Zooarchaeology studies animal remains from archaeological contexts to investigate relationships among people, animals and environments. It can help explain hunting, herding, fishing, food preparation, exchange and other activities, but a surviving bone is not a direct record of a meal. Its meaning depends on identification, archaeological context, preservation and the route by which it entered the collection.
The gold standard of zooarchaeology is an explanation whose claims remain connected to those observations and whose uncertainties can be checked. This guide follows that connection from a recovered fragment to a defensible account of past animal use. All numerical assemblages, sample codes and worked scenarios below are invented teaching examples; they do not describe real excavations or establish universal biological thresholds.
This subject belongs to The Gold Standard, where the phrase names a framework for dependable work rather than an accreditation. Its learning goal follows The Gold Standard of Education: understanding should remain useful when the species, site or question changes.
THE GOLD STANDARD · ZOOARCHAEOLOGY
Connect animal remains with past lives
Choose a route through identification, counts, life histories and archaeological explanations. All teaching chapters remain open below.
ROUTE 1 · CHAPTERS 1–4
Follow animals into the evidence
Connect research questions, formation, recovery and identification.
ROUTE 2 · CHAPTERS 5–8
Count and interpret remains
Work through specimens, minimum individuals, anatomy and surface evidence.
ROUTE 3 · CHAPTERS 9–12
Reconstruct animal lives
Examine age, husbandry, domestication and smaller animals.
ROUTE 4 · CHAPTERS 13–16
Combine laboratory and field evidence
Test molecular identification, isotope assumptions and context comparisons.
ROUTE 5 · CHAPTERS 17–20
Repair, transfer and continue
Diagnose weak claims and investigate fish remains in harbour reclamation.
Full chapter index · Try the transfer task · Gold Standard subject hub
Chapter index
Follow animals into the evidence · Chapters 1–4
Count and interpret remains · Chapters 5–8
Reconstruct animal lives · Chapters 9–12
Combine laboratory and field evidence · Chapters 13–16
Repair, transfer and continue · Chapters 17–20
CHAPTER 1 OF 20 · Follow animals into the evidence
1. Begin with the relationship you want to explain
An animal can matter to people while alive, at death and after death. A sheep may supply milk or wool before its carcass supplies meat, skin and bone. A fish may be caught locally, preserved elsewhere or brought into a settlement as a traded product. An animal bone can become a tool, move with rubbish or be redeposited during construction. These possibilities make zooarchaeology a study of relationships and processes, not simply a catalogue of species.
Start by turning a broad interest into a comparison. Instead of asking whether a settlement kept animals, ask whether two securely dated phases differ in the ages and body parts represented among identified cattle. That question states the animal group, the observations and the unit of comparison. It also reveals what could invalidate the comparison: unlike recovery methods, mixed deposits or a very small sample.
Suppose a report lists cattle, sheep and fish from a harbour. It establishes their presence in the recovered assemblage under the report’s identification rules. It does not yet establish the share of calories supplied by each, the residence of livestock nearby or the ethnicity of the consumers. Each further conclusion requires another argument.
Readers new to context and archaeological explanation can begin with The Gold Standard of Archaeology. Keep one sentence beside every proposed interpretation: “The observation that would most strongly change this explanation is…” A useful investigation should be able to finish that sentence.
CHAPTER 2 OF 20 · Follow animals into the evidence
2. Follow the animal through the formation of an assemblage
Consider an invented coastal settlement. Fish are caught offshore; some are cleaned on the beach; selected portions reach houses; dogs disturb discarded material; later building work moves part of the midden; excavation samples only one corner. The final collection has passed through biological, cultural, physical and analytical filters. A species missing from it may have been absent, discarded elsewhere, destroyed, overlooked or left unidentified.
Draw the sequence in words before counting: living population, animals acquired, portions transported, material deposited, remains surviving, material recovered, specimens identified. At each transition, ask which animals or body parts could be favoured. A large jaw and a tiny vertebra do not have equal chances of discovery during hand collection.
This sequence is not a numerical correction formula. Multiplying guessed survival percentages produces a precise-looking answer without recovering lost information. Instead, use the sequence to identify which comparisons remain defensible. Two contexts excavated with identical methods may still differ in preservation, while two well-preserved contexts may have been collected differently.
