Paul Cohen: Forcing, Independence, and the Mathematical Multiverse

Since Georg Cantor first revealed the staggering landscape of different infinite sizes in the 19th century, one question has stood out for its deceptive simplicity and profound difficulty: is there an infinity between the size of the whole numbers and the size of the real numbers? This question, known as the Continuum Hypothesis (CH), became a central challenge for mathematics, occupying the first spot on David Hilbert's famous list of problems for the 20th century. For decades, the greatest minds in logic and set theory struggled to either prove or disprove it from the standard Zermelo-Fraenkel axioms (ZFC), the bedrock of modern mathematics. The truth, as it turned out, was more subtle and revolutionary than a simple "yes" or "no".

This article chronicles the resolution of the Continuum Hypothesis problem, a story in two brilliant parts. It reveals how the axioms we take for granted do not dictate a single mathematical universe, but rather allow for a multitude of them. In the first chapter, ​​"Principles and Mechanisms,"​​ we will delve into the ingenious constructions of Kurt Gödel and Paul Cohen. We will explore Gödel’s orderly "constructible universe" where the Continuum Hypothesis is true, and then unpack Cohen’s groundbreaking technique of "forcing," which allowed him to build new mathematical realities where the hypothesis is false. The second chapter, ​​"Applications and Interdisciplinary Connections,"​​ examines the seismic impact of this independence. We will investigate the parallel story of the Axiom of Choice and explore the startling trade-offs that arise in different mathematical worlds, ultimately confronting the pluralistic vision of mathematics that Cohen's work unveiled.

Imagine you are standing before a grand, infinite staircase. The first step is labeled '0', the next '1', '2', and so on, for all the whole numbers. This first flight of stairs goes on forever, and its "size" is the first kind of infinity, what mathematicians call ​​aleph-naught​​, or $\aleph_0$. Now, look at a simple line segment, say from 0 to 1 on the number line. How many points are on it? It seems like there should be more points than there are whole numbers, and indeed, a beautiful proof by Georg Cantor in the 19th century showed that this is true. The infinity of points on a line, called the ​​continuum​​ and denoted by the symbol $\mathfrak{c}$, is a "larger" infinity than $\aleph_0$.

Cantor proved that $\mathfrak{c}$ is the same size as the set of all possible subsets of the natural numbers, a size we write as $2^{\aleph_0}$. He also showed that there is a well-ordered ladder of infinite sizes, the aleph numbers: $\aleph_0, \aleph_1, \aleph_2, \dots$, where each is the very next possible size of infinity after the one before it. We know for certain that $\mathfrak{c}$ is larger than $\aleph_0$. The very next size of infinity after $\aleph_0$ is, by definition, $\aleph_1$. So, we know that $\mathfrak{c} \geq \aleph_1$. This led Cantor to ask a deceptively simple question that would haunt mathematics for a century: Is the size of the continuum exactly the next size of infinity? In other words, is it true that $2^{\aleph_0} = \aleph_1$?

This assertion is the famous ​​Continuum Hypothesis (CH)​​. It proposes that there is no intermediate size of infinity squeezed between the infinity of the whole numbers and the infinity of the points on a line. For decades, the greatest minds in mathematics tried to prove it or disprove it from the standard axioms of set theory—the Zermelo-Fraenkel axioms with the Axiom of Choice (ZFC)—but all attempts failed. The truth was far more subtle and profound, and it took two of the 20th century's most brilliant logicians to uncover it.

The first breakthrough came in 1940 from Kurt Gödel, the same logician who had earlier stunned the world with his Incompleteness Theorems. To show that the Continuum Hypothesis could not be disproved from ZFC, Gödel needed to construct a universe of mathematics where ZFC was true and CH was true. If such a world is possible, then CH can't be a falsehood, because there is a consistent reality in which it holds.

How do you build a mathematical universe? Gödel's idea was one of elegant minimalism. Instead of assuming all the wild and woolly sets that might exist, he built a universe from the ground up with a very strict rule: at each stage, you are only allowed to add sets that are explicitly definable using a first-order formula with parameters from the sets you have already built. Think of it as building a house using only the tools and materials you can precisely describe and create from what you already possess. Nothing extra, nothing mysterious.

