Notes on Hume’s Treatise

by G. J. Mattey

Book 1


Of the UNDERSTANDING
PART 3
Of knowledge and probability.

Sect. 2. Of probability; and of the idea of cause and effect.

Context

In the previous section, the author had singled out four of the seven philosophical relations as being the objects of knowledge and certainty. What they have in common is that they depend solely upon the related ideas, such that any change in the relation requires a change in the ideas. In the present section, the author will consider the remaining three philosophical relations, which can change without a change in the related ideas. Although the title of the section indicates that these relations are ones of probability, that notion is not mentioned within the section itself.

Background

The idea of cause and effect is treated by Locke in Book II, Chapter XXVI of his Essay. According to Locke, the idea is based on experience of change in things and the observation that they are the result of the activity of another being. “In the notice that our senses take of the constant vicissitude of things, we cannot but observe that several particular, both qualities and substances, begin to exist; and that they receive this their existence from the due application and operation of some other being. From this observation we get our ideas of cause and effect. That which produces any simple or complex idea we denote by the general name, cause, and that which is produced, effect.” This view differs from that of the author, who holds that we cannot observe “application and operation” of any being.

The Treatise

1. As the preceding section has explained all that is needed about the relations which depend entirely upon the ideas related, it remains to examine the other three relations, of identity, the situations of time and place, and causation. Each of these relations is such that they can be different even while the related ideas are the same. The ideas may be related while the relation in the objects is absent, and the ideas may not be related when the relation of the object is present.

2. The author begins with a blanket statement about the nature of reasoning. All that it involves is comparing two or more objects and finding what relations they have to one another. (The relations may themselves be “constant” or “inconstant.”) [For example, two objects might remain constantly contiguous over a period of time, or they might be inconstantly increasing or decreasing their distance from each other. The relation here is that of situation in space.] There are three possible conditions under which comparison can be made:

If both objects are present to the senses, there is no reasoning, but only perception, “a mere passive admission of the impressions thro’ the organs of sensation.” Thus there is no reasoning regarding our “observations” concerning identity and relations of time and place. [For example, if I fix my gaze on an object uninterruptedly, I may perceive its identity over the time in which I perceive it. Or I may see two objects with no apparent object between them. But observation does not amount to knowledge, because it remains possible that the objects differ while the ideas remain the same. Two objects might appear contiguous because of the viewing angle when they are not in fact touching.] Causation alone is a relation that cannot be observed, but which takes the mind from the existence of a perceived object to that of an object, preceding it or following it, that is itself unperceived. If we are to reason about identity and spatial or temporal relations, the reasoning must involve the relation of causality. For example, if we go beyond observation and infer that two objects are invariably adjacent or distant, it is because we think there is a secret cause connecting them or separating them. And if we observe an object for some time, break off the observation, then upon resuming it find a perfect resemblance, we may only infer that it is an identical object if we think there is a cause that keeps the object the same. Given that there is a cause, we think that if we had had a continuous perception of the object, it would have appeared the same throughout. Only if we believe that there is a sustaining cause can we have any assurance that the object endured and was not replaced by a resembling one. “Whenever we discover such a perfect resemblance, we consider, whether it be common in that species of objects; whether possibly or probably any cause cou’d operate in producing the change and resemblance; and according as we determine concerning these causes and effects, we form our judgment concerning the identity of the object.”

3. Thus, it seems that causation is the only one of the three relations which sanctions inference beyond what is observed. “This relation, therefore, we shall endeavour to explain fully before we leave the subject of the understanding.” [In fact, the explanation occupies the vast bulk of Part III.]

4. If the author’s procedure of investigation is to be followed, the next question is that of the origin of the idea of causation. We cannot reason “justly” [with justification] without an understanding of our idea of cause, and this requires finding its origin “and examining the primary impression, from which it arises.” [This is in accordance with the principle that every idea is a copy of an impression or impressions.] If we have a clear impression, the idea will be clear, and if the idea is clear, “all our reasoning” will be clear as well.

5. So we may ask of any two objects, one of which is deemed a cause and the other an effect, what is the origin of the idea of the relation of cause and effect that is supposed to hold between them. One possibility is that the relation is to be found in the qualities that the two objects have themselves. Suppose that object A has quality Q, the impression of which gives rise to the idea of A as a cause of B. The author asserts that whatever Q might be, there are objects C which do not possess Q yet are causes nonetheless. [Perhaps Q is thought to be motion, for example. Motion does not account for such causal relations as that of an idea of the beauty of a possession of mine causing pride in myself.] A second argument is that every object is in fact considered to be a cause or to be an effect. But there is no single quality that holds of every object. So, there is no single quality that can itself account for an object’s being a cause or an effect.

