1: \documentclass[groupedaddress, beta
2: 3),10pt,twocolumn,bibnotes,prl,showpacs]{revtex4}%
3: \usepackage{amssymb}%
4: \usepackage{amsmath}%
5: \setcounter{MaxMatrixCols}{30}%
6: \usepackage{amsfonts}%
7: \usepackage{graphicx}
8: %TCIDATA{OutputFilter=latex2.dll}
9: %TCIDATA{Version=4.00.0.2312}
10: %TCIDATA{CSTFile=revtex4.cst}
11: %TCIDATA{Created=Sunday, June 30, 2002 10:32:08}
12: %TCIDATA{LastRevised=Wednesday, August 20, 2003 21:18:28}
13: %TCIDATA{<META NAME="GraphicsSave" CONTENT="32">}
14: %TCIDATA{<META NAME="DocumentShell" CONTENT="Articles\SW\REVTeX 4 (Test Version)">}
15: %TCIDATA{Language=American English}
16: \newtheorem{theorem}{Theorem}
17: \newtheorem{acknowledgement}[theorem]{Acknowledgement}
18: \newtheorem{algorithm}[theorem]{Algorithm}
19: \newtheorem{axiom}[theorem]{Axiom}
20: \newtheorem{claim}[theorem]{Claim}
21: \newtheorem{conclusion}[theorem]{Conclusion}
22: \newtheorem{condition}[theorem]{Condition}
23: \newtheorem{conjecture}[theorem]{Conjecture}
24: \newtheorem{corollary}[theorem]{Corollary}
25: \newtheorem{criterion}[theorem]{Criterion}
26: \newtheorem{definition}[theorem]{Definition}
27: \newtheorem{example}[theorem]{Example}
28: \newtheorem{exercise}[theorem]{Exercise}
29: \newtheorem{lemma}[theorem]{Lemma}
30: \newtheorem{notation}[theorem]{Notation}
31: \newtheorem{problem}[theorem]{Problem}
32: \newtheorem{proposition}[theorem]{Proposition}
33: \newtheorem{remark}[theorem]{Remark}
34: \newtheorem{solution}[theorem]{Solution}
35: \newtheorem{summary}[theorem]{Summary}
36: \newenvironment{proof}[1][Proof]{\textbf{#1.} }{\ \rule{0.5em}{0.5em}}
37: \begin{document}
38: \title{Unifying Entanglement and Nonlocality as a Single Concept: Quantum Wholeness}
39: \author{Zeng-Bing Chen}
40: \author{Sixia Yu}
41: \affiliation{Department of Modern Physics, University of Science and Technology of China,
42: Hefei, Anhui 230026, People's Republic of China}
43: \author{Yong-De Zhang}
44: \affiliation{Department of Modern Physics, University of Science and Technology of China,
45: Hefei, Anhui 230026, People's Republic of China}
46: \author{Nai-Le Liu}
47: \affiliation{Department of Modern Physics, University of Science and Technology of China,
48: Hefei, Anhui 230026, People's Republic of China}
49: \thanks{To whom correspondence should be addressed. E-mail: zbchen@ustc.edu.cn}
50:
51: \begin{abstract}
52: Although entanglement is widely recognized as one of the most fascinating
53: characteristics of quantum mechanics, nonlocality remains to be a big
54: labyrinth. The proof of existence of nonlocality is as yet not much convincing
55: because of its strong reliance on Bell's theorem where the assumption of
56: realism weakens the proof. We demonstrate that entanglement and quantum
57: nonlocality are two equivalent aspects of the same quantum wholeness for
58: spacelike separated quantum systems. This result implies that quantum
59: mechanics is indeed a nonlocal theory and lays foundation of understanding
60: quantum nonlocality beyond Bell's theorem.
