Dunja Dreesens

93 to each other, for example, regarding effectiveness, efficiency or ethical concerns, is not clearly articulated. Nonetheless, guideline developers do recognise that other types of knowledge are often used and somehow integrated in practice, particularly when discussing the evidence and formulating recommendations, often called ‘judgement’ or ‘considered judgement’ (232, 233). This is the traditionally less clearly described or analysed black box part of the process that new initiatives try to shed a light on, such as NICE’s structured tables linking evidence to recommendations and the GRADE Evidence to Decision frameworks (232). Table 1: Alternative types of reasoning to evade the problem of induction TYPE OF REASONING* (WITH EXAMPLES OF KEY SCHOLARS) SHORTHAND DESCRIPTION EXPLANATION Bayesian evasion (Bayes, Hacking) Learning from experience This type of inductive inference agrees with Hume that we cannot predict the future perfectly, but that we can learn from our experiences reasonably well. This allows us to do more and better predictions. This type of reasoning can update current beliefs with information from frequent events (informing prior probabilities and likelihood ratios). However, because we can learn from a single event too, this approach is suited for the individual case scenario (218). Abduction (Peirce) Reasoning to the best explanation Abduction makes inferences by updating beliefs leading to the best explanation (234). Where Bayesian evasion takes prior probabilities as a given (at least as some argue), which may not be the case, abduction does not. It introduces the consideration of theory and mechanism in the act of inferring (235). Mechanistic/deterministic How things appear to work This type of reasoning makes an inference based on a mechanism. Illari et al (226) define a reasoning mechanism as consisting ‘of entities and activities organised in such a way that they are responsible for the phenomenon’. Falsification (Popper) Trial and error Popper (236) 30 agreed with Hume: we cannot say anything about the future, there are only theories that cannot even be proven. At best, we can only prove that they are wrong (falsifiable). This ‘anti-inductivist’ reasoning suggests to continue using a certain theory or practice and adjust if they fail. Precautionary principle In case of uncertainty about the future prevent harm The precautionary principle, often used in environmental decision making and occupational health, favours to take preventive action in the face of uncertainty when making an inference. It puts ‘the burden of proof to the proponents of an activity; exploring a wide range of alternatives to possibly harmful actions; and increasing public participation in decision making’ (227, 237). Chapter 5