Confirmation and the ravens paradox 1 Seminar 3: Philosophy of the Sciences Wednesday, 21 September 2011 1.

Slides:



Advertisements
Similar presentations
General Argument from Evil Against the Existence of God The argument that an all-powerful, all- knowing, and perfectly good God would not allow any—or.
Advertisements

Hume’s Problem of Induction 2 Seminar 2: Philosophy of the Sciences Wednesday, 14 September
Hoare’s Correctness Triplets Dijkstra’s Predicate Transformers
CAN I KILL MY YOUNGER SELF? Time Travel and the Retro-Suicide Paradox Peter B. M. Vranas The University of Michigan 15 September 2000.
Best Practice Precepts [... next] Arguments Arguments Possibility of the Impossible Possibility of the Impossible Belief, Truth, and Reality Belief, Truth,
Chapter 1 Critical Thinking.
Abby Yinger Mathematics o Statistics o Decision Theory.
NOTE: CORRECTION TO SYLLABUS FOR ‘HUME ON CAUSATION’ WEEK 6 Mon May 2: Hume on inductive reasoning --Hume, Enquiry Concerning Human Understanding, section.
The Problem of Induction Reading: ‘The Problem of Induction’ by W. Salmon.
Critical Thinking: Chapter 10
Introduction/Hume’s Problem of Induction Seminar 1: Philosophy of the Sciences 6 September
7 INVERSE FUNCTIONS.
QM Spring 2002 Business Statistics Introduction to Inference: Hypothesis Testing.
Common knowledge: application to distributed systems Caesar Ogole, Jan Gerard Gerrits, Harrie de Groot, Julius Kidubuka & Stijn Colen.
Me Talk Good One Day When Language and Logic Fail to Coincide.
March 10: Quantificational Notions Interpretations: Every interpretation interprets every individual constant, predicate, and sentence letter of PL. Our.
Lec 6, Ch.5, pp90-105: Statistics (Objectives) Understand basic principles of statistics through reading these pages, especially… Know well about the normal.
Lecture 6 1. Mental gymnastics to prepare to tackle Hume 2. The Problem of Induction as Hume argues for it 1. His question 2. His possible solutions 3.
The Problem of Induction Reading: ‘The Problem of Induction’ by W. Salmon.
Mathematical Induction Readings on induction. (a) Weiss, Sec. 7.2, page 233 (b) Course slides for lecture and notes recitation. Every criticism from a.
1 Arguments in Philosophy Introduction to Philosophy.
RESEARCH IN EDUCATION Chapter I. Explanations about the Universe Power of the gods Religious authority Challenge to religious dogma Metacognition: Thinking.
C OURSE : D ISCRETE STRUCTURE CODE : ICS 252 Lecturer: Shamiel Hashim 1 lecturer:Shamiel Hashim second semester Prepared by: amani Omer.
POSC 202A: Lecture 1 Introductions Syllabus R Homework #1: Get R installed on your laptop; read chapters 1-2 in Daalgard, 1 in Zuur, See syllabus for Moore.
Can Big Questions Be Begged?. Fallacies are mistakes in inference, BUT Begging the question is not a mistake in inference. Is it a fallacy at all? Robinson.
MGF 1107 Mathematics of Social Choice Part 1a – Introduction, Deductive and Inductive Reasoning.
Introduction to Proofs
1 Chapter 7 Propositional and Predicate Logic. 2 Chapter 7 Contents (1) l What is Logic? l Logical Operators l Translating between English and Logic l.
Basics of Probability. A Bit Math A Probability Space is a triple, where  is the sample space: a non-empty set of possible outcomes; F is an algebra.
APPLICATIONS OF DIFFERENTIATION Indeterminate Forms and L’Hospital’s Rule APPLICATIONS OF DIFFERENTIATION In this section, we will learn: How to.
1 Derivation Schemes for Topological Logics. 2 Derived Logics What Are They? Why Do We Need Them? How Can We Use Them? Colleague: Michael Westmoreland.
Chapter 1 Logic Section 1-1 Statements Open your book to page 1 and read the section titled “To the Student” Now turn to page 3 where we will read the.
1 Knowledge Representation. 2 Definitions Knowledge Base Knowledge Base A set of representations of facts about the world. A set of representations of.
LOGIC AND ONTOLOGY Both logic and ontology are important areas of philosophy covering large, diverse, and active research projects. These two areas overlap.
Copyright © Curt Hill Mathematical Logic An Introduction.
“Facts are not science – as the dictionary is not literature” –Martin H. Fischer If science is not facts, what is it?
1 Science!. 2 Science Suppose you knew nothing about science. How would you explain how it rains? Suppose someone did not believe your explanation. Could.
Invitation to Critical Thinking Chapter 9 Lecture Notes Chapter 9.
READING #4 “DEDUCTIVE ARGUMENTS” By Robert FitzGibbons from Making educational decisions: an introduction to Philosophy of Education (New York & London:
Critical Thinking. Critical thinkers use reasons to back up their claims. What is a claim? ◦ A claim is a statement that is either true or false. It must.
PHILOSOPHY OF LANGUAGE Some topics and historical issues of the 20 th century.
HEMPEL’S RAVEN PARADOX A lacuna in the standard Bayesian solution Peter B. M. Vranas Iowa State University PSA’02, 8 November 2002.
The Problem of the External World Kareem Khalifa Philosophy Department Middlebury College.
Philosophy of science What is a scientific theory? – Is a universal statement Applies to all events in all places and time – Explains the behaviour/happening.
Proof And Strategies Chapter 2. Lecturer: Amani Mahajoub Omer Department of Computer Science and Software Engineering Discrete Structures Definition Discrete.
WHAT IS THE NATURE OF SCIENCE?
Chapter 7: Induction.
The Foundations: Logic and Proofs
PHI 208 Course Extraordinary Success tutorialrank.com
Computer Science cpsc322, Lecture 20
DISCRETE MATHEMATICS CHAPTER I.
Infinitesimal Confirmation
The second Meeting Basic Terms in Logic.
If You Aren’t Dong Arguments, You Aren’t Doing Evidence
Chapter 3 Philosophy: Questions and theories
Today’s Topics Universes of Discourse
Confirmation The Raven Paradox.
The Foundations: Logic and Proofs
Arguments.
The Paradox of the 99-foot Man
Inductive and Deductive Logic
Theory & Research Dr. Chris Dwyer.
Computer Security: Art and Science, 2nd Edition
POSC 202A: Lecture 1 Introductions Syllabus R
Hypothesis Testing and Confidence Intervals (Part 2): Cohen’s d, Logic of Testing, and Confidence Intervals Lecture 9 Justin Kern April 9, 2018.
Computer Science cpsc322, Lecture 20
EQ: What is the goal of science?
INFORMATIVE ABOUTNESS
A POCKET GUIDE TO PUBLIC SPEAKING 5TH EDITION Chapter 24
Presentation transcript:

