Bourbaki

Scholium

Tag 06B7, permanent. Cite it as [Bourbaki, Tag 06B7].

Algebra, chapter IV, § 2. Zeros of polynomials

— Let A be an integral domain and f∈A[(Xi)i∈I]f \in A[(X_i)_{i \in I}]. Th. 2 provides a convenient means of proving that f=0f = 0. It suffices to consider an infinite integral domain E containing A as subring; if we can show that f((xi))=0f((x_i)) = 0 for all (xi)∈EI(x_i) \in E^I (or even for those (xi)∈EI(x_i) \in E^I at which a finite number of given non-zero polynomials do not vanish) then it follows that f=0f = 0. If A itself is not infinite, we can for example take E to be the ring A[X]A[X] or its field of fractions.

Once we have proved the relation f=0f = 0, we can clearly deduce that f((yi))=0f((y_i)) = 0 for all (yi)∈FI(y_i) \in F^I where F is any unital associative and commutative A-algebra whatsoever; in particular F may be finite or non-integral.

In other words, the proof of the identity f((xi))=0f((x_i)) = 0 when the xix_i run over an infinite integral domain containing A as subring (with the possible restriction that gk((xi))≠0g_k((x_i)) \neq 0 for 1≤k≤m1 \leq k \leq m, where the gkg_k are non-zero polynomials) implies the same identity when the xix_i run over any unital associative and commutative A-algebra.

In particular let f∈Z[(Xi)i∈I]f \in Z[(X_i)_{i \in I}]. If f((xi))=0f((x_i)) = 0 when the xix_i run over Z\mathbf{Z} (with the possible restriction that gk((xi))≠0g_k((x_i)) \neq 0 for 1≤k≤m1 \leq k \leq m, where the gkg_k are non-zero elements of Z[(Xi)]\mathbf{Z}[(X_i)]), then we have the same identity when the xix_i run over an arbitrary commutative ring.

Bourbaki puts the proof directly after the statement, so what is above is the statement and its proof together.

Cited by

Nothing.

Cites

Transcribed from the printed edition, 2003, Springer. Not translated.