Frequency Seriation
Also known as: Frequency Seriation Dating, Battleship-Curve Seriation, Proportional Seriation
Frequency seriation is a relative-dating technique that orders archaeological assemblages in time by the changing proportions of the artifact types they contain. Its premise is that any cultural type is introduced, gradually becomes popular, peaks, and then declines, so that the relative frequency of a type traces a single rise-and-fall curve through time. By rearranging the rows of a type-by-assemblage abundance table until every type's frequency forms one continuous unimodal sequence, the analyst recovers an ordering interpreted as chronological. Drawn as horizontal bars, these curves take the lens or 'battleship' shape that gives the method its popular name. Frequency seriation grew out of W. M. F. Petrie's sequence dating and was formalized for proportional data by mid-twentieth-century Americanists such as James A. Ford, becoming a backbone of culture-historical chronology before absolute dating was widely available.
Key highlights
- Produces a fine-grained relative ordering from artifact counts alone, without any absolute dates.
- Uses proportions, so it is robust to differing assemblage sample sizes and exploits gradual quantitative change.
- Yields an interpretable graphical result — the battleship-curve diagram — that exposes anomalous assemblages.
- Maps cleanly onto multivariate ordination (correspondence analysis), giving a reproducible, computer-checkable solution.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use frequency seriation when you have many discrete assemblages, each yielding reasonably large counts of a set of consistently classified artifact types, and you want a relative chronology — especially where absolute dates are sparse or expensive. It is most powerful for abundant, rapidly changing material such as decorated pottery or pipe stems within a single cultural tradition and region, where popularity-driven turnover is plausible. It is poorly suited to small assemblages with unstable proportions, to types whose abundance is driven by function or activity rather than fashion, to mixed or palimpsest deposits, and to comparisons across regions where types diffuse at different times. Frequency seriation should be cross-checked against stratigraphy and absolute dating whenever possible, since the popularity assumption can fail.
Strengths & limitations
- Produces a fine-grained relative ordering from artifact counts alone, without any absolute dates.
- Uses proportions, so it is robust to differing assemblage sample sizes and exploits gradual quantitative change.
- Yields an interpretable graphical result — the battleship-curve diagram — that exposes anomalous assemblages.
- Maps cleanly onto multivariate ordination (correspondence analysis), giving a reproducible, computer-checkable solution.
- Recovers only relative order and not direction or calendar age until anchored by external dating.
- Depends on the popularity assumption: types whose frequency is driven by function, trade, or reuse violate unimodality.
- Requires assemblages that are contemporaneous internally, regionally comparable, and free of mixing or redeposition.
- Small assemblages give unstable proportions, and several orderings may fit the data nearly equally well.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
How does frequency seriation differ from contextual seriation?
Frequency seriation orders assemblages by the relative proportions (frequencies) of types and assumes each type follows a single rise-and-fall popularity curve. Contextual (occurrence) seriation orders units only by the presence or absence of types, assuming each type has a continuous span of use so its occurrences fall in an unbroken block. Frequency seriation needs larger, countable assemblages and exploits quantitative change; contextual seriation works with presence-absence data, as in W. M. F. Petrie's original sequence dating. The two are complementary techniques within the same culture-historical toolkit.
Does seriation tell you the actual dates of assemblages?
No. On its own, seriation produces only a relative ordering and does not even reveal which end is older. It must be oriented and calibrated with external information — a stratigraphic superposition, a radiocarbon date, or a historically dated type. Once one or more assemblages in the sequence are absolutely dated, the relative order can be converted into an approximate chronology, and undated assemblages placed in the sequence inherit an estimated age range.
What can make a frequency seriation fail?
The method assumes assemblages are internally contemporaneous, regionally comparable, and that type abundance is driven by popularity rather than by function, trade, curation, or reuse. It fails when assemblages are mixed or redeposited, when types serve activities that vary in intensity across sites, when samples are too small for stable proportions, or when comparing regions where the same type peaks at different times. Such violations produce multimodal curves and ambiguous orderings, which is why results are cross-checked against stratigraphy and absolute dates.
Sources
- 1.Lyman, R. L., & O'Brien, M. J. (2006). Measuring Time with Artifacts: A History of Methods in American Archaeology. University of Nebraska Press.ISBN 9780803280526
- 2.Renfrew, C., & Bahn, P. (2016). Archaeology: Theories, Methods, and Practice (7th ed.). Thames & Hudson.ISBN 9780500292105
You have read it. What now?
Cite this page
ScholarGate. (2026, June 23). Frequency Seriation. ScholarGate. https://scholargate.app/archaeology/frequency-seriation