Reply to “Counterfactual communication not achieved yet – A Comment on Salih et al. (2022)”

Hatim Salih, Jonte R. Hance, Will McCutcheon, John Rarity · 2025

Read Reply to “Counterfactual communication not achieved yet – A Comment on Salih et al. (2022)” by Hatim Salih, Jonte R. Hance, Will McCutcheon, John Rarity (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-2507-19563.txt.

Introduction

Hatim Salih and three co-authors reply to a critic of their claim that messages can pass without particles crossing, in a 2025 paper. The exchange concerns counterfactual communication, a quantum scheme where two parties swap a message while, by set tests, no particle takes the path. A comment by Popescu had argued their 2022 proof fails. The reply says the comment missed the exact claim and trimmed the scheme past fairness.

Their claim covers post-selected particles only, judged by two named tests for a particle's path. On those terms, they hold that earlier theory and lab work already back the scheme. The protocol can, rarely, send long binary messages with high accuracy. The authors reject the charge that the dispute is one of interpretation. They frame it as a settled technical point.

Catalog record

CREATORHatim Salih, Jonte R. Hance, Will McCutcheon, John Rarity
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE14 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-19563

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Reply to “Counterfactual communication not achieved yet – A Comment on Salih et al. (2022)” Hatim Salih, Jonte R. Hance, Will McCutcheon, John Rarity Abstract In his Comment on our recent paper “The laws of physics do not prohibit counterfactual communication”, npj Quantum Information (2022) 8:60, Popescu argues that the claims of the paper are invalid. Here, we refute his argument, showing that it is based on ignoring the specifics of what we set out to prove (that counterfactual communication is possible for post-selected particles , and more specifically in these cases is not prohibited by the weak trace or consistent histories criteria for particle path)), followed by an unwarranted simplification of the protocol. Moreover, the Comment’s excursion into interpretation is misplaced. Our communication protocol is a precisely defined one that allows two remote parties, albeit rarely, to communicate an arbitrarily long binary message, with arbitrarily high accuracy. This is not a matter of interpretation—as the concrete example given in our paper in question illustrates. As for our overarching claim that no particles are exchanged in the course of this communication, we have already demonstrated this both theoretically and experimentally, in the postselected case we consider, as per the weak trace and consistent histories criteria for path of a quantum particle. The simplest issue with Popescu’s comment Popescu2025CommentLaws is that it mistakenly infers that we are saying our protocol salih2022laws is counterfactual even for failed runs (those where no information is transferred from Bob to Alice, and the photon ends up in detector D_{3} ). This is patently wrong—in our abstract, and throughout the body of our paper, we repeatedly state we are talking about the postselected case (when Alice’s photon arrives either in her detector D_{0} or her detector D_{1}) . At the beginning of our Results Section, we even explicitly say that one of the two aims of the protocol we construct in our paper construct one where “It can be shown unambiguously that Alice’s post-selected photons have never been to Bob.” Popescu’s Comment instead treats it as though we are referring to counterfactuality even outside this postselected case. Further, the toy protocol in our paper is one modified from Salih et al’s 2013 protocol Salih2013Protocol specifically designed to show that the weak trace Vaidman2013Past and consistent histories Griffiths2002ConsistentTheory criteria for particle path do not prohibit counterfactual communication, given these are the criteria on which criticisms of Salih et al’s original protocol had been based at the time of publication (see e.g., Vaidman2015 ; Vaidman2016CommentState ; Vaidman2014a ; Wander2021Analyzing ; Griffiths2016 . While our toy protocol is formed of one outer interferometer cycle of Salih et al’s 2013 protocol, there is a reason this protocol was originally give for two or more outer cycles—this one-outer-cycle toy protocol can give erroneous “0-bit” clicks, while in two-or-more outer-cycle versions of Salih et al’s protocol, this is not the case—whenever Alice receives a bit, it is exactly the same bit-value Bob intended to send. While we reply here to Popescu’s criticisms of the toy protocol, we wish to make clear the toy protocol is designed specifically so we could show that counterfactual communication is not prohibited by the weak trace and consistent histories approaches, given the complexity of analysis with these tools, and so lacks features (e.g., interference between inner and outer path of outer interferometer) which Salih et al’s 2013 protocol possesses, which more easily rebut Popescu’s criticisms. (Note also that, while we show in salih2022laws that counterfactual communication is possible by the weak trace and consistent histories criteria of particle path, this does not mean we advocate these as criteria for the path of a quantum particle—indeed, Salih2023Wormholes challenges the consistent histories criterion, and Hance2023Weak challenges the weak trace criterion (with further discussion in Vaidman2023CommentWeak ; Hance2023ReplyWeak ). One could quite easily come up with alternative definitions for particle presence, which would lead to different conclusions about what protocols are, and are not, “counterfactual”. Similarly, even if you accept a given criteria for a quantum particle being in a certain location, a protocol being counterfactual by said criteria does not rule out there being indirect influences between Alice and Bob—indeed, one could argue that there must be some indirect influence/interference effect, otherwise it would be impossible to communicate using such a counterfactual protocol. Hance2024CFBAIG , again by one of the authors, considers whether indirect influences in scenarios like this may be caused by back-action effects onto the system by the meter during measurement. We next note that the caption of Fig. 1 of our paper salih2022laws indicates clearly that detector D_{0} could have been placed immediately after the topmost beamsplitter—meaning that interference only takes place in the concatenated interferometers on the right. The author of the Comment takes several paragraphs to acknowledge this. The caption of Fig. 1 explains the reason for placing D_{0} after the bottom-most beamsplitter: so that the setup exactly includes the equivalent of one outer cycle of Salih et al ’s Michelson-type (polarisation-based) protocol, laid-out sequentially in time Salih2013Protocol . This allows the conclusions drawn for this single-outer-cycle protocol to be applicable to any of the outer cycles of Salih et al’s 2013 counterfactual communication protocol, as does reference Salih2023Wormholes . Perhaps not fully aware of how that 2013 (polarisation-based) counterfactual communication protocol works, the Comment disputes this applicability. The important point to note, however, is that when the inner interferometers are not blocked—i.e., bit ‘0’, which is the case where counterfactuality had been questioned—no interference should take place in the outer interferometers, just as in the present single-outer-cycle protocol. The Comment goes on to suggest a threefold simplification of our Protocol. First, the Comment suggests that the topmost polarising beam-splitter PBS and the polarisation rotator HWP1 can be removed, and replaced with a biased coin to decide whether to fire the photon into the concatenated interferometers on the right or not. This is correct—–and something that has already been shown in an “After the Paper” (on Springer Nature’s Physics Community) less-technical explanatory article on our paper in question BehindThePaper . The next simplification the Comment makes is to remove the coin toss altogether. It says: “Then even the coin toss is not necessary. If Alice and Bob want to perform the original protocol for N runs with probability p of live runs, they can simply use the basic apparatus for pN runs, then turn it “off” and go home both, with Alice guessing “no blocking” for the rest of the (1-p)N runs, while Bob doesn’t even have to attempt transmitting anything.” When referring to N runs, does the Comment mean one run per bit choice by Bob? In this case, having got rid of our use of the coin, the space of possible N -bit whole messages that can be accurately communicated with some non-zero probability is limited to essentially one N -bit message only: the message where all pN\approx N bits have bit-value ‘1’. In all other cases, the post-selection would fail (the photon would end at D_{3} ), so the message would need to be disregarded. Contrast this with our protocol, where any N -bit string from Bob can be entirely and accurately communicated, with some non-zero probability (that goes down as

Read the complete text