After having defined necessary connection in Sections 3 and 6, Hume proceeds to further explore the concept of necessity, wanting to determine how the mind draws the connection. This discussion takes place in Section 14, Book 1, Part III of the Treatise. Section 14 elaborates on ideas first raised in Section 2. Hume will continue to show that necessary connection comes about due to us experiencing constant conjunctions between two objects.
Section 14: Of the Idea of Necessary Connection
Hume starts by again reinforcing a principle he has previously declared. There is nothing in any object that gives us the idea of relation to another object. Even if we were to see the two objects contiguous to one another, with one object having priority in time, this by itself is not enough for the mind to draw any connection between the two objects. Only after a frequent repetition of events A and B does the mind start to develop the idea of the two events being necessarily connected. Hume will use the word ‘power’ to describe the strength of the relation that exists between two objects.
To explain what he means by ‘power’, Hume believes it would be a waste of time to use other terms such as ‘efficacy’ or ‘agency’ or ‘connection’. All these words have a similar meaning, so we won’t get very far in understanding what ‘power’ really means. Reason alone can do nothing to ‘rationalise’ about the connection, since reason does not give us any original idea. Again, we must turn to our impressions obtained from experience.
If two objects are presented to us for the first time, we have no reason to think that they are connected. One instance alone is usually not enough to arrive at the idea of cause and effect. Only after several repeated instances do we begin to conceive of a connection. The multiplicity of events is the power that strengthens the connection. So power is not based on new ideas and impressions, but rather experiencing the one idea repeated over and over again.
Even so, why does the multiplicity have this effect? Simply experiencing the same conjunction over and over again tells us nothing new about the objects themselves. Each instance on its own has no influence on any other instance. So this continues to prove that any idea of necessary connection does not come from the items themselves, no matter how many times the conjunction is repeated.
Instead, the multiplicity results in a new impression in the mind. The several instances of conjunction strengthen the power of the original impression. This happens purely in the mind, so must be another impression again. Hume concludes that because reason cannot provide this impression, it must come from the imagination. Based on contiguity and precedence of cause before effect, the idea of one objects causes the mind to form the idea of the other. This cannot come from any one impression on its own, but from a mental process by which the power of an idea is strengthened by multiple instances.
Hume finishes Section 14 with a couple of definitions of cause and effect, which help to tie together the previous discussion. The relationship of cause and effect could be based on a philosophical relationship (a comparison of two ideas) or a natural one (an association between two ideas).
The first definition of cause (the natural one) proceeds as follows: all events that resemble c (the cause) will be followed by events that resemble e (the effect). Hume objects to this definition, thinking that it gives the impression that once we’ve discovered a cause, this suggests that it should apply to every single instance across time and space. The definition therefore introduces an undesired ‘foreign element’. In other words, any claim to causality can only ever be general. But this argument largely depends on how the original claim for causality is made. Given that you can phrase your causal claim in whatever way you wish (eg ‘today at 2pm, I saw smoke coming from a nearby park, so assumed there was a fire’), this doesn’t suggest that every claim has to be a general determination.
As for the second definition of cause and effect (the philosophical connection based on necessary connection), Hume is concerned that it also introduces a ‘foreign element’ into the equation, that of the state of mind of the person making the casual claim. Given that necessary connection IS a product of the mind according to Hume, it is difficult to see how we can get around this problem.
Overall, the two definitions are similar to each other, and any differences are technical in nature. However, they do give rise to a number of related ideas. Can there ever be a real event in nature without any association being formed in the mind? The answer would have to be ‘yes’. We have to assume that a physical world continues to function around us, and this is subject to the laws of nature. Just because no one has ever formed an association of an event doesn’t stop nature from continuing to function. Life clearly started on this planet without anyone being around to observe the event.
Similarly, can there ever be associations in the mind that do not relate to real events? Again, the answer has to be ‘yes’. Hume himself acknowledges that we can create new ideas in our head by combining existing ideas. Simply creating the connection (‘the cow jumped over the moon’) is no proof that the event really did occur.
Hume’s argument in Section 14 can be summarised in five parts:
- Simply observing the repeated occurrence of an event does not provide necessary connection
- But there is something about the repeated occurrence that gives us the impression of necessary connection
- The repetition has no effect on actual events that take place
- The repetition instead creates a new impression in the mind
- Therefore, the custom of observing repeated instances is the basis of necessary connection.
Section 15: Rules by which to Judge Causes and Effects
Hume summarises the previous discussion by proposing a list of eight rules. The first three rules summarise the main requirements of cause and effect: (1) contiguity, (2) the cause must be prior in time to the effect, and (3) there must be a constant union between cause and effect.
After having established a general rule about resemblance (‘all events that resemble A are followed by events that resemble B’), Hume notes that we then seek instances of A that have some resemblance to our current idea. It may be the case that we do come across an ‘A-like’ event, but this is not followed by B. We must therefore assume that we misunderstood the original cause-and-effect relationship.