The Gold Standard of Taphonomy develops the processes between an organism’s life and the evidence available for study. In a zooarchaeological report, formation history belongs beside the behavioural interpretation. If redeposition is plausible, say that the bones may describe material brought into the deposit rather than activity at the exact place where they were found.
CHAPTER 3 OF 20 · Follow animals into the evidence
3. Recover evidence at the scale of the question
Recovery design determines which questions a collection can answer. A project interested in small fish needs a method capable of retaining their diagnostic remains. A project interested in an articulated animal needs spatial relationships recorded before those relationships are dismantled. The general professional framework is set out in Historic England’s Animal Bones and Archaeology handbook, which connects project planning with recovery, analysis, publication and archiving.
For an original comparison, imagine two trial samples taken from the same mixed deposit. Sample A contains 20 litres of sediment and yields 8 identifiable fish specimens after coarse screening. Sample B contains 20 litres and yields 48 after a finer recovery procedure. Their observed densities are 0.4 and 2.4 specimens per litre. The sixfold difference cannot sensibly be read as six times more fishing: the samples were treated differently.
The trial suggests that recovery affects the evidence, but it does not establish a universal correction factor of six. Different species, sediment textures and fragmentation patterns may respond differently. A proper follow-up would use a documented sampling design and retain information about fractions, mesh sizes and processing losses.
Keep volume and recovery method attached to each sample. “Forty fish bones” without either field is much less useful than forty identifiable specimens from a stated volume and fraction. Record unsuccessful processing too. A sample that could not be separated reliably should not quietly become an ordinary zero in the final dataset.
CHAPTER 4 OF 20 · Follow animals into the evidence
4. Identify the specimen before naming the species
Identification works through anatomical comparison. What element is present? Which portion survives? Can its side be determined? Which observed features distinguish plausible taxa? A reference collection lets an analyst compare real variation rather than match a fragment to the most similar picture. Historic England describes its zooarchaeology programme and comparative collection as resources supporting this work.
Imagine specimen Q17 preserves a joint surface and part of a shaft. The analyst recognises an element but cannot distinguish two closely related animals with confidence. Recording the shared taxonomic group is more informative than forcing a species name. It preserves the positive identification and the unresolved boundary. “Unidentified” should likewise distinguish an unidentifiable fragment from one that has not yet been examined.
An identification record can hold the specimen ID, context, element, side, portion, taxon, confidence and reference used. Photographs with scales are valuable for review, but a photograph may not show all the three-dimensional features or surface conditions needed for a final decision. Re-examination must remain possible when the claim matters.
Use a second analyst on a deliberately selected subset, including difficult specimens, rather than only on easy examples. Agreement on obvious bones does not measure reliability at the decision boundary. When two analysts disagree, preserve their observations and reasoning before settling the final category. That disagreement may reveal a weak reference, an inconsistent rule or a genuinely ambiguous specimen.
CHAPTER 5 OF 20 · Count and interpret remains
5. Count specimens without turning them into animals
NISP, the number of identified specimens, is a count of specimens assigned under an explicit identification and counting protocol. It is not a count of animals. An element broken into several recognisable pieces may contribute several records unless a stated refitting or counting rule combines them. A fragment broken beyond recognition may disappear from a taxon-specific count altogether.
In an invented sample, cattle have 60 identified specimens and sheep/goat have 40. Within these two categories, cattle account for 60/(60 + 40) = 60% of NISP. The denominator excludes unidentified mammal material and all other animals. The correct statement is therefore about the identified specimen sample, not about 60% of livestock, meat or meals.
Now suppose ten of the sheep/goat specimens each break into three still-identifiable pieces during an unfortunate handling event, and the records count each resulting fragment. Sheep/goat NISP rises from 40 to 60; cattle’s share falls to 50%, although no animal has been added. This is an arithmetic demonstration of dependence on the counting unit, not a recommendation for treating modern damage as independent archaeological evidence.
The methodological literature explicitly examines these choices, including Lyman’s discussion of the history of NISP and MNI. Preserve original specimen relationships, note recent breaks and state the convention. A percentage becomes interpretable only after its numerator, denominator and exclusions have been explained.
