On the importance of induction for knowledge

Ignaz Semmelweis, 1860

Ignaz Semmelweis worked as a doctor at the Vienna General Hospital between 1844 and 1848. He observed that a large proportion of the women in the First Maternity Division who were delivered of their babies contracted a serious and often fatal illness known as puerperal or childbed fever. In the adjacent Second Maternity Division of the same hospital, which accommodated almost as many women as the First, the death toll from childbed fever was much lower.

Semmelweis began his efforts to resolve the problem by considering various explanations that were current at the time; some of these he rejected as incompatible with well-established facts; others he subjected to specific tests.

One widely accepted view attributed puerperal fever to ‘epidemic influences’ spreading over districts. But Semmelweis questioned how such influences could have affected the First Division for years and not the Second. At the same time, there was hardly a case in the city of Vienna or in its surroundings. Semmelweis noted that women who had given birth in the street on their way to the hospital had a lower death rate than the average for the First Division.

Semmelweis noted that overcrowding could not have been the cause since it was a greater problem in the Second Division, partly because of the efforts of women to avoid the First Division. He also considered the diet and general care of the patients in the two Divisions and found there were no differences.

A commission found in 1846 that rough examination by the medical students was the cause, but Semmelweis rejected this view because the injuries resulting naturally from the birth process are much more extensive, and the examinations by the midwives in the Second Division, though done in much the same manner, didn’t have the same dangerous results. Even after the number of medical students was halved and their examinations of the women were reduced to a minimum, the mortality eventually rose to levels higher than before.

Psychological explanations were attempted (the visits of a priest to deliver the last sacraments to dying women was thought to have had a bad effect on the patients). Semmelweis tested this idea and found that the mortality in the First Division did not decrease. He altered the womens’ position of delivery for birth but again found no alteration in the mortality.

In 1847 a fatal accident suffered by a colleague – a cut on a finger from a scalpel used in an autopsy, resulting in the same symptoms he had observed in the victims of puerperal fever – gave Semmelweis the clue he needed. He realised that ‘cadaveric matter’, introduced by the scalpel, had caused the doctor’s illness and death. He realised that the doctors and students had been the carriers of the infection, moving directly to the wards after performing autopsies, having only superficially washed their hands.

Semmelweis again tested his idea by reasoning that if he were correct, his instruction that all medical students should wash their hands in a solution of chlorinated lime before making an examination, could prevent the fever. The mortality from the fever immediately began to decrease.

In further support of his hypothesis, Semmelweis noted that it accounted for the fact that the mortality in the Second Division was consistently so much lower: the midwives there did not engage in anatomical dissection. The hypothesis also explained the lower mortality among the ‘street births’: these women were rarely examined after admission. Again, the hypothesis accounted for the fact that the victims of the fever among the newborn babies were all among those whose mothers had contracted the disease during labor, when the infection could be transmitted to the baby before birth.

Semmelweis broadened his hypothesis by first sterilising his hands, then examining a woman suffering from cervical cancer, and then proceeding to examine a number of other women in the same room after only routine washing. Nearly all of them died of puerperal fever. He concluded that the fever can be also caused by ‘putrid matter derived from living organisms’.

Logical deduction usually moves from the general to the particular and requires that if the premises are true, the conclusion must be true and that to accept the premises and to deny the conclusion is to contradict oneself. It depends on a priori reasoning. Its aim is to produce certain truth.

Induction (the general method of science) usually moves from the particular to the general (from the observed to the unobserved, from the past to the future or from partial experience to claims about general experience). Even if the premises are true, the conclusion might be false, and there is no contradiction in accepting the premises and denying the conclusion. It is based on a posteriori reasoning and is the method by which a general law or principle is inferred from observed particular instances. The problem with inductive argument is that the conclusion is not guaranteed truth (the Absolute Truth beloved of metaphysicians), even though all one’s observations may all be correct.

In order to find the cause of the fever, Semmelweis collected data, analysed it and formed and tested various hypotheses on that basis. These hypotheses were thought up to account for observed facts. Further, they gave no deductively conclusive evidence, but only more or less strong confirmation for their applicability. Semmelweis did not restrict his study to events at his hospital but to better understand these events, took into consideration births before the mothers arrived at the hospital and the incidence of puerperal fever outside the hospital. Because he proceeded inductively, Semmelweis was able to retain, reshape or discard premises in the light of incorrect conclusions. Even when he had identified the cause of the fever’s spread in the hospital, and did a further testing of his hypothesis using ‘putrid matter derived from living organisms’, not all of the women subjected to it died.

In cognition, induction and deduction are not self-sufficient methods, but are closely interconnected and interdependent.

