The Resultant, Episode 5: Inside the Episode

The double-product form for the resultant:

(1)

implies Fact 3:

- res
(_{x}*y*_{0})=0 ⇔ [the line*y*=*y*_{0}passes through an intersection of*E*and*F*, or*a*(_{m}*y*_{0})=*b*(_{n}*y*_{0})=0].

*provided* it makes sense to evaluate the *u _{i}*‘s and

*v*‘s at

_{j}*y*

_{0}. If you’ll grant me that, then here goes the argument. In one direction: if res

*(*

_{x}*y*

_{0})=0 and

*a*(

_{m}*y*

_{0})≠0 and

*b*(

_{n}*y*

_{0})≠0, then some factor

*u*(

_{i}*y*

_{0})–

*v*(

_{j}*y*

_{0}) must equal 0. (What if

*a*(

_{m}*y*

_{0})≠0 but

*b*(

_{n}*y*

_{0})=0? See below.) Let

*x*

_{0}=

*u*(

_{i}*y*

_{0})=

*v*(

_{j}*y*

_{0}). Plugging into the factored forms for

*E*(

*x*,

*y*) and

*F*(

*x*,

*y*):

*E*(*x,y*) = *a _{m}*(

*y*)(

*x–u*

_{1})···(

*x–u*) (3a)

_{m}*F*(

*x,y*) =

*b*(

_{n}*y*)(

*x–v*

_{1})···(

*x–v*) (3b)

_{n}gives *E*(*x*_{0},*y*_{0})=*F*(*x*_{0},*y*_{0})=0, so (*x*_{0},*y*_{0}) is an intersection of *E* and *F*. This argument is easily reversed to prove the other direction.

This *near-proof* or *plausibility proof* has some wrinkles worth investigating. First off, it seems from eq.(1) that *a _{m}*(

*y*

_{0})=0

*or b*(

_{n}*y*

_{0})=0 should force res

*(*

_{x}*y*

_{0})=0—why do we need them

*both*to be zero? The simplest example exhibiting the flaw in this ointment:

*x*=

*y*,

*xy*=1, i.e.,

*E*(

*x*,

*y*)=

*x–y*,

*F*(

*x*,

*y*)=

*yx–*1. When

*y*

_{0}=0, the leading coefficient of

*F*is 0. The double product is (

*y*–(1/

*y*))=(

*y*

^{2}–1)/

*y*. The resultant is

*y*times that, that is

*y*

^{2}–1; thus plugging in

*y*

_{0}=0 gives res

*(*

_{x}*y*

_{0})=–1. If we hadn’t canceled

*y*first. we’d have gotten 0·∞.

A more elaborate example: *E*(*x*)=*ax*^{2}+*bx*+*c*, *F*(*x*)=*dx*+*e*. Using the approximation , valid for |*b*|≫|*ac*|, the roots of *E*(*x*)=0 are approximately

Keeping *b* and *c* fixed (with *b*≠0) while *a*→0, one of the roots of *E*(*x*)=0 tends to –*c*/*b* while the other wanders off to infinity. *F*(*x*)=0 has the fixed root –*e*/*d*. When *a*=0, *E* morphs into a linear equation, sharing a root with *F* precisely when *c*/*b=e*/*d*, i.e., when *be*–*cd*=0. We’ve seen this before: it’s the Sylvester determinant:

But if we plug *a*=0 into the Sylvester determinant for the original pair *E* and *F*, we have:

Let’s look at the general case, but let’s get rid of those pesky leading coefficients by dividing them out:

*E*′(*x*) = *x ^{m}* + (

*a*

_{m–1}/

*a*)

_{m}*x*

^{m–1}+···+ (

*a*

_{0}/

*a*) (3a′)

_{m}*F*′(

*x*) =

*x*+ (

^{n}*b*

_{n–1}/

*b*)

_{n}*x*+···+ (

^{n}*b*

_{0}/

*b*) (3b′)

_{n}Say the original coefficients are polynomials in *y*. We’ve introduced poles at any of the roots of *a _{m}*(

*y*)=0 or

*b*(

_{n}*y*)=0. If, say,

*b*(

_{n}*y*

_{0})=0, then the

*actual*degree of

*F*(

*x*,

*y*

_{0}) in

*x*is less than

*n*, but (you might say) its

*formal*degree is still

*n*. (Some people do say this.) Whereas

*F*(

*x*,

*y*

_{0})=0 has fewer than

*n*roots,

*F*′(

*x*,

*y*

_{0})=0 is blessed (or cursed) with additional roots at infinity.

Now, if *a _{m}*(

*y*

_{0}) is

*not*0, then all the roots of

*E*′(

*x*,

*y*

_{0})=0 are finite. If we’re asking after common roots, we can ignore the infinite roots of

*F*′(

*x*,

*y*

_{0}).

How does the resultant view this? The roots at infinity bedevil the double product in eq.(1), but the leading coefficient factors smooth the troubled waters. If we use the actual degrees, we’re rewarded with a smaller Sylvester determinant. As long as *a _{m}*(

*y*

_{0})≠0 and

*b*(

_{n}*y*

_{0})=0 (or vice versa), the “formal” and “actual” determinants give the same verdict for the existence of common roots: one equation has only finite roots, the other just some extra roots at infinity. If both

*a*(

_{m}*y*

_{0}) and

*b*(

_{n}*y*

_{0}) are 0, then we have common roots at infinity, and det(

*S*) is 0.

