. Fixing all the PDN truths also fixes the
normative/evaluative truths, but there are no logically necessary and
sufficient conditions for normative/evaluative terms in PDN terms.
Note the analogy with
consciousness (e.g., the perception of yellow), which seems to supervene on
naturalistic truths about non-conscious processes and events.
What is the disanalogy with "yellow"?
Normative/evaluative
judgments are not simple perceptual judgments.
The judgments are affected by purely descriptive, naturalistic
background information.
Moore recognizes that there is a disanalogy,
but holds that supervenience applies in both
cases. Because of supervenience
Moore believes that there are no exceptionless moral principles that provide a
PDN sufficient condition for moral rightness/wrongness.
For any example of a morally wrong possible act with property
PDN1, there is another possible act with property PDN1
that is not wrong.
MACKIE'S CHALLENGE:
NO OBJECTIVE NORMATIVITY
Mackie calls his view “moral
skepticism”, but skepticism is an epistemological thesis (that we have no way
of knowing or justifiably believing moral truths, if there are any). He actually defends both an epistemological
thesis (moral skepticism) and a metaphysical thesis (moral anti-realism).
Mackie's moral skepticism (and
moral anti-realism) are second order views, not a
first order views. What is the
difference? (Is Mackie correct that the
two levels are "completely independent"(549)?)
The challenge has two parts:
I. Our ordinary moral judgments make a claim to
objective values: "objective, intrinsic
prescriptivity"(553)
A. Both ethical non-cognitivism and ethical naturalism are inadequate. Why?
B. According to
Mackie, Moore was right about the commitments of ethical language. Why?
(Mackie agrees that if there were objective moral properties,
they would have to be non-natural.
Mackie is a naturalist: He denies
that there are any non-natural properties.)
II. There are no objective values
("Error Theory")—moral anti-realism; or at least, we have no way of
knowing anything about them (moral skepticism).
Two arguments:
A. The Argument from Relativity. Not the important one.
B. The Argument from
Queerness. This is the argument that has
been most influential.
The argument has two parts:
(A) Metaphysical. Objective values would be "entities or
qualities of a very strange sort, utterly different from anything else in the
universe"(555).
Plato's Form of the Good:
the end has "to-be-pursuedness somehow
built into it"(555-556).
Objective principles of wrongness: a wrong act "would have not-to-be-doneness
somehow built into it"(556).
We will use Mackie's terms to state the problem. Objective values would require objective
to-be-pursuedness (TBP). Objective right and wrong would require
objective to-be-doneness (TBD) and objective not-to-be-doneness (-TBD).
(B) Epistemological. For us to be aware of objective
TBP, TBD, and -TBD, "it would have to be by some special faculty of moral
perception or intuition, utterly different from our ordinary ways of knowing
everything else"(555).
Intuitionism.
THE OBJECTIVIST RESPONSE:
"COMPANIONS IN GUILT"
(1) Richard Price's
list: "essence, number, identity,
diversity, solidity, inertia, substance, the necessary existence and infinite
extension of time and space, necessity and possibility in general, power, and
causation"(555).
(2) Anything else to
add? Non-moral normativity:
Rationality of belief and of action in
non-moral contexts.
Compare Mackie's discussion of hypothetical imperatives on p.
557. Do they have objective normativity?
1. Moral Realism and Anti-Realism
Moral Realism (MR): There are normative truths about what one
morally ought or ought not to do.
(There is some disagreement among moral realists on whether or not these
truths depend on one's situation). These
truths apply to all rational beings (at least, when they are in relevantly
similar situations).
Moral Anti-Realism
(MAR): There are no normative truths
about what one morally ought or ought not to do. (The advocate of MAR typically provides an
explanation of why it seems to us that there are such truths).
2. Practical Reason Realism and Anti-Realism
Practical Reason Realism
(PRR): There are normative truths about
what it is rational to do (which typically depend on one's situation). These truths apply to all rational beings (in
relevantly similar situations).
Practical Reason Anti-Realism
(PRAR) (Extreme Humeanism): There are no normative truths about what it
is rational to do. (The advocate of PRAR
typically provides an explanation of why it seems to us that there are such
truths.) PRAR implies MAR.