61:
62: \end{abstract}
63: \pacs{03.65.Ud, 03.67.-a, 03.65.Ta}
64: \pacs{03.65.Ud, 03.67.-a, 03.65.Ta}
65: \maketitle
66:
67:
68: Since the classic works of EPR (Einstein, Podolsky and Rosen \cite{EPR}) and
69: Bell \cite{Bell,Bell-book}, quantum nonlocality and entanglement have been
70: recognized as two crucial notions in modern understanding of quantum
71: phenomena. As entanglement is an essential \textquotedblleft
72: resource\textquotedblright\ in practical applications of quantum information
73: \cite{QIT}, various aspects \cite{QIT,Vedral,Horodecki-e,PhysToday}\ of it
74: have been extensively studied in recent years. However, our current
75: understanding \cite{Barrett}\ of nonlocality is heavily based on Bell's
76: theorem \cite{CS-exp,Peres-book,Werner-rev,Laloe}. Since Werner's seminal work
77: \cite{Werner}, many elegant ideas
78: \cite{Popescu,Gisin-hidden,Zuk-Horo,Peres-coll,Werner-rev,Horodecki-e,Barrett-POVM}%
79: \ have been proposed to understand nonlocality of mixed states. Yet, the
80: relationship between nonlocality and entanglement for mixed states is still
81: unclear \cite{Horodecki-e,Werner-rev,Barrett-POVM}\ and remains one of the
82: most challenging problems in the field. Here we first give a generic locality
83: condition (i.e., a condition that should be satisfied by any local theory and
84: can thus be regarded as a proper definition of locality) for any multipartite
85: system (with its subsystems being spacelike separated). We then prove that the
86: locality condition, also underlying Bell's inequalities, is satisfied if and
87: only if the states of any spacelike separated quantum system are
88: non-entangled. As such, any entangled state is quantum mechanically nonlocal,
89: i.e., entanglement and quantum nonlocality are two side of the same attribute
90: of entangled systems. It is anticipated that the result may be a starting
91: point for penetrating the mystery of nonlocality of nature in general and
92: quantum nonlocality in particular.
93:
94: The concept of entanglement is well justified in the sense that one, at least,
95: has a definition of entangled states \cite{Werner}. Quantification
96: \cite{Vedral} and manipulation \cite{PhysToday} of entanglement are hot topics
97: of fundamental interest in the fields of quantum mechanics and quantum
98: information theory; various separability criteria \cite{Horodecki-e} have been
99: found to classify quantum states into entangled and separable ones. By
100: contrast, quantum nonlocality still remains, to a large extend, an intuitive
101: notion. Roughly spoken, it usually means the impossibility of simulating
102: certain quantum predictions by local realistic theories \cite{Barrett}. As is
103: well known, local realism represents a world view \cite{CS-exp} which states
104: that a physical system has local objective properties, independent of any
105: observations on other spacelike separated systems. By assuming locality and
106: realism, the celebrated Bell inequalities (as well as their various
107: generalizations \cite{CS-exp,Werner-rev,Gisin-Peres,Collins,Chen}) can be
108: derived and impose an upper bound on correlations of the results of the
109: \textquotedblleft Bell experiments\textquotedblright. Astonishingly, the upper
110: bound is violated quantum mechanically by a large class of entangled states
111: \cite{CS-exp,Werner-rev,Gisin-Peres,Collins,Chen}, as confirmed by many
112: experiments \cite{CS-exp,Aspect,Pan-GHZ}. To Bell \cite{Bell-book}, the
113: quantum violations of Bell's inequalities imply that quantum mechanics is a
114: nonlocal theory. This attitude is largely accepted in literature and serves as
115: the basis of our current understanding of nonlocality.
116:
117: However, the attitude is not unquestionable. Anyway, Bell's inequalities are
118: derived from two underlying assumptions: locality and realism. Thus, the
119: experimentally confirmed conflict between quantum mechanics and local realism
120: could be equally interpreted as implying that quantum mechanics is a local
121: non-realistic theory. On the one hand, it is natural for some authors (see,
122: e.g., Ref. \cite{Braunstein}) to refute realism, rather than locality, in
123: quantum mechanics. On the other hand, there are also attempts \cite{Stapp} to
124: prove the nonlocality of quantum mechanics without explicitly assuming
125: realism. Yet, the proof is based on some counterfactual reasonings that are
126: controversial (see Ref. \cite{Vaidman} and references therein). Currently,
127: whether quantum mechanics is indeed a nonlocal theory remains an open
128: question, whose answer can be approached only when we have a solid foundation
129: of nonlocality.
130:
131: Now it is a well established fact that biparticle
132: \cite{Gisin-Peres,Collins,Chen} and multiparticle \cite{Popescu-PLA} entangled
133: pure states always lead to certain violation of Bell-like inequalities. This
134: convincingly indicates that there may exist a close and direct relationship
135: between entanglement and nonlocality. Yet, for mixed states the situation is
136: very puzzling
137: \cite{Werner,Popescu,Gisin-hidden,Zuk-Horo,Peres-coll,Werner-rev,Horodecki-e,Barrett-POVM}%
138: . Werner \cite{Werner} first demonstrated that there are entangled states (the
139: Werner states) that do not violate any Bell inequality. Subsequent works show
140: that the Werner states possess \textquotedblleft hidden
141: nonlocality\textquotedblright\ \cite{Popescu,Gisin-hidden}, which, however,
142: can be uncovered only by invoking generalized measurements
143: \cite{Popescu,Zuk-Horo,Peres-coll,Barrett-POVM} instead of the standard von
144: Neumann measurement. Despite of many elegant efforts
145: \cite{Popescu,Gisin-hidden,Zuk-Horo,Peres-coll,Werner-rev,Horodecki-e,Barrett-POVM}
146: along the line of the \textquotedblleft Bell paradigm\textquotedblright, there
147: are still some open problems \cite{Horodecki-e,Werner-rev,Barrett-POVM} on the
148: relationship between nonlocality and entanglement. The difficulty encountered
149: in our current understanding of nonlocality might suggest that Bell's theorem
150: is not sufficient for uncovering nonlocality of quantum mechanics. Then a
151: fundamental problem arises here: In what sense we classify quantum states into
152: local and nonlocal ones?
