Chapter Fourteen: How Big Is Infinity?

Germany and Britain, 1870–1910 CE

For most of human history, infinity was not a mathematical object. It was a warning.

Greek philosophers had argued about it. Theologians had attached it to God. Poets and mystics had used it as a word for what exceeds the mind. Even mathematicians, when they encountered it, tended to handle it cautiously and at arm’s length. Euclid let lines extend indefinitely but did not treat the infinite as a completed thing. Calculus used processes that approached infinity or zero without ever quite arriving. The Kerala mathematicians had worked with infinite series astonishingly early, but always in the service of finite results. Infinity was something one moved toward, not something one laid on the table and counted.

In the late nineteenth century, a German mathematician named Georg Cantor did something that many of his contemporaries considered reckless and some considered obscene. He treated infinite collections as objects in their own right. Then he asked a question so simple that it sounds almost childish: how many things are there in an infinite set?

The answer he found was not “infinitely many” in the vague old sense. It was more precise and much stranger. Some infinite sets, he showed, are larger than others. This was one of the most shocking ideas in the history of mathematics. It did not merely extend arithmetic. It changed what counting means. And because it changed what counting means, it shook the foundations of the subject.

Cantor Did Not Begin With Philosophy

It is tempting, in retrospect, to tell this story as though Cantor sat down determined to conquer infinity directly. He did not. Like many major mathematical revolutions, this one began inside a technical problem. The problem came from trigonometric series.

By the nineteenth century, after Fourier, mathematicians had become intensely interested in representing functions as sums of sines and cosines. A trigonometric series looks like:

a₀ + a₁cos(x) + b₁sin(x) + a₂cos(2x) + b₂sin(2x) + ...

This kind of expression had already proved extraordinarily useful in the mathematics of heat, waves, and periodic motion. But there was a subtle question lurking behind the technique: if a function can be represented by such a series, is that representation unique? That sounds narrow. It was not.

To answer it, Cantor was forced to think carefully about the sets of points at which two supposedly equal trigonometric series might differ, or at which certain limiting behaviours might occur. The issue led him into the study of point sets on the real line, derived sets, accumulation points, and the structure of infinite collections. What began as analysis slowly turned into set theory.

This is worth noticing because it shows, once again, how mathematics develops. Cantor did not wander into infinity because he was temperamentally mystical. He was pushed there by a concrete research problem about functions. The abstraction came later, as it so often does. First the technical obstruction appears. Then someone realises that the obstruction is pointing toward an entirely new landscape.

By the time Cantor had followed that path to the end, he was no longer just studying trigonometric series. He was studying the architecture of the infinite itself.

When Counting Stops Being Obvious

Counting seems, at first, like the safest operation in mathematics.

If you have three apples and two apples, you have five apples. If one shelf holds ten books and another holds twelve, the second shelf holds more. The natural numbers — 1, 2, 3, 4, 5, and so on — arise so early in human life that they feel less like inventions than like reflexes. Mathematics begins here for almost everyone, with the intuition that more objects means a larger number.

That intuition works perfectly well for finite collections. If one set can be paired off exactly with another, item by item, and nothing is left over on either side, then the two sets have the same size. If one set has leftovers after every possible pairing, then it is larger.

The trouble begins when the sets are infinite.

Take the natural numbers:

1, 2, 3, 4, 5, 6, ...

Now take the even numbers:

2, 4, 6, 8, 10, 12, ...

At first glance, the even numbers ought to be fewer. They are only part of the natural numbers. Every even number is a natural number, but not every natural number is even. Surely the whole must be larger than the part.

And yet you can pair them perfectly:

and in general:

n ↔ 2n

Every natural number has exactly one even number paired with it, and every even number has exactly one natural number paired with it. Nothing is left over. By the item-by-item criterion that works perfectly in the finite case, the two sets have the same size.

This is the first point at which infinity stops behaving like a stretched-out version of ordinary arithmetic. In the finite world, a proper part is always smaller than the whole. In the infinite world, a proper part can match the whole exactly.

Galileo noticed this paradox in the seventeenth century and treated it as a sign that ordinary ideas of greater, less, and equal become unstable when the infinite is involved. Cantor took the opposite lesson. If the old language becomes unstable, then a new language is needed.

That language begins with the idea of a one-to-one correspondence.

The Right Definition of Size

Cantor’s fundamental move was to define size, for arbitrary sets, by pairing rather than by counting in the everyday sense.

