Ethnobotany studies relationships between people and plants: how people recognise, name, grow, gather, transform, exchange and understand them. Its strongest explanations connect an identified plant with a particular practice, the people who know that practice, and the conditions under which it works. A list of useful species is a starting point; a careful account explains how a relationship is maintained, changes or breaks down.
The gold standard is to make those connections traceable without stripping knowledge from its context or claiming more than the evidence shows. This guide develops that standard through plant identification, material processing, food systems, interviews, collections and original worked cases. All numerical teaching cases below are hypothetical. They demonstrate reasoning, not findings about an actual community. The broader Gold Standard subject hub introduces the series; its Education parent explains why understanding should survive a change of context.
THE GOLD STANDARD · ETHNOBOTANY
Connect plants, practices and people
Choose a route through plant identity, material use, living knowledge and the evidence needed to connect them. All teaching chapters remain open below.
ROUTE 1 · CHAPTERS 1–4
Connect plants, names and evidence
Define the relationship and establish what the plant record supports.
ROUTE 2 · CHAPTERS 5–8
Follow people and practices
Explore participation, observation, material processing and food systems.
ROUTE 3 · CHAPTERS 9–12
Trace diversity and knowledge
Examine cultivation, markets, transmission and the units being counted.
ROUTE 4 · CHAPTERS 13–16
Test conclusions and records
Work through sampling, harvest budgets, health claims and collections.
ROUTE 5 · CHAPTERS 17–20
Repair, transfer and continue
Diagnose weak explanations, solve a new case and build a learning route.
Full chapter index · Try the transfer task · Gold Standard subject hub
Chapter index
Connect plants, names and evidence · Chapters 1–4
Follow people and practices · Chapters 5–8
Trace diversity and knowledge · Chapters 9–12
Test conclusions and records · Chapters 13–16
Repair, transfer and continue · Chapters 17–20
CHAPTER 1 OF 20 · Connect plants, names and evidence
1. Start with a relationship, not a list
Imagine a basket displayed beside a living plant. The label names the species and says that its stems are used in weaving. Useful questions immediately appear. Which stems are selected? Does the maker use them fresh or dried? How are they split? Who teaches the technique? Is the material gathered, cultivated or purchased? Can another plant replace it, and what changes when that happens?
These questions connect three different objects of investigation: the organism, the practice and the social setting. Identifying the organism cannot explain all the steps in making the basket. Interviewing the maker cannot, by itself, resolve a difficult botanical identification. Testing the finished basket cannot reveal who is entitled to share a design. Each inquiry supplies a different part of the explanation.
A useful working record is therefore a relationship with several fields: plant identity; plant part; action; purpose; participant or knowledge source; place and time; and evidence. Add uncertainty to the field where it belongs. You might be certain that a strip came from a leaf, uncertain of the species, and confident that the maker described a particular drying method. A single confidence score would hide that structure.
The subject connects naturally with Botany, which examines plants as organisms, and Anthropology, which examines human life in context. Ethnobotany asks what becomes visible when these perspectives investigate the same relationship. Its subject matter includes urban kitchens and commercial supply chains as well as farms, forests and historical collections.
The first repair for a weak explanation is usually to replace “this plant is useful” with a more precise sentence: “these people use this part, prepared in this way, for this purpose, under these conditions.” That sentence exposes the questions a bare list leaves unanswered.
CHAPTER 2 OF 20 · Connect plants, names and evidence
2. Treat knowledge as situated and changing
“The community knows this plant” can conceal substantial differences. A grower may recognise planting material, a trader may distinguish grades, and a craftsperson may judge how a strip bends. A child may know a plant through a story while an adult knows its seasonal availability. None of these forms of knowledge is automatically a complete substitute for another.
An original comparison helps. Suppose two people are shown a bundle of dried leaves. One identifies the market name immediately but cannot identify the living plant. The other identifies the living species but has never seen the processed bundle. Calling one knowledgeable and the other ignorant would miss what each task measured. The tasks involve different cues and different histories of practice.
