On the Weak Point of the Stronger Uncertainty Relation Published in Academia Quantum , 2025, vol. 2, No. 1, (Academia.edu Journals), doi:10.20935/AcadQuant7557
Krzysztof Urbanowski · 2025
Read On the Weak Point of the Stronger Uncertainty Relation Published in Academia Quantum , 2025, vol. 2, No. 1, (Academia.edu Journals), doi:10.20935/AcadQuant7557 by Krzysztof Urbanowski (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-09201.txt.
Introduction
Krzysztof Urbanowski questions a claim made for a newer uncertainty relation in quantum physics, in a paper published in two thousand twenty-five. The relation concerns the sum of variances of two observables that cannot be measured sharply at the same time.
Some papers present this sum rule as stronger than the classic Heisenberg-Robertson relation. Urbanowski tests the claim on a clean case: a system sitting in an eigenstate of one of the two observables. In that case, he shows, the sum rule gives exactly the same results as the classic relation. Worse, it places no bound at all on the variance of the other observable. The paper warns against reading more power into the newer relation than it holds. It is a short, pointed piece of criticism aimed at a specific weak point.
Catalog record
| CREATOR | Krzysztof Urbanowski |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 12 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-09201
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
On the Weak Point of the Stronger Uncertainty Relation Published in Academia Quantum , 2025, vol. 2, No. 1, (Academia.edu Journals), doi:10.20935/AcadQuant7557 Krzysztof Urbanowski Abstract We analyze the uncertainty relation for the sum of variances, which is called in some papers, the stronger uncertainty relation for all incompatible observables. We show that this uncertainty relation for the sum of variances of the observables A A and B B calculated for the eigenstate of one of these observables, (say of B B ), contrary to the suggestions presented in some papers, leads to the same results as the Heisenberg–Robertson uncertainty relation, that it does not give any bounds on the variance of A A . 1 Introduction The most general form of the conventional Heisenberg–Robertson (HR) uncertainty relation [ 1 , 2 , 3 ] , looks as follows \Delta_{\phi}A\cdot\Delta_{\phi}B\;\geq\;\frac{1}{2}\left|\langle[A,B]\rangle_{\phi}\right|, (1) where in general \Delta_{\phi}F=\|\delta_{\phi}F|\phi\rangle\|\geq 0, (2) and \delta_{\phi}F=F-\langle F\rangle_{\phi}\,\mathbb{I} , and \langle F\rangle_{\phi}\stackrel{{\scriptstyle\rm def}}{{=}}\langle\phi|F|\phi\rangle is the average (or expected) value of an observable F in a system whose state is represented by the normalized vector |\phi\rangle\in{\cal H} , provided that |\langle\phi|F|\phi\rangle|<\infty . (Equivalently: \Delta_{\phi}F\equiv\sqrt{\langle F^{2}\rangle_{\phi}-\langle F\rangle_{\phi}^{2}} ). The observable F is represented by hermitian operator F acting in a Hilbert space {\cal H} of states |\phi\rangle and here F=A,B . The problem is that in the case of non–commuting observables A,B the right hand side of Eq. ( 1 ) vanishes for |\phi\rangle being an eigenvector of A (or of B ), even if \Delta_{\phi}B (or \Delta_{\phi}A ) is different from zero. To solve this and other possible problems, the uncertainty relations are considered in some papers, which use the sum of standard deviations (or variances) instead of the product of them (see, e. g., Eq. (13) in [ 4 ] and also [ 5 , 6 , 7 , 8 , 9 ] , and others). One of the proposed solutions to this and