The Falling Dominoes of the Principle of Proof by Mathematical Induction, #1.

(Wikipedia, 2025)

TABLE OF CONTENTS

1. Introduction

2. The Definition of Mathematical Induction

3. Wang’s Paradox and the Predicate “Small”

4. Edward Nelson’s Critique of Mathematical Induction

5. Mathematical Induction Meets Gödel’s Incompleteness Theorems

6. Omega-Consistency and Its Role in Gödel’s Theorems

7. A Foundational Skepticism about Mathematical Induction

8. A Foundational Skepticism about Mathematical Induction

9. Conclusion

This essay will be published in three installments; this one, the first, contains sections 1-3.

But you can also download and read or share a .pdf of the complete text of this essay, including the REFERENCES, by scrolling down to the bottom of this post and clicking on the Download tab.

The Falling Dominoes of the Principle of Proof by Mathematical Induction

1. Introduction

In this essay, we present a skeptical critique of proof by the principle of mathematical induction, as part of a series of articles challenging some fundamental aspects of classical mathematics, in order to illustrate the doctrine of epistemic humility.

The doctrine of epistemic humility, in its most basic form, posits that human knowledge is inherently limited and fallible. It challenges the notion that we can attain rational justification about a wide range of beliefs, emphasizing instead the provisional and often uncertain nature of our understanding. At its core, epistemic humility encourages a cautious and open-minded approach to knowledge acquisition. It suggests that many things we commonly assume to be secure, well-founded, or universally true are, upon closer inspection, far less certain than we imagine. This can apply to scientific theories, historical accounts, ethical principles, or even received mathematical propositions, as to be discussed in this paper.

The doctrine doesn’t necessarily deny the possibility of knowing anything but rather urges us to be aware of the boundaries of our cognitive capacities and the potential for error in our reasoning and observations. It fosters an attitude of intellectual modesty, acknowledging that our perspectives are limited, our evidence is often incomplete, and new information and arguments can always emerge to challenge existing beliefs. In essence, epistemic humility is a call to recognize the vastness of what we don’t know, alongside what we do know.

More specifically, we challenge the epistemic foundations of mathematical induction (MI) by exposing its reliance on ω-consistency, a property unprovable within Peano Arithmetic due to Gödel’s Second Incompleteness Theorem. Then, through a novel reductio ad absurdum and insights from Wang’s paradox and Nelson’s finitism, we argue that mathematical induction’s intuitive “certainty” masks fundamental uncertainties, aligning with the doctrine of epistemic humility. This critique not only questions a cornerstone of classical mathematics but also highlights the need for philosophical reflection on mathematical proof. This critique runs parallel to David Hume’s famous challenge to empirical induction. As an incidental point, if our argument is sound, then it illustrates that philosophy is not a mere “under-laborer” to the formal and natural sciences, but possesses genuine epistemic force, capable of challenging even long-received claims.

2. The Definition of Mathematical Induction

Let P(n) be a statement about natural numbers. Mathematical induction is a method of proof that shows P(n) holds for all natural numbers n, by following two steps:

Base Case: Prove that P(0) (or P(1), depending on the version) is true.

Inductive Step: Assume P(k) is true for an arbitrary natural number k (this is called the inductive hypothesis), and then prove that P(k + 1) is also true.

Conclusion: If both steps are proven, then P(n) is true for all n in ℕ (the natural numbers).

In logical notation: If P(0) is true, and for all k ∈ ℕ, P(k) → P(k + 1), then for all n ∈ ℕ, P(n) is true.

Mathematical induction is justified by the Peano axioms, which define the structure of the natural numbers. These axioms are:

0 is a natural number.

Every natural number n has a successor S(n), which is also a natural number.

0 is not the successor of any natural number.

If S(m) = S(n), then m = n. (Successors are unique.)

(Axiom of Induction)

If a property P holds for 0, and P(k) → P(k + 1) for all k ∈ ℕ, then P(n) holds for all n ∈ ℕ.

