
Why progress can keep beginning
In The Beginning of Infinity: Explanations That Transform the World, physicist David Deutsch asks why people can understand a universe far larger than their own experience. His answer connects scientific discovery with technology, culture, politics, and moral learning. People create explanations. By finding and correcting errors in those explanations, they can change what they understand and what they can do.
The book’s central claim is that this process has no necessary final stage. There need not be a last discovery or a perfect society with nothing left to improve. Yet progress can stop when people destroy the conditions that support it. The infinity in the title means an open-ended capacity for improvement, rather than a prediction that history will automatically get better. The publisher’s description places this argument across science, civilization, art, and political institutions.
What makes an explanation good?
Deutsch’s key test is whether an explanation is difficult to change while still explaining the same facts. In a strong explanation, the details do necessary work. A weak explanation permits endless changes because its details have little connection to the result.
He illustrates the difference with the seasons. A story about a goddess’s grief can connect winter with sadness. But a storyteller can change the gods, their motives, or their actions without losing the seasonal pattern. When awkward evidence appears, another invented detail can rescue the story.
Earth’s tilted axis gives a different kind of account. As Earth travels around the Sun, each hemisphere receives changing angles and durations of sunlight. The explanation connects geometry, sunlight, and the opposite seasons in the two hemispheres. Changing these parts changes what the theory predicts.
The point is broader than testability. A prediction can be testable and still rest on an arbitrary story. Deutsch wants explanations whose parts constrain each other. His TED talk on explanation introduces this argument.
Knowledge grows through guesses and criticism
Following philosopher Karl Popper, Deutsch rejects the idea that observations simply pour knowledge into a passive mind. People must invent possible accounts of what they observe. They then criticize those accounts and test them where experiments are possible.
This is fallibilism: any of our ideas may contain errors. It does not mean that every idea is equally good. We can prefer an explanation because it solves problems that its rivals cannot solve. We can also discover faults without first possessing a final, certain account of reality.
Deutsch also rejects reducing science to prediction. He argues that explanations can describe a real world beyond immediate experience. His official glossary makes these distinctions explicit.
Consider an everyday example, offered here as an application. A machine fails whenever a room becomes hot. Recording that pattern helps predict a failure. Explaining that a component expands and breaks an electrical contact suggests where to inspect it and how to redesign it. Criticism can target the proposed mechanism. Does cooling that component prevent the failure? Does a different contact solve it? The explanation creates opportunities that the pattern alone does not provide.
Explanations reach beyond their starting point
An explanation can solve more than the problem that first prompted it. Deutsch calls this its reach. A theory developed in one setting may apply elsewhere because it describes something general about reality.
The related idea of a jump to universality concerns a change in capability. A system can move from handling a restricted collection of tasks to handling every task within a defined class. Positional number notation gives an accessible example. A finite set of digits, combined with rules about position, can represent arbitrarily large whole numbers. Writers need more space for longer numbers, but they do not need a new basic symbol for each larger quantity.
Deutsch connects this idea with computers and human understanding. A universal computer has a general repertoire, even though any actual machine has limited memory and speed. He argues that people are universal explainers, with no fixed subject boundary beyond which understanding becomes impossible in principle. He develops these connections in his interview with Micah Redding, collected in Bold Conjectures.
Universality does not supply answers on demand. Time, materials, education, and the difficulty of a problem still matter. Its significance is that a present failure need not reveal a permanent ceiling.
Creativity changes what counts as a resource
Deutsch applies the same reasoning to material limits. The usefulness of a substance depends on what people know how to do with it. A stock of material and a capacity to use it are different things.
His discussion of resources challenges forecasts that hold technology fixed while projecting needs far into the future. In his 2011 interview on KERA’s Think, he points to uranium ore: its significance changed when people learned about radioactivity. He also distinguishes cultural ideas that spread through criticism from ideas that prevent their own criticism.
An application is to ask what makes a shortage difficult. Is the obstacle a physical law, a missing technique, a cost, or a rule that blocks experimentation? Each answer leads to different work. Calling all four an absolute limit hides useful distinctions.
This question also needs a deadline. A possible future technique cannot repair a water system today. The practical value of Deutsch’s argument is to encourage investigation into alternatives. It does not establish that an alternative will arrive before damage occurs.
A society needs ways to correct itself
The book extends error correction into public life. A culture can preserve ideas by allowing people to question them. It can also preserve ideas by making questions shameful or dangerous. Deutsch calls these contrasting patterns rational and anti-rational memes. Here, a meme is an idea that helps cause its own transmission, rather than an internet joke.
For Deutsch, the political question is how people can remove bad rulers and bad policies peacefully. Following Popper, he values institutions that permit correction over hopes of identifying permanently wise leaders. In his interview about the book, he links freedom of dissent with the ability to test proposals and change course.
This makes institutional design part of knowledge creation. A policy can fail even when its supporters are sincere. A useful institution lets people expose that failure without first defeating a claim of infallibility.
The same test can guide a workplace. Can someone report a defect without punishment? Can a failed project lose support? Can a team explain why it changed its mind? These are applications of the argument, rather than a management program prescribed by the book.
Optimism demands work
Deutsch’s optimism treats problems as unavoidable and solutions as possible. Solving one problem creates new questions. Better knowledge can also reveal faults that were previously invisible. The absence of a final resting point is part of progress.
This is a demanding position. It gives people responsibility for creating solutions and protecting inquiry. The book’s introduction presents the search for good explanations as the common source of theoretical and practical progress.
A useful reading of this optimism is to replace resignation with a more precise question: what knowledge would make a difference? That question can turn an overwhelming difficulty into a field of investigation. It still leaves the hard work of finding, checking, and applying an answer.
Where the argument needs caution
The book combines arguments with different levels of support. Its account of scientific criticism does not, by itself, establish every wider claim about culture or the future.
Deutsch also defends the Everett, or many-worlds, account of quantum theory and argues for objective truths in aesthetics and morality. In the book interview, he identifies these as parts of his position. Readers should treat them as arguments to assess, rather than settled conclusions borrowed from the success of physics.
Several further limits follow from the argument itself:
- Possible in principle is different from available in time. A solution may require resources or discoveries that a particular community cannot obtain before a crisis.
- A good explanation does not choose every priority. People must still decide whose problems receive attention and how to handle costs imposed on others.
- Broad historical categories can conceal differences. The contrast between static and dynamic societies needs care when applied to particular institutions, places, and periods.
- Explaining a mechanism remains difficult. A confident story can sound tightly connected while overlooking another cause. Independent evidence and serious alternatives still matter.
These cautions preserve the book’s most useful challenge: examine the means by which an idea can be corrected. When a belief, institution, or project prevents criticism of itself, it weakens the process that could improve it. When it makes criticism productive, it keeps further discovery possible.
Sources
- Penguin Random House: book and author information.
- Official book site: introduction and glossary.
- David Deutsch: “A new way to explain explanation,” TEDGlobal 2009.
- Elliot Temple’s interview with Deutsch about the book.
- KERA Think interview, August 29, 2011: community-posted transcript. The opening post transcribes the interview; later discussion is separate.
- Bold Conjectures, Volume I: collected Deutsch interviews, edited by Logan Chipkin, especially “Micah Redding: Humanity’s Infinite Reach.” These later interviews clarify related ideas; they are not excerpts from the selected book.


