Ma · Mathematics

Why Is the Jacobian of a Polynomial Isomorphism a Nonzero Constant?

Published By the 9to5Science editorial team

At a glance

  1. 1. Use the inverse

    Let ψ be the polynomial inverse of φ.

  2. 2. Apply the chain rule

    Differentiate ψ composed with φ to get a product of Jacobian matrices equal to the identity.

  3. 3. Take determinants

    The Jacobian determinants multiply to 1, so Jφ is a unit in the polynomial ring.

  4. 4. Identify the units

    The only units in a polynomial ring over a field are nonzero constants.

Advertisement

Assume the usual setting: ϕ\phi is a polynomial isomorphism of affine nn-space over a field kk, and JϕJ_\phi denotes the determinant of its Jacobian matrix. The key point is that an isomorphism has a polynomial inverse. Differentiating the composition of the map and its inverse shows that JϕJ_\phi has a multiplicative inverse in the polynomial ring k[x1,…,xn]k[x_1,\ldots,x_n]. The units of that ring are exactly the nonzero elements of kk, so JϕJ_\phi must be a nonzero constant.

Apply the chain rule to the inverse

Write the coordinate functions of ϕ\phi as

ϕ(x1,…,xn)=(f1(x),…,fn(x)),\phi(x_1,\ldots,x_n)=(f_1(x),\ldots,f_n(x)),

where each fif_i is a polynomial. Because ϕ\phi is an isomorphism, there is an inverse polynomial map ψ=ϕ−1\psi=\phi^{-1}, with ψ∘ϕ\psi\circ\phi equal to the identity map.

For a polynomial map F=(F1,…,Fn)F=(F_1,\ldots,F_n), its Jacobian matrix is

DF(x)=(∂Fi∂xj(x))i,j,D F(x)=\left(\frac{\partial F_i}{\partial x_j}(x)\right)_{i,j},

and its Jacobian determinant is JF(x)=det⁡DF(x)J_F(x)=\det D F(x). The chain rule for the composition gives

D(ψ∘ϕ)(x)=Dψ(ϕ(x))Dϕ(x).D(\psi\circ\phi)(x)=D\psi(\phi(x))D\phi(x).

Since ψ∘ϕ\psi\circ\phi is the identity, its Jacobian matrix is the identity matrix. Taking determinants yields

1=Jψ(ϕ(x))Jϕ(x).1=J_\psi(\phi(x))J_\phi(x).

Both factors are polynomials in x1,…,xnx_1,\ldots,x_n. Thus JϕJ_\phi has a polynomial multiplicative inverse: namely, Jψ∘ϕJ_\psi\circ\phi.

Why this forces a constant

In a polynomial ring over a field, a polynomial is invertible only if it is a nonzero constant. For example, if nonzero polynomials pp and qq satisfy pq=1pq=1, then the total degree of the product is the sum of their total degrees. The product has degree zero, so both pp and qq must have degree zero. Consequently,

Jϕ∈k×,J_\phi\in k^\times,

where k×k^\times denotes the nonzero elements of $k. This proves that the Jacobian determinant is a non-zero constant.

The same argument can be written as a direct calculation. For any point xx, the chain rule gives

Dψ(ϕ(x))Dϕ(x)=I,det⁡(Dψ(ϕ(x)))det⁡(Dϕ(x))=1.\begin{aligned} D\psi(\phi(x))D\phi(x)&=I,\\ \det(D\psi(\phi(x)))\det(D\phi(x))&=1. \end{aligned}

The determinant is therefore not merely nonzero at each point: it is invertible as a polynomial, which is the stronger fact needed to conclude it is constant.

Worked example

Consider the polynomial map ϕ:A2→A2\phi:\mathbb{A}^2\to\mathbb{A}^2 defined by

ϕ(x,y)=(x,y+x2).\phi(x,y)=(x,y+x^2).

Its inverse is ψ(x,y)=(x,y−x2)\psi(x,y)=(x,y-x^2), so ϕ\phi is an isomorphism. Its Jacobian matrix and determinant are

Dϕ(x,y)=(102x1),Jϕ(x,y)=1.D\phi(x,y)= \begin{pmatrix} 1&0\\ 2x&1 \end{pmatrix}, \qquad J_\phi(x,y)=1.

The inverse has Jacobian determinant 11 as well, consistent with the chain-rule identity. More generally, a polynomial automorphism can have any nonzero constant as its Jacobian determinant; for instance, an invertible linear map has determinant equal to the determinant of its matrix.

Common mistakes and scope

  • Showing only that the determinant is nonzero at every point: Pointwise nonvanishing alone is not the argument. The chain rule shows that the determinant has a polynomial inverse. That makes it a unit in the coordinate ring.
  • Assuming a polynomial map with nonzero constant Jacobian is automatically an isomorphism: The proof establishes a necessary condition, not a general converse. In characteristic zero and dimension at least two, the converse for polynomial maps of affine space is the Jacobian conjecture, an open problem. In positive characteristic it can fail: over a field of characteristic pp, the one-variable polynomial x+xpx+x^p has derivative 11 but is not a polynomial automorphism.
  • Using the statement without specifying what JJ means: Here JJ is the determinant of the Jacobian matrix of a polynomial map An→An\mathbb{A}^n\to\mathbb{A}^n. For maps between more general varieties, the appropriate Jacobian or determinant depends on the varieties and the chosen local coordinates; the affine-space argument applies directly in the stated setting.

Spotted an error? Send a correction. Read how we write and check articles in our editorial policy.