Philosophy Club

What is a Cause?

Introduction

We live in a world of causation. The notion of causation is basic to our understanding of how the world works. Causes are everywhere. Here are just a few examples:

1. Smoking causes lung cancer.
2. The cause of the fire is not yet known.
3. What were the causes of the First World War?
4. The remark caused him to blush.

Causation occurs in a wide variety of circumstances, legal, medical, historical, economic and in our day to day life. We think and talk in terms of causation. But have you ever wondered: What exactly is a cause? What is the nature of causation? One of the most well known answers to these questions was given be David Hume.

Hume's theory of causation

According to Hume there are three essential characteristics of causation. If A and B are events, then A causes B only if,

1. A and B are “spatially contiguous” ie. are in contact.
2. A occurs before B.
3. A and B are “constantly conjoined”.

To explain this in more detail: A and B stand for such events as “the firing of the gun”, “the collapse of the bridge” and so on.

According to (1), if A causes B then A and B must be in contact. This contact may be direct as in “Striking the match caused it to ignite”, or indirect as in “The gravitational attraction of the moon causes the movement of the tides”.

(2) is clear enough - causes must precede their effects.

However (3) needs some explaining. According to Hume, (1) and (2) were not sufficient. One vital ingredient was missing - necessary connection. If A causes B then A and B are necessarily connected. If A occurs then B must occur. If the match is struck with sufficient force, then it must light; if the iron bar is heated to the right temperature then it must expand. This idea of necessary connection between cause and effect is often expressed in terms of causal powers. Thus, cause A is necessarily connected to effect B in virtue of A's having a causal power.

For instance, a match contains a causal power to ignite, an acid contains a causal power to corrode other substances. Causal powers are the basis of necessary connections between cause A and effect B. But, Hume asked, what exactly are causal powers and necessary connections? We don't observe them. All we observe is the match being struck and then lighting, the blue litmus paper being placed into the acid and then turning red, the first billiard ball colliding with the second billiard ball and the second ball moving off, and so on.

In these situations we don't observe any necessary connections or causal powers. So why believe them to be an essential ingredient in the notion of causation? Hume's answer was this: The idea of necessary connection is a construct of the mind. Repeated observations of A's being followed by B's produces in our minds the idea that B's must follow A's.

For instance, we observe many instances of a match being followed by its igniting, and as a result we form the idea of a necessary connection between match being struck and its lighting. Hume's point is that there are no necessary connections or causal powers in nature; they are simply a construct of the human mind.

“Necessity, then, is the effect of this observation [ie of several resembling instances], and is nothing but an internal impression of the mind, or a determination to carry our thoughts from one object [eg a match being struck] to another [match lighting].”

For Hume, all that exists in nature is a constant conjunction of resembling instances such as matches being struck followed by their lighting. In terms of our example, Hume's theory of causation can be explained in this way: The striking of the match is the cause of its lighting in virtue of these following three facts:

1. The striking of the match is “contiguous” with its lighting - there is contact.
2. The striking of the match precedes its lighting.
3. All instances of striking a match with the required force are regularly followed by (“constantly conjoined with”) the match lighting.

Some difficulties for Hume's theory

1. The problem of singular causal statements: some causes and their effects are unique - there are no resembling instances. Consider, for example, “The dropping of an atomic bomb on Nagasaki caused the Japanese to surrender”. There are no resembling instances or regular connections in this case. The cause and its effect are unique occurrences in history, never to be repeated. Yet there seems to be no doubt that this is a genuine case of causation. This suggests that there may be causation without constant conjunctions of resembling instances. Perhaps, contrary to Hume, the central idea of causation is of one event A producing or bringing about another event B, and this may occur without A's regularly being followed by B's.
2. Hume's three conditions for causation may be necessary but do not seem sufficient. For example, think about the connection between low and high tides at a certain place or between night and day. The event “Low tide at Sydney Harbour at 6 am” doesn't cause the event “High tide at Sydney Harbour at 12.15 pm”. Nor does the occurrence of day cause the occurrence of night. The actual causes are, respectively, the gravitational attraction of the moon, and the rotation of the earth on its axis. Thus it appears there are cases where, for two events, there is contact, temporal precedence and constant conjunction but no causation.

Some questions

1. What, if anything, is missing from Hume's account? If, contrary to Hume, necessary connections and causal powers do exist, then what are they? How could we establish their existence?
2. Some properties such as brittleness and solubility do seem to attribute causal powers. But can we explain away the apparent existence of causal powers like this: To say that glass is brittle is to say nothing more than “If glass is hit with sufficient force it will break”? In other words, every property such as brittle, soluble, flammable, malleable and so on can be analysed by means of a suitable conditional statement like the one given for “brittle”. Thus, Hume can be saved from causal powers. Does this work?
3. Is it possible to have a universe containing exactly the two following events: A match being struck followed by its igniting? If so, then Hume is wrong in thinking that causation requires laws and generalisations (“constant conjunctions”), because in such a universe there will not be any regularities since there are no other events.

Reference

D. Hume, A Treatise of Human Nature.

R. Neurath