Two sets have the same cardinality if their elements can be matched one-to-one and onto. In modern language, there exists a bijection between them.

This sounds formal, but the underlying idea is simple. If every object in Set A can be paired with exactly one object in Set B, and every object in Set B gets paired with exactly one object in Set A, then the sets are equally large as collections, whether they are finite or infinite.

For finite sets, this reproduces the ordinary notion of size.

For infinite sets, it produces surprises.

The natural numbers and the even numbers have the same cardinality.

So do the natural numbers and the odd numbers.

More startlingly, as Cantor would show, so do the natural numbers and the rational numbers:

all the fractions of the form

p/q

where p and q are integers and q ≠ 0.

This is harder to believe. Between any two integers there are only finitely many integers, but between any two integers there are infinitely many rational numbers. Between 0 and 1 alone there are:

1/2, 1/3, 2/3, 1/4, 3/4, ...

densely packed without end. The rationals feel vastly more numerous than the naturals. Cantor showed they are not. The lesson is severe. Density and size are not the same thing. A set can be densely packed into every interval and still be countable. Infinite sets require sharper distinctions than intuition first provides.

The Rationals Are Countable

Let us see how Cantor’s argument works, because this is the kind of reasoning that changed mathematics.

Imagine listing the positive rational numbers in a grid. Put the numerator across the top and the denominator down the side:

1/1   2/1   3/1   4/1   ...
1/2   2/2   3/2   4/2   ...
1/3   2/3   3/3   4/3   ...
1/4   2/4   3/4   4/4   ...
...

Every positive rational number appears somewhere in this infinite table.

Now sweep through the table diagonally:

1/1
2/1, 1/2
1/3, 2/2, 3/1
4/1, 3/2, 2/3, 1/4
...

As you move from diagonal to diagonal, you eventually reach every position in the grid. If you skip duplicates such as:

2/2 = 1/1
2/4 = 1/2

and keep only fractions in lowest terms, you obtain a sequence:

r₁, r₂, r₃, r₄, ...

that lists every positive rational number exactly once.

Once such a list exists, the rationals can be paired with the natural numbers:

1 ↔ r₁
2 ↔ r₂
3 ↔ r₃
...

So the rational numbers are countable.

This is an astonishing result the first time one sees it. The rational numbers seem to overflow the number line. Between any two rationals there are infinitely many more rationals. They are everywhere dense. And yet they can all be placed in a single list.

Countable infinity, in Cantor’s sense, does not mean “sparse” or “widely separated.” It means only that the set can be matched with:

1, 2, 3, 4, 5, ...

This is the first infinite cardinality. Cantor denoted it:

ℵ₀

pronounced aleph-null.

So the natural numbers, the even numbers, the odd numbers, and the rational numbers all have cardinality:

ℵ₀

That alone would have been enough to secure Cantor a place in mathematical history. But it was only the beginning.

The Real Numbers Are Not Countable

If the rationals can be listed, what about the real numbers?

By the real numbers we mean the full continuous number line: integers, fractions, irrationals like:

√2, π, e

and all the infinitely many others filling the gaps between them.

The Greeks had discovered that irrationals exist. The nineteenth century, through Dedekind and others, had given the real numbers a rigorous construction. What Cantor asked was whether all real numbers can be listed in sequence the way rationals can. His answer was no. This is the central discovery of set theory. There is more than one infinite cardinality. The infinity of the real numbers is strictly larger than the infinity of the natural numbers.

Cantor’s proof, the diagonal argument, is one of the most beautiful in mathematics.

Suppose, for contradiction, that all real numbers between 0 and 1 can be listed:

r₁ = 0.a₁₁a₁₂a₁₃a₁₄...
r₂ = 0.a₂₁a₂₂a₂₃a₂₄...
r₃ = 0.a₃₁a₃₂a₃₃a₃₄...
r₄ = 0.a₄₁a₄₂a₄₃a₄₄...
...

Here each aᵢⱼ is a decimal digit.

Now build a new number by changing the diagonal digits. Look at:

a₁₁, a₂₂, a₃₃, a₄₄, ...

and define a new decimal:

s = 0.b₁b₂b₃b₄...

where each bₙ is chosen to differ from aₙₙ. For instance, one may say:

if aₙₙ ≠ 5, let bₙ = 5
if aₙₙ = 5, let bₙ = 4

Then the new number s differs from r₁ in the first decimal place, from r₂ in the second decimal place, from r₃ in the third, and so on. So it differs from every number on the list in at least one place.

