Contextual Seriation
Also known as: Occurrence Seriation, Sequence Dating, Incidence Seriation
Contextual seriation, also called occurrence or sequence seriation, is a relative-dating method that orders discrete archaeological units — typically graves or closed deposits — using only the presence or absence of artifact types within them. Its logic is the lifespan assumption: each type is introduced, used continuously for some span, and then disappears, so the contexts in which a type occurs should form an unbroken stretch of the sequence. By permuting the rows and columns of a presence-absence matrix until every type's occurrences cluster into a single contiguous block, the analyst recovers a one-dimensional ordering interpreted as time. The technique originates with W. M. F. Petrie's sequence dating of Egyptian predynastic graves and remains a standard tool for chronology where only incidence data, not abundances, are available.
Key highlights
- Works from presence-absence data alone, so it handles small closed finds like grave goods where counts are unreliable.
- Has a clean combinatorial formulation (the consecutive-ones property) that makes optimality well defined.
- Maps directly onto correspondence analysis, giving a reproducible solution and a built-in one-dimensionality diagnostic.
- Naturally identifies units and types that violate the temporal pattern, flagging mixing or non-chronological structure.
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 contextual seriation when your dating units are discrete closed finds — graves, caches, sealed deposits — described by which types they contain rather than by reliable counts, and you need a relative chronology. It is the appropriate choice when assemblages are too small for stable proportions, making frequency seriation unreliable, and when types have reasonably well-bounded lifespans. It is less suitable for open or mixed deposits, for long-lived utilitarian types that span the whole sequence and therefore carry no ordering information, and for situations where presence is driven by wealth or function rather than date. Results should be oriented and checked against stratigraphy and absolute dates.
Strengths & limitations
- Works from presence-absence data alone, so it handles small closed finds like grave goods where counts are unreliable.
- Has a clean combinatorial formulation (the consecutive-ones property) that makes optimality well defined.
- Maps directly onto correspondence analysis, giving a reproducible solution and a built-in one-dimensionality diagnostic.
- Naturally identifies units and types that violate the temporal pattern, flagging mixing or non-chronological structure.
- Yields only relative order, with no direction or absolute age until anchored externally.
- Long-lived or ubiquitous types span the whole sequence and contribute no ordering information.
- Presence-absence discards abundance information that frequency seriation could exploit for finer resolution.
- Non-temporal structure — wealth, function, region — can masquerade as a gradient and produce spurious orderings.
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
Why use presence-absence instead of counts?
Because the dating units in contextual seriation are usually small closed finds such as single graves, where the number of objects of each type is tiny and statistically unstable, but whether a type is present is clear and reliable. Presence-absence data are therefore more trustworthy for these contexts than proportions. When assemblages are large enough to give stable proportions, frequency seriation extracts more temporal information; the two methods are chosen according to the nature of the data.
What is the consecutive-ones property and why does it matter?
The consecutive-ones property holds when the rows of a presence-absence matrix can be reordered so that, in every column, all the ones appear in one unbroken block. It is the mathematical expression of the assumption that each type was in use over a single continuous time interval: correctly time-ordered, every type's occurrences must be contiguous. A matrix with this property is perfectly seriable, and the row order is the chronology. Real data only approximate it, so the analyst minimizes the embedded gaps that break the blocks.
How is contextual seriation related to correspondence analysis?
Correspondence analysis applied to an incidence matrix arranges both units and types along latent axes that summarize their co-occurrence structure. When the data are dominated by a single temporal gradient, the first axis scores order the units chronologically and reproduce the consecutive-ones blocks, so the ordination and the combinatorial seriation agree. A pronounced arch in the plot of the first two axes — the horseshoe effect — is the diagnostic signature of a genuine one-dimensional seriation, which is why CA is the standard modern engine for the method.
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). Contextual Seriation. ScholarGate. https://scholargate.app/archaeology/contextual-seriation