The Non-Triviality Principle and Objections to Its Application

In Chapter 6, six objections to the application of the Non-Triviality Principle in the triviality argument are examined. According to the first objection, the Non-Triviality Principle does not apply to the kind of facts referenced in the triviality argument. According to the second and third objections, the triviality argument depends on what are claimed to be false assumptions about causation—respectively, that causation comes in degrees and that probabilistic causation implies that causation is scalar. The fourth objection is that the relation that matters varies in strength with the strength of the causal connection, but the triviality argument wrongly assumes otherwise. The fifth objection is that the triviality argument works only if reasons externalism is true, but reasons internalism is true. The sixth objection is that the triviality argument fails if particularism or brutalism applies to what matters in survival. None of these objections, it is argued, hit their targets.

What Probabilistic Causation Should Be

Irrespective of degrees of belief, there are cases where the world itself seems to work in a probabilistic way. But there are different accounts of what this means. One view is that the facts of probability are fixed ultimately by the relative frequency of an occurrence. A propensity theory, however, says that real-world probabilistic facts are what generate any such frequencies. The latter is associated with epistemic humility: because there is no guarantee that the generated frequencies match the real facts of probability, we cannot know for sure what those probabilities are even if we know all the facts of frequency. Furthermore, the strength of a propensity cannot be probabilistically defined since the former might be doubled indefinitely while probabilities are measured on a bounded scale.

Time in probabilistic causation: direct vs. indirect uses of lexical causative verbs

It is traditionally assumed that lexical causative verbs (e.g. kill) express direct causationonly, while periphrastic (bi-clausal) causatives (e.g. cause to die) may also express indirectcausation. In favour of this constraint, Fodor famously observed that the (change of) state introducedby lexical causative verbs is not accessible for separate adverbial modification by temporal(or manner) adverbials. In this paper, I present old and new arguments against the direct causationconstraint under the definitions of directness of Fodor and Wolff. I then propose a new definitionof directness in terms of ab-initio causal sufficiency framed in Kvart’s probabilistic account ofsingular causation. I argue that directness so redefined is an implicature rather than an entailmentof lexical causative verbs, which enables me to account for old and new data. Furthermore, I accountfor why the constraint on separate modification by temporal adverbials can be relaxed witheventuality-denoting subjects.Keywords: lexical causative verbs, direct vs. indirect causation, causal sufficiency, probabilistictheories of causation, semantics/pragmatics interface.

Probabilistic Causation

This chapter will explore a variety of projects that aim to characterize causal concepts using probability. These are, somewhat arbitrarily, divided into four categories. First, a tradition within philosophy that has aimed to define, or at least constrain, causation in terms of conditional probability is discussed. Secondly, the use of causal Bayes nets to represent causal relations, to facilitate inferences from probabilities to causal relations, and to ‘identify’ causal quantities in probabilistic terms is discussed. Thirdly, efforts to measure causal strength in probabilistic terms are reviewed, with particular attention to the significance of these measures in the context of epidemiology. Finally, attempts are discussed to analyze the relation of ‘actual causation’ (sometimes called ‘singular causation’) using probability.