153:
154: In this work we provide answers to the above-mentioned open questions. This is
155: achieved by understanding quantum nonlocality at a deeper level going beyong
156: Bell's theorem. As one might intuitively envision, nonlocality inherent in
157: entanglement is a purely quantum phenomenon. Meanwhile, realism underlying
158: Bell's theorem is a world view that logically has nothing to do with quantum
159: mechanics. Thus, in order to reveal nonlocality as a fundamental feature of
160: nature, it is necessary to discard the premise of realism; what concerns us
161: first is the condition that any local theory must satisfy.
162:
163: Local causality (or simply, locality) means that the experimental results
164: obtained from a physical system at one location should be independent of any
165: observations or actions made at any other spacelike separated locations.
166: Unfortunately, the locality condition
167: \cite{Bell,Bell-book,Werner-rev,Shimony,assume} was previously given in
168: company with realism. Since here one needs a locality condition without any
169: pre-assumption other than locality, the condition must be expressed by
170: quantities that are experimentally observable for localists.%
171: %TCIMACRO{\FRAME{ftbpFU}{2.2328in}{1.91in}{0pt}{\Qcb{The spacetime diagram of
172: %two spacelike separated particles. The particles A and B are located in two
173: %coloured locations which are spacelike separated. $W$ denotes the overlap of
174: %the backward light cores of the particles A and B; the remaining parts of the
175: %backward light cores of the particles A and B are denoted, respectively, by
176: %$U$ and $V$, which are spacelike separated such that no causal signal can
177: %connect them.}}{\Qlb{figure}}{one-fig.eps}%
178: %{\special{ language "Scientific Word"; type "GRAPHIC";
179: %maintain-aspect-ratio TRUE; display "USEDEF"; valid_file "F";
180: %width 2.2328in; height 1.91in; depth 0pt; original-width 2.5638in;
181: %original-height 2.1873in; cropleft "0"; croptop "1"; cropright "1";
182: %cropbottom "0"; filename 'one-fig.eps';file-properties "XNPEU";}}}%
183: %BeginExpansion
184: \begin{figure}
185: [ptb]
186: \begin{center}
187: \includegraphics[
188: height=1.91in,
189: width=2.2328in
190: ]%
191: {one-fig.eps}%
192: \caption{The spacetime diagram of two spacelike separated particles. The
193: particles A and B are located in two coloured locations which are spacelike
194: separated. $W$ denotes the overlap of the backward light cores of the
195: particles A and B; the remaining parts of the backward light cores of the
196: particles A and B are denoted, respectively, by $U$ and $V$, which are
197: spacelike separated such that no causal signal can connect them.}%
198: \label{figure}%
199: \end{center}
200: \end{figure}
201: %EndExpansion
202:
203:
204: In order to obtain the desired locality condition in mathematical terms, let
205: us consider two particles A and B in the spacetime diagram shown in Fig. 1.
206: They are located, respectively, in two spacetime regions denoted by 1 and 2,
207: which are spacelike separated. According to local causality, events occurring
208: in the backward light core of a particle (e.g., particle A or B) may affect
209: the events (e.g., detecting measurement results) occurring on the particle.
210: Thus, events in the overlap $W$\ of the backward light cores of the two
211: particles may be \textquotedblleft common causes\textquotedblright%
212: \ \cite{Bell-book} of the events in regions 1 and 2, though events in region 1
213: should not be causes of events in region 2 (and vice versa) as required again
214: by local causality.
215:
216: Without loss of generality, one can assume that the observables of A and B
217: have only a finite number of outcomes. Then consider local measurements on the
218: two particles. By the standard rule of probability, the requirement of
219: locality can be mathematically expressed as
220: \begin{equation}
221: P(a_{i},b_{j}\left\vert C)\right. =P(a_{i}\left\vert C)\right.
222: P(b_{j}\left\vert C)\right. \label{local}%
223: \end{equation}
224: for all observed results in the spacelike separated regions 1 and 2 and for
225: any common causes $C$ in $W$. Here $P(a_{i},b_{j}\left\vert C)\right.