Separate knowing a name, recognising an organism, explaining a procedure and performing it. A person may accurately describe a sequence but struggle with its timing. Another may perform a skilled action without expressing every adjustment verbally. An interview and a demonstration can reveal different aspects of the same practice, provided participants choose to share them.
Knowledge also changes. A household may substitute a purchased ingredient when access to a gathering place disappears. A maker may adopt a new tool while retaining a familiar design. Such changes deserve investigation on their own terms. They do not automatically demonstrate either cultural loss or progress. Ask what was retained, what became easier, what ceased to be possible, and whose assessment supports the conclusion.
UNESCO’s explanation of intangible cultural heritage places transmitted knowledge and skills within living community practice. For an ethnobotanical study, the practical consequence is to record the date and circumstances of an account. An account from one period should not silently become a timeless statement about everyone associated with a place.
CHAPTER 3 OF 20 · Connect plants, names and evidence
3. Keep plant names, parts and categories distinct
A vernacular name is evidence about language and classification. It is not automatically a one-to-one scientific identifier. One local term may cover several botanical species; several terms may refer to one species in different places, stages or uses. Preserve the original term, its language and the speaker’s explanation before deciding how it maps onto a scientific name.
Consider an invented material called “river cane”. At two markets the same term might refer to different plants selected for similar weaving qualities. Combining every sale under one species would turn a commercial category into a false biological identity. Conversely, replacing all local terms with a single scientific name could erase distinctions between desirable and unsuitable material within that species.
Plant parts matter just as much. Root, stem, leaf, bark, fruit and seed describe different structures. A processed strip may hide the features needed for identification. A botanical name attached to a living display does not establish the identity of a nearby manufactured object unless the connection is documented. Follow the chain from organism to harvested part to processed material.
Use separate columns for the name as recorded, the proposed scientific identification and the reason for that identification. If an expert later revises the name, retain the earlier record with its date. The revision then becomes an intelligible improvement rather than a mysterious contradiction.
The Kent Ethnobotanical Herbarium explicitly connects voucher specimens with vernacular and scientific names while recognising that their correspondence can be difficult. Its approach illustrates why an ethnobotanical record needs both botanical evidence and the participant’s categories. Neither should be quietly substituted for the other.
CHAPTER 4 OF 20 · Connect plants, names and evidence
4. Build a traceable record of plant identity
A voucher is a preserved reference specimen associated with a record. Its value is that another investigator can revisit the identification rather than rely entirely on a name typed into a spreadsheet. Kent’s herbarium combines specimens with information about morphology, habitat, names and cultural significance. A photograph or processed product can also be evidence, but its limitations must be stated.
For a teaching exercise, imagine record P017. It links a permitted photograph of a living plant, a field description, an identifier’s assessment and an interview entry. The interview entry refers to P017, not merely “the green plant”. If identification remains at genus level, retain that resolution. Do not supply a species name because a database requires a non-empty cell.
Three distinctions prevent common errors. First, a specimen documents the plant collected at a particular time; it does not prove that every product with the same market label came from that species. Second, a correct identification does not verify every claim made about the plant. Third, a specimen cannot recover a relationship that was never recorded: who provided the name, what the term meant, or how the material was used.
Collection itself requires appropriate permission and a justified purpose. A beginner can practise with institutionally published specimen records rather than collecting plants. In an actual project, agree on collection, deposition, access and associated information before taking material. Some locations or knowledge may require restricted handling rather than public display.
A strong record can remain incomplete. “Leaf photographed; reproductive features unavailable; identification unresolved” is useful evidence. An unsupported species-level label is a hidden error that can spread through every later comparison. Precision means expressing the resolution the evidence supports.
CHAPTER 5 OF 20 · Follow people and practices
5. Agree on participation before gathering information
Ethnobotanical research concerns people as well as plants. The International Society of Ethnobiology Code of Ethics sets out principles including informed consent, active participation, confidentiality, reciprocity and acknowledgement. It treats collaboration as extending across research, rather than beginning and ending with an interview signature.