CHAPTER 6 OF 20 · Count and interpret remains
6. Calculate minimum individuals and test aggregation
MNI asks for the smallest number of individuals consistent with the observed material under stated assumptions. Repeated diagnostic parts, side, age and other incompatibilities can constrain that minimum. It is a lower bound constructed from the assemblage, not an estimate of the full number of animals once present.
Consider securely identified, non-refitting proximal femoral portions of one taxon. Context A contains four left and one right; context B contains one left and four right. Ignore age and size differences for this simplified calculation. Each context separately requires at least four animals. If the contexts are genuinely independent populations, their minima sum to eight.
If the contexts are parts of one redeposited assemblage and are pooled, there are five left and five right portions. On this evidence alone, the pooled minimum is five. The arithmetic has not discovered that three animals disappeared. It has changed the grouping rule and allowed left and right parts from different contexts to belong to the same individuals.
Choose grouping from archaeological relationships, not from the answer you prefer. Report sensitivity when grouping is uncertain: “The minimum is five when pooled; separate context minima sum to eight.” Neither value is a census. The specialist guide to MNI, fragmentation and aggregation bias extends this reasoning. A useful exercise is to ask what new evidence would force the pooled minimum above five: an incompatible age class or a further duplicated portion might do so.
CHAPTER 7 OF 20 · Count and interpret remains
7. Read body-part patterns as competing explanations
A collection dominated by heads and feet may suggest early carcass processing, but that interpretation is not automatic. Survival, transport, craft use and identification can also change anatomical representation. The question is whether several observations fit the proposed activity better than the alternatives.
Suppose an invented yard has many horn-core fragments and little limb material. One proposal is horn working. A stronger investigation would look for repeated removal patterns, production debris, unfinished objects and spatial organisation. If the same fragments occur in imported levelling material, the yard’s function becomes less secure. If only skull elements were routinely retained during excavation, the collection itself cannot establish original anatomical dominance.
Anatomical comparisons must also respect the number of each element in an animal. Raw totals of vertebrae and femora are not directly comparable as fractions of complete bodies. Analysts therefore use clearly defined element-based measures where appropriate. Those measures still depend on identification, portion overlap and recovery; they do not remove formation history.
Experimental research on hammerstone-broken long bones shows why fragmentation and identifiable portions matter for behavioural interpretation. For the yard example, write an observation-only account first, then two competing activity explanations and one formation explanation. Choose a further observation that could discriminate among them. This prevents a familiar pattern label from doing the work of evidence.
CHAPTER 8 OF 20 · Count and interpret remains
8. Distinguish modification from the cause assigned to it
A mark is first a physical feature: its location, shape, orientation and relationship to the bone surface can be described. Calling it a cut mark, tooth mark or excavation scratch adds a causal interpretation. That interpretation should depend on diagnostic features, appropriate comparison and the condition of the surface.
Imagine 100 recorded specimens, but only 40 have surfaces preserved well enough for a particular mark to be assessed. Eight of those 40 show the feature. Its observed frequency among assessable specimens is 8/40 = 20%. Reporting 8/100 = 8% as if every surface were equally observable would conceal the assessment problem. Both totals can be reported, provided their meanings are separated.
Fragmentation introduces another complication. A single modified element can become several marked specimens, or a mark can remain on just one fragment. An experimental and archaeological comparison of NISP and element-based cut-mark frequencies addresses this measurement issue. Use the unit that fits the question and retain enough records for alternative summaries.
Burning likewise requires context. Heated bone might relate to cooking, fuel use, accidental fire or waste disposal; colour alone should not be treated as a complete temperature history. In the teaching sample, the defensible initial result is a frequency of a defined observable feature among assessable surfaces. The explanation of human action comes after the microscopic, spatial and depositional comparisons.
CHAPTER 9 OF 20 · Reconstruct animal lives
9. Build an age profile from bounded observations
Tooth eruption, tooth wear and skeletal fusion provide different kinds of age information. They are not interchangeable clocks. Eruption and fusion refer to developmental events; wear reflects use over time and can vary with diet and other conditions. Once an element is fully fused, that observation alone cannot distinguish all later adult ages.
Reference choice matters. A study of known-age Svalbard reindeer mandibles developed and tested tooth eruption and wear recording schemes, including a relative scheme requiring calibration for absolute ages. Its value here is methodological: a scheme needs a defined population, observable criteria and a test of performance. It should not simply be transferred to a different species because the teeth look broadly similar.