3. Theoretical Reason Realism and Anti-Realism
Theoretical Reason Realism
(TRR): There are normative truths about
what it is rational to believe (which typically depend on one's
situation). These truths apply to all
rational beings (in relevantly similar situations).
Theoretical Reason
Anti-Realism (TRAR): There are no
normative truths about what it is rational to believe. (The advocate of TRAR typically provides an
explanation of why it seems to us that there are such truths.)
4. Normative Anti-Realism (NAR): There are no normative truths. NAR implies TRAR, PRAR, and MAR.
OBJECTIVE PRESCRIPTIVITY
AND OBJECTIVE VALUES
A. Non-Moral Properties
Objective Non-Moral To-Be-Pursuedness [or Not-To-Be-Pursuedness]: This would be
a property of goals that it would be irrational, though not necessarily
immoral, not to pursue [or to pursue] or a non-moral constraint on the goals to
be pursued. For example the Strong or
Weak Norm of Transitivity is a potential norm of Non-Moral To-Be-Pursuedness, because it is a rational constraint on
preferences (goals).
Objective Non-Moral
To-Be-Doneness [or Not-To-Be-Doneness]: This would be a property of actions that it
would be irrational, though not necessarily immoral, to fail to perform [or to
perform], in the appropriate circumstances.
For example, the Instrumentalist Norm is a potential norm of non-moral
to-be-doneness.
B. Moral Properties
Objective Moral To-Be-Pursuedness [or Not-To-Be-Pursuedness]: This would be
property of goals that everyone morally should [or should not] pursue. For example, act utilitarians believe that
the goal of maximizing overall utility is a moral goal that everyone should
pursue.
Objective Moral
To-Be-Doneness [or Not-To-Be-Doneness]: This would be a property of actions that
everyone morally should perform [or should not perform]. For example, Kant thought that his
categorical imperative was a moral norm that all rational agents should obey,
regardless of whether they had any inclination to do so.
C. Epistemological Properties
Objective To-Be-Believedness [Not-To-Be-Believedness]: This would be
a property beliefs that everyone should believe [or
should not believe] in the appropriate circumstances. For example, the Law of Non-Contradiction is
a potential norm of not-to-be-believedness.
Hume’s Practical Reason Anti-Realism
According to Hume, what can
reason do?
(1) It judges relations of
ideas through abstract reasoning.
(2) It judges relations of
cause and effect through reasoning based on experience.
“Reason is the discovery of
truth or falsehood. Truth or falsehood
consists in an agreement or disagreement either to the real relations of ideas, or to real
existence and matter of fact”(503).
According to Hume, what can't
reason do?
(1) "Reason alone can
never be a motive to any action of the will"(501).
(2) Reason "can never
oppose passion in the direction of the will"(501).
What is Hume's argument for
(1)? Do you agree?
What is Hume's argument for
(2)? Do you agree?
"Reason
is, and ought only to be the slave of the passions, and can never pretend to
any other office than to serve and obey them"(501).
According to Hume, there are
only two (inexact) senses in which a passion can be called unreasonable:
(1) when
founded on the supposition of the existence of objects that do not really
exist;
(2) when
means are chosen that are insufficient to the designed end.
But don't these examples show
that reason can, at least, oppose the passions?
Does this sound like a slave?
“The moment we perceive the
falsehood of any supposition or the insufficiency of any means, our passions
yield to our reason without any opposition”(502).
Millgram
on Hume
Hume is generally regarded as
an instrumentalist about practical
reason—that is, that the only kind of practical reasoning is means-end
reasoning.
[Is this the only kind of
practical reasoning that Hume discusses?
What about matters of existence?]
Millgram argues that this is a mistake. Hume should be understood as denying that
there is any kind of practical reasoning at all. Hume is a skeptic about practical reason (or
Practical Reason Anti-Realist).
What is Millgram's
explanation of Hume's skepticism?
Hume's psychology allowed
mental states to "have either [propositional] contents or motivational
force, but not both"(69).