Consider an original project scenario. A community garden group agrees to help produce a bilingual guide to publicly shared food plants. That agreement does not automatically authorise recording every conversation, publishing individual names, uploading precise locations, or adapting the information for commercial advertising. Each is a different use with different consequences.
A workable project description states what participants are being invited to do, what will be recorded, who will see the results and what will happen if they decline a question. It should also make room for participants to correct a mistaken plant name or contextual interpretation. Explain the actual project in language people can understand; technical vocabulary does not make an agreement more meaningful.
This chapter’s garden example is a design exercise, not a universal consent form. The appropriate process depends on the people, institutions and circumstances involved. Where knowledge has collective responsibilities, a conversation with one willing person may not settle every question about publication. Where an individual prefers anonymity, a colourful quotation is not a reason to override that preference.
Useful collaboration also asks what the work is for. A guide returned in an accessible format may help participants more than a dataset they cannot use. Budgeting for review, translation or a locally useful output changes the project from the beginning. If the agreed scope cannot be respected, revise the project rather than treating disclosure as an obstacle to overcome.
CHAPTER 6 OF 20 · Follow people and practices
6. Ask questions that separate recall from observation
“What plants do you use?” can produce an interesting list, but it leaves the activity, period and meaning of use unspecified. A participant might describe current household practice, something remembered from childhood, a specialised trade or something heard from someone else. Record those distinctions instead of merging them into a single yes.
An original interview sequence might begin: “Which plant materials have you worked with during the past month?” A follow-up asks about one volunteered example: “Could you describe how you selected that material?” Another asks whether the account concerns something personally done or observed. The purpose is clarity, not an interrogation or a test the participant is expected to pass.
Prompts change what can be recalled. A photograph, a specimen, a finished object and a walk through a garden each make different cues available. If one group sees photographs and another answers without prompts, their totals are not directly comparable as measures of general knowledge. Record the prompt and its order. A recognition task and a free-recall task answer different questions.
Observation can illuminate a procedure that words leave compressed. For example, “dry until ready” may involve judgements about feel, sound, flexibility or weather. With permission, ask what signs guide the decision. Do not infer an exact drying duration from one demonstration when conditions vary.
Keep the participant’s account separate from the researcher’s interpretation. A field note can contain three short sections: what was said; what was observed; and what remains a hypothesis. This makes disagreements productive. If a later visit shows a different method, you can investigate season, purpose or personal preference without rewriting the first visit as though it never happened.
CHAPTER 7 OF 20 · Follow people and practices
7. Explain a material through its transformation
A basket is not simply a plant in another shape. Between the living organism and the finished object are decisions about selection, cutting, sorting, splitting, drying, storage and assembly. Which steps occur, and in what order, must be established for the actual practice under study. A generic process diagram cannot replace that evidence.
Consider a hypothetical workshop comparing two batches of plant strips. Batch A is wider and more variable in thickness. Batch B is narrower and more uniform. If baskets made from B are more consistent, several explanations remain: the raw material, the preparation, the maker’s skill, or the design. A comparison that changes all four cannot identify one cause.
Start with the property that matters to the maker. “Strong” may mean resisting a pull, surviving repeated bending, keeping its shape under a load, or remaining useful after wetting. These are different outcomes. Translate the question into a defined observation while preserving the maker’s own criterion. A laboratory measurement is most helpful when it addresses the actual demand placed on the object.
Processing creates further trade-offs. Removing more material might produce a finer strip but reduce the amount available for manufacture. A quicker method might save labour while increasing variability. Recording only the final object hides these choices and makes substitution look easier than it is.
Kew’s Economic Botany Collection links plant materials and objects with records searchable by plant, place, part and use. Its emphasis on accompanying labels and archives offers a useful lesson: material evidence and documentary history belong together. To explain an object, ask both how it was made and what supports the account of its origin.
