betaFeedback

WorksMathematics

A shape of maximal degree 150 in the plane Jacobian problem: the point (18/5,8) is not a regular corner

Abstract

In the list of Guccione, Guccione, Horruitiner and Valqui of the possible shapes of a counterexample to the plane Jacobian conjecture with degrees at most 150, one shape of maximal degree 150 consists of the standard (3,2)-pairs whose chain of corners is (10,40), (18/5,8), (8/5,3). We show that (18/5,8) is not a regular corner of any (3,2)-pair in K[x^{+-1/5},y]; hence, by a theorem of the four authors, no such pair exists. As in the note hexagon:2611.00709 on the degree pair (126,84), the proof of the theorem uses only results of Guccione-Guccione-Valqui, applied to every root of the edge polynomial, together with elementary arithmetic, which is reproduced by an ancillary script. The argument has not yet been examined by specialists.

Statement

Throughout, KK is an algebraically closed field of characteristic zero, L(l)=K[x±1/l,y]L^{(l)} = K[x^{\pm1/l},y], [P,Q]=PxQy−PyQx[P,Q] = P_xQ_y-P_yQ_x and (m,n)=(3,2)(m,n)=(3,2). We use the notation of [1]: vρ,σv_{\rho,\sigma}, ℓρ,σ\ell_{\rho,\sigma}, st⁡ρ,σ\operatorname{st}_{\rho,\sigma}, en⁡ρ,σ\operatorname{en}_{\rho,\sigma}; the interval of directions I=](1,−1),(1,0)]I=](1,-1),(1,0)]; (m,n)(m,n)-pairs in L(l)L^{(l)} and their regular corners [1 Definitions 4.3 and 5.5]. For a regular corner ((a/l,b),(ρ,σ))((a/l,b),(\rho,\sigma)) of (P,Q)(P,Q) we write ℓρ,σ(P)=xk/lp(z)\ell_{\rho,\sigma}(P)=x^k/lp(z) with z:=x−σ/ρyz:=x^{-\sigma/\rho}y, and we distinguish, as in [1 Section 5, before Remark 5.9]: Case I, [ℓρ,σ(P),ℓρ,σ(Q)]≠0[\ell_{\rho,\sigma}(P),\ell_{\rho,\sigma}(Q)]\ne0; Case II, the bracket vanishes and pp has at least two distinct linear factors, the factor zz included (this is how we read the condition # factors(p)>1(p)>1 of [1]); Case III, the bracket vanishes and p=α(z−β)rp=\alpha(z-\beta)^r with β≠0\beta\ne0. One always has v1,−1(st⁡ρ,σ(P))≠0v_{1,-1}(\operatorname{st}_{\rho,\sigma}(P))\ne0 (loc. cit., by [1 Theorem 2.6(4)]); Case II is called II.a) if this number is negative and II.b) if it is positive.

Theorem 1.1. No (3,2)(3,2)-pair in L(5)L^{(5)} has a regular corner at the point (18/5,8)(18/5,8).

Corollary 1.2. There is no standard (3,2)(3,2)-pair (P,Q)(P,Q) whose chain of corners, in the sense of [2 Theorem 2.20], contains Ai=(18/5,8)A_i=(18/5,8) with li=5l_i=5. In particular the shape with chain (10,40)(10,40), (18/5,8)(18/5,8), (8/5,3)(8/5,3) and maximal degree 150 of [2 Section 6] does not occur.

