Consistent Histories solipsism?

[JamieSalaor]

TL;DR Summary

Consistent Histories interpretation question on relativism / solipsism

I have recently finished reading Fay Dowker and Adrian Kent's 1994 paper On the Consistent Histories Approach to Quantum Mechanics.

As I understand it this interpretation of QM is not very popular, collecting only two votes from the 2016 survey - Surveying the Attitudes of Physicists Concerning Foundational Issues of Quantum Mechanics.

In the paper it appears as if they believe this interpretation implies solipsism (which may be why it is so unpopular).

However, I'm not entirely sure what they mean by this.
There seems to be two definitions of solipsism; epistemological and metaphysical.
With epistemological saying that we cannot be sure about the true nature of reality, but not outright denying other minds, and with metaphysical straight up denying other minds and external reality.

I have noted that in the paper they believe solipsism to be an unreasonable position. Which I too agree with (along with most other philosophers and physicists).

I have had a couple of brief chats with supporters of the Consistent Histories interpretation. They deny this framework entails solipsism.

They instead believe that their framework implies a form of 'relativism'.
For example:
If observer A uses framework S, and observer B uses framework S', then each observer draws his own conclusions on the history of the system (in our example, A will say that at t1 the particle was in the interval [0.1-0.2], and B will insist that the particle at t1 had energy E=13.6), but they can't combine their results into a unique, consistent picture of reality.

However, since these observers with their different frameworks can still communicate their contrasting results and in turn disagree. Surely if observers can disagree then there can be no metaphysical solipsism. Just a lack of real knowledge about the outside world? Therefore, since observers still communicate this cannot imply metaphysical solipsism.
Instead, they have argued it is relative.
I see this as comparable to relativity of simultaneity, how we can disagree on the timing of an event...
Therefore, I can see how this may imply epistemological solipsism, in the sense that through this interpretation we all come out equally clueless about the true nature of reality...
But I don't think it denies other observers...

Does anyone who properly understand the interpretation have anything to say?

Would appreciate any comments

 

[Doc Al]

You might want to take a look at this recent thread (if you haven't already): https://www.physicsforums.com/threads/consistent-histories-interpretation-history.997837/

 

[JamieSalaor]

You might want to take a look at this recent thread (if you haven't already): https://www.physicsforums.com/threads/consistent-histories-interpretation-history.997837/

I did yeah. I didn't really get a straight answer from that thread. I also agree that consistent Histories does not at all deny the existence of dinosaurs...
I think the original poster was really struggling to interpret the interpretation...

Then again so am I

Cheers though (:

 

[Morbert]

CH is not solipsistic. The conclusions drawn from one framework do not invalidate conclusions drawn from another framework, so long as the inference processes belong to the boolean algebra of the respective frameworks. I can be confident that any conclusion draw from a framework will be reliable even while acknowledging other frameworks are valid. I do not need to suppose that I am the only being that exists.

However, since these observers with their different frameworks can still communicate their contrasting results and in turn disagree.

