Agile InsightsBriefing
Black and white photograph of a dark workbench: an upright hardback with a yellow ribbon, reading glasses, two pocket watches showing different times labelled Two accounts, and a map with a compass labelled No view from nowhere. Title: Rovelli's World.

Making You Think 7 · Primer and discussion · About 15 minutes

Rovelli's World, by Bas van Fraassen

There may be no view from nowhere, only accounts from somewhere.

#people

Part 1: Primer · About 6 minutes

Download the MP3 · Linda and Leonard, AI voices, read a script written from the sources below.

Part 2: Discussion · About 7 minutes

Download the MP3 · Linda and Leonard, AI voices, read a script written from the sources below.

Based on essays and our own research. The paper, free from the author.

In this episode

01Accounts from somewhereIn Carlo Rovelli's relational quantum mechanics, a thing has properties only relative to another system, and an observer is any physical system, not a person.
02Relative, not arbitraryLike the speed of a coffee on a train, a state relative to an observer is still governed by rules that are the same for everyone.
03No view from nowhere neededVan Fraassen's answer to the obvious objection: Rovelli describes the form every account must take, a transcendental point of view, not the world from outside.
04The gap, and the fixWhen observers compare notes, nothing guaranteed agreement. Van Fraassen proposed a rule that accounts can never flatly contradict, and Rovelli and Adlam's 2023 cross perspective links credit him.
05Decide what counts as a contradictionWithout a master account, the work moves to comparison: write down what would count as a contradiction between two records, and make sure someone checks.
The Additional Postulate
Van Fraassen's Additional Postulate Two observers each hold an account of the same system. A third observer watches both. The accounts may differ, but a yellow link marks the rule that they can never flatly contradict each other. Third observer watches both Observer A its account Observer B its account System CANNOT CONTRADICT The accounts may differ. They must not flatly contradict.

Our diagram of van Fraassen's Additional Postulate: two observers' accounts of the same system may differ, but when a third observer brings them together, they must not flatly contradict.

Your turn

Three questions to ask yourself

Think of an account you hold at work that you are sure is right. From where, exactly, is it right?
When two honest accounts of the same event differ, which one do you call the truth, and why that one?
What would count as a real contradiction between two of your records, rather than a difference?

Two things to try, alone, with a pen

10 minutesThree accountsWrite one event from three positions and sort differences from conflicts.
5 minutesThe comparison ruleDecide what would count as a contradiction between two records, and who would notice.
Open the activity sheets (print or save as PDF)

About Bas C. van Fraassen

Born in Goes, the Netherlands, in 1941, he moved to Canada as a teenager and taught at Yale, Toronto, Southern California and Princeton, where he is emeritus, then San Francisco State. He is known for constructive empiricism and received the Rolf Schock Prize in 2026.

Other writing (5)
The Scientific Image (1980) · Oxford University Press
The case for constructive empiricism.
Laws and Symmetry (1989) · Oxford University Press
Against laws of nature as the heart of science.
Quantum Mechanics: An Empiricist View (1991) · Oxford University Press
His reading of quantum theory, including the modal interpretation.
The Empirical Stance (2002) · Yale University Press
Empiricism as a stance rather than a doctrine.
Scientific Representation (2008) · Oxford University Press
Paradoxes of perspective in how science represents the world.
Sources and links (10)
Based on essays and our own research · The paper, free from the author
Where this episode comes from.
Rovelli's World (2010) · The paper (PDF)
Van Fraassen's paper in Foundations of Physics, free from his Princeton page.
Relational Quantum Mechanics (1996) · arXiv
Rovelli's original paper.
Information is physical: cross-perspective links (2023) · arXiv
Adlam and Rovelli's added rule, which credits van Fraassen.
Space is blue and birds fly through it (2018) · arXiv
Rovelli on the objections and van Fraassen's answers.
A Quintet of Quandaries (2021) · arXiv
Pienaar's five no-go theorems for relational quantum mechanics.
Qubits are not observers (2021) · arXiv
Brukner's objection.
Assessing relational quantum mechanics (2022) · arXiv
Muciño, Okon and Sudarsky's critical assessment.
Relational Quantum Mechanics · SEP
The Stanford Encyclopedia of Philosophy entry.
On the Equality of All Things · Amazon Australia
Rovelli's 2026 book, which takes the idea further.