For a fictional set of 24 independent ageable mandibles, suppose six are assigned to an early class, twelve to a middle class and six to a late class. The observed distribution is 25%, 50%, 25%. If four additional mandibles remain unclassified, they must remain visible in the report. They are not automatically distributed in the same proportions as the classified sample.
The profile describes the ageable remains recovered under these rules. Moving from that sample to a herd-management strategy requires preservation, sampling and context arguments. Use the dedicated age-at-death guide for further practice distinguishing a developmental observation, an age interval and a historical interpretation.
CHAPTER 10 OF 20 · Reconstruct animal lives
10. Test husbandry explanations without reducing animals to one use
An age profile can help investigate management, but several economic arrangements may generate overlapping patterns. Animals can provide different products at different ages, and households can acquire animals raised elsewhere. A settlement’s slaughter sample is not necessarily the mortality record of a herd kept at that settlement.
Return to the 24 mandibles. An analyst proposes that the twelve middle-class animals indicate a specialised production strategy. Before accepting that claim, ask whether the middle class spans a much longer interval than the early class, whether fragile younger remains were lost, and whether the deposit combines centuries. Equal bar widths on a chart can disguise unequal time intervals and unequal chances of preservation.
Now add a small group of older animals with joint changes. Their presence may be compatible with prolonged working lives, but compatibility does not establish a specific task. Sex, body size, disease, terrain and the location of changes may matter. Nor does a bone abnormality directly measure how people felt about an animal.
A real example of combining methods is the Houten-Castellum cattle study, which uses archaeological and isotope evidence alongside age information. The general lesson is to compare lines of evidence that answer different parts of the management question. State separately what the bones reveal about age, what other measurements suggest about life history, and what the archaeological setting supports about human decisions.
CHAPTER 11 OF 20 · Reconstruct animal lives
11. Separate domestication, management and identification
Identifying a bone to a taxon does not by itself reveal the animal’s relationship with people. Domestication concerns changes across populations and generations; management concerns human practices affecting animals. Hunting, feeding, moving, breeding and keeping animals can overlap historically without becoming one simple category.
Consider an invented series of pig bones from three phases. Average bone dimensions decline, but the samples also differ in age, sex composition and which elements were measured. A claim of domestication based only on the overall mean would be premature. The first repair is to compare like elements and comparable biological groups, then examine context and independent evidence.
A second temptation is to treat a single “domestic-looking” specimen as proof that an entire community practised intensive husbandry. One animal could have been exchanged or moved over a long distance. A population-level explanation needs an appropriate sample and chronology. Conversely, a lack of a dramatic size change does not demonstrate that people never influenced animal movement or feeding.
The Hardinxveld-Giessendam study of suids illustrates a research design combining zooarchaeology and stable isotopes to investigate animal exploitation. Read such studies for how their evidence is assembled, not as a universal decision key. In the invented sequence, the useful conclusion may initially be narrower: measured dimensions differ between specified samples, and several biological and historical explanations remain open.
CHAPTER 12 OF 20 · Reconstruct animal lives
12. Give fish, shellfish and small animals their own questions
Large mammal bones should not silently define the whole subject. Fish, shellfish, birds and small vertebrates have different anatomical structures, identification resources and preservation problems. A recovery procedure adequate for large joints may be poorly suited to tiny diagnostic remains. A study must state which animal groups it can actually examine.
In a fictional cave sample, small mammal jaws occur with digestive surface alteration and a restricted skeletal pattern. Predator accumulation is a hypothesis worth testing; human consumption is not the default merely because artefacts occur nearby. The site may preserve overlapping histories from people and non-human animals. The specialist microvertebrate guide develops this distinction.
Fish otoliths offer a different route into life history. Their growth structures can be investigated using methods appropriate to the species and preparation, but a visible band is not automatically an annual boundary. Age, season of capture and season of site occupation are separate claims. Even a secure seasonal catch can have been stored or transported before deposition. Continue with archaeological otolith analysis for these distinctions.
For a mixed assemblage, resist putting unlike counts into one pie chart labelled “diet.” A shell valve, a fish vertebra and a mammal bone are different units with different relationships to animals and edible quantities. Compare within clearly defined analytical categories before attempting a broader reconstruction.