For a concrete specialist route, the henequen fibre vocabulary guide follows the connection from plant to extracted fibre and made object. Use its process terms to distinguish an action from a property: describing how a fibre is prepared is different from establishing how the finished material performs.
CHAPTER 8 OF 20 · Follow people and practices
8. Follow a food plant through a whole system
Food is a useful way to see why a species list is insufficient. The same plant can enter different dishes or products depending on the part selected, its maturity, preparation and storage. Availability also depends on labour, land, planting material, transport and decisions about when to harvest. Ethnobotany connects these relationships without assuming that one household represents an entire region.
Enset provides a documented example. In its January 2025 account of research in Ethiopia, Kew describes work with Addis Ababa University and partners on changing farming systems and crop diversity. The account connects cultivation with processing, tools, consumption and cultural practice. It shows why understanding a crop requires attention to the people who maintain and use it.
The existing enset cultivation vocabulary guide offers a focused route into terms for this system. Read such terms relationally: a word for planting material connects to propagation; a term for a tool connects to an action; a processing term connects raw material to a usable product. Knowing the vocabulary is helpful, but it does not itself confer practical competence.
For an original classroom exercise, map a familiar purchased food backwards. Begin with the item as eaten, then identify the processed ingredient, the plant part and the crop. Mark every stage supported by a label or reliable record and every stage still unknown. “Bought at this market” describes a point of sale, not necessarily a place of cultivation.
Do not turn a cultural description into an improvised recipe for an unfamiliar plant. An account may omit critical details because its intended audience already knows them. The purpose of the map is to understand the chain and its evidence, including where information is missing.
CHAPTER 9 OF 20 · Trace diversity and knowledge
9. Distinguish species, varieties and maintained diversity
Counting species gives one view of a food system. It does not measure all the variation people maintain within a crop. Two gardens may contain the same species while differing in the planting material, named varieties, management and traits valued by growers. A species inventory would miss these differences.
In a hypothetical garden, growers distinguish six named forms of one crop. That is evidence of six locally recognised categories. It is not yet evidence of six genetically distinct lineages. Two names might refer to the same material under different circumstances; one name might cover material that varies. Investigate the correspondence rather than declaring either local naming or genetic analysis the sole definition of useful diversity.
Ask what differences matter in practice. A grower may distinguish harvest timing, processing effort, storage qualities or performance in a particular site. These criteria can explain why a form is maintained even when it is not the highest-yielding option in a single trial. Conversely, a valued name can persist while the material associated with it changes.
Propagation adds another layer. Seed-derived plants and vegetatively propagated plants have different routes of reproduction and movement. Record how planting material is obtained, selected and exchanged instead of inferring the route from the plant’s common name. For a particular crop, botanical and cultivation sources are needed to establish the details.
A conservation question should specify the object being conserved. Is it a population, a named planting line, a technique, access to a growing place, or the ability to pass knowledge to another generation? These may support one another, but saving a packet of material does not automatically maintain every relationship that made it useful. The practical plan should match the intended outcome.
CHAPTER 10 OF 20 · Trace diversity and knowledge
10. Trace markets without inventing origins
A market is an excellent place to notice plant diversity and a difficult place to infer its history. A trader may know product grades and customer preferences in great detail while knowing little about the original harvest site. A label can be accurate for commerce without containing enough information for botanical research.
Imagine a bundle labelled “mountain fibre”. That phrase could describe an origin, a grade, a style or a brand. Begin by recording the label exactly. Ask what the term means in that transaction and what evidence supports the answer. Keep a seller’s account distinct from a verified supply record. Neither should be silently converted into a precise geographic coordinate.
Substitution is another important question. A maker may accept a different plant because it is available, affordable or suitable for a particular design. The substitute may preserve one function while changing another. To study the change, compare the relevant qualities and document the reasons people give. Price alone cannot establish why a material was chosen.
Markets also shape visibility. A frequently sold product is easy for a visitor to encounter. A rarely traded household plant may remain largely invisible. A market inventory therefore describes what was offered at selected places and times, not the full range of plant relationships in the surrounding population.