This last axiom—the induction axiom—is the foundation of all proofs by induction. It ensures that the natural numbers are the smallest set containing 0 and closed under the successor function.

Example: Sum of the First n Natural Numbers

We want to prove the following proposition:

For all n ∈ ℕ,

1 + 2 + 3 + … + n = n(n + 1) / 2

Proof by Induction

Let P(n) be the statement:

1 + 2 + 3 + … + n = n(n + 1) / 2

1. Base Case (n = 1)

Left-hand side:

Right-hand side:

1(1 + 1) / 2 = 2 / 2 = 1

So P(1) is true.

2. Inductive Step

Assume P(k) is true for some arbitrary k ∈ ℕ. That is,

1 + 2 + 3 + … + k = k(k + 1) / 2 (this is the inductive hypothesis)

We must show that P(k + 1) is also true, i.e.,

1 + 2 + … + k + (k + 1) = (k + 1)(k + 2) / 2

Start with the left-hand side of P(k + 1):

1 + 2 + … + k + (k + 1)

= [k(k + 1) / 2] + (k + 1) (by the inductive hypothesis)

= (k(k + 1) + 2(k + 1)) / 2

= (k + 1)(k + 2) / 2

So P(k + 1) is true.

Conclusion

Since both the base case and the inductive step have been proven, by the principle of mathematical induction,

P(n) is true for all n ∈ ℕ. ■ Q.E.D.

What could possibly be wrong with that?

3. Wang’s Paradox and the Predicate “Small”

Wang’s paradox is a philosophical challenge to the unrestricted use of mathematical induction, particularly when applied to vague predicates, i.e., terms that lack a precise, fixed definition.

Let “small(n)” be a predicate meaning “n is a small natural number.” At first glance, this seems reasonable—after all, 0, 1, 2… all feel “small.”

Now suppose we attempt to prove:

For all n ∈ ℕ, small(n) (i.e., “every natural number is small”)

We try to use mathematical induction:

Base Case:

small(0) — Certainly true. Zero is small.

Inductive Step:

Assume small(k) is true for some arbitrary k ∈ ℕ.

Then it seems plausible that small(k + 1) is also true—after all, if 100 is small, 101 isn’t much bigger.

Conclusion:

Therefore, by induction: small(n) is true for all n ∈ ℕ.

But this reduces to an absurdity: All natural numbers are small.

That includes numbers like 1,000,000 or 10^100, which clearly are not “small” in any ordinary or useful sense.

This contradiction reveals the core problem: the predicate “small” is vague. It lacks a precise, fixed extension—there’s no clear cut-off between “small” and “not small.”

In formal terms, mathematical induction requires that the predicate P(n) must be well-defined and precise. If P(n) is vague, the inductive step becomes invalid, because it assumes a continuity of meaning that doesn’t exist.

Wang’s paradox exposes a weakness, or limit, in applying induction outside of formal logic: Induction works only when the predicate is sharply defined over the domain. For vague predicates, the logic “If P(k), then P(k + 1)” can break down, because “closeness” in everyday meaning doesn’t imply logical entailment. In short: vagueness breaks induction.

Mathematicians often dismiss paradoxes like Wang’s by pointing out that predicates such as “small” or “large” are informal and seldom used rigorously in mathematical proofs. They argue these terms are merely linguistic conveniences, not genuine mathematical concepts (Wang, 1953; Dummett, 1975). However, this easy dismissal overlooks the fact that mathematics contains several other predicates and notions that are inherently vague yet fundamental to its practice. These predicates are not so easily brushed aside, and their vagueness raises deep foundational questions that demand serious scrutiny (Avigad, 2010; Bishop, 1967). Directly below are some key examples of such predicates where the line between precision and ambiguity is blurred, offering possible grounds for mathematical skepticism, but not via a critique of the principle of mathematical induction.