Therefore s is a real number between 0 and 1 that is not on the supposedly complete list.

Contradiction.

So no such list can exist.

The real numbers are uncountable.

This proof is devastatingly simple. It does not depend on complicated analysis or delicate algebra. It depends only on the logic of listing and the construction of a counterexample that slips past every attempted enumeration.

With that, Cantor had shown that infinity comes in layers.

There are infinitely many natural numbers.

There are infinitely many rational numbers, but no more of them than naturals.

There are infinitely many real numbers, and strictly more of them than naturals.

The continuum is a larger infinity.

Even Algebra Does Not Fill the Line

A real number is called algebraic if it is a solution of some polynomial equation with integer coefficients. Thus:

√2

is algebraic because it satisfies:

x² - 2 = 0

The golden ratio is algebraic. So are all rational numbers. By contrast, numbers like:

π, e

are transcendental: they are not roots of any such polynomial equation.

For a long time, transcendental numbers were rare and shadowy. Liouville had shown in the nineteenth century that they exist, but they still felt exceptional. Algebraic numbers, by contrast, looked substantial and respectable. They are the numbers produced by ordinary polynomial equations, and polynomial equations had driven so much of mathematical history that one might easily imagine they cover most of the important territory.

Cantor showed otherwise. There are only countably many algebraic numbers.

The reason is simple once one sees it. There are only countably many polynomial equations with integer coefficients, because each such equation is determined by a finite string of integers, and the set of all finite strings of integers is countable. Each polynomial has only finitely many roots. A countable collection of finite sets is still countable. Therefore the algebraic numbers are countable.

But the real numbers are uncountable. So most real numbers are not merely irrational. They are not even algebraic. They are transcendental.

This is one of the most dramatic reversals in mathematics. The numbers that arise from the whole classical machinery of algebra — equations, radicals, roots, coefficients, polynomials — form only a countable sliver of the continuum. The typical real number lies beyond algebra entirely.

It is hard to overstate how strange that is. By the late nineteenth century, algebra was one of the subject’s great triumphs. And yet, from the viewpoint of the continuum, algebraic numbers are sparse. The vast majority of reals cannot be reached by solving polynomial equations at all.

The exceptional had become the ordinary.

Most Numbers Are Irrational

One consequence of Cantor’s work is so striking that it changes how one sees the number line.

The rational numbers are countable.

The real numbers are uncountable.

Therefore, the irrational numbers — the real numbers that are not rational — must themselves be uncountable. In fact, they make up “almost all” real numbers in the crude cardinal sense. The rationals, for all their familiarity and calculational usefulness, form only a tiny countable subset of a vastly larger continuum.

This gives precise force to something the Greeks had dimly glimpsed when they discovered √2. The neat numbers are not the norm. They are exceptions.

The number line is not sparsely decorated by a few inconvenient irrationals. It is overwhelmingly irrational. Fractions are scattered through it like a thin countable dust.

This was a conceptual reversal of the kind mathematics occasionally produces. Rational numbers feel manageable because they can be written exactly. Irrationals feel exotic because many of them cannot. But from the point of view of the continuum, it is the rationals that are special and limited.

The familiar numbers are not the typical ones.

That is a disturbing thought, and a fertile one.

Infinity Makes More Infinity

Cantor did not stop with the real numbers.

Once one admits one infinite cardinality and then a larger one, a natural question arises: can this process continue? Yes. In fact it never ends.

Cantor proved a general theorem: for any set, the set of all its subsets has strictly larger cardinality than the set itself.

If a set is called A, its set of all subsets is called the power set:

P(A)

For a finite example, if:

A = {1, 2}

then:

P(A) = {∅, {1}, {2}, {1,2}}

so a 2-element set has a power set with 4 elements.

Cantor showed that this phenomenon persists infinitely. The proof has the same flavour as the diagonal argument. Suppose you try to assign to each element a of a set A a subset f(a) of A, hoping to list all subsets this way. Now form a new subset:

S = {a in A : a is not in f(a)}

Then S cannot be equal to any f(a). If it were equal to f(k) for some k, ask whether k is in S. If it is, then by definition it is not in f(k) = S. If it is not, then by definition it is in f(k) = S. Contradiction. So no attempted listing of all subsets can succeed.

The power set of the natural numbers therefore has larger cardinality than the natural numbers themselves. Indeed, it has the cardinality of the continuum. And then the power set of that set is larger still.