Based on essays and our own research, checked on 21 October 2026. The author's argument is paraphrased, not reproduced. Check the original before quoting.

Our fact check · 14 claims: 11 confirmed, 2 unverified

Before we record, every claim the hosts make is checked against the sources (this table was last updated 21 October 2026). Confirmed means a primary source or two independent reports; Reported means one party says so; Partly right, Disputed, Unverified and Not yet known mean what they say, and the hosts say so on air. This is the table as we left it.

  1. ConfirmedPaper: Found Phys 40:390–417; online 11 Jul 2009; DOI 10.1007/s10701-009-9326-5Author's PDF; DOI; Dedicated to Jeffrey Bub; written at SFSU
  2. ConfirmedRovelli's "main observation" wordingRovelli 1996, arXiv quant-ph/9609002; paper p. 396; Quote under 15 words
  3. Confirmed"Observer" means any system, not a personPaper §1.1
  4. ConfirmedRovelli describes the form of any description, not the worldPaper pp. 397–398; "transcendental" is van Fraassen's word
  5. ConfirmedAdditional Postulate (§5.4), conceded not to follow from RovelliPaper pp. 412–414
  6. ConfirmedEPR result: ROV knows "opposite" but not whichPaper §6; No empirical difference
  7. ConfirmedThe O/P example is a Wigner's friend set-upRovelli 1996 §II; Muciño et al. 2022; In substance; Van Fraassen does not use the name for it
  8. ConfirmedAdlam & Rovelli 2023 add cross-perspective links and credit van FraassenArXiv 2203.13342, §4; They cite the Discusiones Filosóficas 2010 version of his paper
  9. ConfirmedRovelli (2018) says the objections were answered, "in particular" by van FraassenArXiv 1712.02894; Calls his reading "marked empiricist"
  10. ConfirmedPienaar, Brukner, Muciño et al. critiques as summarisedArXiv 2107.00670; 2107.03513; 2105.13338; Muciño's venue (Synthese) Reported
  11. Section title alludes to NagelPaper; Nagel 1986; Allusion; Nagel is not in the references
  12. ConfirmedBio dates, posts, Lakatos Award 1986, Schock Prize 2026SFSU CV; LSE; KVA; Hempel Award 2012 Reported
  13. Unverifiedvan Fraassen discussed in Helgoland—; Don't say it
  14. UnverifiedMaudlin as an RQM critic—; He attended the Princeton seminar on the first draft; nothing more

Want to know more?

Sign up for our podcast at agile-insights.com.au.

Read the transcript · Part 1: Primer

Linda Here is the idea for today. There may be no view from nowhere, only accounts from somewhere. So the question that matters is not which account of an event is the true one. It is whether the accounts agree when you bring them together. That idea comes from a short, sharp paper called Rovelli's World, by the philosopher Bas van Fraassen, published in 2010. By the end of this part you will know what it says about physics, why it is not anything goes, and the one gap it found. This is Making You Think, episode seven.

Leonard And this one is a little different, because it is a paper, not a book, and it came from a footnote. The physicist Carlo Rovelli has a new book out, On the Equality of All Things. In the preface he thanks van Fraassen as an early champion of his ideas. So who is he? He was born in the Netherlands in 1941, moved to Canada as a teenager, and taught at Yale, Toronto, Southern California and Princeton, where he is now emeritus. He is one of the leading philosophers of science, best known for The Scientific Image, from 1980. And this year he was awarded the Rolf Schock Prize, one of the major honours in philosophy.

Linda To follow the paper, you need one idea from Rovelli first. It is called relational quantum mechanics, and he proposed it in 1996. Quantum mechanics is our best theory of very small things. It has always had a puzzle at its heart about what happens when something is measured. Rovelli's answer is that a thing has no properties on its own. It only has properties relative to something else it has interacted with. His starting observation, in his own words, is that "different observers may give different accounts of the same sequence of events". And one point to hold onto: an observer here means any physical system at all. Not a person, not a mind. A particle can be an observer.

Leonard That sounds strange, but is it so different from ordinary life? If I am on a train holding a coffee, the coffee is perfectly still to me and doing a hundred kilometres an hour to someone on the platform.

