Construction of p-adic Numbers

In the familiar world of mathematics, the distance between two numbers is measured along the straight, infinite road of the real number line. This concept of distance, based on absolute value, seems natural and absolute. But what if it isn't? What if there are entirely different, equally valid ways to define what it means for two numbers to be "close"? This question opens the door to the strange and powerful universe of p-adic numbers, a system built not on a number's overall size, but on its properties of divisibility by a single prime number. This article explores the construction of this fascinating mathematical object.

First, under ​​Principles and Mechanisms​​, we will journey through the revolutionary ideas that underpin the p-adic world. We will define a new way to measure distance using the p-adic valuation and explore the bizarre, non-archimedean geometry that results. We will then detail the formal construction of the p-adic numbers, $\mathbb{Q}_p$, as the completion of the rational numbers, a process that mirrors the construction of the real numbers themselves. Following this, in ​​Applications and Interdisciplinary Connections​​, we will see why this abstract construction is not merely a mathematical curiosity. We will discover how p-adic numbers serve as a powerful microscope for number theorists, providing tools to solve ancient problems about whole numbers, forging connections to fractal geometry, and revealing a hidden harmony that unites the entire landscape of mathematics.

Imagine you’re a cartographer. Your job is to map the world of numbers. You're familiar with the number line, a straight, infinite road where numbers are spaced according to their size. The distance between any two numbers, $x$ and $y$, is simply $|x-y|$. This seems like the only natural way to measure distance. But what if it isn't? What if there are entirely different, equally valid, ways to define what it means for two numbers to be "close"? This is the revolutionary idea that gives birth to the world of $p$-adic numbers.

Instead of focusing on a number's overall size, let's fix a prime number, say $p=5$, and focus on its divisibility by $5$. We could say a number is "small" if it's divisible by a large power of $5$. For instance, $50 = 2 \times 5^2$ is divisible by $5^2$, while $15 = 3 \times 5^1$ is only divisible by $5^1$. In this new view, $50$ is "smaller" than $15$. And a number like $3$, which isn't divisible by $5$ at all, is considered "large".

We can make this precise with the ​​p-adic valuation​​, denoted $v_p(x)$. For any rational number $x$, $v_p(x)$ is simply the exponent of $p$ in its prime factorization. For example, if $x = 450/7 = (2 \cdot 3^2 \cdot 5^2)/7$, then $v_5(450/7) = 2$, $v_3(450/7) = 2$, and $v_2(450/7) = 1$. For any other prime $q$ (like $7$ or $11$), $v_q(450/7)$ is non-positive (in this case, $v_7(450/7) = -1$). A larger valuation means the number is more divisible by $p$.

From this valuation, we define a new kind of size, the ​​p-adic absolute value​​: $|x|_p = p^{-v_p(x)}$. If $x=0$, we set $|0|_p=0$. Notice the minus sign in the exponent! This inverts our intuition: a high power of $p$ (large valuation) leads to a small absolute value. For $p=5$:

• $|25|_5 = 5^{-v_5(25)} = 5^{-2} = \frac{1}{25}$
• $|5|_5 = 5^{-v_5(5)} = 5^{-1} = \frac{1}{5}$
• $|3|_5 = 5^{-v_5(3)} = 5^{-0} = 1$
• $|1/5|_5 = 5^{-v_5(1/5)} = 5^{-(-1)} = 5$

This leads to a mind-bending consequence. What happens when we multiply a number by $p$? In the real world, it gets bigger. In the $p$-adic world, it gets smaller! The distance from $px$ to $0$ is $|px|_p = |p|_p |x|_p = \frac{1}{p}|x|_p$. The map $f(x)=px$ is not an expansion; it's a ​​contraction​​ that shrinks everything by a factor of $\frac{1}{p}$. This simple fact is the key to the unique character of the $p$-adic universe.