C. Hempel. “Scientific Inquiry: Invention and Test” in Philosophy of Natural Science, Prentice-Hall, 1966

*   *   *

As Lenin wrote:

‘From living perception to abstract thought, and from this to practice, – such is the dialectical path of the cognition of truth, of the cognition of objective reality. Kant disparages knowledge in order to make way for faith: Hegel exalts knowledge, asserting that knowledge is knowledge of God. The materialist exalts the knowledge of matter, of nature…’

V.I.Lenin, Collected Works, Vol., 38 (Philosophical Notebooks), Progress Publishers, Moscow, 1976, 171.


David Hume, ‘The problem of induction’ and the workingman’s sausage


David Hume is credited with developing induction. He argued that our inductive generalisations rest on the assumption that unobserved cases will follow the patterns that we discovered in observed cases but that one cannot make induction deductively certain. We cannot know that nature is uniform.

Hume distinguished between relations of ideas and matters of fact. He regarded deductive reasoning as demonstrative (involving certainty, established through relations of ideas) and inductive reasoning as probable (involving matters of fact, known by experience).

Hume denied the objective character of causality, arguing that there is no necessary relationship between cause and effect.

He insisted that philosophy cannot go beyond experience and that there is no solution to the ‘problem of induction’ which he set as a syllogism. He asked how we can move from a first premise that all observed A’s have been B to the conclusion that all A’s without restriction have been, are, and will be B.

He thought it can be addressed from two perspectives – that of redundancy (‘All A’s have been observed’), which is disqualified because it cannot be an argument from experience, and that of ‘the principle of the uniformity of nature’.

He also ruled this out because it could only be known to be true by a question-begging appeal to arguments of the very kind here in question (Nature has always been uniform. How does one know? Because it has always been uniform.). Further, such a premise would have to imply that all the A’s experienced by anyone constitute in all respects a perfectly representative sample of A’s.

Hume’s ‘moral’ conclusion was that argument from experience must be without rational foundation.

‘He seems nevertheless to have felt few scruples over the apparent inconsistency of going on to insist, first, that such argument is grounded in the deepest instincts of our nature, and, second, that the rational man everywhere proportions his belief to the evidence – evidence which in practice crucially includes that outcome of procedures alleged earlier to be without rational foundation…Argument from experience should be thought of not as an irreparably fallacious attempt to deduce conclusions necessarily wider than available premises can contain, but rather as a matter of following a tentative and self-correcting rule, a rule that is part of the very paradigm of inquiring rationality – that one would think that other A’s have been and will be the same, until and unless a particular reason is discovered to revise these expectations.’1

Lenin wrote that the sophism of idealist philosophy is that it regards sensation as being not the connection between consciousness and the external world, but a barrier between the two, not an image of the external phenomenon, but the sole entity.2

Whence arises the relations of ideas in deduction – from other ideas, or from within a form of matter which is utterly part of the world?

The (theoretically) absolute truth in nature is approached (the deepening of truth, inseparable from change and uncertainty) through a compound of relative truths by a process of sensory experience, brain processing of that experience and the testing of the resulting ideas in practice.

And this debate, which ultimately traces to that concerning the precedence of matter (objective reality) or consciousness as thought over the other, is directly related to ideology.

‘The workingman who eats sausage and receives a hundred sous a day knows very well that he is robbed by the employer and is nourished by pork meat, that the employer is a robber and that the sausage is pleasant to the taste and nourishing to the body. Not at all, say the bourgeois sophists, whether they are called Pyrrho, Hume or Kant. His opinion is personal, an entirely subjective opinion; he might with equal reason maintain that the employer is his benefactor and that the sausage consists of chopped leather, for he cannot know things-in-themselves.

The question is not properly put, that is the whole trouble…In order to know an object, man must first verify whether his senses deceive him or not…The chemists have gone deeper – they have penetrated into bodies, they have analysed them, decomposed them into their elements, and then performed the reverse procedure, they have recomposed them from their elements. And from the moment that man is able to produce things for his own use from these elements, he may, as Engels says, assert that he knows the things-in-themselves. The God of the Christians, if he existed and if he had created the world, could do no more.’3


1. In A. Flew, Ed.,  A  Dictionary of Philosophy, London: Pan, 1984, 172

2. V.I.Lenin, Materialism and Empirio-Criticism, Critical Comments on a Reactionary Philosophy, 1908, Progress, Moscow 1977, 38

3. Ibid., 185-186; Paul Lafargue, “Le matérialisme de Marx et l’idéalisme de Kant”, Le Socialiste, February 25, 1900.