An honest chain-of-custody diagram can contain an unknown segment. For example: known shop, documented wholesaler, cultivation site unresolved. That is more useful than a seamless story built from assumptions. Later evidence can fill the gap; meanwhile, readers know exactly which claim the record can support.
CHAPTER 11 OF 20 · Trace diversity and knowledge
11. Study transmission through tasks and opportunities
If knowledge is learned through practice, opportunities to practise matter. A person who no longer has access to a growing place may remember names and procedures while being unable to demonstrate them. Someone newly entering a trade may quickly learn commercial distinctions without acquiring every skill associated with cultivation.
Consider a hypothetical comparison between older and younger participants. Older participants name more plants during an unprompted interview. It would be premature to conclude that all plant knowledge is disappearing. The groups may differ in occupation, language, access to gardens or familiarity with the interview format. Younger participants might recognise processed foods or market categories that the first question never elicited.
A better design defines several tasks and records relevant experience. Ask about naming, recognition, explanation and performance separately. Use the same prompts when comparing equivalent tasks. Where appropriate, include participants’ accounts of how they learned and when they last used the knowledge. This cannot remove every difference, but it makes the comparison interpretable.
Transmission also involves correction. A beginner may know the sequence of a craft but fail to notice when material is unsuitable. An experienced maker’s intervention reveals a decision rule: what cue was noticed, why it mattered and how the action changed. Recording that exchange can be more informative than filming an uninterrupted successful performance.
Avoid treating a recording as a complete replacement for learning with a practitioner. A video can preserve visible steps yet omit touch, judgement, feedback and access to materials. A useful educational output identifies those limits. It might say, “This resource documents the sequence; supervised practice is still needed to learn the material judgement.” That is a precise account of what has and has not been transferred.
CHAPTER 12 OF 20 · Trace diversity and knowledge
12. Count people, reports and categories separately
Quantitative summaries become misleading when their units change unnoticed. A participant can mention one plant several times, describe several uses of it, or repeat a use reported by another participant. These are different observations. Define the unit before choosing a denominator.
In an original teaching example, 40 participants answer the same question about one plant. Eighteen report personally using it in the stated period. Across those 18 people, the records contain 27 participant–use-category pairs because some describe more than one category. The coding scheme recognises four distinct categories overall.
| Quantity | Calculation | What it describes |
|---|---|---|
| Participants reporting use | 18 ÷ 40 = 45% | Share of this interviewed sample reporting use |
| Use-category pairs per participant | 27 ÷ 40 = 0.675 | Mean coded pairs across all participants |
| Use-category pairs per reporting user | 27 ÷ 18 = 1.5 | Mean coded pairs among the 18 reporting users |
| Distinct coded categories | 4 | Breadth under this particular coding scheme |
The figure 45% does not mean the plant works for 45% of people. It describes reports in a sample under a particular question. The figure 1.5 does not mean each person knows one and a half plants. It is an average number of coded categories among users of one plant.
Coding decisions also affect results. If “wrapping food” and “lining a steaming container” are combined into one category, the category count may fall without any participant changing their practice. Preserve the codebook and explain merged categories. If you use a published ethnobotanical index, state its exact definition and follow the cited method; a familiar acronym is not a substitute for a denominator.
CHAPTER 13 OF 20 · Test conclusions and records
13. Check whether the sample mix changes the conclusion
A pooled percentage can move because the people sampled changed. This is especially important when occupations or access to plants differ. The following original example shows a reversal without any arithmetic error.
Two surveys ask whether a fibre plant was used during the same defined period. Within each survey, participants are divided into craft workers and other adults. The figures are hypothetical and do not represent a real population.
| Survey and group | Reporting use | Interviewed | Reporting share |
|---|---|---|---|
| A: craft workers | 18 | 20 | 90% |
| A: other adults | 12 | 80 | 15% |
| B: craft workers | 64 | 80 | 80% |
| B: other adults | 2 | 20 | 10% |
| A: all participants | 30 | 100 | 30% |
| B: all participants | 66 | 100 | 66% |
Survey B contains many more craft workers, who have higher reporting shares in both surveys. The overall increase from 30% to 66% therefore cannot establish an increase in use within comparable groups. With an illustrative equal weighting of the two groups, A becomes (90% + 15%) ÷ 2 = 52.5%, while B becomes (80% + 10%) ÷ 2 = 45%.