This new way of measuring distance, $d_p(x,y) = |x-y|_p$, creates a geometry that defies our everyday experience. The $p$-adic absolute value doesn't just satisfy the standard triangle inequality, $d(x,z) \le d(x,y) + d(y,z)$. It satisfies a much stronger version, the ​​ultrametric inequality​​: $|x+y|_p \le \max(|x|_p, |y|_p)$ This means the "length" of one side of a triangle is never longer than the maximum of the lengths of the other two sides. This seemingly small change has bizarre consequences. For instance, in any $p$-adic triangle, the two longest sides must have equal length. In other words, ​​all triangles are isosceles​​!

Imagine an open ball $B(x,r)$, which is the set of all points $y$ such that their distance to the center $x$ is less than $r$. In our familiar geometry, the center is a special point. Not so in the $p$-adic world. Here, ​​any point inside an open ball is also its center​​. This strange property means that two balls are either completely disjoint or one is entirely contained within the other. There is no such thing as a partial overlap. The $p$-adic world is not a smooth, continuous landscape but a strangely structured, hierarchical, almost fractal-like space.

Just as the rational numbers $\mathbb{Q}$ are full of "holes" from the perspective of the real number line (for example, $\sqrt{2}$ is missing), they are also incomplete with respect to this new $p$-adic distance. For example, using the $7$-adic metric, one can construct a sequence of rational numbers that get closer and closer to a "square root of 2," even though no such rational number exists. This sequence is a ​​Cauchy sequence​​—its terms eventually get arbitrarily close to each other—but it doesn't converge to a point within the rational numbers.

To fill these holes, we perform a procedure called ​​completion​​. It’s the very same procedure that is used to construct the real numbers $\mathbb{R}$ from $\mathbb{Q}$. The idea is to declare that the limits of these Cauchy sequences exist, and that the new, "complete" space consists of these limits. Formally, we define a new number to be a Cauchy sequence of rational numbers. To avoid multiple representations for the same limit, we say that two Cauchy sequences are equivalent if their difference converges to zero. The field of ​​p-adic numbers​​, $\mathbb{Q}_p$, is precisely this set of equivalence classes of Cauchy sequences.

This new field, $\mathbb{Q}_p$, contains the original rational numbers $\mathbb{Q}$ (represented by constant Cauchy sequences) and is "complete"—every $p$-adically Cauchy sequence now has a limit within $\mathbb{Q}_p$. And just as any real number can be approximated by a sequence of rational numbers, any $p$-adic number can be approximated by a sequence of rationals, typically by truncating its infinite $p$-adic expansion. This means the rational numbers are ​​dense​​ in the $p$-adic numbers, forever weaving our familiar number system into the fabric of this strange new one.

There is another, beautifully algebraic, way to construct these numbers. Think of a number not as a single point, but as a system of information that becomes more and more precise. A ​​p-adic integer​​ (an element of the subring $\mathbb{Z}_p \subset \mathbb{Q}_p$) can be seen as a sequence of integers $(x_n)_{n\ge1}$ where $x_1$ is defined modulo $p$, $x_2$ is defined modulo $p^2$, $x_3$ is defined modulo $p^3$, and so on. The key is that this sequence must be ​​compatible​​: each piece of information must refine the previous one. For instance, the value modulo $p^2$ must be consistent with the value modulo $p$, meaning $x_2 \equiv x_1 \pmod{p}$.

This construction, called an ​​inverse limit​​, gives us a powerful way to think about $p$-adic numbers as infinite strings of digits in base $p$. And it leads to some truly astonishing results.

Consider the geometric series $1 + p + p^2 + p^3 + \dots$. In the real numbers, this series diverges wildly for any $p \ge 2$. But in the $p$-adic world, the terms get smaller and smaller, since $|p^n|_p = p^{-n} \to 0$. The series converges! And what does it converge to? The old high school formula still works: $1 + p + p^2 + \dots = \frac{1}{1-p}$ So the simple rational number $\frac{1}{1-p}$ is a $p$-adic integer whose base-$p$ expansion is just an endless string of ones: $\dots 111_p$.

