The Beginning of Infinity

Deutsch, a physicist and a founder of quantum computation, argues that scientific progress comes from seeking good explanations — those that are hard to vary while still accounting for what they explain — rather than from accumulating observations.

Hard-to-vary is his criterion and it does real work. A bad explanation can be adjusted to fit any outcome; a good one breaks if you change it, which is why it is testable.

A good explanation is one you cannot easily adjust. That is the whole criterion.

Devmark’s reading — not a quotation from the book

From this he derives an optimism that is not temperamental but structural: problems are soluble, all evils are due to insufficient knowledge, and there is no fundamental limit to what people can come to understand.

He applies the framework to the multiverse, to computation, to the evolution of culture through memes, to aesthetics and to political institutions — arguing that good political systems are those that make error correction easy.

This is an ambitious book by a physicist writing well outside physics, and the chapters on memetics, aesthetics and moral philosophy are considerably more contested than the epistemology.

His Popperian position — that induction plays essentially no role in science — is a minority view among philosophers of science, and readers should know it is a position rather than a consensus.

The epistemology is a defended minority position, not the settled view it sometimes reads as.

Devmark’s reading — not a quotation from the book

Genuinely difficult and genuinely rewarding, and long. Deutsch writes clearly but the arguments are dense and build across chapters.

It is among the most intellectually ambitious popular books of the century, and its confidence is either the main attraction or the main obstacle.

It is also a sequel of sorts to his earlier The Fabric of Reality, though it stands alone; readers who find the first three chapters hard going generally find the book opens up considerably after them.

Deutsch's criterion is that a good explanation is hard to vary while still accounting for what it is meant to account for. A myth can be adjusted freely and therefore explains nothing; a scientific theory has components that cannot be altered without breaking it. This single test does most of the work in the book.

A bad explanation can be adjusted to fit anything. That is exactly what makes it useless.

Devmark’s reading — not a quotation from the book

He argues that knowledge does not in fact come from observation, because observation is always theory-laden and because no quantity of data ever selects a single theory on its own. Knowledge grows instead by conjecture followed by criticism, with experiment functioning as a way of killing bad guesses rather than generating good ones. The position is inherited from Popper and pushed considerably further than Popper took it.

A recurring theme is that good explanations turn out to apply enormously far beyond the narrow problems that produced them: arithmetic devised for counting livestock ends up describing particle physics, and rules discovered in one laboratory on one planet hold across the observable universe. Deutsch treats this unreasonable and unearned reach as the central and most astonishing fact about human knowledge.

He argues directly and at length against the fashionable view that humans and the Earth are cosmically insignificant specks, claiming instead that people are universal explainers and therefore occupy a genuinely special position in the universe: capable, in principle and given enough knowledge, of understanding and physically transforming anything whatsoever that is not forbidden outright by the laws of physics.

The book's optimism is a stated principle rather than a temperament: all evils are caused by insufficient knowledge, progress is unbounded, and any problem not forbidden by the laws of physics can be solved given the right knowledge. It is a bold claim and he defends it at length.

The chapters on the many-worlds interpretation of quantum mechanics, on sustainability as a fundamentally mistaken goal, and on voting systems and political philosophy are all considerably more contested than the epistemology that precedes them, and Deutsch presents every one of them with exactly the same confidence. Readers should stay alert to the point at which he moves from argued ground onto advocacy.