Proof. By items (3) and (6) (or (9)) of [2 Theorem 2.20], (Ai,(ρi,σi))(A_i,(\rho_i,\sigma_i)) is a regular corner of the (3,2)(3,2)-pair (Pi,Qi)(P_i,Q_i) in L(li)L^{(l_i)}, which contradicts Theorem 1.1 when li=5l_i=5. For the chain (10,40)(10,40), (18/5,8)(18/5,8), (8/5,3)(8/5,3), which is (A0,A1,A2)(A_0,A_1,A_2) in the notation of that theorem: l0=1l_0=1, and the corner (A0,(ρ0,σ0))(A_0,(\rho_0,\sigma_0)) is of Case II (item (6)). In Case II.a) item (7) gives l1=1l_1=1, hence A1∈Z×N0A_1\in\mathbb{Z}\times\mathbb{N}_0, which is false; so we are in Case II.b) and item (8) applies: A1A_1 lies on the line of direction (−σ0/ρ0,1)(-\sigma_0/\rho_0,1) through (k0/m,0)(k_0/m,0), and so does A0=1men⁡ρ0,σ0(P)A_0=\frac{1}{m}\operatorname{en}_{\rho_0,\sigma_0}(P). Since A0−A1=(32/5,32)A_0-A_1=(32/5,32), we get (ρ0,σ0)=(5,−1)(\rho_0,\sigma_0)=(5,-1) and l1=lcm⁡(5,1)=5l_1=\operatorname{lcm}(5,1)=5.

Remark 1.3. If (P,Q)(P,Q) is a (2,3)(2,3)-pair with a regular corner, then (Q,P)(Q,P) is a (3,2)(3,2)-pair with the same regular corner [1 Definition 4.3, Proposition 4.6(2),(3), Corollary 5.7(2),(3)]; so Theorem 1.1 also holds for (2,3)(2,3)-pairs. The list of [2 Section 6] gives only one of the two orders of (m,n)(m,n); for the same chain with (m,n)=(2,3)(m,n)=(2,3), of degrees (100,150)(100,150), [2 Theorem 2.20] gives (2,3)(2,3)-pairs (Pi,Qi)(P_i,Q_i) in L(5)L^{(5)} with a regular corner at (18/5,8)(18/5,8), which is excluded in the same way.

The tools

We use the following statements, for an (m,n)(m,n)-pair (P,Q)(P,Q) in L(l)L^{(l)} and a regular corner ((a/l,b),(ρ,σ))((a/l,b),(\rho,\sigma)).

(T1) [1 Proposition 5.19]: the point (a/l,b)(a/l,b) is the first entry of a regular corner of Case I or of Case II of a (possibly different) (m,n)(m,n)-pair in L(l)L^{(l)}; in Case I, l−a/b>1l-a/b>1; in Case II, gcd⁡(a,b)>1\gcd(a,b)>1. The directions of these corners are not larger than (ρ,σ)(\rho,\sigma) (proof of that proposition, which uses [1 Proposition 5.17(3)]).

(T2) Case I. By [1 Proposition 1.13], vρ,σ(P)+vρ,σ(Q)=ρ+σv_{\rho,\sigma}(P)+v_{\rho,\sigma}(Q)=\rho+\sigma; moreover en⁡ρ,σ(P)=m(a/l,b)\operatorname{en}_{\rho,\sigma}(P)=m(a/l,b) by definition and en⁡ρ,σ(P)=mnen⁡ρ,σ(Q)\operatorname{en}_{\rho,\sigma}(P)=\frac{m}{n}\operatorname{en}_{\rho,\sigma}(Q) [1 Corollary 5.7(2)]. Hence

m(alρ+bσ)=mm+n(ρ+σ).(1)m\left(\frac{a}{l}\rho+b\sigma\right)=\frac{m}{m+n}(\rho+\sigma). \tag*{(1)}

which determines (ρ,σ)(\rho,\sigma), and vρ,σ(P)=mm+n(ρ+σ)v_{\rho,\sigma}(P)=\frac{m}{m+n}(\rho+\sigma), vρ,σ(Q)=nm+n(ρ+σ)v_{\rho,\sigma}(Q)=\frac{n}{m+n}(\rho+\sigma). By [1 Proposition 5.13], either 1mst⁡ρ,σ(P)=(a′/l,0)\frac{1}{m}\operatorname{st}_{\rho,\sigma}(P)=(a'/l,0) with a′∈Na'\in\mathbb{N} (statement and proof), or {st⁡ρ,σ(P),st⁡ρ,σ(Q)}={(k/l,0),(1−k/l,1)}\{\operatorname{st}_{\rho,\sigma}(P),\operatorname{st}_{\rho,\sigma}(Q)\}=\{(k/l,0),(1-k/l,1)\} with k∈Nk\in\mathbb{N}.

