I Qbism Vs. Copenhagen

I read a few articles about Qbism, but feel like I don't understand the difference from Copenhagen. One author claims the difference is that each observer has his own view of the world (and own probabilities in qbism). I thought that was also the case in Copenhagen. E.g. the old Wigner's friend setup when the friend makes an observation and updates his probabilities accordingly while Wigner, before observing the larger system, still has a superposition with less "precise" probabilities (until he measures and updates). Is qbism somehow different here? Is there another example that clearly makes the distinction?


Science Advisor
Insights Author
Gold Member
And it's also important to note that there are never contradictions between different observers concerning the probabilities for a given situation (state preparation).
Why QBism is not the Copenhagen interpretation

QBism maintains that my understanding of the world rests entirely on the experiences that the world has induced in me throughout the course of my life. Nothing beyond mypersonal experience underlies the picture that I have formed of my own external world. This is a statement of empiricism. But it is empiricism taken more seriously than most scientists are willing to do.
Lived experience is where we start from and where all must link back to, like a guiding thread



Science Advisor
Insights Author
Gold Member
In “’B’ is for Bohr” (https://arxiv.org/abs/1905.07118), Ulrich Mohrhoff suggests that the "B" in QBism should stand for Bohr.
I think to write a paper like this is an art of its own. The author managed to be even more obscure than good old Bohr or even Heisenberg himself. To associate this gibberish with Kant or even Pauli is an insult against these people... SCNR.


Science Advisor
Insights Author
2018 Award
A related thread
In particular, Section 5 of the linked paper says the following:
"5 Isn’t it just the Copenhagen intepretation?
This has been said a lot through the years, and we continue to hear it today. Some-
times, it’s said that QBism is trying to be more Copenhagen than the Copenhagen
interpretation itself. As if QBism had a fever, and the only prescription were more
Copenhagen! But the idea that there ever was a unified “Copenhagen interpretation”
— i.e., that the definite article is remotely applicable — was a myth of the 1950s.
Trying to exceed “the” Copenhagen interpretation in any respect is to race against a
QBism does not have, for example, Bohr’s emphasis on “ordinary language” [11],
whatever that might mean. Nor does it have the quantum-classical cut of Heisenberg, the classical laboratory equipment of Landau and Lifshitz [12], the public experimental records of Pauli [13], the essentially ontic state vectors of early Bohm [14], or the frequentism of early von Neumann [15]. Unlike van Kampen, QBism does not presume that the vanishing of interference terms will solve all riddles [16]. Unlike Wheeler, QBism does not posit that all observers should ideally have the same information about a system and thus the same quantum state for it [13, footnote 9]. There simply is not a way to summarize this overflow of differences by claiming that QBism is “more Copenhagen”.
At one point, the Wikipedia article on QBism claimed that it “is very similar to the Copenhagen interpretation that is commonly taught in textbooks”. What does this even mean? First, as we noted, there’s no such thing as “the” Copenhagen interpretation. In addition, claiming that “the Copenhagen interpretation” is “commonly taught in textbooks” conflates the early developers of quantum theory and the varied modern expositions of it into a vague mishmash. Asher Peres’ textbook is more instrumentalist than the undergraduate standards; the Feynman Lectures handle probability in a less
frequentist way than Peres. Are all common textbooks Copenhagen, or is Copenhagen that which is commonly taught in all textbooks? Better to strike the term “Copenhagen interpretation” from our lexicon going forward and instead be precise about what views we mean!"
Last edited:
Thanks for the replies. A couple followup questions.

And it's also important to note that there are never contradictions between different observers concerning the probabilities for a given situation (state preparation).
Don't Wigner and his friend have different probabilities before the friend interacts with Wigner (after the friend performs a measurement)? Or maybe that's not considered a contradiction because they haven't interacted?

Why QBism is not the Copenhagen interpretation

Lived experience is where we start from and where all must link back to, like a guiding thread

I agree that one's understanding of the world, including any predictions they make, must be based on their personal experience. Isn't that how an experimentalist has always used QT (and in fact Newtonian, ... all theories)? There's no alternative is there?


Science Advisor
Gold Member
Note I use quotation marks for concepts intended to be loose to save tedious levels of exposition.

As mentioned above there can be a lot of variation among what counts as "Copenhagen" both among the founders of QM and within modern day expositions. Matt Leifer uses the phrase "Copenhagenish" for this broad family of views. QBism is certainly within this family, so first let us say what they have in common and then the variations.

