A few quick links:
- Philip Ball at Quanta has a nice article on “Quantum Darwinism” and experiments designed to exhibit actual toy examples of the idea in action (I don’t think “testing” the idea is quite the right language in this context). What’s at issue is the difficult problem of how to understand the way in which classical behavior emerges from an underlying quantum system. For a recent survey article discussing the ideas surrounding Quantum Darwinism, see this from Wojciech Zurek.
Jess Riedel at his blog has a new FAQ About Experimental Quantum Darwinism which gives more detail about what is actually going on here.
- This year’s TASI summer school made the excellent choice of concentrating on issues in quantum field theory. Videos, mostly well worth watching, are available here.
- This month’s Notices of the AMS has a fascinating article about Grothendieck, by Paulo Ribenboim. It comes with a mysterious “Excerpt from” title and editor’s note:
Ribenboim’s original piece contains some additional facts that are not included in this excerpt. Readers interested in the full text should contact the author.
- I’ve finally located a valuable Twitter account, this one.
If someone learns about the “additional facts” on Grothendieck (and Peter doesn’t mind) could they please share their findings in the comments section?
Thanks in advance!
Yeah it would be worth asking Ribenboim if they are willing to publicly share the unedited version, or otherwise whatever was removed. If so I would also be keen to see what was removed.
I’m really happy with Ball’s article. Finally someone has explained correctly what decoherence (and Quantum Darwinism) is and what it isn’t, that it does explain how interference becomes unobservable and the classical results stable, and that it cannot explain the emergence of a unique outcome.
The idea that decoherence somehow solves the measurement problem or explains the collapse of the wavefunction is a common misconception even among people working in the field, and this article might help dispel it.
“The idea that decoherence somehow solves the measurement problem or explains the collapse of the wavefunction is a common misconception.”
It does explain how several possible classical worlds emerge from one quantum world.
It remains to explain the selection of a single classical outcome.
This problem is “solved” in MWI by postulating that all possible outcomes are realized.
In this way, decoherence provides a detailed, microscopic description of the way MWI could work.
Mateus, I think decoherence does a lot more than you suggest to solving the measurement problem. True, it cannot predict which eigenstate will be observed (that remains probabilistic), but it can explain why only a single state (one of the eigenstates) is observed, so it can explain why “interference becomes unobservabke” and “the classical results stable”.
We can understand it mathematically. Eigenstates (which Zurek calls “pointer states” in the article) are more robust because if you apply an operator to an eigenstate, the state remains unchanged, so “classical results stable”. In contrast, any state which is not an eigenstate gets destroyed by the random state phase during the decoherence process, so “interference becomes unobservable”.
As I say, decoherence cannot predict which eigenstate will be observed, but decoherence comes pretty darn close to solving the measurement process. Certainly to the point where it appears that there is no “big unsolved mystery” here, no infinity of parallel universes created every time I measure anything.
All,
Comments about quantum Darwinism welcome, but not the usual arguments about more general issues in measurement theory. About MWI in particular, I’m likely to write something within the next couple months about Sean Carroll’s forthcoming book, and that would be a better time to discuss MWI.
Another Anon,
The point is not predicting the outcome (solving the measurement problem is not about predicting the outcome), but explaining why only a single outcome is observed.
As Ball points out, decoherence only eliminates interference between alternatives, it doesn’t single out any alternative, deterministically or otherwise. It couldn’t, as decoherence can only delete off-diagonal elements of the density matrix, and to single out an alternative one needs to delete diagonal elements.
I’m afraid our host is getting annoyed by the constant discussion of MWI here, so maybe it’s better to leave it for the post on Carroll’s book.
The article may have alluded to this problem…
“Horodecki and other theorists have also sought to embed QD in a theoretical framework that doesn’t demand any arbitrary division of the world into a system and its environment…”
Some folks seem to believe this need for division makes QD entirely circular:
“…’classical’ pointer states do not emerge unless a key aspect of classicality has been tacitly assumed from the beginning…quantum Darwinists smuggle in classicality via their partitioning of the universe into distinguishable systems of interest that interact with mutually randomized environmental subsystems.”
https://physicstoday.scitation.org/doi/full/10.1063/PT.3.2760
What think you? Fatally flawed or not?
LMMI:
I think the ultimate hope of the proponents of quantum Darwinism is that, given a quantum system with dynamics, you can recover the partition into different subsystems that gives pointer states, and that this partition will be unique.
Has this been demonstrated yet? I don’t believe so, but it certainly doesn’t seem out of the question.
Finally, I suspect that there are probably some quantum dynamics which don’t actually give pointer states. I don’t know what quantum Darwinism says about these, but it seems to me that they should be investigated from a quantum Darwinistic point of view.
About circularity: maybe the goal is to show that if we start from a classical-looking universe like ours and apply a quantum evolution, classicality is preserved (for instance, we will never observe the proverbial deal-and-alive cat)?
This looks like a reasonable goal.
If on the other hand they want to show that the “classical pointer states” emerge from some arbitrary (non “classical looking”) initial quantum state, then it looks like a much more dubious goal to me.
Is it the first or second goal that the darwinists have in mind? Or something else entirely?
Pascal: I don’t know whether all the quantum Darwin people have the same goal.
Certainly, if you start with a high-temperature thermal (i.e., completely random) initial state, then it will stay thermal. But if you start with a low-energy state, you might be able to prove something.
Mateus Araújo,
as you write, decoherence cannot explain why only a single outcome is observed. The deeper reason for this is that quantum mechanics in the traditional interpretations has not even the means to express what it means to have a single outcome.
To explain the occurrence of single outcomes one needs an interpretational device that can talk about this. This is done in my thermal interpretation (already twice mentioned in Peter’s blog); see the 5 preprints listed at
https://www.mat.univie.ac.at/~neum/physfaq/therm/
Essentially, decoherence tells roughly the same the same story as the thermal
interpretation, but only in statistical terms, whereas the thermal interpretation refines this to a different, more detailed story for each single case. This is possible since in the thermal interpretation, outcomes are defined as macroscopic expectations approximating the microscopic quantities to be measured, and q-expectations are always single-valued. This makes a big difference in the interpretation of everything!
One comment re the Quanta article on quantum Darwinism. The statement at the end:
“Spectrum broadcast theory (which has only been worked through for a few idealized cases) ”
is not exactly correct in the following sense. Spectrum Broadcast Structures (SBS) have been theoretically found by the Gdansk group in *all* the models where quantum Darwinism (qD) was earlier predicted, apart from the NV centers (we are working on that). And in several more (QED, gravitational decoherence). Actually one of the pillars of the SBS program is to check as many models as doable to gather a theoretical evidence in their favor. So the SBS “has only been worked through for a few idealized cases” to the very much the same extent as qD. An arXiv search on my name will give the relevant papers. I wrote a comment to Quanta but it didn’t seem to get through to the Editors.