The Grand Design
Stephen Hawking and Leonard Mlodinow, 2010
Executive Summary
Twenty-two years after A Brief History of Time ended by asking what place there is for a creator, Hawking returned with a co-author and a verdict. The Grand Design is no longer than the first book and carries much less argument per page. It is also far more confident, and the confidence is the subject of this review. Its claim is that the universe explains itself: the laws do not need to be set, only to be possible, and once they are possible a universe of our kind will happen.
The engine of that claim is the best idea in the book. Hawking and Mlodinow call it model-dependent realism, and it says that a theory is a model plus rules connecting the model to observation, so that asking which of two successful models is “really” true is a mistake. That is a serious position, competently deployed against the goldfish and the flat-earther, and it is also philosophy of science of a kind the book’s first chapter has just announced dead.
The argument of this review is that the book converts why-questions into how-answers and then reports the why-questions as answered. It sets out three of them in chapter one, states four criteria for a good model in chapter two, and then never holds its own candidate theory against those criteria. Where the 1988 book taught a reader to feel the difference between a boundary and a place, this one closes the difference by assertion, and the famous sentence about gravity is the assertion.
The Architecture, and the Three Questions
Eight chapters, about 198 pages of text, released on 7 September 2010, straight to the top of the Publishers Weekly hardcover nonfiction list. The co-author is Leonard Mlodinow, a physicist and writer who had already worked with Hawking on A Briefer History of Time in 2005, so the publisher’s line that this was Hawking’s “first major work in nearly a decade” is right, while the common claim that it was his first co-authored book is not. Mlodinow has described how it began: after the earlier collaboration he suggested a book on Hawking’s recent thinking during one of Hawking’s annual visits to Caltech, and Hawking said yes, and then they argued about the outline “for the next year.”
| Ch. | Title | What it does |
|---|---|---|
| 1 | The Mystery of Being | Poses the questions; declares philosophy dead |
| 2 | The Rule of Law | What a law of nature is; the four criteria for a good model |
| 3 | What Is Reality? | Model-dependent realism; the goldfish |
| 4 | Alternative Histories | Quantum mechanics; Feynman’s sum over histories |
| 5 | The Theory of Everything | M-theory, and what it is not |
| 6 | Choosing Our Universe | The early universe; the sum over histories applied to the whole |
| 7 | The Apparent Miracle | Fine-tuning; the anthropic principles; the multiverse |
| 8 | The Grand Design | Spontaneous creation; the conclusion |
The book opens by promising to answer three questions, set as three lines of display type at the end of the first chapter’s argument:
Why is there something rather than nothing?
Why do we exist?
Why this particular set of laws and not some other?
The first of these is Leibniz’s question, and the book does not attribute it, though it does credit it properly as the Ultimate Question of Life, the Universe, and Everything before promising something better than “42.” The stated purpose is direct: “The purpose of this book is to give the answers that are suggested by recent discoveries and theoretical advances.”
It is worth holding that promise next to the first page of chapter one, where the authors dispose of a discipline:
Traditionally these are questions for philosophy, but philosophy is dead. Philosophy has not kept up with modern developments in science, particularly physics.
Two sentences, no argument, and then eight chapters spent doing the thing they have just declared dead.
5 Core Arguments
Model-dependent realism is the book’s real contribution, and it is a philosophy. The definition arrives in chapter 3 and is precise: “There is no picture- or theory-independent concept of reality. Instead we will adopt a view that we will call model-dependent realism: the idea that a physical theory or world picture is a model (generally of a mathematical nature) and a set of rules that connect the elements of the model to observations.” The operational rule that follows is the honest one: “it is pointless to ask whether a model is real, only whether it agrees with observation.” Read as an account of what physicists actually do, this is defensible and clarifying. Read against the book’s own opening, it is an embarrassment, because the position has a name in the literature. Craig Callender, reviewing the book for New Scientist, put the problem exactly: “Having declared that ‘philosophy is dead’, the authors unwittingly develop a theory familiar to philosophers since the 1980s, namely ‘perspectivalism’.”
