Consequences of the Field Axioms

The nine field axioms and their basic consequences are the shared foundation beneath enormous swaths of mathematics and its applications. In Ordered Fields, one adds an order relation compatible with $+$ and $\cdot$, and the zero-product property becomes essential for proving that squares are nonnegative and that the reals cannot be totally ordered in a way compatible with $\sqrt{-1}$. In The Rational Numbers as an Ordered Field, verifying the field axioms for $\mathbb{Q}$ (built from $\mathbb{Z}$ via formal fractions) is exactly where cancellation is used to show the fraction construction is well-defined. In finite field theory, $\mathbb{F}_p$ (integers mod a prime $p$) satisfies all nine axioms — critically, (M4) holds because $p$ is prime, which is precisely why $\mathbb{Z}/6\mathbb{Z}$ above fails to be a field. This underlies the Diffie–Hellman and RSA-adjacent constructions in cryptography, error-correcting codes (Reed–Solomon codes over $\mathbb{F}_{2^n}$), and algebraic geometry over finite fields. In linear algebra, the zero-product property is exactly what makes the theory of a vector space's dimension work over any field. In numerical analysis and computer arithmetic, the fact that division by zero is undefined rather than merely inconvenient traces directly back to axiom (M4)'s restriction to nonzero elements, combined with the theorem above showing that allowing $0^{-1}$ would force $1=0$.

Ask the tutor: Why does the proof of the zero-product property fail if we only assume $F$ is a ring, and what extra structure exactly fixes it?

The field axioms describe two layers of structure that mesh together. Addition alone (axioms A1–A4) makes $F$ into what is called an abelian group: a set where you can combine, undo, and reorder additions freely. Multiplication restricted to the nonzero elements (axioms M1–M4) forms a second abelian group. The single axiom gluing the two together is distributivity, and it is genuinely doing all the interaction work — without it, addition and multiplication would be two unrelated structures living on the same set with no connection between them. The logical dependency chain in this article looks like a short ladder. Cancellation laws sit at the bottom because they need only the group axioms for one operation at a time. Uniqueness statements sit above them, since "there is only one identity" is proved by using cancellation, or by pitting two candidate identities against each other directly. The fact $0 \cdot a = 0$ needs distributivity to bridge the two operations. And the zero-product property sits at the top, because it needs all of this machinery: distributivity to relate $0\cdot a=0$ to the product $ab$, and the multiplicative inverse axiom applied delicately to $a \neq 0$.

Notice exactly where hypothesis $a \neq 0$ enters: axiom (M4) only guarantees an inverse for nonzero $a$, so part (ii) cannot even get started without it. This is not a minor technicality — the counterexample below shows precisely what collapses when it's dropped.

This proof uses only (A2) and the defining property of an identity — no cancellation needed, since it is a direct "race two candidates against each other" argument. The identical argument, with the roles of $+$ and $\cdot$ swapped, proves uniqueness of $1$.

These four uniqueness results justify the definite-article notation "$-a$" and "$a^{-1}$" and "the zero element" and "the unit element" used everywhere in field theory: the axioms only assert existence, but existence plus these theorems gives exact-one-ness, which is what the notation presupposes.

Note the proof technique: it introduces the same quantity $0\cdot a$ on both sides of an equation and invokes additive cancellation — exactly the tool built in the first theorem. This is the first point in the chain where distributivity (D) is essential; nothing about $0\cdot a$ could be derived from the additive and multiplicative group axioms in isolation, since those describe $+$ and $\cdot$ as unrelated operations.

Example 1 (using cancellation to solve an equation abstractly). Let $F$ be an arbitrary field and let $x, y, z \in F$ satisfy

Suppose further that $w \in F$ satisfies $y \cdot w = z \cdot w$ and $w \neq 0$. Show $y = z$, using only the theorems proved above, and identify each step's justification. Solution. From $x+y=x+z$, additive cancellation (Theorem: Cancellation Laws, part (i)) with the common term $x$ gives

That already answers the question — we did not even need the second hypothesis. But let's use it as a consistency check. Since we've shown $y=z$, certainly $y\cdot w = z \cdot w$ holds trivially (substitute $z$ for $y$), matching the given data. So the extra hypothesis $y\cdot w = z\cdot w$, $w\neq 0$ is consistent with, but not needed to derive, $y=z$; multiplicative cancellation on $y\cdot w = z\cdot w$ (valid since $w \ne 0$) gives the same conclusion $y=z$ redundantly. This illustrates a subtlety: in a field, once one piece of given data pins down an equality, other given data must be consistent with it — a field has no room for contradictory equalities among cancellable quantities. Each step used: additive cancellation (part (i)), then, for the redundant check, multiplicative cancellation (part (ii)) with $w \neq 0$ playing the role of "$a\neq0$" in that theorem.

Example 2 (deriving a genuinely new identity: $(-1)\cdot a = -a$, then applying zero-product). Let $F$ be a field. Prove that for every $a \in F$, $(-1)\cdot a = -a$ (where $-1$ is the additive inverse of $1$, and $-a$ is the additive inverse of $a$). Then use this, together with the zero-product property, to solve for all $x \in F$ satisfying

in the specific field $F = \mathbb{Q}$, where "$3$" denotes $1+1+1$. Solution. First, the identity. By (A4), $1 + (-1) = 0$. Multiply both sides by $a$:

By the absorbing lemma, the right side is $0$. By (D) and (M2)/(M3) on the left,

So $(-1)\cdot a$ is an additive inverse of $a$: it satisfies $a + \big((-1)\cdot a\big) = 0$. By uniqueness of additive inverses, $(-1)\cdot a = -a$. Now apply this to the equation. Write $u = x + (-1)\cdot 3 = x - 3$ and $v = x + (-1) = x - 1$ (using the identity just proved to rewrite $(-1)\cdot 3$ as $-3$ and $(-1)\cdot 1 = -1$). The equation is $u\cdot v = 0$. By the zero-product property (proved above, valid in any field, in particular $\mathbb{Q}$), either $u = 0$ or $v=0$:

Both candidates lie in $\mathbb{Q}$, so the full solution set is $\{1, 3\}$. Note how the argument is entirely axiom-driven up to the identification "$-1 \cdot a = -a$"; only the very last step ("solve $x - 3 = 0$") uses that $\mathbb{Q}$ actually contains an element called $3$ that is the sum $1+1+1$, and that this sum is nonzero in $\mathbb{Q}$ — a fact that will be justified once $\mathbb{Q}$ is shown to satisfy the field axioms in The Rational Numbers as an Ordered Field.