Elizabeth Anscombe continues her argument of causality. After having dismissed the arguments of Hume and other philosophers, she will now start her own argument as to what causality is.
Rather than believing cause and effect to be some Hobbes-like logical connection (if A, then B) or some Humean notion of necessary connection, Anscombe introduces the idea of effects being ‘derived’ from causes without implying any form of generalisation or natural law. For instance, parenthood is a causal relationship between parent and child, as the child has derived (come from) the parent. As a further analogy, Anscombe mentions the idea of someone travelling from a place, for example ‘Bob came (travelled from) London’. When we hear this statement, it does not necessarily follow that the following propositions are true:
- That Bob ALWAYS comes from London
- That people travel from London in general
- That if Bob came from somewhere, it had to be London
- Given that Bob exists, a London has to exist.
Anscombe is claiming that ‘comes from’ in the sense of travel is how we should view cause-and-effect. We should see effect B as something that comes from cause A, without giving rise to any form of generalisation.
Anscombe concludes her essay with two objections the Humean might make to the travel analogy, and then follows this with counter-arguments of her own. The first objections relates to the Humean claim that we can never observe causality in any single individual case. The Humean would argue that the travel analogy does not apply as travel is observable, whereas cause-and-effect does not exist in the objects themselves (rather the connection is created as an inference in the mind). Anscombe gives two replies to this objection:
- A Humean talks of ‘finding’ something in perception which is different from ‘finding’ causality between two objects in terms of necessary connection. However, the Humean cannot provide any evidence that ‘finding’ exists in one case (necessary connection), but not the other sense (perception). The Human is being selective in how they define the idea of ‘finding’.
- The Humean admits that we perceive causality when we observe bodies in motion (such as the billiard balls about to strike). If the Humean accepts that perception is sufficient to observe the motion, then why can’t they also accept that perception is sufficient to observe the causality as well?
In addition to the first objection, Anscombe notes that the word ‘cause’ is very broad and represents many concepts in English (scrape, push, carry, make, throw, drove, drank, etc). Our understanding of cause therefore derives from language itself. We don’t need a technical definition of cause to understand what the concept means. We understand cause intuitively when we say that B was caused by A.
The second objection to the Humean point-of-view is that we can rarely make a universal generalisation when we experience an individual case. There are always exceptions and special circumstances in the real world. When used in any universal law, the word ‘always’ needs to be treated with scepticism. For example, even a simple fact such as ‘water boils at 100 degrees’ is not strictly true in all cases. The boiling point is lower on top of a mountain.
As an example of her argument, Anscombe notes the following case: when an alarm bell sounds (A) somebody wakes up (B). But even if this occurs often, it cannot be true in all cases. The alarm might be timed to go off a certain time each day, even when nobody is in the house. Also, the people in the house could have already woken up via other means (for example, by receiving a phone call).
When defining causality, Anscombe makes a distinction between a ‘law of nature’ (a scientific law that does contain an element of necessary connection), and making predictions in normal conditions. But as pointed out with the water boiling example, even laws of nature often need to be framed as applying to specific conditions only.
We should view ‘universal generalisations’ of cause and effect in the same way. If water doesn’t boil when the temperature happens to be 100 degrees, we must look for some other cause to explain why the expected effect didn’t occur, rather than insisting that the connection is necessary.
Anscombe’s essay is useful in that she applies cause and effect to the real world, whereas Hume’s approach is more theoretical and abstract. Sometimes philosophers make the mistake of thinking from the armchair, without giving proper recognition to what we experience in practice (being an empiricist, Hume should accept this point). Anscombe is correct to point out that humans suffer from ‘tunnel vision’ and will happily make conclusions based on a few scanty pieces of evidence, without asking themselves whether the conclusion is really true. This can be seen in how the news media report on certain events (eg migrants attempting to stowaway on vehicles), which might give the viewer an impression that all migrants act like this.
To summarise Anscombe’s position:
- Hume argues that we need to experience several instances of a conjunction before form the idea of necessary connection. Anscombe suggests that we can perceive the connection even after a single observable instance.
- Hume claims that necessary connection gives rises to no exceptions, whereas Anscombe indicates that it is almost impossible to find generalisations that are truly universal.