A situation where a sequence of homomorphisms ...→A_{i-1}→A_i→A_{i+1}→... fails to be exact at some position i, meaning the image of the incoming map is not equal to the kernel of the outgoing map; homology groups measure this failure.

Principle

Exactness captures the perfect balance of images and kernels; failure of exactness signals obstruction, measured by nonzero homology, and reflects hidden extension, torsion, or relational data not captured by the maps.

Demonstration

Consider the sequence Z --×2--> Z --×3--> Z. The image of the first map is 2Z while the kernel of multiplication by 3 is 0, so image≠kernel and the sequence fails to be exact at the middle term.

Misapplication

Using sequences as though exact without verifying — for example applying derived functor vanishing or splitting arguments on nonexact sequences — leads to incorrect homological conclusions.

Consequence

Nonzero homology groups record obstructions to lifting, splitting, or solvability; they indicate extensions, detect torsion, and obstruct naive computations in derived functors and spectral sequences.

Reversal

Exact sequences, where images and kernels match at every position, allow decomposition, lifting, and clean calculations of derived functors; exactness is a hypothesis that enables many structural theorems.

Boundary

Pertains to chains of homomorphisms in algebra, topology and geometry; failure can be local or global and must be distinguished from lack of surjectivity or injectivity alone since exactness combines both properties.

Semantic Tension

Between exactness as an ideal condition and practical calculations: short exactness, left- or right-exactness of functors, and the presence of derived functors create a lattice of partial exactness notions.

Synthesis

Failure of exactness is the concrete homological signal that images and kernels misalign: homology measures the obstruction, which encodes extension and torsion data and warns against naively transferring properties that require exact sequences.