寄生的反事實依賴與典型因果關係

Parasitic Counterfactual Dependence and Paradigmatic Causation

Hitchcock（2007）提出一個修正版的反事實因果分析，他認為：自足因果網絡中的反事實依賴是典型的因果關係，非自足因果網絡中的反事實依賴則是寄生的反事實依賴；寄生的反事實依賴是否為因果關係，Hitchcock持不可知的保留態度。本文論證，Hitchcock以「非自足的因果網絡」來界定寄生的反事實依賴並不恰當，也未成功捕捉到寄生與非寄生反事實依賴的直覺差異。首先，有些直覺意義下的寄生反事實依賴，仍可在自足因果網絡中成立，而為典型因果關係。其次，因果網絡之自足與否，是由網絡內上下級變元的預設值／偏差值連動關係來界定，它連結於因果篩選的議題；反事實依賴之寄生與否，則取決於影響其成立的某個網絡外變元為偏差值或預設值，這涉及是否提供完滿的解釋。自足的因果網絡，即提供了判斷寄生反事實依賴是否為因果關係的標準。我們也援引一些法院判決顯示，法律實務將某些寄生反事實依賴視為因果關係，並據此來對違反義務的不作為究責。

Hitchcock (2007) proposes a revised counterfactual analysis of causation. On his view, counterfactual dependence in a self-contained causal network is paradigmatic causation; counterfactual dependence in a causal network that is not self-contained is parasitic dependence. Hitchcock remains agnostic on whether or not parasitic dependence is genuine causation. We argue that it is inadequate to define parasitic dependence in terms of non-self-contained network, nor does the definition successfully capture the intuition of explanatory difference between parasitic and non-parasitic dependence. First, some cases of parasitic dependence in the intuitive sense are counterfactual dependence in a self-contained network, thus belonging to the paradigmatic cases of causation. Second, whether or not a causal network is self-contained is defined by the default/deviant covariance of variables in the network and involves the issue of causal selection. By contrast, whether or not counterfactual dependence is parasitic is contingent upon the fact that some relevant variable outside the network takes a deviant or default value, and related to the issue of satisfactory explanation. Finally, the notion of self-contained network has provided the criteria for judging whether a parasitic dependence relationship is causal. We also refer to some legal decisions to show that certain cases of parasitic dependence are regarded by the court as actual causation in order to attribute responsibility over duty-breaching omissions.