David Hume on the Problem of Induction
An overview from An Enquiry Concerning Human Understanding
Introduction to Hume
David Hume (1711–1776) was a Scottish philosopher best known for his skepticism and empiricism. He thought of philosophy as being an experimental and inductive science of human nature. He was influenced by empiricists like John Locke and George Berkeley. He made significant contributions to the development of theories such as logical positivism and utilitarianism and to the fields of theology, cognitive science, analytic philosophy, and philosophy of science. Two of his major works are A Treatise of Human Nature (1740) and An Enquiry Concerning Human Understanding (1758).
Problem of Induction
Hume questioned — how can the past or the present be a guide for the future?
Hume was a skeptic. He said that there is no way for us to ‘standardize’ things that occur in nature. Bringing into question the legibility of all sciences, Hume talked about the importance of keeping a neutral mind and not letting presuppositions derived from past instances be a guide for predicting future instances.
The scientific method is synonymous with the inductive method. Even though scientific principles and laws imply universality, they cannot be called cartesian because there is no assurance (empirical, epistemological, or logical) that they are infallible. Because if there is a gap between the premises and the conclusion (and there is, as induction provides only probability, not certainty), then how can it be believed?
The success of past predictions in no way guarantees the success of future predictions. There is no means to achieve logical certainty about the future.
What is Induction?
The process of reasoning in which it is inferred that if something holds true in all observed cases, then it can be inferred to hold in all cases (universally). Here, experiences from the past and present are used to come to generalizations that are then expected to occur in the future.
The Problem…
Science claims to be the finder of general laws that are applicable temporally and spatially. But it is evident that scientists have not (and cannot) tested every incident that the particular laws claim to predict (as being universal) before forming a general law.
The Problem of Induction is the philosophical issue of determining the truth of scientific laws and (what is accepted as) knowledge from a finite number of observed instances.
For example — All kids that I know enjoy playing with Lego. Therefore, all kids enjoy playing with Lego.
But what justification can there be for making such an inference? How can philosophers of science use the inductive method to determine laws of science when there is an undeniable gap between the premises and the conclusion?
Induction is not logically tenable. Hume pointed out that there is no rational basis for the justification for the inductive method, since its claims are, at best, highly probable, but never certain.
Why, then, do we rely on Induction?
Hume, as a behavioral philosopher, pointed out that we are creatures of habit. We expect repetition. He states that the idea of a ‘necessary’ connection comes to us when the repeatedly observe their occurrence together, or in a similar manner. But this repetition of instances cannot validate the idea of a necessary connection or cause, no matter how significant the observed repetitions may be. Multiplication is, he claimed, logically superfluous, as repetition does not involve rational thinking.
Induction is psychologically natural to us. Even though it has no rational basis, it is used by us in all our matters. We are conditioned this way, and we are habitual in associating ideas with repetition. On the loss of this mechanism (of association of ideas), our survival would become difficult, claimed Hume. It is these conditions of the problem that demand the impossibility of any solution.
Aristotle and Bacon on Deduction/Induction
Aristotle was perhaps the first person who attempted to explain the meaning of deduction and induction in his work — The Organon (the tool. with ‘tool’ referring to ‘deduction’). He used the deductive method for understanding nature. He called Socrates the ‘originator’ of induction. He claimed that deduction is the only method that can be used to obtain knowledge as deduction alone provides us with certainty, whereas induction only suggests probabilities. Induction, for him, was a logical method, described as “a progression from particulars to a universal.”
Francis Bacon, in his work — The Novum Organon (the new tool, with ‘new tool’ referring to ‘induction’) pointed out the importance of induction as induction (and not deduction) can lead to the attainment of knowledge. In fact, Bacon was the one who introduced the inductive method as the scientific method. He put together observation, induction, and experimentation, and thus pointed out a new way of attaining knowledge.
Empiricist Influence
Descartes discarded sense perception as a valid source of knowledge, as he claimed that logical certainty could not be ascribed to it. John Locke, often considered to be the first empiricist, in contrast to Descartes, believed that knowledge can be achieved solely through sense perception, but that this (present) knowledge is contingent upon past experiences.
It should be noted that deduction has a place in empiricism when it relates to fixed, albeit abstract, mathematical concepts. Deduction, however, is incapable of leading to any (new) addition to knowledge.
In the title of his work An Enquiry Concerning Human Understanding, Hume uses the word ‘enquiry’ instead of ‘knowledge’ because of the lack of certainty in categorizing any given thing as ‘knowledge’, as explained in his work.
Objects of Enquiry
As per Hume, there are two kinds of objects in the world — one, as existing in the mind, and two, as existing in the external world.
i) Objects in Mind: These consist of ideas and relations of ideas and are derived from impressions. They are certain and incontingent and hence cartesian, for example-mathematical ideas. Demonstrable scientific propositions that can be known intuitively (since they are a priori, they can be known from the mere operation of thought) fall into this category. Here, experience is not a requirement or a factor. Denying them would lead to a contradiction. For example — The sum of the internal angles of a triangle is 180 degrees. (The opposite of this is contradictory)
ii) Objects in the External World: These consist of matters of fact that are dependent on universal principles and are obtained by observation, experience and cause-effect relationship. They are uncertain and contingent and hence cannot be known in the cartesian sense. The natural sciences are an example of this type. The contrary of matters of fact is conceivable and would not lead to a contradiction. For example — The sun will rise tomorrow. (The opposite — The sun will not rise tomorrow, is conceivable)
Objects in the mind are a priori, and objects in the external world (which have been claimed to be discoverable by induction) can never be known for certain.
Hume argued that the conclusions reached by the inductive approach are matters of fact. By the means of inductive inference, the conclusion produced pertains to what remains unobserved. So, such conclusions extend our ‘knowledge’ beyond what has been observed by senses.
On Cause and Effect
Hume said that the constant conjunction of p and q does not necessarily imply a causal connection between them. There is a misassociation between cause and effect in this sense. Here, he emphasizes the difference between (p.q) and (p⊃q), as explained in terms of formal logic. In opposition to what Kant claimed, Hume equated all necessary conditions with a-priori propositions.
Conclusion
The problem of induction has a lot of philosophical significance and many philosophers since Hume have attempted to provide a solution to it. But the majority agree that the problem might have no solution, and some even question its significance. They either question the basis of the argument or claim that no significant skeptical conclusion is laid by it. However, there are those who believe this problem poses a significant threat to the conception of knowledge.
“(If Hume’s Problem of Induction cannot be solved, then)..there is no difference between sanity and insanity.” — Bertrand Russell
Justification for the applicability of the inductive method in science by invoking the ‘universality of nature’ or ‘universal principles’ fails as it leads to the fallacy of begging the question (‘it is the way it has always been’ is a circular justification).
An important point to note is that Hume did not ask us to do away with induction. He acknowledged the viability of its use in day-to-day life. He merely pointed out how it is not the best scientific method because of it cannot deliver certainty.
You can also read Response to Hume’s Problem of Induction.