The book sets four criteria for a good model and then never applies them to its own candidate. Chapter 2 is the most useful chapter in the book, and its list is the reason: “A model is a good model if it: 1. Is elegant 2. Contains few arbitrary or adjustable elements 3. Agrees with and explains all existing observations 4. Makes detailed predictions about future observations that can disprove or falsify the model if they are not borne out.” The authors then run Aristotle against the list and find him wanting. They never run M-theory against it, although their own description of M-theory fails the fourth criterion by construction: “M-theory is not a theory in the usual sense. It is a whole family of different theories, each of which is a good description of observations only in some range of physical situations.” The governing image is a collection of overlapping flat maps, which is an image of incompleteness offered as a credential. Peter Woit, a Columbia mathematician who has spent a career on string theory’s problems, made the omission explicit: “The fact that ‘M-theory’ satisfies none of these criteria is not remarked upon.” A book that publishes its own test and then declines to take it has told you where its confidence comes from.
The creation argument answers a how while claiming a why. The sentence that travelled the world appears at the end of chapter 8: “Because there is a law like gravity, the universe can and will create itself from nothing in the manner described in Chapter 6. Spontaneous creation is the reason there is something rather than nothing, why the universe exists, why we exist.” Note first what the newspapers dropped: not “such as gravity” but “like gravity,” and the trailing clause that sends the reader to chapter 6. The clause matters because it locates the argument in quantum cosmology, where it belongs, rather than in metaphysics, where it does not belong and cannot be defended. Note second the form of the claim. A law of nature is a description of regularities, and a description is not a cause. Saying that gravity explains why there is something rather than nothing is like saying that grammar explains why there are novels. The book has a how-answer, an excellent one, and it has labelled it a why-answer.
The God argument is a regress argument, and it is the strongest thing in the book. After the gravity sentence comes the line the press quoted: “It is not necessary to invoke God to light the blue touch paper and set the universe going.” That is a claim about necessity, not existence, and in an interview the same week Hawking put the narrower version plainly: “God may exist, but science can explain the universe without the need for a creator.” The argument behind it is not physical at all but logical, and it appears earlier, in the chapter’s restatement of the three questions: “It is reasonable to ask who or what created the universe, but if the answer is God, then the question has merely been deflected to that of who created God.” That is a clean regress, it has been made before by better philosophers than either author, and it survives whatever happens to M-theory. It is also the one place where the book’s philosophy is doing real work while the book denies having any.
The multiverse is presented as a consequence rather than an escape, which is right, and the consequence rests on something unproved. The authors are careful here, and the care is worth crediting: “That multiverse idea is not a notion invented to account for the miracle of fine-tuning. It is a consequence of the no-boundary condition as well as many other theories of modern cosmology.” That defuses the standard charge of ad hocness. The difficulty is in the last chapter, where the same chain of reasoning is left visibly conditional: “If it is finite, and this has yet to be proved, it will be a model of a universe that creates itself.” A consequence of an unproved conjecture is not evidence. It is a promissory note, and the book is honest about that in the footnote clause and silent about it in the conclusion.
A Respectful Disagreement
The published objections divide into two kinds, and they are worth keeping separate because only one of them is fatal.
The fatal one is the missing test. Callender put the empirical half of it in a sentence: in the absence of theory, the appeal to a vast ensemble of universes “is nothing more than a hunch doomed, until we start watching universes come into being, to remain untested.” Hamish Johnston, writing on Physics World’s blog the week of publication, said the same about the book’s theology: “there is currently little experimental evidence to back up M-theory. In other words, a leading scientist is making a sweeping public statement on the existence of God based on his faith in an unsubstantiated theory.” Woit went further and compared the result to the thing Hawking spent a career opposing: “the kind of pseudo-science going on here and being promoted in this book isn’t obviously any better than the faith-based explanations of how the world works favored by conventional religions.” I would not go that far. M-theory is a serious research programme with serious mathematics behind it, and calling it pseudo-science is a rhetorical overreach that costs Woit the argument. But the underlying point, that the book’s chain of inference has two untested links and reports them as closed, is correct.
