The Gold Standard · How X Works
One animal can become many fragments, while many animals can collapse into one minimum
Faunal counts look concrete, but they depend on how fragments are identified, sided, zoned and grouped. The minimum number of individuals can change when contexts are pooled or separated. This guide shows what each count can support and where the denominator quietly moves.
Choose a reading route
- Start from first principles — define the unit and the threat.
- Follow the worked example — make the hidden arithmetic visible.
- Test transfer — change the context and check the rule.
- Audit the evidence — inspect records, assumptions and failure points.
- Use the practical checklist — apply the method to new work.
Chapter contents
Foundations: Chapters 1–3
Method: Chapters 4–6
Transfer and limits: Chapters 7–9
Connections and use: Chapters 10–12
1. Counts are analytical constructions
Number of identified specimens, or NISP, counts identified fragments. Minimum number of individuals, or MNI, estimates the smallest number of animals needed to account for the observed elements under stated grouping rules. Neither directly equals the number originally deposited or consumed.
NISP is additive and transparent but sensitive to fragmentation. MNI can reduce some fragment inflation but depends strongly on element, side, age, sex, size and aggregation. Good reporting keeps both the observation and the rule visible.
First principle
Define the assemblage and counting unit before comparing abundance.
2. Fragmentation changes the evidence base
Butchery, cooking, trampling, weathering, carnivore action, excavation and laboratory processing can break bone. Several fragments may refit to one specimen; fragments from the same anatomical zone may overlap or remain separate.
High fragmentation can raise NISP if every identifiable fragment is counted, yet it can also lower identifiable counts when diagnostic features disappear. Record completeness, size, refits, burning, gnawing and recovery method. Fragmentation is not a simple correction factor.
| Process | Possible effect on NISP | Possible effect on MNI |
|---|---|---|
| Breakage with identifiable zones | Increase | May remain stable |
| Breakage removing landmarks | Decrease identified count | May decrease evidence |
| Refitting | Recombines fragments | Clarifies individual |
| Fine recovery | Adds small fragments | May reveal new side/age |
← Previous · Contents · Next →
3. MNI requires a grouping rule
A common MNI procedure finds the most abundant non-duplicable element category after considering side and relevant attributes. Five left distal tibiae require at least five individuals. Yet the result depends on whether all layers, phases or features belong to one analytical group.
Age, sex, size and pathology can raise the minimum when apparently duplicate elements cannot belong to the same animal. These criteria must be applied consistently and described. A spreadsheet maximum without anatomical logic is not an MNI.
- State taxon and identification level.
- State element, portion and side.
- Record age, size, sex or pathology used.
- Name the spatial and chronological aggregation.
- Explain how refits and articulations were handled.
← Previous · Contents · Next →
4. Worked example: pooling can lower the total
A fictional site has two secure contexts. Context A contains three left and two right sheep distal tibiae, giving MNI 3. Context B contains two left and three right distal tibiae, also giving MNI 3. Summing context-level MNI gives 6.
| Grouping | Left distal | Right distal | MNI |
|---|---|---|---|
| Context A | 3 | 2 | 3 |
| Context B | 2 | 3 | 3 |
| Pooled A+B | 5 | 5 | 5 |
| Sum of separate MNI | — | — | 6 |
Pooled MNI is 5 because the maximum side count is five. The contexts may genuinely represent six or more animals, but MNI is a minimum under each grouping rule. Neither 5 nor 6 is universally correct; the defensible value follows the archaeological question and context integrity.
← Previous · Contents · Next →
5. Choose aggregation before seeing the answer
If contexts are contemporaneous refuse from one activity, pooling may be justified. If they are separated phases, buildings or depositional episodes, pooling can erase change. Very narrow grouping can inflate summed MNI because the same individual may be counted in several contexts after dispersal.