Linda Same coffee, two answers. That is exactly the comparison Rovelli draws, and van Fraassen starts there too. Speed has always been relative to a frame of reference, and nobody thinks that makes speed meaningless. Rovelli says the same is true of a thing's state, which in physics is the summary of what you can predict about it. Here is the example the paper works through. One observer measures a particle and gets a definite result. A second observer, outside, who has not interacted with either of them yet, describes the first observer and the particle together as still undecided. Physicists will recognise a version of the puzzle called Wigner's friend. Rovelli's answer is that both descriptions are correct, each relative to its observer.

Leonard But then the obvious objection. If every description is relative to somebody, what is Rovelli doing when he describes the whole situation, both observers at once? That sounds like a view from nowhere.

Linda That is van Fraassen's first big move, and it is the heart of the paper. His section is even called Is There a View from Nowhere? His answer is no, and Rovelli does not need one. Rovelli is not describing the world from outside. He is describing the form that any observer's description has to take. Think of the difference between a map and the rules for making maps. Every map is drawn from somewhere, at some scale, for some purpose. But you can still say true things about what every map must do. Van Fraassen calls this "a transcendental point of view", meaning it is about the conditions any knowledge has to meet, not about the world in itself.

Leonard So relative does not mean arbitrary. There are rules everyone's account has to follow, even though there is no master account...

Linda Yes, and that is the line between this and anything goes. Then van Fraassen finds a real gap. Suppose two observers each have their own account. What happens when they compare notes? Nothing in Rovelli's original version quite guarantees that their accounts will fit together. Worse, what one observer has recorded might not show up in anyone else's account at all. You cannot fix it by imagining an all seeing observer who knows everything, because that would bring back the view from nowhere. So he proposes an extra rule, which he calls an Additional Postulate. The accounts held by different observers, and by a third observer watching them both, can never flatly contradict each other.

Leonard And did anyone take him up on it?

Linda That is the nice ending. In 2023, Rovelli himself, with the philosopher Emily Adlam, added a new rule to his theory, which they call cross perspective links. If one observer later checks another's record, and the record is still intact, the results will match. They write that it "has something in common with an earlier proposal by Van Fraassen", and then they differ on one point. He made agreement depend on a third observer watching. They make it hold between the two observers directly. So the gap he found in 2009 was real, and the fix he suggested pointed the right way.

Leonard Before Part 2, a question for you. Think of an account you hold of something at work that you are sure is right. Now ask yourself: from where, exactly, is it right?

Linda In Part 2, the critics, the strongest objection to Rovelli's whole approach, and what an idea from physics might have to teach anyone who keeps records. See you there.

Read the transcript · Part 2: Discussion

Linda The idea again. There may be no view from nowhere, only accounts from somewhere, so what matters is whether the accounts agree when they are compared. A quick recap of the terms. Relational quantum mechanics is Carlo Rovelli's view that a thing only has properties relative to another system. An observer is any physical system, not a person. Bas van Fraassen's transcendental reading says Rovelli describes the form every account must take, not the world from outside. And the Additional Postulate is van Fraassen's rule that different observers' accounts can never flatly contradict each other. Today, the strengths, the critics, and what any of this might mean off the physics bench.

Leonard The strength first, and I think it is a quiet one. A lot of writing about quantum theory either mystifies it or waves the puzzles away. Van Fraassen does neither. He states the objections as sharply as he can, such as the claim that two observers could catch each other in a contradiction. Then he shows, step by step, why they dissolve. Rovelli clearly valued that. In a 2018 essay he says the apparent inconsistencies in his view "have all been convincingly answered", and he points in particular to van Fraassen's arguments. When the author of a theory sends you to a critic's paper for the defence, the critic has done something right.

Linda And the other strength is honesty about the gap. Van Fraassen says plainly that his extra rule does not follow from Rovelli's own principles. He also admits it sits uneasily with something Rovelli had said: that observers can only compare information by physically exchanging it. He offers the rule anyway, as the price of keeping the accounts consistent. Then, fourteen years later, Rovelli and Emily Adlam add a rule that does a very similar job and credit him for the earlier version. That is how ideas are meant to move between philosophy and physics, and it does not happen often.