This is easy to see from the determinant itself, using expansion by minors down the first column. Say *a _{m}*(

*y*

_{0})≠0 and

*b*(

_{n}*y*

_{0})=0. Letting

*m*=2,

*n*=3 as usual, to ease formatting, we have:

with the “formal” determinant on the left, and the “actual” one on the right. (At least if *b*_{2}(*y*_{0})≠0. If not, rinse, repeat.) So one determinant is 0 if and only if the other is.

OK, how about the other dodgy aspect of the plausibility proof, evaluating the *u _{i}*‘s and

*v*‘s at a value

_{j}*y*

_{0}of

*k*? The

*u*‘s and

_{i}*v*‘s belong to an algebraic extension

_{j}*L*of

*k*(

*y*). We

*sort of*have an evaluation map from

*k*(

*y*) to

*k*, at least when we steer clear of poles. It seems like one

*should*be able to extend this to a suitable map from

*L*to

*k*, since

*k*is algebraically closed. When

*k*=ℂ, the theory of Riemann surfaces handles the matter. For algebraically closed

*k*in general, I gather Puiseaux series step into the breach. Kendig discusses this in §§3.3–3.4.

While we’re on the topic of alternate proofs, Lang’s *Algebra* concludes the justification of eq.(1) (det(*S*)=double product) this way, adjusting his notation to agree with ours:

From (2) we see that [the double product] is homogeneous and of degree

nin [thea_{i}‘s], and homogeneous and of degreemin [theb_{j}‘s]. Since [det(S)] has exactly the same homogeneity properties, and is divisible by [the double product], it follows that [det(S)=c(double product)] for some integerc. Since both [det(S) and the double product] have a monomial [a_{m}^{n}b_{0}] occurring in them with coefficient 1, it follows that^{m}c=1, and our proposition is proved.

To my mind, the logic suffers from two gaps, although I think I see how to fill them. Recall the corresponding spot in the proof from Episode 4. We were looking at the special case with *a _{m}*=

*b*=1. We’d set

_{n}*R*=

*k*[

*u*

_{1}

*,…,u*,

_{m}*v*

_{1},…,

*v*]. We’d shown

_{n}, with *h*∈*R.*

Let’s write *R _{ab}* for the subring

*k*[

*a*

_{m}_{–1}

*,…,*

*a*

_{0}

*,*

*b*

_{n–1}

*,…,b*

_{0}] of

*R*. Note that det(

*S*) belongs to

*R*. If we knew that the double product and

_{ab}*h*also belonged to

*R*, we’d be home free.

_{ab}Galois theory shows that the double product belongs to *R _{ab}*. The equivalent expression belongs to

*R*[

_{ab}*u*

_{1}

*,…,u*], and is symmetric in the

_{m}*u*‘s; that should do the trick. We know that

_{i}*h*results from dividing one element of

*R*by another, without remainder in

_{ab}*R*. This ought to force

*h∈R*Ideals, Varieties, and Algorithms (Cox, Little, and O’Shea) provides a multivariable polynomial division algorithm; I haven’t checked the details, but I think this offers what we need.

_{ab}.So we can validate Lang’s argument, but it takes a bit of work. The approach in Episode 4 avoids these potholes, and as a bonus, introduces the “packing with *t*‘s” technique we’ll need later.

One last matter. We’ve already made use of the formula *PE+QF*=det(*S*), for some *P*∈*R _{n}*[

*x*],

*Q*∈

*R*[

_{m}*x*]. Proving this took some effort. It’s much easier to derive this result:

*PE+QF*=*D
*for some

*P*∈

*R*[

*x*],

*Q*∈

*R*[

*x*], and

*D*∈

*R*

Here’s the proof. Let *K* be the fraction field of *R*, as usual. Then *K*[*x*] is a PID. So if *E* and *F* are coprime, then for some *p,q*∈*K*[*x*],

*pE+qF*=1

Since the coefficients of *p* and *q* all belong to *K*, we can clear denominators and get *PE+QF*=*D.* Note that *D*≠0. When *R*=*k*[*y*], *D* is a polynomial in *y* alone, and the y-coordinates of the intersections of *E* and *F* are roots of *D*.

Now let’s think about the ideal (*E*,*F*) = {*PE+QF* : *P,Q*∈*R*[*x*]}. Intersecting this with *R* gives an ideal of *R*. When *R*=*k*[*y*], *R* is a PID, so (*E*,*F*)∩*k*[*y*] is principal. It’s nonzero because *D* belongs to it. In other words, we *can* eliminate *x* from the equations *E*(*x*,*y*)=0 and *F*(*x*,*y*)=0 and get something non-trivial.

We might as well write *D* for a generator of (*E*,*F*)∩*k*[*y*]. I used to believe that res* _{x}*(

*y*) was such a generator. Not true. We’ve seen an example already in Inside Episode 2 (with

*x*and

*y*switched): the two ellipses

*x*

^{2}+

*y*

^{2}=0 and 2

*x*

^{2}+

*y*

^{2}=0. Easy congruence calculations show that

*x*

^{2}≡ 0 mod (

*E*,

*F*), and as we saw way back in post 6, this amounts to showing that

*x*

^{2}belongs to (

*E*,

*F*). In fact (

*E*,

*F*)∩

*k*[

*x*] is generated by

*x*

^{2}. But res

*(*

_{y}*x*) , as we saw, is

*x*

^{4}.