Asymptotic measure of size of a subset A of the natural numbers given by the limit d(A)=lim_{x→∞} |A∩[1,x]|/x when that limit exists; ranges in [0,1] and quantifies proportion of integers in A.

Principle

Compare counting function of A to linear growth x to capture long-term frequency; natural density treats each integer equally and demands stabilization of relative frequency.

Demonstration

Examples: even integers have natural density 1/2; multiples of 3 have density 1/3; the set of primes has density 0 because π(x)/x→0; perfect squares have density 0.

Misapplication

Assigning a natural density to sets where the limit does not exist (oscillatory sets) or confusing natural density with logarithmic or Banach densities yields misleading conclusions about 'size'.

Consequence

When it exists, natural density provides a straightforward probabilistic interpretation: a random integer in {1,...,x} falls in A with probability ≈ d(A) for large x; it informs asymptotic counting and heuristic models.

Reversal

The reversal is to use nonuniform weighting (e.g., logarithmic density) so that density depends on weighting of integers; such reversals change which sets are large and which are negligible.

Boundary

Requires existence of the limit; excludes many important number-theoretic sets with oscillatory or sparse behavior for which natural density fails to exist; does not reflect distribution inside residue classes or local irregularities.

Semantic Tension

Tension with logarithmic density, upper/lower densities, and natural density: different notions can assign different 'sizes' to the same set, making choice of density crucial to correct interpretation.

Synthesis

Natural density is the limit proportion of integers in a set measured by simple counting up to x; it is the basic uniform-frequency notion for subsets of naturals that exists only when relative counts stabilize asymptotically.