When trying to identify causality, do we assume “nearness” between cause and effect?

Question

When asking people what causes what, it seems that they assume that causality has something to do with "temporal nearness" and "space nearness". That is:

• If I turn on the light switch and the lamp turn on a short time after that, then the light switch is really what is turning the lamp on.

But using this, if I turn on the light switch and a person dies, then we could deduce that this effect also happens because I turned the light switch. In a scenario with a lot of people, it could be probable that one person would die every time you turn the light switch. And hence, that conclusion could be deduced.

• If I turn the light switch on and the lamp turn on and I believe there is a connection between the light switch and the lamp - that is, they are near, connected by something - then the light switch is really what is turning the lamp one but this has nothing to do with the people dying because there is no connection.

Are there formal names to these 2 kinds of nearness? I'd like to read more about it.

Answer 1

First, we need to distinguish between the proximate and mediate causes, because we would call throwing a ball to be the (mediate) cause of a broken window even though the throwing hand never came into contact with the window. But it is indeed common to assume that proximate causes, the elementary links of causal chains, do involve spatio-temporal proximity. Cognitive psychologists believe that the concept of cause initially forms in infants from observing that making something move requires them to contact it, and then is reinforced by observing motion transfer through contact. So the "intuition" behind causation does presuppose the spatio-temporal nearness, a.k.a. contiguity, and even a material connection, contact. Here is Mandler in On the Birth and Growth of Concepts:

We know that infants perceptually differentiate self-starting from starting with contact at least by 6 months of age (Leslie, 1982)... Pace Hume’s belief that we cannot see causality, there is evidence that we can, or at least we see one of the major components of a commonly experienced causal relation, namely, the transfer of motion from one moving object to another. White (1988) proposed that the powerful sense of causality perceived when, say, a billiard ball strikes another, comes from the short duration of iconic storage (about 250 msec).

The iconic store is a large-capacity sensory store that holds visual information prior to attentive processing; it is continuously refreshed, enabling the temporal integration, which makes us see motion as continuous... We see a causal relation when a conflict exists between two types of continuity cues. Spatial discontinuity between two objects says there are two objects, whereas continuous motion suggests there is only one. The conflict is resolved by perceiving the causal sequence as the transfer of motion from one object to the other.

However, the transfer of motion example rather confirms Hume's view that what we directly observe is not the "causing", but one type of event following another with regularity. Thus, causation plays a role similar to theoretical entities, that of unifying experiences, except it formed even before scientific practice made introduction of such entities systematic and deliberate. Hume thought that it is formed through a subconscious psychological mechanism, "the law of association". Once this "theoretization" is accomplished the presence or absence of causation is determined based on holistic considerations of forming a coherent picture of environment rather than just induction on individual observations. Some previously admitted regularities, such as thunder following lightning, are reclassified as having common cause rather than directly causative, others are dismissed as coincidences or superstitions. At this stage proximity still remains a requirement for proximate causes.

In science, however, the notion of causation is further subsumed under instantiation of a general predictive law, or a combination thereof, classically expressed by various differential equations. For example, we can say that one ball causes another one to move upon contact because that is what the laws of mechanics predict. At this point it becomes logically possible to have proximate causes without spatio-temporal contact, and indeed Newton introduced gravitational action at a distance that ostensibly did not require it. This was however in sharp conflict with the pre-existent everyday notion of proximate cause, and the following passage from his letter to Bentley shows just how conflicted he was about it before acquiescing to "hypotheses non fingo" [I feign no hypotheses] stance:

It is inconceivable that inanimate Matter should, without the Mediation of something else, which is not material, operate upon, and affect other matter without mutual Contact... That Gravity should be innate, inherent and essential to Matter... is to me so great an Absurdity that I believe no Man who has in philosophical Matters a competent Faculty of thinking can ever fall into it.Gravity must be caused by an Agent acting constantly according to certain laws; but whether this Agent be material or immaterial, I have left to the Consideration of my readers.