226: $\ ($i,j=1$, $2$, $3$,\ldots) is a joint probability of measuring any
227: observables in region 1 (with outcome $a_{i}$) and in region 2 (with outcome
228: $b_{j}$) conditioned on a given element in the set $C$ of all common causes in
229: $W$; $P(a_{i}\left\vert C)\right. $\ [$P(b_{j}\left\vert C)\right. $] is the
230: local probability of getting the outcome $a_{i}$ ($b_{j}$)\ when measuring the
231: observable in region 1 (region 2) conditioned on the given common cause in
232: $C$. Thus, equation (\ref{local}) has a physically apparent interpretation in
233: accord with locality: The given common cause can affect the probabilities with
234: regard to particles A and B; conditioned on the same cause, measurements in
235: region 1 and region 2 must be mutually independent effects. Importantly, the
236: common causes in our analysis represent physically a set of certain individual
237: events whose occurrences can be assigned corresponding probabilities.
238:
239: Denoting the joint probability of getting outcomes $a_{i}$ and $b_{j}$ as
240: $P(a_{i},b_{j})$ and the probability of common causes in $W$ as $P(C)$ (see
241: Fig. $1$), then the rule of conditional probability gives
242: \begin{equation}
243: P(a_{i},b_{j})=\sum_{C}P(a_{i},b_{j}\left\vert C)\right. P(C)\label{pab}%
244: \end{equation}
245: where the summation may also mean integration, if necessary. Let us denote the
246: set of common causes in $W$ by $C=\{\mu\left\vert \mu=1,2,3,\cdots\right.
247: \}$, with $\lambda_{\mu}\geq0$ being the probability for the cause $\mu$ to
248: occur and $\sum_{\mu}\lambda_{\mu}=1$. Thus, for a given cause $\mu$, equation
249: (\ref{local}) becomes $P(a_{i},b_{j}\left\vert \mu)\right. =P(a_{i}\left\vert
250: \mu)\right. P(b_{j}\left\vert \mu)\right. $.
251:
252: Obviously, equation (\ref{pab}) can be assigned an operational meaning.
253: Namely, one can either measure directly $P(a_{i},b_{j})$, or monitor the
254: common causes first. Conditioned on a specific common cause ($\mu$, say) being
255: detected, $P(a_{i},b_{j}\left\vert C)\right. $\ is then measured. Theory of
256: probability insures that the two procedures are equivalent. Of course, for a
257: localist, equation (\ref{local}) should be right.
258:
259: Some further comments in support of the locality condition (\ref{local}) are
260: noteworthy. First of all, for localists the probabilities in the locality
261: condition are all measurable quantities and can be measured, in principle,
262: with an arbitrary accuracy. Given a statistical ensemble, experimenters can
263: always observe the relative frequencies among all the outcomes; After the
264: number of the observations tends to infinity, the relative frequencies will
265: tend to the true probabilities of the corresponding outcomes. Moreover and
266: more importantly, the locality condition given in a probabilistic terms is
267: presented without resorting to any specific theory (realistic or quantum) as
268: it only involves observable quantities; theory enters the picture when one
269: predicts the probability of each outcome. This feature of the locality
270: condition enables one to test locality versus any specific theory, i.e., to
271: test whether nature is local \cite{Chen-test}. For instance, when one uses
272: hidden-variable theories to assign the probabilities in equation
273: (\ref{local}), the locality condition (\ref{local}) reduces exactly to Bell's
274: locality condition \cite{Bell,Bell-book,Werner-rev,Shimony,assume}. Then
275: following Bell's reasoning yields the usual Bell inequalities which must be
276: satisfied by any local realistic theory \cite{Bell,Bell-book}. Bell's locality
277: condition \cite{Bell-book} has been well justified in various aspects in the
278: context of Bell's inequalities \cite{Werner-rev,Shimony,assume} and is now
279: widely accepted.
280:
281: The generic locality condition (\ref{local}) allows one to use quantum
282: mechanics, instead of hidden-variable theories, when predicting the local
283: probabilities [e.g., $P(a_{i}\left\vert \mu)\right. $] appearing in the
284: locality assumption. As such, the two particles A and B must then be described
285: locally by standard quantum mechanical rules. Whether the events in $W$\ are
286: described quantum mechanically or not is unimportant in the subsequent
287: consideration. When described quantum mechanically, the events in $W$\ may be
288: represented by quantum states of an \textquotedblleft
289: ancilla\textquotedblright\ that can be any kind of quantum systems possibly
290: coupled to both A and B. And in this case, the ancilla will be
291: \textquotedblleft traced out\textquotedblright\ as one is concerned only with
292: particles A and B.
