Anscombeの哲学（続く）

著者が次に原因と結果の接続に疑います。
One example the author provides is the contraction of a malady. Given exposure to the cause of some plague the disease manifests itself, one supposes the cause to be the said exposure. However, given the exposure, a doctor cannot predict that the disease will necessarily ensue. The next argument endeavors to separate causes from causation. An analog is the esoteric notion of "valency", which cannot be clarified and yet has been observed in scientific investigation. It may be possible to note the presence of "causes" without stating what causation is.
次に原因として動く事が必要さと関係がないと述べられます。　Laws of nature are used to explain cause and effect (eg the growth of a plant) by deriving the effect from the cause. However, there is no necessity or universality involved which demonstrates that the cause is the source of the effect. This seems tenuous to me, as I do not see how derivation occurs without acknowledging the universality (causality) of the law of nature.
The notion that "A comes from B" does not dictate that every item similar to A comes from something similar to B. If there is no necessity/generality which leads to the association of cause with effect, then perhaps the association is due to perception and the constant connection between perceptions.
The author then asserts that we experience causality for the first time by learning language. Even without the word "cause" itself, causal concepts are embodied by words like "knock" and "eat". Thus, simply by speaking, we are unconsciously expressing causes without implying any necessity.
An aside is made about Cartesian scepticism, which I assume questions how any perceptions are made evident in the mind. The argument is one of Hume, it seems. Thus, perception of "cutting" (a cause) may not actually be linked to physical sensations in the same way that the perception of "red" (based on light frequencies striking the retina) is. So the Hume argument seems to be that causes are obtained by some other connection in the mind besides perception.
The next argument made by Anscombe is that universal generalizations expressing causal relations are not possible, e.g. "Motion in one body upon impulse is followed by motion in another." might be stated upon observing two billiard balls colliding. However, this would not hold true when a billiard ball strikes a wall. Thus, any general statement must include all cases where the effect does not follow the cause, and such an enumeration of all cases is impossible. The only correct way to state such a relationship is "if cause A occurs and effect B does not, there must be some cause of its not doing so." Such cause and effect relationships differ from "laws of nature", which are generally true in all circumstances included in the law. For example, Newton's law of inertia would not be stated as "If a body accelerates without any force acting on it, then there must be some cause of its doing so." It is stated as "A body upon which no force acts will not accelerate". Examples of cause and effect relationships which are not universal are those expressed in medicine, since the medical profession seeks to deal with phenomena in an uncertain world.
The author next attempts to prove by example that exact determinations of effects are impossible. The model is one of a ball which can drop into any one of multiple vertical pipes. Newtonian physicists (and Hume, it seems) would state that the motion is absolutely deterministic. However, all measurements of velocity, position, etc have some margin for error. If we desire the probability that the final outcome we predict is the correct one be very high, the margin for error must be infinitesimally small. In fact it must be smaller than what is physically possible.