Abstract

I outline a heretofore neglected difference between manipulation scenarios and merely deterministic ones. Plausible scientific determinism does not imply that the relevant prior history of the universe is independent of us, while manipulation does. Owing to sensitive dependence of physical outcomes upon initial conditions, in order to trace a deterministic history, a microphysical level of analysis is required. But on this level physical laws are time-symmetrically deterministic, and causality, conceived asymmetrically, disappears. I then consider a revised scenario to resurrect the threat of manipulation even in the presence of time-symmetry and sensitive dependence upon initial conditions. To do so we posit a Designer-manipulator containing all the information of the manipulated and time-symmetrically related to him. The new scenario violates special relativity, but even waiving that objection, the scenario cannot meet its requirements. I argue that the Designer lacks agency enough to manipulate the target, both because her information lacks the robustness required to constitute knowledge of what she does, and because it leaves no room for desires for specific results.

摘要

我概述了操纵场景和纯粹确定性场景之间迄今为止被忽视的区别。看似合理的科学决定论并不意味着宇宙的相关先前历史独立于我们，而操纵则不然。由于物理结果对初始条件的敏感依赖性，为了追踪确定性的历史，需要进行微观物理水平的分析。但在这个层面上，物理定律是时间对称的确定性的，并且不对称地设想的因果关系消失了。然后，我考虑修改后的场景，即使存在时间对称性和对初始条件的敏感依赖，也能重新恢复操纵的威胁。为此，我们假设一个设计操纵器包含被操纵者的所有信息并与他时间对称地相关。新的场景违反了狭义相对论，但即使放弃这一反对意见，该场景也无法满足其要求。我认为设计师缺乏足够的代理来操纵目标，这既是因为她的信息缺乏构成她所做的事情的知识所需的稳健性，也因为它没有为实现特定结果的愿望留下空间。