293:
294: Thus in the present context, quantum mechanics can predict these
295: probabilities, namely, it guarantees that \cite{Peres-book}, for the
296: probabilities $P(a_{i},b_{j})$, $P(a_{i}\left\vert \mu)\right. $\ and
297: $P(b_{j}\left\vert \mu)\right. $, there are corresponding density operators
298: $\rho_{AB}$ for the two-particle system, $\rho_{A\mu}$ for A and $\rho_{B\mu}%
299: $\ for B such that $P(a_{i},b_{j})=Tr(\rho_{AB}\hat{P}_{Ai}\hat{P}_{Bj}%
300: )=\sum_{\mu}\lambda_{\mu}P(a_{i},b_{j}\left\vert \mu)\right. $,\ $P(a_{i}%
301: \left\vert \mu)\right. =Tr(\rho_{A\mu}\hat{P}_{Ai})$\ and $P(b_{j}\left\vert
302: \mu)\right. =Tr(\rho_{B\mu}\hat{P}_{Bj})$. Then equation (\ref{pab})
303: immediately gives
304: \begin{equation}
305: \mathrm{Tr}(\rho_{AB}\hat{P}_{Ai}\hat{P}_{Bj})=\sum_{\mu}\lambda_{\mu
306: }\mathrm{Tr}(\rho_{A\mu}\rho_{B\mu}\hat{P}_{Ai}\hat{P}_{Bj})\label{roab}%
307: \end{equation}
308: Here $\hat{P}_{Ai}=\left\vert a_{i}\right\rangle \left\langle a_{i}\right\vert
309: $ and $\hat{P}_{Bj}=\left\vert b_{j}\right\rangle \left\langle b_{j}%
310: \right\vert $\ are the projection operators corresponding to the outcomes
311: $a_{i}$ and $b_{j}$, respectively. Due to the arbitrariness of $\hat{P}_{Ai}$
312: and $\hat{P}_{Bj}$, one must have (see also \cite{note})
313: \begin{equation}
314: \rho_{AB}=\sum_{\mu}\lambda_{\mu}\rho_{A\mu}\rho_{B\mu}\label{sep}%
315: \end{equation}
316: which represents the state of the two particles allowed by the locality
317: condition (\ref{local}). The state $\rho_{AB}$ is a convex sum, with weights
318: $\lambda_{\mu}$, of direct products of the local density operators $\rho
319: _{A\mu}$ and $\rho_{B\mu}$. Different set of common causes leads to different
320: convex sum of the local density operators.
321:
322: Two remarks are in order here. The state $\rho_{AB}$ in equation (\ref{sep})
323: is called \textquotedblleft classically correlated\textquotedblright%
324: \ \cite{Werner}. Apparently, this classical correlations can only originate
325: from the `common history' of the particles A and B, as can be seen from the
326: spacetime diagram (Fig. $1$). In the particular case where the observed
327: results in $U$ or/and $V$ are independent on $W$, then one always has
328: $\rho_{AB}=\rho_{A}\rho_{B}$. In this case, $\rho_{A}$ and $\rho_{B}$ contain
329: already a complete specification of the states of the corresponding particles;
330: supplementary information from $W$ is redundant, similarly to the classical
331: case considered by Bell \cite{Bell-book}.
332:
333: Inverting the above reasoning, it is easy to see that any state given in
334: equation (\ref{sep}) satisfies locality. Thus, for the two spacelike separated
335: particles locality is a necessary and sufficient condition for the states
336: having the form specified in equation (\ref{sep}). Consequently, the state
337: (\ref{sep}) is local by definition. On the contrary, any state that cannot be
338: written as equation (\ref{sep}) is nonlocal.
339:
340: Importantly, since the above reasoning does not require any specification of
341: the particles, it is general enough that it is valid for any biparticle
342: systems. Furthermore, its generalization to multiparticle systems is
343: straightforward, and the resulting states satisfying locality are still given
344: by a convex sum of direct products of the local density operators, similarly
345: to equation (\ref{sep}).
346:
347: Surprisingly, the form of $\rho_{AB}$ in equation (\ref{sep}) is just the
348: \textquotedblleft mathematical\textquotedblright\ definition of separable
349: (i.e., non-entangled) two-particle states, as is suggested by Werner
350: \cite{Werner} and now widely accepted. Meanwhile, $\rho_{AB}$ is local by
351: definition as it comes from the \textquotedblleft physical\textquotedblright%
352: \ criterion of locality, which clarifies the physical content, and justifies
353: Werner's definition, of separable states from another perspective.
354: Consequently, the two basic notions---entanglement of quantum states and
355: nonlocality of measured results---are equivalent for spacelike separated
356: systems; local measurements performed on entangled (separable) states give
357: nonlocal (local) results. This result is a further support of the viability of
358: the locality condition (\ref{local}). When the particles A and B are not
359: spacelike separated, the locality condition given in equation (\ref{local})
360: can be reasonably called Einstein's separability condition. In this broader
361: sense, locality is identical to Einstein's separability when the particles are
362: spacelike separated, and Einstein's separability is a necessary and sufficient
363: condition for the separability of states for any quantum systems.
