The Gold Standard · Archaeology
Make the decision path visible
Twenty pottery fragments do not necessarily represent twenty pots. One broken vessel can produce many sherds, while a single rim fragment may preserve only a small portion of another. A reliable comparison begins by stating which quantity is being counted.
This guide separates sherd count, weight, rim equivalents and minimum vessel counts. It works through original numbers, shows why aggregation changes a minimum count, and gives a practical recording route. The examples are fictional, not excavation findings.
Choose the route closest to your decision
Chapter contents
Choose the quantity · Chapters 1–3
Work the comparison · Chapters 4–6
Interpret the assemblage · Chapters 7–9
Audit and continue · Chapters 10–12
CHAPTER 1 OF 12 · CHOOSE THE QUANTITY
1. Start with the archaeological question
Ask whether you want to compare fragmentation, the representation of vessel forms, the amount of pottery recovered or an estimate of vessels represented. These are different questions. No single number answers all of them.
A high sherd count might reflect intensive use, greater breakage, better recovery or several processes together. Before interpreting activity, specify the assemblage boundary, recovery method, fabric and form categories, and the counting rules. The recorded quantity should remain visible whenever a percentage is quoted.
CHAPTER 2 OF 12 · CHOOSE THE QUANTITY
2. Sherd count records fragments under a stated convention
A sherd is a fragment, not a vessel. Count according to a consistent specialist protocol, including how fresh excavation breaks are reunited. CIfA guidance recommends counting freshly broken sherds as one. Record such treatment so a later analyst can reproduce the total.
Fragment counts remain useful. Combined with weight, they can describe changes in average fragment size. But twice as many fragments need not mean twice as many original vessels. A fragile fabric can produce a different count from a tougher one under similar disturbance.
Evidence check
Check whether the scope, owner and decision consequence are explicit before continuing.
CHAPTER 3 OF 12 · CHOOSE THE QUANTITY
3. Weight measures recovered material, not pot numbers
Weight adds another view of the assemblage but is affected by wall thickness, vessel size, preservation and recovery. A few heavy storage-jar pieces can outweigh numerous thin tableware fragments. Neither measurement is wrong; they measure different properties.
Use comparable preparation and weighing procedures. Record whether retained residue, sediment or conservation treatment could affect measurements. Report counts and weights together where appropriate, and avoid interpreting a weight share as a direct household vessel share.
| Measure | Direct quantity |
|---|---|
| Sherd count | Recorded fragments |
| Weight | Recovered mass |
| Rim equivalent | Represented circumference |
| Minimum vessels | Least compatible count |
CHAPTER 4 OF 12 · WORK THE COMPARISON
4. A rim equivalent expresses a proportion of circumference
A rim equivalent sums the surviving proportions of vessel rims, normally measured with a rim chart under a stated protocol. A 10% rim contributes 0.10; a 25% rim contributes 0.25. The total describes represented rim circumference in whole-rim units.
It is not an inventory of intact pots. A sum of 1.00 can arise from one complete rim or small portions of many vessels. Specialist guidance and vessel regularity matter: an irregular prehistoric vessel may not support the same measurement procedure as a regular wheel-thrown rim.
CHAPTER 5 OF 12 · WORK THE COMPARISON
5. Calculate without changing the denominator
Suppose four non-overlapping rim portions contribute 10%, 15%, 25% and 20%. Their sum is 70%, or 0.70 rim equivalents. Do not divide by the number of rim sherds: that produces average rim completeness per sherd, a different quantity.
Keep original percentages alongside the sum. If a piece has an uncertain diameter or orientation, mark the uncertainty and follow the project’s specialist convention. A confident-looking decimal does not remove ambiguity in fitting a small fragment to a chart.
Evidence check
Ask a second reviewer to reconstruct the example from the preserved evidence and stated assumptions.
CHAPTER 6 OF 12 · WORK THE COMPARISON
6. Worked example: the ranking reverses
Assemblage A has 40 sherds and rim portions of 10%, 15%, 25% and 20%, giving 0.70 equivalents. Assemblage B has 20 sherds and rim portions of 40%, 35% and 25%, giving 1.00. A leads on fragments; B leads on represented rim circumference.
The reversal is not a contradiction. It shows why “more pottery” needs a defined measure. Neither total alone establishes the number of people, meals or years of use. A defensible report gives both measures, then investigates recovery, vessel types and depositional history.
| Assemblage | Sherds | Rim equivalents |
|---|---|---|
| A | 40 | 0.70 |
| B | 20 | 1.00 |
CHAPTER 7 OF 12 · INTERPRET THE ASSEMBLAGE
7. Avoid double-counting a reconstructed rim
Document which pieces join and which rim segments overlap in the reconstruction. Joining fragments do not become several independent vessels. Conversely, non-joining pieces with similar fabric are not automatically separate vessels.