Answer 2

At least since David Hume, the father of the modern philosophical discussion of causality, it is customary to associate causality with contiguity in space and time, and with succession in time.

Contiguity means not merely nearness, but a kind of continuity. Immediate (or "proximate") cause and effect are supposed not only to be near to actually touch each other. For example, it is not enough that the light switch is near the bulb. A continuous electrical circuit must be connecting them.

In such ways, mediate cause and effect are supposed to be connected by a chain of immediate causes and effects, such that each successive cause/effect pair are contiguous.

Succession in time means that the cause is supposed to precede the effect, in time. Such as pressing the light switch precedes (if only by very little) the appearance of the light.

You can read about all that in David Hume's Treatise on Human Nature, in the section Of the Idea of Cause and Effect.

The idea, then, of causation must be derived from some relation among objects; and that relation we must now endeavour to discover. I find in the first place, that whatever objects are considered as causes or effects, are contiguous; and that nothing can operate in a time or place, which is ever so little removed from those of its existence. Though distant objects may sometimes seem productive of each other, they are commonly found upon examination to be linked by a chain of causes, which are contiguous among themselves, and to the distant objects; and when in any particular instance we cannot discover this connexion, we still presume it to exist. We may therefore consider the relation of CONTIGUITY as essential to that of causation; at least may suppose it such, according to the general opinion, till we can find a more [Part IV. Sect. 5.] proper occasion to clear up this matter, by examining what objects are or are not susceptible of juxtaposition and conjunction.

The question, whether there could be a causal connection without contiguity, is debated. At least since Newton's force of gravity, scientists have sometimes accepted the existence of actions from distance i.e. of some causal connections that do not require spatial contiguity.

Answer 3

The example you give reminds me greatly of a XKCD comic:

If you flip a light switch many times, you can find a strong correlation between flipping the light switch and the light turning on. As the comic suggests, this is not a causal relationship yet, just a correlation.

To become a causal relationship, the cause must be necessary and sufficient for the result to occur. The result should not occur without the cause, and if the cause occurs, so should the result. Sometimes we'll weaken this to just sufficient, meaning as long as the cause occurs, so does the result.

As far as the philosophy of science is concerned, this is a predictive act. When you associate the cause and effect, what you are in fact stating is that future causes will yield predictable future effects. This is, of course, impossible to prove, because nobody knows the future. However, in the philosophy of science, one strives to bring forward enough evidence that the events are causal such that one may come to believe they are truly causal through abduction.

Science's experimental methodology is designed to support this. Great care is taken to make it hard to mistake correlation for causality. This is at the heart of falsification. While a person may mistake correlation for causality, a second person repeating the experiment may falsify this mistaken conclusion.

The "nearness" effect is one that comes from our model of the world. We have a pretty good model of the world in our minds, most of the time. Most of the time things go as we expect them to. When they don't, we look for a "proximate cause." Scientifically, this always means a "local" cause in space and time. We're looking for some potion of the system whose state is not the state we predicted, right before we observed the effect. We assume "locality" because science has found a great deal of evidence to suggest it is true (like everything in science, all we can do is create a preponderance of evidence). This is such a well accepted assumption that when quantum entanglement suggested that the laws of the universe may in fact be non-local, this was received with quite a great deal of resistance. However, almost anything in science can be falsified, even deep rooted assumptions like this. With enough evidence, we have shown that quantum non-locality describes the experimental evidence well, while models which assume locality do not.

So those would be the three terms I would consider to answer your question: correlation, causality, and locality.

Answer 4

The first mention of this I have seen is in Aristotles De Anima:

419a15 For Democritus did not speak rightly, thinking that if the intervening space were to become a void, then even if an ant were in the sky it would be seen accurately; for this is impossible.

For seeing takes place when that which can perceive is affected by something. Now it is impossible for it to be affected by the actual colour which is seen; it remains for it to be affected by that which is intervening.

But if it were to become a void, not only should we not see accurately, but we would see nothing at all.

A goes into this in more detail in his Physics.

It's this notion thats later codified into the principle of locality, of which there are a few variants.