(T3) Case II. By [1 Proposition 5.14], with d:=gcd⁡(a,b)d:=\gcd(a,b), aˉ:=a/d\bar{a}:=a/d, bˉ:=b/d\bar{b}:=b/d: d>1d>1, and there is μ∈N\mu\in\mathbb{N} with d∤μd\nmid\mu, μ≤l(bl−a)+1/bˉ\mu\le l(bl-a)+1/\bar{b}, such that (ρ,σ)(\rho,\sigma) is the direction orthogonal to (μaˉ−l,μbˉl−l)(\mu\bar{a}-l,\mu\bar{b}l-l); moreover ρ/gcd⁡(ρ,l)≤b\rho/\gcd(\rho,l)\le b. The point A′:=1mst⁡ρ,σ(P)A':=\frac{1}{m}\operatorname{st}_{\rho,\sigma}(P) lies in 1lZ×N0\frac{1}{l}\mathbb{Z}\times\mathbb{N}_0 (proof of that proposition), below (a/l,b)(a/l,b) on the line of direction (−σ/ρ,1)(-\sigma/\rho,1) through it, and v1,−1(A′)≠0v_{1,-1}(A')\ne0 as recalled above. One has ℓρ,σ(P)=λRm\ell_{\rho,\sigma}(P)=\lambda R^m for a (ρ,σ)(\rho,\sigma)-homogeneous R∈L(l)R\in L^{(l)} [1 Remark 3.1 and Corollary 5.7].

(T4) Case II.a). (A′,Pred⁡P(ρ,σ))(A',\operatorname{Pred}_P(\rho,\sigma)) is a regular corner of (P,Q)(P,Q) [1 Remark 5.11].

(T5) Case II, roots. Let λ∈K×\lambda\in K^\times be a root of pp of multiplicity mλm_\lambda with

mλ≥mlaρ+blσρ+σ.(2)m_\lambda\ge\frac{m}{l}\frac{a\rho+bl\sigma}{\rho+\sigma}. \tag*{(2)}

Then, with l′:=lcm⁡(ρ,l)l':=\operatorname{lcm}(\rho,l) and φ(y):=y+λxσ/ρ\varphi(y):=y+\lambda x^{\sigma/\rho}, the pair (φP,φQ)(\varphi P,\varphi Q) is an (m,n)(m,n)-pair in L(l′)L^{(l')} with a regular corner at (kml,0)+mλm(−σρ,1)(\frac{k}{ml},0)+\frac{m_\lambda}{m}(-\frac{\sigma}{\rho},1), of direction smaller than (ρ,σ)(\rho,\sigma) [1 Proposition 5.18].

The use of (T5) for a root other than the one continuing a chain, followed by (T1) and (T2), is modelled on the proof of [2 Proposition 6.1].

Arithmetic

In this section (m,n)=(3,2)(m,n)=(3,2). For a point (a/l,b)(a/l,b) we say that Case I is arithmetically impossible if, for the direction given by (1), none of the starting points allowed in (T2) exists, and that Case II is arithmetically impossible below (ρ∗,σ∗)(\rho_*,\sigma_*) if no μ\mu as in (T3) gives a direction of II smaller than (ρ∗,σ∗)(\rho_*,\sigma_*). Put

Cμ:=(25,0)+μ(25,1)=(2+2μ5,μ),1≤μ≤8,C_\mu:=\left(\frac{2}{5},0\right)+\mu\left(\frac{2}{5},1\right)=\left(\frac{2+2\mu}{5},\mu\right),\qquad1\le\mu\le8,

so that C8=(18/5,8)C_8=(18/5,8).

