For the preliminary definitions, a block is for a primitive central idempotent , and a module belongs to it when . For a semisimple algebra, the Artin–Wedderburn theorem identifies its blocks with the factors in its product decomposition. Each factor is a block of a finite-dimensional algebra.
We prove Ext separation of finite-length modules. The extension hypothesis is for simples in opposite sets. First, for a simple and a module of finite composition length with factors in , we have . Induct on the length of : for with simple, the long exact sequence of the Ext functor contains
whose outer groups are zero. The same argument works with the two sets interchanged.
Now induct on the length of , with immediate. Take a simple quotient in . By induction, with factors in the respective sets. Suppose ; the other case is symmetric. Quotienting by gives
This splits by the preceding Ext vanishing. Let be the inverse image in of the chosen complementary copy of . Then , , and . Thus has only factors in , while has only factors in .
If two finite-length modules have factors in disjoint sets, any homomorphism between them is zero: a nonzero image would, by the Jordan–Hölder theorem, have a simple factor belonging to both sets. For any submodule of with factors in , projection onto is consequently zero, so . The analogous argument applies to . Hence
In particular both summands are preserved by every endomorphism of .
We prove , the main step in Ext-connected components determine blocks. Suppose a block of a finite-dimensional algebra had more than one component of the graph of nonsplit extensions between its simple modules. Partition its simple types into one component and the union of the others. There are no cross-extensions, so part (a) decomposes the left regular -module as . Both summands are nonzero: every simple -module is a quotient of the regular module and therefore occurs among its composition factors.
The maximality in (a) makes these summands canonical. Every right multiplication is a left -module endomorphism, so it preserves them. Thus they are two-sided ideals, and their projection commutes with both left and right multiplication. Putting gives
Since both summands are nonzero, . This contradicts the primitivity of the central block idempotent . There is therefore just one Ext component within each block, proving that simples in the same block satisfy .
We prove . The central block idempotents split every finite-dimensional module into its block components. A nonzero indecomposable representation, in particular an indecomposable projective module, has only one such component; otherwise these give a nontrivial direct-sum decomposition.
Every Jordan–Hölder factor of a module in a block belongs to that block, since its central idempotent acts as identity on submodules and quotients. Thus any two simple factors of the same projective indecomposable are in the same block. Applying this along the proposed chain yields .
We prove . Consider one nonsplit short exact sequence
with simple. Lift the canonical surjection from the projective cover through , obtaining . Its image maps onto . The intersection is either zero or , since is simple. If it were zero, would identify with , splitting the extension. Hence , and surjectivity onto then implies .
Therefore both and occur as Jordan–Hölder factors of the indecomposable projective . Replacing every edge in an Ext chain by this observation proves (ii). Together with the preceding two implications, this proves
The repeated word “projective” in the printed version of (ii) has no mathematical effect.

Articles by others on the same topic (0)

There are currently no matching articles.