What Is the Problem of Induction? Can It Be Solved?

...

Step 1: Form a hypothesis

Step 2: Derive predictions from the hypothesis

Step 3: Test your predictions via experimentation

Indeed for Popper science was based on conjecture and refutation, in that we posit a theory and test our predictions, if the “predicted effect is definitely absent, then the theory is simply refuted”[9] and we form a new hypothesis. However Popper argued that positing a theory with a high probability was actually contrary to the goals of science and ineffectual. Rather a highly falsifiable theory is preferable and if it withstands sustained criticism, that theory is arguably better corroborated than others, but it is in no way supported or verifiable. This is because there is an infinite amount of conjectures that one can make, and there is also an infinite number of ways in which these conjectures can be refuted via experimentation. For Popper therefore science does not rely on induction at all, because “the criterion of scientific status of a theory is its falsifiability,”[10] which relies solely on deduction. Popper’s point therefore is that the problem of induction it is not an issue that concerns science, as science relies on conjecture and refutation, which in turn undermines the scope of the problem of induction.

Another response to the problem of induction comes from Peter Strawson who offers a semantic solution, which argues that induction is rational by definition. For Strawson the idea of providing a justification for the rationality of induction was simply incoherent, and was like asking whether a triangle has three sides. Indeed Strawson argued that we need not provide a justification for the rationality of induction, because “it is an analytic proposition… to have a degree of belief in a statement which is proportional to the strength of the evidence.”[11] Therefore when we say that a scientific law or rule is ‘rational,’ we are saying that it draws conclusions based on a certain number of repeated observations. Indeed for Strawson therefore the idea that induction is rational was an analytic truth, because it is true by definition. For someone to question the rationality of induction would be like asking “is the law legal,”[12] as ‘legal’ means ‘being in accordance with the law,’ and it is thus absurd to question the general legality of the law. As a result Strawson argued that there is no problem of induction as induction is rational by definition, and to question its rationality is simply incoherent.

Moreover another response to the problem of induction comes from Hume himself, who argues that while no justification can be provided for the rationality of induction, it cannot be proved to be irrational either. Hume argued it was clear that no satisfactory justification for the reliability of induction could be provided, but for Hume this did not mean that we should therefore do away with induction. This is because arguably the “inductive principle is at any rate not capable of being disproved”[13] by a positive argument. While induction is arguably no more logical than say astrology, it is not an irrational practice. Indeed induction as a principle is not incoherent or contrary to reason as no argument can be put forward, which actually proves the irrationality of induction. Moreover Hume argued that our belief in the principle of induction does not stem from reasoned argumentation, but rather from “the psychological mechanism of custom.”[14] Indeed for Hume we have come to adopt inductive inference as a principle because our experience of causation shows a correlation of one event following another. For Hume our experience of this pattern leads us instinctively to accept the principle of induction, as it “is the psychological basis of all forms of inductive inference.”[15] Therefore while Hume agreed that our belief in the uniformity of nature was not rational, he also argued that we adopt the principle out of habit, and because there is no positive argument against induction, we have no compelling reason to reject it.

However there are a number of issues with various responses to the problem of induction, one of which is that all the responses ultimately fail to actually solve the problem. The problem of induction lies in the fact that it cannot seemingly provide any satisfactory justification for the uniformity of nature, and therefore the reliability of induction. However all the responses to the problem fundamentally fail to provide a deductive or a priori justification for induction, and therefore fail to address the crux of the issue. For example Popper’s solution that science is based on conjecture and refutation is problematic on two grounds. Firstly Popper’s solution suggests that no scientific theory no matter how ‘corroborated,’ can ever be proved to be true or justifiable. Indeed Popper’s solution does not only fail to explain why corroborated theories are preferable and more reliable than a falsified one, but also reduces all science and its predictions to mere speculation. Secondly Popper’s solution only serves to save science from the problem of induction, as it does not address the inductive inferences we make in our daily lives, nor does it actually attempt to directly solve the problem.

Furthermore Strawson’s semantic solution to the problem is arguably nothing more than a technicality in our language that is inadequate in actually justifying induction. Indeed Strawson attempts to justify induction on the grounds that it is rational by definition and therefore an analytic truth. However this claim is obviously problematic, as we can easily imagine a world where reading tea leaves is seen as a rational practice. Indeed the people of that world would argue that reading tea leaves is a perfectly rational practice, and therefore to say ‘reading tea leaves is rational’ would be an analytic truth. However it would be immediately clear to anyone of our world that this practice is not rational by any means, regardless of whether it is an ‘analytic truth.’ Therefore it is clear that Strawson’s semantic solution is not an adequate response to the problem of induction, as just because something is rational by definition does not therefore make it rational, which therefore suggests induction cannot be justified on semantic grounds.

To conclude inductive arguments differ from deductive arguments because the premises of inductive arguments do not necessarily entail the conclusion, but rather suggest support for the conclusion. The problem of induction was raised by David Hume, who argued that we have no deductive justification for believing in the uniformity of nature and therefore the future reliability of induction.