6. Since there is no quality in objects that is sufficient for their being deemed causes or effects, we must look elsewhere to explain why we assign that relation to objects. It could only be some relation among objects. But which one or ones? The author finds one relation that is universally thought to hold between cause and effect: the cause and the effect are contiguous in space. While it may appear that an object’s state is changed by a distant object, we commonly find that there is an intervening chain of contiguous objects that act as causes, and if we do not find one, we suppose that there is one. [For example, Newton believed that gravitational attraction at a distance must be explained by action through intervening matter.] So contiguity will be considered essential to the relation of causation, at least until the matter is examined more fully in Part IV, Section 5.

7. Also essential is the relation of priority in time of the cause before the effect. Some claim that causality can be contemporaneous, i.e. that the cause and effect exist at the same time. [Kant believed that there is a contemporaneous relation of “community” which is similar to causality in that it explains why physical objects behave the way they do. See the “Third Analogy” in the Critique of Pure Reason.] But, 1) experience apparently contradicts this, and 2) an inference may be made to rebut the claim of simultaneous causality. The author begins with the generally accepted claim that “an object, which exists for any time in its full perfection without producing another, is not its sole cause.” [For example, suppose that billiard ball A is touching billiard ball B, but neither of them is moving. If A is to move B, it must be in virtue of A’s being struck by another ball C, or being moved by something else.] If the object does produce another, it is because it has an “energy” that it “secretly possest,” that was activated by another object. Now suppose that any cause may be contemporaneous with its effect. And suppose that A at t1 is a prior cause of B at t2. Then at t1, A did not exist in its full perfection, since B was not produced at t1. In that case, A is not the sole cause of B, given the principle agreed to above. So nothing prior could be a cause, in which case all causes are contemporaneous. “The consequence of this woul’d be no less than the destruction of that succession of causes, which we observe in the world.” As if this were not enough, a collateral consequence would be that there would be no succession, since all causes of change would be contemporaneous. And if there is no succession, there is no time, since succession is the basis of the idea of time, as was argued in Part II, Section 3, paragraph 10, where it was claimed that the idea of time “can plainly be nothing but different ideas, or impressions, or objects dispos’d in a certain manner, that is, succeeding each other.”

8. It would be good if this argument succeeds, but even if it does not, its conclusion, that causes are prior to their effects, can be supposed to be true. But whether it is true or not will be found to be “of no great importance.”

9. Recall that the goal was to discover why in particular instances, one object is to be deemed the cause and the other the effect. Observation of the relations of contiguity and priority is not sufficient for this. For example, we observe that one body strikes another and that its motion precedes that of the second body with no detectable time in between. This does not establish that the motion of the first causes that of the second.

10. We might try to remedy this deficiency by saying that the motion of the first body produces that of the second. But this poses a dilemma: either the notion of production is the same as that of causing, in which case his explanation is circular, or that notion must be defined, which the author challenges anyone to do.

11. But this throws us back on contiguity and priority as the only two bases of the causal relation, and this will not do. We can always think of contiguous and prior objects as not standing in a causal relation. Now the author supplies the missing third relation, that of necessary connection, “and that relation is of much greater importance, than any of the two above-mention’d.” [This relation will not be fully explained until Section 14.]

12. How is the idea of a necessary connection to be found in the idea of an object? We must find the impression that gives rise to the idea of necessary connection, but where is it to be found? It cannot be a known quality in the object [since it was shown in paragraph 5 that no known quality can account for causality, of which necessary connection is a component]. The only relevant relations of objects are contiguity and priority, which by themselves are not sufficient to make up the relation of causality. One option might be to dispense with the copy principle and declare that the idea of causality is not copied from an impression. But to reject a principle as firmly established as the copy principle would be to take that principle to lightly. We must at least try to find another approach before giving it up.

13. Since we have no clue at this point regarding the origin of the idea of a necessary connection, the best approach would be to consider some related questions in the hope that they will unexpectedly provide an answer. This is similar to looking in nearby areas for something that we think we have lost in the immediate surroundings. We can only hope that we will find it in an unexpected place. [And it will be found in a very unexpected place, as is detailed in Section 14.] The author proposes two questions that might yield a hint as to the origin of the idea of causation.

14. “First, For what reason we pronounce it necessary, that every thing whose existence has a beginning, shou’d also have a cause. [This question will be addressed in Section 3.]

15. “Secondly, Why we conclude, that such particular causes must necessarily have such particular effects; and what is the nature of the inference we draw from the one to the other, and of the belief we repose in it.” [The answer will be found in Sections 4 through 10.]

16. Although emphasis has been and will be on the idea of the causal relation derived from impressions of sensation, impressions of reflection admit no less of the relation of cause and effect, connecting our passions. “Passions are connected with their objects and with one another; no less than external bodies are connected together.” So there is a common relation of cause and effect between external and internal objects.

The Enquiry

In Section IV, Hume makes the negative claim that the relation of cause and effect cannot be found in any quality of objects. “No object ever discovers, by the qualities which appear to the senses, either the causes which produced it, or the effects which will arise from it.” He discusses the relation of necessary connection in great detail in Section VII. However, he does not decompose the relation of cause and effect into those of contiguity, priority and necessary connection.

[ Previous Section | Next Section | Treatise Contents | Text of the Treatise ]