Eleminating chance through small probabilities
2nd edition, revised and expanded
by William A. Demski and Winston Evert
A landmark of the intelligent design movement, The Design Inference revolutionized our understanding of how we detect intelligent causation. Originally published twenty-five years ago, it has now been revised and expanded into a second edition that greatly sharpens its exploration of design inferences. This new edition tackles questions about design left unanswered by David Hume and Charles Darwin, navigating the intricate nexus of chance, probability, and design, and thereby offering a novel lens for understanding the world. Using modern concepts of probability and information, it exposes the inadequacy of undirected causes in scientific inquiry. It lays out how we infer design via events that are both improbable and specified. Amid controversial applications to biology, it makes a compelling case for intelligent design, challenging the prevalent neo-Darwinian evolutionary narrative. Dembski and Ewert have written a groundbreaking work that doesn’t merely comment on contemporary scientific discourse but fundamentally transforms it.
Table of Contents
Foreword by Michael Egnor Introduction to the Second Edition
Chapter 1: The Challenge of Small Probabilities
1.1 Historical Backdrop
1.2 The Reach of Chance
1.3 Life in the Short Run
1.4 Chance as a Side Effect of Intelligence
1.5 From Chance Elimination to Design
Chapter 2: A Sampler of Design Inferences
2.1 Intellectual Property Protection
2.2 Forensic Science
2.3 Data Falsification in Science
2.4 Financial Fraud—The Madoff Scandal
2.5 Randomness 2.6 Cryptography
2.7 SETI—The Search for Extraterrestrial Intelligence
2.8 Directed Panspermia
Chapter 3: Specification
3.1 Patterns That Eliminate Chance
3.2 Minimum Description Length
3.3 Events and Their Corresponding Patterns
3.4 Recognizing Patterns That Signal Design
3.5 Prespecifications
3.6 Specification-Induced Rejection Regions—The Idea
3.7 Specification-Induced Rejection Regions—The Math
3.8 Modes and Tails of Probability Distributions
Chapter 4: Probabilistic Resources
4.1 Calculating Probabilistic Resources
4.2 Relative Versus Absolute Probabilistic Resources
4.3 Variations in Absolute Probabilistic Resources
4.4 Avoiding Superexponentiality by Not Miscounting
4.5 Minimizing Absolute Probabilistic Resources
4.6 Universal and Local Probability Bounds Chapter
5: The Logic of the Design Inference
5.1 The Man with the Golden Arm
5.2 The Generic Chance Elimination Argument (GCEA)
5.3 Key Concepts and Predicates
5.4 The Design Inference as a Deductive Argument
5.5 From Design to Agency
5.6 The Explanatory Filter
5.7 Probabilistic Modus Tollens
Chapter 6: Specified Complexity
6.1 A Brief History of the Term and Idea
6.2 Description Length 6.3 Practical Approximation of Description Length
6.4 Specification and Complexity 6.5 Frequentist Interpretation
6.6 Bayesian Interpretation
6.7 Contextual Factors
6.8 Examples Chapter
7: Evolutionary Biology
7.1 Insulating Evolution against Small Probabilities
7.2 Resetting Darwinian Evolution’s Bayesian Prior
7.3 John Stuart Mill’s Method of Difference
7.4 The Challenge of Multiple Simultaneous Changes
7.5 What to Make of Bad Design? 7.6 Doing the Calculation
7.7 Where to Look for Small Probabilities in Biology?
Epilogue: Beyond the Design Inference—Conservation of Information The Basic Idea of Conservation of Information The Challenge of Evolutionary Computing The Mathematics of Conservation of Information
Appendix A: A Primer on Probability and Information
A.1 The Bare Basics of Probability
A.2 Conditional Probability and Independence
A.3 Bayes’ Theorem
A.4 The Bayesian Approach to Statistical Inferences
A.5 The Fisherian Approach to Statistical Inferences
A.6 Random Variables and Search
A.7 The Privileged Place of Uniform (or Equi-) Probability
A.8 The Universality of Coin Tossing for Probability
A.9 Information as Constraint on Contingency
A.10 Shannon Information
A.11 Connecting Information to Probability and Computation
Appendix B: Select Related Topics
B.1 New Specifications from Old
B.2 Transformations of Specifications
B.3 Perturbation Neighborhoods
B.4 The Uniformizability (or Normalizability) Problem
B.5 Stopping and Waiting Times
B.6 Probabilistic Hurdles to Irreducible Complexity in Biology
B.7 The Origination Inequality
B.8 The Drake Equation
Appendix C: The First Edition of The Design Inference
C.1 Endorsements Appearing in the Hardback First Version
C.2 Subsequent Endorsements for the First Edition
C.3 Preface to the First Edition
C.4 Acknowledgments for the First Edition
C.5 Transitioning from the First to the Second Edition Endnotes Bibliography Index
Advance Praise
Ever since Darwin, most scientists have adopted a principled view by which they reject out of hand any non-naturalistic explanations. This works perfectly well in the physical sciences, but less so in biology where, due to the incredible complexity of biological systems, appearance of design is overwhelming. Yet, by appealing to this metaphysical principle, intelligent design (ID) ideas are automatically rejected; natural selection and random mutations are viewed as the only acceptable explanation of the mechanism by which biological evolution takes place. Though I take an agnostic position on ID, I have no doubt that its main proponents, Behe and Dembski, have brought to light important challenges to the reigning neo-Darwinian version of evolutionary biology. This second edition of Dembski’s classic The Design Inference is well argued and eminently readable. The appendix provides the reader with a short, effective, introduction to the probabilistic and statistical methods used throughout the book. The authors give plenty of well-motivated, non-biological examples on how specified events of small probability lead to a convincing inference of intelligent design. The same arguments become controversial only when applied to biology! I don’t see how any open-minded scientist can ignore this important book. —