CHAPTER 13 OF 20 · Combine laboratory and field evidence
13. Use ZooMS to improve identification, not erase sampling
Zooarchaeology by Mass Spectrometry, or ZooMS, uses peptide-mass fingerprints from preserved proteins, commonly collagen, to investigate animal origin. It can make some morphologically unidentifiable material informative. The University of York’s palaeoproteomics programme describes the approach and its applications. Resolution depends on preserved information and the reference markers available; not every result distinguishes a species.
Imagine 200 unidentified fragments, of which 20 large, well-preserved pieces are selected for analysis. Fifteen produce usable results and ten of those are assigned to taxon X. Ten out of fifteen describes successful identifications in that selected subset. It does not establish that two-thirds of all 200 fragments belong to X. Selection and analytical success may both vary among taxa and preservation conditions.
A better design states the target population, selection method, failed results and limits of identification. If the goal is taxonomic composition, a convenient set of visually promising fragments may answer a different question from a stratified or otherwise justified sample. New molecular labels should remain connected to their physical specimen IDs and existing morphological records.
A 2023 double-blind comparison of morphology and ZooMS provides an empirical model for comparing methods rather than assuming one is always correct. The site’s ZooMS specialist guide explains group-level identifications and their limits. If two methods disagree, investigate references, contamination, preservation and recording before replacing one label automatically.
CHAPTER 14 OF 20 · Combine laboratory and field evidence
14. Interpret isotopes through tissues and baselines
An isotope measurement is a chemical observation made on a specified material. Its archaeological interpretation depends on the tissue, its formation history, preservation and relevant environmental or dietary comparisons. Animal movement, feeding and habitat can be investigated, but a measured value does not name a birthplace or provide a menu.
Use a deliberately simplified mixing exercise to see why assumptions matter. Suppose two hypothetical dietary endmembers have carbon-isotope values of −22‰ and −12‰. Assume equal relevant carbon contribution per unit of the quantity being mixed, a known tissue adjustment already applied, two sources only and approximate linear mixing. An adjusted value of −18‰ gives −18 = f(−12) + (1 − f)(−22), so f = 0.4.
The answer is a 40% contribution under this model, not a general claim that 40% of an animal’s food mass came from one source. If the second endmember were −14‰ instead, f would be 0.5. If there were a third source, one measurement would generally not determine all proportions. If the relevant tissue weighted dietary components differently, the simple interpretation would change again.
The stable-isotope specialist article develops these evidence limits. In practice, begin with the biological question and sampling window, then decide which measurements and local comparisons could discriminate among explanations. Do not begin with a colourful chemical plot and invent a story for every cluster.
CHAPTER 15 OF 20 · Combine laboratory and field evidence
15. Combine ancient DNA and chronology with the physical record
Molecular evidence can address questions that morphology leaves unresolved, but it introduces its own requirements. An ancient-DNA result needs authentication, appropriate controls and an account of analytical uncertainty. A genetic relationship is also not identical to a management relationship: genetically related animals need not have been raised in the same household or handled in the same way.
The ancient-DNA teaching guide connects contamination, molecular damage and reference interpretation. At subject-root level, the key practical principle is to maintain one traceable chain from the excavated specimen through every sample, laboratory result and revised identification. A molecular result without a secure specimen-context connection cannot repair that missing archaeological relationship.
Chronology creates another boundary. The age of an animal’s death, the time its remains were deposited and the date of the surrounding activity can differ. A bone may be residual in a later deposit or intrusive into an earlier one. A direct date can constrain the sampled material, but the proposed historical event still requires an association argument.
Imagine a dated bone from levelling fill beneath a building. Even a reliable result does not automatically date the building’s construction to the animal’s death. The fill might contain older refuse. Geoarchaeology helps examine the movement and formation of that fill. Combining methods means checking how their objects and timescales relate, not simply counting the number of techniques that appear in a report.