Equal weighting is not automatically the correct population estimate. It simply holds the mix constant to expose what pooling concealed. A population estimate would need an appropriate sampling design or justified weights. The repair begins by stating the question: change among comparable groups, change in the population as constituted, or the experience of a particular group? These questions need not have the same answer.
CHAPTER 14 OF 20 · Test conclusions and records
14. Separate a harvest budget from ecological sustainability
A material budget answers how much is used or lost. Sustainability asks whether a harvesting system can continue while maintaining the relevant populations and ecological functions. The first can inform the second, but it cannot replace evidence about growth, survival, recruitment and other pressures.
Suppose a hypothetical cooperative receives 120 kilograms of fresh plant material. Sorting removes 15 kilograms. Processing yields 42 kilograms of usable material from the remaining 105 kilograms. The sorting retention is 105 ÷ 120 = 87.5%. Processing yield relative to sorted input is 42 ÷ 105 = 40%. Overall usable yield is 42 ÷ 120 = 35%.
If a new order requires 70 kilograms of usable material, applying the same overall yield gives 70 ÷ 0.35 = 200 kilograms of fresh input. That is a planning estimate under unchanged material quality and processing conditions. It is not permission to harvest 200 kilograms from any site, and it is not a sustainable quota.
To address the ecological question, identify what is removed and what remains. Harvesting a whole plant, removing a stem and taking some leaves are biologically different actions; consequences depend on the species and method. Record the number and condition of plants, harvested parts, interval between harvests and relevant outcomes over time. A short observation period may miss delayed effects.
The social question remains separate too. An ecologically plausible supply does not establish who has access to it, how labour is paid or whether benefits are distributed as intended. A credible account can place a mass balance, an ecological assessment and an account of access alongside one another without pretending that one number settles all three.
CHAPTER 15 OF 20 · Test conclusions and records
15. Keep reported remedies separate from tested health claims
Plants used in health practices are legitimate subjects of ethnobotanical research. Recording a reported use, however, is different from demonstrating that a preparation is effective and safe for a particular medical condition. Respectful description and careful assessment can coexist.
An original evidence exercise makes the distinction clear. Imagine 25 interviewees independently describe a preparation. This supports a claim about reported knowledge or practice within the study. Suppose a laboratory extract subsequently changes an outcome in a cell experiment. That adds evidence about that extract under those experimental conditions. It does not establish that the preparation people described produces the same exposure or improves health outcomes in humans.
The identity of the plant, the part, preparation, composition, amount and route all matter when interpreting such research. A finished product may differ from the material described in an interview. An observed association between use and recovery also leaves alternative explanations, including the course of the condition and other care received. These gaps should be named rather than bridged by enthusiasm.
NCCIH’s Herbs at a Glance separates what research says about particular botanicals from cautions and side effects. Its explanation of why natural does not necessarily mean safer or better is a useful companion when reading claims about plant products.
For public writing, choose verbs that match the evidence: participants reported; researchers identified; an experiment observed; a clinical study assessed. Avoid silently upgrading “used for” into “treats”. This guide documents ways of reasoning about plant knowledge; it does not provide remedy instructions, doses or an identification guide for consuming unfamiliar plants.
CHAPTER 16 OF 20 · Test conclusions and records
16. Read gardens, museums and databases as evidence
A public collection can teach relationships that a living plant alone cannot show. An object, a specimen and a label can connect morphology with material use and historical context. Yet each has a provenance: somebody selected, described and displayed it. Ask what the record establishes and which voices or stages of production remain absent.