Disagreement implies the two have combined their incompatible frameworks together and have therefore violated the inference rules of the respective boolean algrebas. But they can always construct a framework and algebra that contains both of their measurement outcomes. I will use a simple example. Say there are two particles $a$ and $b$ prepared in the state $|\omega_a^+,\omega_b^+\rangle$ at time $t_0$ (i.e. both have a spin-$\omega$ of 1/2). Alice measures the spin-$X$ of particle $a$ with apparatus $A$, and Bob measures the spin-$Z$ of particle $b$ with apparatus $B$. Consider the times $t_1$ before measurement, and $t_2$ after measurement. To build an appropriate histories framework, we build a Hilbert space for these times $$\mathcal{H} = \mathcal{H}_{t_1}\odot\mathcal{H}_{t_2}$$The symbol $\odot$ just denotes $\otimes$ for spaces at different times. Each space at a given time is a tensor product of the space for the particles and the apparatuses $$\mathcal{H}_{t_i} = \mathcal{S}_{t_i}\otimes\mathcal{A}_{t_i}\otimes\mathcal{B}_{t_i}$$I'll use Griffiths's notation: a projector is written with square brackets, and identity operators are implicit. E.g.\begin{eqnarray*}
|X_a^+ ,X_b^+\rangle\langle X_a^+ ,X_b^+| =\left[X_a^+ ,X_b^+\right] = \left[X_a^+,X_b^+\right]\otimes I_\mathcal{A}\otimes I_\mathcal{B}
\end{eqnarray*}So let's say Alice constructs a framework appropriate for her measurement. It might consist of the two histories\begin{eqnarray*}
\left[X^+_a,X^+_b,A\right]&\odot&\left[A^+\right]\\
\left[X^-_a,X^-_b,A\right]&\odot&\left[A^-\right]
\end{eqnarray*}She can infer spin-$X$ properties of the particles at time $t_1$, based on whether or not her apparatus returns $+$ or $-$ at $t_2$. Similarly, Bob constructs a framework appropriate for his measurement
\begin{eqnarray*}
\left[Z^+_a,Z^+_b,B\right]&\odot&\left[B^+\right]\\
\left[Z^-_a,Z^-_b,B\right]&\odot&\left[B^-\right]
\end{eqnarray*} and he can draw conclusions about spin-$Z$. However, if they want to draw conclusions based on both of their measurements, they cannot simply combine these two frameworks, as they will not be decoherent/consistent. But this does not mean they cannot use a framework that includes both of their measurement results. E.g. The set \begin{eqnarray*}
\left[X^+_a,Z^+_b,A,B\right]&\odot&\left[A^+,B^+\right]\\
\left[X^+_a,Z^-_b,A,B\right]&\odot&\left[A^+,B^-\right]\\
\left[X^-_a,Z^+_b,A,B\right]&\odot&\left[A^-,B^+\right]\\
\left[X^-_a,Z^-_b,A,B\right]&\odot&\left[A^-,B^-\right]
\end{eqnarray*} is consistent, and includes both of their measurement results, so any inference they draw using this framework will be valid.

This is certainly odd, and not a feature of classical physics (where history frameworks can always be combined through some common refinement), but if one finds the ontological implications troublesome, they can always simply interpret the history frameworks as realms of valid logical inferences from initial data, without attaching ontological meaning to the histories.

 

[JamieSalaor]

Amazing answer thank you, that actually makes a lot of sense. It's nice to be able to communicate with people who actually understand this stuff.
I have a quick question referring to the following statement
However, if they want to draw conclusions based on both of their measurements, they cannot simply combine these two frameworks, as they will not be decoherent/consistent. But this does not mean they cannot use a framework that includes both of their measurement results.

To understand this qualitatively may I describe a scenario?
Say Alice measures the spin of her particle S and Bob measures the spin of his particle S'. Alice measures spin up, Bob measures spin down.
They then meet and discuss their measurements.
If we want to represent this process through the Consistent Histories Interpretation, would it then be necessary to construct a framework where both their measurement results are illustrated.

Am I right in saying frameworks are a mode of describing the world? They are a useful tool to ensure consistency.

As I see it this does not imply solipsism at all. Observers still communicate and other observers exist. It is just necessary to formulate interactions into suitable frameworks.

If I grossly misunderstood this please correct me.
The mathematics is beyond me (I'm a philosophy student)
I am just attempting to understand the ontology and I suppose the metaphysics of this interpretation.

So would their be any chance if I am completely off the mark that you could explain this qualitatively.
Thank you

 

[Demystifier]

but if one finds the ontological implications troublesome, they can always simply interpret the history frameworks as realms of valid logical inferences from initial data, without attaching ontological meaning to the histories.

That's my main conceptual problem with CH. If the initial data is ontological, then how can it be that the logical inferences from those data are not ontological?

 

[atyy]

So would their be any chance if I am completely off the mark that you could explain this qualitatively.

The ontology is that there is no single underlying reality. This is sometimes stated as not having 'one "real" fine-grained history'. You can read the background in https://arxiv.org/abs/1106.0767, which tries to correct this, but at the price of having negative (!) probabilities.

 

[JamieSalaor]

The ontology is that there is no single underlying reality. This is sometimes stated as not having 'one "real" fine-grained history'. You can read the background in https://arxiv.org/abs/1106.0767, which tries to correct this, but at the price of having negative (!) probabilities.

Fine grained ontologies can vary. But course grained ontologies as described there are consistent?
Us as IGUS's are large macroscopic objects, therefore we all experience pretty much the same reality?
Which is why there is so much consistency of results.

Still seems as if the interpretation does not imply solipsism.
Just a lack of objectivity in fine grained histories

 

[atyy]

Fine grained ontologies can vary. But course grained ontologies as described there are consistent?
Us as IGUS's are large macroscopic objects, therefore we all experience pretty much the same reality?
Which is why there is so much consistency of results.

