B5: Statistical Causation, Intervention, and Freedom

项目B5基于因果贝叶斯网理论（Spirtes et al. 2000），研究有关人类行为和自由的形而上学问题。这些问题的答案取决于所认可的因果关系的具体理论（参见第10节）。希区柯克，2012年）。从归纳形而上学的观点来看，使用因果贝叶斯网理论作为因果关系的背景理论似乎很有前途：该理论的核心可以通过对某些经验现象的最佳解释的推理来支持，并且该理论的几个版本是经验可检验的（Schurz & Gebharter 2016）。项目B5特别关注以下四个研究问题（RQ）。在RQ 1中，我们问有多少自由和什么样的自由是可能的因果贝叶斯网络框架。RQ 2是关于我们必须假设有多少自由来解释经验现象，例如通过干预手段，特别是在实验和随机化中发现因果关系的成功案例。在对最佳解释的推理的驱动下--归纳形而上学最重要的方法之一--我们发展了一种普遍的、跨学科的自由理论。在RQ 3中，我们研究了当人们添加关于自由人类干预的不同假设时，因果贝叶斯网理论的不同版本所隐含的经验后果。最后，在RQ 4中，我们从因果贝叶斯网理论中重建了关于人类意志和行动自由的文献中的突出论点，并从因果关系这一特定理论的角度评估了它们的有效性和前提的可验证性。在这样做的时候，我们特别关注支持和反对自由和非决定论的兼容性的论点。

Project B5 investigates metaphysical questions concerning human action and freedom on the basis of the theory of causal Bayes nets (Spirtes et al. 2000). The answer to such questions depends on the specific theory of causation endorsed (cf. Hitchcock 2012). Using the theory of causal Bayes nets as one's background theory of causation seems promising from the viewpoint of Inductive Metaphysics: The theory's core can be backed up by an inference to the best explanation of certain empirical phenomena and several versions of the theory are empirically testable (Schurz & Gebharter 2016). Project B5 is specifically interested in the following four research questions (RQs). In RQ1 we ask how much freedom and what kinds of freedom are possible in light of a causal Bayes net framework. RQ2 is about how much freedom we have to assume to explain empirical phenomena such as successful cases of causal discovery by means of interventions, especially in experimentation and randomisation. Driven by an inference to the best explanation--one of the most important methods of Inductive Metaphysics--we develop a general and interdisciplinary theory of freedom. In RQ3 we investigate which empirical consequences are implied by different versions of the theory of causal Bayes nets when one adds different assumptions about free human interventions. Finally, in RQ4 we reconstruct prominent arguments from the literature on the freedom of human will and action within the theory of causal Bayes nets and evaluate their validity and the plausibility of their premises from the viewpoint of this particular theory of causation. In doing so, we focus especially on arguments for and against the compatibility of freedom and indeterminism.