A mapping from finitely encoded inputs (commonly finite strings or natural numbers) to outputs for which there exists a finite, mechanically executable procedure that, on every valid input, halts and produces the correct output.

Principle

Computability is the existence of an effective finite procedure: a single fixed recipe that transforms any allowed input into its output in a finite number of deterministic steps.

Demonstration

Addition on natural numbers implemented by a Turing machine: given two numeral encodings, the machine halts with the numeral encoding of their sum.

Misapplication

Asserting a real-valued function is computable without specifying an encoding for reals or a termination criterion, or treating numerical approximation routines as exact computability claims.

Consequence

When a function is computable one can mechanize its evaluation, prove decidability properties for its graph, and embed it in formal reductions between decision problems.

Reversal

A noncomputable function: no finite algorithm exists that halts with correct outputs for all valid inputs.

Boundary

Applies only to functions with effective finite encodings of inputs/outputs; excludes unrepresented real functions, oracle or hypercomputational models, and claims that probabilistic approximations imply exact computability.

Semantic Tension

Often confused with decidability of a set (membership decision) or with numerical approximability; computable function concerns explicit output generation for each input, not just set membership or approximation quality.

Synthesis

A computable function is a mapping between finitely representable objects for which a single finite mechanical procedure yields exact outputs for every permitted input.