So there is no largest infinity. There is an endless hierarchy:

ℵ₀ < continuum < larger infinities < still larger infinities < ...

The infinite is not a single foggy beyond. It is an ascending arithmetic landscape.

This is one of the rare moments in mathematics where an idea genuinely outruns ordinary language. The natural response is disbelief, not because the logic is obscure but because the conclusions violate inherited habits of thought. We are used to the idea that infinity is what lies beyond all finite numbers. Cantor showed that even beyond that there is structure, order, comparison, and growth.

He had made infinity into arithmetic.

Cantor Against Common Sense

Not everyone was pleased. Leopold Kronecker, one of the leading mathematicians in Germany and an advocate of a more finitist and arithmetic-minded approach, became Cantor’s bitter opponent. Kronecker distrusted completed infinities and thought mathematics should remain grounded in the integers and in explicitly constructive operations. He is often associated, somewhat unfairly in simplified retellings, with the remark that God made the integers and all else is the work of man.

Cantor, by contrast, was willing to accept infinite totalities as legitimate mathematical objects if they were defined clearly enough and reasoned about consistently enough. This was a philosophical as well as a mathematical divide. Was mathematics about objects that can be constructed step by step, or could it include completed infinite sets given all at once?

The dispute was not merely personal, though it became personal too. Kronecker’s hostility hurt Cantor professionally and emotionally. Cantor suffered repeated periods of severe mental illness, including hospitalisations. It would be careless and simplistic to blame this on mathematics or on Kronecker alone; human distress is rarely that tidy. But the intellectual resistance was real, and the emotional cost to Cantor was clearly heavy.

The tragedy here is familiar by now. Mathematics often honours its revolutionaries after first making them miserable.

Cantor knew he was changing the subject at a fundamental level. He also knew that many of his contemporaries regarded his work as metaphysical excess masquerading as mathematics. Yet he persisted, because the proofs held. However strange the conclusions looked, the arguments were unyielding.

That is one of the deepest habits of the subject. When rigorous reasoning and intuition clash, mathematics eventually sides with the reasoning.

The Continuum Hypothesis

Once Cantor had shown that the natural numbers and the real numbers have different cardinalities, another question naturally arose:

Is there any infinite size strictly between them?

That is, is there a set whose cardinality is larger than:

ℵ₀

but smaller than the continuum?

Cantor believed the answer was no. This claim is called the continuum hypothesis.

It is easy to state and extraordinarily difficult to settle. Like the parallel postulate in geometry or the quintic in algebra, it became one of those deceptively plain questions that expose the depth of a subject.

What matters for this chapter is not the later technical fate of the continuum hypothesis, but the fact that it emerged so quickly and naturally from Cantor’s new arithmetic of infinity. Once sets and cardinalities exist, the infinite becomes a terrain with its own landmarks and mysteries. The continuum hypothesis was the first great unsolved problem in that terrain.

Hilbert would later place it first on his famous 1900 list of problems for the twentieth century. That was a sign that set theory was no longer a strange side-path. It had moved to the centre of mathematics.

But before it could stabilise there, it had to survive a more serious crisis.

Russell’s Paradox

Cantor had given mathematicians a powerful new language: sets, subsets, cardinalities, infinite hierarchies. It was natural, in the first flush of enthusiasm, to treat sets rather loosely. A set was simply any collection of objects satisfying some condition. That sounds harmless. It was not.

In 1901, Bertrand Russell discovered a paradox that exposed a contradiction at the heart of naive set theory.

Consider the set:

the set of all sets that do not contain themselves as members.

Call this set R.

Now ask:

Does R contain itself?

If it does contain itself, then by definition it should not contain itself, because R is supposed to contain only those sets that do not contain themselves.

If it does not contain itself, then by definition it should contain itself, because it satisfies the condition for membership.

So either way:

R ∈ R  if and only if  R ∉ R

Contradiction.

This was not a puzzle about wording. It was a structural disaster. The naive principle “any definable collection is a set” had generated inconsistency.

And inconsistency in foundations is intolerable. If contradiction is allowed into the basic language of mathematics, then in principle anything can be proved.

The shock was profound. Cantor’s infinite paradise, as Hilbert would later call it, now seemed to rest on unstable ground.

The subject had gone from exhilarating expansion to foundational alarm in a generation.

Foundations Become a Problem

This was the point at which mathematics began to turn its full attention on itself.

For centuries, mathematicians had been willing to work with methods whose foundations were not completely secure. Calculus had done this successfully. Infinite series had done it earlier still. Usually the rigor came later and cleaned up the practice.