Perhaps even more shocking is the number whose digits are all $p-1$: $(p-1) + (p-1)p + (p-1)p^2 + \dots$ This series also converges, and its sum is simply ​​-1​​. This means that in any $p$-adic system, the number $-1$ has the representation $\dots(p-1)(p-1)(p-1)_p$. For $p=2$, this means $...111_2 = -1$. For $p=5$, $...444_5 = -1$. This is not an approximation; it is an identity.

These expansions can reveal hidden structures in familiar numbers. The seemingly innocuous fraction $\frac{1}{3}$, when viewed in the $2$-adic numbers, becomes the limit of a geometric series $\frac{1}{1-(-2)} = \sum_{n=0}^\infty (-2)^n$, which corresponds to the endlessly alternating base-2 expansion $\dots 0101011_2$.

This new arithmetic isn't just a curiosity; it's a powerful tool. One of its most celebrated applications is ​​Hensel's Lemma​​, which is essentially a $p$-adic version of Newton's method for finding roots of polynomials.

The idea is remarkable: if you can find an approximate solution to a polynomial equation modulo $p$, you can often lift it to an exact solution in the $p$-adic integers. The process involves starting with your root $a_1 \pmod p$ and iteratively refining it to find a root $a_2 \pmod{p^2}$, then $a_3 \pmod{p^3}$, and so on, with each step uniquely determined as long as the derivative at the root is non-zero. This generates a compatible sequence $(a_n)$ which, by the inverse limit construction, defines a true root $a \in \mathbb{Z}_p$ where $f(a)=0$. This allows number theorists to move from finite, modular arithmetic to the infinite, continuous world of the $p$-adics, turning approximate solutions into exact ones.

For a long time, it seemed there was one world of numbers centered on the real number line, governed by the usual absolute value. Now we see there is an infinite family of other worlds, the $p$-adic fields, one for each prime $p$. A fundamental result known as ​​Ostrowski's Theorem​​ states that this is all there is. Up to equivalence, every possible way of defining an absolute value on the rational numbers falls into one of two categories: the familiar real absolute value, or a $p$-adic absolute value for some prime $p$.

This places the real numbers and all the $p$-adic numbers on an equal footing. They are the fundamental completions of the rational numbers, the different ways to fill its gaps. But the story doesn't end there. These seemingly separate worlds are deeply interconnected by a single, beautiful equation known as the ​​Product Formula​​. For any non-zero rational number $x$, $|x|_\infty \cdot \prod_{p \text{ prime}} |x|_p = 1$ Here, $|x|_\infty$ is the usual absolute value. This formula says that if you measure the "size" of a rational number in every possible world—the real world and every $p$-adic world—and multiply them all together, the result is always exactly $1$. A gain in size in the real world must be perfectly balanced by a loss of size across the various $p$-adic worlds. It is a stunning statement of conservation, revealing a hidden harmony that unites the entire landscape of number theory. The construction of $p$-adic numbers is not just an exercise in abstraction; it is the discovery of an essential part of this beautiful, unified structure.

Now that we have painstakingly built these peculiar $p$-adic number systems, a fair question is: what for? Are they merely a clever intellectual exercise, a sterile world of numbers where "big" and "small" have been turned on their heads, or do they offer us a new window onto the universe of mathematics? The true test of a physical law, or a mathematical idea, is not just its internal elegance, but what it can do. What new phenomena can it explain? What old problems can it solve?

It turns out that these strange number systems, far from being a mere curiosity, are an essential tool, a new kind of microscope for peering into the deepest structures of mathematics. They forge surprising connections between disparate fields—from the geometry of fractals to the most profound questions about whole numbers that have puzzled mathematicians for centuries. Let us take a tour of this new universe and see what wonders it holds.

Our first stop is the world of geometry and analysis. We are used to the real number line, a continuous, seamless thing. The $p$-adic world, as we have seen, is utterly different. It's a "dust" of disconnected points. How can one possibly do calculus or geometry in such a space? The key is that it is a complete metric space, which means sequences that should converge actually do converge to a point within the space. This is all we need to get started.