Confirmation and the ravens paradox 1 Seminar 3: Philosophy of the Sciences Wednesday, 21 September

Required readings Peter Godfrey Smith. Theory and Reality. Section (can be downloaded from HKU library) Clark Glymour ‘Why I am not a Bayesian’ (on course website) 2

Optional readings Paul Horwich ‘Wittgensteinian Bayesianism’ (on course website) J. A. Cover and Martin Curd ‘Commentary on confirmation and relevance’, section 5.1, pp (on course website) Hawthorne and Fitelson. ‘The paradox of confirmation’, Philosophy Compass pp (can be downloaded from HKU library) 3

Tutorials Tutorials will start on this Friday 23 September Class 1: 1 PM - 2 PM seminar room 305 Class 2: 4 PM – 5 PM seminar room 305 Required reading: ‘The Problem of Induction’, Section I, Chapter 7 of Richard Feldman’s book Epistemology pp (on course website) Required reading and seminar handouts must be brought along to tutorials 4

Questions to be addressed Q1) What is it for evidence E to confirm hypothesis H (that is, what is it for E to be evidence in favour of H)? Q2) Which propositions confirm which propositions? Q3) Which propositions do scientists and ordinary people take to confirm which propositions? Note: Skeptics like Hume might still be interested in Q3 5

Instances of Q2 Does a being a black raven confirm all ravens are black? Does a being a white shoe confirm all ravens are black? Does there was scratching noises coming from the cupboard last night and the cheese in the cupboard has now disappeared confirm the cheese was eaten by a mouse? Do our data concerning changes in temperature and climate confirm the theory of man-made global warming? 6

The instantial theory of confirmation (ITC) For any predicates F and G, and any name a, i) Fa.Ga confirms  x(Fx  Gx) (All Fs are Gs); and ii) Fa.~Ga disconfirms  x(Fx  Gx). Def: ‘.’ means ‘and’ Problem 1 with ITC: There is no obvious way to extend ITC to deal with plausible cases of confirmation like c) and d) described on slide 6. 7

Problem 2: The ravens paradox (Carl Hempel) Let ‘R’ symbolise ‘is a raven’, and ‘B’ symbolise ‘is black’. (1)By ITC, ~Ba.~Ra confirms  x(~Bx  ~Rx) (2)  x(~Bx  ~Rx) is necessarily equivalent to  x(Rx  Bx) (3) By (1), (2) and (EQ), ~Ba.~Ra confirms  x(Rx  Bx) 8

The ravens paradox (cont) (EQ) If E confirms H1, and H1 is necessarily equivalent to H2, then E confirms H2 (PC) therefore follows from (ITC) and (EQ). But how can a white shoe provide evidence that all ravens are black?? (PC) That a is non-black and non-raven confirms that all ravens are black 9

