A formal language (symbols such as O, o, Θ, Ω, ω) for classifying the growth rates of functions by their limiting behaviour as an argument tends to infinity or to a singular limit.

Principle

Compare dominant terms by discarding lower-order contributions and constant factors to characterize long-run or small-parameter scaling; relations are defined via limits or bounding inequalities.

Demonstration

For f(n)=3n^2+5n and g(n)=n^2, f(n)=Θ(g(n)) because lim sup f(n)/g(n) is bounded above and below by positive constants; thus they share quadratic growth.

Misapplication

Using Big-O to assert precise values for finite inputs or to compare functions outside the specified limit regime (e.g., treating O(1) as a specific constant rather than a bounded class).

Consequence

Provides a concise, order-only classification that guides algorithm selection, scaling analysis, and limiting approximations while suppressing irrelevant constants.

Reversal

Exact equality or pointwise bounds; reversing asymptotic comparison requires producing uniform finite-range inequalities or exact formulas rather than limit relations.

Boundary

Only quantifies growth in a limiting regime (usually argument → ∞ or parameter → 0); does not provide convergence rates if limits do not exist or for oscillatory functions without additional constraints.

Semantic Tension

Tension with probabilistic or average-case statements: asymptotic notation describes worst-case or order behavior, not necessarily typical or distributional behaviour unless specified.

Synthesis

Asymptotic notation is a limit-based shorthand that classifies functions by dominant growth or decay, enabling comparison of their leading-order behaviour while omitting lower-order structure.