Singapore’s Centre for Ethnobotany uses plant materials, artefacts and displays to explore links between plants and people. The existing Centre for Ethnobotany and Seed Bank visitor guide provides a practical route for planning a visit; check the official page for current arrangements.
For a focused learning visit, choose one displayed relationship. Record the botanical name supplied, the material or object shown, the stated use and the place or period described. Then write one question the display answers and one it leaves open. Do not collect or taste plants to complete the exercise. The task is to read evidence carefully.
A database requires similar attention. A blank field may mean unknown, not recorded, not applicable or deliberately withheld. Those meanings are not interchangeable. If a source has no locality detail, do not infer a location from the collector’s address. If a record contains a historical identification, look for a subsequent determination before treating the name as current.
Objects can outlive the relationships that produced them. That makes collections valuable, but it also makes their limits important. A preserved basket may allow close inspection of construction while its makers’ explanations remain missing. State both facts. A modest, traceable interpretation is more useful than an elaborate narrative unsupported by the surviving record.
CHAPTER 17 OF 20 · Repair, transfer and continue
17. Diagnose weak explanations and repair the specific fault
Ethnobotanical errors often arise from crossing a boundary without noticing: name to species, report to observation, observation to cause, or sample to population. Repair is most effective when it identifies that crossing precisely.
If two sources disagree about a plant, first compare the material, names and identification evidence. They may concern different species under one vernacular term. If they concern the same species, compare plant part, maturity, preparation, place and purpose. A disagreement about use can be real without either source being wholly unreliable.
If an interview count seems surprisingly low, examine the question and opportunity to respond. Did participants understand the same time period? Were they asked about personal use or general knowledge? Did the prompt show a living plant while the activity usually involves a processed product? A better prompt may resolve a mismatch without changing the participant’s underlying knowledge.
If a claim of decline rests on a single visit, specify what actually declined. Fewer sellers at one market could reflect season, a changed location, supply disruption or reduced demand. Repeated observations and relevant accounts are needed to distinguish these explanations. “Not encountered during this visit” is not equivalent to “no longer used”.
If a proposed substitute performs poorly, revisit the function. Did it fail during splitting, drying, assembly or use? Replacing a species while keeping every processing step fixed may overlook why the original method worked. The next investigation should target the stage where evidence indicates failure.
Finally, check whose conclusion is being reported. “Participants described this change as beneficial” and “the researcher measured a shorter production time” are different statements. Both can be true while leaving costs elsewhere in the system. Good repair preserves these distinctions instead of flattening them into a single verdict.
CHAPTER 18 OF 20 · Repair, transfer and continue
18. Transfer task: a plant guide meets a new market
The changed situation
You are reviewing a hypothetical public guide to a weaving plant. The original village study interviewed 30 people. Twenty-one reported using material called “silver reed”, and a voucher linked the village material to Species A. The guide says, “Silver reed is used by 70% of residents and is Species A.”
A new market study interviews 20 traders. Sixteen sell bundles called “silver reed”. Of ten bundles examined by a qualified botanist, six are Species A and four are Species B. The remaining bundles were not examined. Traders say the name refers to a material grade. The project has permission to publish aggregate findings but no permission to identify individual traders.
Decide which claims can be retained, which need repair, and what should be checked before combining the studies. Calculate the relevant observed proportions. Then write a public conclusion of no more than three sentences that does not reveal individual identities or imply a population estimate the studies cannot support.
Answer and reasoning
The village proportion is 21 ÷ 30 = 70% of interviewed people reporting use. Without a sampling basis for generalisation, “70% of residents” overstates it. The voucher supports the identity of the linked village material, not every material called “silver reed” elsewhere.
The market proportion is 16 ÷ 20 = 80% of interviewed traders reporting sales. Selling is different from personally using, and traders are a different sampled group. A rise from 70% to 80% is therefore not evidence that use increased. Among the ten examined bundles, 6 ÷ 10 = 60% were Species A and 40% Species B. Those percentages describe the examined subset; they should not be assigned to unexamined bundles without a justified sampling argument.