Mental states are either a
kind of picture (and thus can be true or false) or a kind of feeling (and thus
can have motivational force. But no
feelings have pictures pictures and no pictures have
feelings.
Hume the Practical Reason Skeptic [Anti-Realist]
"'Tis not contary to reason to
prefer the destruction of the whole world to the scratching of my finger. 'Tis not contrary to reason for me to chuse
my total ruin, to prevent the least uneasiness of an Indian or person wholly unknown to
me. 'Tis as
little contrary to reason to prefer even my own acknowledg'd
lesser good to my greater, and have a more ardent affection for the former than
the latter"(502).
Hume's Moral Noncognitivism
Cognitivism with respect to a certain kind of discourse is the
view that the discourse is propositional (and thus that statements in the
discourse are true or false).
Non-Cognitivism
with respect to a certain kind of discourse is the view that the discourse is
not propositional (and thus that statements in the discourse are neither true
nor false). Hume is generally regarded
as a non-cognitivist.
Hume's main claim: Moral judgments are motivating. Reason is motivationally inert. Therefore, moral judgments cannot be the
product of reason alone.
Hume's examples: the ungrateful sapling; incest; willful
murder.
Hume's Argument Against the
Metaphysical Version of the So-Called Naturalistic Fallacy
There is no way of deriving an ought
(normative judgment) from an is
(purely descriptive judgment) [without some premise connecting the two and
there are no such premises that are true].
Korsgaard's
Defense of Reason as a Source of Motivation
What does K mean by skepticism about practical reason?
What is the distinction
between Motivational Skepticism and Content Skepticism about practical reason?
Classical Internalism (Falk, Frankena,
Nagel): Knowledge (or the truth and
acceptance) of a moral judgment implies the existence of a motive (not
necessarily overriding) for acting on that judgment.
Korsgaard's InternalismK
Requirement on practical reasons:
Practical-reason claims, if they are really to present us with reasons
for action, must be capable of motivating rational persons. (Note:
This is not what Williams means by internalismW
about practical reasons.)
Does Korsgaard
think that classical internalism is true? Hint: Does
the Korsgaard Internalism
Requirement imply that rational considerations always succeed in motivating us?
Korsgaard’s
Key Idea
All reasons “motivate” us to
the extent that we are rational.
Theoretical reason (reasons
for belief): Consider the “force” that
we feel to change a belief when we recognize a reason for change. Call this quasi-motivation.
Recognizing a reason for
belief produces conviction (call this motivation
to believe) in a theoretically rational person.
Consider the change in
motivation that comes from recognizing that coming to class is necessary to get
an A in this course, when you desire to get an A in this course.
Recognizing an instrumental
or hypothetical reason produces motivation
in a practically rational person.
Korsgaard and the desire to be rational. Consider the analogous internalism requirement on theoretical reason.
Sensitivity to Reasons and Responsiveness to Reasons
Suppose that there is a
reason R for X to do act A and X recognizes that R is a reason for X to do A.
Classical internalism
implies that X will be motivated to do A.
Korsgaard wants to help us to realize that this is not so. It is possible to recognize that R is a
reason to do A (this is a kind of sensitivity
to reason R) but not to be motivated to do R (motivation is produced by responsiveness to reason R), if one is
not rational.
True irrationality: “failure to respond
appropriately to an available reason.”(12)
A truly (I wish she had said deeply)
irrational person would be sensitive
to reasons but not responsive to them.
We can imagine this kind of true (deep) irrationality for theoretical
reasons and for nonmoral practical reasons.
What does Korsgaard
mean by saying “the necessity is in the law, and not in us”(25/91)?
Korsgaard thinks we will agree that true (deep) irrationality
is possible and, thus, that rationality requires both sensitivity and
responsiveness to reasons. But the
necessity is not in us, because we could lack responsiveness.
By analogy, if there are
moral reasons, recognizing them would produce motivation in a reasonable
person. This would be an operation of pure practical reason (because not based
on the contents of our SMS).
So the issue cannot be
whether reason can produce motivation (it can), but whether there are moral
reasons that apply to all rational beings regardless of the contents of their
SMS.
