A quantitative property of an archaeological assemblage that combines taxonomic richness (the number of distinct artifact categories) and evenness (the relative abundances of those categories) to express how varied and uniformly represented artifact types are within a defined recovery unit.

Principle

Assemblage diversity is determined jointly by category richness and relative abundance; differences in measured diversity may reflect variation in past activities, duration or intensity of deposition, post‑depositional mixing, or sampling and classification strategies.

Demonstration

Illustrative scenario: Two 1 m2 excavation spits yield 100 artifacts each. Unit A contains 10 artifact types with roughly equal counts; Unit B contains 10 types but 80% belong to one type. Measuring richness alone would equate them, but a combined richness+evenness metric shows Unit A has higher assemblage diversity, suggesting broader or more evenly distributed activity signatures in Unit A than in Unit B.

Misapplication

Interpreting raw counts as directly comparable without controlling for sample size, recovery method, or classification resolution (e.g., lumping versus splitting categories) leads to misleading diversity estimates; apparent differences may be methodological rather than behavioral.

Consequence

When correctly used with transparent sampling and classification, assemblage diversity can guide inferences about activity breadth, site function, or depositional processes; when misused it can produce false contrasts between contexts and misdirect research focus.

Reversal

Under very small sample sizes, highly fragmentary assemblages, or when recovery methods differ between contexts, diversity metrics lose discriminating power and may reflect sampling noise; in some cases specialized activity areas produce low diversity that nonetheless indicates intense, focused use rather than low occupational importance.

Boundary

Clearly within: a stratigraphic unit for which artifact counts and a defined classification scheme are available and sampling methods are recorded. Boundary case: two adjacent spits with unequal recovery techniques (hand‑collection vs sieving) where interpretation depends on adjusting for recovery bias. Clearly outside: a qualitative statement that a site is 'diverse' without accompanying counts, categories, or measurement method.

Semantic Tension

Tension exists between treating diversity as an index of past behavioral complexity (signal) and treating it as a function of sampling, classification, and post‑depositional processes (noise); resolving this requires explicit methodological control and corroborating evidence.

Synthesis

Assemblage diversity is a sample‑dependent, methodological construct: it can indicate meaningful differences in past behavior or formation processes, but only when measured with consistent classification and sampling protocols and interpreted alongside taphonomic and contextual information.