This process, carried out over a transfinite hierarchy of stages, creates a sleek, well-behaved inner world within the larger universe of all sets. This world is called the ​​constructible universe​​, denoted by the letter ​​L​​. Gödel showed that this universe $L$ is a perfectly valid model for all the axioms of ZFC. But what about CH?

In the sparse, disciplined world of $L$, there isn't enough "definitional freedom" to create a profusion of strange sets. Every set in $L$ is there for a reason, pinned down by a formula. Gödel used this feature, through a powerful technical tool called the ​​Condensation Lemma​​, to show something remarkable. He proved that in $L$, not just CH, but the ​​Generalized Continuum Hypothesis (GCH)​​ is true. GCH is the sweeping statement that for every infinite cardinal $\kappa$, the next largest size of infinity is precisely the size of its power set, or $2^\kappa = \kappa^+$. The CH is just the first instance of this beautiful, regular pattern.

The proof, in essence, shows that any constructible subset of an infinite set of size $\kappa$ must appear by the $\kappa^+$-th stage of the hierarchy. This puts a strict upper limit on how many such subsets there can be. There are simply not enough stages in the construction to produce more than $\kappa^+$ subsets, so $(2^\kappa)^L = \kappa^+$. In this universe of perfect order, the ladder of infinities has no unexpected gaps; each rung is as close to the next as it could possibly be.

By constructing $L$, Gödel demonstrated that if ZFC is consistent, then ZFC + CH must also be consistent. The Continuum Hypothesis could not be disproved. This was the first half of the puzzle.

The second, and arguably more revolutionary, half of the puzzle was solved by Paul Cohen in 1963. To prove CH was independent, he had to show that it could not be proven from ZFC either. This required building a universe where ZFC holds, but CH is false. Gödel had built a minimalist world; Cohen needed to find a way to add new, exotic objects to a world without shattering its axiomatic foundations.

His method, called ​​forcing​​, is one of the most powerful and counter-intuitive ideas in modern mathematics. Imagine you want to add a new real number to your universe. This number has a binary expansion, an infinite sequence of 0s and 1s. But you can't just write it down, because if you could, it would have been in your universe all along! Cohen's genius was to describe this new object incompletely.

Forcing works by defining a set of "conditions" or "finite approximations" of the new object we want to add. For a new real number, a condition might be a finite string of 0s and 1s, like "the first 10 digits are 0110101001...". We don't specify the rest. We then imagine a "generic" object that satisfies a collection of these conditions but has no other special properties that could be described in the old universe. It is "generic" in the sense that it avoids being pinned down by any pre-existing definition.

To build a model where CH is false, Cohen needed to make the continuum much larger than $\aleph_1$. A classic example is to build a model where $2^{\aleph_0} = \aleph_2$. To do this, he started with a model where CH was true (like Gödel's $L$) and used forcing to "add" $\aleph_2$ new real numbers. The forcing conditions are carefully chosen so that this massive injection of new objects doesn't break the rules of ZFC. One of the miracles of the method is a property called the ​​countable chain condition (ccc)​​, which ensures that the process doesn't accidentally mess with the infinite ladder itself; it doesn't "collapse" cardinals, making $\aleph_2$ suddenly the same size as $\aleph_1$.

The result is a new, expanded mathematical universe, a "generic extension" of the old one, which is a perfectly valid model of ZFC. But in this new universe, there are $\aleph_2$ real numbers. Therefore, CH is false.

Cohen's work, combined with Gödel's, was a thunderclap. It showed that the Continuum Hypothesis is ​​independent​​ of the ZFC axioms. It can be neither proven nor disproven within standard mathematics. The axioms of ZFC, our bedrock foundation for mathematics, do not determine the size of the continuum.