Set theory was different because it threatened the whole edifice at once. Sets had become the language in which large parts of modern mathematics could be reformulated. Analysis, topology, algebra, and logic were all beginning to rely on set-theoretic notions. If the language itself was inconsistent, then the danger spread everywhere.

The response was not to abandon sets, but to discipline them. Ernst Zermelo and later Abraham Fraenkel and others reformulated set theory axiomatically, restricting the kinds of set formation allowed. Instead of saying “every definable collection is a set,” one specified explicit axioms governing what sets may exist and how new ones may be formed from old ones.

This was, in a deep sense, Euclid all over again. When intuition becomes dangerous, mathematics retreats to axioms.

The pattern repeats across the centuries. Greek geometry becomes rigorous through axiomatic order. Calculus is rebuilt through limits. Set theory, after Cantor and Russell, is rebuilt through formal axioms.

Each time the subject grows more powerful, it also becomes less innocent.

Why Cantor Changed Everything

Cantor changed mathematics in at least three ways.

First, he made infinity precise. What had been philosophical atmosphere became mathematical structure.

Second, he showed that infinite size can be compared, ordered, and distinguished. Infinity was not one thing but many.

Third, he forced mathematics to confront foundational questions explicitly. Once sets became central, the subject could no longer avoid asking what counts as a legitimate object and what kinds of reasoning are safe.

This is why Cantor belongs with Euclid, Newton, and Galois as a revolutionary of the mathematical imagination. Euclid made proof central. Newton made change calculable. Galois made symmetry structural. Cantor made the infinite countable, uncountable, and hierarchy-rich all at once.

He also changed the emotional atmosphere of mathematics. After Cantor, the subject no longer seemed confined to the finite and the concrete, nor even merely to the continuous and the geometric. It had entered a region where the basic objects were collections, mappings, infinities, and logical possibilities. Modern mathematics became not only broader but stranger.

That strangeness was not decorative. It would shape the twentieth century.

The Price of the Infinite

There is a tension running through this whole history that becomes especially sharp here.

Mathematics advances by abstraction. That is one of the book’s central arguments. Objects that begin as tools for practical problems become, over time, detached from their origins and studied on their own terms. Complex numbers, functions, non-Euclidean spaces, groups, sets: each step moves farther from immediate necessity and deeper into structure.

And again and again, the abstractions turn out to be useful.

But usefulness is not the only consequence. Abstraction also destabilises. It forces mathematics to inspect its own assumptions. The more general the concepts become, the more urgently the subject must ask whether it still knows what it is doing.

Cantor’s work is a perfect example. It opened a magnificent new domain and nearly provoked a foundational crisis at the same time. It enlarged mathematics and made it less secure. That is not a contradiction. It is often how serious intellectual progress feels from the inside.

The infinite, once invited in, would not behave politely.

What the Infinite Revealed

By the early twentieth century, mathematics knew something it had never known before. The infinite is not merely the unfinished. It is not merely what lies beyond every finite stage. It has internal structure. It has arithmetic. It has paradoxes. It has levels. That was Cantor’s revelation.

The old language of the infinite had been mostly negative: boundless, endless, immeasurable, beyond. Cantor replaced it with a positive mathematics. One can compare infinite sets. One can prove one larger than another. One can formulate questions about their relations. One can be surprised by rigorous theorems about them.

This is one of the moments when mathematics most clearly shows its power to outgrow intuition without abandoning reason. No one would have guessed, from ordinary experience, that the rationals and the integers have the same cardinality while the reals have a larger one. No one would have guessed that the power set operation creates an endless ladder of ever larger infinities. But once the proofs are seen, the conclusions become unavoidable.

Mathematics had once begun, in this book, as a technology for handling grain, tax, land, and calendars. By Cantor’s time it had become capable of taking the oldest metaphysical word in human thought — infinity — and turning it into a precise field of research.

That is an extraordinary arc.

It is also not the end of the story. The more precise the foundations became, the more urgent new questions grew. Could mathematics prove its own consistency? Could every truth be formally derived from axioms? Could the infinite paradise Cantor opened be made permanently safe?

The twentieth century would discover that these questions, too, had answers stranger than anyone expected.

In the next chapter, the abstractions of geometry return to the physical world. If light travels at the same speed for everyone, then space and time cannot be what Newton thought they were. Minkowski and Einstein would turn geometry into physics, and the universe itself would become a mathematical object.