364:
365: This fact may open up an exciting perspective for understanding nonlocality
366: (or more generally, Einstein's inseparability) and entanglement as a unified
367: concept in quantum mechanics, i.e., quantum wholeness: Entanglement represents
368: the mathematical inseparability of quantum states, while nonlocality
369: physically manifests itself in the correlations of certain measurement
370: results. Thus, one has to accept the existence of a quantum weirdness in
371: nature: Quantum entanglement induces exotic influences for a composite system
372: even when the constituent parts are spacelike separated. Here one is
373: confronted with a situation where an entangled quantum system must be regarded
374: as a holistic entity; any attempt to describe the entangled system locally
375: must fail for certain quantum predictions.
376:
377: As we proved, locality permits only the classically correlated (i.e.,
378: separable) states (\ref{sep}) in quantum mechanics. Or equivalently, whatever
379: the set of common causes is, the locality assumption (\ref{local}) cannot be
380: fulfilled by any entangled state. However, entanglement is ubiquitous in
381: quantum mechanics as well as in practical quantum information processing as an
382: essential resource. Thus, locality is in conflict with quantum mechanics,
383: namely, quantum mechanics is definitely a nonlocal theory. We believe that we
384: have for the first time proved in simple terms the intrinsic nonlocal feature
385: of quantum mechanics in a clear-cut way, without resorting to the realism
386: assumption \cite{Bell,Bell-book} or other counterfactual reasonings
387: \cite{Stapp,Vaidman}. Actually, the realism assumption used in deriving Bell's
388: inequalities is redundant and even detrimental for the purpose of uncovering
389: quantum nonlocality \cite{Chen-test}.
390:
391: One might wonder whether the apparent nonlocality of quantum mechanics could
392: be used for superluminal signaling. If this were the case, then quantum
393: mechanics would violate the relativistic causality which forbids any
394: superluminal causal action. Fortunately, quantum mechanics in its current form
395: is still in a \textquotedblleft peaceful co-existence\textquotedblright\ with
396: relativity \cite{Shimony} in the sense that nonlocality does not lead to
397: superluminal information transmitting. To see this, recall that the locality
398: assumption (\ref{local}), according to Shimony \cite{Shimony}, consists of two
399: independent factors: (i) the outcome independence and (ii) the parameter
400: independence, which demand that any measurement outcome of particle A should
401: be independent, respectively, on the outcomes and on the experimental settings
402: of the spacelike separated particle B. Explicit calculation shows that quantum
403: mechanics violates the outcome independence, which nevertheless cannot be used
404: for superluminal information transmitting \cite{Peres-book,Shimony}. However,
405: violation of the parameter independence may imply superluminal communication
406: \cite{assume,Shimony}. To prove the parameter independence (see, e.g.,
407: \cite{Shimony,Laloe}) in the two-particle case (the multiparticle
408: generalization is straightforward), it is sufficient to prove \cite{assume}
409: $P(a_{i}\left\vert b)\right. \equiv\sum_{j}P(a_{i},b_{j})=P(a_{i})$ for any
410: chosen setting $b$.\ Actually, this is always true as\ $\sum_{j}P(a_{i}%
411: ,b_{j})=Tr(\rho_{AB}\hat{P}_{Ai}\sum_{j}\hat{P}_{Bj})=Tr(\rho_{A}\hat{P}%
412: _{Ai})=P(a_{i})$. Here we have used the simple fact that $\sum_{j}\hat{P}%
413: _{Bj}=\hat{I}_{B}$ ($\hat{I}_{B}$ is the unit operator for particles B);
414: $\rho_{A}=Tr_{B}(\rho_{AB})$ is the reduced density operator for particle A.
415: Thus, quantum mechanics respects the parameter independence, implying that
416: quantum nonlocality cannot result in superluminal signaling.
417:
418: Based on the equivalence between nonlocality/inseparability of measurements
419: and entanglement of states, quantum nonlocality has acquired the same solid
420: basis as quantum entanglement. The two equivalent notions are two distinct
421: aspects of the same quantum wholeness. This fact, being interesting in its own
422: right on fundamental issues of quantum mechanics, might be important in
423: quantum information science, where manipulating entanglement is a vital task
424: for processing information. This practical impulse has greatly enriched our
425: current knowledge on entanglement. We anticipate that future works on
426: entanglement and nonlocality may be mutually promoted to deepen our
427: understanding of the weird quantum wholeness.
428:
429: We thank Jian-Wei Pan and Lu-Ming Duan for useful discussions. This work was
430: supported by the National Natural Science Foundation of China, the Chinese
431: Academy of Sciences and the National Fundamental Research Program. N.-L.L. is
432: also supported by the Scientific Research Foundation for the Returned Overseas
433: Chinese Scholars by the State Education Ministry of China and the USTC
434: Returned Overseas Scholars Foundation.
