Pyramidal groups are solvable
ProvedBFFPyramidal.isSolvable_of_isPyramidalEvery pyramidal group is solvable.
Indeed, in the defining chain each step has index equal to the least prime factor of the order of the preceding term, and a subgroup of least prime index is normal; so the chain is a composition series with cyclic factors of prime order, whence the group is solvable. This is why the class of pyramidal groups sits between the supersolvable groups (all of which are pyramidal) and the solvable groups.
Formalization note. Solvability is Mathlib's Group.IsSolvable. The trivial group is included and is solvable.
import Mathlib import Definitions.Def_BFFPyramidal_pyramidal
namespace BFFPyramidal
theorem isSolvable_of_isPyramidal {G : Type*} [Group G] [Finite G] (hG : IsPyramidal G) :
Group.IsSolvable G := by
sorry
end BFFPyramidal
Read-back
What the Lean code literally says, in plain math · claude-opus-5
Read-back: BFFPyramidal.isSolvable_of_isPyramidal
The declaration asserts a single implication, universally quantified over a type together with two typeclass structures on it.
Binders. There is one implicit type variable (implicit: the reader/user never supplies it explicitly; it is inferred). There are two instance-implicit hypotheses: carries a group structure, and is finite (Mathlib's Finite, i.e. is in bijection with some ; this is the non-constructive finiteness predicate, not a chosen enumeration). There is then one explicit hypothesis, named , asserting that is pyramidal in the sense of the accompanying definition, unfolded below. The conclusion is that is solvable. Note that the finiteness instance is not merely an ambient convenience: it is required even to state the hypothesis, since the pyramidality predicate itself takes a finiteness instance as an argument. So the claim is:
Unfolding "pyramidal". is the following existential statement. There exist a natural number and a function — a total function on all of , so is defined for every , including , but only some of its values are constrained — such that
- , the whole group viewed as a subgroup of itself;
- , the trivial subgroup ;
- for every natural number with (strict inequality), both of the following hold:
- , i.e. is contained in as a subgroup; and
- .
Here is Mathlib's relative index: by definition it is the index of the subgroup inside , i.e. the number of cosets of in , as a natural number, with the convention that this value is when that index is infinite (which cannot occur here, being finite). Under the accompanying containment this equals the ordinary index ; the containment is stated separately and is not implied by the index equation. Next, denotes Nat.card of the subgroup regarded as a type — its cardinality as a natural number (again for an infinite type, which cannot occur here). And is Mathlib's least-prime-factor function on , with the total-function conventions and ; for it is the smallest prime dividing . No truncated subtraction or division appears anywhere in the statement; the only "junk value" conventions in play are those just listed for , Nat.card, and the index.
So the hypothesis says: admits a descending (not necessarily strictly descending) chain of subgroups in which each successive index equals the smallest prime dividing the order of the previous term.
Degenerate and boundary cases implicit in this. The case is allowed: then conditions (1) and (2) together force , i.e. is the trivial group, and condition (3) is vacuous (there is no ). The hypothesis is thus satisfiable and in particular holds for the trivial group, for which the conclusion is also true; the theorem is therefore not vacuous, and the hypothesis is not unsatisfiable. Because the chain need not descend strictly, and because , once the chain reaches at some index the step condition at later indices is satisfied by (index ); hence is not determined by and may be taken arbitrarily large. Values for are entirely unconstrained. The values of at indices are constrained only by (1)–(3); in particular nothing requires to be normal in or in , and nothing directly requires to be proper. Since is finite, always, so the convention is never reached, but the case is reachable exactly when is trivial.
Unfolding the conclusion. Group.IsSolvable G is Mathlib's solvability predicate: there exists a natural number with the -th term of the derived series of equal to the trivial subgroup, , where and is the commutator subgroup of the previous term. Equivalently, has a finite chain of normal-in-the-previous subgroups descending to with abelian quotients. For the trivial group this holds with .
Proof status. The proof body is sorry: the statement is asserted but not proved in this artifact, and the file will compile only with a sorry warning. No part of the statement is stated as an iff, an uniqueness claim, or a strict inequality other than the index bound noted above.
Confirmed by the mission captain (proposal self-audit).