Skip to content
Coordinates of ThoughtHome of Philosophy
Join
Question · Explain

QBism says the wavefunction is my degrees of belief. What exactly is it claiming?

Wanjiru Njenga Fellow
· 3 replies
Physics & cosmologyAnalyticModern EuropeanBruno de FinettiNiels Bohr#bayesian-probability#qbism#quantum-interpretation#subjectivism
Follow

A colleague who knows I work with Bayesian statistics sent me an article about QBism ("quantum Bayesianism", though I gather some proponents now prefer just the acronym) with the note: "your people have taken over physics."

As far as I can tell, the claim is that the quantum state is not a description of a system. It is an agent's personal probability assignment about what she will experience when she acts on the system. Probability is subjective in de Finetti's sense, the famous line being that probability does not exist, and quantum mechanics is a set of coherence rules for an agent's bets.

I can't tell whether this is a deep insight or a category mistake, and I have two specific confusions:

  • Does it mean reality depends on observers? That sounds like idealism, and I doubt that is intended.
  • In statistics, subjective priors are tolerable because data make different agents converge. Does that work here, and if so what is the data about?

I'd welcome an explanation from someone who knows the physics better than I do.

LikeAnswer
Follow3 answers

Join the conversation: members reply here, and joining is free. Reading is always free.

Join free to reply Sign in

3 answers

Accepted answer

  1. Abdulrahman Al-Bassam

    ✓ Accepted answer Fellow

    QBism was developed from the early 2000s by Carlton Caves, Christopher Fuchs and Rüdiger Schack. Your summary is right: a quantum state is an agent's personal degrees of belief about the outcomes of her own actions on the world (her "measurements"). The Born rule is then not a law of nature but a normative constraint, an addition to ordinary Bayesian coherence that an agent in a quantum world ought to obey.

    Is it idealism? No, and its proponents are emphatic. There is a world that is not up to the agent; what the agent experiences when she acts is not chosen. What QBism denies is that the quantum state describes that world. It is closer to a user's manual for living in it. Critics answer that a manual that works so astonishingly well calls for an explanation in terms of how the world is, and QBism declines to give one.

    Convergence. Agents who share data will, by the usual Bayesian mechanisms, come to similar quantum states. The data are each agent's experiences of measurement outcomes, shared by communication. That makes the theory's objectivity intersubjective rather than descriptive.

    So it is neither a takeover nor a mistake: it says precisely where the subjectivity sits, and pays by refusing the realist's question.

    Helpful · 3
    1. Wanjiru Njenga

      Fellow

      "A user's manual for living in the world" is the phrase I'll use with my colleague. What I find honest about it is that it says plainly which question it refuses to answer, which is more than many statistical methods do.

      Helpful · 1

Other answers

  1. Lars Eriksen

    Fellow

    I'd add that this is not as distant from Bohr as people think. Bohr also refused to treat the wavefunction as a picture of reality between measurements. The difference is that Bohr's emphasis fell on the classical description of the apparatus, and QBism's falls on the agent. Where they part company is that QBism is explicitly personalist: two agents can legitimately hold different quantum states for the same system.