CHAPTER 16 OF 20 · Combine laboratory and field evidence
16. Work a complete two-context comparison
The invented Harbour Lane investigation asks whether animal use changed between two phases. Both samples were processed with the same stated recovery protocol, but their preservation differs. Cattle and sheep/goat are the only categories included in the following NISP percentage calculation; other identified and unidentified material remains separately recorded.
| Observation | Earlier context | Later context |
|---|---|---|
| Processed sediment volume | 40 litres | 80 litres |
| Cattle NISP | 30 | 48 |
| Sheep/goat NISP | 20 | 72 |
| Total NISP in these two categories | 50 | 120 |
| Specimens assessable for the selected surface feature | 25 | 60 |
| Specimens bearing that feature | 5 | 6 |
Cattle’s share falls from 30/50 = 60% to 48/120 = 40%. Yet cattle specimen density also falls from 30/40 = 0.75 to 48/80 = 0.60 per litre, while sheep/goat density rises from 0.50 to 0.90. These summaries describe a change in the recovered material. They do not by themselves establish changes in meat consumption.
The selected surface feature occurs on 5/25 = 20% and 6/60 = 10% of assessable specimens. That apparent decline needs a comparability check: were the same elements, surface areas and diagnostic criteria assessable in both contexts? A proportion is not rescued merely by choosing a smaller denominator.
The next investigation should examine fragmentation, taxonomic identification rates, depositional duration and anatomical composition. If the later context is a secondary rubbish fill accumulated over much longer, direct behavioural comparison weakens. A responsible interim conclusion reports the numerical changes, specifies the sampled material and identifies what prevents a stronger account of human behaviour.
CHAPTER 17 OF 20 · Repair, transfer and continue
17. Diagnose an attractive but weak interpretation
When an interpretation fails, identify the broken connection before collecting more data. More specimens cannot solve every problem. A thousand fragments with unknown context may remain less informative for a phase comparison than a smaller, securely associated set.
- If a report equates specimen percentages with animal percentages, repair the counting-unit explanation and calculate complementary measures where justified.
- If a species disappears after a recovery-method change, compare recovery and identification before proposing a dietary taboo.
- If an MNI total changes dramatically between reports, inspect refitting, side, age and aggregation rules before assuming a biological disagreement.
- If an age profile is described as a specific husbandry system, test preservation, unequal age intervals and alternative supply arrangements.
- If an isotope cluster is called an imported herd, inspect local baselines, tissue timescales and other explanations for variation.
In Harbour Lane, suppose the two contexts were assessed by different analysts. A useful repair is a blind reassessment of a selected mixed subset using agreed criteria. Do not show the analyst which historical narrative each specimen is supposed to support. Record both agreements and disagreements, then decide whether the original difference remains meaningful.
Another repair is narrowing the claim. “The later recovered assemblage contains a higher sheep/goat NISP share” may survive when “the population switched to sheep meat” does not. Narrowing is a positive result when it accurately locates the evidence. It tells the next researcher what is established and where a new observation could make progress.
CHAPTER 18 OF 20 · Repair, transfer and continue
18. Transfer task: a fish assemblage in harbour reclamation
An invented harbour excavation recovers fish remains from two adjacent reclamation fills. Fill R contains 6 left and 2 right diagnostic otoliths of one taxon; fill S contains 2 left and 6 right. Assume each fish has one left and one right of the selected otolith type, all identifications and sides are secure, none refit, and no size or age incompatibility raises the minimum. The fills may have come from one earlier midden.
R also has 18 identified fish specimens from 30 litres of processed sediment. S has 60 from 60 litres. R was processed with a coarser procedure than S. Twelve visually promising unidentified fragments from S were sent for ZooMS; eight gave usable results, six assigned to the same taxon as the otoliths.
Before reading the answer, state the separate and pooled minima, calculate observed specimen densities, and write one claim that the molecular subset supports. Then explain why this evidence does not establish either a rise in fishing intensity or eight separate fish meals. Name the next observation that would most improve interpretation of the fills.
Answer and reasoning
Each fill has a separate MNI of six. Their minima sum to twelve only under an appropriate independence assumption. Pooled, there are eight left and eight right otoliths, so the simplified minimum is eight. If the fills share a source midden, separate minima may double-count individuals whose remains were divided during movement.
Observed densities are 18/30 = 0.6 and 60/60 = 1.0 identified specimens per litre. Unequal recovery prevents a clean inference about original abundance, even before the reclamation history is considered. Six of eight successful ZooMS results in the selected subset support the stated taxon assignment; they do not establish a 75% share in all of S.