similar problems are the so–called "stronger uncertainty relations". Namely, there were proposed in [ 6 ] two new uncertainty relations, stronger than the HR one, based on the sum (\Delta_{\phi}A)^{2}+(\Delta_{\phi}B)^{2} . The first of them has the following form, (\Delta_{\phi}A)^{2}+(\Delta_{\phi}B)^{2}\geq\pm i\langle[A,B]\rangle_{\phi}+\left|\langle\phi|(A\pm iB)|\phi^{\perp}\rangle\right|^{2}, (3) where \langle\phi|\phi^{\perp}\rangle=0 and \|\;|\phi^{\perp}\rangle\|=1 . The second one looks as follows (\Delta_{\phi}A)^{2}+(\Delta_{\phi}B)^{2}\geq\frac{1}{2}\left|\langle\phi^{\perp}_{(A+B)}|(A+B)|\phi\rangle\right|^{2}. (4) (where |\phi_{(A+B)}^{\perp}\rangle is defined using the Aharonov–Vaidman identity [ 10 ] : F|\psi\rangle=\langle F\rangle_{\psi}\,|\psi\rangle+\Delta_{\psi}F\,\,|\psi^{\perp}_{F}\rangle and here \langle\psi|\psi^{\perp}_{F}\rangle=0 and \langle\psi_{F}^{\perp}|\psi^{\perp}_{F}\rangle=1 ). In [ 6 ] one can find a statement that it is guaranteed that these new relations will be nontrivial even in the case that the state for which it is considered is either an eigenvector of A or of B , with [A,B]\neq 0 : In [ 6 ] on page 2 in the sentence preceding Eq. (4) the authors of this paper state that "A second inequality with nontrivial bound even if |\phi\rangle is an eigenstate either of A or of B is" the inequlity numbered here as ( 4 ). Similar statements are supported in many publications, e.g. [ 7 , 8 , 9 ] and other papers using the results published in [ 6 ] . The question is: Are statements of this type true? Unfortunately, the answer is negative. 2 Results By analyzing the "first" and "second" inequalities ( 3 ), ( 4 ) we show here, that statements of the kind presented in Sec. 1 are untrue. Let’s assume that [A,B]\neq 0 and |\psi\rangle=|\psi_{b}\rangle is a normalized eigenvector of B for the eigenvalue b , that is that B|\psi_{b}\rangle=b|\psi_{b}\rangle . This immediately implies that \langle\psi_{b}|B|\psi_{b}\rangle=b and then that \delta_{\psi_{b}}B|\psi_{b}\rangle\equiv 0 . As a result we have that \Delta_{\psi_{b}}(B)=0 and \langle\psi_{b}|[A,B]|\psi_{b}\rangle\equiv 0 . The second term of the right–hand side of the inequality ( 3 ) can be rewritten as follows \langle\psi_{b}|(A\pm iB)|\psi_{b}^{\perp}\rangle=\langle\psi_{b}|A|\psi_{b}^{\perp}\rangle\pm i\langle\psi_{b}|B|\psi_{b}^{\perp}\rangle, (5) where \langle\psi_{b}|\psi_{b}^{\perp}\rangle=0 and \left\|\;|\psi_{b}^{\perp}\rangle\right\|=1 . There is \langle\psi_{b}|B|\psi_{b}^{\perp}\rangle =\langle\psi_{b}^{\perp}|B|\psi_{b}\rangle^{\ast}=b\,\langle\psi_{b}^{\perp}|\psi_{b}\rangle^{\ast}\equiv 0 . So, the only non–zero part in ( 5 ) is \langle\psi_{b}|A|\psi_{b}^{\perp}\rangle . The Aharonov–Vaidman identity [ 10 ] allows the following transformation, \displaystyle\langle\psi_{b}|A|\psi_{b}^{\perp}\rangle \displaystyle= \displaystyle\langle\psi_{b}^{\perp}|A|\psi_{b}\rangle^{\ast} (6) \displaystyle= \displaystyle\left(\langle A\rangle_{\psi_{b}}\,\,\langle\psi_{b}^{\perp}|\psi_{b}\rangle+\Delta_{\psi_{b}}A\,\,\langle\psi_{b}^{\perp}|\psi_{b,A}^{\perp}\rangle\right)^{\ast} \displaystyle\equiv \displaystyle(\Delta_{\psi_{b}}A)\;\langle\psi_{b,A}^{\perp}|\psi_{b}^{\perp}\rangle. (Here \langle\psi_{b}|\psi_{b,A}^{\perp}\rangle=0 , but \langle\psi_{b}^{\perp}|\psi_{b,A}^{\perp}\rangle\neq 0 ). Finally, from ( 3 ) for |\phi\rangle=|\psi_{b}\rangle we get (\Delta_{\psi_{b}}A)^{2}\geq(\Delta_{\psi_{b}}A)^{2}\;\left|\langle\psi_{b,A}^{\perp}|\psi_{b}^{\perp}\rangle\right|^{2}. (7) We will now focus on the analysis of the properties of the inequality ( 4 ). Using the parallelogram law in the Hilbert space it has been proved in [ 6 ] that the inequality ( 4 ) is equivalent to the following one, (\Delta_{\phi}A)^{2}+(\Delta_{\phi}B)^{2}\geq\frac{1}{2}\left(\Delta_{\phi}(A+B)\right)^{2}. (8) Here \Delta_{\phi}(A+B)=\left\|\delta_{\phi}(A+B)|\phi\rangle\right\| and \delta_{\phi}(A+B)|\phi\rangle\equiv\delta_{\phi}A|\phi\rangle+\delta_{\phi}B|\phi\rangle . Indeed, using the Aharonov–Vaidman identity, one finds that (A+B)|\phi\rangle=\langle\phi|(A+B)|\phi\rangle\,|\phi\rangle+\Delta_{\phi}(A+B)|\phi_{(A+B)}^{\perp}\rangle, (9) which means that \langle\phi^{\perp}_{(A+B)}|(A+B)|\phi\rangle\equiv\Delta_{\phi}(A+B), (10) This last Equation shows that inequalities ( 8 ) and ( 4 ) are equivalent. Let’s now analyze properties of the inequality ( 8 ). Let’s assume again that |\phi\rangle=|\psi_{b}\rangle and use some results leading to the inequality ( 7 ). Elementary calculations show that in this case, \Delta_{\psi_{b}}(A+B)\equiv\Delta_{\psi_{b}}(A) . This in turn means that the inequality ( 8 ) takes the following form, (\Delta_{\psi_{b}}A)^{2}\geq\frac{1}{2}\left(\Delta_{\psi_{b}}(A)\right)^{2}, (11) for |\phi\rangle=|\psi_{b}\rangle . By assumption A and B have not common eigenvectors. Therefore \Delta_{\psi_{b}}(A)>0 . This means that we have two solutions of the inequality ( 7 ). The first is: 1\geq\left|\langle\psi_{b,A}^{\perp}|\psi_{b}^{\perp}\rangle\right|^{2}, (12) and the second is: (1-\left|\langle\psi_{b,A}^{\perp}|\psi_{b}^{\perp}\rangle\right|^{2})\,(\Delta_{\phi}A)^{2}\geq 0, (13) which implies that (\Delta_{\psi_{b}}A)^{2}\geq 0. (14) Let us now analyze solutions of the inequality ( 11 ). There are two absolutely trivial solutions: The first is 1>1/2 , and the second is (\Delta_{\psi_{b}}A)^{2}\geq 0, (15) which is identical to the solution ( 14 ) for the inequality ( 7 ). 3 Discussion Results ( 7 ), ( 11 ), and ( 12 ) — ( 15 ) show that the right hand side of inequalities ( 3 ), ( 4 ) are non–zero when |\phi\rangle=|\psi_{b}\rangle . The question that arises is as follows: Is this really a nontrivial lower bound? Does it mean that the right–hand side of ( 3 ) and of ( 4 ) is different from zero or that these inequalities allow us to obtain useful information about the variance of the observable A in the case considered in this note. In my opinion the only reasonable answer for this question is to choose the second possibility. The only useful conclusions that can be drawn from ( 7 ) and ( 11 ), resulting from the considered example of the state |\phi\rangle=|\psi_{b}\rangle , are that (\Delta_{\psi_{b}}A)^{2}\geq 0 and they are absolutely trivial and useless: The same information was already known earlier — it results directly from the definition ( 2 ). In [ 6 ] , and [ 7 , 8 , 9 ] and in many other papers it is stated that in order to obtain a true lower bound, one should choose the maximum value of one of the two lower bounds obtained from inequalities ( 3 ) and ( 4 ). Unfortunately, in the case considered this procedure also does not provide any useful information about the lower bound for (\Delta_{\psi_{b}}A)^{2} . The only conclusion is (\Delta_{\psi_{b}}A)^{2}\geq 0 . Note that the similar result can be obtained from the relation ( 1 ) for the considered example. Indeed, if |\phi\rangle=|\psi_{b}\rangle , then as already shown above \langle\psi_{b}|[A,B]|\psi_{b}\rangle\equiv 0 and \Delta_{\psi_{b}}(B)=0 . Thus, the inequality ( 1 ) takes the following form: \Delta_{\psi_{b}}A\,\cdot\,0\,\geq\,0 . The product \Delta_{\psi_{b}}A\,\cdot\,0 is a real number only