Williams on Internal and External Reasons
1. Subjective Motivational Set (SMS) = desires,
dispositions of evaluation, patterns of emotional reaction, personal loyalties
and commitments (105).
2. Sound Deliberative Route. For Williams, practical reasoning includes
instrumental reasoning (sub-Humean model), but also
includes other kinds of reasoning, including:
thinking how the satisfaction of elements in S can be combined;
resolving conflicts of motivation; and finding constitutive solutions (104).
3. InternalW
Reasons.
Consider the agent’s subjective motivational set, SMS. Remove from it any member D that is dependent
on a false belief; or that is dependent on another member that is dependent on
a false belief.
Call this the subject’s corrected
subjective motivational set, CSMS.
(Def) A has an internalW reason
to Ф iff There are one or more members of A's CSMS
from which there is a sound deliberative route to Ф-ing.
Williams’s InternalismW
about Practical Reason
Define an externalW reason as one that is not internalW.
Williams’s internalismW about practical reasons: All practical reasons are internalW
in the above sense. There are no externalW reasons.
Why does Williams think that
the example of Owen Wingrave makes it plausible that
all practical reasons are internalW in his
sense?
The key idea: Reason can produce motivation, but only when
there is something in the agent’s SMS to deliberate
from, to reach the new motivation (109).
ExternalW reasons statements are all false, because they depend
on reason being able to produce motivation that is not related (by a sound
deliberative route) to motivation already in the agent’s SMS.
Williams’s Picture of
Practical Reasoning
The analogy to reasoning
about belief: Practical reasoning is
like inferential reasoning. Both are
governed by hypothetical principles. In theoretical reasoning, modus ponens is a hypothetical
principle, because it only tells you that if
you believe p and (if p then q) that you should believe q. Similarly, the instrumental principle is a hypothetical principle, because it only
tells you that if you desire E or have E as an end, that you should take the
appropriate means to achieve E. For
Williams, all practical reasoning is hypothetical, though not necessarily
instrumental.
According to Williams, practical
reasoning is hypothetical, though not necessarily instrumental, because takes
us from an agent’s original SMS (SMS1) to a new SMS (SMS2)
that is conditional on the contents of the original SMS1. (Of
course, it eventually produces actions.)
The Issue Between
Williams and Korsgaard (Kant)
Williams and Korsgaard agree
that reason can produce motivation, if it has something with motivation to
start with. Williams rejects Hume’s PRAR
(108).
Korsgaard thinks reason can do something more. Korsgaard thinks
that there are categorical reasons for action, reasons that don’t depend on the
contents of one’s SMS.
Continue the analogy with
reasons for belief: Some people think
that there are some beliefs that it is rational for us to believe regardless of
what other beliefs we have. Reason can give
us some necessarily true beliefs (e.g., Law of Non-Contradiction), beliefs that
we do not need to infer from any other beliefs.
No matter what else one believes, it is not rational to believe a
contradiction. So the Law of Non-Contradiction would be a categorical principle of theoretical reason.
Kant/Korsgaard
thinks that practical reasoning can give us categorical reasons and, if we are
rational, motivation (to act in accordance with the categorical reasons),
independent of the actual content of our SMS.
Williams denies this.
[Note that Kant/Korsgaard think that we must have some motivation for reason to work on. Kant also
believes that we must have some beliefs for it to be rational for us to believe
the Law of Non-Contradiction. But it
does not matter what those beliefs are.
If we have any rational beliefs at all, then it is rational for us to
believe the Law of Non-Contradiction.]
Categorical practical reasons
are practical reasons that do not depend on the contents of the agent’s SMS. This is Kant’s conception of the categorical
imperatives of morality.
Main Issue between Williams
and Korsgaard:
Are there categorical practical
reasons that apply to all rational beings regardless of the contents of their
SMS?
NAGEL ON OBJECTIVE VALUES:
THE METHOD OF OBJECTIVE
REFLECTION
I. The Difference Between
the Personal and the Impersonal Point of View.