What happens when we study motion in this space? Consider a simple "dynamical system" on the $p$-adic integers $\mathbb{Z}_p$, where we repeatedly apply a function and see where we end up. Let's take one of the simplest possible functions, an affine map $T(x) = ax+b$. On the real line, iterating such a map can lead to all sorts of behaviors—the point can fly off to infinity, settle on a fixed value, or oscillate. But in the $p$-adic world, the bizarre "all triangles are isosceles" geometry imposes a surprising rigidity. For any two points $x$ and $y$, the distance between their images is $|T(x) - T(y)|_p = |a|_p |x-y|_p$. This means the map shrinks all distances by a fixed factor $|a|_p$. If this factor is less than one—which happens if $a$ is a multiple of $p$—the map is a "contraction." The Banach fixed-point theorem then tells us something wonderful: no matter where you start, repeatedly applying the map will always lead you to one unique, unmoving fixed point. The motion is perfectly, unerringly predictable.

This very property of contraction is what allows us to build fractals in the real world. Think of the famous Sierpinski gasket, built by repeatedly shrinking a triangle and placing copies of it inside the original. We can play the same game in $\mathbb{Z}_p$. Using a set of contracting maps, known as an Iterated Function System, we can define a unique set, called an attractor, which is a union of shrunken copies of itself. For example, on the 2-adic integers $\mathbb{Z}_2$, the two maps $f_1(x) = 2x$ and $f_2(x) = 2x+2$ create an attractor. And what is this exotic fractal set? It is nothing other than the ball $2\mathbb{Z}_2$—the set of all even 2-adic integers! This is a beautiful revelation: the very structure of the $p$-adic integers is inherently fractal. The ball $p\mathbb{Z}_p$ is made of $p$ smaller, identical disjoint balls, namely $p^2\mathbb{Z}_p$, $p^2\mathbb{Z}_p+p$, $\dots$, $p^2\mathbb{Z}_p+(p-1)p$. This self-similarity is built into their very bones.

The topology of $p$-adic spaces holds even greater surprises. The space $\mathbb{Z}_p$ is "totally disconnected"—it is like a fine dust of points, with no paths or curves connecting them. The real interval $[0,1]$, by contrast, is the epitome of connectedness. Surely, these two worlds are irreconcilable. Yet, the magic of continuity can bridge this gap. It is possible to construct a continuous function that takes the totally disconnected space of 4-adic integers, $\mathbb{Z}_4$, and maps it onto the entire connected interval $[0,1]$. This is a "space-filling curve" of a most extraordinary kind, painting a solid line from a starting point of disconnected dust.

This profound difference in topology explains why some of our most trusted mathematical tools fail when we cross from the familiar archimedean world to the non-archimedean. The Fundamental Theorem of Algebra, which guarantees that any polynomial has a root in the complex numbers, has a famous and beautiful proof using the idea of "winding numbers." You draw a huge circle in the complex plane, see how the polynomial wraps its image around the origin, and show that this wrapping prevents the image from ever missing the origin. But if you try to replicate this argument in the $p$-adic numbers, you fail at the first step. In the totally disconnected space $\mathbb{Q}_p$, you cannot even draw a loop! Any continuous path from a connected interval like $[0,1]$ into $\mathbb{Q}_p$ must be constant. The entire concept of winding numbers becomes meaningless. Understanding why a proof fails is often as enlightening as understanding why it succeeds. It forces us to recognize that a proof is not just an abstract sequence of deductions, but something that lives and breathes within the specific topological environment of its host space.

For all their geometric strangeness, the prime motivation for inventing $p$-adic numbers was, and remains, the study of the ordinary whole numbers. The central idea is the ​​local-global principle​​, which is a philosophy, a strategy for solving problems about integers or rational numbers.