Still seems as if the interpretation does not imply solipsism.
Just a lack of objectivity in fine grained histories

I'm not sure what solipsism is, but I'm already lost at the part where there is no single underlying reality.

 

[Demystifier]

The ontology is that there is no single underlying reality.

Is it different from "The ontology is that there is no ontology"?

 

[Demystifier]

Perhaps the most concise explanation of CH is: Copenhagen-like complementarity, but without measurements.

 

[JamieSalaor]

Solipsism is the nutty idea that nothing exists except the self...
I don't think anyone actually believes in it. I don't think it's there to believe in. It's obviously false.
I think usually when an interpretation implies solipsism something is wrong with it or it's incomplete.
However, some interpretations that put consciousness on a pedestal aren't far off...

 

[atyy]

Solipsism is the nutty idea that nothing exists except the self...
I don't think anyone actually believes in it. I don't think it's there to believe in. It's obviously false.
I think usually when an interpretation implies solipsism something is wrong with it or it's incomplete.
However, some interpretations that put consciousness on a pedestal aren't far off...

Well, Copenhagen allows one to say what is real and what is not, and Copenhagen can exclude nonlocality by denying the reality of at least some other observers. Copenhagen doesn't necessarily deny an underlying reality, but can be interpreted more as a limitation on what we can predict. However, some flavors of Copenhagen are close to solipsism, since they can deny the reality of other observers. Consistent histories is sometimes called "Copenhagen done right" so maybe that's where the idea that CH is close to solipsism comes from. I myself like Copenhagen as a nice practical interpretation, but I also like a single underlying reality, so I usually think CH is "Copenhagen done wrong".

 

[JamieSalaor]

Well, Copenhagen allows one to say what is real and what is not, and Copenhagen can exclude nonlocality by denying the reality of at least some other observers. Copenhagen doesn't necessarily deny an underlying reality, but can be interpreted more as a limitation on what we can predict. However, some flavors of Copenhagen are close to solipsism, since they can deny the reality of other observers. Consistent histories is sometimes called "Copenhagen done right" so maybe that's where the idea that CH is close to solipsism comes from. I myself like Copenhagen as a nice practical interpretation, but I also like a single underlying reality, so I usually think CH is "Copenhagen done wrong".

I don't really think Copenhagen denies other observers. It just says that things don't have defined states until measured. Since basically anything can do the measuring, most classical things are in classically definable states due to decoherence. Like how the moon exists when we are not looking, due to decoherence...
I think as you say some Copenhagen flavours, such as von-Neuman-Wigner which posits that the consciousness causes collapse could be seen as solipsism as they're very idealistic frameworks.
But these ideas are largely disregarded, and I think for good reason... It's excessively anthropocentric and if anything arrogant.

 

[Demystifier]

I don't think anyone actually believes in it. I don't think it's there to believe in.

Solipsism is there to make clear why some philosophical stance looks untenable: "But if you really believe that, you in fact believe in solipsism."

 

[JamieSalaor]

Solipsism is there to make clear why some philosophical stance looks untenable: "But if you really believe that, you in fact believe in solipsism."

Yes exactly.
I'm sure it's possible to formulate QM in such a way that solipsism would be the case. But nobody would take it seriously. It is nonsense. But it's definitely interesting to think about, but I don't think it's worth more than that.

 

[Demystifier]

I myself like Copenhagen as a nice practical interpretation, but I also like a single underlying reality

Then my instrumental Bohmian mechanics should be perfect for you.

 

[Demystifier]

Yes exactly.
I'm sure it's possible to formulate QM in such a way that solipsism would be the case. But nobody would take it seriously. It is nonsense. But it's definitely interesting to think about, but I don't think it's worth more than that.

Then you might enjoy my
https://arxiv.org/abs/1112.2034

 

[Demystifier]

... so I usually think CH is "Copenhagen done wrong".

Or perhaps, CH is "Copenhagen overdone".

 

[JamieSalaor]

Then you might enjoy my
https://arxiv.org/abs/1112.2034

You know what is really funny.
Just the other week when doing some research for a piece of coursework I came across a lecture on YouTube covering this exact topic...
I'm assuming you're the author and presenter then? Really interesting piece.
I gathered it's not for taking seriously, but for illustrating that conclusions that deny non-locality and realism are broadly unacceptable because they can lead to daft ideas like solipsism (assuming you demand hidden variables).

 

[EPR]

This discussion brings up the old bag of unknowns...