This revealed that ZFC does not describe a single, unique mathematical universe. Instead, it describes a "multiverse" of possible realities. In Gödel's elegant, constructible universe $L$, CH is true. In Cohen's generic extension, CH is false. By tweaking the forcing method, one can create models where $2^{\aleph_0}$ is $\aleph_3$, or $\aleph_{17}$, or even more exotic cardinals.

This discovery opened up a whole new landscape for mathematicians to explore. If ZFC is not enough, what other axioms could we add? This led to the study of "forcing axioms" like ​​Martin's Axiom (MA)​​ and the ​​Proper Forcing Axiom (PFA)​​. These are powerful principles that decide many questions left open by ZFC. For example, it is consistent to have MA and $\neg$CH (in fact, MA is often studied with the assumption that $2^{\aleph_0} = \aleph_2$). Even more strikingly, PFA, whose consistency requires assuming the existence of very large "large cardinals," actually implies that $2^{\aleph_0} = \aleph_2$.

Paul Cohen's legacy was not just the solution to a century-old problem. It was the radical idea that the world of mathematics is richer and more varied than anyone had ever imagined. He gave us not just an answer, but a tool to build new worlds and a license to explore the vast, uncharted territory of the mathematical multiverse.

After a journey through the intricate machinery of forcing, it’s natural to ask: what did Paul Cohen’s profound discovery actually do for mathematics? The answer is as beautiful as it is unsettling. Cohen didn't just solve a problem; he revealed that the world of mathematics is not a single, predestined reality. Instead, it is a vast, branching multiverse of possibilities, and his work gave us the keys to explore it. By proving the independence of the Axiom of Choice and the Continuum Hypothesis, Cohen showed that we, as mathematicians, stand at a fork in the road. The axioms we choose are the fundamental laws of our mathematical universe, and different choices lead to vastly different, yet equally valid, worlds.

For nearly a century, the Axiom of Choice (AC) was a powerful, if controversial, tool. It seems innocuous enough: given any collection of non-empty bins, you can always pick one item from each bin. Yet, this simple idea has far-reaching consequences. It is the invisible hand that organizes much of modern mathematics, appearing in many equivalent and powerful forms, such as the Well-Ordering Principle, which asserts that any set can be lined up in a neat, well-ordered sequence, and Zorn's Lemma, a crucial tool for proving existence theorems throughout algebra and topology.

With AC, the universe of sets is a tidy and well-behaved place. For instance, it guarantees that any two sets can be compared in size—for any sets $A$ and $B$, either there’s an injection from $A$ to $B$ or one from $B$ to $A$. There are no "incomparable" infinities. This axiom also underpins fundamental theorems in other fields. In linear algebra, the statement that every vector space has a Hamel basis—a set of building blocks from which every vector can be uniquely constructed—is a direct consequence of the Axiom of Choice. Without AC, this cornerstone of algebra is not provable.

But this power comes at a price. The choices guaranteed by AC are non-constructive; the axiom asserts a choice function exists without telling you how to find it. This leads to consequences that clash with our physical intuition. A classic example comes from measure theory, the mathematical formalization of concepts like length, area, and volume. Using AC, one can construct a so-called Vitali set, a subset of the real number line so bizarre and scattered that it cannot be assigned a meaningful "length" or Lebesgue measure. AC gives us powerful tools like Hamel bases, but it also forces us to accept the existence of these mathematical "monsters."

For decades, mathematicians wondered if these monsters were an unavoidable feature of reality. Cohen's work, and subsequent results by logicians like Robert Solovay, provided a stunning answer: no. Using techniques inspired by forcing, Solovay constructed a model of set theory (consistent relative to the existence of certain large cardinals) in which the Axiom of Choice is false. And in this brave new mathematical world, every single subset of the real numbers is Lebesgue measurable. The monsters are gone!

This seems like a wonderful improvement. But what is the cost of living in such a tidy universe? The very same theorems that depended on AC now fail. In the Solovay model, the vector space of real numbers over the rationals, $\mathbb{R}$ over $\mathbb{Q}$, has no Hamel basis. Furthermore, the tidiness of cardinal comparison vanishes. This universe is populated by pairs of infinite sets that are fundamentally incomparable in size—you cannot map one into the other injectively in either direction.

