In 1934, Popper, then living in Vienna, published his book The Logic of Scientific Discovery. In this work, Popper completely rejects the method of induction, insisting that all conclusions must be drawn from logical deduction. Popper specifically rules out the method of induction, based on observation. To qualify for Popper’s certificate of "science-worthiness," a theory must be internally consistent, must not be a tautology, and must make predictions that can be tested. Moreover, he maintained that the results of a test cannot verify a theory, only falsify it.
All of this sounds very nice, and is in complete accord with the method of formal logic. But it has got very little to do with the actual practice of science. One physicist commented wryly that Popper’s ideas were strategically sound but tactically indefensible, in other words, fine in (formal logical) theory, but, like an umbrella full of holes—useless precisely for the purpose for which it was intended.
Induction (from the Latin inducere, to lead in) is another method of reasoning. It was already known to Aristotle, but achieved wide acceptance during the Renaissance, when it was championed by Bacon and Galileo. As a form of reasoning, induction proceeds from single facts to general propositions. Men and women have always made such generalisations on the basis of their experience, often reaching correct conclusions, sometimes not.
Let us consider an example of inductive reasoning. A child burns its hand on a flame, and draws the conclusion, on the basis of experience that it is not a good idea to get too close to fire. "Fire (in general) burns." That is an inductive reasoning—from the particular to the general. In this case, the conclusion is perfectly valid and rather useful. But consider another example. A turkey is visited every morning by a nice old lady with a bag of corn in her hand. The turkey, by the method of inductive reasoning, might very well conclude that the kind lady means food. This conclusion is drawn from the same experience repeated many times—364 times, to be exact. Then, one morning, the farmer’s wife appears with a butcher’s knife in her hand. Here the turkey’s inductive logic proves to be somewhat defective, and does not really help it to clarify its existential dilemma!
Scientific induction, like its popular equivalent, also consists of drawing conclusions from a whole class based on the number of elements of that class. But here the grounds for conclusion are provided by the discovery of essential connections between the elements studied, which show that the given feature must be possessed by the whole class. The task of discovering these necessary connections involves detailed observation. Thus, induction signifies experimental study of things, in such a way that we pass from single facts to generalisations.
The method of deduction is, on the face of it, the exact opposite of induction. Deduction consists of proving or inferring a conclusion from one or more premises by the laws of logic. The deductive method does not set out from particular experiences, but from so-called axioms, which are assumed to be correct from the start. This is the traditional method of mathematics, for example classical geometry, based on the axioms of Euclid, which were for centuries supposed to represent absolute truths, valid for all time, under all circumstances. Deductive reasoning therefore proceeds from the general (law) to the particular.
The struggle between induction and deduction goes back to the 17th century, to the different approaches adopted by two great scientific thinkers—Bacon and Descartes. The Englishman Bacon was the father of empiricism, and the method of inductive reasoning, which attempts to derive theories from observed facts alone. In Bacon’s case, the obsession with observation proved fatal; he died of bronchitis as a result of an early experiment in refrigeration, involving stuffing a chicken with snow.
Descartes approached science from a diametrically opposite standpoint. Taking Euclid’s geometry as his model, he attempted to develop consistent and coherent theorems derived from pure reason, without recourse to the unreliable evidence of the senses. His method was that of rationalism, which became the main tradition in France. Bacon’s empiricism triumphed on the other side of the Channel. Both men, in different ways, advanced the cause of science, and both made important discoveries.
However, neither deduction nor induction on their own are capable of grasping the whole picture. The problem with Bacon’s method is that the facts do not select themselves. You need an initial theory (a hypothesis) even to decide what observations to make in the first place. Moreover, the results of induction always have a more or less provisional character. For example, a person who had observed a hundred swans might draw the conclusion that all swans were white. This is an inductive conclusion. But it would be wrong, because some swans are black. Engels makes the point that "The empiricism of observation alone can never adequately prove necessity." (The Dialectics of Nature, p. 304.)
We therefore did not have to wait for Sir Karl to point out the limitations of inductive logic. However, to deny induction altogether is to jump from the frying pan into the fire. Induction plays a necessary role in science, as well as in everyday life. Is it really necessary for somebody to drink all the water in the sea before being prepared to admit that sea water is salty? Popper’s attempt to eliminate induction from science shows a lamentable ignorance both of the true relationship between deduction and induction, and of how science works in real life.
Until the end of the 19th century, the deductive method was used almost exclusively in mathematics. Not until the 20th century were attempts made to apply it to fields such as physics, biology, linguistics, sociology, etc. Despite all the impressive claims made on its behalf, experience shows that the axiomatic-deductive method is quite limited in what it can achieve. The controversy between induction and deduction is pointless, since, in practice, induction always exists together with deduction. Neither is self-sufficient as a method, but, in dialectical materialism, they are combined as different aspects of the process of cognizing reality, which are inseparably connected, and determine each other.
The Economist article already mentioned goes on to criticise Popper’s rejection of the inductive method:
"A number of philosophers also question Popper’s rejection of induction. The use of induction, they say, is logically unsatisfactory but inescapable. Deductions about the real world are only as good as the assumptions about the real world on which they are based. These assumptions rest on induction, as does the scientist’s interpretation of the experimental results that test the conclusions drawn from them. Both in forming a hypothesis and in interpreting tests of it, a scientist makes the basic assumption that nature will behave in other places and at other times as it behaves here and now. That is an inductive assumption." And it continues:
"Dr. Jennifer Trusted is one British philosopher who puts induction in perspective. Induction, she says, is essential but not sufficient for knowledge of the real world. The same could be said for deduction."
This last observation is absolutely correct, and goes to the heart of the matter. Neither induction nor deduction, taken on its own, is sufficient. It is necessary to combine them, which is just what dialectics does. Deduction is also a conclusion, and therefore induction is also a kind of deduction. On the other hand, all deductions are, in the last analysis, derived from material reality. This is true even of axioms, which are supposed to be the products of "pure theory." For example, Euclid’s axiom that a straight line is the shortest distance between two points is clearly the result of long experience and observation. Engels explains the one-sidedness of both induction and deduction, when taken in isolation, and also explains the dialectical relation between them:
"Induction and deduction belong together as necessarily as synthesis and analysis. Instead of one-sidedly lauding one to the skies at the expense of the other, we should seek to apply each of them in its place, and that can only be done by bearing in mind that they belong together, that they supplement each other." (The Dialectics of Nature, p. 302.)
Yorum Gönder