# A note on different definitions of momentum disturbance

- 841 Downloads
- 4 Citations

## Abstract

It was recently shown that it is possible to construct an error–disturbance relation having the same form as Heisenberg’s original heuristic definition (Busch et al., Phys Rev Lett 111:160405, 2013), in contrast to the formula proposed by Ozawa which we and others previously confirmed experimentally. With Ozawa’s definitions of measurement error and disturbance, a relation of Heisenberg’s form is not in general valid, and a new error–disturbance relationship can be derived. Here we explain the different physical significance of the two definitions, and suggest that Ozawa’s definition better corresponds to the usual understanding of the disturbance that Heisenberg discussed.

## Keywords

Quantum mechanics Quantum information Quantum measurement## 1 Introduction

Recently, a paper by Busch et al. [7] appeared claiming to invalidate this work, arguing in favour of a different set of definitions of error and disturbance, and proving that Heisenberg’s original expression was in fact rigorously correct by these definitions. In what follows, we will clarify the origin of this seeming contradiction. In short, our reading of [7] is that the disturbance of Busch et al. is a property of the measurement device (being maximized over all input states) and is thus “state independent”. As a result, their work shows that any measurement which is capable of achieving a measurement precision of \(\Delta {X}\) must impart a momentum disturbance of \({\hbar }/(2\Delta {X})\) on *some* state; however, there are many states which could pass through the measurement apparatus undisturbed. Ozawa’s result, on the other hand, demonstrates that when a measurement is done on a particular state, with a precision of \(\Delta {X}\), the momentum disturbance *to that state* need not be as large as \({\hbar }/(2\Delta {X})\)—i.e. it is “state dependent”.

It is important to make the distinction between state-dependent and independent relationships for several reasons. First, the original Robertson relation is state dependent. By this we mean that the variance of both observables (\(\Delta {X}\) and \(\Delta {P}\)), and in general the right-hand side of the relationship (the expectation value of their commutator, \(\langle [\hat{X},\hat{P}]\rangle \)) is calculated for a given state—and not maximized over all states as are Busch et al.’s definitions of error and disturbance. As such, it might be expected that an error–disturbance formulation of Heisenberg’s ideas would also be state dependent. Second, in a quantum key distribution setting, entropic forms of the Robertson relation have been used in security proofs of certain protocols [8, 9]. In such scenarios, a state-dependent form of the relationship is essential, since the performance of a protocol must be quantified on a given run. This is not to say that a state-independent formulation has no use, but it is why a state-dependent error–disturbance relationship could prove essential. Such a state-dependent relationship is what we [6], the Hasegawa group [5], and now several others [10, 11, 12, 13, 14] have confirmed experimentally. The point of our present paper is to explain clearly how the difference of definitions arises. We will also point to some specific examples of disturbance which we argue are better captured by Ozawa’s definition than by Busch et al.’s.

## 2 Busch et al.’s disturbance

*some states*measure position to an accuracy of \(\Delta {X}\) must be able to disturb the momentum of

*some state*by at least \(\Delta {P}={\hbar }/(2\Delta {X})\) (while other states could be disturbed significantly less). On physical grounds, it is clear that this must be the case: if a state begins in a momentum eigenstate and \(\hat{X}\) is then measured to a precision of \(\Delta {X}\), the final state would be required to possess an uncertainty in momentum, and that uncertainty could only come from the measurement. Ozawa instead asks whether or not a measuring device must disturb the momentum of

*every state*by such a large amount. As has been shown experimentally, it need not do so [5, 6, 10, 11, 12].

## 3 Ozawa’s disturbance

Consider the position-disturbance of a process (not necessarily a measurement) which simply flips a particle’s position wave function, taking \(\hat{X}\) to \(-\hat{X}\). Ozawa’s definition, applied to position, is \(\eta _\mathrm{o}(\hat{X})= \langle (\hat{U}^\dag \hat{X} \hat{U}-\hat{X})^2\rangle ^\frac{1}{2}\). For this process, \(\hat{U}^\dag \hat{X}\hat{U}=-\hat{X}\), so we have \(\eta _\mathrm{o}(\hat{X})=2\langle \hat{X}^2\rangle ^\frac{1}{2}\). This is at least equal to \(2\Delta {X}\). (The exact equality holds if \(\langle \hat{X}\rangle =0\).) But note that in the case of a symmetric wavefunction the position probability distribution will not change, therefore, the unmaximized disturbance defined by Busch et al. will be zero. This is reminiscent of the debate over the momentum disturbance required (or not) to destroy double-slit interference [17, 18]; our group recently applied weak measurements in that case to show how non-zero disturbance can exist even when measures of “average disturbance” vanish [19].

