This article was published on the Jan 07 edition of Philosophy Today, newsletter of the University of London's Society for Applied Philosophy.

Abstract

The ‘Problem Of Induction’ has been a recurrent theme in epistemological discussions since its popularizing by David Hume in the 18th century. This article seeks to demonstrate that the alleged ‘Problem’ is more of a terminological flaw than a justified criticism. This is achieved by averring that objections to Inductive arguments can only be gained through contrary experiences which, themselves, present alternate Inductive arguments. If Induction is refuted via Induction, the criticism is rendered irrelevant.

1. The ‘Problem’ Defined

Like the background musician in a singing concert or the referee in a sports event, Induction has oft been accorded less credit for its importance in the realm of epistemology. Philosophical circles highlight its reputed flaw, coining it the ‘Problem Of Induction’ as a reflection of their sentiments on its inherent qualities.

The ‘Problem’ Of Induction has been a recurrent theme in epistemological discussions since its popularizing by David Hume in the 18th century. A typical Inductive argument takes the following form:

All ‘A’s have led to ‘B’s under all observable conditions in the past.
> All ‘A’s will lead to ‘B’s in the future.

According to Hume, we should cast doubt over any argument of this structure because, in his words,

“… the knowledge of this relation is not, in any instance, attained by reasonings a priori; but arises entirely from experience…It is impossible, therefore, that any arguments from experience can prove this resemblance of the past to the future; since all these arguments are founded on the supposition of that resemblance.”

In the shadow of such profoundness I shall attempt to affirm Induction’s role in our perpetual quest for certain knowledge, and demonstrate that the alleged ‘Problem’ Of Induction is an unjustified criticism of the Inductive method.

2. The Cause For Doubt

The bulk of criticisms is attributable to the allowance for an exception- even in the most seemingly certain of scenarios- in the spirit of scepticism. We can be exceedingly sure that the sun will rise tomorrow, yet we cannot rule out the minuscule possibility that it does not.

A common vignette highlighting the Problem Of Induction is Bertrand Russell's Inductivist Turkey. This parable draws a parallel between Induction and how a turkey, fattened by its owner, is led to believe its life of prosperity will persist- only to be slaughtered at Christmas. Yet, how can the turkey know that it faces a gruesome fate, based on its uniform experience of being well-fed?

The commonality with both examples is how events can happen against observer extrapolation. This is justification for discrediting Induction’s ability to foresee future events. However, we must question how the sceptic justifies his objection. In truth, antitheses of Induction themselves are grounded on observed exceptional events in the past.

In the argument for the sun’s not rising, scientists have witnessed how stars die or planets stray from their orbits—the laws of physics and previous observations with other stars validate a possibility where our sun does not rise. In fact, new discoveries in quantum theory have shown that even the laws of physics do not stay constant in all space and time. As for the Inductivist Turkey, there have existed previous cases of turkeys slaughtered for Christmas- that is why we entertain the possibility that our well-fed protagonist is destined for the knife.

Such exceptions are conceivable because in the past, similar odds-defying events have occurred, proving the possibility of their re-occurrence in future. Experience tells us such atypical events can happen, though with minute probability. In short, an Inductive argument has been used to justify the ‘Problem’ Of Induction. In other words, examples that serve to illustrate the flaw of Induction are conjured through Induction itself!

3. And Then Some

Can objections to Inductive arguments be raised from reasons other than contrarian experiences? Let us entertain the prospect of using mathematics to object. We begin by assuming in this universe, stars never extinguish, nor planets stop rotating. First we assert that it is never certain if our sun will continue rising forever, simply because it has always risen before. Any Inductive argument predicting the continuity of the sun’s rising is assumed flawed, based on Hume’s theory, though it is contrary to empirical evidence. Such an assertion can only be conceived through mathematics—projecting complex cosmological patterns to predict the sun’s not rising despite the lack of observed cases.

However, abstract mathematics differs from applied mathematics. While the former can deduce a multitude of possible outcomes without the need for empirical anecdotes, only the latter possesses practical utility for actual phenomena. When using math to project cosmological patterns, scientists must first understand how the atoms of the sun and Earth behave. Such information can only be derived, once again, from empirical sources. In fact, Hume himself believes that only experience, not logical deduction, can teach us about the behavior of our surrounding environment. Consequently in our example, scientists are making unspoken Inductions that the laws of physics will stay constant, in order to invoke mathematics against the method of Induction itself.