The basic core of all these views is that quantum mechanics only provides you with probabilities ##P(E)## to "record" (in quotation marks since this will be part of the variation below) an outcome ##E## witnessed at the macroscopic level. They reject both:
  1. The existence of any variables, commonly denoted ##\lambda##, which would determine the outcome

  2. Thinking of the quantum state itself as a true aspect of the system completely independent of the observer
A note on terminology. These views are often called AntiRealist. This is meant in a technical philosophical sense, not the "they say nothing exists outside of our experiences man" moronic sense one encounters on the internet. It means they all reject the view that QM tells one about microscopic systems "in and of themselves". A better phrase I think is Non-Representational, i.e. the quantum formalism does not directly represent microscopic systems. For anybody not holding these views I think this is their most "shocking" feature.

Now within this general framework we have three points of possible disagreement:
  1. How should one think of the probabilities ##P(E)##. In a Subjective Bayesian, Objective Bayesian, Frequentist, Informational or "naive" fashion? Naive meaning the way people often apply probability theory where they might think Frequentism makes more sense in one situation, Bayesianism in another

  2. What exactly do we mean by record a value, i.e. what counts as an event in this formalism?

  3. When applying quantum theory we often have a cut. We treat the instruments as classical, but the system under study as quantum. What is the fundamental meaning of this?
Sometimes the views on (1) (2) and (3) are essentially independent, in QBism they're related

Variation on (1):
Even in classical probability all the points of view in (1) are espoused. So in a sense this is a debate external to QM, though relevant to it.

QBism takes a very strong stance on point (1) saying that probabilities should be viewed in the Subjective Bayesian school of thought à la de Finetti, Savage or Ramsey. Jeffrey Bub, Richard Healey and many others prefer to view the probabilities of QM to have an information theoretic content which is closely related to the Objective Bayesian school. The Ensemble view of people like Ballentine takes a frequentist view.

Bohr and many of the early founders were not so concerned with how exactly you should think of probability. One can see Bohr use Bayesian and Frequentist points whenever he felt exposition involving one or the other was simpler (see the papers of Arkady Plotnitsky for more on this).

Regardless of what interpretation one takes of probability theory, it will have to be modified somewhat from its classical form due to how probability works in QM. However this isn't the thread for that.

Variation on (2):
This is where most of the differences are encountered. Even Bohr varies over the course of his career on this. Having one view in earlier essays such as the 1929 "The Atomic Theory and the Fundamental Principles underlying the Description of Nature", before coming to his final view in the 1940s seen in such essays as the 1949 "Discussion with Einstein on Epistemological Problems in Atomic Physics".

Some such as Caslav Brukner and Pascual Jordan see the event as being the microscopic system gaining the property being measured. If you measure ##S_z## then the particle now actually has ##S_z = +\frac{1}{2}##, though other properties cannot be so described and remain indeterminate. Bohr's earlier views are like a more limited version of this. The particle has or can be thought of has having ##S_z = +\frac{1}{2}## during the course of the experiment alone.

QBism considers the recording of the event to be when the agent applying quantum theory notices the results of interacting with the system. This ties into their view of probability in (1). However since there are no hidden variables and the formalism is not deterministic what is noticed is fundamentally new.

Others such as Rudolf Haag and Berthold-Georg Englert take the view that since the event ##E## in quantum probabilities ##P(E)## is the result of some event on the microscopic level, tracing it all the way down it ultimately refers some fundamental microscopic occurrence. Though one can say little of this aside from that it causes the macroscopic impression we see. Decoherent histories is similar, but considers the event to be whenever a classical concept is sensible to use as signaled by decoherence.

Bohr's final views were that the events refer purely to macroscopic outcomes. They're properties of the device, not the microscopic system's in any sense. This is what his call for a redefinition of "phenomena" was about.

Variation on (3):
I won't spend too long on this. Heisenberg thought the cut was a pragmatic matter, it could be shifted up and down as one included more degrees of freedom in the quantum treatment.

Bohr thought the cut was more fundamental than being a purely pragmatic concern for philosophical reasons. Ultimately in a scientific account we speak in classical terms about the equipment using terms like "momentum", "position" etc. Since Bohr was already rejecting "position" etc as actual properties of the particles these necessarily refer to the device. This meant we are forced by language to retain some classical elements, the cut is a linguistic/epistemological necessity.

QBists think the cut is a fundamental one in the application of probability theory. The cut is between the agent, those things they leaves outside of the application of probability and those things they are forming beliefs about.

The QBist views on (1), (2) and (3) are all closely related.
Last edited:
A note on terminology. These views are often called AntiRealist.
Justly challenge naive realism.

I would rather speak of another form of realism that takes into account the subject and therefore human scaffolding in the construction of our knowledge.


Want to reply to this thread?

"Qbism Vs. Copenhagen" You must log in or register to reply here.

Physics Forums Values

We Value Quality
• Topics based on mainstream science
• Proper English grammar and spelling
We Value Civility
• Positive and compassionate attitudes
• Patience while debating
We Value Productivity
• Disciplined to remain on-topic
• Recognition of own weaknesses
• Solo and co-op problem solving

Hot Threads