II. Issue: Not are there objective
normative entities (Plato's Forms), but are there objective normative
reasons?
"Whether what we have
reason to do or want can be determined from a detached standpoint toward
ourselves and the world”(135).
III. Nagel's Epistemology: Not Proof or Refutation, but Normative
Explanation and Consideration of What is Most Plausible (i.e., What It Makes
the Most Sense to Believe).
Objective Values and
Objective Reasons
The key idea: To believe in objective values is not to
believe in a weird kind of entity, but to believe in objective reasons for
action.
Kinds of reasons:
(a) broad
vs. narrow
(b) external
vs. internal. What is Nagel’s conception
of an internalN
reason? “It depends on the existence of
an interest or a desire in someone”(136).
Nagel is an internalistN, because he believes that all
practical reasons are internalN
reasons. Note that internalismN
is different from classical internalism, internalismW, and internalismK. You are responsible for understanding those
three forms of internalism, but not internalismN.
Two Kinds of Objective Reason
Agent-neutral reason: "If a reason can be given in a general
form which does not include an
essential reference to the person to whom it applies, it is an agent-neutral reason"(136).
Agent-relative reason: "If on the other hand the general form
of a reason does include an essential
reference to the person to whom it applies, it is an agent-relative reason"(136).
Example of
a headache or other pain.
1. Is pain at least an agent-relative
(dis)value?
2. Is pain an agent-neutral (dis)value? Why does Nagel think it is self-evident? What does Nagel think is crazy?
The commitment to
objectivity: In reasoning, whether
theoretical, practical, or moral, we typically assume there is a correct answer
that we can be mistaken about. To
vindicate objectivity, we must try to understand what it is that we might be
mistaken about.
How does Nagel respond to
Mackie?
Hampton's Defense of
Objective Normativity
Hampton's
"companions in guilt" strategy.
Recall that theoretical
reason seems to be able to come up with categorical norms of epistemic rationality
that apply to everyone regardless of what they actually believe (e.g., the Law
of Excluded Middle or the Law of Non-Contradiction)
Recall that Williams denies
that there are any categorical norms (or normative truths) of practical reason
that apply to everyone regardless of the contents of their SMS. According to Williams, in practical
reasoning, as illustrated by instrumental reasoning, the results depend on what
we start with (the contents of the SMS).
Practical reasons are hypothetical, not categorical.
Hampton wants to show us that
most instrumentalists (not Williams) are mistaken to think that they can avoid
being committed to any objective norms (or truths) of practical reason and even
Williams is mistaken to think that he can avoid being committed to any
categorical norms (or truths) of practical reason. Instrumentalism itself is committed to
non-instrumental, categorical norms of practical reason.
Hampton’s Argument:
(1) Expected Utility Theory
is the best realization of the Instrumental Theory.
(2) Other attempts to cast
doubt on Expected Utility Theory Fail:
(a) Rescher and deliberation about
ends.
(b) Ruddick on maternal thinking.
(c) Slote on satisficing rather
than maximizing.
Bounded rationality
literature does not support Slote.
(d) Frank on commitment.
His defense of commitment (and thus morals) is instrumentalist.
(e) Gauthier’s “constrained maximization” and McClennen’s “resolute choice”. Also an instrumental
defense of morality.
(3) Hampton’s Two Criticisms
of Instrumentalism from within Expected Utility Theory.
(a) Even Instrumentalism
requires objective norms of practical reason.
Note what she calls the Consequentialist Norm, we will call the Instrumentalist Norm: “Act so as to perform the most effective
means to a desired end”(233).
What is the point of the
example of the curmudgeon?
Why does Hampton think that
the instrumental model of practical reasoning has seemed “metaphysically
benign” (229)?
(b) Hampton’s Deeper
Point: EUT requires non-instrumentalist
(and, we would add, non-hypothetical) objective norms of practical reason.
“To be able to reason
instrumentally, we must be able to reason non-instrumentally”(232). What does this mean?
[Note that Hampton
misinterprets Williams as denying that we can ever criticize an act as
irrational if the agent is not motivated to do the act (233).]