And what of the Inductivist Turkey? This rather convincing tale overlooks the fact that the turkey has had no prior knowledge that fattening diets lead to the slaughterhouse, and we can assume it has never seen such incidents happen to other turkeys. If the turkey had possessed prior knowledge of why it was fed well, it would never have assumed its continued well-being. Conversely, we can predict its eventual fate through our experiences from past Christmases, and make a valid, Inductive objection to the turkey’s own Inductivist assumption.

A final, albeit flippant, foray to discredit Induction is mustered thus- A sceptic may posit that we cannot be indubitably certain that ‘A’ will always be ‘B’- despite the lack of supporting evidence- simply for the sake of scepticism. However, sceptical criticism is never targeted directly at Induction, but at arguments and statements in general. Even those as elementary as ‘I am a mortal’ are subject to a sceptic’s attack. The very nature of a sceptical standpoint casts minuscule uncertainty over statements of all natures, not just Inductive ones. So it seems the sceptic’s assail is extraneous in our debate on the merits of Induction.

Even if we engage the sceptic at this game we may yet gain the upper hand. This is achieved by posing the rejoinder ‘For what reason do you posit that my knowledge of the future is not indubitably certain?’ The sceptic can only respond by citing Inductivist errors derived from to previous situations and cases. Only experience has demonstrated the flaws of Induction in those other cases, as events contrary to their Inductive assertions have occurred and contributed to the experience. Without experience, according to Hume himself, we can never know that ‘A’ leads to ‘B’; similarly only experience teaches us that sometimes ‘A’ does not lead to ‘B’.

Our banter with the sceptic has consequently been bounded into our original argument in this article. If the sceptic asserts that Induction is unjustified and this scepticism can only be derived Inductively, the criticism translates to: ‘It is unjustified to believe that Induction is unjustified’- meaning that Induction is justified!

4. In A Nutshell

Therefore, we realize that the alleged ‘Problem Of Induction’ suffers from a terminological flaw. The ‘Problem’ should not be perceived as such, but rather, a neutral quality. Many have regarded Induction’s quality in bad light, and this can chiefly be attributed to judging Induction with the wrong standards. Nobody expresses how a tree is a ‘poor animal’ because the tree cannot walk, since ‘walking’ is an animal quality and not a plant quality. Yet both life-forms are dependent on each other to complete the ecosystem.

Ditto for Induction vis-à-vis Deduction. We regard Induction a ‘Problem’ because we judge it using Deduction’s criteria of guaranteeing future knowledge, which is not an inherent quality of Induction. Moreover, there have even been times when even deductive statements were proven wrong:

The ground beneath and around me is flat.
> The Earth is flat.

If even Deductions can be disproved, are we warranted to demand guaranteed knowledge from Inductive statements?

5. Final Words

Conventionally, persons who seek knowledge use Induction to validate their initial hypotheses. When astronomers aspired to discover new planetary systems, they constructed an Inductive argument for their existence, using our sun as an example. Discoveries of planets around other stars then served to confirm the Inductive premise. It was with this assumption that astronomers combed the vicinity of nearby stars for the faintest traces of orbiting bodies, and found them.

Surely, Induction is how we seek new knowledge. It is firmly embedded in the way scientists and ordinary people think, and has emerged as a pragmatic method of acquiring discoveries. Used in conjunction with Deduction, it forges a partnership to fill inherently different but equally vital roles. Induction’s role in epistemology has proven invaluable, and perhaps even irreplaceable in times forthcoming.

Bibliography

Hume, D. (EHU) (1777). An Enquiry concerning Human Understanding (3rd. ed.), edited by Nidditch, P. N. (1975). Clarendon Press.

Fogelin, R. J. (1993). Hume’s scepticism. In The Cambridge Companion to Hume, edited by Norton, D. F. (1993), 90-116. Cambridge University Press.

B. Russell. (1957) The Problems of Philosophy. Williams and Nogate.