Lemma 3.1. (a) At (18/5,8)(18/5,8), l=5l=5: Case I is arithmetically impossible (direction (39,−17)(39,-17)); the directions allowed in Case II are (5,−2)(5,-2) and (15,−4)(15,-4), and none of them is smaller than (5,−2)(5,-2).

(b) At C1=(4/5,1)C_1=(4/5,1), l=5l=5: l−a/b=1l-a/b=1 and gcd⁡(4,1)=1\gcd(4,1)=1.

(c) At C2=(6/5,2)C_2=(6/5,2), l=5l=5: Case I is arithmetically impossible (direction (9,−5)(9,-5)); the only direction allowed in Case II below (5,−2)(5,-2) is (2,−1)(2,-1).

(d) At C4=(10/5,4)C_4=(10/5,4), l=5l=5: Case I is arithmetically impossible (direction (19,−9)(19,-9)), and so is Case II below (5,−2)(5,-2).

(e) At C5=(12/5,5)C_5=(12/5,5) and C7=(16/5,7)C_7=(16/5,7), l=5l=5: Case I is arithmetically impossible (directions (24,−11)(24,-11) and (34,−15)(34,-15)), and gcd⁡(12,5)=gcd⁡(16,7)=1\gcd(12,5)=\gcd(16,7)=1.

(f) At (7/10,1)(7/10,1), l=10l=10: Case I is arithmetically impossible (direction (8,−5)(8,-5)), and gcd⁡(7,1)=1\gcd(7,1)=1.

(g) At (14/5,5)(14/5,5), ℓ=5\ell=5: Case I is arithmetically impossible (direction (24,−13)(24,-13)), and gcd⁡(14,5)=1\gcd(14,5)=1.

Proof. Direct computation, checked by the ancillary script; for instance in (c), (1) reads (65−15)ρ+(2−15)σ=0(\frac{6}{5}-\frac{1}{5})\rho+(2-\frac{1}{5})\sigma=0, so (ρ,σ)=(9,−5)(\rho,\sigma)=(9,-5), v(P)=125v(P)=\frac{12}{5}, v(Q)=85v(Q)=\frac{8}{5}, and 95⋅3a′=125\frac{9}{5}\cdot3a'=\frac{12}{5}, 95k=85\frac{9}{5}k=\frac{8}{5}, 95k=125\frac{9}{5}k=\frac{12}{5} have no integer solution. □

By (T1), (T2), (T3) and Lemma 3.1:

none of the points C1C_1, C5C_5, C7C_7, (14/5,5)(14/5,5) (ℓ=5\ell=5) is the first entry of a regular corner of a (3,2)(3,2)-pair in L(5)L^{(5)}, and C4C_4 (ℓ=5\ell=5) is not the first entry of a regular corner of direction smaller than (5,−2)(5,-2).

Lemma 3.2. No (3,2)(3,2)-pair in L(5)L^{(5)} has a regular corner at C2=(6/5,2)C_2=(6/5,2) of direction smaller than (5,−2)(5,-2).