Imagine you are a detective investigating a crime—say, whether a Diophantine equation like $x^2 + y^2 = z^2$ has an integer solution. The local-global principle suggests you look for clues "locally." First, you check the scene in the real numbers, $\mathbb{R}$. Does the equation have a real solution? (Yes, obviously). Then, you dust for fingerprints in every $p$-adic world, for every prime $p$. Does the equation have a solution in $\mathbb{Q}_2$? In $\mathbb{Q}_3$? In $\mathbb{Q}_5$? And so on. The hope, known as the Hasse Principle, is that if you find a solution in every local environment, then you are guaranteed to find a global solution in the integers or rational numbers.

For many simple equations, this principle works perfectly. But the real excitement is when it fails. There are equations for which we can painstakingly construct a solution in the real numbers and in every single $\mathbbQ_p$, yet no rational solution exists!. It is as if our detective finds consistent clues at every separate location, but they are impossible to reconcile into a single, coherent global story. These "counterexamples" are not failures; they are signposts to a deeper, more subtle structure. They tell us that there are "obstructions" to finding global solutions that are invisible in any single local picture and only emerge when we consider all the local worlds at once.

To manage this army of local worlds, mathematicians devised a truly magnificent structure: the ​​ring of adeles​​, $\mathbb{A}_{\mathbb{Q}}$. The adele ring is a single object that packages $\mathbb{R}$ and all of the fields $\mathbb{Q}_p$ together. It is like a grand control panel with dials for every completion of the rational numbers. This allows number theorists to phrase questions about the local-global principle and its obstructions in a unified, elegant language.

This $p$-adic machinery is not just for philosophical statements; it provides powerful computational tools. A celebrated result known as ​​Chabauty's method​​ uses $p$-adic analysis to prove that certain algebraic curves (solutions to polynomial equations) have only a finite number of rational points. The idea is breathtaking: inside a high-dimensional $p$-adic space, the curve is a one-dimensional "thread." The group of rational points of a related object (the Jacobian) forms a "sheet" of a certain dimension. If the rank of this group is less than the genus of the curve, the dimension of the sheet is smaller than the ambient space. This allows one to construct $p$-adic analytic functions that are zero on the sheet. When you restrict these functions to the one-dimensional thread of the curve, they can only be zero at a discrete, finite set of points! Since all rational points must lie in this set, there can only be finitely many. It is a stunning use of a $p$-adic analytic "scalpel" to pin down the elusive rational solutions to an ancient problem.

This theme of a "parallel universe" runs through the highest levels of modern number theory. For nearly every major analytic tool we have in the real/complex world for studying number theory, a $p$-adic analogue has been found.

• On an elliptic curve, we can measure the "arithmetic complexity" of a rational point using the real-valued ​​canonical height​​. Amazingly, for each prime $p$, there exists a parallel ​​$p$-adic height pairing​​, constructed using $p$-adic integration. Both of these heights give rise to "regulators," and these regulators appear at the heart of the respective real and $p$-adic versions of the Birch and Swinnerton-Dyer conjecture—one of the seven Millennium Prize Problems. The structure of rational points on the curve is conjectured to be intimately linked to the behavior of both a complex $L$-function and a whole family of $p$-adic $L$-functions.
• ​​Leopoldt's conjecture​​ asks a seemingly simple question: if we take a set of "multiplicatively independent" numbers in a number field (like the fundamental units), do they remain independent when we view them through a $p$-adic microscope? The answer is believed to be yes, and this is equivalent to the non-vanishing of a certain determinant called the ​​$p$-adic regulator​​. This conjecture, which is still open for general number fields, lies at the foundation of Iwasawa theory and our understanding of the arithmetic of number fields.

From the geometry of fractals to the deepest conjectures in number theory, $p$-adic numbers have proven to be an indispensable tool. They are a testament to the fact that in mathematics, creating a new structure, a new way of looking at things, can unveil connections and solve problems that were previously untouchable. They show us that the world of numbers is far richer, stranger, and more beautiful than we could have ever imagined from our limited vantage point on the real number line.