The other objection is the philosophical one, and it is aimed at the first page. Christopher Norris, in Philosophy Now, adapted Keynes: “those scientists who claim to have no use for philosophy are most likely in the grip of a bad old philosophy or an insufficiently thought-out new one that they don’t fully acknowledge.” Leemon McHenry made the case in a 2025 article with an economy worth borrowing: “Hawking himself, in venturing into the territory of philosophy by proposing his own theory about physical theories, is doing philosophy,” and “There is no escape from philosophy in the same way that there is no escape from some conceptual framework by which we interpret our experience.” Both are right, and both are saying something narrower than “philosophy is undead.” They are saying that a man who defines reality operationally and then argues about what exists has not escaped the field. He has taken a position inside it.
Here is my own disagreement, and it is with the book’s structure rather than its physics. The Grand Design is a machine for making the transition from how to why look like a deduction. Chapter 1 poses three questions that are not answerable by measurement. Chapter 2 supplies a criterion (agreement with observation) that only measurement can satisfy. Chapter 8 announces that the questions are answered. At no point does the book tell the reader which of the three questions its apparatus can actually handle. My view is that it handles one of them, the third, and handles it well: “why this particular set of laws” is a question about selection, and the anthropic principle, renamed by the authors with unusual candour as a “selection principle,” is a real answer to a selection question. The first question is not a selection question. It is a question about why there is anything at all, and spontaneous creation is a mechanism, and mechanisms presuppose something to operate on: a law, a vacuum, a quantum rule. The book knows this. It says so in the clause about chapter 6 and then writes the confident sentence anyway.
That is not a small rhetorical sin. It is the reason the book sold. The blurbs, the serialisation, and the front pages all carried the verdict and dropped the conditional, and the conditional is where the honesty lives.
Bottom Line
Read chapter 2 for the four criteria and chapter 3 for model-dependent realism. Those twenty-odd pages are the book’s contribution, and they are genuinely useful: the account of why a theory is a model rather than a picture is the clearest short statement of it in popular science, and the goldfish is a better teaching device than anything in A Brief History of Time. Read chapter 8 to see what happens when the apparatus is asked to carry an argument it was not built to carry.
Skim chapters 5, 6, and 7. The M-theory chapter is an apology for a theory the authors cannot define, and the middle chapters are the standard tour of inflation, fine-tuning, and the cosmic microwave background, competently done and available in better company.
The book is also shorter of substance than its predecessor, in a way that reviewers noticed at the time: Woit counted “about 100 pages of actual text, interspersed with lots of color graphics and cartoons,” and no references of any kind. For a book whose central move is to cite theoretical advances as authority, the absence of citations is not a formatting choice. It is how a claim stops being checkable.
If you want one Hawking book, read A Brief History of Time instead. It asks the question this one claims to have answered, and it is better for leaving it open. The Grand Design is what happens when the answer arrives before the argument, and the prose carries the strain: the earlier book’s spare and careful voice is replaced here by a confidence that has to keep announcing itself.
“If the theory is confirmed by observation, it will be the successful conclusion of a search going back more than 3,000 years. We will have found the grand design.” — Stephen Hawking and Leonard Mlodinow, The Grand Design (Bantam, 2010), ch. 8, full text
Sources
- Stephen Hawking and Leonard Mlodinow, The Grand Design (Bantam, 2010), full text
- Craig Callender, “Stephen Hawking says there’s no theory of everything,” New Scientist, September 8, 2010
- Peter Woit, “Hawking Gives Up,” Not Even Wrong, September 7, 2010
- Christopher Norris, “Hawking contra Philosophy,” Philosophy Now, issue 82 (2011)
- Leemon B. McHenry, “On the Prospects for a Unified Theory of Physics: A Belated Response to Stephen Hawking,” Science & Philosophy 13(2), 2025
- Hamish Johnston, “M-theory, religion and science funding on the BBC,” Physics World blog, September 8, 2010
- CNN, Larry King Live transcript, September 10, 2010
- BBC News, “Stephen Hawking: God did not create Universe,” September 2, 2010
- Publishers Weekly, hardcover nonfiction bestsellers, September 27, 2010
PRH | huffmanwrites.org | © Philip Huffman