435:
436: \begin{thebibliography}{99} %
437: \bibitem {EPR}A. Einstein, B. Podolsky, and N. Rosen, Phys. Rev. \textbf{47},
438: 777 (1935).
439:
440: \bibitem {Bell}J. S. Bell, Physics (Long Island City, N.Y.) \textbf{1}, 195 (1964).
441:
442: \bibitem {Bell-book}J. S. Bell, \textit{Speakable and Unspeakable in Quantum
443: Mechanics} (Cambridge University Press, Cambridge, 1987).
444:
445: \bibitem {QIT}D. Bouwmeester, A. Ekert, and A. Zeilinger, \textit{The Physics
446: of Quantum Information} (Springer-Verlag, Berlin, 2000).
447:
448: \bibitem {Vedral}V. Vedral, M. B. Plenio, M. A. Rippin, and P. L. Knight,
449: Phys. Rev. Lett. \textbf{78}, 2275 (1997).
450:
451: \bibitem {PhysToday}B. M. Terhal, M. M. Wolf, and A. C. Doherty, Phys. Today
452: \textbf{56}, 46 (April 2003).
453:
454: \bibitem {Horodecki-e}M. Horodecki, P. Horodecki, and R. Horodecki, in
455: \textit{Quantum Information: An Introduction to Basic Theoretical Concepts and
456: Experiments}, edited by G. Alber \textit{et al}. (Springer Tracts in Modern
457: Physics, Berlin, 2001).
458:
459: \bibitem {Barrett}J. Barrett, Phys. Rev. A \textbf{64}, 042305 (2001).
460:
461: \bibitem {CS-exp}J. Clauser and A. Shimony, Rep. Prog. Phys. \textbf{41}, 1881 (1978).
462:
463: \bibitem {Peres-book}A. Peres, \textit{Quantum Theory: Concepts and Methods}
464: (Kluwer, Dordrecht, 1993).
465:
466: \bibitem {Werner-rev}R. F. Werner and M. M. Wolf, Quantum Inf. Comput.
467: \textbf{1} (3), 1 (2001).
468:
469: \bibitem {Laloe}F. Lalo\"{e}, Am. J. Phys. \textbf{69}, 655 (2001).
470:
471: \bibitem {Werner}R. F. Werner, Phys. Rev. A \textbf{40}, 4277 (1989).
472:
473: \bibitem {Popescu}S. Popescu, Phys. Rev. Lett. \textbf{74}, 2619 (1995).
474:
475: \bibitem {Gisin-hidden}N. Gisin, Phys. Lett. A \textbf{210}, 151 (1996).
476:
477: \bibitem {Zuk-Horo}M. \.{Z}ukowski, R. Horodecki, M. Horodecki, and P.
478: Horodecki, Phys. Rev. A \textbf{58}, 1694 (1998).
479:
480: \bibitem {Peres-coll}A. Peres, Phys. Rev. A \textbf{54}, 2685 (1996).
481:
482: \bibitem {Barrett-POVM}J. Barrett, Phys. Rev. A \textbf{65}, 042301 (2002).
483:
484: \bibitem {Gisin-Peres}N. Gisin and A. Peres, Phys. Lett. A \textbf{162}, 15 (1992).
485:
486: \bibitem {Collins}D. Collins, N. Gisin, N. Linden, S. Massar, and S. Popescu,
487: Phys. Rev. Lett. \textbf{88}, 040404 (2002).
488:
489: \bibitem {Chen}Z.-B. Chen, J.-W. Pan, G. Hou, and Y.-D. Zhang, Phys. Rev.
490: Lett. \textbf{88}, 040406 (2002).
491:
492: \bibitem {Aspect}A. Aspect, Nature (London) \textbf{398}, 189 (1999).
493:
494: \bibitem {Pan-GHZ}J.-W. Pan, D. Bouwmeester, M. Daniell, H. Weinfurter, and A.
495: Zeilinger, Nature (London) \textbf{403}, 515 (2000).
496:
497: \bibitem {Braunstein}S. L. Braunstein and C. M. Caves, Phys. Rev. Lett.
498: \textbf{61}, 662 (1988).
499:
500: \bibitem {Stapp}H. P. Stapp, Am. J. Phys. \textbf{65}, 300 (1997).
501:
502: \bibitem {Vaidman}L. Vaidman, Found. Phys. \textbf{29}, 615 (1999).
503:
504: \bibitem {Popescu-PLA}S. Popescu and D. Rohrlich, Phys. Lett. A \textbf{166},
505: 293 (1992).
