← 返回 Avalaches

哥本哈根诠释中的“cut”长期将经典与量子图景分开:经典变量在测量前后是确定的,而量子波函数在测量前只给出概率描述。然而当代实验已能在尺度只有几纳米的中观体系到普通光学显微镜可见物体中观察到量子效应,因此量子行为向经典行为的边界已不能再被视为纯粹约定。文章回顾了既有解释:哥本哈根将塌缩视作经典表述的一部分,主张自发塌缩机制,de Broglie–Bohm 的导波模型,以及 Hugh Everett 的多世界解释,它们都只能局部说明测量谜题,未能完全解释为何会出现稳定的单一经典现实。

Zurek 与 H. Dieter Zeh 主张,这一过渡可由标准量子形式主义中的普遍纠缠推出。任意量子体系与环境的相互作用,包括光子与苹果表面分子的散射,都会产生环境纠缠并形成信息外流通道。退相干(decoherence)将叠加态相位扩散到大量环境自由度,使体系本身的量子效应在可观测上迅速消失。对一粒尘埃来说,典型退相干时间约为 10^-31 秒,约为光穿越一个质子所需时间的百万分之一(10^-6)。并非所有量子态都同样脆弱;“指针态”(如位置或电荷)可在环境中稳定留下多份信息拷贝而不被模糊,Zurek 将其称为“适者”,并将其在放大和筛选中的机制 formalize 为 quantum Darwinism。

Zurek 与 Jess Riedel(2010)估计,在一微秒内,阳光光子可对单个尘埃粒子留下约 1,000万 份位置印记,后续实验也显示只需很少数环境片段就能恢复体系的大部分信息,之后信息量便趋于饱和。该理论进一步断言这些印记最终会收敛到同一结果,因此不同观察者共享单一经典记录而非真正分歧的结果分支。于是,一个被环境冗余拷贝包围的观测对象获得“相对客观存在”(extanton)。Zurek 进一步把波函数性质称为“epiontic”(兼具 epistemic 与 ontic):退相干前所有分支仍是数学上的可能性,但只有被选择的状态进入可观测现实。仍未解决的问题包括:为何某一具体结果被选择、何时变为不可逆,以及如何更严格检验;Sally Shrapnel 与 Renato Renner 等批评者仍认为退相干前的基底及观察者一致性问题尚未闭合。

The article revisits the quantum-to-classical transition. Classical and quantum pictures were historically separated by the Copenhagen “cut”: classical variables are definite while the quantum wave function is probabilistic until measurement. Yet current experiments probe from mesoscales of only a few nanometres to microscope-visible objects and still observe quantum effects, so the boundary between domains cannot remain a mere convention. Competing interpretations—Copenhagen collapse-as-classical bookkeeping, spontaneous-collapse proposals, de Broglie–Bohm pilot waves, and Hugh Everett many worlds—each explains part of the puzzle but not the emergence of one stable classical reality.

Zurek and H. Dieter Zeh argue that this transition follows from standard quantum formalism via ubiquitous entanglement. Any interaction between a quantum system and its surroundings, including photons with apple-surface molecules, produces environment entanglement and an information channel outward. Decoherence then spreads superposition phases into many environmental degrees of freedom, making quantum effects in the system itself effectively unobservable. For a dust grain, typical decoherence occurs in about 10^-31 s, about one-millionth of the time for light to cross a proton. Not all states are equally fragile: “pointer states” (such as position or charge) can imprint environmental copies robustly without being smeared away. Zurek terms these states the fittest and formalizes their amplification and selection as quantum Darwinism.

Zurek and Jess Riedel (2010) estimated that sunlight photons can leave about 10 million positional imprints of one dust grain within one microsecond, and experiments later show that only a few environment fragments can already recover most object information, after which information content saturates. The theory further claims all such imprints converge to the same outcome, so observers agree on a single classical record and avoid true branching. Thus an observed object, surrounded by redundant environment copies, gains relatively objective existence (extanton). He later calls the state description “epiontic,” partly epistemic and partly ontic: all branches remain as mathematical possibilities before decoherence, but only selected ones become observed reality. Remaining issues remain—how a specific outcome is selected, when this becomes irreversible, and how decisively to test it; critics like Sally Shrapnel and Renato Renner call the pre-decoherence substrate and inter-observer consistency still open.

2026-03-01 (Sunday) · fee1bb92bc62c104635fbd96c7425d7a8a32c760