Preserve object identifiers and the basis of each association. Physical joins, distinctive decoration, fabric and rim form provide different strengths of evidence. A photograph of the reconstructed group and a reproducible calculation are more useful than an unexplained final vessel total.
CHAPTER 8 OF 12 · INTERPRET THE ASSEMBLAGE
8. Minimum vessel counts depend on the grouping boundary
A minimum number of vessels is the least number consistent with the distinctions used in the analysis. It depends on which fragments can reasonably belong together. Define the analytical unit and diagnostic criteria before comparing totals.
Imagine one distinctive pot represented in each of two adjacent contexts. Each context has a minimum of one, but a secure cross-context join can make the combined minimum one rather than two. Minimum counts are therefore not automatically additive across contexts. Keep the context-level record even when creating a combined interpretation.
Evidence check
Verify that the control still reaches the intended decision under a changed but plausible context.
CHAPTER 9 OF 12 · INTERPRET THE ASSEMBLAGE
9. Changed-context transfer: a pottery production deposit
A kiln dump can contain production waste rather than ordinary household discard. Repeating the household interpretation because the same counting method was used would confuse measurement with context.
The arithmetic of rim proportions transfers; the behavioural meaning does not. Ask about wasters, vessel size range, selection, transport and deposition. Historic England’s production-site guidance treats quantification as part of a wider investigation, not a machine that turns fragments directly into production output.
| Unit | Minimum vessels |
|---|---|
| Context 1 | 1 |
| Context 2 | 1 |
| Combined, secure same-pot join | 1, not automatically 2 |
CHAPTER 10 OF 12 · AUDIT AND CONTINUE
10. Compare like with like before computing percentages
A form’s percentage can change depending on whether the denominator is all sherds, diagnostic sherds, weight or total rim equivalents. Label the denominator and show unclassified material. Compare assemblages with comparable recovery and recording procedures.
If tiny rim fragments were omitted from one deposit but estimated in another, apparent form differences may partly be procedural. Reconcile the rules or report sensitivity. Do not hide a changed recording threshold inside a cultural explanation.
CHAPTER 11 OF 12 · AUDIT AND CONTINUE
11. Check the inference and its limits
Rim equivalents do not correct every preservation or recovery bias. Minimum counts depend on recognition and grouping. Sherd counts are sensitive to fragmentation. Weight favours heavy material. Use complementary measures to identify disagreement, not to select whichever number tells the preferred story.
Practical check: could someone reconstruct your table from the fragment register? Can they identify joins, excluded material and uncertain estimates? If not, improve the evidence trail before explaining consumption or trade.
Evidence check
Confirm that versions, notices and unresolved limits remain visible in the final record.
CHAPTER 12 OF 12 · AUDIT AND CONTINUE
12. Report a compact, inspectable result
Give the assemblage boundary, recovery procedure, fabric and form categories, fragment counts, weights, rim-equivalent method, vessel-grouping criteria and important uncertainties. Explain any ranking that changes between metrics.
CIfA’s pottery recording toolkit describes the principal quantification choices. Oxford Archaeology’s recording system supplies an operational example of rim-percentage recording, while Historic England places quantification within pottery-production research. They guide method selection; local material still needs qualified specialist judgement.
| Field | Keep visible |
|---|---|
| Boundary | Contexts included |
| Method | Counting convention |
| Join | Object identifiers |
| Percentage | Named denominator |
Frequently asked questions
Does 1.00 EVE mean one pot?
No. It is a whole-rim-equivalent amount, potentially assembled from parts of several vessels.
Can I add minimum counts across contexts?
Not automatically; cross-context associations can make the combined minimum smaller than the sum.
Which metric is best?
Choose for the question and use complementary metrics to expose fragmentation and material biases.
Does a larger assemblage prove more intensive occupation?
Not by itself. Recovery, breakage, vessel type and deposition must also be examined.
Continue through verified routes
Core routes
Archaeology subject root · Gold Standard of Education · Gold Standard directory · How X Works hub
Specialist routes
- The Gold Standard of Archaeology | Assemblage Integrity, Refitting and Mixed Deposits
- The Gold Standard of Archaeology | Recovery Fractions, Sieve Mesh and Small-Find Bias
- The Gold Standard of Archaeology | Residue Contamination, Conservation Treatments and Chemical Attribution