Conclusion: There are objective norms of practical reason
and some are not instrumental (I would say “not hypothetical”): “A norm purports to give us [or anyone]
reason to act, choose, or believe [or, I would add, to change our beliefs or
preferences] as the norm directs no matter what other reasons or motives we
have, where this reason is supposed to be decisive in some circumstances”(228).
Non-Moral Norms of To-Be-Pursuedness (Rational Constraints on Preferences, Desires,
or Goals)
Strong Norm of
Transitivity: One ought not to have intransitive
preferences—for example, one ought not to have preferences of the following
kind: A > B, B > C, and C > A.
Weak Norm of Transitivity: If one
discovers an intransitivity in one's preferences, one
ought to eliminate it.
In-Class Questions
In
answering the following question, you should suppose that you have been
diagnosed with cancer and are given a choice between two treatment options,
surgery (S) or radiation therapy (R).
You should rank the two options based solely on the information
provided. The three possibilities
are:
(1)
S > R (you prefer surgery to radiation therapy); (2) R > S (you prefer
radiation therapy to surgery) or (3) S = R (you are indifferent between the two
forms of treatment).
1. You have a choice between Surgery (S) or Radiation Therapy (R):
Surgery: Of 100 people having surgery 90 live through
the post-operative period, 68 are alive at the end of the first year and 34 are
alive at the end of five years.
Radiation
Therapy: Of 100 people having radiation
therapy all live through the treatment, 77 are alive at the end of one year and
22 are alive at the end of five years.
In answering the following
question, you should suppose that you have been diagnosed with cancer and are
given a choice between two treatment options, surgery (S) or radiation therapy
(R). You should rank the two options
based solely on the information provided.
The three possibilities are:
(1) S > R (you prefer
surgery to radiation therapy); (2) R > S (you prefer radiation therapy to
surgery) or (3) S = R (you are indifferent between the two forms of treatment).
2. You have a choice between
Surgery (S) or Radiation Therapy (R):
Surgery: Of 100 people having surgery 10 die during
the surgery or the post-operative period, 32 die by the end of the first year
and 66 die by the end of five years.
Radiation Therapy: Of 100 people having radiation therapy, none
die during treatment, 23 die by the end of one year and 78 die by the end of
five years.
1. You have a choice between Surgery (S) or Radiation Therapy (R):
Surgery: Of 100 people having surgery 90 live through
the post-operative period, 68 are alive at the end of the first year and 34 are
alive at the end of five years.
Radiation Therapy: Of 100 people having radiation therapy all
live through the treatment, 77 are alive at the end of one year and 22 are
alive at the end of five years.
2. You have a choice between
Surgery (S) or Radiation Therapy (R):
Surgery: Of 100 people having surgery 10 die during
the surgery or the post-operative period, 32 die by the end of the first year
and 66 die by the end of five years.
Radiation Therapy: Of 100 people having radiation therapy, none
die during treatment, 23 die by the end of one year and 78 die by the end of
five years.
Strong Norm of Invariance Under Equivalent Descriptions: Let D1 be a
description of a choice situation of agent S.
Let D2 be an alternative description which is equivalent to D1 (i.e.,
they both contain the same information, simply stated differently). S's preferences should be invariant over the
two descriptions. For example, the
following is irrational: (1) Given only information
D1 about a choice satiation, S would prefer A to B; and (2) Given
only information D2 about a choice situation, S would prefer B to A.
Weak Norm of Invariance Under Equivalent Descriptions: Let D1 be a
description of a choice situation of agent S.
Let D2 be an alternative description which is equivalent to D1 (i.e.,
they both contain the same information, simply stated differently). If S knows that the two descriptions are
equivalent and that his/her preferences are not invariant over the two
descriptions, S should change his/her preferences to preserve invariance.
2-PERSON PRISONERS'
DILEMMA
(Higher numbers
represent more preferred outcomes.)
B. Sequential 2-Person Prisoners' Dilemma
3,3
C
C
D
1, 4
4,1
C
D
Player #1 Player #2 D
2,2
The
Sequential Two-Person Prisoners' Dilemma.