« AnteriorContinuar »
SCIENCE AND THE PECULIARITIES OF THE RELATIONS
THAT IT TRACES.
The conviction that there is a distinction between circumstances and things is only a starting-point in our search for
the uniformity that we believe to pervade the Twofold
work of science.
world. The vast differences between one com
plex group of phenomena and another are not explained to our satisfaction by the mere general statement that they are all due to differences between things and the circumstances in which they are placed. The difference between one complex situation and another is always perfectly definite, and what we want to find out is the precise difference in the things and in the circumstances that accounts for this definite difference in the situations as a whole. This is the business of Science, and its work has two sides: (1) Observing as much as we can, and finding out from what we observe the general laws according to which things always act; and (2) inferring further from the concrete combination of circumstances or events that we observe and from the general laws that we have discovered what must be the concrete state of affairs where we cannot, or cannot yet, observe. As a result of various concrete observations made by himself and others Newton discovered and proved the law of gravitation; and an astronomer who knows this law and also observes the position of some heavenly body at various intervals is able to tell where it was before he saw it, when it will come within a certain distance
of the earth or the sun, and what will happen to it and to them when it does. In this way any one who was master of an absolutely perfect science would be able (1) to find every law in the universe, and then (2) starting with the present, to go back indefinitely and tell the history of the past and to look forward just as far and tell the story of the future. In the following chapters we shall speak first of the logical method pursued by scientists in the discovery of general laws, and afterwards of the application of these laws for the discovery of particular concrete facts. But before we begin it is desirable to say a few words more about the nature of the identity and causal interaction assumed in all such investigations.
To tell whether one object resembles another we need only look at the two and compare them, and in the same way we can often tell by direct observation whether one event succeeds another; but to tell whether one Peculiarities
of individual object is identical with another, or whether some identity
and causal given event is the cause of another, simple interaction. inspection is not sufficient. If we lose sight of a thing for a single instant how can we tell that it has not been removed and another put in its place ? Indeed, until we know something more about it than its outward appearance, how can we even be sure that such a substitution has not taken place by some jugglery before our very eyes ? So also with the causal relation between one thing and another; however certain we may be that a change we observe in one thing followed immediately after a change we observed in something else, how can we be sure that it was caused by this rather than by a change in some third thing, miles away perhaps, that we did not happen to observe at all? No one can directly observe either the identity of an object with itself or its causal action upon something else, and therefore our identifications and causal explanations have to be reached by guesswork in the first instance, and if they are afterwards proved to be correct the proof has to be indirect.
Another thing to notice about these relations of identity and causal interaction is the conviction we feel that they are always present. Since they are aspects of the uniformity of nature that we believe in, we have a right to believe that at every moment in the past there were things identical with the things of the present, and that there are and always have been causal relations to other things that help to explain their present conditions; and the same is true, mutatis mutandis, of the future.
Still another peculiarity of these two relations is their exclusiveness. In this respect identity and causal interaction are very different from resemblance and succession. The house I live in can be similar to each one of a hundred others, wholly regardless of the place and time of their existence, and its similarity to one does not interfere in the slightest with its similarity to some other. It is true too that the house that keeps out the rain to-day can be identical with the house that was bathed in sunshine yesterday and none the less identical on that account with the house that was covered with snow two months ago, for these relations of identity and causation go back from instant to instant without a break to all eternity; but the house cannot be identical with one existing in some other place or state at the same time. There is only one thing at a time with which it can be identical, and to be identical with this means not to be identical with some other. The same exclusiveness is found also in the case of causation. The house we have been talking about can exist at the same time as an unlimited number of others, and an unlimited number of things might have been acting before or during its erection; but if the house was built by John Smith it could not have been built by anybody else; and if a spark from a certain locomotive destroyed it, it is quite certain that an earthquake or a stroke of lightning did not. It is perfectly true, to be sure, that if we try to explain absolutely everything about the house as it exists at a given instant the explanation will have to include
a statement about absolutely everything in the universe the instant before. Thus, inasmuch as the position in space of every object depends upon that of every other, the position of the house would certainly be affected by a boy throwing a stone in China. And if we use the word 'cause 'with reference to some single aspect or relation of things that we are trying to explain it is true also that even this single relation may be due to the co-operation of several causes. John Smith may not have built the house alone. But when causes co-operate in this way no one of them is a complete cause. If anything is the complete cause of a condition, so far at least as one instant of time is concerned, it is its sole cause. A resemblance or coexistence or succession can belong at once to many things without diminishing the share of any. A relation of identity or causal interaction can not.
Our knowledge, or rather our assumption, that these relations of individual identity and causal interaction always exist and are always exclusive is very helpful in our search for uniformity in nature. If we have any way of showing that this, that, and the other thing of one instant is not identical with the thing of another, the first of these assumptions gives us a right to infer that some other—perhaps the only one left-is. In the same way, if we have any means of showing that this, that, and the other thing had nothing to do with some present condition of a thing that we are inquiring about, our assumption gives us the right to infer that something else had. In this way we can apply the ‘method of exhaustion ' in our search for particular uniformities. So with the second assumption, if we know that this thing of one instant is identical with a given thing of another, we know perfectly well that we need not look for any other thing of the same instant to be identical with it. Here again the same is true, mutatis mutandis, of causation. In this way our search for uniformity is simplified and shortened.
The fact that causes may be endlessly complex gives one
of our two relations—that of causal interaction—still another peculiarity, which is very important. The relatively simple relations between two or more things that we are accustomed to pick out and call relations of cause and effect are not independent; for the sequence of what we call the effect upon what we call the cause' is at the mercy of other causal relations. If two things are similar their similarity is neither increased nor diminished by their likeness to other things or by the likeness of other things to each other. John does not look any the less like James because he looks like Henry also, or because Henry looks like Thomas, In the same way, if the line AB meets the line AC at an angle of 45 degrees you can draw as many more lines as you like to the point A from as many different directions as you like, without affecting the size of the angle in the slightest.
As a result of this we are justified in absolutely ignoring the existence of Henry when we are discussing the resemblance of John and James, and in ignoring all the other lines which meet at A when discussing the relations of AB and AC.
This possibility of considering two relations regardless of everything else in the world is what makes the problems of deduction and of geometry so relatively simple. But with causation all this is changed. John can beat James in a fight if they are left alone, but if Thomas warns James to run or takes part too he cannot; the sun will keep a comet in a certain path if they are left alone, but if Jupiter happens to come too near, the comet may swing out of its path; and
Whether A is similar to B is a mere question between A and B; but whether A will cause B to act in a certain way is a question that also involves C and D and E; and when we are trying to find out what A will make B do we must know whether C and D and E are present and how they are acting. We cannot ignore them. If they are not affecting the relations of A and B we must make sure that they are not; and this, of course, makes our problem much more complex. Deduction and geometry can neglect irrelevant