Pre-specify spatial and temporal units using stratigraphy and formation evidence. Report sensitivity across at least one reasonable alternative when the conclusion depends on aggregation.
| Question | Possible grouping | Risk |
|---|---|---|
| Site-wide taxon presence | Whole compatible phase | Hides local variation |
| Household provisioning | Feature or building | Dispersed carcass crosses units |
| Change through time | Phase | Broad phase mixes episodes |
← Previous · Contents · Next →
6. Use NISP and MNI together
NISP supports element profiles, modification frequencies and transparent totals. MNI gives a conservative individual minimum. Other measures such as MNE, weight, biomass or diagnostic-zone counts may address different questions. Agreement among measures strengthens interpretation; disagreement reveals where method matters.
Percentages based on MNI should not be compared with percentages based on NISP as if they share a denominator. Provide raw counts and calculation units. When sample size is small, one additional element can change an MNI proportion sharply.
Practical check
Recalculate the main taxon ranking under NISP, MNI and a documented diagnostic-zone measure. Explain any reversal.
← Previous · Contents · Next →
7. Transfer test: commingled fish bone
Move the method to a midden with thousands of fish fragments. Many vertebrae cannot be sided, and repeated elements vary along the body. MNI based on a paired cranial element may be low if heads were removed elsewhere, while vertebral counts capture processing waste differently.
The transferable principle is not to force one measure across every assemblage. Choose elements and units that fit preservation, recovery and the behavioural question. Transfer succeeds when the measure is biologically and archaeologically interpretable.
A small diagnostic bone can establish presence without supporting abundance.
← Previous · Contents · Next →
8. Misconceptions and limits
- “MNI is the actual herd size.” It is a minimum under a rule.
- “NISP always exaggerates abundance.” Fragmentation can also destroy identifiability.
- “Pooling more data improves accuracy.” It can lower MNI and erase phases.
- “Summed MNI is always safer.” Narrow units can count one dispersed individual repeatedly.
- “The highest count is the best estimate.” A measure must fit the question and formation history.
Species identification, preservation and recovery differ among taxa. No counting method fully corrects differential survival. Interpret abundance alongside taphonomy, context and sampling.
← Previous · Contents · Next →
9. Evidence audit
Audit identification codes, sided zones, refits, phase assignments and calculation scripts. Reproduce the MNI from the specimen table. Check whether left/right labels, age classes or context filters changed during analysis. Retain rejected joins and uncertain identifications rather than silently recoding them.
| Audit question | Evidence | Warning sign |
|---|---|---|
| Can MNI be rebuilt? | Specimen-level table | Only final total retained |
| Was grouping justified? | Stratigraphy and protocol | Contexts pooled after results |
| Was fragmentation recorded? | Completeness and refit fields | NISP treated as individuals |
| Are measures comparable? | Denominator labels | Mixed MNI and NISP percentages |
← Previous · Contents · Next →
10. Questions readers often ask
Should MNI be calculated by context or phase?
Use the smallest archaeological unit that supports the question without repeatedly counting a dispersed individual; show sensitivity if both are plausible.
Can NISP be added across contexts?
Yes, when identification and recovery are comparable. Interpretation still needs formation and fragmentation checks.
Does refitting reduce NISP?
It can if fragments are recombined as one specimen under the recording protocol. Preserve both the raw fragment history and final analytical unit.
← Previous · Contents · Next →
11. Reading routes and ecosystem connections
Start with the Gold Standard of Archaeology and Teaching Zooarchaeology. Use the guide on assemblage integrity and refitting before selecting aggregation units.
The companion guides on recovery fractions and sieve bias and bulk samples and rare-find detection show how evidence enters the count. Continue through learning that transfers into life, The Gold Standard and the How X Works hub.
← Previous · Contents · Next →
12. Sources and practical checklist
Authoritative guidance includes Historic England’s Animal Bones and Archaeology: Recovery to Archive, its key reference supplement, and its report on sampling and sieving for animal bones. These resources connect recovery, identification, quantification and archiving.
- Define the assemblage boundary.
- Keep specimen-level observations.
- Record element, zone, side and attributes.
- State how MNI was calculated.
- Report NISP and MNI with denominators.
- Test reasonable aggregation choices.
- Tie abundance claims to taphonomy and recovery.