Proof. By (T1) and Lemma 3.1(c) there would be a (3,2)(3,2)-pair (P′,Q′)(P',Q') in L(5)L^{(5)} with a regular corner (C2,(2,−1))(C_2,(2,-1)) of Case II. Put z:=x1/2yz:=x^{1/2}y. The point A′A' of (T3) is C2−h(1/2,1)C_2-h(1/2,1) with h∈{1,2}h\in\{1,2\} and lies in 15Z×N0\frac{1}{5}\mathbb{Z}\times\mathbb{N}_0, so h=2h=2 and A′=(1/5,0)A'=(1/5,0). Thus ℓ2,−1(P′)=x3/5p(z)\ell_{2,-1}(P')=x^{3/5}p(z) with deg⁡p=6\deg p=6, p(0)≠0p(0)\ne0. By (T3), ℓ2,−1(P′)=λ′R3\ell_{2,-1}(P')=\lambda'R^3 with R=x1/5r(z)∈L(5)R=x^{1/5}r(z)\in L^{(5)}, deg⁡r=2\deg r=2, r(0)≠0r(0)\ne0; since x1/5zi=x(2+5i)/10yix^{1/5}z^i=x^{(2+5i)/10}y^i lies in L(5)L^{(5)} only for ii even, rr is a multiple of z2−cz^2-c with c≠0c\ne0, and p=λ′(z2−c)3p=\lambda'(z^2-c)^3. A root λ\lambda of z2=cz^2=c has mλ=3m_\lambda=3, and the right-hand side of (2) is 35⋅6⋅2−2⋅51=65\frac{3}{5}\cdot\frac{6\cdot2-2\cdot5}{1}=\frac{6}{5}. By (T5) some (3,2)(3,2)-pair in L(10)L^{(10)} has a regular corner at (15,0)+(12,1)=(710,1)(\frac{1}{5},0)+(\frac{1}{2},1)=(\frac{7}{10},1), which contradicts (T1), (T2) and Lemma 3.1(f). □

Proof of Theorem 1.1

Assume that some (3,2)(3,2)-pair in L(5)L^{(5)} has a regular corner at A1:=(18/5,8)A_1:=(18/5,8). By (T1) and Lemma 3.1(a) there is a (3,2)(3,2)-pair (P,Q)(P,Q) in L(5)L^{(5)} with a regular corner (A1,(ρ,σ))(A_1,(\rho,\sigma)) of Case II, with (ρ,σ)∈{(5,−2),(15,−4)}(\rho,\sigma)\in\{(5,-2),(15,-4)\}. Let A′A' be the point of (T3).

The direction (15,−4)(15,-4). Here A′=A1−h(4/15,1)A'=A_1-h(4/15,1) with 0<h≤80<h\le8, and A′∈15Z×N0A'\in\frac{1}{5}\mathbb{Z}\times\mathbb{N}_0 forces h∈{3,6}h\in\{3,6\}: A′=(14/5,5)A'=(14/5,5) or A′=(2,2)A'=(2,2). The second is excluded by v1,−1(A′)≠0v_{1,-1}(A')\ne0. In the first, v1,−1(A′)<0v_{1,-1}(A')<0, so we are in Case II.a) and (14/5,5)(14/5,5) is the first entry of a regular corner by (T4), contrary to (3).

The direction (5,−2)(5,-2). Put z:=x2/5yz:=x^{2/5}y. Now A′=A1−h(2/5,1)=Cb′A'=A_1-h(2/5,1)=C_{b'} with b′=8−h∈{0,…,7}b'=8-h\in\{0,\ldots,7\} (and C0:=(2/5,0)C_0:=(2/5,0)). The line of the edge meets the first axis at (2/5,0)(2/5,0), so ℓ5,−2(P)=x6/5p(z)\ell_{5,-2}(P)=x^{6/5}p(z) with p=z3b′p~(z)p=z^{3b'}\widetilde p(z), p~(0)≠0\widetilde p(0)\ne0, deg⁡p~=3(8−b′)\deg\widetilde p=3(8-b'). Since ℓ5,−2(P)\ell_{5,-2}(P) is a cube by (T3),

p~(z)=λ′∏i(z−ci)3μi,ci∈K× distinct,μi≥1,∑iμi=8−b′.\widetilde p(z)=\lambda'\prod_i(z-c_i)^{3\mu_i},\qquad c_i\in K^\times\text{ distinct},\quad\mu_i\ge1,\qquad\sum_i\mu_i=8-b'.