Cohen’s legacy, then, is not just a proof, but a choice. Do we prefer a universe where every vector space has a basis, at the cost of accepting sets with no defined volume? Or do we prefer a universe where every set has a volume, but where fundamental algebraic structures may not exist and infinities can be incomparable? Cohen proved that Zermelo-Fraenkel set theory cannot decide for us.

The plot thickens further still. The choice is not merely a binary switch between "AC on" and "AC off." Research following Cohen's breakthrough has revealed a rich hierarchy of weaker choice principles, painting a spectrum of possibilities between the austere world of ZF and the wild one of ZFC.

For instance, the Axiom of Dependent Choice (DC) allows for making a sequence of choices where each choice depends on the previous one, like finding an infinite path through a tree where every node has a successor. The Axiom of Countable Choice (CC) is weaker still, guaranteeing a choice function only for a countable collection of sets.

This is not just a technicality; it has profound implications for the practice of mathematics. It turns out that a vast portion of classical analysis, including the Baire Category Theorem and the proof that a countable union of countable sets is countable, does not require the full, controversial power of AC. The more palatable Axiom of Dependent Choice is sufficient. This allows for a "middle way": a mathematical universe where analysis works as expected, but the most bizarre paradoxes of full AC (like the non-existence of a universal measure on all subsets of $\mathbb{R}$) can be avoided. Cohen’s methods allow us to build such universes, giving mathematicians a framework to ask precisely how much "choice" is needed to prove a given theorem.

Cohen's most famous result, of course, concerns the Continuum Hypothesis (CH). Here, his work stands in beautiful contrast to that of Kurt Gödel. Years before Cohen, Gödel had shown that CH is consistent with the ZFC axioms. He did this by constructing an "inner model" of set theory, the constructible universe $L$.

The universe $L$ is a minimalist's dream. It is built from the ground up, level by level, using only what is explicitly definable. It is a slim, elegant, and astonishingly orderly reality. This inherent orderliness means there isn't much "room" in $L$. As it turns out, there is just not enough space to fit an infinity between the size of the integers and the size of the real numbers. The very structure of $L$, governed by definability and a powerful principle called condensation, ensures that the number of constructible real numbers is the very next infinity after the integers, $\aleph_1$. In $L$, the Continuum Hypothesis is not an axiom, but a provable theorem.

Gödel showed you can live in a universe where CH is true. Then came Cohen. Using forcing, he showed how to do the opposite. Starting with a minimalist universe like $L$, he invented a way to "force" it to grow, systematically adding "generic" sets. He showed how to construct new, "fatter" mathematical universes—universes that were still perfectly valid models of ZFC, but in which vast collections of new real numbers existed.

In one such universe, the number of reals could be $\aleph_2$. In another, it could be $\aleph_{517}$. Subsequent work by Easton showed that the behavior of the continuum could be almost anything not ruled out by basic theorems. Cohen’s work revealed that the real number line, a concept at the heart of geometry, calculus, and physics, has no definite, provable cardinality within our standard axiomatic framework. Its "size" is one of the unfixed parameters of our mathematical reality.

Paul Cohen’s legacy is a paradigm shift in our understanding of mathematical truth. He took set theory from a single, rigid foundation and transformed it into a dynamic, creative exploration of possible worlds. His work has profound implications for the philosophy of mathematics, challenging the Platonist view of a single, immutable reality that we discover. Instead, it suggests a more formalist or pluralistic perspective, where we construct different, consistent realities and explore their unique properties.

This perspective has found echoes in other fields. The methods of model theory and formal logic, honed to a fine edge by the challenges of independence proofs, are now indispensable tools in theoretical computer science for reasoning about the semantics of programs and the limits of computation.

Ultimately, Cohen left us not with a final answer, but with a more profound question. The goal is no longer just to determine what is true, but to understand what is true in which universe, and why. This transforms mathematics into an even grander adventure. We are not just explorers of a single continent, but architects and travelers of an entire multiverse, free to build new worlds and discover the beautiful, strange, and unexpected truths they hold.