Eight minimum individuals are not eight meals. A meal can include parts of several fish, one fish can supply several eating events, and remains can originate outside consumption. The strongest next observation depends on the question, but evidence about the fills’ source and relationship—such as secure stratigraphic information, refits or other provenance indicators—would directly test whether they represent independent activity. The changed setting makes depositional history central to the answer.
Is zooarchaeology the same as archaeozoology?
Both terms are used for the study of animal remains in archaeology. Research traditions and emphases can vary, so read the question, material and methods rather than inferring a rigid difference from the label. This guide uses zooarchaeology for the broad field.
Can a bone tell us exactly what people ate?
A bone may contribute to evidence of food acquisition or processing, especially when identification, modification and context agree. It cannot normally reconstruct an exact menu on its own. Some foods leave little bone evidence, some bones come from non-food activities, and preparation or disposal may occur elsewhere.
Does the most common species provide the most meat?
Not necessarily. Specimen counts depend on anatomy, fragmentation, identification and survival. Animals also differ in body size and edible yield. Estimating dietary contribution requires explicit assumptions and uncertainty; a NISP percentage should not be silently relabelled as a calorie percentage.
Is a molecular identification always more reliable?
Different methods have different strengths and failure modes. A molecular result can resolve some fragments while leaving related taxa grouped or producing no usable signal. Morphological expertise can provide anatomical information that a taxonomic fingerprint alone does not supply. Investigate disagreements rather than treating either method as infallible.
Can I learn the reasoning without excavating or sampling bones?
Yes. Begin with documented teaching datasets, authorised reference resources and published reports. Recalculate totals, inspect definitions and compare alternative interpretations. Handling archaeological collections or taking samples requires the relevant permissions, conservation planning and specialist supervision. Most of the reasoning exercises in this guide can be completed with a spreadsheet and the stated assumptions.
CHAPTER 20 OF 20 · Repair, transfer and continue
20. Build a practical learning and research route
For a first learning session, complete the NISP and MNI calculations, then explain aloud why their answers differ. For a second, use Harbour Lane to produce three short paragraphs: observations, interpretations and unresolved alternatives. For a third, attempt the reclamation task without copying the earlier grouping rule. Success is recognising which assumption changed, not merely obtaining the arithmetic.
The existing Teaching Zooarchaeology guide provides a complementary instructional route. Continue through the linked specialist articles when your question concerns age, molecular identification, isotopes or small animals. Use How Archaeology Works for the larger evidence process, and the How X Works hub to connect this subject with other scientific mechanisms.
For a real project, a useful result package contains the research question, context and sample definitions, recovery information, identification rules, specimen records, calculation methods, limitations and a traceable archive. A reader should be able to reconstruct a reported percentage without guessing which records were excluded. Destructive analysis should have a stated purpose and preserve the connection between the original specimen, the sampled portion and the results.
Sources and further reading
The links below are primary research or institutional method resources checked for this guide on 23 October 2026. Their particular populations, materials and geographical scopes matter. They support methods and comparisons; the invented calculations in this guide are original exercises, not extracted research findings.
- Baker and Worley, Animal Bones and Archaeology: Recovery to Archive, Historic England, 2019 — professional project guidance.
- Historic England zooarchaeology programme, updated 27 March 2026 — research and comparative resources.
- Lyman, observations on NISP and MNI — quantitative-unit choices.
- Hammerstone fragmentation and specimen identifiability study — experimental evidence relevant to body-part interpretation.
- NISP and MNE in cut-mark analysis — effects of measurement units.
- Van den Berg and colleagues, reindeer tooth-wear and eruption study, 2021 — tested age-estimation procedures.
- Houten-Castellum cattle-management study, 2021 — combining zooarchaeological and isotope evidence.
- Hardinxveld-Giessendam suid study, 2022 — investigating exploitation with multiple evidence types.
- University of York palaeoproteomics and ZooMS — protein-based identification.
- Morin and colleagues, double-blind morphology and ZooMS comparison, 2023 — empirical comparison of identification methods.
Before finishing any explanation, return to the specimen and ask whether another reader can follow the same path from observation to claim. If that path depends on an uncertain identification, a changing denominator or an unresolved depositional relationship, make the dependency visible. That is how animal remains become useful historical evidence without being asked to carry a story they cannot establish.