A response to the paradox Since ‘(  xFx).(  x(Fx  Gx))’ is a better symbolisation of ‘All Fs are Gs’ than ‘  x(Fx  Gx)’, ITC should be replaced with ITC*. ITC*) For any predicates F and G, and any name a, i) Fa.Ga confirms (  xFx).(  x(Fx  Gx)); and ii) Fa.~Ga disconfirms (  xFx).(  x(Fx  Gx)). 10

A problem with this response Given (SPC), (ITC*) entails (PC). (SPC) If E confirms H1, and H1 entails H2, then E confirms H2. Argument: (1)By ITC*, ~Ba.~Ra confirms (  x~Bx).(  x(~Bx  ~Rx)) (2)(  x~Bx).(  x(~Bx  ~Rx)) entails  x(Rx  Bx) (3) By (1), (2) and (SPC), ~Ba.~Ra confirms  x(Rx  Bx) 11

Hempel’s response to the paradox (PC) is true. It seems false because a) We falsely think that ‘  x(Rx  Bx)’ is only about ravens, when in fact it is about all objects, as it is reformulation as ‘  x(~Rx v Bx)’ reveals. b) We falsely think that PC is false, because we fail to distinguish it from the false PC*. (PC*) That a is non-black and non-raven confirms that all ravens are black, given the background information that a is a non-raven. 12

Two kinds of confirmation relation 3-place: E confirms H relative to background knowledge K 2-place: E confirms H absolutely Connection between them: E confirms H absolutely iff E confirms H relative to no information (or relative to a logical truth T) ITC is intended as a theory of absolute confirmation 13

The hypothetico-deductive theory of confirmation (HDT) i)E confirms H if E can be divided into two parts, E1 and E2, such that a) E1 does not entail E2, but b) the conjunction of H and E1 does entail E2. ii)E disconfirms H if E entails ~H iii)Otherwise E neither confirms or disconfirms H In favour of HDT: HDT fits many episodes in the history of science well. 14

Problems with HDT Problem 1: HDT entails PC, and hence faces the ravens paradox Problem 2: HDT cannot account of confirmation of statistical theories such as the hypothesis that anyone who smokes has a 25% chance of developing lung cancer. 15

Problem 3: Irrelevant conjunction Suppose evidence E, made up of E1 and E2, is such that i) E1 does not entail E2, but ii) H.E1 does entail E2. Then H.S.E1 entails E2, where S is any hypothesis at all. Hence, according to HDT, E confirms H.S. Moreover, by SPC, E confirms S. But S can be anything at all! 16

The probability raising theory of confirmation PRT for absolute confirmation: i)E confirms H iff P(H|E) > P(H) ii)E disconfirms H iff P(H|E) < P(H) where ‘P(H)’ means ‘the probability of H’, and ‘P(H|E)’ means ‘the probability of H given E’. 17

The probability raising theory of confirmation (cont) PRT for relative confirmation: i)E confirms H relative to background knowledge K iff P(H|E.B) > P(H|K) ii)E disconfirms H relative to background knowledge K iff P(H|E.B) < P(H|K) 18

Quantitative probability raising theories of confirmation Def: c(H,E,K) = the degree to which E confirms H relative to background knowledge K A popular account of c among PRT theorists: D) c(H,E,K) = P(H|E.K) - P(H|K) 19

Good’s response to the raven paradox Good’s claim: Whether E=Ra.Ba confirms H=  x(Rx  Bx) relative to K depends on what the background knowledge K is. Example: E won’t confirm H if K is the knowledge that either i)There are 100 black ravens, no non-black ravens and 1 million other birds ii)There are 1000 black ravens, 1 white raven, and 1 million other birds 20

Good’s example E won’t confirm H if K is the knowledge that either i)There are 100 black ravens, no non-black ravens and 1 million other birds ii)There are 1000 black ravens, 1 white raven, and 1 million other birds In this case P(E|H.K) < P(E|~H.K), from which it can be proved that P(H|E.K) < P(H|K). 21

Good on absolute confirmation Good also claimed that it might be that Ra.Ba fails to confirm  x(Rx  Bx) absolutely. Discuss unicorn case. 22

The standard Bayesian strategy to solve the ravens paradox Show that given plausible assumptions about our background knowledge, Ra.Ba confirms  x(Rx  Bx) relative to K more than ~Ra.~Ba. The result if established can then be used to explain why (PC) seems false. 23

Hawthorne and Fitelson’s attempt Given the assumptions about K given by (K-ass), H+F show that the following theorem holds. K-ass: i) P(H|Ba.Ra.K), P(H|~Ba.~Ra.K), and P(~Ba.Ra|K) aren’t 0 or 1; and ii) P(~Ba|K) > P(Ra|K). Theorem: If P(H|Ra.K) ≥ P(H|~Ba.K), then P(H|Ba.Ra.K) > P(H|~Ba.~Ra.K). 24