In this paper, criticisms are made of the main tenets of Professor Mellor's argument against ‘backwards’ causation. He requires a closed causal chain of events if there is to be ‘backwards’ causation, but this condition is a metaphysical assumption which he cannot totally substantiate. Other objections to Mellor's argument concern his probabilistic analysis of causation, and the use to which he puts this analysis. In particular, his use of conditional probability inequality to establish the ‘direction’ of causation is shown to (...) be in error. 1I am indebted to Drs H. Krips, L. O'Neill and to the anonymous referee for their suggestions and critical comments on earlier drafts. (shrink)

The aim of this paper will be to try and show why there could not be any acceptable counter-stipulation-example to this analytic truth. The overall strategy will be to articulate the conceptual centrality of this analytic truth by showing that a change in it will cause absurdities to break out in certain neighboring concepts with which the concept of cause has logical liaisons, such as action, possibility, intention, deliberation, memory, responsibility, and punishment: and when we try to protect ourselves from (...) these absurdities by altering these latter concepts as well, we will have twisted our concept of causality, along with these other concepts, out of recognition. In other words, it will be argued that to give up this analytic truth will require such a drastic overhauling of the basic concepts of our conceptual system that they will not be a recognizable analog of their former self. Obviously, I cannot give any precise criterion for just when an altered concept ceases to be a permissible analog of its former self. While there is room for disagreement about these matters, we can reach agreement upon the logical connections between the concept of causality and these other neighboring concepts, so that we can determine what changes must be made in the latter so as to accommodate a change in the former. Whether we then wish to say that these changes are so extensive that these concepts are not permissible analogs of their former selves is of secondary importance. However, I do believe that it can be shown that the extended concept of causality which permits a cause to be later than its effect requires such a drastic change in neighboring concepts that there will be little room for disagreement as to whether it is a legitimate analog to the original concept of causality. (shrink)