506:
507: \bibitem {Shimony}A. Shimony, in \textit{Quantum Concepts in Space and Time},
508: edited by R. Penrose and C. J. Isham, (Clarendon Press, Oxford, 1986).
509:
510: \bibitem {assume}A. G. Valdenebro, Eur. J. Phys. \textbf{23}, 569 (2002).
511:
512: \bibitem {Chen-test}Z.-B. Chen, S. Yu, and Y.-D. Zhang, quant-ph/0307143.
513:
514: \bibitem {note}One can prove the following fact: If $\left\langle
515: \phi\right\vert \left\langle \psi\right\vert \hat{O}\left\vert \phi
516: \right\rangle \left\vert \psi\right\rangle =0$\ for any product state
517: $\left\vert \phi\right\rangle \left\vert \psi\right\rangle $ for a
518: two-particle system (generalization to multiparticle cases is
519: straightforward), then the two-particle operator $\hat{O}\equiv0$. To this
520: end, let $\left\vert i\right\rangle $ be any state of particle A, and
521: $\left\vert k\right\rangle $ and $\left\vert l\right\rangle $ any two
522: orthonormal states of particle B. Then for the product states $\left\vert
523: \varphi_{1}\right\rangle =\left\vert i\right\rangle \otimes\frac{1}{\sqrt{2}%
524: }\left( \left\vert k\right\rangle +\left\vert l\right\rangle \right) $ and
525: $\left\vert \varphi_{2}\right\rangle =\left\vert i\right\rangle \otimes
526: \frac{1}{\sqrt{2}}\left( \left\vert k\right\rangle +i\left\vert
527: l\right\rangle \right) $, one necessarily has $\left\langle \varphi
528: _{1}\right\vert \hat{O}\left\vert \varphi_{1}\right\rangle =\frac{1}{2}\left(
529: \left\langle ik\right\vert \hat{O}\left\vert ik\right\rangle +\left\langle
530: ik\right\vert \hat{O}\left\vert il\right\rangle \right. \left. +\left\langle
531: il\right\vert \hat{O}\left\vert ik\right\rangle +\left\langle il\right\vert
532: \hat{O}\left\vert il\right\rangle \right) $\ and $\left\langle \varphi
533: _{2}\right\vert \hat{O}\left\vert \varphi_{2}\right\rangle =\frac{1}{2}\left(
534: \left\langle ik\right\vert \hat{O}\left\vert ik\right\rangle +i\left\langle
535: ik\right\vert \hat{O}\left\vert il\right\rangle \right. \left.
536: -i\left\langle il\right\vert \hat{O}\left\vert ik\right\rangle +\left\langle
537: il\right\vert \hat{O}\left\vert il\right\rangle \right) $. Since
538: $\left\langle \varphi_{1}\right\vert \hat{O}\left\vert \varphi_{1}%
539: \right\rangle =\left\langle \varphi_{2}\right\vert \hat{O}\left\vert
540: \varphi_{2}\right\rangle =\left\langle ik\right\vert \hat{O}\left\vert
541: ik\right\rangle =\left\langle il\right\vert \hat{O}\left\vert il\right\rangle
542: =0$, one obtains from the above equations that $\left\langle ik\right\vert
543: \hat{O}\left\vert il\right\rangle =0$. Particularly, choosing $\left\vert
544: i\right\rangle $ to be the eigenvectors of the Hermitian operator
545: $\left\langle k\right\vert \hat{O}\left\vert l\right\rangle +\left\langle
546: k\right\vert \hat{O}\left\vert l\right\rangle ^{\dagger}$ for particle A and
547: using $\left\langle ik\right\vert \hat{O}\left\vert il\right\rangle =0$, it is
548: easy to see that $\left\langle k\right\vert \hat{O}\left\vert l\right\rangle
549: +\left\langle k\right\vert \hat{O}\left\vert l\right\rangle ^{\dagger}=0$.
550: Similarly, one can prove $\left\langle k\right\vert \hat{O}\left\vert
551: l\right\rangle -\left\langle k\right\vert \hat{O}\left\vert l\right\rangle
552: ^{\dagger}=0$. Therefore, $\left\langle k\right\vert \hat{O}\left\vert
553: l\right\rangle =0$, which means that $\hat{O}\equiv0$. In particular, for
554: $\hat{O}=\rho_{AB}-\sum_{\mu}\lambda_{\mu}\rho_{A\mu}\rho_{B\mu}$, $\left\vert
555: \phi\right\rangle =\left\vert a_{i}\right\rangle $ and $\left\vert
556: \psi\right\rangle =\left\vert b_{j}\right\rangle $, equation (\ref{roab})\ and
557: the result proved in this note then give equation (\ref{sep}).
558: \end{thebibliography}
559:
560:
561: \end{document}