---
title: Kraus's Error in Jacobian Conjecture
url: https://www.emergentmind.com/topics/kraus-s-error
type: topic
---

# Kraus's Error in Jacobian Conjecture

Kraus’s Error refers to the critical, historically significant mistake in L. Kraus’s 1884 attempted proof of what is now called the Jacobian Conjecture. Kraus’s reasoning correctly anticipated many later algebro-geometric strategies, but the flaw in his final analytic argument remains the core obstacle to current approaches. The error is centered on the control of ramification at infinity—specifically, the unjustified assumption that certain pole-parametrizations admit finite, nonzero derivatives at points corresponding to infinity on the Riemann sphere. The ramifications of the error have persisted for nearly 140 years, leaving the central obstruction in the path of a complete algebro-geometric resolution of the conjecture [2512.23614].

## 1. Formulation of Kraus’s Theorem

Kraus considered two polynomials $p(x, y), q(x, y) \in \mathbb{C}[x, y]$ satisfying a constant Jacobian condition:
$$
J(p, q) := p_x q_y - p_y q_x = \text{const} \neq 0.
$$
He asserted that the associated polynomial map
$$
\Phi: \mathbb{C}^2 \to \mathbb{C}^2,\quad (x, y) \mapsto (p(x, y), q(x, y))
$$
is bijective with a polynomial inverse. Equivalently, for $F(x, y) = p$, $G(x, y) = q$ and $\Delta(x, y) = \det \left [ \begin{smallmatrix} \partial F/\partial x & \partial F/\partial y \\ \partial G/\partial x & \partial G/\partial y \end{smallmatrix} \right ]$, Kraus claimed that if $\Delta(x, y)$ is a nonzero constant, then $F(x, y) = \eta, G(x, y) = \xi$ can be solved for $x, y$ as polynomials in $\xi, \eta$ [2512.23614].

## 2. Modern Reconstruction of Kraus’s Argument

Kraus’s argument can be structured in the following logically interdependent steps:

### A. Normal Form and Resultant

Transform $p, q$ so both are monic in $y$: $p(x, y) = y^{\deg p} + \ldots$, $q(x, y) = y^{\deg q} + \ldots$. Define the $y$-resultant
$$
R(x; u, v) := \operatorname{Res}_y(p(x, y) - u, q(x, y) - v)
$$
as a polynomial in $x$ with coefficients in $\mathbb{C}[u, v]$. Known results (Formanek, Adjamagbo–van den Essen, Płoski) establish that if $R$ has $x$-degree $n = 1$ and $J$ is constant, then $\Phi$ is a polynomial automorphism.

### B. Field Degree and Irreducibility

Abhyankar–McKay–Wang’s refinement implies
$$
R(x; u, v) = \alpha H(x; u, v)^\ell
$$
where $H$ is irreducible, and $\ell = [\mathbb{C}(x, y): \mathbb{C}(x, p, q)]$. By Formanek’s theorem, $J(p, q) = 1$ implies $\ell = 1$, so $R$ is irreducible. For each $c \in \mathbb{C}$, $R_c(x, v) = R(x; c, v) \in \mathbb{C}[x, v]$ remains irreducible and of fixed $x$-degree $n$.

### C. Analytic Argument at Finite Critical Values

For fixed $c$, the equation $R_c(x, v) = 0$ defines an $n$-valued analytic function $x = f(v)$ on $\mathbb{P}^1_v$, ramified only over finitely many roots of the discriminant and possibly at $v = \infty$. Local Puiseux expansion around any finite branch point $v_0$ shows, using the constancy of $J(p, q)$ and Cramer’s rule, that $v(t) \mapsto t$ is invertible, eliminating ramification at finite $v$.

### D. Ramification at Infinity

At infinity, one uses Newton–Puiseux parametrizations for the branches of the curve $p(x, y) = c$:
$$
x = t^m,\qquad y = u_\nu(t^m) = t^m(\ldots + o(1)), \quad \nu=0,\ldots, m-1.
$$
Defining $w_\nu(t) = q(x(t), y(t))$, expansion yields $w_\nu(t) = b_0 + b_1 t^{-1} + \cdots$, so $v = b_0$ is a “branch point at infinity.” Kraus’s analytic continuation argument attempts to resolve ramification at such $v$ by substituting $s = 1/t$ and invoking the inverse function theorem at $s = 0$.

## 3. The Fatal Gap: Ill-Defined Derivative at the Singularity

Kraus incorrectly claimed that $x(s) = s^{-m}$ possesses a well-defined, nonzero derivative at $s = 0$. However, $x(s)$ is singular at $s = 0$; $s = 0$ is an essential singularity or pole rather than an ordinary regular point. The derivative $x'(0)$ does not exist as a finite number, invalidating Kraus’s analytic argument. The consequence is the breakdown of control over ramification of $x = f(v)$ at $v = b_0$ (the “point at infinity”) [2512.23614].

## 4. Persistent Obstruction: Ramification at Infinity

The significance of Kraus’s error is that the inability to control ramification at infinity is precisely the main unsolved problem in all algebro-geometric approaches to the Jacobian Conjecture. The vanishing of the leading coefficient $r_n(u, v)$ of the resultant $R$ corresponds to the non-properness locus of $\Phi$, i.e., the possibility that there exist sequences $(x_i, y_i) \to \infty$ with $(p, q)(x_i, y_i) \to (u_0, v_0)$. Ensuring properness or “no ramification at infinity” requires that the leading coefficient never vanishes, a requirement for which current methods are inadequate. Kraus’s incorrect treatment of the singular parametrization at infinity anticipates this enduring difficulty [2512.23614].

## 5. Illustrative Example of the Failure

A model example highlighting the flaw is the resultant
$$
R_c(x, v) = (v - c)^m x - 1,
$$
with solution $x = 1/(v - c)^m$. Expanding near $v = c$, i.e., setting $s = v - c$, gives $x(s) = 1/s^m$, which again possesses a pole at $s = 0$. Any argument requiring $x'(0)$ is inapplicable: the apparent pathway for analytic continuation at the “branch point” fails, precisely paralleling Kraus’s reasoning breakdown [2512.23614].

## 6. Enduring Influence and Contemporary Perspective

Kraus’s 1884 argument is notable for its anticipation of descendant techniques—usage of normalization, resultants, branched covers, Puiseux expansions, and analytic function theory. Despite the single fatal mistake, the obstruction to controlling “infinite branches” (or ramification at infinity) flagged by Kraus remains the central unsolved hurdle in the algebraic and analytic study of polynomial automorphisms with constant Jacobian. All current algebro-geometric strategies must ultimately grapple with this singular point of failure. No subsequent advances have circumvented this precise difficulty, placing Kraus’s error at the heart of the continuing challenge posed by the Jacobian Conjecture [2512.23614].

Source: https://www.emergentmind.com/topics/kraus-s-error