An adequate conclusion is: “In the village study, 21 of 30 interviewees reported using material called silver reed, and the linked voucher identified village material as Species A. In the market study, 16 of 20 interviewed traders reported selling that grade; the ten examined bundles included six of Species A and four of Species B. These studies describe different activities and samples, so their percentages do not establish a change in use or a single botanical identity for the trade name.”
Before combining records, define the unit, activity and sampling method; retain the vernacular grade separately from botanical identity; and mark unexamined material as unresolved. The permission to report aggregate findings does not expand because individual details would make a livelier story. The transfer succeeds when the method adapts to the changed meaning of the name and the changed unit of observation.
Is ethnobotany only about medicinal plants?
No. Food, fibres, dyes, construction materials, cultivation, exchange, ritual, naming and other relationships all provide relevant questions. Medicinal use is one area where the distinction between documenting a practice and testing an outcome is especially important. A useful introductory project can focus entirely on non-medical materials or public collections.
Is it only concerned with Indigenous or rural communities?
No. People–plant relationships also occur in cities, households, gardens, shops and industries. The research question determines the setting. Work involving Indigenous knowledge has particular histories, responsibilities and community protocols that should not be erased by treating every setting as interchangeable.
Can a photograph replace a voucher?
A photograph can document valuable visible features, and in some circumstances it may be the appropriate available evidence. Whether it supports identification depends on what is visible and what distinctions must be resolved. State the evidence used and its limits; do not describe an image as a deposited physical specimen.
Does widespread agreement show that a plant is effective?
Agreement can show a shared report, classification or practice within the conditions studied. It does not by itself test a causal or medical claim. The next investigation depends on the claim: a material property, a crop outcome and a health effect require different evidence.
Should all documented knowledge be made public?
No. The scope of sharing is part of the relationship and the agreement under which information was obtained. A public collection, a restricted record and a decision not to record something are different possible outcomes. More disclosure is not automatically better research.
How can beginners contribute without overreaching?
Begin with a bounded question and existing public evidence. Compare how a museum labels a material with how a botanical collection records its identity. Practise separating observation, report and inference. A small, accurate account with explicit gaps is a better foundation than an encyclopaedic list whose identities and sources cannot be traced.
CHAPTER 20 OF 20 · Repair, transfer and continue
20. Build a learning route that ends in a defensible explanation
Choose a question small enough to answer. For a school or personal project, one displayed plant material is often sufficient. Produce a one-page account connecting the plant, the part, the transformation, the use and the evidence. Add a short section explaining what remains unknown. The quality test is whether another reader can follow each connection without being asked to trust an unsupported leap.
For a botanical route, return to The Gold Standard of Botany and practise describing structures before attempting fine identification. For a cultural route, use The Gold Standard of Anthropology to examine context, meaning and the limits of generalising from a small encounter. The enset vocabulary article offers a concrete specialist case in which cultivation and processing terms belong to a larger food system.
For a quantitative route, rebuild Chapters 12–14 with a clearly labelled invented dataset. Change one denominator, sample mix or processing yield and explain why the conclusion changes. Then attempt Chapter 18 without looking at the answer. This tests the Education principle of transfer: understanding is demonstrated when a familiar method is adjusted to a genuinely different situation.
Source trail and further study
The linked institutional sources serve different purposes. The ISE Code supports the discussion of research relationships; Kent’s herbarium illustrates botanical and cultural documentation; Kew’s collection explains routes into material evidence; Kew’s enset account supplies a specific research example; NParks provides the local public-learning setting; UNESCO explains living heritage; and NCCIH distinguishes evidence about botanical health products from assumptions based on natural origin. Sources were checked on 23 October 2026. The numerical examples and transfer case are original teaching scenarios, not reported field results.
Continue through How X Works to investigate the mechanisms behind a chosen material or food system, or return to the Gold Standard subject hub. Carry forward one discipline: keep the organism, the practice, the people and the evidence connected. That is what allows a plant relationship to be understood, questioned and responsibly explained.