To make it clear why it can be useful to have a measure of disturbance which does not vanish in such cases, imagine a related but more complicated situation: a symmetric, localized, one-dimensional position wavefunction, in the presence of dispersion. Due to the dispersion, this wavefunction will broaden, and after some time the higher momentum components will be predominantly on one side of the wavefunction and the lower momentum components near the other side. If nothing is done, the wavefunction will continue to broaden. However, if this wavefunction is flipped about its center, it will begin to narrow: clearly an observable effect. However, immediately after the process (of flipping the wavefunction) both the position and the momentum probability distributions are unchanged. While these distribution-based measures vanish, there is clearly a sense in which individual particles must have had their positions and/or momenta disturbed in order for their future evolution to be modified; this disturbance is captured by Ozawa’s definition.

## 4 Ozawa’s relationship

To understand Ozawa’s relationship (Eq. 5), let us first consider why it is that Eqs. 2 and 1 are often confused—a confusion coming from an understandable mistake. If we measure the position of a particle with some associated error, \(\epsilon (\hat{X})\), the post-measurement state is collapsed into a state having a wavefunction with a spread of at most \(\epsilon (\hat{X})\). If this is the case, then since the state must satisfy Eq. (1), it must have a width in momentum of \(\Delta {P} \ge {\hbar }/(2\epsilon (\hat{X}))\). The mistake is to assume that this momentum uncertainty is all due to the measurement disturbance (i.e., \(\eta (\hat{P}) = \Delta {P}\)); then Eq. (2) follows immediately. What this argument misses is that the final state may possess momentum uncertainty from the initial state as well as from the measurement interaction. If the particle’s momentum is already sufficiently uncertain to satisfy the Robertson relationship, then there is no obvious requirement for the measurement to disturb it further.

^{1}(Fig. 1a). So, we see that we can perform a measurement with finite \(\epsilon (\hat{X})\) which does not disturb the particle’s momentum and thus \(\epsilon (\hat{X})\eta (\hat{P})=0\), in contradiction with Eq. (2). Although the final momentum uncertainty must satisfy the Robertson relation (with the small post-measurement position uncertainty), some of this final uncertainty may come from the initial uncertainty (\(\Delta P\)) rather than relying on a contribution from the measurement (\(\eta (\hat{P})\)).

## 5 Conclusions

In summary, we have discussed Ozawa and Busch et al.’s different definitions of error and disturbance, and their resulting constraints. Busch et al.’s definitions are maximized over all states and are thus state-independent, while Ozawa’s definitions are state-dependent, applying to any state that is measured. The reason for the different definitions may depend on the point of view: Ozawa’s state-dependent definition is more appropriate when considering the disturbance a given quantum state will experience, while the state-independent approach of Busch et al. may be more natural for characterizing a measurement apparatus without reference to specific states. When applied to the disturbance, Busch et al.’s state-dependent definition describes the disturbing power of a measuring device, quantifying how much the measurement could disturb some hypothetical state. We have also pointed to a situation which Busch et al.’s unmaximized disturbance would assign a value of zero, but that Ozawa’s disturbance would better quantify. Finally, we contend that Ozawa’s definition is closer in spirit to the disturbance typically associated with Heisenberg’s microscope than the definition of Busch et al.

Since the original submission of this paper (to another journal, which rejected it based on the criticisms of the very workers with whose work we were taking issue and whom we had requested not be used as referees, and despite our responses thereto), several other works have appeared attempting to clarify the different definitions of precision and disturbance and their role in discussions of the interpretation of the uncertainty principle and measurement–disturbance relations; see for instance [23, 24, 25, 26].

## Footnotes

- 1.
For the finite slit example with \(\epsilon (\hat{X})>\Delta {X}\), Busch et al.’s unmaximized disturbance for the state under consideration is clearly zero, since the probability distribution is unchanged (although their maximized disturbance will be non-zero). Ozawa’s disturbance is also zero. This follows from the fact that \(\hat{U}\psi (x)=\psi (x)\) (which is due the compact support of our initial wavefunction) and Eq. (3).

## Notes

### Acknowledgments

We thank Howard Wiseman and Holger Hofmann for valuable discussions.