Has anyone established what constitutes a measurement device, a measurement and what is being measured(theoretical wave functions)? How do theoretical wave functions(not to be confused with classical electromagnetic waves!) interact with other theoretical wave functions to cause an instant 'collapse' that many here claim happens everywhere all the time?

Where does the confidence come from in hand waving propositions most don't have a good grasp of? Esp. ones that make no assumptions at all.

The CI makes very few claims and zero assumptions - it's bound to survive the longest.

 

[Morbert]

That's my main conceptual problem with CH. If the initial data is ontological, then how can it be that the logical inferences from those data are not ontological?

This feature (propositions that do not describe facts themselves, but connect preparations to resultant facts through logical inferences) is analogous to the way time-evolution of a not-real wavefunction can connect a real preparation to real possible measurement outcomes in textbook CI. A CI proponent would not commit to the wavefunction as real, and a CH proponent would not commit to every proposition in the logic as describing a possible fact.

But CH logics are more versatile than CI equivalents. E.g. They can be extended to include either the preparation process or the measurement process. The permit intuitive reasoning within their framework. They can be applied to closed systems. They offer a sharper account of measurement scenarios etc.

In this sense CH has more modest ambitions than other projects like Bohmian Mechanics or Many-Worlds.

 

[Morbert]

Has anyone established what constitutes a measurement device, a measurement and what is being measured(theoretical wave functions)?

CH gives quite precise answers to these questions. A measurement scenario is the establishment of a logical equivalent between a property of a microscopic (or macroscopic) system, and a property of the measurement apparatus that constitutes a decoherent irreversible datum/record. I.e. If $\{a_i\}$ are possible properties of the measured system and $\{\epsilon_i\}$ are possible data produced by the measurement apparatus, then we have a measurement scenario of the data are are irreversible and if $$p(a_i) \approx p(\epsilon_i) \approx p(a_i \land \epsilon_i)$$

 

[Demystifier]

You know what is really funny.
Just the other week when doing some research for a piece of coursework I came across a lecture on YouTube covering this exact topic...
I'm assuming you're the author and presenter then? Really interesting piece.
I gathered it's not for taking seriously, but for illustrating that conclusions that deny non-locality and realism are broadly unacceptable because they can lead to daft ideas like solipsism (assuming you demand hidden variables).

Yes. But I don't like how I sound on this YouTube video, my accent is much worse than I like to think.

 

[Demystifier]

In this sense CH has more modest ambitions than other projects like Bohmian Mechanics or Many-Worlds.

In your opinion, what is the advantage of CH over Bohmian mechanics or many worlds?

 

[Morbert]

In your opinion, what is the advantage of CH over Bohmian mechanics or many worlds?

I haven't followed the developments of either too closely but:

I doesn't seem to be clear how probabilities enter into the MWI. (See section C here https://arxiv.org/pdf/quant-ph/0405161.pdf ). In CH, facts are assumed to obtain probabilistically, and probabilities are a ready feature of the language.

CH seems to be automatically applicable to any quantum theory. Once the physics of a Hamiltonian and a state space are worked out*, you can readily go from state space to sample space to event algebra to boolean logic of CH. To quote Hartle: "Decoherent (or consistent) histories quantum mechanics (DH) is logically consistent, consistent with experiment as far as is known, consistent with the rest of modern physics such as special relativity and quantum field theory, general enough or cosmology, and generalizable to apply to semiclassical quantum spacetime." My understanding is there is still work to do re/ generealiseability of Bohmian mechanics.

*We could even presumably write down a consistent histories interpretation of Bohmian mechanics but it would likely be redundant.

 

[Morbert]

The ontology is that there is no single underlying reality. This is sometimes stated as not having 'one "real" fine-grained history'. You can read the background in https://arxiv.org/abs/1106.0767, which tries to correct this, but at the price of having negative (!) probabilities.

One subtle point here. Ontological flavours of CH still only posit a singular reality. Gell-Mann and Hartle sometimes call history frameworks "realms" which certainly doesn't help matters, but all history frameworks still describe the same reality. It's just that this reality cannot be completely described with a single logic. This is still a deal breaker for some people, but still worth emphasising that CH is a "one world" interpretation.

 

[Demystifier]

It's just that this reality cannot be completely described with a single logic.

That's very hard to digest, that "reality cannot be completely described with a single logic". Is there some analogy or example unrelated to quantum mechanics?