Sergiu Klainerman, Higgins Professor of Mathematics, Princeton University, member of the National Academy of Sciences
Darwinists have long asserted any appearance of design in life is the result of natural unintelligent processes that had no end in mind. Any suggestion that there is a designer is merely a primitive “god of the gaps” argument from ignorance. But Drs. Dembski and Ewert show in an accessible and testable way that intelligent design is not an argument from ignorance—life itself contains empirically, verifiable evidence for design. And the evidence is mathematically overwhelming. Some will use their designed minds to continue to resist the conclusions of this brilliant tour de force, but given the evidence I suspect any resisters will be either stubborn ideologues or really bad at math.
— Frank Turek, President of CrossExamined.org, author and speaker
This new and expanded edition of The Design Inference follows in a long line of books over the last quarter century by distinguished mathematician William Dembski, renowned as the leading intelligent design (ID) specialist in the world. This book is another important step along the way to validating intelligent design as a mainstream and scientifically robust alternative to Charles Darwin’s nineteenth-century philosophy of natural selection. A key question this book addresses is: Does natural selection have sufficient creative power to account for the immense complexity and information-richness of life? To date, mainstream science has failed to answer this question. An appeal to faith in natural selection’s information-creating powers—in the continued absence of clear confirming evidence of such—remains the current leading answer for information creation. But in some countries, such as Brazil, ID is making dramatic inroads as a sub-discipline within biology. In other countries where a more traditional ruling scientific orthodoxy holds sway, ID is a target of scientific censorship. It is a well-known adage, “Where there’s smoke, there’s fire.” There is certainly a lot of smoke surrounding this ID-versus-natural-selection controversy. This book gives us the clearest picture yet of the fire, and thereby takes the science of ID one step closer to validation as a rigorous and scientifically robust explanation for the immense information-richness of life.
— Andrew Ruys, Professor of Biomedical Engineering (Retired), University of Sydney
In his book, Six Great Ideas, the philosopher Mortimer Adler stated: “There would be stars and atoms in the physical cosmos with no human beings or other living organisms to perceive them. But there would be no ideas as objects of thought without minds to think about them.” Stars and atoms are the venue of physics. During the last century, the vast majority of physicists have made peace with the notion that the universe had a beginning, and that any attempt to assign a cause or mechanism or prior state to that beginning lies beyond the reach of the natural sciences. Darwinists, however, persist in the hubris of believing that they have fully resolved how the chance assemblage of an exponentially lengthy sequence of statistically impossible events could produce life in all its variations. Any objections to this conclusion are met with censorship, derision, and a profound obliviousness to the mathematical hurdles confronting the Darwinian view. We have sadly reached an anti-scientific point in many circles where even openly thinking about an alternative explanation is viewed as heresy. In this second edition of The Design Inference, Dembski and Ewert present a formidable probabilistic and information-theoretic method for determining whether design, rather than chance, was the cause of an observed event. They then apply this method to the intricate forms we find in biology. With devastating mathematical precision, the book demonstrates that any complex event having both a briefly described specification and a small probability of occurrence—that is, small in light of all available probabilistic resources—must logically be attributed to design rather than chance. This edition also incorporates further mathematical refinements, particularly in the account of specified complexity. It updates many of the references. And it convincingly refutes the various objections raised since the publication of the original version. It is remarkable that the question of design, ubiquitous in everyday experience, is met with such ferocious resistance when it comes to thinking about the origin of living organisms, which represent the ultimate in specified complexity. Minds open to the issues raised in this book will be able to fruitfully engage the debate over biological origins. In this greatly revised and expanded edition, opponents of design have a new and unenviable challenge to surmount.
— Terry Rickard, PhD, Engineering Physics, University of California, San Diego