Leonard Now the critics, and they are serious. The physicist Jacques Pienaar published what he called a quintet of quandaries in 2021, five theorems arguing that relational quantum mechanics cannot keep everything it wants at once. Roughly, it cannot treat every system as an equal observer and still have the facts relative to each observer behave the way the theory needs. Rovelli and a colleague replied the following year. The physicist Caslav Brukner wrote a paper titled Qubits are not observers. His argument is that a simple quantum system cannot really hold knowledge the way an observer does. And a group of philosophers, Muciño, Okon and Sudarsky, concluded that the view does not solve quantum theory's puzzles and creates some new ones.

Linda They are serious. So it is worth being clear about the status of all this. Relational quantum mechanics is one interpretation of quantum theory among several. No experiment yet tells them apart. When we say Rovelli's world, we mean the world as it would be if his interpretation is right, which is exactly how van Fraassen frames his own paper. He is not saying physics has proved this. He is asking what kind of world it would be, and whether it hangs together. And his answer is that it mostly does, with one rule added.

Leonard Here is where I want to push a little. When I hear no view from nowhere, part of me hears relativism. Everyone has their own truth, so nobody can be wrong. And I do not think that is a comfortable place for a listener who has to make decisions.

Linda I understand the worry, but I think the paper is the opposite of that. Relativism says every account is as good as any other. Van Fraassen's version says every account is from somewhere, and all of them must obey the same rules, and when they are compared they must not contradict each other. That is a stricter standard, not a looser one. You lose the master account. You keep the discipline. In fact you need more of it, because there is no referee at the end to sort it out for you...

Leonard That I can live with. And it is where I think the idea travels, carefully. Van Fraassen wrote about physics, in 2009, and he says nothing about organisations or computers, so this next part is ours. Think of any organisation that runs on records kept by different systems: finance, operations, a set of A I agents each logging what it did. Each record is an account from somewhere. There is often no master record at all. And the control that actually matters is reconciliation. When two records are brought together, do they agree? And has anyone written down what would count as a contradiction?

Linda There is a link to an earlier episode, too. In Making You Think five, Barry Oshry argued that where you stand in an organisation shapes what you see. Van Fraassen's reading of Rovelli is the same lesson at the scale of physics. So a question for you. When two honest accounts of the same event at work differ, which one do you call the truth, and why that one?

Leonard Two activities, five to ten minutes each, with a pen and paper. The first is called Three accounts. Pick one recent event at work. Write it down three times. Once from your position. Once from the position of another person who was involved. And once from someone who only ever saw the paperwork. Then mark the places where the accounts merely differ, because they stood in different places. And mark the places where they truly conflict, where both cannot be right. Most people find far more of the first than the second.

Linda The second is called The comparison rule. Pick two records at work that ought to agree. A calendar and a timesheet. A support ticket and the incident report. A system's log and what the customer was told. Write down what, exactly, would count as a contradiction between them. Then write down who would ever notice if one happened. If the answer to the second question is nobody, you have found the place where van Fraassen's rule is not being checked.

Leonard So what to carry away. First, an account being from somewhere does not make it wrong, and it does not make it a matter of taste. Second, without a view from nowhere, the work moves to comparison: the accounts must not contradict each other, and someone has to check. Third, a good critic can end up shaping the theory they criticise, as van Fraassen did. The paper is free on his Princeton page, and Rovelli's new book takes the same idea a great deal further.

Linda If you'd like to know more, sign up for our podcast at agile-insights.com.au.

Leonard Thanks for listening.

More on #people

Episode 34 · Primer and discussionWhen the test stops telling you anything: hunger beats tenureEpisode 32 · Primer and discussionAgentic video is real: what to try, and what to checkEpisode 31 · Primer and discussionSpeeding up planning decisions with AI: who checks the checker?Making You Think 12 · Primer and discussionThe Game Changer, by Jason FoxEpisode 13 · Primer and discussionFairness under pressure: a fair AI decision has to hold when someone leans on itMaking You Think 9 · Primer and discussionArtificial Intimacy, by Sherry Turkle

Every episode tagged #people

All of Making You Think (13) · Follow Making You Think in Apple Podcasts or any podcast app

Everything from Agile Insights: briefings, Making You Think and the reference series