 

[EPR]

Yup. Wouldn't causality fall apart with no single logic? Everyone has seen events for which no causes can be attributed, but this is quite rare. Certainly more rare than observing things pop out of nowhere into existence. In this regard, this reality is consistent to a very high level.

 

[Morbert]

Yup. Wouldn't causality fall apart with no single logic? Everyone has seen events for which no causes can be attributed, but this is quite rare. Certainly more rare than observing things pop out of nowhere into existence. In this regard, this reality is consistent to a very high level.

One of the primary motivations of the CH approach is to accommodate the deterministic regularities of our classical world in probabilistic quantum theories. See e.g. this paper by Gell-Mann and Hartle (preprint). Far from breaking causality, the logics of CH are considered in order to enable a broader discussion of regular causal relations.

 

[Morbert]

That's very hard to digest, that "reality cannot be completely described with a single logic". Is there some analogy or example unrelated to quantum mechanics?

Nothing immediate springs to mind. In classical physics we can build multiple valid but incompatible logics if we want, but there is always some logic built from a common refinement of the sample spaces associated with these logics. The Spekkens toy theory in a sense artificially restricts this common refinement, but is interpreted as an epistemic limit rather than an ontological one.

 

[EPR]

Decoherence on its own does not provide or guarantee classically consistent past/history where causal relationships with the present are maintained.

Yet - we always find causal connection of the decohered present with the decohered past(be it dinasaur bones or remnants of the second World War)

 

[JamieSalaor]

One subtle point here. Ontological flavours of CH still only posit a singular reality. Gell-Mann and Hartle sometimes call history frameworks "realms" which certainly doesn't help matters, but all history frameworks still describe the same reality. It's just that this reality cannot be completely described with a single logic. This is still a deal breaker for some people, but still worth emphasising that CH is a "one world" interpretation.

Then after all this why was CH accused of being solipsistic?
From the sounds of things it definitely isn't...

 

[PaoloDiM]

Summary:: Consistent Histories interpretation question on relativism / solipsism

I have recently finished reading Fay Dowker and Adrian Kent's 1994 paper On the Consistent Histories Approach to Quantum Mechanics.

As I understand it this interpretation of QM is not very popular, collecting only two votes from the 2016 survey - Surveying the Attitudes of Physicists Concerning Foundational Issues of Quantum Mechanics.

In the paper it appears as if they believe this interpretation implies solipsism (which may be why it is so unpopular).

However, I'm not entirely sure what they mean by this.
There seems to be two definitions of solipsism; epistemological and metaphysical.
With epistemological saying that we cannot be sure about the true nature of reality, but not outright denying other minds, and with metaphysical straight up denying other minds and external reality.

I have noted that in the paper they believe solipsism to be an unreasonable position. Which I too agree with (along with most other philosophers and physicists).

I have had a couple of brief chats with supporters of the Consistent Histories interpretation. They deny this framework entails solipsism.

They instead believe that their framework implies a form of 'relativism'.
For example:
If observer A uses framework S, and observer B uses framework S', then each observer draws his own conclusions on the history of the system (in our example, A will say that at t1 the particle was in the interval [0.1-0.2], and B will insist that the particle at t1 had energy E=13.6), but they can't combine their results into a unique, consistent picture of reality.

However, since these observers with their different frameworks can still communicate their contrasting results and in turn disagree. Surely if observers can disagree then there can be no metaphysical solipsism. Just a lack of real knowledge about the outside world? Therefore, since observers still communicate this cannot imply metaphysical solipsism.
Instead, they have argued it is relative.
I see this as comparable to relativity of simultaneity, how we can disagree on the timing of an event...
Therefore, I can see how this may imply epistemological solipsism, in the sense that through this interpretation we all come out equally clueless about the true nature of reality...
But I don't think it denies other observers...

Does anyone who properly understand the interpretation have anything to say?

Would appreciate any comments

Interesting post, thanks. Would a discussion of language here be too far afield here? For example, how is any disagreement measured, whether between persons or between observations, but through language? Does the same string of words refer to the same mental object (a theory in QM, for example)?
If this is too far afield, pardon the intrusion.

 

[JamieSalaor]

Interesting post, thanks. Would a discussion of language here be too far afield here? For example, how is any disagreement measured, whether between persons or between observations, but through language? Does the same string of words refer to the same mental object (a theory in QM, for example)?
If this is too far afield, pardon the intrusion.

I think yes, this is a one world interpretation so I'd assume when we talk about the same thing, we actually talk about the same thing. We truly communicate.
If that's what you mean
Pardon any misunderstanding