1. Convergence in Analysis

The concept of convergence is central to analysis, traditionally defined precisely via the ε-δ criterion (Rudin, 1976). Yet, intuitively, convergence means a sequence gets “arbitrarily close” to a limit—a notion inherently vague (Bridges and Richman, 1987). In constructive or intuitionistic mathematics, where explicit construction and verification are required, this vagueness becomes problematic. The classical approach depends on infinite processes and idealized limits, raising skepticism about the actual existence or meaningfulness of limits beyond formal definitions.

2. “Almost Everywhere” in Measure Theory

The term “almost everywhere” means that a property holds everywhere except on a set of measure zero (Folland, 1999). Although “measure zero” is a formal notion, ignoring exceptions on such sets can be seen as a kind of vagueness in application. This vagueness matters foundationally because it can mask pathological cases and subtleties (Fremlin, 2000). Skeptics question whether neglecting measure-zero exceptions is always justifiable or whether it sometimes hides significant mathematical complications.

3. “Sufficiently Large” in Number Theory

Number theory frequently uses the phrase “for sufficiently large n, property P holds,” without specifying what “sufficiently large” means in practice (Hardy and Wright, 2008). This lack of explicit quantification allows for flexibility but also introduces ambiguity. In many cases, effective bounds are unknown or uncomputable. For finitist or constructive perspectives that demand explicitness and computability, this predicate’s vagueness is a significant problem (Troelstra and van Dalen, 1988).

4. “Well-Defined” Functions in Abstract Mathematics

Proofs sometimes assert that a function or construction is “well-defined” without fully verifying it. What counts as “well-defined” can depend on context, and in complex mathematical settings, this assumption can conceal logical gaps or ambiguities (Munkres, 2000). This vagueness challenges the “certainty,” or at least epistemic security mathematicians often claim about their constructions.

5. “Randomness” in Probability and Algorithmic Theory

Although probability theory rigorously formalizes randomness, the intuitive notion remains vague. Concepts like algorithmic randomness and Kolmogorov complexity provide formal frameworks, yet the meaning and interpretation of randomness remain philosophically and mathematically subtle (Li and Vitányi, 2008; van Lambalgen, 1996).

Wang’s paradox reveals that mathematical induction’s validity hinges on the assumption that predicates are sharply defined over an infinite domain. Yet, Gödel’s theorems and Nelson’s finitism (to be discussed in the next section) suggest that this infinite domain—the natural numbers as a completed totality—may itself be an epistemically suspect assumption. The vagueness revealed by Wang’s paradox is a rhetorical flourish, a clever artifact of applying induction to an obviously informal predicate like “small,” and that does not threaten the principle of mathematical induction as it is used in rigorous settings. However, this response misses the deeper point. The real concern is not that induction stumbles on sloppy language, but that its validity depends on predicates being sharply defined, and we often assume this without proof. Foundational terms such as “computable,” “well-defined,” or even “effective” are routinely used in formal reasoning, yet they carry conceptual weight that exceeds their formal casings. For example, whether a function is well-defined can depend on contextual judgments that are not always formally articulated. Similarly, the notion of computability presumes agreement on models of computation—for example, the Church-Turing thesis—which is philosophically rather than strictly mathematically grounded. These are not trivial ambiguities but indicators of epistemic vagueness: the kind that operates not at the linguistic level but at the level of foundational assumptions. Mathematical induction’s reliability relies on these assumptions being valid, but in many cases, their status is unprovable within the system itself. Thus, the specter of vagueness haunts even the most apparently precise applications of mathematical induction, not as a semantic glitch, but as an inherent epistemological vulnerability. The problem is not that the predicates are visibly vague, but instead that their justification is uncritically presumed. In this light, Wang’s paradox is not a one-off oddity but a vivid signal of deeper instabilities lurking behind the façade of formal rigor. This issue will be explored in a paper presently under preparation.

Download