The right-hand side of (2) is 35⋅18⋅5−8⋅5⋅23=2\frac{3}{5}\cdot\frac{18\cdot5-8\cdot5\cdot2}{3}=2, so every root of p~\widetilde p, of multiplicity 3μi≥33\mu_i\ge3, satisfies (2). A root with μi=8\mu_i=8 would force b′=0b'=0 and p=λ′(z−ci)24p=\lambda'(z-c_i)^{24}, which has a single linear factor and is not Case II; so μi≤7\mu_i\le7. By (T5), for every ii some (3,2)(3,2)-pair in L(5)L^{(5)} has a regular corner at CμiC_{\mu_i}, of direction smaller than (5,−2)(5,-2). By (3) and Lemma 3.2, μi∉{1,2,4,5,7}\mu_i\notin\{1,2,4,5,7\}, so μi∈{3,6}\mu_i\in\{3,6\} for every ii. Then 8−b′∈{3,6}8-b'\in\{3,6\}, that is, A′=C5=(12/5,5)A'=C_5=(12/5,5) or A′=C2=(6/5,2)A'=C_2=(6/5,2). In both cases v1,−1(A′)<0v_{1,-1}(A')<0, so we are in Case II.a) and, by (T4), A′A' is the first entry of a regular corner of (P,Q)(P,Q) of direction smaller than (5,−2)(5,-2). This contradicts (3) for C5C_5 and Lemma 3.2 for C2C_2. □

Remark 4.1. The corner C3=(8/5,3)C_3=(8/5,3), which ends the chain of [2], passes the tests used here: Case I is arithmetically possible there, with direction (2,−1)(2,-1). The contradiction comes from the other roots of the edge polynomial and from the starting point of the edge.

Scope

  • Proved here, relative to the cited results: Theorem 1.1 and Corollary 1.2. The inputs are (T1)–(T5), that is [1 Theorem 2.6(4), Remark 2.8, Definitions 4.3, 5.5, Propositions 1.13, 4.6, 5.13, 5.14, 5.17(3), 5.18, 5.19, Corollary 5.7, Remarks 3.1, 5.11, and the discussion of the cases before Remark 5.9], and [2 Theorem 2.20] for the corollary. The reference [2] is an arXiv preprint; of its Theorem 2.20 we use items (3), (6), (8), (9), and item (7) only at the first corner, where l0=1l_0 = 1. Three of these inputs are taken from proofs rather than statements, as indicated in § 2: the monotonicity of the directions in (T1), the integrality of a′a' in (T2), and the lattice of A′A' in (T3).

  • Verified by exact computation: Lemma 3.1 and the numerical values in the proofs (ancillary file verify_150.py in the folder anc; Python standard library only).

  • Not claimed: a statement on the plane Jacobian conjecture itself, or on the other shapes of [2]; a new bound on the degree of a counterexample (the published one, [3], concerns degrees below 125 and does not involve this shape). The completeness of the list of [2] is not re-examined here.

  • The argument has been checked only by the author with AI assistance (Claude, Anthropic); it has not been validated by professionals.

References

Independent researcher, France

References

  1. [1]C. Valqui, J. A. Guccione, J. J. Guccione, On the shape of possible counterexamples to the Jacobian Conjecture, J. Algebra 471 (2017), 13–74, doi:10.1016/j.jalgebra.2016.08.039; arXiv:1401.1784v3. Author order as in the journal; statement numbers refer to the arXiv version v3.DOI
  2. [2]J. A. Guccione, J. J. Guccione, R. Horruitiner, C. Valqui, Some algorithms related to the Jacobian Conjecture, arXiv:1708.07936v1.arxiv.org/abs/1708.07936
  3. [3]J. A. Guccione, J. J. Guccione, R. Horruitiner, C. Valqui, Increasing the degree of a possible counterexample to the Jacobian Conjecture from 100 to 108, arXiv:2204.14178.arxiv.org/abs/2204.14178

Paper details

vlnc:W_xkgjtte8
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Anthropic Claude Fable

View statement
Looking for related papers…
Contents