## References

- 1.Robertson, H.P.: The uncertainty principle. Phys. Rev.
**34**, 16 (1929)CrossRefGoogle Scholar - 2.Kennard, E.H.: Zur quantenmechanik einfacher bewegungstypen. Z. Phys.
**44**, 326 (1927)MATHGoogle Scholar - 3.Weyl, H.: Gruppentheorie Und Quantenmechanik. Hirzel, Leipzig (1928)MATHGoogle Scholar
- 4.Ozawa, M.: Universally valid reformulation of the heisenberg uncertainty principle on noise and disturbance in measurement. Phys. Rev. A
**67**, 042105 (2003)CrossRefGoogle Scholar - 5.Erhart, J., Sponar, S., Sulyok, G., Badurek, G., Ozawa, M., Hasegawa, Y.: Experimental demonstration of a universally valid error-disturbance uncertainty relation in spin measurements. Nat. Phys.
**8**, 185 (2012)CrossRefGoogle Scholar - 6.Rozema, L.A., Darabi, A., Mahler, D.H., Hayat, A., Soudagar, Y., Steinberg, A.M.: Violation of heisenberg’s measurement-disturbance relationship by weak measurements. Phys. Rev. Lett.
**109**, 100404 (2012)CrossRefGoogle Scholar - 7.Busch, P., Lahti, P., Werner, R.F.: Proof of heisenberg’s error-disturbance relation. Phys. Rev. Lett.
**111**, 160405 (2013)CrossRefGoogle Scholar - 8.Berta, M., Christandl, M., Colbeck, R., Renes, J.M., Renner, R.: The uncertainty principle in the presence of quantum memory. Nat. Phys.
**6**(9), 659 (2010)CrossRefGoogle Scholar - 9.Prevedel, R., Hamel, D.R., Colbeck, R., Fisher, K., Resch, K.J.: Experimental investigation of the uncertainty principle in the presence of quantum memory and its application to witnessing entanglement. Nat. Phys.
**7**(10), 757 (2011)CrossRefGoogle Scholar - 10.Weston, M.M., Hall, M.J.W., Palsson, M.S., Wiseman, H.M., Pryde, G.J.: Experimental test of universal complementarity relations. Phys. Rev. Lett.
**110**, 220402 (2013)CrossRefGoogle Scholar - 11.Baek, S.Y., Kaneda, F., Ozawa, M., Edamatsu, K.: Experimental violation and reformulation of the Heisenberg’s error-disturbance uncertainty relation. Sci. Rep. 3 (2013)Google Scholar
- 12.Ringbauer, M., Biggerstaff, D.N., Broome, M.A.: Fedrizzi Alessandro, C., White, A.G.: Experimental joint quantum measurements with minimum uncertainty. Phys. Rev. Lett.
**112**, 020401 (2014)CrossRefGoogle Scholar - 13.Kaneda, F., Baek, S.Y., Ozawa, M., Edamatsu, K.: Experimental test of error-disturbance uncertainty relations by weak measurement. Phys. Rev. Lett.
**112**, 020402 (2014)CrossRefGoogle Scholar - 14.Sulyok, G., Sponar, S., Erhart, J., Badurek, G., Ozawa, M., Hasegawa, Y.: Violation of heisenberg’s error-disturbance uncertainty relation in neutron-spin measurements. Phys. Rev. A
**88**(2), 022110 (2013)CrossRefGoogle Scholar - 15.Werner, R.F.: The uncertainty relation for joint measurement of postion and momentum. Quantum Inform. Comput. 4, 546 (2004). arXiv:Quant-ph/0405184
- 16.Lund, A.P., Wiseman, H.M.: Measuring measurement-disturbance relationships with weak values. New J. Phys.
**12**, 093011 (2010)CrossRefGoogle Scholar - 17.Scully, M.O., Englert, B.G., Walther, H.: Quantum optical tests of complementarity. Nature
**351**, 111 (1991)Google Scholar - 18.Storey, P., Tan, S., Collett, M., Walls, D.: Path detection and the uncertainty principle. Nature
**367**, 626 (1994)Google Scholar - 19.Mir, R., Lundeen, J., Mitchell, M., Steinberg, A., Wiseman, H., Garretson, J.: A double-slit which-way experiment addressing the complementarity-uncertainty debate. New J. Phys.
**9**, 287 (2007)Google Scholar - 20.Di Lorenzo, A.: Correlations between detectors allow violation of the heisenberg noise-disturbance principle for position and momentum measurements. Phys. Rev. Lett.
**110**, 120403 (2013)Google Scholar - 21.Branciard, C.: Error-tradeoff and error-disturbance relations for incompatible quantum measurements. Proc. Natl. Acad. Sci.
**110**(17), 6742 (2013)Google Scholar - 22.Branciard, C.: Deriving tight error-trade-off relations for approximate joint measurements of incompatible quantum observables. Phys. Rev. A
**89**, 022124 (2014)Google Scholar - 23.Dressel, J., Nori, F.: Certainty in heisenberg’s uncertainty principle: Revisiting definitions for estimation errors and disturbance. Phys. Rev. A
**89**, 022106 (2014)Google Scholar - 24.Korzekwa, K., Jennings, D., Rudolph, T.: Operational constraints on state-dependent formulations of quantum error-disturbance trade-off relations. Phys. Rev. A
**89**, 052108 (2014)Google Scholar - 25.Buscemi, F., Hall, M.J.W., Ozawa, M., Wilde, M.M.: Noise and disturbance in quantum measurements: An information-theoretic approach. Phys. Rev. Lett.
**112**, 050401 (2014)Google Scholar - 26.Busch, P., Lahti, P., Werner, R.F.: Heisenberg uncertainty for qubit measurements. Phys. Rev. A
**89**, 012129 (2014)Google Scholar