If `p ≡ ±1 (mod 7)` then `f mod p` is *not* irreducible: `p` splits.
ProvedCyclicCubic.not_irreducible_mod_of_pm_oneaether-catalogapplications
If p ≡ ±1 (mod 7) then f mod p is not irreducible: p splits.
theorem CyclicCubic.not_irreducible_mod_of_pm_one(hpm : (p : ZMod 7) = 1 ∨ (p : ZMod 7) = 6) :
¬ Irreducible (fpoly (ZMod p)) := by sorry
Formalization Note Transplanted verbatim from the Aether Catalog source Applications/CyclicCubicTypeChannel/Splitting.lean; the statement is byte-identical to the source declaration, elaborated with autoImplicit disabled in the platform environment.
Preamble
-- Thm stub generated from Applications/CyclicCubicTypeChannel/Splitting.lean
import Mathlib
import Definitions.Def_Applications_CyclicCubicTypeChannel_Splitting
/-
# The cyclic cubic field `ℚ(ζ₇ + ζ₇⁻¹)` has exactly two splitting types
## Context (FACT round-32 #3, "THE-CYCLIC-CUBIC-IS-FULLY-PINNED", paper 122)
The real subfield `K = ℚ(ζ₇ + ζ₇⁻¹)` of the seventh cyclotomic field is the
cyclic cubic field of conductor `7`. Its ring of integers is `ℤ[α]` with
`α = ζ₇ + ζ₇⁻¹` a root of
`f(X) = X³ + X² − 2X − 1` (discriminant `49`).
By Dedekind's factorisation criterion the splitting type of an unramified
rational prime `p ≠ 7` in `K` is read off from the factorisation of `f mod p`.
This file proves, from scratch and with no number-field machinery, the complete
arithmetic law behind the experiment:
* `CyclicCubic.root_iff` — for a prime `p ≠ 7`, `f` has a root in `ZMod p`
**iff** `p ≡ ±1 (mod 7)`;
* `CyclicCubic.splits_completely` — one root forces three *distinct* roots
(the map `x ↦ x² − 2` cycles them), so `f mod p` either splits completely or
is irreducible: **only two types**;
* `CyclicCubic.irreducible_mod_of_not_pm_one` — the inert case;
* `CyclicCubic.resDeg_congr` — the residue degree is a function of `p mod 7`
alone: the arithmetic form of **full pinning**;
* `CyclicCubic.irreducible_rat`, `CyclicCubic.minpoly_zeta_add_inv` — `f` is
irreducible over `ℚ` and is the minimal polynomial of `ζ₇ + ζ₇⁻¹`, so `K`
really is a cubic field.
The hard direction ("a root forces `p ≡ ±1`") and the hard existence direction
("`p ≡ −1` forces a root") are both proved by transporting the question into
the group `GL₂(𝔽_p)`: the companion matrix of `Y² − xY + 1` has order `7`
exactly when `x` is a root of `f`, and Cauchy's theorem supplies an order-`7`
matrix in the converse direction, whose trace is then forced to be a root of
`f` by a Cayley–Hamilton recursion.
-/
open Matrix Polynomial
open CyclicCubic
/-! ## The defining cubic -/
/-! ## The `y + y⁻¹` substitution -/
/-! ## Small decidable facts about `ZMod 7` -/
/-! ## `2 × 2` matrix toolkit -/
variable {R : Type*} [CommRing R]
/-! ## The splitting criterion -/
variable (p : ℕ) [hp : Fact p.Prime]
/-! ## One root forces three: only two types -/
variable {R : Type*} [CommRing R] [IsDomain R]
/-! ## Irreducibility: the inert type -/
variable (p : ℕ) [hp : Fact p.Prime]Formal statement
theorem CyclicCubic.not_irreducible_mod_of_pm_one(hpm : (p : ZMod 7) = 1 ∨ (p : ZMod 7) = 6) :
¬ Irreducible (fpoly (ZMod p)) := by sorrySource